TT 7050022

1. F. Golubev

VISCOSITY OF GASES
AN~D GAS MIXTURES
A Handbook
(Vyazkosl:' gazov i gazovykh smesei Cspravochnoe rukovodstvo»

Gosudarstvennoe Izdate1' stvo

Fiziko-Matematicheskoi Literatury
Moskva 1959

Translated from Russian

Israel Program for Scientific Translations

Jerusalem 1970

REPRODUCED BY
NATIONAL TECHNICAL
INFORMATION SERVICE ---------------
us DEPARTMENT OF COMMERCE
.. SPRINGFIELD. VA. Z2161
 TT 70- 50022
Published Pursuant to dn Agreement with
THE U. S. DEPARTMENT OF THE INTERIOR
and
THE NATIONAL SCIENCE FOUNDATION, WASHINGTON, D. C.

Copyright © 1970
Israel Program for Scientific Translations Ltd.
IPST Cat. No. 56 80

Translated by R. Kondor
Edited by D. Slutzkin

Printed in Jerusalem by Keter Press

Binding: Wiener Bindery Ltd., Jerusalem

A vailable from the

U. S. DEPARTMENT OF COMMERCE

IX/2/3

\
 Table of Contents

Preface . . . . . . . . . . . . . . . . . . . 0 • • • • • • • • • • • • • • • • • • • • • • • • • • •• vii

Chapter I. VISCOSITY OF GASES. • • • • • • • • • • • • • • • . • • . • • • • . • . • • 1

1. General concepts and dimensions. • • • • • • • • • • • • • • • • • • • • • • • 1
2. The kinetic theory of gas viscosity. • • • • • • • • • • • • • • • • • • • . • • 3
3. Detennination of the mean velocity of molecules. • • • • • • • • • • • • 6
4. Determination of the mean free path of molecules •••••••••••• 9
5. Analogy between viscosity and other molecular transport processes
(thermal conductivity and diffusion). • • • • • • • • • • • • • • • • • • • • • 13

Chapter II. MEt\SUREMENTS OF GAS VISCOSITY. • • • . • • • • • • • • • • • •• 15

1. Capillary Inethod. • • • • • • • • • • • • • • . • • • • • • • • • • • • • • . • •• 15
2. The Golubev and Petrov viscosimeters •••••••••••••••••••• 33
3. Method of measuring the fall of a weight in the sample gas. • • • • •. 47
4. Method of damping the torsional oscillations of a disk suspended
in the gas saln pIe. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •. 50
5. Concentric cylinder method. • • • • • • • . • • • • . • • • . • • • • • • . •• 55
6. Viscosimeter for liquefied gases ••••• • • • • • . • • • • • • • • • • • .• 57

Chapter III. EXPERIMENTAL RESULTS. • • • • • • • • • • • • • • • • • • • • • • • •• 61

1. Viscosity of gases at atmospheric pressure. • • • • • • • • • • • • • • • •• 61
2. Viscosity of gases at reduced pressures • • • • • • • • • • • • • • • • • • •• 61
3. Viscosity of gases at high pressures •••••••••••••••••••••• 68
4. Viscosity of water and stearn •••••••••••••••••••••••••• 97
5. Viscosity of vapor and liquid on the saturation line. • • • • • • • • . •• 104
6. Viscosity of liquefied gases ••••••••••••••••••••••••••• 112
7. Reference data on the viscosity of liquids •••••••••••••••••• 122

<:':hapter IV. THE VISCOSITY OF GASES AT ATMOSPHERIC PRESSURE 133

1. The viscosity of gases at aITnospheric pressure as a function of

telnperature • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • . • • • . .• 133
2. Applica.tion of the law of corresponding states to the determination
of the temperature dependence of the gas viscosity at atmospheric
pressure~ • • • • • • • • • • • • • • • • • • • • • • • • • • • • . • • • • • • • • • •• 143

iii
Chapter V. THE VISCOSITY OF COMPRESSED GASES. • • • • • • • • • • • • •• 151

1. The pressure dependence of the gas viscosity at different

temperatures ••••••••••••••••••• " • • • • • • • • • • • . • • • •• 151
2. Derivation of equation for the viscosity of compressed gases •••••• 159
3. Application of the equation for the viscosity of compressed gas
to experim ental results •••••••••• " • , • • • • • • • • • • • • • • • •• 167
4. The determination of the thermal pressure. • • • • • • • • • • • • • • • •• 170
5. The density dependence of the viscosity of gases. • • • • • • • • • • • •• 177
6. The application of the law of corresponding states to the deter-
mination of the gas viscosity at different temperatures and pressures. 181

Chapter VI. THE VISCOSITY OF GAS MIXTURES. • • • • • • • • • • • • • • . •• 187

1. The viscosity of gas mixtures at atmospheric pressure •••••••• , •• 187

2. The calculation of the viscosity of gas mixtures at atmospheric
pressure ••••• , , • • • • • . • • • • • • • • • • • • • • • • • • • • • • • • • •• 201
3. The viscosity of gas mixtures at high pressures • • • • • • • • • • • • • •• 208
4. The calculation of the viscosity of gas mixtures at high pressures • •. 222

Appendix 226

Bibliography ••••••••••••••••••••••••••••••••••••••••••• 232

List of Abbreviations 238

iv
 This reference book on the viscoSity of gases and gas mixtures gives a survey
of contemporary experimental methods for measuring the viscosity of gases under
various conditions. Several types of viscosimeters are described, and data are
included on the viscosity of numerous gases and gas mixtures over wide ranges of
pressure and temperature.
The published data and calculation methods for the viscosity coefficients of
gases and gas mixtures are analyzed and generalized.
The book is intended for research workers, designers, a.nd engineering person-
nel in chemical industry.

Preceding page blank

VI
 From the relationship which (as already mentioned)
exists between the viscosity and other physical or che-
mical properties, it can be concluded that internal fric-
tion plays an important role in molecular mechanics.
D. I. Mendeleev

PREFACE

The molecular transfer processes in gases (and liquids), viscosity

(internalfriction), thermal conductivity, and diffusion, are very importantfor
explaining phys ical and phys icochemical phenom ena, and also for solving m any
chemical-engineering problems involving mass, momentum, and heat transfer.
The subject of this book is the viscosity of gases and gas mixtures.
The viscosity of gases and gas mixtures and its temperature and pressure
dependence are fundamental constants characterizing the physical properties
of substances.
The most widely studied parameter of a molecular transfer process
is the viscosity eoefficient. It partially determines the coefficients of
diffusion and thermal conductivity, which can be calculated from equations
of the molecular-kinetic theory.
The problem of the viscosity of gases and gas mixtures at high pressure
is discussed at length. Gases under high pressure have densities and some
::>ther physical properties which are intermediate between those of the
gaseous and liqui.d states. It is thus possible to reveal general patterns
::>f molecular transfer phenomena in gases and liquids, and to obtain more
3.ccurate ideas on their mechanism. We shall deal not only with the gaseous
state but also with the critical region and the liquid state, which is
iiscussed only in so far as is necessary to give the reader a better under-
standing of the gaseous state and its continuous transition to the liquid.
Information on the viscosity of gas and gas mixtures is important for
:1Umerous tasks in industrial practice (design, technology, and calculation
I)f processes taking place under atmospheric and high pressures) and the

development of industrial processes. This is because formulas for

hydraulic resistance and heat exchange contain the viscosity coefficient
as one of the most important parameters.
A large number of experimental and theoretical papers have been
published, and numerous experimental data on the viscosity of gases and
i~as mixtures over wide temperature and pressure ranges are available.
:3oviet scientists have been especially successful in carrying out
,~omprehensive measurements and developing the theory of viscosity.
This book describes methods and instruments for determining the
'Tiscosity of gases under very different experimental conditions. Experi-
mental data for almost all the gases studied are given in generalized
tables and graphs suitable for practical use. Data are given on the
'riscosity of gas mixtures, especially at high pressures. The book gives
equations and methods for calculating the coefficients of viscosity most
Imitable for the experimental results.
When selecting the material, special emphasis was placed on its use in
laboratory viscosimetry of gases and liquids and in engineering calculations.

vii
Chapte-r I

VISCOSITY OF GASES

1; General concepts and dimensions

Newton /1/ was the first to form a hypothesis preceding the law of
internal friction (Principia, Lib. II, Section IX).
"On the Circular Motion of Liquids
Hypothesis
That the resistance which arises from the lack of slipperiness of the
parts of the liquid, other things being equal, is proportional to the velocity
with which the parts of the liquid are separated from one another.
Proposition L1. Theorem XXXIX
If a solid i.nfinitely long cylinder in a uniform infinite liquid revolve
round its axis with uniform motion, and the liquid be caused to revolve by
the impulse thereof only, but every part of the liquid persevere uniformly
in its motion; I say that the periodic times of parts of the liquid are as their
distances from the axis of the cylinder." Later, Petrov accurately
formulated the law of internal friction /2/.
In accordance with this law, a force (referred to unit area of contact)
exists between two adjacent layers in a direction opposite to their motion.
This force is proportional to the change in velocity in a direction perpendi-
cular to the motion

(1 )

where !!.W is the change in the velocity of two adjacent layers at a distance
D.z from one another. The corresponding differential equation is

(la)

The proportionality factor TJ is termed the coefficient of internal friction

or the coeffici.ent of dynamic viscosity; hereafter, the viscosity coefficient.
If follows from (1) that TJ is equal to the tangential forces per unit area
of either of two horizontal planes, one of which is fixed, while the other
moves, if the velocity gradient between them is unity.
In the cgs system the unit of viscosity is defined as the viscosity of a
substance in which at a velocity gradient of 1 cm / sec per 1 cm a friction
force of 1 dyne acts per square centimeter. The unit of viscosity is termed
a poise and its hundredth part a centipoise (cpoise).
 Thus, in the cgs system the viscosity coefficient is expressed as

In the mks system the unit of viscosity is the viscosity of a substance in

which at a velocity gradient of 1 m / sec per 1 m a friction force of 1 kg acts
per square meter,
The reciprocal of the viscosity is termed fluidity, and is designated
by:p:

1
<p=-.
~
(2 )

In the cgs system the fluidity is expressed as

[;,01'::1::2'''-':
" g •

The ratio of the viscosity coefficient to the density is termed the

coefficient of kinematic viscosity:

u=l (3)
p

In the cgs system v is expressed in

The unit of kinematic viscosity is termed a stoke.

The three coefficients shown above depend on the type and state (P, T)
of the gas.
The tables below show the conversion factors between the different
viscosity units.

Dynamic viscosity

Conversion into

poise kg hrlm'

Micropoise .••••..•. 10-6 2.833'10- 1'

Millipoise .••••..•. 10-' 2.833'10-'
Centipoise ' .••••.... 10-' 2.833'10- 8
Poise .•..•..••..•• I 1 2.833'10- 6
,
'1!
kg sec/m . . . . • . • . . 98.1 2.777'10-4
kg hr/m' ' ••....... 3.5316.10 5 1

Conversion of British units into metric:

1 pound/foot scc=14.882g/cm 58C; 1 pound sce/sq foot=4.882 kg sec/m'=
= 478.6 gI em sec.

2
 Kinematic viscosity

Conversion into
2
stokes m /hr

Micro~;toke 10 -6 3.&.10- 7
Milliswke 10-3 3.6,)0-4
Centistoke ••.•••••• 10-2 3.6'10 -3
Stoke •..••••••..• 1 3.6.10-1
m2;hr .•..•..... 2.777 1
m2;se" ....•..•.. l'lO" 3.6.10 3

2. The kinetic: theory of gas viscosity

We shall here briefly explain the fundamentals of the kinetic theory of

gas viscosity. The reader will find a more comprehensive treatment in
books on the kinetic theory of gases /3/.
The principles of the kinetic treatment were developed by Lomonosov
in the eighteenth century.
Lomonosov considered" a cause sufficient to generate heat is in the thermal
movement of matter" and, further, "evidently, the individual air atoms approach
one another and collide with the nearest ones within insensible moments of time,
and when they are in contact other atoms which have bounced away from each
other collide with the nearest ones and again bounce away; thus constantly repuls-
ed from one another by frequent mutual collisions, the atoms tend to disperse
from one another."
The kinetic theory of gases was then developed in the second half of tl).e
nineteenth century only, thanks to the work of Clausius, Maxwell and
Boltzmann. The theory is still being improved in our time. It is
successfully employed to explain some physical properties of gases.
In accordance with this theory, gas is an assembly of molecules which
are separated from one another by distances larger than the molecule size.
The molecules move randomly and collide with one another, traversing a
certain path between two successive collisions. The forces of interaction
between the molecules are sufficiently small to
be neglected, with the exception of those at
z the instant the molecules collide. The
/ thermal energy is considered to be the energy
of the moving gas molecules, the temperature
is given by their speed, and the gas pressure
is the result of impacts of the molecules
:c against the walls of the vessel.
This makes it possible to explain many prop-
FIGURE 1. Speed distribution in erties (including the viscosity) of real gases.
the moving gas layers. Let uS assume that a gas flowing in direction x
is separated into layers (Figure 1). With
increase in the height of the layer, its velocity increases. The flow pattern
can be imagined by superposing a directed motion for all the molecules on
the random (chaotic) motion.

3
 Let us determine the momentum transferred by molecules in time Lh
through a circular area L'l.s considered to be isolated in the bulk of the gas.
This area is parallel to the direction of velocity of the gas (Ox- direction)
and perpendicular to the velocity gradient (Oz -direction). Let us first
determine the number of molecules traversing /:;s in 6:;:.
Let us assume, for the sake of simplicity, that in this random motion
the gas molecules have the same speed C, equal to the arithmetic mean of
all the absolute speeds of all the molecules, and that of the total number
of molecules an equal number moves at each i.nstant in three mutually
perpendicular directions. Since in each direction the molecules can
move forward and backward, one sixth of the total number of molecules
move in each direction at each instant.
It is easily seen that in time /:;, only those molecules pass through /:;s
which are separated from it by the distance traversed by the molecules
in .l.:. This distance is c/:;~, i. e., one sixth of the molecules enclosed in
c
a cylinder with base /:;s and height /:;~ pass through the surface from the
top and bottom.
Hence the volume of the cylinder is CLh b.s. If n is the number of
molecules, 1/6 ncb.d.s molecules traverse from the top to the bottom (and
from the bottom to the top). Every molecule transports not only the mom-
entum me (m is the mass of the molecule) but also the momentum mw.
Since w 1 is larger than 1V 2 , an excess momentum !!.z", is transported from
the top to the bottom, equal to

If we designate by w the velocity of the layer, situated on the actual

surface !::,.s, in the direction x, the speeds WI and W 2 of the upper and lower
layers, separated from !!.s by a distance equal to the mean free path':' lof
the molecules, will equal

and

!J.zv
W2 = w---s:zl.
Hence we obtain

By inserting this value into the equation for /:;zw we obtain

1 - /!'w
b.z rn = :3 mncl Tz b.s b.: .

• [The mean free path is the distance traveled by the mOlecui~s between two successive collisions.]

4
 We know from mechanics that the impact (impulsive force) (the product
of the force and time of its action) is equal to the change in momentum
(the product of the mass and velocity change). Therefore

1 - Ilw
F Lh = tJ.z m =:3 mncl-Ilz tJ.s tJ.~.

For F acting upon unit area we have

1 -,1w
F=:3 mncI Tz .

By compari.ng this equation with (1), with transition to the limits for
tJ.w and tJ.z at mn = p, where p is the gas density, we obtain
1 -
'f/="3 pcl . (4)
Thus, the effect exerted by one moving layer on the other is explained as fol-
lows: molecules fall from the upper layer to the lower one, and these move in the
direction x more rapidly than the molecules of the lower layer. Hence, the upper
layer, as it were, attempts to accelerate the motion of the lower layer in direction
x. Conversely, molecules with a lower speed in direction x drift fromthe lower
layer to the upper one. They thus, as it were, decelerate the motion of the whole
upper layer.
Due to such molecule transfer, the total momentum in direction x for the
upper layer somewhat decreases and for the lower layer somewhat increases.
This momentum change characterizes the viscosity coefficient. The
coefficient is numerically equal to the momentum transferred in the moving
gas in unit time through unit area, when the velocity gradient is unity per
unit length. At the same time it is numerically equal to the force acting
on unit area of the gas layer under friction for the above velocity gradient.
Equation (4) was first derived by Maxwell.
Depending on various assumptions of scientists, other equations have
been derived, which differ from (4) in the constant multiplicand. Therefore,
we obtain a more general equation if a constant k is substituted for 113:

'f/ = kpcl. (5 )

Enskog derived the following accurate equation for molecules considered

to be absolute spheres:

5 _ /MRT 1
"1=16 1.016 V-~-.- nd"

where d is the diameter of the molecule, M the molecular weight, and R the
gas constant. If

1= __1__
V2rcnd2
and c= 2 -Vr ~MT
..
•

then

'f/ = 0.499pcl. (5a)

5
 We do not intend to explain the strict mathematical theory of molecule
transfer phenomena developed by Enskog /4/ and Chapman /5/. The
theory is too comprehensive to be treated in this practical handbook.
Of the three magnitudes p, I and c c
in (5), is pressure-independent,
p is directly proportional to the gas pressure and I inversely proportional.
For gas under moderate but not too Iowa pressure, pi is constant, and
hence 11 is pressure-independent. When the pressure decreases, the number
of particles and hence the number of molecules transferring the momentum
from one layer to the other decreases per unit area. On the other hand,
the mean free path of the molecules becomes longer and for this reason the
molecules drift into the given layer without impacts from the more remote
layer moving at a higher speed in direction x. As a result of these two
oppOSing factors, layer-to-layer momentum transfer remains constant.
The pressure-independence of the gas viscosity was experimentally
confirmed for certain pressure ranges only. The limits of these ranges
are given by the following conditions.
Firstly, because of the very large number of molecules per unit volume,
the vast majority of collisions take place in the bulk, and only a negligible
number of molecules impact the walls of the vessel.
However, as the pressure decreases, the relative number of collisions
between the molecules and the walls will increase. But when the path
traversed by the molecules between two successive collisions is of the
same order or larger than the distance between the vessel walls, i. e.,
the distance between the surfaces of exchange of momentum, then most
of the molecules will move from one wall to the other without undergoing
collision on the way.
Consequently, the mean free path of the molecules will no longer depend
on the pressure, but will become approximately constant, and will be
determined by the linear dimensions of the vessel only. At the same time,
the number of momentum-transferring particles decreases with decrease
in pressure in proportion to the pressure or density. If this is taken into
account, it follows from (5) that high rarefactions (in the vacuum range)
decrease the viscosity coefficient of the gas, and at P -+ 0, this also tends
to zero.
Secondly, at high pressures, when the forces of interaction between the
molecules considerably influence the state of the gas, and when the size
of the molecules in comparison with the mean free path can no longer be
neglected, the viscosity coefficient also becomes pressure-dependent.
Furthermore, ~ follows from (5) that the viscosity should be temperature-
dependent, since c is proportional to the squ3.re root of the Kelvin
temperature. However, in practice it was found that for imperfect gases
the temperature dependence differs appreciably from that predicted by
theory.
Let us now discuss in detail the mean velodtyc and the mean free path
I of molecules, the decisive parameters for molecular transfer in gases.

3. Determination of the mean velocity of molecules

Because of the random motion of molecules, their velocities differ

greatly in both direction and magnitude. Thus, to determine the mean

6
velocity (arithmetic mean of the absolute molecular velocity), we must
know the la'v of velocity distribution.
Maxwell was the first to formulate such a law for the equilibrium state,
when all the parts of the gas are at the same temperature, and there is a
completely determined time-independent distribution of molecules in
accordance with their velocity.
The analytieal form of Maxwell's law is

dnc= n(c)dc = 4 .. n (27t~T rz

3 1 me'},
e -kT-z-.C2 dc. (6)

Here dnc is the number of molecules per unit volume, with a velocity
between c and c·+dc, and n is the total number of molecules. The arith-
metic mean of the velocities is obtained by dividing the sum total of
the velocities of all the molecules present in the given volume by the
total number of molecules. The mean velocity can be expressed by the
following equation:

- If
co

e=li ednc· (7)

o

By substituting the value for dnc from (6), we obtain

3 co me')
C= 41' (~)
2dT
2f e-2k'f c dc. 3

Since j~e _ me!!

2kT c3
2k?T2
de = ----;n2 ,
o

(8)

Boltzmann's constant k and the gas constant are connected by

kNo=R,

where No is the Avogadro number, and mN o = M is the molecular weight of

the gas. Therefore (8) can be written

-,= 1.6 .rIff

V M' (9 )

Other concepts used in the kinetic theory of gases are the mean-square
and the most probable velocities.
If the sum of the squares of all molecular velocities in a given volume
:~s divided by the number of molecules, and the square root of the result
c:aken, we obtain the mean-square velocity, which in analogy with (7) is
defined by

- If
co

c2 = -n e2 dn c· (10)
o

7
By substituting the value for dnc from (6), we obtain

-2 - 4-" ( 2-;;kF
c --
3
m )" • e liT c4 d c.f-
(Y) mc2

o
Here

Thus,

V2-1
c - ~3·rkT
.1 V -,nl" ,
or

V~=173-r~ . (11)

The most probable velocity ep is defined as the velocity with which the
largest number of molecules move at any given instant, or in other words
when the distribution function of the molecules with respect to the velocity
nee) is a maximum. To find e when this function is a maximum, the
explicit form of the derivative of n (e) with respect to e must be equated to
zero, and the equation obtained solved for e.

fI(l)

__ --- e
I~,, ."_"_~_~--~~_~_~-~~&L---~
'i-'---,,--- Average ve,ocity
\ Mean-square velocity
Most probable velocity

FIGURE 2. Distribution of gas mole-

cules in accordance 'v,,rith their velo-
cities.

Thus, for (6) we obtain

3 mc2
4'm (----'!!-.-)" e -",1' 2c (1 -- ..!!!....
21tkT 2kT
c2) = 0

or
me'
e . e - nT" (1 - -.!'!....
2kT
C2 ) -
-,
0

8
whence

(12 )

(The second solution c = 0 has no physica.l meaning.)

It follows from (9), (11), and (12) that the mean-square velocity has the
highest value, and is followed in descending order by the average and the
most probable velocities.
Figure 2 shows the graph of the distribution function of the molecules,
and the magnitudes of these three velocities are compared.

4. Determinatlon of the mean free path of molecules

Let us assume that a molecule A travels with a constant velocity between

stationary molecules of the same type, and collides with them. At the
instant of collision, the centers of the molecules are separated by a distance
d, equal to the molecular diameter. Let us substitute a disk of radius d,
in position normal to c,
for the traveling molecule, and points in the same
position as the molecular centers for the stationary molecules. If during
its travel the disk encounters (or contacts) such a point, we can assume
that a collision takes place and the disk changes its direction.
Thus, during its travel the disk will circumscribe several cylinders
with bases 'ITd 2 and heights proportional to c~i' C~2' C~3, • • • , where
~l' ~2' ~3 , • • • are the times the molecules travel between two successive
collisions.
If a disk circumscribes N cylinders per sec, the volume circumscribed
will be

'ITd 2 ;;~I + ltd 2

C~2 + ... + ltd 2
C:C N =
='ITd2c(~1 +~2+

since

We have not taken into consideration that the volumes of the individual
cylinders partially overlap at the points of collision. However, if the path
is considerably longer than d, this overlap is a very small fraction of the
volume, since the linear parts of the path are large in comparison with
those parts where the cylinders overlap.
The number of collisions of molecule A with other molecules is given by
c
the number of molecular centers present in the volume 'ITd 2 circumscribed
by the disk.
If n molecules are present in unit volume, the number of collisions z per
unit time is given by

(13 )

9
 Hence, the mean free path of a molecule, i. e., the mean distance
traveled by a molecule between two successive collisions, is

1=£=_1_. (14 )
z2 7td n

Clausius assumed that not only a single molecule A but all gas molecules
travel in all possible directions, and derived the formula

3
1= 4"d zn' (15 )

Under given temperature and pressure conditions the lengths of the free
paths of the molecules of a gas (like their vE,locities) have very differing
values, and there is a certain distribution of molecules in accordance with
their free paths. If we take this into account we get a more accurate
expression for I.
Let a fraction of the molecules no: of the total number n travel a path
equal to or greater than x (without collision) and another (smaller) fraction
of the molecules n(xdx) travel a path equal to or greater than (x+dx) without
collision.
By subtracting n(x+dx) from nx , the number of particles nx traveling
distance dx decreases; let us denote this decrement by dnr.. It is pro-
portional to d x, since the collision probability on a small section will
increase with increase in length. Moreover, dn x is also proportional to
nx ' Thus,

(16 )

where a is a proportionality constant.

By integrating (16) we obtain

Innx=-"-x+c.

The integration constant is determined from the conditions nx = n at

x=O. Hencec=lnn. Thus,

(17 )

The magnitude e- ac is the probability of the event that a given molecule

on its path of length x collides with another molecule.
Let us determine a. If a single molecule completes z free paths per
unit time, then z., = ze-- ar paths are longer than x and Z(x+dx) = ze- a (,C+dl') are
longer than x+dx. The number of paths longer than x but shorter than
x + dx will be equal to
Z",-Z(x-'-dxl =ze-ax-ze- a (.vl-dx) ,=ze--a.r(l-e- ad .c ).
By expanding e- a d.r we obtain

a dx a dx 2 2
e- adx = 1--1!-+--w-+

10
If we omit the higher-order terms from the last two expressions for the
number of free paths completed by the molecule in unit time with lengths
between x and x+dx, we obtain

Zx - zx+dx == ze--ct:t'o: dx, (18)

. where r;.e-a x dx denotes the probability that a molecule has a free path length
larger than x but smaller than x+dx. By mllltiplying (18) by x, we get
the sum total of the free paths of lengths between x and x+dx:

z(J.e->·"x dx.

By integrating this equation for x from zero to infinity, we obtain the' over-
all lengths of all the free paths completed by the molecule per unit time,
i. e., a magnitude numerically equal to the velocity of the molecule:

Zo..r e->J:x dx = c.
o
Here

and hence
z
-=c.
"
Since C/7 is equal to the average length of the free path I,

1 (19 )
o.=z·
From (17) and (19) we obtain

(20)

Equation (20) gives the distribution law of the lengths of the free paths,
and from this we can calculate the number of molecules of the total
number n with a free path equal to or larger than x. If, for example, we
:mbstitute 1 for x, then

n n
nl =e = 2.718 = O.3679n.

i. e., about one third of the molecules present in the given volume have a
free path equal to or larger than I.
Maxwell used (20) to obtain the following equation for the mean free path
of the molecules:

/= ___1_.
V
2 "d 2 n (21 )

11
 In a mixture consisting of two different gases the mean free path of each
molecule is determined as follows.
Let n, be the number of molecules of the first type, and n2 the number of
molecules of the second type per unit volume, d, and d 2 the diameters of
the first and second types of molecules, i. e., the distances between the
centers at the instant two molecules of the same type collide, s the distance
between the centers at the instant two molecules of different types collide
(we assume that s = t
E., ~), c, and C2 the velocities of the molecules, and m1
and m2 their masses.
The probability of the event that a certain molecule A will not collide
with another molecule of the same type as it traverses x is equal to
n
e -V2 ,di
1'" The probability of the event that the molecule will not collide
with a molecule of the second type on the same path is given by

Hence, the probability of the event that the molecule on its path will
not collide with a molecule of either the same or different type is given by

-"It

e
( ,-, ..Vr-:r )
}' 2 dinl + 1 + c~ s'n 2 x

If I, is the mean free path of a molecule of the first type, we obtain

(22 )

Similarly,

(23 )

Since both the gases are under the same conditions, the following
relationship holds:

where M2 and M1 are the molecular weights of the gases. By substituting

the values for c; and c~ into (22) and (23), we obtain

(24)

and

(25 )

12
 If Y] is known we can also calculate the molecular diameter from (5),
since p can be experimentally determined, C is calculated from the above
equation, and d is calculated from I using (21).
Table 1 shows the mean velocity C, the mean free path I, and the
diameter of different gases (all the values in the table are calculated
for STP).

TABLE 1
-------
Gas c, cm/sec I, A d, A
Nitrogen. ·. · . · . 45,430 599 3.1
Argon. · . ... 38,080 635 2.8
Hydrog"n · . . . · . · . .... . 169,200 1123 2.3
Water vapor ....... ·. · . ·. ·... 56,650 404 2.6
Helium •. ... . . · . · . .. 120,400 1798 1.9
Oxygen. · . · . .. . · . 42,510 647 2.9
Krypton •.· . · . .. · . ·. · . 26,410 487 3.2
Methane. ·. · . .. · . 60,060 493
Carbon dioxide. · . .. . . 36,250 397 3.2
Carbon monoxide · . · . .. . · . 45,450 584 3.2
Chlorine. · . · . · . ·. · . · . 28,560 287 3.6
Mercury (:lOO°C) ·. 17,000 217 3.6

~). Analogy between viscosity and other molecular transport

processes (thermal conductivity and diffusion)

Gas viscosity is directly related to other molecular transport processes,

diffusion and thermal conductivity. These three phenomena are analogous
because all involve the transport or transfer of a certain magnitude
(momentum in viscosity, mass in diffUSion, and thermal energy in thermal
c:onductivity). The transport continues until the given magnitude is uni-
formly distributed over the whole gas. These processes are irreversible;
their result is the equalization of the densities, temperatures, and
velocities, and hence the equilibrium state with a minimum free energy
and maximum entropy. According to the kinetic theory, the mechanism
of all three processes is formally the same, and therefore the equations
involved also agree formally.
For viscosity, we have the equation

then the corresponding equation for the diffusion is

D-
-32.cl (26 )

13
and for thermal conductivity

(27 )

A more rigid treatment leads to the replacement of the number l/3 by

another constant.
The following equations give the relationships between l'], D and >C,

D=r:t.!l
p , r:t.=1.3, (28)

(29 )

The proportionality constant € is approximately equal to 2.5 for mono-

atomic gas, 1.9 for diatomic, and 1.75-1.5 for tri- and polyatomic. It
was shown, however, that E: is temperature-dependent to some extent.
Equation

1
e=4(9y-5), (30)

where 'Y agrees with the experimental results /3/.

Experiments on the viscosity and thermal conductivity of gas at high

pressures show that € is also pressure-dependent.

14
Chapter II

MEASUREMENTS OF GAS VISCOSITY

Numerous papers contain accurate data on the viscosity of gas and gas
mixtures obtained at atmospheric pressure and different temperatures.
There are fewer measurements for the high-pressure range, but these
are still fairly numerous. The determination of gas viscosity at high
temperatures involves difficulties, and the results available are less
accurate than those at atmospheric pressure. There is still room for
research in this field. For this reason we shall describe viscosimetric
methods for the high-pressure range in great detail.
The following three general methods are most widely applied:
1) capillary method (the viscosity is determined in a gas flowing through
a capillary);
2) falling weight method (a weight falls in the gas to be studied);
3) a method based on attenuation of the torsional oscillation of a disk
suspended in the sample gas.
Only method :[ is strictly justified by theory. The theoretical basis
was experimentally confirmed.

1. Capillary m,ethod

We distinguish two types of flow phenomena in tubes: laminar (jetlike)

and turbulent (eddy) flows.
The dimensionless Reynolds number (Re) is the decisive parameter
for laminar and turbulent flow. The Reynolds number can be considered as
a magnitude expressing the ratio of the inertial force to the viscous force
in the gas flow. If the Reynolds number is comparatively small, laminar
flow predominates. In this case viscous forces equalizing the flow
predominate over inertial forces. If, on the other hand, Re is large, the
flow is turbulent. Laminar flow abruptly passes into turbulent at a certain
Re. At the same time, the drag of the flow abruptly changes. It was
found that the critical Reynolds number for straight, smooth tubes is

Re = w.,d = const ::::: 2300. (31 )

where w is the mean linear speed of the gas, d is the diameter of the tube,
md v is the kinematic viscosity. It can be seen that, other conditions

15
being equal, the critical flow velocity corresponding to the transitioYl from
laminar to turbulent increases with decrease in the tube diameter.
Therefore, if a gas streams through a
capillary, laminar flow is obtained, even
- -- - - - - - - - - - --I~-''''''\
when the gas speed is high. Since in laminar
/! \ p flow the transverse component of the velocity
----·-·-H--·--t -? vector is zero, and only the component
" ,
" \ , ! , ,II parallel to the axis of the tube differs from
------------- .........
~
zero, the mathematical problem is not
e----L - - - - very complicated. Thus, the velocity
distribution along the tube diameter can be
FIGURE 3. Derintion of the Hagen determined.
- Poiseuille equa tion.
Let us consider a capillary of radius r ,
(Figure 3) and length L. The variable
distance of a certain point from the axis
will be designated by r. Let the pressure on both capillary ends be PI and
P2. The pressure gradient Pl- P2 is responsible for the gas flow through
the tube, and because of the frictional force between the gas and the tube
walls the flow is steady- state. Figure 3 shows a cylindrical ring lying
inside the capillary, with inner and outer radii of rand r+dr, respectively.
The area of the base of this ring is equal to 27frdr. Thus, a pressure
force equal to

K = (P1-P2) 2r:rdr (32 )

acts on the gas present inside the ring. The gas layer adjacent to the
inner side of the ring moves more rapidly than the gas layer in the bulk
of the ring, and conversely the gas layer adjacent to the exterior of the
ring moves more Slowly. Hence, the mass of the gas enclosed in the ring
will be accelerated on the inner side (at a distance r from the axis) and
retarded on the outer side (r+dr) by the frictional force. Let us designate
the first force by Ri and the second by Ra; then

(33 )

(34 )

In these equations w is the mean (for the ring) flow velocity of the gas,
and Fr=271'rL and Fr~dr=271'(r+dr)L are the inner and outer lateral
surfaces of the ring. The + and - signs denote that Ri accelerates and Ra
retards the flow, or Ri> 0, Ra<O, since ~~ is negative. In steady-state
flow the forces K, Ri , and Ra are in equilibrium. Therefore the
equilibrium state can be described by

16
Let us expand (!!dWr ) r+dr into a Taylor series and retain only the first two
terms. Thus,

The left-hand side of (35) is constant for any given case. The equation
can be transformed into

~
dr
(r dW)
dr
=_ PI - P2
~L
r
to make it integrable. After integration

W=- P1
4
-;l' r2+AIg r+B, (36)

where A and B are the integration constants. At r = 0, i. e., on the tube

axis, W = 0, then A = 0; B is a function of the conditions on the
tube walls. If the gas does not slide along the walls, w r , = 0, and

B = PI-P2
4~L
r2
I"
(37 )

Thus, the solution of the differential equation is given by

_ PI- P, (
W---;r,;L r - rl"
2 2) (38 )

The volume of gas flowing per sec through the capillary is obtained by
integrating (38) over the whole cross- sectional area of the capillary (i. e.,
for values of , from zero to '1):

(39 )

Equation (39) is termed the Hagen -Poiseuille equation.

This is the fundamental equation for the capillary viscosimeter. For
the viscosity determination the capillary dimensions, the pressure
gradient over the capillary, and the gas volume flowing through it per
unit time must be known. •
If we multiply both sides of (39) by p (the mean density corresponding
to PI +2 P2) ' we obtain

(40)

for the mass of gas flowing through the capillary in time '"
For a perfect gas where p = ~j ,

M
ror; (PI - p,) mp (41)
8"L "W"

17
 In the practical application of the Hagen -·Poiseuille equation corrections
must be applied. These are determined by the ratio of the capillary radius
to the mean free path of the gas molecules. Thus, if the mean free path is
of the same order of magnitude as the radius, or larger, flowing gas does
not adhere to the tube wall, but slips along it with a certain velocity differing
from zero. Thus, we must assume a finite frictional force between the gas
and the tube wall proportional to the velocity gradient of the gas near the
wall /6/. In this case constant Bof (36) will no longer be determined by
(37). To obtain equations describing the conditions near the tube wall
(boundary conditions), let us again consider a cylindrical ring bounded on
the outside by the tube walls. (The outer radius of the ring is equal to r1
and the inner to r,- dr.) In this ring the gas is driven by the 'pressure
equal to

(42 )

K is in equilibrium with the two frictional forces Ri (at a distance of

r=r,-dr from the axis) and Ra(at a distance r=r,). Thus:

Ri =-TjPr,-dr( ~;) . (43 )

Tj-dr

(44)

Here Wo = w r , is the velocity of the gas near the tube walls (boundary layer).
The minus sign denotes that Wo is positive in contrast to ~; in (34).
The proportionality constant 'fIa is termed the coefficient of internal
friction. Thus, as before, we obtain

27!r (Pl- pz) dr-TjL2" (rl-dr) [(~il, - (~:~t drJ--

-Tja 2 "'1 L1vr, =0. (45)

The terms containing dr and dr 2 are infinitely small compared to the other
terms, and are therefore neglected. For the velocity of the boundary layer
we obtain

(46)

To determine B, (46)is compared with (36), in which A is equated to zero.

It is easy to see that

(47 )

The magnitude (= Tj/Tja is termed the coefficient of slip.

To calculate the gas volume v flowing through the capillary per second,
(47) must be integrated over the cross-sectional area of the capillary:

(48)

18
 For the mass of gas, from (41) we obtain

7tr~ (p 1 - P2) mp ( 4~ ) (49)

M= 8~L RT 1 +-,:;:- .
A kinetic analysis of the coefficient of slip shows / 3b / that this magni-
tude is of the same order as the mean free path of the molecule. Therefore,
slip is effective only ifl is the same (or close) order of magnitude as the
capillary radius. This occurs at very low pressures, since I is inversely
proportional to p. Thus the coefficient of slip is also inversely proportional
to p (the experiments of /7/ and /8/ confirm this). For this reason,
« is sometimes replaced by
a
~, where a is an empirical constant.
~ p
Thus, (49) can be written as

(50)

The pressure drop in the capillary leads to an "expansion" (change in the

specific volume) of the flowing gas. For this reason another correction
must be introduced into the Hagen-Poiseuille equation. For the liquid
state this correction need not be introduced, since liquids are almost
incompressible. For gases under high pressure, when the pressure drop
/:l.p in the capillary is very small compared to the mean pressure to which
the gas is subjected, the correction can also be neglected.
Meyer /9/ took into account the "expansion" of the gas in the capillary
and gave the following equations:

(51 )

when at a measured pressure PI volume v l flows through the capillary, and

(52)

when at measured pressure P2 volume v 2 flows through it.

Meyer aSSumes that the gas flows under isothermal conditions, which
is usually true since the capillary is thermostated during the measuring
process. Fischer /10/ believes that a polytropic process occurs, and
notes that the viscosity coefficients calculated from (51) and (52) are some-
what high. Furthermore, Meyer assumed that the behavior of the gas is
governed by the Boyle - Mariotte law. Trautz and Weizel /11/ took into
account the deviations of real gases from the perfect gas law, and derived
a more accurate formula. This was later used by Trautz and Narath /12/
to calculate the corrections for measurements carried out by earlier
workers; in some cases the correction was as high as 3.80/0.
A correction for the kinetic energy must be introduced for gases under
pressures above atmospheric (especially high pressures), and for liquids.
Hagen and Foiseuille observed that (39) holds for gas flow in fairly narrow
and long tubes only. In shorter tubes considerable deviations from this
equation were found.

19
 When Hagen /13/ processed his data, he assumed that

(53)

holds.
The pressure drop corresponding to the first term of the right-hand side
of the equation is caused by the frictional resistance, and that corresponding
to the second term by the production of the kinetic energy of flow (~~ is
the mean flow velocity).
Hagenbach /14/ was the first to give an acceptable interpretation of this
phenomenon, and for this reason the term "Hagenbach's correction for the
living force" is still used [in Russian}.
When a gas flows from a reservoir through a tube, its potential energy
is partially consumed to overcome friction, and the other part is trans-
formed into the kinetic energy of the moving gas particles.
By dividing the right-hand and left-hand sides of (39) by the cross-
sectional area F = rtr;, we obtain the mean flow velocity in the capillary

whence

(54)

For the maximum velocity (on the capillary axis), from (38) we obtain

(55 )

From the above two equations, w max = 2 w',

i. e., for laminar flow the gas
velocity along the capillary axis is twice the mean velocity of flow. Let
us designate the additional pressure drop due to the conversion into kinetic
energy of the moving particles by /:'P"
The potential energy obtained by multiplying b.Pl by the volume rate (vol-
ume of gas flowing out of the capillary per sec), i. e., /:'Pl7triw, corresponds
to /:'Pl' By comparing this energy with the kinetic energy of the gas jet
discharged, we obtain

(56)

From (38) and (55), we have

ri - r? - ri - r2
w=wmax--o-=2w--o- ,
'1 ri

By inserting this into the right-hand side of (56), we obtain

j'27:P ~3 r dr = 7:pw 3r;. (57 )

o

20
 From (56) and (57):

(58)

Thus, equation (54) supplemented by the term accounting for the kinetic
energy correction can be written as

(59)

For a long time the problem as to when this correction must be applied,
and to what degree it improves the result,- was left open. Some authors /15/
believed that the correction is to be intro-
duced only if the compressed gas or

~
---------"------------!/!I
~-~---------------------r liquid jet is freely discharged into air. If,
I:
:' :
I
f
I P
fJf~
Ps.._:E--------~
P,-P2
pwt.
--...L ____ ___ L
on the other hand, the jet is discharged
into a compressed gas or liquid the

AiA / If
~~~h , B
potential energy lost is compensated
completely or to a large extent by the
pressure increase (Bernoulli equation).
Other authors thought that if a gas flows
out of a capillary into a gas subjected to
a similar pressure, a free jet is formed,
FIGURE 4. Pressure distribution in gas and its energy is used for eddy formation
flow through a capillary. or converted into heat, and therefore
the correction must be applied. Erk /16/
studied this problem in detail. His
results showed that a correction even higher than the theoretical must be
used. Thus, instead of the factor 2 in (59), the number 2.2 should be
applied.
The following flow pattern for gas through a capillary is the most
probable.
In the re gion A to I (Figure 4) the gas is accelerated to a velocity
prevailing at the inlet into the capillary; the dynamic pressure Pd increases by
pf (w is the mean velocity of the gas flow in the capillary), and the static
pressure Ps drops by the same magnitude.
The overall pressure in this section remains constant, Pg=Pa+Ps = const.
The friction is neglected because of the large cross section of the tube
from which the gas flows into the capillary.
However, at the instant the gas enters the capillary the velocity distri-
bution in the stream does not yet correspond to the Hagen -Poiseuille
law, i. e., the velocities are not parabolically distributed over the cross
section of the capillary. But because the kinetic energy of laminar flow
is p~2, in our case we must increase the kinetic energy of the flowing gas
by p~2 to obtain the parabolic velocity distribution in the stream. Therefore
Pd increases by P~ and Ps correspondingly decreases. This is shown by
the section IK on the diagram. Here the internal friction becomes effective,
and more energy is required to overcome this (in the initial section "eddies"

21
are formed) than in laminar flow. The magnitude y in Figure 4 corresponds
to this energy increment. In the laminar flow region (staring from point K)
Pd remains constant in the capillary but Ps and Pg decrease linearly in
accordance with the Hagen - Poiseuille law.
Finally, at the outlet of the capillary in the gas jet inside the wide out-
let tube Pa is lost by friction (evolution of heat), Ps remains constant,
and Pg drops to the value of the static pressure Ps prevailing at the
capillary outlet.
Thus, the pressure gradient measured in the capillary is higher than
that predicted by the Hagen-Poiseuille law by pw 2 +y. By combining both
corrections, pW2+ y= tnpW 2 , we can write

(60 )

and for the viscosity

1Crfl: rnpv
'1]= 8Lv (Pl-P2)-S;[-:;:· (61 )

Various authors /17-20/ give different values of m (between 1.10-1.16).

The value of m which agrees best with the experimental results is 1.12.
If we take into account Hagenbach's correction, and consider the
"expansion" of the gas flowing through the capillary as isothermal, then
from (51) and (52) the viscosity can be written as

(62 )

and

(63 )

In (62) and (63) all the main corrections to be applied in the capillary
method are accounted for (corrections for kinetic energy, slip, and the
expansion of the gas flowing through the capillary). The relative magnitude
of each correction will vary with the experimental conditions. It can be
stated in general that at low gas pressures the last two corrections must be
taken into account, and the first can be neglected. Conversely, at high
gas pressures only the kinetic energy correction is important.
It appears from the above that corrections involving flow conditions at
the inlet and outlet of the capillary have no rigid theoretical basis. It
should be mentioned, however, that in viscosity measurements by the
capillary method these corrections can be either experimentally determined,
or the conditions adjusted so that they are very small.
The capillary method was used in viscosity measurements for many
gases and gas mixtures. The data are accurate, and the results obtained
by various workers are in good agreement.
We shall now describe the most typical capillary viscosimeters.

22
 The Schulze viscosimeter /21/. This capillary viscosimeter can be
used for measurements at constant pressure. It is shown in Figure 5. The
gas sample is in two glass cylinders a and b, which are successively
connected to a mercury-filled vessel. On the other side they are connected
to capillary k and a mercury seal m. Vessel c is shifted along a calibrated
strip d. When the mercury flows into the glass cylinders a and b, the
displaced gas is forced through the capillary k.
The pressure on both sides of the capillary is measured by water gages.
During the measurement the pressure gradient is kept constant at
approximately 6 cm H 20 by adjusting screw e. The gas volume flowing
through the capillary is recorded by two fused-in contacts g and f. When
these are in contact with the mercury, an electric bell rings. The
cylinders a and b and the capillary k are immersed in a thermostat during
the experiment. The devices for pressure equalization, filling the gas,
etc., are not shown in the figure.

FIGURE 5. The Schulze viscosimeter.

The Trautz and Weizel viscosimeter /11/. The instrument (Figure 6)

consists of two vessels, an upper (200cm 3 ) and a lower (250cm 3 ), joined
by aU-tube. The capillary k in thermostat t2 is connected by a three -way
stopcock to the lower vessel, in thermostat t 1 , and to the gas reservoir.
A measured volume of mercury is introduced into the instrument in an
amount sufficient to fill the whole lower bulb and half the upper bulb.
The gas leaving the reservoir displaces the mercury from the lower bulb
to the upper. The stopcock is then turned so that the gas can flow through
the capillary to the atmosphere. First, the mercury level decreases to

23
mark m! in the upper bulb. At this moment a stopwatch is started. The
watch is stopped when the mercury level in the lower bulb reaches mark m 2
on the 5-mm-diameter tube. The time measured by the stopwatch
corresponds to the time in which a known amount of gas flows under a
given pressure.

FIGURE 6. The Trautz and Weizel viscosimeter.

The method gives relative rather than absolute results. However,

Trautz and Weizel calibrated the instrument, and calculated graphically
and analytically the corrections necessary for determining the absolute
viscosity coefficients. It is easy to see that lor a given instrument and
a given amount of mercury, the mean pressure gradient and the volume of
the gas displaced are a function of the design of the instrument and hence
constant; this facilitates the relative viscosity measurements.
By this method Trautz and his co-workers measured the viscosity of
many gases and gaseous mixtures over a wide temperature range.
The Sigwart viscosimeter /22/. The Sigwart viscosimeter for the
determination of the viscosity of water and steam at various temperatures
is shown in Figure 7.
The steam flows from boiler F through platinum capillary K (a quartz
capillary may be used). The boiler is a cylindrical vessel with the follow-
ing dimensions: length 455mm, inside diameter 40mm, outside
diameter SOmm, total volume 475cm 3 . The capillary K is inserted into
a high-pressure vessel D closed at both ends by steel caps A and B. These
are fixed to vessel D by bolts C. Two tubes p pass through A and B to the
ends of the capillary. They measure the pressure gradients at the ends of
the capillaries. The tubes are fixed to a steel cross piece E, fitted with
an intermediate valve u to equalize the pressure at the ends of the capillary,
if necessary. Two flexible capillaries lead from the cross pieee to a
differential gage l:>p. The pressure drop at the capillary ends is determined
here as on an analytical balance. The differential gage is only schematically
drawn in the figure. The reader will find a detailed description in
Schmidt's paper /23/.

5680 24
 b

FIGURE' 7. The Sigwart apparatus and viscosimeter:

F-boiler; D-high-pressurevessel with caps A and B; K-capiHary
made of quartz or platinum; E- steel cross piece with intermediate
valve U; Llp - differential gage.

The steam ps.ssing through the capillary enters the condenser through
choking valve b. The condensate is collected in a vessel, and the amount
e,f steam deterrr:ined.
The left end of the capillary is hermetically joined to vessel D by a
Epecial device e, so that steam can pass from the right-hand side of
vessel D to its left-hand side through the capillary only. On the right-hand
Eide of the vessel there is a clearance between capillary K and socket H.
Therefore, vapor or liquid pressure is transferred to the outside surface
cf the capillary over its whole length.
The following capillary dimensions were used: quartz length L= 348.8mm,
nean diameter d m = O.3889mm, platinum length L = 370.4mm, mean
diameter d m = O.5484mm.
The other parts of the equipment can be seen in the figure.
In the experiment we determine the amount of steam passing through
t.1e capillary in unit time at a given 6.p, temperature, and pressure.
Sigwart calculated the viscosity coefficients from these data by the Hagen-
Poiseuille equation, introducing a correction for the kinetic energy.
The disadvantage of the Sigwart viscosimeter is the very complicated
differential gage. It needs careful centering and calibration, and special
skill is required.
With this apparatus Sigwart determined the viscosity of water between
116.9 and 371.4"C: at pressures of 10.5 to 253kg(cm 2 , and the viscosity of
steam between 276.8 and 382.9°C at pressures of 25.5 to 270kg/cm 2 •

25
 The Rankine viscosimeter /24/. Rankine /24/ designed and used a very
simple and accurate viscosimeter (Figure 8). It has often been used to
determine the viscosity of gases.
A capillary K and a wider glass tube D form a closed ring. A drop of
mercury E falls in tube D, held in a strictly vertical position. The weight
of the mercury creates a certain pressure, which forces
the gas filling tube D to flow through the capillary. The
direction of the mercury drop can be reversed by turning
the instrument through 180 around the horizontal axis.
0

c It can be seen on the figure that the upper half of the

viscosimeter is similar to the lower one, so that the
D instrument is completely symmetrical about the axis
A
(shown on the figure by the dotted line). The viscosimeter
is placed into a wide glass tube C. To attain the tempera-
E
ture necessary for the experiment, tube C is filled with a
thermostatic fluid, which enters through tube P and leaves
p through tube Q. The stopcocks T at the ends are used
to fill the viscosimeter and blow gas through it.
We must accurately know the gas volume between marks
B A and B, and the weight of the mercury drop. The time
the mercury drop falls is measured by a stopwatch.
This viscosimeter is more suitable for relative measure-
ments' since for absolute determinations several corrections
must be introduced. The most important is the correction
for the effect of surface tension on the movement of the
mercury drop. The pressure difference c"p is not
FIGURE 8. The accurately equal to the weight of the mercury divided by
Rankine viscosi-
the cross section of tube D, since the upper surface of
meter.
the falling drop is less curved than the lower one. Con-
sequently, the capillary forces are not equal at the ends of
the capillary and their resultant brakes the fall. In
earlier experiments Rankine found the pressure drop caused by surface
tension. He introduced mercury drops of different weights, and measured
their time of fall.
A better method was employed by Kuenen and Visser /25/; here a
constant weight of mercury is employed, and the time is measured during
which the mass falls as one, two, or three drops (i. e. the total mercury
mass is split into the required number of drops). The authors plotted
1/,1' vs. the number of drops n, and found an approximately linear
relationship. The same method was employed by Rankine and Smith /26/,
who found that the action of the capillary forces is very dependent on the
temperature, but only slightly on the nature of the gas with which the
mercury is in contact.
It should be noted that the Hagen-Poiseuille equation is accurately
valid for the Rankine viscosimeter only in cases when the flow rate
through the capillary is independent of the position of the drop in the tube.
For an incompressible liquid, the falling speed of the drop along the whole
tube is constant, and for a compressible one it is variable. Rankine showed
that if the total gas volume forced out by the mercury drop is symmetrically
distributed over the volume of the viscosimeter, w is constant (1/ 2700 ),
provided that the length of the drop is less than 5 cm and the gas is under

26
atmospheric pressure. If the gas volume is not symmetrically distributed
in the upper and lower halves of the viscosimeter, corrections must be
introduced.
The following rule applies to this and other capillary viscosimeters:
the length-to-diameter ratio of the capillary should be as large as possible
to reduce the end effects to a minimum.
Rankine recommends the following dimensions: length of capillary 50 em,
diameter of capillary 0.2 mm, diameter of tube for mercury 3 - 3.5 mm,
distance between marks 30 cm.
It is very di.fficult to clean and dry the instrument, because the capillary
and the tube in which the mercury falls are parallel, and therefore the
solvent or dry gas flows very slowly through the capillary. It is also very
difficult to remove any mercury which may accidentally enter the capillary.
The Comings, Mayland, and Egly viscosimeter /27/. This viscosi-
meter consists of a glass capillary K and tube C joined in parallel
(Figure 9). The dimensions of one of the
instruments used by the authors are shown in
the figure, the dimensions of the second viscosi-
meter are similar.
The mercury drops weigh 1.5 to 5 g. Above
and below tube C two tungsten wires are soldered
(contacts) and the time of fall of the mercury
drops between the two contacts is measured.
The viscosimeter is fixed in a steel block (see
figure), and placed into a high-pressure vessel
which can be turned through 180 around the
0

axis passing through the center of the instrument

and the axis perpendicular to it. Thus, the
viscosity can be measured while the mercury
falls from both ends of tube C. To facilitate
the reversing of the instrument, the gas is
supplied through flexible long coils able to with-
stand high pressure. During the measurement
the glass viscosimeter is filled with the gas
sample, and the space surrounding it, i. e., the
high-pressure vessel, is filled with nitrogen
FIGURE 9. The Comings,
to generate a compensating pressure. The
Mayland, and Egly viscosimeter.
inlet gas tube is heated to prevent condensation
of some of the gases at the high pressures
employed. The high-pressure vessel has seven
openings at one of its ends. Through one of these the gas sample is fed,
and through the others are inserted two thermocouples and the two electrodes
to the tungsten contacts. The other end of the high-pressure vessel has two
holes for the third thermocouple, All three iron-constantan thermocouples
are used to measure the nitrogen temperature at points near the ends, and
at the center of the viscosimeter, Nitrogen is introduced through a hole on
the lateral side of the high-pressure vessel.
For a complete removal of the gas left in the viscosimeter from a
previous experiment, the instrument is exhausted several times at certain
time intervals, and then filled with the gas sample.

27
 Since the authors were not able to design 8. symmetric viscosimeter
for high-pressure conditions (the volume of one half was four times the
volume of the other), they had to introduce corrections for the variable
falling speed of the mercury drop through the tube. This correction,
however, did not exceed 0.010/0 of the measured value, on average, and
therefore it was neglected. Also, the correction for the end effect
(Hagenbach correction) could be neglected, since the length-to-width
ratio of the capillary was large (3670 and 3850 for the two models,
respectively).
The surface tension effect was determined as follows.
The coefficient Ct used in the correction, which expresses the relationship
between the weight of the drop, equal to m Peg - ~(m is the mass of the mercury
PH,
drop and PHg and PG are the densities of the mercury and the sample gas),
and its falling time T,

8vbL .
where k=~ "" const, was determmed by two methods. First a can be
"gr
determined by the method of Kuenen and Visser using

m ( -~- ~
PHg- Pc) -o.n=k·- (65 )
PHg , '

where n is the number of individual parts into which the mercury drop
is split. This method should not be applied if the viscosimeter is placed
in a nontransparent jacket.
In the second method a can be determi.ned by working with a single
drop. The viscosimeter is placed at various angles to the horizontal,
u)

so that from (65) we obtain

(66)

From this equation we determine a from the various experimental data

for 0) •
It is implicitly assumed in the second method that a remains constant
with variation in 0). This assumption is apparently justified since both
methods give identical values fora.
To determine a, the mercury and the walls of the viscosimeter must
be very clean, since only under these conditions a has a minimum fairly
constant value during the experiments.
Errors are possible due to incorrect determination of the viscosimeter
dimensions, and the corrections used. The parameters (p, t, ~) may be
inaccurately determined. The authors consider that the overall error
does not exceed 10/0 of the absolute value of the viscosity coefficient.
With this instrument Comings, Mayland, and Egly determined the
viscosity of CO 2 at 40°C in the pressure range 1-137 atm. The authors
take this determination to calibrate the instrument, since the viscosity of

28
 carbon dioxide has been measured with sufficient accuracy by other
scientists. The good agreement between the data and those of Phillips is
considered proof of the reliability of the method. Later, Comings and
his co-worker.s determined the viscosity coefficients of methane (30 to 95°C,
1 to 171 atm), ethylene (30 to 95°C, 1 to 171atm) and propane (30 to
104.5°C, 1 to 41.8atm).
Phillips /28/ employed a similar method at high pressures to find the
viscosity of CO 2 (20 to 40°C, 1 to 120 atm). Nasini and Pastonesi /29/ used
the method for the viscosity of air at 14°C at pressures up to 200 atm.
The Shugaev and Sorokin viscosimeter /30/. The great advantages of
this viscosimeter are its very simple design and its suitability for high
temperatures. A diagram is shown in ::!'igure 10. The seamless tube
of the instrument is of chrome -molybdenum
steel and bent as accurately as possible over a
radius R = 306 mm. The block and flanges are
of stainless steel. There is a hole drilled at the
bottom part of the block, at the center of the tube
arc. An axle with hardened steel knife edges is
keyed in this hole. The knife edges rest on
supports, so that the tube assembly can swing
with very little friction, like the beam of a
balance, over an angle of 120°. Before the tube
is filled with sealing liquid, the· center of gravity
of the tube assembly can be adjusted by weights
attached to the axis.
The internal space of the tube assembly is a
FIGURE 10. The Shugaev and
closed system, with a sealing liquid of a low
Sorokin viscosimeter.
vapor pressure (tin or Wood's alloy) within its
bottom arc. The space above the sealing liquid
is divided into two by a tightly screwed-in
platinum capillary. After the instrument has been centered, filled with
the gas or vapor specimen to be tested, and screwed tight, a pulley with
a thread carrying a weight is set onto its axle. Under the action of the
constant torque created by the weight, the tube assembly will turn over
3. certain angle, The sealing liquid will be displaced inside the arc, thus
rarefying the gas in one half of the tube and compressing it in the other.
The gas driven by the resulting pressure gradient will flow through the
Jlatinum capillary. Once the in strum ent has been carefully centered and
:he angular velocity made constant, the work of gravity of the pulley-
3uspended weight will be equal to the work of internal friction resulting
:'rom the gas flow through the capillary, plus additional losses accounted
:'01' in Hagenbach's correction. In the experiment, the time it took the
assembly to deflect to the right or left over a certain angle, under the
action of the pulley-suspended weight, was measured. The viscosity was
calculated from the Poiseuille equation, with appropriate corrections. To
attain the temperature required for the experiment, the instrument was
placed in a thermostat.
All data, except the time interval of gas flow from the capillary (time
of tube-assembly deflection over a certain angle), were determined from
the geometrical dimensions of the intstrument. The instrument was used
by the authors tG measure the viscosity of water vapor at various pressures
and temperatures.

29
 Later, the same principle was applied by Timrot /31/, who improved the
design of the viscosimeter and used it to determine the viscosity of water
over wide temperature and pressure ranges (up to oOO°C and 300 kg/cm 2,
respectively).
The Michels and Gibson viscosimeter /32/. The glass viscosimeter B
(Figure 11) is set within a high-pressure vessel A, which also contains
mercury; vessel A is placed in a thermostat. The gas is introduced into
the viscosimeter through valve V and tube S; there must be equal pressure
inside and outside the viscosimeter. After the viscosimeter has been
filled to a certain pressure, valve V is closed. The
space around the viscosimeter, i. e., the free space
in high-pressure vessel A, is filled with oil. To
a
compress the gas in the viscosimeter, a hydraulic
pressure pump connected at p increases the oil
pressure above the mercury and causes it to rise in
the glass. When the mercury reaches contact
A 1, the pressure is accurately measured; it is
measured again when the mercury reaches contact 2;
thereafter the gas is rapidly compressed by the
mercury, which almost reaches contact 3. At this
stage the pressure is measured again.
After the rapid compression the mercury column
in tube II of the viscosimeter will be much higher
than in tube I. The increased pressure will drive the
gas through capillary ab; as a result, the mercury
level will fall in II and rise in I.
The time it takes the mercury to fill the volume
between contacts 3 and 4 in I is measured by a stop-
watch. This volume is known. The mercury level
in II is determined from the compressibility of the
FIGURE 11. The gas, since the pressure is accurately measured,
Michels and Gibson
and the volume of each individual part of the viscosi-
viscosimeter.
meter is fixed by calibration. The capillary used by
the authors was 80 cm long and had a diameter of
about 0.1 mm. The viscosity of nitrogen was measured
at 25, 50 and 75°C, in the pressure range of 13.67 to 965.7 atm (kg/cm 2 ).
Later, Gibson used the same instrument to measure the viscosity of hydrogen
at 25°C and pressures up to 300 atm.
The Michels and Gibson viscosimeter is very difficult to make and
operate, since all parts must be calibrated, the compressibility of the
tested gas must be known, many corrections introduced in the calculations,
and the temperature kept constant with great accuracy for a long period.
The last factor is due to considerable heating of the sample gas under rapid and
strong compression when the mercury rises from contact 2 to contact 3,
so that a long time is needed for the gas to revert to the temperature of
the thermostat.
The Rankine viscosimeter for vapors /33/. Saturated vapor may
condense on capillary walls, and this may distort the results of viscosity
measurement. Therefore, prescribed temperature conditions must be
carefully maintained in work with saturated vapors.

30
 Rankine used a vapor viscosimeter of very simple design and
convenient to operate. The liquid of the vapor for viscosity measurements
is placed in the two U tubes, A ane. B (Figure,12). One end of each tube
is sealed, and the other two ends, each with a preheating section, are joined
by capillary k. The U tubes, as well as the preheating sections together
with the capillary, are placed in separate thermostats.

FIGURE 12. The Rankine viscosimeter for vapors.

It may be seen from the figure that the instrument is symmetrical;

consequently measurements may be carried out irrespective of whether the
vapor flows through the capillary from left to right or vice versa. Each U
tube is graduated after accurate volume calibration. In the sealed ends
of the tubes the respective pressures are equal to the saturated vapor
pressures Ps, and Ps, at the temperature of their respective thermostat, plus
c·r minus the pressure difference indicated by the difference in liquid level.
In case, for example, the vapor flows from left to right, the temperature &2
cf the right-hand thermostat must be lower than the saturation temperature
c r the temperature of any other part of the viscosimeter. Under such
conditions the following pressure distribution is created at the ends of the
capillary:

left-hand side PI = P" - gp lh l'

right-hand side P2 = Ps, + gP2h2'
Here, PI and P2 are, respectively, the liquid densities in the left and right
U tubes. Rankine calculated the viscosity from

(67 )

where To is the temperature of the colder thermostat, T the temperature of

the thermostat with the capillary, m the mass of the vaporized liquid, 1: the
time during which mass m flows through the capillary, and Po and Po,
r(,spectively, the vapor densityandpressure, at temperature To (in Rankine ' s
e;,periments T" was O°C).
Formula (67) is not quite accurate because it was derived under the
allsumption that the vapors follow the Boyle -Mariotte law. Also, no

31
corrections for slip, kinetic energy, or pressure drop due to vapor
expansion when heated from r 2 to &, were made. It should be noted
that such corrections would be insignificant /34/ .
The Khalilov viscosimeter /35/. The instrument designed by Khalilov
for measuring the viscosity of saturated vapors is shown in Figure 13. It
consists of a U tube with a 15 em distance between limbs A and B. The ends
of the U are joined by bends CT to the ends of capillary K. Sockets 1 and 2
are used for cleaning the instrument and producing a vacuum. The
instrument is filled with the sample substance, approximately up to line
L-L. The level difference in the U limbs (which drives the vapor through
the capillary) is created by inclining the instrument at an angle of 30 - 35°,
and returning it to its original position after a certain time. The sample
liquid itself functions as a plunger.

FIGURE 13. The Khalilov viscosimet~r for vapors.

The pressure gradient at the ends of the capillary is equal to the

pressure created by the liquid column:

!:::.P =Pl - pz = hpg.

Here p = P, - Pv (PI is the density of the liquid, pv is the density of the

vapor), and g is the acceleration of gravity.
In a special study Khalilov found that the presence of a very small
fraction of noncondensable gases (air) almost completely eliminates the
effect of condensation on !:::.p, and therefore the above equation for the
pressure difference is accurate.
The volume of vapor paSSing through the capillary is numerically equal
to the volume change in the liquid in the limb of the U tube, i. e.,
v = "R 2h, where R is the radius of the U tube. Thus, to determine !:::.p and v we
measure h for a given R. When calculating the viscosity coefficient from
the Hagen - Poiseuille equation, corrections were made for the kinetic
energy.
The following equation was derived by Khali.lov:

(68)

32
Here P = PL -fly, Px is the density of the phase whose viscosity is to be
determined, il, and I, are the distances between the menisci and the equi-
librium line LL at the initial and final instants of observation. The other
symbols in (68) have the same designations as above.
Khalilov found the viscosity of saturated vapors of mercury, methanol,
ethanol, propanol, n-hexane, and n-heptane. The author considers that
his viscosimeter is also suitable for measuring the viscosity of superheated
vapors. Here, however, another liquid must be used as the sealing sub-
stance, and its meniscus is observed. It is evident that it is best to use
a sealing substance which forms a vapor with a negligible pressure at the
experimental temperature.
Later Khalilov /36/ designed two types of similar viscosimeters to
measure the viscosity of superheated vapors at high pressures. In one of
these the mobon of the sealing liquid was observed by the contact method,
and in the other visually.
The viscoelimeters consisted of a stainless steel rectilinear tubular
frame with removable capillaries; mercury is recommended for the sealing
liquid. The difference in the levels of the sealing liquid is created by
inclining the viscosimeter to an angle of 25 - 30°. After a certain time
interval has elapsed, the instrument is returned to its original (fixed)
position. From the signals triggered off when the liquid passes the contact
(contact viscosimeter) or the mark (visual viscosimeter), the time taken
for a given volume of the sample substance (gas or liquid) to flow through
the capillary is measured.

2. The Golubey and Petrov viscosimeters /37,73,122/

First design /73/. The setup in which the viscosimeter of the first design
is used is shown in Figure 14. The gas sample is introduced from
cylinder 4, by compressor 5, into high-pressure vessels 6. Then the gas
proceeds via filter 7 through control valve 1 into heat-exchanger coil 8
made of stainless-steel tubing (inside diameter 1.5 mm, outside diameter
5 mm). In the heat exchanger the gas is heated to the temperature of the
thermostat; it then enters test tube 10, in which it flows through a glass
capillary. The ends of the test tube are connected to a differential pressure
gage 11. At the inlet to the test tube an Fe -constantan thermocouple 12 is
set in a textoli.te packing.
During filling of gas into the system, or during pressure release,
valve 3 must be open, while during the experiment it must be tightly closed.
After the gas has emerged from the test tube, the pressure is reduced by
valve 2 to atmospheric. Then the volume of the gas is measured (at high
flow rates, the volume is measured with gas meter 14, and at low rates
with gas burette 15). The gas volume used at test temperature and pressure
is calculated from the gas volume flowing through the capillary during a
certain time interval, and measured at atmospheric pressure. For the
calculation pvt data are used, which must be known for the gas studied.
Before the gas enters the meter or the burette it must pass through
water-flow cooler 13, which cools ,it to room temperature.

33
 FIGURE 14. Setup for determining gas viscosity at high pressures
(first design):

1) valve at test-tube inlet; 2) valve at test-tube outlet; 3) inter-

connecting valve; 4) gas cylinder; 5) gas compressor; 6) high-
pressure vessel; 7) filter; 8) hea t-excha.nger COil; 9) thermo-
stat; 10) test tube; ll) differential pressure; 12) thermocouple;
13) water cooler; 14) gas meter; 15) glass measuring burette.

At the time of reading the gas is uniformly fed into the test tube from
vessels 6 through valve 1, so that the pressure of an equal amount of gas is
reduced by valve 2. The pressure and the pressure drop in the capillary
are kept constant.
Thus, the gas volume that flows through the capillary in unit time under
test conditions at a certain pressure gradient is determined. The
viscosity coefficient at test temperature and pressure is calculated from
these values by the Poiseuille equation, using the known geometrical
dimensions of the capillary with the conventional corrections',
The experimental device (Figure 15) is a cylinder made of an approxi-
mately 1-m-long steel tube, 11 mm inside diameter and 25 mm outside
diameter. One end of the tube is connected to four-way piece 6, and the
other to three-way piece 7. Both pieces ha.ve drilled apertures for the
inlet and outlet of the gas, connection to the differential pressure and
insertion of the thermocouple,
Glass capillary 1 placed inside the device is held at one end by the
stuffing box and kept in position by centering rings 2, The packing of
stuffing box 3 is leather or thin asbestos cord, impregnated with bakelite
varnish. It is clamped between two steel collars ,4 by screw socket 5.
The differential pressure gage (Figure 16) consists of a small high-
pressure vessell with a screw socket 2. This is connected to a three-way
piece 3, through which gas is supplied, and an electric lead is inserted by
means of cartridge 4 with textolite packing 5,
At the bottom o(the screw socket there is a stuffing box consisting of two
metal collars 7, packing 8 (leather or asbestos), and screw socket 9, The
stuffing box tightly clamps glass tube 10. This tube is widened at its upper
part to prevent overflow of mercury from the gage; a glass tube 11 with
apertures 12 is connected to the widened part.

34
 Gas
I For cartridge
To differential with thermo-
pressure ga ge-=M~;:::::}F;[JcouP Ie
- - 6

14
2
. ____ I
16
14 t'07791\\-* 15
12

II f1

FIGURE It,. Experimental de- FIGURE 16. Differential pressure

vice: gage:

1) glass c,.pillary; 2) centering 1) high-pressure vessel; 2) screw

rings; 3) ,;tuffing box; 4) steel socket; 3) steel three-way
collar; 5) screw socket; 6) steel piece; 4) packing cartridge;
four-way piece; 7) steel three- 5) textolite packing; 6) electric
way piece. lead; 7) steel collar; 8) stuffing
box packing; 9) screw socket;
10) glass tube; 11) wide glass
tube; 12) apertures; 13) thick
silver wire; 14) thick platinum
wire; 15) thin platinum wire;
16, 17) glass tubes.

The electric lead 6 consists of three sections: a thick silver wire 13

(introduced through cartridge 4), a small section of thick platinum wire 14,
c.nd a measuring section made of thin platinum wire 15. The electric lead
is inserted into glass tubes 16 and 17 so that it is possible to fix the thin
platinum wire 1.5, protect the silver wire against mercury, and readily
remove the lead from the gage, together with the 3-way piece. The ends
c,f the glass tube 17 have apertures 12.
The lower end of the electric lead is immersed in mercury. This acts
a.s a mobile contact, and is earthed to the body of the high-pressure vessel.
The electric circuit includes the following, connected in series: storage
batteries; resistor box; electric lead; earth. By means of a parallelly
connected galvanometer, the voltage drop in the circuit (electric lead-
earth) is measured. During the experiment the pressure between the ends
of the capillary raises the mercury level in the tube 10. The mercury
Eubmerges the thin platinum wire 15, and partly reduces the resistance,

35
 so that the voltage in the electric section decreases. Any pressure
gradient produces a corresponding decrease in voltage, which is recorded
by the galvanometer. The differential pressure is calibrated by a water
column at atmospheric pressure. The conversior: of the results to hi.gh
pressure (up to 1000 atm) does not lead to errors, since the changes in
mercury voLlme and electric resistance are very small in this pressure
range. In our experiments the measuring sections were 50 mm long
in one case, and 100mm in another.
The voltage drop over the length of these sections was recorded by a
galvanometer (with a scale of 250 divisions). Consequently, when the length
of the measuring section was 50mm, each
scale division corresponded to 0.2 mm Hg,
and when it was 100mm, to 0.4mm Hg.
Second design /122/. A setup in which the
differential manometer and the capillary are
connected is shown in Figure 17. The gas
sample is introduced from the high-pressure
vessel (into which it is supplied by a
compressor) via valve 1 into heat-exchanger
coil 6. Here it is heated to the thermostat
temperature. High-pressure vessel 4
together with screw socket 5 and heat
exchanger 6 are placed in an oil thermostat.
While the system is being filled with gas,
intermediate valve 2 must be open for the
gas to enter the viscosimeter by both inlets 7.
Under these conditions the mercury in the
viscosimeter will be under the same pressure
FIGURE 17. Setup for determining from both capillary ends during gas filling
gas viscosity at high pressures (sec- or discharging.
ond design): During the test the gas is reduced by valve
1) inlet valve; 2) intermediate 3 to atmospheric pressure (intermediate
valve; 3) outlet valve; 4) high- valve 2 must be hermetically closed).
pressure vessel; 5) screw socket of Simultaneously with the reduction, gas is
vessel 4; 6) heat-exchanger coil; fed through valve 1 in such an amount that
7) high -pressure inlets; 8) measur-
the pressure in the viscosimeter remains
ing burette; 9) glass capillary;
10) glass tube with widened portion
constant and the pressure drop is constant
11 and bulb 12; 13) outlet from bulb during reading. The gas volume flowing
12; 14) filter; 15,16) electric leads; through the capillary during the test is meas-
17) apertures in glass tube 10; ured by burette 8 after reduction, and then
18) thick-walled glass tube. converted to the test conditions.
At constant pressure and uniform gas
outflow through valve 3 (which also means
uniform gas flow through the capillary) the pressure drop in the viscosi-
meter becomes constant. It is recorded by a galvanometer measuring the
voltage on the measuring section of the platinum wire.
The glass viscosimeter consists of capillary 9 and tube 10. The latter
has a widened portion 11 at its top and bulb 12 at its bottom. The bulb
communicates with the surrounding medium through aperture 13. The
ends of the capillary are fitted with filters 14 made of loose asbestos
fibers which protect the capillary against contamination. Tube 10 has two

36
 electric leads, 15 and 16, of platinum wire (0.5mm diameter). Along
the axis of the tube a thin platinum wire (d = 0.1 mm) is stretched with a
platinum spring at its bottom. This wire serves as the measuring section.
The top of the viscosimeter (thick-walled tube 18) is hermetically
connected to screw socket 5 of pressure vessel 4 by means of an asbestos-
cord packing soaked with water glass and talcum (at low temperatures
a rubber tube may be used instead.)
Variations in test temperature change the absolute magnitude of the
resistance of the platinum -wire measuring section, thus displacing the
zero point of the galvanometer scale. For this reason, whenever the test
temperature i.s changed, the galvanometer arrow must be adjusted to the
earlier selected zero point. This is easily achieved by adjustiilg the series
connected resistor (see the layout of the first design).
The galvanometer scale is calibrated for measuring the pressure drop
in the viscosimeter by means of a water column at atmospheric pressure.
To do this, a pressure gage (a U -tube
filled with water) is connected in parallel
to the viscosimeter placed in the high-
pressure vessel. Gas is passed through
the viscosimeter.
Third design /122/. Some difficulties
occur in experiments with setups of the
first and second designs. It is necessary
3 to calculate the gas volume under the
experimental conditions from the volume
measured at atmospheric pressure, since
pvt data of the gas are not always available
for the required temperature and pressure
ranges. It is also difficult to keep the
pressure and pressure drop on the
capillary constant during the whole experi-
ment.
Figure 18 shows a setup with which it
is not necessary to measure the gas volume
flowing through the capillary and the
pressure drop on the capillary.
The setup is placed in an oil thermostat.
It consists of high-pressure vessel 4 con-
taining the glass viscosimeter. The screw
FIGURE 18. Setup for determining gas socket 5 is fitted with two gas tubes, and
viscosity at high pre:;sures (third design): electric lead 6 also passes through. The
setup is filled through valve 1 while inter-
1) inlet valve; 2) intermediate valve;
3) outlet valve; 4) high-pressure vessel;
mediate valve 2 is open. During the
5) screw socket; 6) electric leads; experiment valves 1 and 2 are closed,
7) glass capillary; 8) filter; 9) packing and valve 3 is open. The amount of gas
cartridge; VI and V,- glass bulbs; is thus reduced so that a certain difference
ro, fI' f2t lSI f4 - platinum contacts with exists between the pressure in the space
electric resistors. surrounding the glass bulbs V! and V ~
and inside the bulbs. This is easily
achieved, sinee the pressure becomes rather slowly equalized over the
capillary.

37
 The gas is fed until the mercury fills the bulbs V 1 and V 2 and reaches
contact r 4 , Then, valve 3 is closed (the experimental pressure is measured
only after the valve is closed).
The pressure drop between the ends of the capillary forces the gas to
flow through the capillary and the mercury sinks to the bottom. The time
taken for the mercury to flow from r3 to r 2 (from V 2 ) and from r 2 to r l
(from V I) is measured with a stopwatch. Two bulbs are used to increase
the reliability of the experiment and to control the flow conditions through
the capillary.
Since the dimensions of the capillary, the volumes of the glass bulbs
VI and V 2 corresponding to the gas volumes flowing through the capillary at
the test pressure and temperature, and the pressure drop, are previously
determined and are constant, we can calculate the viscosity coefficient from
the outflow time.
Contacts r l , r 2 , r3 and r 4 with the resistors connected in series help
to determine the mercury level by means of the galvanometer. Filters 8
made of loose asbestos fibers protect
To galvano-
From high- the capillary against contamination.
meter pressure stor- The glass viscosimeter is connected to
r=~~~~~~~~~;a~ge~tank the screw socket of the high-pressure
vessel by a rubber tube at low tem-
peratures (up to 100°C), and by a thin
To measuring
burette tin cylinder at high ones. The cylinder
is wound at the top by an asbestos cord
impregnated with water glass and tal-
cum. A small platinum cylinder can
be used for this purpose, with one end
fused to glass and the other to the
steel outlet of the screw socket,
Fourth design /37/, To determ ine
the physical properties of gases it is
most convenient to have an apparatus
with which it is possible to measure
simultaneously the compressibility and
viscosity of gases and gas mixtures at
high pressures. A diagram of such
an apparatus is shown in Figure 19.
The apparatus consists of a high-
FIGURE 19. Setup for determining gas pressure vessel 4 containing a glass
viscosity at high pressures (fourth viscosimeter (piezometer-viscosimeter)
design): with screw socket 5.
1) inlet valve; 2) intermedia te valve;
The screw socket is fitted with two
3) outlet valve; 4) high-pressure vessel; inlets: one is joined via the viscosi-
5) screw socket; 6) electric leads; 7) pack- meter capillary to bulbs VI and V 2 , and
ing cartridge; 8) textolite cone; 9) glass the other to the outside space surround-
capillary; 10) glass tube; 11) filter; ing the viscosimeter. The electric lead
12) heat exchangers; 13) aperture. 6 passes through the second inlet. It
is held in position by a textolite cone 8.
Heat exchangers are placed in front of the inlets. They are made of high-
pressure tubing with an internal diameter of 5 mm.

38
 The apparatus is filled with the sample gas through valve 1 while the
intermediate valve 2 is open. After the required pressure is reached,
valves 1 and 2 are closed. Then, the gas is allowed to pass from the glass
bulbs V I and V 2 through valve 3 into the atmosphere. The pressure in the
bulbs is thus decreased, and the mercury rises in bulbs V I and V 2 to the
platinum contacts which record the mercury level on the galvanometer. We
begin to measure the volume of gas displaced from the glass
bulbs when the mercury reaches contact r 2 • With the aid of
intermediate valve 2 the gas is trapped in the "fork" at r 2 •
The complete filling of V I is indicated by contact r 3' and
of V 2 by two contacts r 4 and rs (to increase the accuracy).
The volume of gas displaced from the bulbs is measured
in a dry measuring burette.
This is the final step in the determination of the
compressibility coefficient at Pi (initial pressure in the
experiment) and at the temperature of the thermostat in which
the apparatus is immersed.
The mercury is then raised to contact r 6 , and valve 3 is
closed. The viscosity coefficient is determined at a certain
final pressure Pf of the experiment, and the pressure drop
Pi - Pf corresponds to the pressure decrease in the apparatus
when the gas is displaced from r 2 and r 6 • It should be noted
that in this case the volume of displaced gas is 1/60 of the
total volume of the high-pressure vessel. The intermediate
FIGURE 20. valve 2 is opened so that a closed circuit is formed in which
Glass piezo- the difference in the mercury levels in the viscosimeter
meter-visco-
represents the pressure drop along the capillary. The
simeter (fourth
design).
mercury level in the bulbs falls under the action of gravity,
so that the gas flows through the capillary.
The time of outflow of the mercury from bulbs V 2 and VI
(after r 4 , r 3, and r 2 are connected) is measured, and from
the result the viscosity coefficient is calculated. All the other parameters
used in the Poi.seuille equation (volume of gas flowing through the capillary,
pressure drop in the capillary) are previously determined and are constant.
Thus, the compressibility coefficient of the gas is determined when the
mercury rises in bulbs VI and V 2 , and the viscosity coefficient when the
mercury falls in these bulbs.
The glass piezometer-viscosimeter is separately shown in Figure 20.
To obtain the required temperature, vessel 4 is placed in an oil thermo-
stat together with screw socket 5 and heat exchangers 12. For temperatures
above 150°C, the whole apparatus is placed in an electric furnace. The
temperature is measured by three mercury thermometers inserted into
apertures in the walls of vessel 4 (the apertures are bored to different
depths along the body of the vessel).
The apparatus of the fourth design was used to determine the viscosity,
and in some cases also the compressibility, of gases and gas mixtures over
wide temperature and pressure ranges.
Fifth design /37/. The main feature of this design is the glass viscosi-
meter (Figure 21) placed into a high-pressure vessel. When the viscosi-
meter is inclined at an angle of 90° i. e., in the horizontal position, mercury
flows through tube 6. But when the viscosimeter is again placed in the

39
vertical position, the mercury remains distributed over it as shown on
the right-hand side of the figure.
To create a difference in the Hg levels (pressure drop at the capillary
ends), the viscosimeter must first be placed i.n the horizontal position by
tilting it through 90 and again rapidly returned to the vertical position.
0
,

_J~ 7 6 9

FIGURE 21. Successive positions of the viscosimeter for creating a differ-

ence in the mercury levels:

1,2,3,4,5) platinum contacts; 6) glass tube of variable cross section;

7) bottom glass cylinder; 8) capillary; 9) filter; v, - measuring space;
V2 - additional space; fl. [2' I3. [4 - electric resistors.

The left-hand side of the figure shows the different cross sections at
different heights of the viscosimeter, and also the main dimensions of one
such viscosimeter. This viscosimeter is very convenient for measuring;
it accelerates the experiment, and considerably simplifies the whole lay-
out, since the number of valves and high-pressure tubing is conSiderably
less than in the types described above. It is suitable for determining
the viscosity of gases and liquids.
The setup of a viscosimeter of the fifth design is shown in Figure 22. It
includes a hydraulic press 3, two intermediate vessels 4,5 (to generate the
pressure), and a high-pressure vessel 6, assembled together with the
electric furnace. The viscosimeter is placed into vessel 6, which can be
rotated around its horizontal axis 9 through 90 0 •
It should be noted that the viscosimeter in the figure is not shown in
the working position, but for clarity rotated around its longitudinal axis
until it is perpendicular to the plane of the diagram.
The viscosimeter and vessel 4 are filled with the gas sample from
cylinder 8 through valve 1. The gas is compressed by the hydraulic press
until it is under the required pressure. By means of the electric furnace 7,
the test temperature is adjusted, and checked by mercury thermometers.
The thermometers are inserted to various depths into apertures bored along
the walls of the high-pressure vessel.

40
 When the viscosimeter is filled, or when the gas is discharged, the
instrument must be in the horizontal position so that the gas can freely
enter the tube during filling or leave it during discharge (see Figure 21).
When the viscosimeter is in the vertical position, the mercury in tube 6
will reach above contact 2. It will then fall under gravity through tube 6
into cylinder 7 and force the gas to flow through capillary 8. The mercury
level in the tube is indicated on the galvanometer by means of platinum
contacts 1,2, 3,4 and 5 connected into the circuit through resistors r 1, r 2 ,
r 3 , and r 4 .
Contact 1 fixes the extreme upper position; the mercury must not rise
above this contact. Contact 2 serves for preliminary adjustment, and
contacts 3 and .<[ are control devices.
The period during which the mercury falls from contact 3 to contact 4
(in other words the time taken for gas volume v1 to flow through the
capillary), and the other parameters required to calculate the viscosity
coefficient are experimentally determined. On the other hand, the
dimensions of the capillary, volume v 1, and the pressure drop over the
capillary are known beforehand, and remain constant. The parameters
used by the authors varied over the following ranges in all the viscosimeters:
pressure grad.ient, 5O'-2O'O'mm Hg; time of outflow, 50' sec or more;
length of capillary, 6O'-3O'O'mm; capillary diameter, O'.I-O'.2mm; volume
of the bulb, I-lO'cm 3 •
In all the designs of the viscosimeters used the mercury temperature
did not exceed 25O' o C. In exceptional cases only, such as in the determina-
tion of the viscosity of methanol, isopropanol, isobutanol and acetone, the
temperature of the mercury was 260' to 27O'oC, since the test temperature
had to be higher than the critical temperature of these compounds. At
higher temperatures the vapor pressure of mercury begins to increase
greatly and its absolute magnitude is large. This may influence the
accuracy of the measurements. Attempts to replace mercury by fused
metals (Wood's alloy, tin) with a vapor pressure which is low at tempera-
tures above 2 !5· Q"C , were unsuccessful, since these fused substances adhered
to the glass (or quartz) walls of the viscosimeters.
Generation and measurement of the pressure. To generate the gas or
liquid pressure we used cylinders filled with the compressed sample gas,
a high-pressure compressor, and a hydraulic press with intermediate
vessels.
For the majority of our measurements we used a hydraulic press. Two
communicating high-pressure vessels (Figure 22) are half-filled with
mercury. Vessel 4 has two contacts at the highest and lowest positions
of the mercury. These positions are fixed by making and breaking the
contacts. The space above the mercury in vessel 5 is filled with oil, and
connected to the hydraulic press. First, the gas is transferred from
cylinder 8 to vessel 4 until the lower contact is not broken. Then, the
outlet valve is closed and the gas is compressed by the mercury (feeding
the oil by the press). If the required pressure is not reached at the highest
position, the operation is repeated several times. If a liquid is used as
the sample instead of a gas, it is much easier to generate pressure.

41
 FIGURE 22. Setup for determining gas viscosity at high pres-
sures (fifth design);

1,2) high-pressure valves; 3) hydraulic press; 4,5) pressure-generating

vessels; 6) high-pressure vessel; 7) electric furnace; 8) gas cylinder;
9) rotational axis of viscosimeter.

The mercury seal protects the sample against contamination by oil.

Nevertheless, the mercury is gradually contaminated by the pressure-trans-
mitting liquid (oil), and must be replaced from time to time. Mercury has
an unfavorable effect on the metal surface; it forms an amalgam and thus
rapidly corrodes it.

TABLE 2. Viscosity (gcm -1 sec -1) of some oils (after Volarovieh)

Pressure, 1(g/ em 2
Oil
T,oC 1 3'-,) 700 1000

Aviation •••••. 22 14.5 -- 56.0 -

32 6.5 11,9 21.3 40.3
I
I 60 1.75 2,92 - , -
Avtol (automo-
bile) ••••••• 25 10,4 17.0 - 40.0
II
, 40 1.5 2.22 3.9- 5.1
Machine (export). 20
I 2.74 -- - -
i
33 I 1.1 2.27 - 8.1
40 0.75
i 1.70 2.78 -
Cylmder •••••• ) 16 26.5
I -- - I 13 1. .0

When the intermediate vessels are used, we must also consisier the
viscous properties of the pressure -transmitting liquid. With increase in
pressure some oils become so viscous that the ball valves of the press
cease to operate and the pressure is transmitted very slowly over the tubes

42
 of small diameter. Therefore, we should know the viscosities of the oil
used in the hydraulic press. The viscosity of some oils at various tempera-
tures and pressures is shown in Table 2.
When working with the compressor we should remember that only gases
or gas mixtures which do not liquefy on compression can be processed.
Since the cylinders of the compressor are lubricated by oil, measures must
be taken to prevent entrainment of the oil
by the compressed gas into the apparatus.
Oil separators are installed at the last
stage of the compressor to separate oil
and suspended matter from the gas, and
2 filters, adsorbers, and dryers to free
the gas from oil and water vapors. Water
J------t::- enters the high-pressure system mainly
from the gas cylinder.
The pressure was measured by spring
4-----~
pressure gages, but only standard gages
and gages for accurate measurements
with fine scale divisions. All the gages
were calibrated at the Moscow State
6---~~ Institute for Standards and Measuring
Instruments (Moskovskii Gosudarstvennio
? Institut Mer i Izmeritel' nykh Friborov)
before and after the determination.
8 The difficulties involved in measuring
at high pressure are also due to the
complicated design of the viscosimeter
and the other individual parts of the
setup. A well-designed setup improves
9---~
the reliability of the apparatus, the
tightness of the joints, and hence the
accuracy of the re sults.
Figures 23, 24, and 25 show component
parts frequently encountered in high-
pressure setups. They are very suitable
for operation under such conditions.
High -pressure apparatus are made of
FIGURE 23. Control valve:
high -quality steel. The brands and the
:l) handle: 2) pin: 3) cap nut: 4) valve corresponding State Standard (GOST) of
,.tem with cone end; 5) clamping socket;
the most widely employed steels, and
I,,, stuffing box: 7) steel ring; 8) body; some of the properties of these steels,
D) nipple.
are summarized in Table 3. It appears
that the most suitable brands are stain-
less steels Ya1 and Ya1T, but it should
be remembered that their thermal conductivity is only 1/10 -1/5 of that
(If other steels. Thus, in apparatus made of Yal and of YalT steels, the
E:teady thermal state is established very slowly.

43
 ~ Pressure gage ::::>
5

a b c

FIGURE 24. High -pressure joint:

a - connection by cone: 1) cone, 2) cap nut, 3) nipple: b - con-

nection by lenticular piece: 1) nipple, 2) connecting nut,
3) lenticular piece; c - connection of pressure gage: 1) cone,
2) cap nut. 3) connecting nipple. 4) packing. 5) nipple of
pressure gage.

TABLE 3. Steels for high-pressure apparatus

No. OST I Tensile strength. Yield point. Relative

Brand ! Designa tion
or GOST kg/mm 2 kg/mm2 elongation, 0/0
i
St.5 380-41 I E 16 50- 62 27 17-15
18KhNVA 4543-48 , KhNVTs 120 100 i 10
18KhNMA 4543-48 - 90
! lO
120
I
ZOKhMA 4543-48 ZOKhNZM 90 -130 80-105 I
5-12
OKhNZM i 4543-48
i
I EI179 100-160 84-138 13-10
I I
1Kh18N9 I
I
5543- 50 I Ya 1 55 25 40
1Kh18N9T 5543- 50 YA 1 T 55 25 40
2Kh13
ShKh12
I 5632- 51
801- 47
Zh2. EZh2
EShKh12
85 65 10

Determination of the geometric dimensions of the viscOSimeter. The

selection of the dimensions of the viscosimeter. especially the accuracy of
the determination of the capillary radius, is very important in the measure-
ment of the viscosity by this method, since the viscosity equation has a term
with the capillary radius to the fourth power.

44
 The mean capillary radius was calcu-
lated from the weight of mercury filling
2
the capillary. Since the diameters and
3 lengths of the capillaries are rather
small, in our experiment we had to fill
the capillary many times with Hg and to
5 measure the length of the mercury column.
This operation should be carried out
with a capillary somewhat longer than that
required for the viscosimeter. However,
care must be taken that the mercury
thread fill the part of the capillary which
7 will be used.
In the calibration we corrected for
the temperature at which the length
of the mercury thread is measured, and
for the volume of the mercury meniscus.
Of course, the capillary and the
8 mercury used were carefully cleaned
and dried.
A capillary tube must be selected
with a cross section equal over the
a
whole length. The shape of the
cross section is determined at both
FIGURE 25. Electric leads:
ends of the capillary with a microscope.
a -1) electrode; 2) clamping nipple; It is desirable that the capillary should
3j.body of packing cartridge; 4) ring;
have a circular cross section, but in
5) textolite (teflon) insulation; b-l) el-
ectrode; 2) cap nut; 3) bushing; 4) ring;
fact in the majority of cases the shape
5) textolite (teflon) insulation; 6) body of is conical or elliptic. In this case
packing cartridge; 7) capillary; 8) cone. the Poiseuille equation must be
corrected.
Thus, if the capillary is conical with
a circular cross section of radius r, at one end and T2 at the other, the
term r 4 in the Poiseuille equation is calculated from

(69)

Knibbs /18/ expresses the conicity by

(70)

In this case if the conicity changes uniformly along the capillary

(2r=r 1 +r2), the Poiseuille equation becomes

_"rldp
' f j - Svh
(1 _!212+~14_
3 9
)
•... (71 )

45
 If the capillary is elliptic in cross section, with semiaxes a and b, then

2a3b'
r4-
- a2 +b 2 •
(72)

Knibbs determined the ellipticity from

(73)

and '1) from

r4 I1 p (I_4 2+7 4 __
_7t
'1)- Svh e e ... ). (74)

If a conical capillary has an elliptic cross section, then it is obtained

from

(75)

where r 1 and r 2 are the mean radiuses of the end cross sections, and
l = aa-b
+ b' where a and b are the arithmetic values of the semiaxes of the
elliptic cross sections at the capillary ends.
Knibbs gives the following equation in this case:

(76)

For very exact measurements the effect of the ellipticity must be

taken into account also when determining the term in the Poiseuille equation
which depends on the kinetic energy:

mpv ab
87th' 1 • (77)
"2 (a 2 +b2 )
In designing the viscosimeter, it is advisable when calculating its
dimensions to take into consideration the range over which the viscosity
of the sample changes. The following points must be observed: 1) the
dimensions of the viscosimeter must be suitable for the internal dimensions
of the high-pressure vessel into which the viscosimeter is placed; 2) the
time of outflow of the mercury must not be too long, since the experiment
would take an excessive time, but not too short or otherwise errors of
measurement cannot be avoided; 3) over the whole range studied the flow
must be laminar.
We have already mentioned that laminar c:onditions prevail in the
c:apillary if the Reynolds number is less than 2300. We have

Re = wdp,
'1

where w is the linear velodty in em/sec:; p is the gas density in g/cm 3 ;

d is the capillary diameter in cm; and '1) is theyiscosity of the gas in

46
 4v
For the given viscosimeter we can write W= rcd 2-r,' where v
is the gas volume flowing through the capillary in time ~ (equal to the
constant volume of the reservoir of the viscosimeter. We then have

Re = 4vplO7 (78)
1td~t

By introducing the known or assumed values for p. "/.~, into (78), we obtain
the dimensions d and v for the given case. With this formula we can
determine also the flow conditions for any experiment if the viscosimeter
dimensions are known.

3. Method of rnE~asuring the fall of a weight in the sample gas

In this method the gas moving along a falling cylindrical weight or a

rolling sphere must pass through the annular clearance formed between
the falling or :rolling weight and the inside wall of the tube.
The resista.nce against the fall of a cylindrical weight consists of three
components: 1) resistance due to the shape of the jet and the ends of the
weight; 2) resistance corresponding to the outflow of liquid or gas through
the annular clearance and the falling rate of the weight determined by the
cross-sectional area of this clearance; and 3) resistance caused by
viscous braking because of the relative motion of two cylindrical walls. If
an elongated falling cylinder is used, resistance 1) in the narrow annular
clearance is very small and can be neglected.
Lawaczek /38/ proposed the following equation for calculating the
viscosity coeffieient:

(79 )

where p, and P2 are the densities of the falling weight and liquid or gas;
a and b are the radiuses of the tube and the falling cylinder, and ~ is the
time the cylinder falls through distance s.
Since it is very difficult to measure (a-b) with sufficient accuracy, and
since this value varies in different parts of the tube, the instrument is
calibrated with a gas or liquid of known viscosity. The instrument constant

is determined when resistance 1) is also taken into account.

Below we shall describe an instrument based on this principle.
The Stackelbeck viscosimeter /39/. The instrument consists of a
horizontal, accurately calibrated glass tube of about 12 mm diameter
(Figure 26). It is filled with the sample gas. A cylindrical weight falls
through the tube under gravity. The diameter of the falling weight is
smaller than the diameter of the glass tube, so that a narrow annular

47
 clearance is formed through which the
Valve displaced gas flows to the top. It is
Gas inlet evident that with increase in the gas
viscosity the falling rate of the weight
decreases.
If we assume that the flow in the
annular clearance slit is laminar, we
can determine the absolute viscosity
by measuring the time a weight falls
Solenoid
through an accurately determined
distance. The glass tube is placed
into a steel cylinder, which is closed
at the bottom. The inner diameter of
the tube is determined with an accuracy
'I
of up to 0.01 mm. Since the upper end
:! Falling weight
of the tube is open, the pressure in the
tube is equal to that in the surrounding
medium. The falling rate of the weight
is determined by electric contacts
to a rapidly recording ammeter. A
solenoid is installed in the upper part
which attracts the falling weight. When
the current is interrupted in the
solenoid, the weight falls, and at the
Contact end of the tube it touches the contact.
The falling rate is measured with a
stopwatch. The viscosimeter is filled
FIGURE 26. The Stackelbeck viscosimeter. with the gas sample through a valve
placed at the top of the apparatus. The
pressure is measured with a spring
pressure gage. The steel cylinder is mounted on a movable device so that
it can be turned through 180 to return the weight to the solenoid. The
0

viscosimeter is placed into a thermostat held at the test temperature.

Stackelbeck uscd the instrument to determine the viscosities of CO 2,
NH 3 , CH 3Cl and S02 at pressures and temperatures cha.racterizing the
liquid and gaseous states of these substances.
It should be noted that the falling weight method is reliable for gases
at low falling rates only. In this case the losses due to friction between
the centering projections of the falling weight and the tube walls will be
relatively small (these losses cannot be avoided and may differ in each
experiment). The clearance between the fal1ing weight and the wall should
be very small, which in practice is difficult to realize. For liquids, on
the other hand, which have a much higher viscosity than gases, the method
is successfully employed since the falling rate will be low even if the clear-
ance between the falling weight and the tube wall is large. Moreover, the
falling weight with its centering projections slips along the liquid film only,
which decreases the friction, or at least makes it fairly constant. For
this reason a similar method employed by Bridgman /40/ to determine
the pressure-dependent viscosity of liquids gave sufficiently reliable results
after the introduction of appropriate corrections.

48
 On the other hand, for gases the mechanical friction of the falling weight
will depend to a large extent on the overall drag. The drag cannot be
accurately estimated, and therefore the falling weight method cannot give
accurate results. This is shown, for example, by the discrepancy between
Stackelbeck's data for gaseous ammonia and the results obtained by the
capillary method. A comparison will be given later.
The Sage and Lacey viscosimeter /41/. These authors determined the
viscosity of methane, propane, n-butane and isobutane at various tempera-
tures and pressures by the rolling ball method. This method was earlier
used by Flowers /42/ and Hersey /43/ to determine the viscosity of liquids.
The principle of the method is the determination of the time required for
ball B (Figure 27) to roll under gravity through a certain, accurately fixed
distance along an inclined cylindrical tube A, with a diameter close to that
of the ball.
No rigid mathematical theory for such motion has so far been elaborated.
The equation used to calculate the viscosity coefficient is empirical, and
does not completely describe the hydro-
dynamics of gas flow in the clearance
between the rolling ball and the tube.
The authors who used this method
consider that there isa linear relation-
ship between the viscosity and (PI-P2)"
where PI and P2 are the densities of
the ball and gas, respectively.
In the Sage and Lacey apparatus
(Figure 27) a special device makes it
possible to turn the viscosimeter
through any given constant angle of
inclination to the horizontal to move
the ball between the contacts. In the
experiments of these authors the angle
FIGURE 27. The Sage and Lacey viscosimeter: of inclination was 15°, and the diameters
E - high-pressure vessel; al' az and as - valvesj of the tube A and ball B were 12.705
D- contacts fixing the initial and final positions and 12.59mm, respectively. For
of the ball B in viscosimeter A: T - thermostat. experiments on gases the clearance
between the ball and the tube wall must
be very small and uniform. This
requires very careful treatment of ball and tube surfaces. In calibration
and in the actual experiments the gas flow around the ball must be laminar,
since only then '>1 is a linear function of (PI - pz) 7. In their calculations the
authors used the magnitude P2/'t'>1 proportional to or identical with Re, and found
that when p/''>1 > 2.5 (p lin g!cm 3 , 't in sec, and "f) in micropoise), the gas
flow becomes turbulent.
Since with increase in pressure the gas density increases at an appreci-
ably higher rate than the viscosity, the parameter P2/'t'>1 rapidly increases
with pressure. Therefore, research workers using this viscosimeter to
determine the viscosity of compressed gases could not prevent the for-
mation of eddies at most of the pressures applied. For this reason, the
authors calibrated their instrument for the turbulent range also. They
measured the Lme of roll of a ball in a gas with known viscosity at
pressures corresponding to turbulent conditions in the instrument. Sage

49
and Lacey worked with carbon dioxide, and used the viscosity data of
Phillips. The results were plotted on a graph representing the depen-
dence of the ratio of the actual viscosity 'fJ to the apparent viscosity <¥ on P2/q.
The apparent viscosity <¥ was determined from the calibration curve plotted
under laminar conditions, when <¥ is a linear function of 't(Pl-P2).
Comings, Mayland, and Egly /27/ showed that such a calibration method
cannot be very accurate. But the attempts of these authors /27/ to obtain
a more accurate correction for the effect of eddy formation in the experi-
ments of Sage and Lacey failed.
A similar instrument was later used by Smith and Brown /44/ to
determine the viscosity of ethane and propane and by Bicher and Katz /45/
for the viscosity of methane, propane, and their mixtures over wide
temperature and pressure ranges. These authors like Sage and Lacey
carried out the majority of their experiments in the turbulent range, since
the values of P2/'t'fJ were appreciably higher than 2.5.
Some of the data obtained by the rolling ball method will be reported and
discussed later. No sufficiently accurate viscosity coefficients could be
found by this method. This is shown, firstly, by the considerable
scattering of the experimental points of every author, secondly, by the
large discrepancies between the results of individual authors, and thirdly,
by the large difference between the results of this method and those of
the capillary method, at present the most accurate and reliable.

4. Method of damping the torsional oscillations of a

disk suspended in the gas sample

Pendulum viscosimeters, in which the damping of the torsional oscillations

of a disk are observed, have been used by numerous research workers
/46 - 54/ to determine the viscosity of gases and their mixtures. If the
disk is immersed into a gas or liquid and forced to oscillate, the torsional
oscillations will be damped. The damping is caused by the viscosity of
the surrounding medium. The amplitude of the oscillations decreases
monotonically; each successive amplitude is related to the previous
by a constant factor. Hence the successive amplitudes form a convergent
geometic series tending to zero. The natural logarithm of the ratio of
two successive amplitudes is termed the logarithmic decrement of damping;
and is designated by L
The logarithmic decrement is proportional to the viscosity. This forms
the principle of the method of viscosity determination proposed by Coulomb.
Thus, the oscillation period and the logarithmic decrement are determined
in the experiment.
Though the theory of the method is still far from clear, the simple
designs of the viscosimeters used and the special advantages of this method
at low and high temperatures have promoted its wide application.
The viscosity equation was first proposed by Maxwell:

(80)

50
where 't is the oscillation period and C a constant dependent on the size of
the viscosimeter (it is determined by calibrating with a gas of known
viscosity, usually air), and k is that part of the logarithmic decrement due
to friction of the accessory parts of the viscosimeter; k must be determined
in a special test.
In the majority of cases when the viscosity was determined by the
oscillating disk method, this equation was used.
However, in 1939, Macwood /55/ derived other equations, and though
these have no dear theoretical foundation, they are more accurate. This
author considers two possible cases.
Case I. The distance D between the swinging disk B and fixed disks A
(Figure 28) is such that the disturbance caused by the disk oscillating in
the viscous medium does not reach the fixed ones. The disturbances are
completely braked in this medium, and the logarithmic decrement must not
be too large.
MacWOOd derived the following equations for this case:

(81)

(82 )

here k is the viscosimeter constant given by k = 7t~4 (where R is the radius

of the disk and f the moment of inertia of the whole suspended system);
'tand '0 are the oscillation periods in the medium investigated, and in
absolute vacuum, respectively; ), and Ao corresponding to 't and 'to are
the logarithmic decrements of damping, p is the density of the medium
investigated, and 2d is the thickness of the disk.

FIGURE 28, Diagram showing the po·

sitions of the swinging (B) and fixed
(A) disks.

Whether or not these equations can be applied to calculate the viscosity

coefficient is determined from the criterion

D-;- "i > 4. (83 )

V 1~

51
 An analysis of the above equations and the experimer:tal results showed
that the equation holds for the viscosities of gases and liquids studied at
high pressures, provided that the D is sufficiently large.
Case II. Distance D is chosen so that the disturbance of the viscous
medium, caused by the torsional oscillations of the swinging disk, is
damped to a small extent only, and almost the total disturbance reaches
the fixed disks. From here it is reflected. Macwood derived the following
equations for this case:

't) = 'f( 1-
4dD~~) ,
----~,-'- (84 )
(
2dD~i'-R
,

'f/ , = -kD
2 [( -,
1: 'to
)" 'J
) . - - )'0 '
"Co
(85 )

It is easy to see that if ~/~o is close to unity, equation (85) is identical

with the Maxwell equation.
Whether or not (84) and (85) are applicable can be found from the
criterion

D'~
V ~ <0,3. (86 )

The second case occurs when the gas viscosity is measured at low
pressures and at a small D.
From criteria (83) and (86) we can always select the distance between
the swinging and fixed disks in the viscosimeter so that equations for
cases I and II hold.
The Macwood equations were experimentally verified by Keesom and
Macwood. The authors measured the viscosity of liquid hydrogen in the
same instrument and under the same conditions, changing D only. The
viscosity was calculated from (84) and (85) when D was O.06cm, and from
(81) and (82) when D was 1.5cm. The viscosity coefficients calculated
agreed within the experimental error.
The Vogel viscosimeter. The viscosimeter of Vogel /46/ is used to
determine the viscosity of gases at low temperatures; it is shown in
Figure 29. It is convenient for high- and low-temperature me:.surements,
since the part of the instrument which must be held at the test temperature
is very small. The wire rotates together with the disk, and its elasticity
is a function of the temperature, but it can be kept at room (constant)
temperature. In this instrument a glass disk d, of thickness 1 mm,
diameter 39.8mm, is used together with nickel wire Ni, length 220mm,
diameter 1 mm.
The other end of the wire is connected to mirror S, and the top of this
is joined to platinum wire Pt of diameter O.05mm, length 200mm. The
platinum wire is in turn connected to control screw e, which serves for
zero -point adjustment. The mirror is viewed through window H. The

52
 fixed disks a are placed above and below the glass
disk, and along their outside circumference are held
in position by clamps F. The gas sample is filled
through tube E. The disk is made to oscillate by
bringing a magnet close to the small astatic
H
magnets ff fixed to the nickel wire.
If
Vogel determined the friction of all the parts of
the suspended system. He worked with two viscosi-
meters with disks made of different materials. The
lengths of the suspended systems and the oscillation
periods were the same in all cases. The viscosity
coefficients were calculated from the Maxwell
equation.
The Gonikberg and Vereshchagin viscosimeter /56/.
The scheme of the installation and the construction of
the instrument are shown in Figure 30. Steel
bomb 4 contains brass viscosimeter 6 in which
there is a suspended system. The latter consists
of a brass disk suspended oveX' its length on
manganin fiber 7 and steel roc 11. Disk 12 hangs
FIGURE 29. The Vogel midway between two fixed disks. Iron rod 17
viscosimeter. (perpendicular to the plane of drawing 1is attached
on the fiber, and placed between the two coils IS.
The magnetic field of these coils generated by the
electric pulseE: i.mparts a rotational momentum to the suspended system.
Then the harmonic oscillations of disk 12 and the whole system are
damped, the logarithmic decrement being constant. At a distance of
2.5m from the bomb a 30-cm scale is placed, the reflection of which in
mirror 19, which is fixed on the fiber of the viscosimeter, is observed in
the telescope after the ray is deflected through a right angle by a
prismatic reflector. The mirror is adjusted by turning screw 5 through
the channel of the obturator of bomb 3. The disk adjustment is observed
prior to the beginning of the experiment through two coaxial sight glasses 9,
which aftcr adjustment are sealed by gaskets. Window S is made of
optical glass and its obturation is effected by an optical contact ground to
2-3 Newton rings. The bomb is placed in thermostat 10 and designed to
withstand pressure of 3000 atm. The bomb is adjusted by three screws 1.
The gas is supplied through capillary 2. Electric lead 13 for thermo-
couple and the thermocouple inside the bomb made it possible to check
the temperature of the gas during the experiment. The figure also shows
the electric lead 16 for transmitting the electric pulse into the coil,
stirrer 14, motor with stirrers 15, and the electric heater 21.
Technical data of the instrument: manganin fiber 7: length 290mm,
diameter 0.05mm; steel rod 11: length 105mm, diameter O.Smm;
mobile disk: diameter 30mm, thickness 1.272mm; distance between
immovable disks 11.7 mm .
The viscosity coefficient was calculated from (SI land (S2 l.
1.0 was taken as 0.0003 (data of Vogel) and ~o was found by extrapolation
to be 17.0 sec. ~ is equal to half the difference of the natural logarithms
of the arc lengths between the extreme deviations of two successive
oscillations. In the experiments of the authors ), was constant to ± 0.50/0.

53
Expression k = 1C~4' where J is the moment of inertia of the whole system,
was determined by standardization with nitrogen at 25°C and 100, 200, 300,
and 400 atm. The results of the standardization show that k fluctuates around
an average value of k = 3.10±0·.60/0. To determine the additional retardation
caused by the mirror, experiments were carried out with diethyl ether.
The values of A were compared when the mirror was immersed in liquid
ether and in ether vapor. There were no differences between these values
within the limits of accuracy.
Gonikberg and Vereshchagin measured the viscosity of ethylene at 24°C
up to pressures of 1000atm.

FIGURE 30. The Gonikberg and Vereshchagin vis·

cosimeter.

54
5, Concentric cylinder method

Among the viscosimeters used for gas determinations, the instrument

developed by Timiryazev /3/ is interesting (Figure 31). The instrument
consists of two coaxial cylinders C 1 and C 2 . The
inner cylinder C 1 with polished outside walls is
suspended on a thin wire to which mirror M is also
fixed. Cylinder C 1 is placed between two fixed guard
rings 01 and 02' Outside cylinder C 2 is set in motion
by the rotation of axis A. The angular momentum
is transferred from the outer cylinder to the inner
=~ through the gas, Thus, a moment of force acts upon
the inner cylinder. which rotates C 1 through a
certain angle. Under steady-state conditions the
forces of internal friction will be balanced by the
c2 torque of the wire

M=/C1., (87 )

where C1. is the angle of twist, which is determined

with the help of a mirror directing the reflected
light ray to the calibrated scale. The scale values
are read through a telescope.
FIGURE 31. The Timir- The torque caused by internal friction is found
yazev viscosimeter. as follows: the gas layer adhering directly to the
inner cylinder has an angular velocity OJ! = O. The
gas layer near the outer cylinder has a velocity OJ2,
and the intermediate layer rotates with an angular velocity between
zero and OJ,.
Let us dete:rmine the transfer of angular momentum in unit time
through a cylindrical surface of radius r and height 1 cm. The momentum
transferred per unit time through cylindrical surface S = 2 "r is equal to

Hence, the transfer of angular momentum through surface S is equal to

(88)

Under steady-state conditions the angular momentum transferred in

unit time through an imaginary cylindrical surface drawn arbitrarily
between the inner and outer cylinders must be constant (Q =const), whence

r3 dO) =a (89)
dr '

const
where a=--~~.

55
 By integrating (89) for r from Rl to r, and accordingly for Ul from zero
to we obtain
Ul r ,

(90)

Since when r = R2 Ul r = W2 , we have

(91 )

The value of a is found from (91)

By substituting this value for a into (90), we obtain

__"'2RiR~
wr -R~-Ri
(_1R;__ ~_)
r2 '
(92)

In these equations Rl and R2 are the radiuses of the inner and outer
cylinders. From (92) we can calculate the angular velocity of any gas
layer at a distance r from the cylinder axis.
By differentiating LOr with respect to T, and substituting the differential
corresponding to r into (88), we obtain

(93 )

Evidently Q also determines the change in the angular momentum per

second near the inner cylinder. Since the increase in angular momentum
per unit time is numerically equal to the moment of force, Q is also
equal to the moment of force acting on unit length of the inner cylinder,
and to the torque of the wire.
Thus

whence

(94)

All the magnitudes on the right- hand side of (94) are known except t, which
can be determined from the torsion modulus of the filament material and the
dimensions of the filament.

5680 56
6. Viscosimeter for liquefied gases

The determination of the viscosity of liquefied gases has some specific

features due to the high vapor pressure and the low temperature. Of the
numerous visGosimeters described in the literature, we shall describe
only the most interesting and typical.
The Shatenl3h.tein, Izrailevich and Ladyzhnikova viscosimeter /57/.
This viscosimeter is a modification of the Ostwald viscosimeter used to
determine the viscosity of liquids. It is shown in Figure 32. Capillary
5, of length 115mm, diameter 0.7mm, is joined to bulbs
1 and 3 and reservoir 6. The volume of bulb 3 (6cm 3 ) is
fixed by marks 2 and 4, etched on the sections of the
capillary of diameter 1 mm. When bulb 1 is emptied,
reservoir 6 is filled, and the liquid level reaches
position 7. When bulbs 1 and 3 are further emptied, the
sample liquid flows over into reservoir 8. Thus, in aU
measurements the mean column height is the same. The
glass viscosimeter is open at the bottom end which is
fixed into a stainless steel socket 9 and closed by a
a screw cap 10. The viscosimeter is placed into a thermo-
stat on a special stand, which makes it possible to turn
the instrument through 180 If the sample liquid in the
0
•

initial position is in reservoir 8, it fills bulbs 3 and 4

6
when the instrument is turned around its axis through
180 0 • If on the other hand, the instrument is brought back
to its original position, the liquid flows under gravity
from bulbs 1 and 3 through the capillary into reservoirs.
The time taken for the liquid between the two marks to
flow through the capillary is measured.
The other data necessary to calculate the viscosity
FIGURE 32. The coefficient from the Hagen -Poiseuille equation (pressure
Shatenshtein, drop at the capillary ends, volume of liquid flowing during
Izra ilevich and the measured time, and the geometic dimensions of the
Ladyzhnikova capillary) are constant, and are previously determined.
viscosimeter.
The main dimensions of the viscosimeter used by the
authors to find the viscosity of liquid ammonia are shown
in the figure.
The Rudenko viscosimeter /58/. The method of the rolling cylinder
was modified by Rudenko as follows. The torsional oscillations of the
~ylinder suspended by a flexible filament and submerged into the liquid
~an be described by the equation for the damping of oscillations:

(- -0) t 2"
o;='f'e T COSyt, (95 )

'",here ex is the deflection angle, 'f' the initial amplitude, and 3 the
damping decrement. Damping is mainly determined by the viscosity of
-he liquid. To a first approximation, it can be assumed that the decrement
of damping is proportional to the viscosity, or 3 co "l. If we de signate the time in
which the amplitude is reduced to a half by T, from (95 land from the expression
Ii co "l we obtain ~1~ = const. Consequently, when determining ~ we obtain a vah;e
for 'fJ. A more detailed study, however, showed that the relationship between the

57
 damping decrement and the viscosity is more complex, so that it is very difficult
to use this method to determine the absolute viscosity. For this reason
Rudenko used his method as a relative one. The viscosimeter was
calibrated with compressed gases of known viscosity.
Figure 33 shows a cross section of the viscosimeter. Solid aluminum
cylinder 4 is suspended on a steel filament of diameter 0.2mm and length
850mm. The end of the filament is pressed into a copper rod of 2 mm
diameter screwed onto the cylinder. The sample is in copper vessel 3,
able to withstand pressures of 60 atm. The viscosimeter is brought into
the vertical position by means of four screws. Then the lower screw
socket of vessel 3 is screwed on its thread and soldered by tin. Mirror 1
is fixed onto the filament opposite a glass window. The mirror-and-scale
method is used to observe the oscillations. The upper part of the
viscosimeter contains a three-way piece through which the gas sample is
condensed. The filament is led out and
centered above the viscosimeter, and soldered
with Wood's alloy. The cylinder is first
inclined away from the initial position by means
=""i1l~=-To gas of a permanent magnet. The cylinder must
(llr;==!;8P'=cy linder first be pressed into an iron cross piece. The
2 test temperature is reached by immerSing
the instrument into a Dewar flask which is
7
cooled by blowing cold air through spiral 6. By
controlling the amount and temperature of the
air blown into the spiral it is possible to keep
the temperature in the Dewar flask constant.
Temperatures between 91 and 112°K were
measured from the vapor pressure of methane,
from 125° to 169.5°K from the vapor pressure
of ethylene, and from 206 to 281°K by an
alcohol thermometer calibrated according to
the melting points of ice and mercury and
the sublimation point of carbon dioxide. Some
points of constant temperatures were obtained
3 by using commercial compressed gases 02' N2
4 and H 2 • The time was measured with a
5 stopwatch. The"t values varied between 17.2
5 and 240 sec. The oscillations of the system
were characterized by the following data: for
an "amplitude" of 8° the linear velocity on
the generatrix of the cylinder was 0.12 cm/sec,
and the oscillation period was 3.4 sec. The
measurements were carried out at amplitudes
FIGURE 33. The Rudenko vis- of 8.6 and 4°. The oscillation period remained
cosimeter.
constant in all the liquids studied.
With this viscosimeter Rudenko determined
the viscosity of nitrogen, oxygen, methane,
ethylene, and air at temperatures close to the critical. In all the experi-
ments the pressure was 0.5 atm above the equilibrium, so that the liquid
could not boil even if there were chance temperature variations.

58
 Gerf ar:.d Gal.kov /59/ used a viscosimeter of the same design but with
a platinum cyj.inder and somewhat different dimensions of the individual
viscosimeter parts. They determined the viscosity of several liquid gases
and numerous gaseous mixtures. Verschaffelt and Nicaise /60/ used a
solid oscillating ball instead of the oscillating
cylinder, and determined the viscosity of liquid
oxygen and nitrogen in a mixture with small amounts
of argon.
The Verkin and Rudenko viscosimeter /61/. The
viscosimeter shown in Figure 34 was used by Verkin
and Rudenko to determine the viscosity of liquids and
gases at constant density but at various temperatures.
Inside the heavy high-pressure cylinder 1 made of
nonmagnetic bronze, a 200mm tube of 4.0mm
diameter is inserted. The tube must be prepared
with the utmost care. The sample gas entering
through channel 5 is condensed in the tube. When
the condensation is complete, the volume of liquid
is shut off at the top by valve 6. Cylindrical
weight 2 moves in the sample liquid, and from the
time it takes to fall from one mark to another
engraved on the cylinder, the viscosity coefficient
of the sample can be calculated. Weight 2 consists
of a cylindrical case of the following dimensions:
length 12mm, outside diameter 3.970± 0.002mm. A
column made of Mishima alloy is squeezed into
the case. The high- pressure cylinder, the tube
filled with the sample, and the casing of the weight
must be made of the same material. Two induction
coils 7 are fixed to the jacket of the cylinder. They
are connected in series with a mirror galvanometer.
When the falling weight passes each of the coils the
galvanometer is deflected. In this way the time
between two zero deflections of the light spot (weight
FIGURE 34. The Verkin
in the center ofthe coil) is measured.
and Rudenko viscosimeter.
The first measurement is taken at the conden-
sation temperature of the sample gas. In this case
the viscosity coefficient must be known. Let the
parameters of the first measurement be "1)0 and ~o" By varying the
temperature in the instrument and measuring the time of fall of the weight
each time, we obtain

, (96)
"'Ip=const = "1]0 ~'

":hat is, the viscosity coefficient at the test temperature and p = const. The
density of the sample liquid is given at the temperature at which it condenses
:tn the instrument. Thus, the method is only relative.
The temperature is measured by two resistance thermometers 8 placed
on the walls of the high-pressure cylinder. The time taken by weight 2
to fall was between 20 and 100 sec in the experiments of Verkin and Rudenko.

59
 The high-pressure cylinder was installed i.nside a metallic Dewar flask
/3, 4, 10/. The weight was set in motion by rotating the flask through
180 around the axis perpendicular to valve 6 and in the plane of the diagram.
0

Liquid nitrogen and liquid oxygen were used as coolants. At higher

temperatures the measurements were taken with the temperature variations
in the instrument. To decrease these variations in the viscosimeter, a
compact lead block 11 was soldered to the upper part of valve 6. The time
during which the temperature of the instrument changed from 90 to 300 K
0

was 16 hours.
It was shown that the sample must be entirely free of solid particles.
Therefore it was distilled several times before it was condensed in the
cylinder.
Verkin and Rudenko used this viscosimeter to determine the viscosity
of liquid nitrogen and liquid argon at various densities over a wide
temperature range. With the same viscosimeter Zhdanova /62/ determined
the viscosity of liquid nitrogen in the range from the density near the
triple point to the density at the critical point. In the experiments of
Verkin and Rudenko, and of Zhdanova, the pressure had to be increased to
3000 -4000 kg/ cm 2 to keep the density of the gas or liquid at a constant
value when the temperature was increased. The cylinder was therefore
made of beryllium bronze, which can withstand such high pressures.

60
Chapter III

EXPERIMENTAL RESULTS

1. Viscosity of gases at atmospheric pressure

Many handbooks /64/ and periodicals give experimental data on the

viscosity of various substances in the gaseous state at atmospheric pressure.
Most of the data. were obtained by the capillary method. The results of
different authors for the same gas are in good agreement.
Table 4 was compiled by graphical and analytical processing of the
numerous experimental data available up to 1959. The experimental data
of Trautz and his co-workers /64/ were especially considered. These
authors measured the viscosity of gases by the capillary method, and their
results are probably the most reliable.
We also used the data of Brawne et a1. obtained by the oscillating disk
method /65/.
The data of these authors for I') are rather more temperature-dependent
than those obtained by the capillary method. The viscosity coefficients are
somewhat lower at low temperatures and higher at high temperatures.
The following general tables were also used: the tables of Frost /64/ for
hydrocarbons, a.nd the table compiled by Golubev /122/, Golubev and Petrov
/63/ and Vargaftik /83/ for other gases.
Table 4 waE: compiled for 96 gases at all the temperatures at which the
viscosity coefficients were measured. Sometimes extrapolated data were
also used. It will be shown in section 2 of this chapter that the gas viscosity
below 760 mm Hg is pressure-independent up to 1- 2 mmHg. Therefore,
in Table 4 the 71 values for some substances are reported for temperatures
corresponding to the saturated vapor pressure of these substances below
760 mmHg.

2. Viscosity of gases at reduced pressures

It has already been mentioned (Chapter 1, section 2) that the viscosity

coefficient of gases begins to be pressure-dependent (it decreases with
decrease in pressure) when the distance between the surfaces over which
the gas flows ii, of the same order or smaller than the mean free path of
the molecules. If the gas is highly rarefied, it slips along the surface of
the solid body. Therefore, in analogy to the coefficient of internal friction
(1')), the magnitude of external friction ('1/,,) was introduced (Chapter II,
section 1).

61
 TABLE 4_ Viscosity (10- 7 g cm- I sec-I) of gases at atmospheric (or lower) TARLE 4 (Continued)
pressure at various temperatures
- I TI~~~-.-~~--.-

, o 0' o o
CJ u
z ~ U Q)
z Z :£'
Q) 0 6 ~ ~ .~ Q) 0
OJ) "o -0W
Il.l 0"0 U Q)
ci "o 0~ Q)
~
'" .0 "~
.;:: I I §
~ 8 be ..0....... ';j 'U H X 0
0 ::1 "0 ::j <lJ"
E
Q)
~ u~J I· ] ~
~"
o01) ""
>.
1-0 CL)
~ r" :>-. N (J) N
6"" g >.... I-< I-< X +-' ....... oj 0 ~.,.-i
'"2 vW/ ~cI:I- 0.,
~
b ..... 1...J.......Lt Z
" -< :;z ~ ~ :r::r: Q Q ~ ~ zz 6 0 ~ u g L~ 3 u:S Z ~ 3: ~
--r- ~---'---+-
"lr~, " ---I
220
210 -200 520
2()0 -190 635
""~ 332 465
-- 190 --180 650 710 I 663
862 1280 710 367 516
-180 930 1400 790 -170 718 785 730
399 565 -160 783 860 800
--170 995 1510 870 430 605
- - 160 I05S 1615 950 160 650 -150 846 935 860 860
-- 150 1117 1720 1025 488 690 -140 907 1010 922 930
-- 140 1176 1819 1105 516 728 -130 967 1085 990 1000
-130 1235 HJ15 1180 542 765 -120 1028 1150 1050 030 I 1064
-110 1082 1218 1109 090 1130
0-,
-- 120 1290 2005 1255 568 802
w -" 110 1345 2095 1325 593 837 -100 1143 1286 1170 130 1198 886
100 HOO 2W2 1405 618 872 75 1285 1152 1312 275 1357 1007 990
75 1526 2388 1585 677 95li 50 1419 1612 1445 400 1510 1126
50 528 1115\"
1640 2WO 1760 733 1035 25 1542 1753 1582 1656 1247 1240
+ 0 1665 1910 1708 1662 1800 1367 1360 883
I
25 1750 2788 1930 788 1110
a 18GO 2975 2085 2330 2110 840 1185 + 20 1766 2026 1812 ! 17491 1899 1463 1460 I
-1- 20 1946 ,3113 221.') 2480 2255 880 1240 1778 2052 1840 1766 1920 1486 1482 975
25 1968 3142
2"
2248 2515 2290 890 1255 50 1883 2182 1954 1872 2035 1607 1595 1065
50 2065 3310 2400 2695 2470 908 1325 75 1986 2310 2068 1980 2156 1716 1712 1157
100 2086 2437 2180 2076 2272 1827 1822 1250
75 2175 3482 2550 2880 2045 985 1390
" 100 2281 3646 2695 3060 2820 1033 1460 150 2278 2674 2391 2271 2475 2043 2040 1435
150 2475 3950 2965 3400 3165 1123 1590 2,152
200 200 2464 2867 2588 2682 2254 224.1 1615
2G72 4248 3223 3725 3481 1213 1715 2622
250 250 2639 3103 2760 2870 2456 2450 1800
2875 4527 3462 4040 3790 1299 1830 2788
300 2800 3310 2942 30.05 2646 2649 1985
400 3118 3686 3275 3090 3400 2994 3030 2350
300 3055 4785 3680 4325 4080 1382 1955
400 3415 5300 4110 4850 4G20 1538 2170
500 3750 5796 4505 5:350 5120 1686 2380 500 3403 4030 3567 3370 3700 3309 3375 2720
600 4070 6255 4870 5840 5600 1828 2580 600 3665 4350 3842 ,3630 4010 3605 3090
700 4370 6689 5215 1965 2780 700 3916 4652 4110 3870 4275 3876 3460
800 4160 4940 4365 4100 453S 4140 3820
800 4660 900 4390 5210 4610 4330 4780 4400 4190
7100 5550 2103 2970
1000 5240 7855 6160 2355 3320
1000 I 4600 I 5470 I 48f)()J 4330 iu507.'i.1 4658 i I 4560
 'T''' '~T r A /;-. __ . , _ "
-'.~.'~u"", -:t" \"'~VLILJ.HUC.UJ TAGLE 4 ((;ontinued)

, 0'
VJ , c _
;; c: s:
'2 <lJ gj' V " - <lJ OJ <lJ c:
~}J <lJ c: '" c c: <lJ
"00 "0
'0." "0 o 8, ~ <lJ "0.~ C 00 "" 00 -
. 'g" .§ o .- 8 .§ "
o "0 ~
:c:(;;"
~ M '" ~
6 - " x 6 8;oj .c .. ~ ~ .S
>-.~ "0 .- M
<lJ ~ ~
.- 0 6 x -8
<lJX " <lJ <lJ (:) .c X -8 N ~
6<lJ " :2~ '" .u>- ~0 u. . . ~8 o
.c X
f-< " VJ
" "0
.- «:2 2u (:) 6 u wU "-'U f-< ;: U u ~£ "".s~ ::c-B::r: X.o ~l £] '" '"
--0-~45
1390 1320 1710 1700 1072
-190 I .3.36 399
14')0 lRJ..1 1R,<{\ 11 ~f\
20 i 1335 1495 - --- ....'-'-::;: ..... '"VVV J.J.VV
-180 I 37.0 446
-170 ' 414 492 25 1360 1520 14.50 1877 1862 1171
-160 453 .')40 50 1465 1645 1580 2040 2020 1264
-150 490 584 75 1570 1770 1705 2200 2172 1358

-140 .128 629 100 1675 1890 17 30 1830 2365 2323 1451
-130 566 G75 150 1882 2135 19 53 2070 2690 2625 1635
-120 603 718 200 2090 2380 21 75 2302 3010 2923 1818
-110 640 761 250 2295 2625 24 )0 2530 3216
-100 G77 806 552 300 2505 2865 26 30 2755 3510

75 855 670 769 915 623 400 2920 3340 30 35

~ 50 955 760 860 1022 703 792 500 3326 3815 35 05
W ,
25 1126 780 864 600 3730 4290 39 35
1053 850 946
+ 0 1158 935 1028 1223 855 941 C
20 1250 1000 1092 1290 915 1010 0 g
, g % u
_ <li
~
" V
~ N
25 1270 102S 1108 1319 929 1027 ......... '"0 ........ "0
~ -0U
>- v c:"0
~~ <20
0. ~
£.;:: <Ii (J 2';::1<lJ
I o .-
50 1390 1110 1185 1110 1020 1108 <YO o 'c ~ U'" ~
.0 0~ _"
....., Coo ........ t: '§ ~ ......>-0_
1180
E
<lJ
<u
~
oC
::a :r: .:; .s -8
.- .c .3::::a~
.c - ~
" '" ~ U
75 1500 1205 1260 1500 1070 f-< ;: U U ~~G~~ g-._-z v ~ () '-' ~-Bc u -5 u
100 1611 1285 1332 1590 1150 1260
1280 1405 0 933 9H9 12 12 1084 9!(; 911 906
150 1830 1463 1472 1750
20 1000 101)2 13 ?7 1170 989 978 966
200 2Q40 1G48 1604 1910 1410 1545 25 1017 1080 13 ;0 1193 1005 994 982
250 2255 1820 1725 2055 1525 1670 50 1100 1173 14 'I 1297 1095 1080 1061
300 2455 1990 1850 2200 1640 1810 75 1182 1264 15 19 1402 1181 1164 1141
400 2820 2337 2080 2480 1900 2049
2700 2140 22(;0 100 12(;;) 135,) 17 )7 1503 1269 1247 1216
500 315,5 2650 2208
150 1430 1537 19 l5 1718 1440 1410 1367
600 3465 2925 2465 2940 2380 200 1590 1725 21 )0 1910 1602 1574 1517
700 3760 3210 2650 250 1750 1895 24 10 2115 1766 1736 1670
800 4040 3480 2820 300 1906 2070 26 '0 2315 1928 1900 1815
900 4305 3740 2980
1000 4540 3990 3135 400 nOG 2420 31 2505 I 2220 2225 2100
51
 TABLE 4 (Continued) TARLE 4 (Continued)

s ..,
, (f)
,cJi' t)j) ~
"-0 0 , ft
.~
, (f) -"j 0 ft c:
~
23 u M (j::r::.. u::r::
~ U
:oU .c:
C
JCfT--J--- ~.
";:! ~ .~ ~ 0t: ~ N ()
a-
(J) U
u: E
<lJ ::l
""go '" z '" C)~
"
"n, 0 "o.g;~ n.
"o .g;- 1J",,"(f) ~wO oN (f)
.D .~ "' M § ~ 'g ~"§ ~ § ~ u "2:g "d
.;:; --84:! .~ .~ C 0 C
E Q; o :r:.., .c: " M :r:
:-sE 's-0
"'~
~
m ~ ""' .c: .., ~ '" :r: .c: M
"'"'" S E ~ DO~:2 ~8~ ~--g .S
'" ~
() ~
~.c:
...<; o..U.. U) if.) I-<
~ b U I-< U
I-< 3 :r:'" u " 3'"~ .2<.X:~UL;..Cl:r:l..<;· ...... N -~:e~~uu
·-e----- I

0 920 11:32 100 . 1330~4221---

755 690 910 1076
20 990 1220 614 742 958 1 ]48 1,,0 1485 1496
25 1008 1240 628 756 972 1167 200 4500 I 2200 I 2253 I 2045 1640 1770
50 1094 1346 901 824 1034 1258 250 5020 2460 I 2187
I 2262 1795 1945
75 1184 1456 973 888 1098 1344 300 5530 2725 2725 2478 I 1950, 2112
!
100 1262 1557 1045 954 1160 1435 400 6570 32561 3190 29081 2260 2468
150 1432 1755 1190 1085 1285 1610 500 7620 3792 3664 3335 5250 2570 28181 5590
200 1600 1333 1218 1410 600 8770 4310 I 4120 3752 'I· 6000 I 2880 3160 6490
250 1766 1472 1349 1535 700 6750 7300
300 1930 1612 1478 1658
~
.j>. 11.)
---- . - .

, ,,~ ~:v W r:S-== ~ ~

~
C)
U
0 Q)~
,r
===T i7--n~ 1 I---~ 1-
... C I W c: c: ('j C1j C
oS
"'>< "'I'
I-<
I'~ ~,~
C'?
>-.
.0
<.()
"~:3- "g00 .9 z ~~~ I ~a.. j ~ ~ i::~ ~o..l::~
~ ~ ] :r:;'"
Be\) a .ol: ~l:: £ :2 ~"t £
" v '§ ~ ~ ~O ~Z d U IN~ 8 U
"
[-<
~
B --C < U U .5:'0' ~§t J:: ;, 121U""I;'00' 21U'
.~--~--
-_.------ :r: " :r:
o 750 682 689 623 638 600
0 1468 67:2 933 20 I 800 735 744 668 685 644
20 1576 737 1002 25 I 811 749 758 680 696 654
25 1602 754 1020 50 878 814 821 737 752 710 691
50 1734 834 1108 830 75 942 881 886 796 805 765 741
75 1865 914 1194 892
100 1006 947 947 850 860 820 768
100 1996 994 1279 955 150 1130 1070 1074 967 969 930 857
150 2260 1150 1452 1084 200 1248 1185 1202 1079 1079 1040 941
200 1305 1208 250 13(i2 1300 1326 1191 1185 1150 1027
250 1·160 1334 300 1475 1420 1450 1295 1291 1260 1110
300 16],3 1462
400 1715 1650 1510 1475
400 1920 1720 500 1940 1880 1725 1685
600 2180 2100 1930 1900
 TABLi: '"* (Connnued) TABLE 4 (Continued)
<Ii <Ii
c: c:
I ~ 1l.i' 1l.i' ~ go} ~i ~~ 0.
'" c: ' .
m U t'$ c: c: c: <J) c: III c: <J) c:: 8 '"x <Ii <Ii
~ u 0.
'" ">.."
..c: c: ~ "-l
~o b..1t) ;] ~o ~ 'E-5:;~ ~~ " 0 0
"
0 N
"
~ N
E~ Ill..-l 0..-10"-l ~<D::lCOOO:::l~oo::l..DOO ~
~
.~ :£
'"E <Ii 'Ox ""N ""'" :r:~ i~~
<J)~ ~~ 9~~~ e~~~~~I~-g~~ >, ~ ~~ "
JO, o ~
f-<C C::U c:U c:U o...U Ci,-Uq)~U~~U b" 2~ UU U U ~u bU :;;;" U~ iSEu
0 ROR n.S.1 693 , ,
20 870 701 745 686
o 784 708 G94 732
25 884 712 758 698
20 840 761 746 786 50 960 770 822 763 660 605
25 855 775 759 801 75 1038 817 886 826 704 653

50 618 586 554 930 841 825 870 100 1113 878 950 891 745 700
150 12b5 986 1077 1008 331 796
75 667 630 592 1002 906 889 938 200 1414 1090 1202 1120 907 888
250 1191 1326 1230 988 976
100 716 674 633 1073 971 952 1006
300 1291 1453 1335 1067 1063
150 810 757 702 1211 1098 1076 1141
400 1482 1690 1545 1228
200 900 843 778 1340 1228 1192 1273 500 1670 1926 1745 1390
0\
V>
250 985 927 SGO 1470 1355 1301 1408 600 1860 2170 1950 1530

300 1072 1007 920 1592 1483 1410 1542 <lJ

<Ii 'iii
, '0 tU
()
0
~
'iii
'0 (J
"c: '0 ":::r: '" :r: B '" ~
~ .s~l I ~ Ql~
'" '" :r:
"'"0 "'" :r: ">-u '" N ~o
"
<J) "a} c: <l) I lQ rl ~ 1-1 "<t' 0. go .0 • ~o
E • ~5 sC?- 8~ '" 0
OJ ;:: ~£
'" ~
..c::r: ££ g- ~I " ()
'" ..., <J) ~
'" ~ '"o :I:~
~ " ..c::I:
~<l)£J ~~
>-'1=:
~v
..c!C
~~U", ~
1-I1-I0)...c::" <l)
~ g~
.-U b ;: ~u wU ~u ~U -<
-
U :;;; ~ wU o:u
c.. p<l)oQ)~oQ)Q)1=: G)~o>" >-.~
8 i ~ '7 =€: ~ E £ .§ ~ ~ 2 3:£ Q) -£ ~ ~ o 870 77;5 715 720 685 690 740
~a ~~u ~~u ~~E ~Eu ~u ~~ 20 935 835 770 774 735 743 797
25 950 850 780 786 750 756 811
o 629 665 639 665 955 808
50 1040 930 850 855 815 870 820 880
20 676 716 686 716 1017 867 75 1130 1010 910 915 880 944 884 952
25 685 729 698 729 1033 879
100 1225 1083 970 945 1016 949 1022
980
50 744 800 754 792 1110 951 150 1393 1230 1111 1112 1076 1160 1078 1162
200 1563 1375 1245 1242 1210 1298 1206
75 801 857 820 854 1188 1020
250 1725 1512 1380 1370 1340 1435 1333
1090 300 1895 1650 1500 1505 1466 1567 1462
100 858 915 886 915 1265
150 972 1050 1102 1040 1417 1230 400 2220 1920 1755 1760 1720 1830 1710
500 2550 2200 2010 2020 1975 2090 1960
TABLE 4 (Continued)

i~ di~o~ li~ hi ?~n~li~ !~

z;~
~3
..s2!.l--'
~o]E]8
-'-' I
:2~
E;~C):Co...N(1):r:
8~J OU OU ~u
--'--0-+1' ---+---+I----:I~I 870
680 :
20
25
I 890
906
778
800
840; 824
I'

858 843
925
942
730
742
50 990 906 I 951 934 1020 ROB 6iO 825
75 1074 1014. 1047 I 1028 1095 870 6.57 912
i
100 ,1158 1120 I 1142' 1120 1168 930 703 1000
150 1325 1332 ,I 1330 i 1304 I'1315 1050, 797 1170
200 1465 1172 i 888 I 1340
250 i I I 1Ii 10 , 1290 I 978 1 1510
300 1750 i 1405 1064 1670
400
I 206,0 i _---'
1640! 1238! 1980
__I

TABLE 5. Viscosity (10- 7 g C'11-

1
sec- 1) at reduced pressures

Hydrogen Deuterium

50.05 'K 20.4 'K 90.15'K I 20.2S'K

p,mmHg 11 p,mmHg '1 f,mmHgj '1 kmmHgJ n

760 394.1 700 11 0.7 1 760 II 5445 i 700 1/

138 .0
7.31 391.0 7.19 111.8 536 5434 I 7.19 1383
1.542 382.3 0.899 110.3 2.469 i :27.8 I 0.440 1 134.5
0.569 356.4 0.120 ,101.6 1.211 1 v187[ 0.1039 .126.2
0.260 319.2
I 00562 i 92.5 0.661 I 4989, 0.0424 '111.4
0.116 259.91 0.0237 74.3 0.151 I 410.41 0.0180 i 83.6
0.0568 185.8 0.0113 52.3 I 0.069 ,301.3 0.007941 54.23

~'~~~~j_ ~~~::J~_~~368 J 2~:

0.0112 56.2
0.00514 2375 1
I
Helium Neon

90.27'K 14.62'K 90.0S'K I 20.42'K

p, mmHg/ 11 p,mmHg n p,mmHgj1l ~,mmHg 11

760 1[910 7GO

! ;)93.71 300 11344 'I' 300 i
359.3
15.4 ~07 3.32
_93.7 9.51I I 13406 I 3.71 !359.3
1876 ! 861.8 0.472
285.4 2.186 11303.4: 0.696 1 3441

0.885 : 816.2 0.1258 , 260.4 : 1.149 11267.7 1 0.217 333.9

0.416 1729.6 0.0528 i 221.0 i 0.600 1209.51' 0.097 316.5
0.228 I
620.9 0.0189 I 142.9 i 0.312 1105.4 0.0246 244.1
0.049 ,275.9 0.00704 64.43) 0.0772 696.5 1 00111165.6
I
0.01 051 66.91
0.0222 3139 i 0.005 92.5
0.0057 36.67 0.0052 77.5! 0.00218' 39.8

66
 The magnitude ~ = _l is termed the coefficient of slip. Knudsen /6/
~l
studied this magnitude; he proved that numerically ( is approximately equal
to the mean free path I of the molecules. Weber /66/ introduced the
empirical coefficient k2 with a value close to unity, and for r, the value
\" = k2i, to make his results the same as those of Knudsen. As long as I is
small compared to the linear dimensions of the system (capillary diameter,
distance between the disks), the effect of slip on molecular transfer
processes can be neglected. When a high vacuum is employed, i. e., I is
commensurable with the dimensions of the system, the effect of slip becomes
considerable. Consequently, with decrease in the dimensions, the pressure
at which the gas viscosity begins to decrease becomes higher. In this case,
the viscosity coefficient is not considered to be the true value but the
apparent, since at infinitely large dimensions of the apparatus the above
considerations no longer have meaning.
Van Itterbeek and Keesom /67/, and later Van Itterbeek and Van Paerilel
/68/, determined the viscosity of different gases at 1 to 2.5 .10-3 mmHg.
The authors employed the oscillating disk method, using glass disks. They
introduced the "apparent" viscosity coefficient r(, which was determined
from the logar:lthmic decrement by the method used for the true viscosity
coefficient at normal pressure. Equation

(97)

is used for high-vacuum conditions, where d, and d z are the distances

between the oscillating and the two fixed disks, and 'lj is the viscosity at
760 mmHg.
The experimental data of Van Itterbeek and Van Paemel for hydrogen,
jeuterium, heLum, and neon are summarized in Table 5, which shows that
below about 1-- 2 mmIIg the viscosity decreases markedly with decrease
in pressure, while from 1- 2 mm to 760 mmHg the viscosity remains
:tlmost constant. The data of these authors can be expressed by an equation
,)f type

~p
'Y/
I
=a,j-rp , (98)

,,,here p is the pressure in cm Hg.

Weber /69/ derived this equation theoretically, and found that the value
a for a disk oscillating between two fixed disks is equal to

(99)

Constant k2 for high-vacuum conditions is equal to unity. The viscosity

coefficients determined by this equation agree well with the experimental
results.
Table 6 gives the earlier data of Vogel for air and carbon dioxide. The
data show a similar viscosity-pressure dependence on rarefaction.

67
 TABLE 6. Viscosity (10- 1 g cm -1 sec -1) of gases at reduced pressures

Air Carbon dioxide

15'C 15'C

P. mm Hg 1] p. mm Hg
I 1]

750 1810 750 1490

380 1828 380 1490
20.5 1817 20 1480
0.978 1770 2 1470
0.956 1730 o.0 c·
1380
0.908 1643

3. Viscosity of gases at high pressures

Many experimental papers have been published on gas viscosity at high

pressures. The viscosity was measured by various methods for numerous
gases in the gaseous and liquid states over wide temperature and pressure
ranges.
Unfortunately, the viscosity coefficients for gases under high pressure
are rather unreliable, unless they were determined by the capillary method.
But even some of the results of this method are incorrect. With other
methods (falling weight, rolling ball, or osdllating disk) only approximate
viscosity coefficients were found. The data of different authors obtained by
the capillary method show fair agreement, but the results of other methods
differ widely. This applies to the agreement between results of the same
method and that between results of the given method and the capillary
method.
Table 7 summarizes all the experimental results on the viscosity of
gases at high pressures published up to 1959.
In the column "Method" the Roman number indicates which design
was used by Golubev and his co-workers /37, 122, 158/, or by other
authors, in the capillary method.
When we determined the viscosity, special attention was paid to the
purity of the substances studied. Nitrogen, oxygen, and argon were very
pure, and contained only traces of contaminants. Ammonia and carbon
dioxide were freed from low-boiling gases by a very simple method. The
gas cylind-er was held vertically with the head at the top, and 40- 50% of
the gas was released: during this operation all the low- boiling impurities
volatilized. The impurity level in ammonia and carbon dioxide did not
exceed 0.1- 0.2%. Methane was prepared by refluxing, and was freed from
C02 traces by alkalifle scrubbing. The purified methane contained only 0.2%
of nitrogen. Ethane was obtained from the town of Grozny, and contained
only propylene as an impurity. This was removed by activated carbon
under pressure. The scrubbed gas contained 99.9% ethane. Propane was
also obtained from Grozny, and contained butane, ethane, nitrogen, and
small amounts of other gases. The gas was rectified under pressure in a
highly efficient column, and the end product was very pure propane (99.95%).

68
TABLE 7. List of papers on the determination of gas viscosity at high pressures

Maximum Temperature
Gas Author Method
pressure, atm range, °C

Hydrogen Gibson /64 b / Capillary 294.7 25

Golubevand Petrov /37/ 800 15-250
IV
Kestin and Pilarczyk /80/ Oscillating disk 70 21
Kuss /74/ Capillary 500 25-75
Michels, Schipper, and Rintoul 1899.3 25-125
/79/
Deuterium lI'lichels, Schipper, and 2149.3 25-75
Rintoul /79/
Helium Kestin and Pilarczyk /80/ Oscillating disk 55 20
Ross and Brown /161/ Capillary 680 (-50)-+25
Nitrogen Michels and Gibson /32/ 965.8 25-75
Boyd /70/ 190 30-70
Golubev and Petrov /37/ Capillary, III, IV, V 800 0-250
Kestin and Pilarczyk /80/. Oscillatory disk 70 21
Ross and Brown /161/ Capillary 680 (-50)-+25
Oxygen Golubev and Petrov /37/ Capillary, III, IV 700 15-100
Argon Golubev and Petrov /37/ Capillary, V 400 0-200
Kestin and Pilarczyk /80/ Oscillating disk 70 20
Air Ruckes /71/ Capillary 39.3 14
Wildhagen /72/ 40 16
Nasini and Pastonesi /29/ 200 14
Golubev /73/ 300 0-100
Kestin and Pilarczyk /80/ Oscillating disk 70 21
Carbon monoxide Golubev and Petrov /37/ Capillary, IV 800 0-250
Carbon dioxide Phillips /28/ Capillary 120 20-40
(C02) Stackelbeck /39/ Falling weight 120 (-15)-40
Comings, Mayland, and Egly Capillary 137 40
/27/
Golubev and Petrov /37/ II, IV, V 800 20-250
Ammonia Stackelbeck /39/ Falling weight 22 (-20)-80
Golubev and Petrov /37/ Capillary 800 30-250
Methane Sage and Lacey /41/ Rolling ball 170 37.7-104.5
Bieher and Katz /45/ 340 25-225
Comings, Mayland, and Egly /27/ Capillary 171 30-95
Meshcherakov and Golubev /158/ Capillary, V 800 (-15)-250
Kuss /74/ Capillary 600 20-80
Ethane Smith and Brown /44/ Rolling ball 340 15-200
Meshcherakov and Golubev Capillary, V 800 (-15)-250
/158/
Propane Sage and Lacey /41/ Rolling ball 137 37.7-104.5
Smith and Brown /44/ The same 340 21.9-120
Bicher and Katz /45/ 340 25-225
Comings, Mayland, and Egly /27/ Capillary 41.8 80-104.5
Golubev and Petrov /37/ Capillary, V 800 25-250
"-Butane Sage, Vale, and Lacey /75/ Rolling ball 136.1 37.8-104.4
[sobutane Sage, Vale, and Lacey /75/ The same 136.1 37.8-104,4
Ethylene Mason and Maass /53/ Oscillating disk near criti-
cal point
Comings, Mayland, and Egly /27/ Capillary 171 30-95
Gonikberg and Vereshchagin /56/ Oscill aring disk 1000 24
Golubev and Petrov /37/ Ca pill ary, V 800 24-150

69
TABLE 7 (Continued)

Maximum Temperature
Gas Author Method
pressure, ann range, °C

Propylene Golubey and Petroy /37/ Capil1ary, V 800 18-250

Methanol Golubey and Petroy /37/ " V 800 150-270
Ethanol Golubev and Petroy /37/ V 1000 0-265
Isopropanol Golubey and Petroy /37/ " V 800 155-260
Isobutanol Golubey and Petroy /37/ " V 800 220-270
Dimethyl ether Golubey and Petrey /37/ " V 800 100-250
Diethyl ether Golubey and Petroy /37/ " V 800 135-250
The same Schroer and Becker /78/ Falling weight 75 184.5-235
Acetone Golubey and Petroy /37/ Capillary, V 800 200-265
Benzene Golubey and Petroy /37/ " V 650 95-248

Propylene was obtained by splitting chemically pure isopropanol on a

catalyst (alumina gel) at 350°C. It contained 99.9% propylene. The ethylene
used was of the same purity. Methanol, ethanol, isopropanol, isobutanol,
acetone, benzene, diethyl ether, and other substances were distilled with a
highly efficient rectifying column. Numerous fractions were obtained. The
middle fraction only, in a very narrow temperature range (the temperature
was almost constant), was taken for the investigation. The purity of the
substances obtained was tested by measuring the specifiC gravity,
refractive index, and boiling point.
Viscosity of different gases. Hydrogen. The viscosity of hydrogen was
determined by several research workers. Gi.bson measured the viscosity at
25°C and up to 294.7 atm. We used the temperature range 15 to 250°C, and
pressures from 1 to 800 atm (Table 8). Kuss /74/ determined this parameter
at 25 to 75°C, at 1 to 500 atm. The determinations of Michels, Schipper
and Rintoul covered the range of pressures from 26 to 1900 atm at 25°C,
and up to 1000 atm at 50, 75, 100, and 125°C (Table 9). All the results were
obtained by the capillary method. Gibson, and also Michels, Schipper and
Rintoul, used the Michels and Gibson viscosimeter /32/. Figure 35 shows
comparative data on the viscosity coefficient of hydrogen, and it can be seen
that there is good agreement between the data of all the authors listed.

TABLE 8. Viscosity Clo- 7 g cm -1 sec -1) of hydrogen (Golubey and Petrey)

~I p. atm
15 25 50 I
I
100
I
150 200 250

I 866 886 935 1030 1125 12'10 1295

50 875 895 943 1040 1130 1215 1300
100 885 905 952 1050 1140 1220 1305
200 910 931 977 1070 1155 1235 1315
300 943 960 1005 1090 1175 1250 1330
400 975 994 1035 1115 1195 1265 1340
500 1010 1030 1140 1215 1283 1355
600 1050 1165 1235 1300 1370
800 1120 1220 1280 1335 1400

70
TABLE 9. Viscosity (10- 7 g cm- I sec-I) of hydrogen (Michels,Schipper,
and Rintoul)

25' C 50' C

p, anTI I I P. atrn I p, anTI I

26.43 894.1 803.8 1143 170.8 966.1
30.60 895.3 905.5 1184 204.6 974.7
34.41 897.5 905.6 1181 235.6 983.3
36.51 895.9 905.6 1184 235.7 980.7
36.52 897.l 1008.7 1225 293.1 996.3
44.91 897.7 1008.7 I 1224 344.0 1013
53.25 898.6 1080.3 1254 452.5 1045
150.3
174.3
915.1
922.2
10993
1201.3
I
I
1262
1304
552.0
652.5
1077
1114
i I I
198.9 927.8 1201.3 1303 652.3 1111
223.2 9350 12014 1305 752.5 1150
223.2 934.4 1300.2 I 1347 853.9 1187
269.8 946.9 1349.6 I 1367 9535 1225
325.5 963.0 1399.5 1386
430.5 999.0 1500.4 1433
430.5
478.7
995.0
1017
1500.4
1599.4 I 1434
1473
498.4 1022 1698.6
I
I 1516
601.9 1062 1799.0 1559
5963 1057 1852.8 1582
678.1 1094 1899.3 1602
6997 1102

75 0 C 100' C 125' C

p.anTI p. atm p. atm

I

43.71 995.3 189.2 1059 209.6 1108

53.09 9971 233.7 J070 2471 1115
61.52 997.7 2772 1079 285.0 1123
180.0 1012 277.3 1078 285.1 1122
221.0 1022 340.0 1093 355.6 1139
254.3 1031 401.2 1109 424.9 1156
2544 1031 501.3 1135 5015 1175
312.6 1044 • 601.7 1164 602.0 1200
370.5 1061 701.8 1194 702.2 1231
4524 1082 700.5 1192 789.8 1255
501.3 1098' 803.9 1226
601.8 1129 901.4 1259
702.0 1162
700.5 1161
803.9 1197
901.4 1230
1000.6 1265

71
 Deuterium. The viscosity of deuterium was determined by Michels,
Schipper and Rintoul /79/ at 25, 50 and 75°C, and at pressures up to 2000 atm,
by the capillary method (Table 10).

141)01-- '---f. i I I l--

~
I f.----
1,35,0 i I I
! 1250 --o- f.---- i
f.3O,'0
! ~ -I' i I l--I-- -----r-T
.. f2.'50
...- ...- r:1
I 1-~ '-i
~-j- Vl
! ;--
r ----Vt:::
~fZ,N
'eff5.0- ,-- ~
;E
i

~'--1!Q!L~vr-
~--
~
--- ~

" ~
';""I~00
~ vf--fr:r:::v~! I
I l-.-~~ L.-- H5!L-:"--~ i I
;:'10.'50~,....
1S
fOo,'Of--"
95.
°v
9[}''Or-
...1---~""-- ~ V
~....-"'...tI
j,~~
~
I
,
p~~_,_-=-!
I
tt 'I

' 1 •
I
I-Gibson (1933)"--
! !

o-Colubev and Petro v (1944)

'-,-, '-Kuss (1953)
I~- and
Michels, schi:f~erj
Rintoul ( 53
85.0
fOO 200 300
iii
700
I I
800
I I
gOO
WO 500 500
,0, ann

FIGURE 35. Viscosity of hydrogen according to different authors.

It WaS found that the ratio between the viscosities of hydrogen and
deuterium is given by "D = 1)H V2 in the temperature and pressure ranges
studied, provided the de~siti~s are the same. The accuracy of this equation
is ± 1% (Table 11). In the 'fj-P diagram for deuterium, the temperature
curves intersect at about 1600 atm, and at 2000 atm the viscosity decreases
slightly with increase in temperature (i. e., the compressed gas behaves
like a liquid).
TABLE 10. Viscosity (10- 7 g cm-! sec-I) of deuterium (Michels,
Schipper, and Rintoul)

25° C

p, atm I p, atm I I p, anll I

23m 1263 149,9 1296 894,5 1681

25.63 1263 169.9 1303 !IlOO.7 1745
31.37 1264 169.9 1302 1000.6 1744
31.17 1263 199.9 1314 1098.4 1806
32.02 1263 227.0 1327 1199.5 1867
37.23 1265 227.0 1.125 12901 1923
43.25 1266 iI 277.,5 1348 1359.1 1964
43.27 1268 324.8 1376 I 1359.0 1962
54,08 1269 324.8 1373 1459.1 2020
62,96 1271 399.7 1408 1556.6 2082
62.98 1272 459.7 i 1412 1658.5 2146
75,92 1274 459.7 1440 17591 2207
87,90 1276 499.6 1457 1809.1 I 2238
87.93 1275 609.7 1521 1859.0 2269
!
106.0 1282 6748 1559 1958.8 2329
121.9 1287 674.6 I
1558 !
121.9 1285 784.9 1618

72
 TABLE 10 (Continued)

p, atm p, atm I I p, aIm I

I II
25.70 1337 132A 1357 800.9 1657
31.44 1337 1581 1365 901.2 1714
34.78 1337 182.2 1374 il 1002.5 1771
'i
34.74 1336 182.2 1372 Ji 1100A 1824
35.02 1337 214.3 1384 I,'I 1100.6 1824
43.01 1337 245.3 1396 I, 1201.2 1883
49.01 1339 245.3 1394 I'·1 1292,4 1936
,
49.02 1340 300.6 1417 II 1382.0 1989
59.15 1341 355.4 1441 1461.7 2028
6887 1343 355.5 1438 !I 1461.0 2026
68.89 1344 430.6 1473 1559.1 2085
IIi'
84.08 1347 I;
5008 , 1505 "I,If 1661.9 2142
97.01 i 1349 5006 1502 I
17623 2109
97003 1348 6009 1557 ;'1 1853.1 2253
1105.1 1354 702.1 1607 19420 2303
II
1324 1359 II 701.0 1604 " 2023.0 2351
'I
i
II',
!

P, ann I p. arm I I p, ann I

I'

27.39 1407 II 170.1 1434 II 898.2 1743

32.21 1406 197.1 1442 1001.8 1797
37.41 1407 il 197.1 1440 1099.8 1853
Ii III,
37.40 1406 232.1 1453 120l.J 1905
I
4523 1407 265.1 1464 II 1200.8 1904
52.57 1409 265.1 1461 1291.6 1955
52.59 1410 323.2 1483 1381.2 2000
62.00 1411 380.2 1505
II 1471.0 2045
73.97 1413 380.1 1504 1560.5 2089
74.58 I 1413 457 ..') 1538 1559.0 2088
!!
90.09 1417 I 540.4 1573 1660.8 2138
104.7 II 540.4
I 1760.8
1419 "
1572 2195
104.7 1417 " GOO.5 1599 1850.7 2245
124.0 1422 700.5 1649 ! 1951.4 2301
142.1 1427 800.3 1697 2051.6 2357
142.1 1425 800.3 1693 2149.3 2415 _

I I II

Nitrogen. The viscosity of nitrogen was measured by Michels and

Gibson by the capillary method, at 25, 50, and 75°C, and up to 965.S atm
°
(Table 12). We measured the viscosity at to 250°C, and 1 to SOO atm
(Table 13, Figures 36 and 37). The results show good agreement.
The viscosity of nitrogen at various temperatures and pressures was
measured by Boyd /70/, but it appears that the determinations were not
sufficiently accurate because the data were very scattered. Boyd used a
rather wide capillary. For this reason the pressure gradient was very
small, so that the accuracy was probably low.

73
TABLE 11. Comparison of the viscosity coefficients (10- 7 g cm- ' sec-')
of deuterium, calculated from the equation 'In, = 'lH, and the -y2
observed values

Den-I 0

2S C [ 50 C
0

H, _I~' H, _I~-,-- ---H-,-_---.-D-,-

~ I~Y2i ~ ~ I~V'21 I~V21
-:r
1

909 '1~:11285 -:~355~,56 -~~:_1~423 1425

125 916, 1295 1295 964 1363 1:165 1012 11431 1434
150 923 11305 \.107' 972 1375 l:m 1020 1442 i 1445
175 933\1319 1322 981 1387 11:191 1030; 14571' 1458
200 942 1332 1338 991 1401' 1406 1040 1471 1472
225 953 1348 1356 1003 1418 1'122 1052 1488' 1488
250 965 1365 1375 1015 1435 1'140 1064 1500 1506
275 978 1383 1395 1029 1455 '11458 1077 1523 1526
300 994 1406 1417 1043 1475 1478 1092 1544 1547
325 1010 '1428 1440 1060 1499 '1501 1108 1567 1570
350 1028 i 1454 1464 1077 1523 , 1526 ,1126 1592 1.596
375 1047 1481 1490 1095 1549 11555 11145 1619 1624
400 1067 1509 1519 1116 1578 i 1584 ! 1166 1649 1652
425 1089 1540 1550 1138 1609 1614 I 1188 1680 1682
450 1112 1573 1582 I 1161 1642 1646 111211 1713 1713
475 1136 1607 1615 1186 1677 1681 11235 1747 1748
500
:~~~ :~~~ :~~~ 1211 1713 1719 " 1260 1782
1788
550
600 1280 1810 1823 1
650 1352 1912 19281 1
700 1432 2025 I 2039 I i
750 1521 1
2151 I, 2171 I
I _ _I_'----_
-------'-_--'----'--_-'---_L_ _ _ _

TABLE 12. ViSCOSity Cl 0 -7 g em -1 sec-I) of nitrogen (Michels and

Gibson)

25° C 50 0 C I 75° C

p, aIm I 1/ p, atml
1
p, ann I~.-I '
p, alln I ~
13.57 1799 II 277.7 2583 15 37 1913 1537 2022
15.36
48.22
1805 320.4
1864 II 364.5
2764 1 57.60
2941
I
, 1 1981
104 5 2088
57,61
1045
20[8
2175
I' ,I
. i 2373
57.59 1884 364.6 2948 1 212.4 212.4 2408
66.98 : 1908 387.6 I 3040 I 320.1 ! 2737 320.3 2746
1
92.31 'i 1981 430.2 3209 : 430.2 3129
1 430,2 3079
104,5 i2005 476.1 3385 541.7 3509 i 541.7 3426
118.6 \ 2049 54L7 3654 630.4 3786
I 630,8 3701
132.7 2094 587.7 1 3820 742,1 4165 742.0 4016
165,5 I 2199 630.4 3997 854.1 4550 I 854,1 4362
174,9 ! 2231 742,0 4404 i 965,8 4913 965,7 4703
189 2282 854.1 4802
212.4 2367 9657 520::;

74
 FIGURE 36. Viscosity of nitrogen as a function of pressure and
temperature.

4500
I

4'0'0'0
"" "''0 1
~I
i
i !
. . . . 1- t-- ' i
,
"
I
"''0 i
~ I
,. 350'0 ~ 1'- i I
Ei<.)
~~ J
be
';- 3000
o
rl
f':: b,

I...;;;: Q3001
~t7t7

I
1
--
_b-: +-- ~
l-

'" 250. i .... l--:::::: ~ ~ ~i

'0 1 ,200 ;::..- ~ ~ p
'OHi:~ ~~
~f?
p
15'00
o 50 100 150 200 250
!,"C
FIGURE 37. Viscosity of nitrogen as a function
of tern perature and pressure.

Ross and Brown /161/ measured the viscosity of nitrogen in the range
from +25 to-50"Catpressuresup to 680 atm (Table 13a). The results of
these authors at 0 and 25°C agree well with our data.
Oxygen. We measured the viscosity of oxygen with a setup of the third
deSign, at temperatures between 15 and 100°C, and pressures up to
152 atm, and with a setup of the fourth design at 15.8 and 54. 8°C, and up
to 700 atm. The values show good agreement (Table 14). For the same
pressures and temperatures the discrepancies do not exceed 10/0.

75
 TABLE 13. Viscosity (10- 7 g cm- I sec-I) of nitrogen (Golubev and
Petrov)

~,I 0 I I I I 1~- I I I
25 50 75 150 200 250

1665 177511880 1985 2090 2280 2460 2640

20 1695 1805 1905 2010 2115 2300 2480 2655
50 1760 1860! 1955 2050 21;'5 2340 2510 2680
100 ,1905 1985 iI 2060 2 -
14" 2230 2400 2565 2730
150 , 2090 12135 I 2180 I 2245 2325 2475 2630 2780
200 2300 2305 2325 I I 2370 2430, 2565 2700 2845
300 2750 ' 2680 I' 2615 I 2655 2675! 2760 2860 : 2980
400 3200 3075 2995 2965 i 2940 I 2970 3025 i 3120
500 3675 3460 3335 3260: 3220 : 3190 3210 i 3275
600 4125 3850 I 3670 : 3565 : 3490 I 3405 3390 13430
700 - 4225 I 3995 i 3865 ! 3765 'I' 3630 3580 3590
800 / -
, 4580 i 4325 ,
_ _ _ _ _ _ _~_ _, :
41 GO i 4030 I 3850 3710 3750

TABLE 13a. Viscosity (10- 7 g cm- 1 sec-I) of nitrogen (Ross and Brown)

~t,oCI + 25.0 0_0 - 2.\.1 - 49.8

p.atm~

35_02 1850 1727 1619 1527

69.05
I 1903 1825 1733 1671
103_07 2008 1901 1818 1769
137.1 2085 2035 1971 1934
205.15 I 2323 2308 2313 2329
273.2 2.564 2604 2627 2712
341.2 2860 2923 2981 308.3
409.23 3103 3207 3329 3473
545.4 3634 3808 4006 4228
681.5 4152 4422 4683 5000

TABLE 14. Viscosity (10- 7 g cm -\ sec-I) of oxygen (Golubev and

Petrey)

r~l,ocl
~I
p, aOn
15 50 100
p.at1n~
15.8
I
I
54.8

1993 2182 2437 1 1997.0 i 2207

22 I 2013 2211 2460 20 I 202~.0 I
2288
2055 2249 2490
50 2093_ 5 1
42 2434
100 2250.0
62 2114 2295 2525 150 2451.5 I 2610
82 2188 2345 2564 200 2688.51
102 2269 2401 2608 300 3220_0
122 2351 2463 26.15 400 3769_0 I
500
600
4330.0
4862.0
I
700 5392.0

Helium. Ross and Brown /161/ measured the viscosity of helium

between + 25 and - 50°C by the capillary method (Table 14 a).
Argon. The viscosity of argon was determined by Golubev and Petrov
from 0 to 200°C at pressures up to 400 atm (Table 15).

76
 TABLE 14a. Viscosity (10- 7 g em-I sec-I) of helium (Ross and Brown)

0.0 -24.8 -SO.I

69.05 1990 18:35 172~ 1601

137.1 1983 1~53 1734 1622
205.15 2007 1876 1758 1651
273.2 2015 1884 1781 1668
409.25 2042 1931 1839 1734
545.4 2103 1993 1880 1794
681.5 i 2154 2068 1971 1879
I I

TABLE 15. Viscosity (10- 7 g em-I sec-I) of argon (Golubev and

Petrov)

''-..~t,OC
p,atm~
I
----~----.-----~--~----.---­

15 25 5U 100 200

1 2095 2190 2250 2400 2695 3225

25 2180 2265 2320 2465 2745 3265
50
I
I 2290 2370 I 2425 2555 2815 3330
100
150
I 2585
2910
2640
2920
,
2680
2935
2780
3010
2995
3195
3440
3570
200 3265 3225 3225 3265 3390 3715
300 4030 3885 3830 3780 3800 4000
400 4865 4625 4510 4355 4220 4290
475 - i - - - - 4520
I I

Air. The viscosity of air at elevated temperatures was first determined

hy Ruckes /71j, However, Wildhagen /72/ showed that in Ruckes'
t~xperiments the flow in the capillary was turbulent. He therefore
redetermined the viscosity of air at pressures between 1 and 40 atm. He
found that the viscosity of air increases by only 4% in this range, while
~lccording to Ruckes the viscosity increase is as high as 20% between 1
cend 15 atm. Wild hagen measured the pressure on two spring gages, and
C'id not employ a differential pressure gage. Since the pressure drop in
the capillary was very low, it was hard to measure, so that the data of
Wildhagen can be considered approximate only. Nasini and Pastonesi /29/
1.; sed Rankine's viscosimeter (mercury- drop viscosimeter), and determined

t':.Je viscosity of air at l4°C and pressures up to 200 kg / cm 2 (Table 16).

Golubev measured the air viscosity on a setup of the first design between
° and 100°C and pressures up to 300 atm (Table 17 and Figure 38). Table 17
contains more accurate data than the earlier paper by Golubev /73/.
Our data are 3- 5% higher than the absolute values of Nasini and
Pastonesi. According to the data in the literature and of our experiments,
the viscosity of air at 14°C and 1 atm is equal to 1786 10- 7 g cm- I sec-I,
while according to N azini and Pastonesi, at the same temperature but at
34.8 kg/cm 2 it is 179510- 7 g cm-Isec- I . In other words, when the pressure
increases from 1 atm to 34.8 kg /cm 2 , the viscosity of air increases by less than
0,5%. This cannot be true, and proves that the absolute viscosity values
rneasured by Nasini and Pastonesi are too low.

77
 TABLE 16. Viscosity (10- 1 g cm- I sec-I) of air at 14°C
(Nasini and Pastones;)
-----------;c----
p. k g /cm
2
! ~ !I p, kg/cm
2
;1_ _ _ __

34.8 1794 117 1989

50 1820 124 2006
59 1840 154 2110
70 1856 170 2157
80.2 1905 181 2211
100.3 1950 200 2242
110.5 1987 -
I
TABLE 17. Viscosity (10- 7 g em-I sec-I) of atr (Golubev)

~!
p, atm
14
I
II) 25 50 90 100

1 1720 1786 1795 1837 1955 2135 2180

20 1816 1823 1980 2170
50 1815 1877 1885 1922 2032 2200 2240
100 1970 2018 2025 2060 2150 2298 2335
150 2165 2190 2195 2215 2280 2390 2420
200 2370 2382 2385 2395 2435 2510 2530
250 2605 2589 2590 2590 2600 2640 2650
300 2860 2820 2815 2800 2780 2800 2810

2900
2800
2700
r--
- 300

250
'u2600
_~2500
I

52400 r-- ~f-- -- I--" -

,)002300
I ..... I-- I-- ..-
V
o 150 l - I-- V ......
rl2200 ...- ~ ....... V ~ ........
p,s:....... ....
!:' 2100

2000 ...- ...... ...... 5,g, ....... V

1900 ....V
...... V
i-""
V
1800
1700
......
,....
0102030 4050607080SVf(l}
t,"C
FlGURE 38. Viscosity of air as 8. function of
temperature and pressure.

Kestin and Pilarczyk /80/ determined the viscosity of hydrogen, helium,

nitrogen, argon, and air, at 1 to 70 atm and 20 and 21°C, using the
oscillating disk method. The authors calculated the results by the Macwood
equations /55/. The measured data were described by empirical equations,
giving the results with an accuracy of± 0.1%. The data of /80/ for hydrogen,
nitrogen, argon, and air agree with our results.

78
 Carbon monoxide. We measured the viscosity of carbon monoxide at
temperatures of 0 to 250°C and pressures up to 800 atm (Table 18). A
comparison of these data with the viscosity coefficients of nitrogen shows
that the numerical values ofry of both gases are similar atthe same pressures
and temperatures. This is in agreement with the fact that the numerical
values of some other physical constants (molecular weight, critical
constants, density, specific heat, and others) of CO and N2 are also
similar, or differ only slightly, when the parameters of state (p and T) are
equal.

TABLE 18. Viscosity (10- 7 g cm -I sec -I) of carbon monoxide

(Golubevand Petrov)

I I I I
'-
p. ~I
ann'
0
I 25 50 100 150 200 250

1 1660 1765 1870 2075 2270 2450 2625

20 1690 1795 1900 2105 2290 2470 2645
50 1750 1850 1945 2145 2320 2500 2670
100 1895 1990 2050 2225 2385 2560 2715
150 2080 2140 2175 2320 2465 2630 2770
200 2300 2305 2315 2430 2550 2710 2830
300 2740 2680 2640 2680 2765 2870 2970
400 3175 3065 2985 2950 2970 3035 -
500 3600 3450 3330 3225 3185 3200 -
600 3825 3660 3490 3405 3400 -
800 4550
I 4300 4010
I 3820 3760 -

The reader should note that the carbon monoxide used by us contained
about 1'10 nitrogen and 0.2 to 0.30/0 hydrogen. The viscosity data given in
Table 18 therefore had to be corrected.

TABL~;::19. Viscosity (10- 7 g cm -I sec -I) of carbon dioxide (Phillips)

83 8230 110.5 7700 120 7880 114,5 6930 112 5710

72 7710 104 7350 112 7410 109 6600 108 5400
:)9 6970 96 6930 104 6950 96 5860 100 4830
~)6 1860 90 6430 93 6270 88 5110 94 4140
~)O 1770 82 5920 87 5860 85 4560 85 2690
40 1660 80 5650 84 5600 80 3610 80 2180
20 1560 76 5290 80 5280 75 2370 70 2000
1 1480 74 4950 76 4480 70 2140 60 1870
73 4780 75 4060 60 1780 40 1760
72 4580 74 2540 40 1740 23,8 1690
70 2290 70 2140 20 1630 1 1570
60 1870 60 1870 1 1560
40 1680 40 1750
20 1590 20 1620
1 1530 1 1550

79
 Carbon dioxide. The viscosity of carbon dioxide at elevated pressures
was measured by Phillips /28/ (at 20 to 40°C, 1 to 120 atm; Table 19).
Later Comings, Mayland, and Egly /27/ measured the viscosity at 40°C and
pressures up to 137 atm. In both cases the capillary method (Rankine
viscosimeter) was used. Stackelbeck /39/ measured the viscosity of CO 2
by the falling weight method at temperatures from - 15 to + 40°C and
pressures up to 120 atm. Golubev and Petrov used the second, fourth, and
fifth designs of the setup to measure the viscosity of CO 2 between 20 and
250°C and pressures up to 800 atm, i. e., the viscosity of liquid carbon
dioxide was also measured.

TABLE 20. Viscosity (10- 7 g cm- 1 sec- 1 ) of carbon dioxide (Golubev

and Petrov)

~I 20 1 1 1 1 1 1 1 1
30 40 60 9)
100 150 200 250
-

1 1 1
1 463 1510 1 560 1 650 179°11 82512045 [22542456
20 ! 1 560 1 590 1 630 - 1 820 1 855'2075122802485
50 1 850 1 830 1 805 - 1 900 1 95°12 140 2 3402 540
70 7320 2500 2 120 - 2030 2040,2 19023852590
80 7680 5310 2470 - 2110 21001222024202610
90 8 000 5 080 3 350 - 1 2200 2 160 2 250 12 450 2 640
100 8280 6500 4875 - 12300 224022901248012665
120 8820 7300 6030, - 2590 24401239°1'2540'2730
150 9520 1 8080 1 6950 4650 3120 2850 25602660 2810
200 1
10400 90901 8000: 6020 419° 38002980[2900,2980
1

1
250
1

11 170\ 9930, 8850' 6980 5130 4700:34803200'3180

300 11820 10610' 9590: 7810 5960 5470'3980135403380
1

400 12960 11780 10 800 9130 7300' 67801 4 980142603890

j

500 14000 12860 1185010 2301 8400 78805920:'49604440

600 15000 13850 12820 11 200: 9 370 88006760;56405020
700 148401372012110110200 96207530162805570
800 15600 14 GOO 12950']] 000 104008250'69006100
____---'------'1 1 _.1_,__

15'0'0'0
14000 ILV V V
V j../
13000 /' V
12000 ~~ V 3QV LV 1::-:::'
V

'uflOOO V V ItO V V
1;l
\
V /'
V
V V /"
-: 10'000
8g'Ooo
\
\ V /v jliY -- V r
to< \
\
1/ / i/ /V 100 V
'7 8000
/'/ / V /'" ~--+----

t// :/~fi-J
::: 7000
",'

6000
V iL k/' /. I~ V ,J:::::::- -----
5000

~V"~""-
4000
3000 i"
50" ~~: i I.
: ' . :
2000 r- .~-- ',-
1000 o '0 L 1 ____ • i
100 200 300 4()J 600 700 8m
Il,atm
FIGURE 39. Viscosity of carbon dioxide as a function of pressure
and temperature.

80
 2000 '50
1000 f attTI
o 20 W 60 8'0 10'0 /2'0 IW fOO /8'0 2fJ!} 22'0 240 Z60
t:C
FIGURE 40. Viscosity of carbon dioxide as a function of
temperature and pressure.

8{}(j,'0

750,'0 V
70U,' 0 - o-Golubev and Petrov
/
,- Phillips
.-Comings
650,'01--- 0- Stackelbeck
./
!
j
6'000

'0
I
5500,'0
/
I
OJ
'0

JOo,'0 I
25(],
j
IJ'
.I' I
2m'0
,.Y.'"
o."..~
1500 ,," • .-0' i I
25 5'0 75 1'0'0 125 15'0 175 2'0'0
,0, atm

FIGURE 41. Viscosity of carbon dioxide at 40·C according

to the data of different authors.

81
 Figure 41 gives a comparison of the viscosity coefficients of C02 at 40°C
and different pressures measured by different research workers. It can be
seen that, in general, the data obtained by the capillary method agree.
Ammonia. Golubev and Petrov used a setup of the fifth design to measure
the viscosity of liquid and gaseous ammonia from 30 to 250°C and at
pressures up to 800 atm (Table 21, Figures 42 and 43).

TABLE 21. Viscosity (10- 7 g em-I sec-I) of ammo"ia (Golubev and Petrov)

~t"C\
p,atm~
30 80 100 120 140 150 160 200 250

I 1040 1218 1287 1358 1430 1465 1498 1642 1816

20 13650 1240 1300 1376 14.'10 1484 1516 1660 1830
40 1330 1410 1483 1515 1545 1683 1850
50 13900 8400 1350 1430 1504 1540 1570 1700 1865
60 8520 1460 1530 1565 1595 1720 1880
70 8620 1560 I 1595 1625 1740 1895
80 8730 1590 1630 1655 1760 1910
100 14300 8920 , 6950 5100 1720 1705 1720 1805 1947
125
14650
9150
9380
7290
7600
5620
6040
2600
4000
1875
2800
I 1830
2100
1890
2000
2000
2060
150
200 15000 9820 8150 6700 5000 4225 3460 2330 2215
300 15800 10620 9000 7580 6200 5550 4980 3400 2700
400 16500 11360 9750 8320 698() 6475 5940 4360 3310
500 8950 7100 - 5100 3920
600 9500 7620 5670 4500
700 10030 8100 6200 5020
800 10520 8590 6690 5420

I
I 30' CVI--'
(WOO
1(,
/3000 I
I
12000 I

---
I
11000
I
7,,10000 --t-I SO V-- ----
'" . /V IQR V I-- I--
- --
00 I
I
7 9000
§ 8000
'V VV 120 1-- I-- l - I--
.,""7000
::: liOOO
"'""5000
4000
'('
I\ /
I
Y
Y
/vr
V,---
YI- V
J~~I--_--
V /1 >/-1~V:l ------
--
i!&- I - IT£L f.--'-- 1-
-- -
I-

3000
I \ /1/ //,,/ ~I J_ 1
2000 . lldy/ ~- ~±J~--~
1000 ~
(DO 200
I 300 400 :~""
5170 600
1
700
~

800
p,atm

FIGURE 42. Viscosity of ammonia as a function of pressure and

temperature.

82
 ~\
~
\~~\\
ISOrA
IFOo.'0 ~ ~\
\~ l\\\
1100.'0
\~ \\
, l'OOrA
_~
u
Q)
9'00.'0
\~ ,,\1\ \ !

'6u 8'00.'0

';" be 7000
\~
\~t\.
l\\ ~
0: ~ .'" ""~
\~ \\ \~ ~!0:~0.1
!
I

o I
~6000
5'00.'0
\:\\ \\\ '\ ~ ~ N'---~
t,{/(}O
'\\\ \ 1"- .......... ~ r--
3000
\\ \ \ ~ ""'-.
~ '--- t---
r--
2000 '<:~oo~ 15a ----
1000 -- -- --- /atm -
20 40 60 80 100 /20 MO 160 /8'0 200 F20 240 250
I, 'C

FIGURE 43. Viscosity of ammonia as a function of

temperature and pressure.

The viscosi.ty of ammonia was also determined by Stackelbeck /39/ by

the falling weight method between 0 and BO°C and pressures up to 22 atm.
However, his data do not agree with ours; they are more pressure-dependent.
This is, apparently, due to the drawbacks of Stackelbeck' s method (see
Chapter II).

TABLE 22. Viscosity (:LO-

1
g cm- I sec-I) of methane (Meshcheryakov and Golubev)

~t,oCI
p,atm~
-15 18 25 50 100 150 200 I 250
-
976 1027 1080 1108 1182 1260 1331 1471 1603 1725
10 991 1040 1090 1120 1193 1270 1340 1477 1611 1728
20 1013 1062 1114 1137 1210 1280 1350 1483
I 1617 1733
40 1070 I,
1115 1153 1180 1247 1312 1377 1512 1634 1745
60 1150 1180 1212 1237 1295 1350 1415 1530 1653 1765
80 1250 1267 1285 1310 1352 1392 1452 1556 1675 1785
100 1387 1375 1370 1395 1417 1445 1497 1580 1700 1809
150 1845 1742 1670 1652 1610 1600 I 1627 1685 1775 1872
200 2255 2100 1965 1930 1830 1787 1787 1807 1867 1947
2617 2422 2235 2187 2002
, 1975 2030
250 2062 1965 1942
300 2937 2715 2515 2455 2300 2205 2142 2085 2075 2120
350 3227 2930 2718 2705 2530 2401 2320 2230 2182 2205
400 3492 3232 3017 2942 2745 2593 2486 2370 2290 2295
500 3955 3690 3445 3365 3145 2955 2815 2635 2500 2470
600 4380 4095 3792 3740 , 3485 3277 3120 2887 2705 2645
700 4782 4470 4170 4087 3802 3545 3395 3130 2910 2820
800 5155 4817 I 4507 4410 4095 3855 3655 3367 3107 2995
! I
I
I

83
 Methane. The viscosity of methane was determined by Meshcheryakov
and Golubev /158/ at temperatures from -15 to 250°C, and pressures up to
800 atm (Table 22 and Figures 44 and 45), by Comings, Mayland, and Egly
/27/ from 30 to 95°C and pressures up to 171 atm, and by Kuss /74/. These
authors used the capillary method. The falling weight method was employed
by Bicher and Katz /45/ at temperatures from 25 to 225°C and pressures
up to 340 atm.

,
S2~Or--+--+-~~~~~~~
.;.
n~~-+--~~~~~+--1--~-4

o 100 200 300 400 500 600 100 800 100 250
p.atm t:C

FIGURE 44. Viscosity of methane as a function FIGURE 45. Viscosity of methane as

of pressure and temperature. a function of temperature and pressure.

In Table 23 the results of the individual authors are compared. The data
of the different authors obtained by the capillary method show good
agreement, but the data of the falling weight method are too high (the
discrepancy is sometimes as high as 150/0).
The viscosity of methane was also determined by Sage and Lacey /41/,
from 37 to 104°C and up to 140 atm. The authors used the rolling ball
method. The results were plotted on a small graph, and so it is difficult
to determine the viscosity coefficients with sufficient accuracy.
In their experiments with methane and other gases Sage and Lacey found
that eddies were frequently formed in the gap between the rolling ball and
the tube walls over which the ball rolls. The same phenomenon was also
detected by other workers. Therefore, when the viscosimeters were
calibrated, operation in the turbulent region was taken into account.
However, Comings, Mayland, and Egly showed that the principle of
calibrating the viscosimeter in the turbulent region is incorrect, and
therefore the results cannot be accurate. In the research of Sage and
Lacey, and of Bicher and Katz, eddie" were formed in the gaps at pressures
only a little higher than 1 atm. Thus, this appears the main reason for the
high results obtained by the method.
Ethane. The viscosity of ethane was determined by Meshcheryakov and
Golubev /158/ in the temperature range from - 15 to 250°C and at pressures
up to 800 atm by the capillary method (Table 24, and Figure 46), and by
Smith and Brown /44/ at temperatures from 15 to 200°C and pressures up
to 340 atm by the rolling ball method. As in the case of methane, the

84
differences between our data and those of Smith and Brown are as high as
15'70 at some points.

TABLE 23. Viscosity (10- 7 g cm- I sec-I) of methane according to the data of different authors

Authors \x\ 1 I 20 / 40 / 60 / 80 /100 /120 1135/150/200 /300 /400 /600

Meshcheryakov and Golubev
Comings, Mayland, and Egly
I 25 1108 1137 11180
1089 1129 1177
1237
1233
1310
1307
1395 1490 1567 11652 11930 2455 2942 3740
1393 1489 1568 - - - - -
Bicher and Katz - 1165 1237 1326 1443 1571 1703 - 1903 2212 2661 - -
Kuss I 1093 1098 1125 1182 1250 1323 1412,1480 1545 1797 2277 273013626
1
Meshcheryakov and Golubev 50 118211210 124711295 1352 141711490115471161011820230027453485
Comings, Mayland, and Egly 1171 1202 1243 1296 1360 143.1 1516,15821- - - - -
Bicher and Katz - 12401301 13751471 1577 1683 - 1847121172556 - -
Kuss 117811183 1207i 1347 1305 1374: 1452,1513i158111786 2217 2616 3416

Meshcheryakov and Golubev 75 1260 128011312 1350 1 "" H<I 1495 1545 16<J(J 17"'1"0512593 3277
Comings, Mayland, and Egly 1271 1296 1329 1374 1420 1479 15461601 - - - - -
Bicher and Katz - 1315 1365 1427 1500 1584 1673 - 18132050.2464 - -
Kuss 1259 126511287 1325 1372 1435 150015471601178212162:25353257

Meshcheryakov and Golubev 100 133111350113771141511145211497/1545111585/162711787121421-1-

Comings, Mayland, and Egly 1~lml~I~I~I~I~I~- - - - -
Bicher and Katz - 137411423 147311533 159J 16801-118042010 2400 - -
1
Meshcheryakov and Golubev 200 1603/1617/1634! 16531167511700/1730/_/1775/11867/2075/-/-
Bicher and Katz - 1631 1665 1702; 174017801821 - 188719972215 - -
1 1

TABLE ~~4. Viscosity (10- 7 g cm -1 sec -1) of ethane (Meshcheryakov and Golubev)

~t,OCI
•
p,atm~
-15 I o 1 21 I 31.5 1 39.5 1 60 1 100 /150 /200 /250

1 806 850 914 945 969 1 030/1 144/1 278 1410 1530
.0 6830 1067 1020 1006 1016 1 070 1 170 1 295 1 417 1540
10 6970 5525 1168 1 108 1071 1105 1201 1318 1422 1550
10 7110 5740 4090 1340 1 229 1165 1239 1 349 1437 1555
iO 7250 5940 4409 3090 1560 1270 1305 1385 1460 1565
iO 7380 6090 4650 3709 ~
1 480 1 385 1429 1485 1575
0 7500 6250 4852 4078 - 1860 1 1500 1487 1508 1585
10 7630 6400 5035 4360 3800 2320 16401 1 553 1 547 1615
H'0
125
150
7880
8170 i
8460
6675
6978
7290
5371 4785
5711 5165
6025 5487
4 245
4 090
5 060
i
3 190 1 980' 1 714 1 660 1 680
3 775 2451 I 95511 775 1 770
, 4 170 2916 2 250 I 930 I 865
2C10
3C10
9000 : 7870 6617 6049 5670 i 4 810 36741 2815 2288 2080
10 e60: 8890 7609 7020 6 665 15 840 4 690 3 725\3 025 2 575
4C10 II 050 9820 8487 7903 7560 6 700 5488 4 470/3 660i 3 100
5C 10 12000 10700 9310 8720 8390 7480 6190 5 165 4270, 3610
6C 10 12900 II 530 10 070 9470 9 110 8150 16800 5735,484914090
7C 10 13720 12300 10790 10 195 9790 884017400 62801533514515
8C10 14500 13060 11496 10 833 10400 9 400 8020 6 790 5 794 5 010
1

Propane. The viscosity of propane was determined by Golubev and

Petrov /37/ between 24.5 and 250 a C and pressures up to 800 atm (Table 25
and Figure 47), and by Comings, Mayland, and Egly /27/ from 30 to 104. 5°e
and pressures up to 41.8 atm (Table 26). The capillary method was used in

85
both cases. The viscosity of propane was also determined by the rolling
ball method over a wide temperature range by Sage and Lacey /41/, Bicher
and Katz /45/, and Smith and Brown /44/.

';' 12000
,;,~10000'~~~-+~~~~~-~~~~~

100 200 300 fro 5170 8ro 700 800

4atm
FIGURE 46. Viscosity of ethane as a function of
pressure and temperature.

A comparison of the data of the different authors (Figure 48) shows that
the results of the capillary method agree well, but there are large
discrepancies between those of the rolling ball method.

p, atm

FIGURE 47. Viscosity of propane as a function of pressure

and temperature.

86
 TABLE 25. Viscosity (10- 7 g cm -1 sec-I) of propane (Golubev and
Petrov)

:~I 245 I I I
52.5 92.7 100.sl 125 11 151 201 I 250

1 8111' 885/ 98711 I 0081 I 0671 1 131 - I 252/1 362

20
35
40
9670,
9950, 7~80.
10050 7S80! -
I
=, I 070 I 12°1 I 175
I 270 1 I 2401 I 260
1
11510 1305 1310
I 285 1 390
I 340 1I 425
1375!1440
45 10150 7680 i 2400 1415' 1365, 1410' 1 460
50 I 10 240 7780 4440! 32501 I 600 1 435' I 450! 1 485
60 : 10 420 7960 4770 41601 2300i I 660' 1 535! I 540
75 10700 8230 5200' 4750\ 3 160 , 2 125 I 735, 1 650
100 II 120 8700 5850' 5450: 4140:' 29651 2 170' 1 900
150 11 920' 9 500 6 840 6 440' 5 320 4 370: ' 3 130' 2 500
'
200 12 640 10 200 7 580 1 7 170 ', 6 160 5 250!
1
3 960: 3 040
300 1113980 11520 I 88701 84501 7440! 6540! 5160, 4060
400 15260 12760: 10050' 9630, 8600' 76601 620015040
500 16530 13980 11200 10 780 9670: 8700 7120' 5880
'
600 . 17770 15170' 12280111 850! 10 6901 9640 , 7940 ' 6600
700 /19040, 16340113280,12840: II 600, 10 510, 8720 7260
800 20300: 17430: 14270: 13800; 12500. II 360: 9500',7930

TABLE 26. Viscosity (10- 7 g cm- I sec-I) of propane (Comings, Mayland,

and Egly)

''', I 'C 1
.~at~~__3_0__~_37_.7__~__54_.4__~ il.l 87.7 104.5

LOO 803 828 885 932 977 1023

4.40 807 831 888 934 978 1023
7.80 811 836 897 946 988 10:36
9.50 814
IL2 845 906 954
14.6 915 960 1009 1054
18.0 932 985
21.4 1007 1075 1084
28.3 1192 1152
35.0 1248
4L8 1569

n-Butane and isobutane. The viscosity of n-butane and isobutane was

measured by Sage, Vale, and Lacey /75/. Though these data are only
approximate, we report them here because they are the only ones available
for these substances (Table 27).
Ethylene. Golubev and Petrov /37/ used the capillary method to determine the
viscosity of ethylene at temperatures from 24 to 150°C and pressures up to
800 atm (Table 28). Comings, Mayland and Egly /27/ used the temperature range
from 30 to 95°C, and pressures up to 171 atm (Table 29). The oscillating disk
method was used by Gonikberg and Vereshchagin /56/ at 24°C and pressures
from 100 to 1000 atm.
A comparison of the data is shown in Figure 49. There is good agreement
between the results obtained by the capillary method, but there are
discrepancies between the data of Gonikberg and Vereshchagin and those of
other authors. This indicates that the oscillating disk method has apparently
not yet been sufficiently adapted to measurements in the high-pressure
range. When Fie disk oscillates, the gas moving with the disk gives rise to
eddies, especially at the instant the direction of oscillation changes. With
increase in the density of the gas, the probability of eddy formation increases.

87
 Of/OO
7000 ./' ./'
5000
/ //)' Vo
15000
A'? V /
WOO J A
~'\;
,~
0

_~ 5000
l1 / ./
,
8 3000 / / "
/
OJ)

'7 4000
/
o
10 b
/ ~o i
~ 2000
3000
fa :'10/ / -Golubev and Petroy
I
"?! I i
<;3 -Comings, ;\1ayland
1000 and EgJy

2000 /1 ,- Sage and Lacey

•. Eicher and Katz

~
V at 125°C
o· Brown and Smith
at 150°C
1000 I I I
I 50 100 150 200 250 3X 350
fl, atm
FIGURE 48. Viscosity of propane according to the data of different authors.

TABLE 27. Viscosity (10- 7 g cm-! sec-I) of gaseous andliquidn-butane

and isobutane (Sage, Vale and Lacey)

n-Butane

~I,ocl 37.78 54.44 71.11 87.78 104.44

p,atm~

1.0 784 826 869 910 947

1.36 787 831 874. 914 953
2.72 821 853 891 933 975
4.08 882 917 958 998
6.80 98:1 1010 1040
10.21 1120 1130
13.61 14860 12480 10340 8360 1250
27.22 15060 12720 105S0 8620
40.83 15300 12950 10810 8870
68.05 15650 13320 11200 9300
102.10 16080 13700 11 650 9790
136.10 16370 13980 11 90~ 10090
Sat yap 838 937 1030 1210 1480
Sat liq 14700 12320 10240 8340

Isobutane

1.0 765 795 825 857 887

1.36 775 802 831 866 898
2.72 792 818 852 887 921
4.08 821 842 873 911 942
6.80 898 925 957 994
10.21 1010 1030 1060
13.61 13440 10710 8960 1170 1140
17.01 13760 1310
27.22 14040 11020 9280 7800
40.83 14560 11320 9560 8110
68.05 15060 11 860 10110 8620
102.10 15540 12520 10650 9130
136.10 12970 11060 I 9540
Sat yap 841 913 I 03_8I 1 344
___S_a_t_li~q__~_1~3_2_00__~_1_0_5_80__~_8~900 ,___7~5~30~~_____

5680 88
 1f00LI
10000
",-
~
-- ---
f--
9000

;-,8000 /
;.-.- /
-- <g£V .... .-- f--
---
u
I--

~
1;;7000
V-
v sQ.y :..--
~6000 V -
()

./
V ._Golubev and Petrov •
.,bC5000
;0;4000
,/' /,/'" --- at 24' and 50°C
,_Gonikberg and Vereshehagi n.
at 24°C

~ V!
I$-
3000 I I .-Comings. Mayland. and Egi y.
!/ at 50'C
2000
1000",L...-
o.1
100 200 300 400 500 600 700 800
fl,atm
FIGURE: 49. Viscosity of ethylene according to the data of different authors.

1 ABLE 28. Viscosity (10 -7 g em -I sec -I) of ethylene (Golubev and

Petrov)

'-......
p,~
I 'C I
-----.-------.------~------,--------

24 50 100 150

1 1023 1106 1260 1400

25 1096 1175 1290 1470
50 1410 1310 1395 1560
60 1813 1400 - 1600
70 2820 1560 1500 1640
80 3357 1760 1575 1685
90 3667 2004 - 1730
100 3906 2420 1734 1780
125 4420 3010 2040 1950
150 4885 3640 2380 2150
200 5524 4380 3027 2600
300 6520 5380 4095 3415
400 7323 6180 4863 4150
500 8090 6890 5540 4770
600 8820 7560 6147 5330
700 9440 8140 6713 5860
800 I 9956 8670 7255 6360

TABLE 29. Viscosity (10- 7 g cm- I sec-I) of ethylene (Comings.

Mayland. and Egly)

~I,ocl 30 40 50 70 95
p,atm~
1.00 1054 1074 1093 1172 1255
5.08 1060 1080 1099 1180 1258
28.'2 1131 1150 1192 1243 1295
48.6 1299 1306 1305 1336 1389
;;5.4 1406 1369 1362 1381 1426
67.3 1775 1592 1505 1474 1498
79.1 2545 2003 1721 1606 1589
89.3 3047 2483 1975 1736 1681
103.1 3592 2997 2396 1919 1818
1,,7.0 4347 -- 3203 2558 2265
11"1.0 - -- I - - I 2773

89
 It is also worth noting that when Gonikberg and Vereshchagin calibrated
their viscosimeter for high pressures with nitrogen, the viscosity of which
had already been measured, they did not determine the conditions of
application of the calibration results to other gases.

TABLE 30. Viscosity (10- 7 g cm- 1 sec-I), of propylene (Golubev and

Perrov)

1 839 942 1032 1075 1212 1340 1468

25 10680 - - 1192 1312 1423 1530
50 11160 7780 5140 2805 1550 1547 1660
75 11620 8280 5880 4970 I 2400 I 1853 1820
100 12050 8700 6400 5500 3405 ' 2294 2060
125 I 12480 9080 6870 6030 I 4100 2782 2350
150 112880 944017280 6470 4600 3260 2695
200 13660 10130 8030 7240 5440 4085 3380
300 15100 11 460 I 9350 8520 6740 5330 4465
400 16400 12700 10460 9650 7850 6350 5420
500 17700 13900 11500 10680 8770 7270 I 6300
600 18940 15000 12500 11620 9670 8055 7050
700 20100 16100 13460 12500 10480 8805 1
7740
800 21300 17100, 14400 13370 11250 9500 8450

Propylene, The viscosity of propylene was determined by Golubev and

Petrov /37/ at temperatures from 18 to 250°C and pressures up to 800 atm
(Table 30 and Figure 50),

220.ViJ
\
200(J,

18'00.V '\ I
J\
,
'0
\~\ ~'\ ~
~~~~ ~
'" <'It{]]
812flJ.'0

';"oIOD(J,V
~~~~ f*:" '-..., ........
rl

:::::
~ ~ t' ';J$- ~: r--..... .............:--.

---
'0
~ ~ ~ r--.... ......... r=::: --::::: :::::::-
.............
6000

'tDa'0
I\,~ ~. ~ .............. ::::
\~ rit. Rr-
S:: ~ ~
2D(J,'0 lanTI I
'0 I -'----
'0 5'0 1'0'0 I.'j{} 20'0 250
t:C
FIGURE 50, Viscosity of propylene as a function of tempera-
ture and pressure.

90
 Alcohols. Golubev and Petrov /37/ determined the viscosity of four
alcohols at pre;;sures from 1 to 800 atm (for ethanol the maximum
pressure was 1000 atm at some temperatures) and at the following
temperatures: methanol 150 to 270°C (Table 31); ethanol 0 to 265°C
(Table 32); isopropanol 155 to 260°C (Table 33); and isobutanol 220 to
272°C (Table :14).

TABLE 31. Viscosity (10 -1 g em -1 sec -1) of methanol (Golubev and

Petrov)

150 200 231 250 270

1 1393 1563 1660 1720 1790

25 - - - 1758 1817
50 14830 8760 - 1796 1849
75 15240 9170 5800 1978 1916
100 15620 9580 6450 3680 2200
125 15940 9990 7100 5200 3400
150 16320 10370 7620 6100 4650
200 17000 11080 8425 7100 5800
300 18200 12280 9670 8435
I
I 7235
400 19310 13330 10 no 9435 8275
500 20330 14280 11630 10370 9170
600 21360 15150 12450 11 135 9 960
700 22360 15000 13240 11855 10565
800 23290 16850 13980 12600 11340
I

r ABLE 32. Viscosity (10 -7 g em -1 sec -1) of ethanol (Go1ubev and Petrov)
~t,OCI 12.2 30 50 75 151 200 235 251 265
p,atm~

,
1 180000 140000 98500 69000 45500 1235 1376 1473 1517 1553
20 - - - - - - - - 1540 1568
50 - - - - -- 18000 9910 - 1640 1650
70 - - - - - 18500 10400 6190 2540 1970
80 - - - - - 18700 10670 6610 4480 2640
90 - - - - - 19000 10 900 6940 5300 3650
100 189000 147000 103000 72 200 48900 19200 11 130 7240 5750 4450
120 - - - - - 19650 11 600 7780 6360 5330
150 - - - - - 20300 12220 8410 7110 6120
200 198000 154 000 108 000 75600 51400 21450 13140 9330 8130 7140
300 1207000 161500 113000 79500 54100 23400 14800 10880 9680 8760
400 216000 169 000 118500 83700 56900 25200 16300 12230 11000 10050
500 226000 176500 124000 88000 59800 27000 17600 13520 12230 I 11 200
600 236500 185 000 130000 92000 63000 28650 18850 14700 13380 12300
800 257500 201500 141500 - 69700 31950 21300 16900 15300 14200
1000 279500 218500 154000 - 76800 - - - - -

----------- I ,I I

91
 No measurements of the viscosity of alcohol vapors have been reported,
but some data have been published on the viscosity of liquid alcohol. Faust
/76/ determined the viscosity of ethanol at different temperatures and
pressures between 1 and 3000 kg / cm 2 • The pressure dependence (up to
12,000 kg / cm 2 ) of the viscosity of 43 liquids (including the alcohols studied
by us) was determined by Bridgman /40/ at 30 and 75°C.

TABLE 33. Viscosity (10- 7 g cm- 1 sec- 1 ) of isopropanol (Go1ubev and

Petrov)

~-I
atm
P. .....
155 200 230 240 260

1 1127 1240 1320 1348 1400

25 18200 1430 1442
50 19190 10790 6260 1808 1664
75 20020 11570 7826 6270 4030
100 20880 12,320 8736 7592 5690
125 21680 13000 9516 8424 6630
150 22460 13680 10170 9152 7410
200 23970 14870 11 280 10300 8,581
300 26780 16900 13130 12090 10400
400 29510 18880 14850 13750 12060
500 32150 20700 15450 15370 13570
600 34580 22 440 180)0 16850 15000
700 36920 24180 19450 18300 16330
800 39260 25870 20£0) I 19710 17630

TABLE34. Viscosity (10- 7 g C111- 1 SeC·-I) ofisobutano1

(Golubev and Petrov)

~I
atm
p. \
220 250 272

20
1
I 1220 1289 1340
1370
40 12440 7465 1660
50 13040 8158 6350
80 13640 8820 7000
100 14210 935b 7605
150 15530 10(20 8850
200 167(0 116'JO 9970
300 19100 13390 11780
400 212,,0 14960 13300
600 2,) 110 18 UO 16100
801 28540 21200 18800

The data of the above authors show considerable discrepancies. For

example, the viscosity of ethanol changes at a temperature of 30°C by a
factor of 2.94 according to Faust, and hy a factor of 2.31 according to
Bridgman, when the pressure increases from 1 to 3000 kg / cm 2 • Thus,
the agreem.ent is not satisfactory.
For this reason Golubev and Petrov checked these data by the capillary
method (the data of Faust were obtained by the capillary method, and those
of Bridgman by the falling weight method). The viscosity of ethanol was
measured at corresponding temperatures and pressures up to 1000 atm.

92
 Table 35 shows the data of Golubev and Petrov, Bridgman, and Faust.
Our data agree with those of Bridgman. It is interesting to note that the
same results were obtained by different methods, and different results by
the same method. Usually the opposite is true in measurements of gases.
We believe that it is possible to use the falling weight method to
determine the viscosities of liquids at high pressures. It should be noted,
however, that the measurements of Faust were carried out as long ago as
1914, and the author did not introduce all the corrections necessary when
the capillary method was applied.

T ABLE 3~;. Changes in the viscosity of ethanol with increase in pressure from 1
to 1000 kg/cm2

Authors
~ 0 12.2
I 30 75

Golube" and Petrov • 1.55 1.56 1.565 1.65

Bridgman - - 1.584 1.642
Faust •• .. - - 1.8 -
--

The viscosity of ethanol (at 1 atm) changes with temperature, and when
':he temperature increases from 30 to 75°C, the viscosity increases by a
:'actor of 2.12, 2.14, and 2.20, according to Golubev and Petrov, Thorpe
and Rogers /77/, and Bridgman, respectively.

8.'000
i .i! J~

()
--Golubev and Petrov
~~6'a!?Or- • - Schroer and Becker
.
E
uw
M
.. !..--
000
x
V
l'

'00
;<
/
I"'"
l,/
.-~ V
170
m M ~ W ~ m m w ~ m
jJ.atm

FIGURE 51. Viscosity of diethyl ether according to the

data of different authors.

Ethers. Golubev and Petrov /37/ determined the viscosity of dimethyl

ether at temperatures between 100.5 and 251°C, and the viscosity of diethyl
e':her between 135.5 and 249°C, at pressures up to 800 atm (Tables 36 and 37)
Data are given by Schroer and Becker /78/ on the viscosity of
d::ethyl ether at temperatures between 184.5 and 235°C and pressures
b,,,tween 2 5 and 7 5 atm. The authors used the falling weight method.
Ir Figure 51 the data of Schroer and Becker /78/ are compared with ours,
and considerable discrepancies Can be seen.

93
 TAGLE 36. Viscosity (10- 7 g em-I sec-I) of di:Tlethyl ether (Golubev and
Petrov)

~t.'cl 100.5 127 133 151 202 251

p.atm~:

1 1170 1248 1272 1319 1468 1610

20 1235 1308 I 32~ 1345 1497 1614
30 1317 1364 1370 1370 1525 1624
40 6260 i 1488 1463 1435 1566 1634
50 6480 1937 1668 1575 1615 1643
60 6680 4110 2640 1876 1705 1665
70 6866 4710 4000 2463 1813 1690
80 7037 5124 4540 :\ 215 I 944 1730
100 7378 51'80 5165 4176 I 2294 1866
125 7750 6 14~ 5713 4850 2883 2170
150 8090 6525 6138 Ei 316 3447 2542
200 8710 7207 6835 G 098 4309 3224
300 9796 {) 348 7980 7270 5543 4370
400 10790 9360 8990 I; 250 6516 5300
500 11 655 10290 9920 'l14C, 7378 5950
600 12524 11 130 10757 !) 960 8153
700 13330 11520 10700 8850
800 14120 12260 11440 9508

TABLE 37. Viscosity (10- 7 g cm- 1 sec-I) of diethyl ether (Golubev and
Petrov)

~t.CI
p. atr~~~1 13S.O IS, 203 249

----- --~-- --~--

1015 1 lJIi 1 ISO 1234 1292

20 12-10 1256 ! 1302
30 8400 1 3% 1317 1333
40 8600 4430 1093 1535 1448
50 88(1) 5068 ' 3720 1984 1668
CO 9011) 5465 4386 2790 1990
70 9210 5797 4827 3627 2427
80 9 -100 (i 107 5190 4110 2945
100 9765 G 603 57('6 4772 3760
125 10200 7160 6324 5394 4464
150 10600 7626 61)20 5920 5020
200 113BO 8400 7638 6758 5920
300 12770 9734 8 ~:68 8144 7285
400 14140 11000 10200 9368 8494
500 15500 12150 11320 10510 9548
600 10580 13210 12370 11500 10510
800 18 ~HO 15220 14370 133EO 12240

If we extrapolate the data of Schroer and Becker to atmospheric pressure

at 225°C, the viscosity coefficient will be at least 2000 X 10-7 g cm- l sec- l •
Our measurements, on the other hand, and those of other authors gave the
viscosity of diethyl ether at 1 atm as 1234>< 10- 7 g cm- I sec-I. Thus, the

94
data of Schroer and Becker are apparently only approximate. For C02 also
Schroer and Becker obtained values which were too high,· and this again
confirms that the falling weight method cannot be used to determine the
viscosity of gases.

TABLE 38. Viscosity (10 -7 g em -1 sec -1) of acetone (Golubev and

Petrov)

-..P.. . . ~. ocl
atm ___________
203 231 240 203.5

I 1220 1290 1315 1375

15 - - 1334 1390
30 - - 1400 1428
45 - - 1770 1612
60 - 5688 4930 2496
75 - 6225 5625 4234
100 9164 6984 6415 5310
150 10110 8010 7365 6540
200 10 900 8850 8340 7460
300 12320 10 240 9700 8850
400 13620 11 380 10850 9955
500 14820 1248) 11 940 11 000
600 15860 i 13490 12920 11 940
800 17850 15360 14600 -
I I

TABLE 39.' Viscosity (10 -7 g em -1 sec -1) of benzene (Golubev and

PE'tro._v~)_ _;--_ _ _---,._ _ _ _ _-.,-_ _ _ _, _ _ _ __

---....... . t 'C I
~,a~
95 153 202 248

1 938 1087 1215 1330

10 27600 16300 1220 1335
20 27850 16550 11500 1340
50 28500 17350 12100 9750
100 29600 18500 13100 II 000
:100 32100 20650 15000 12900
300 34750 22550 16600 14500
400 37500 24400 18100 16000
500 40500 I 26000 19500 17300
650 44400 28600 21400 19200
I

Golubev and Petrov also determined the viscosity of acetone at tempera-

tures between 203 and 263.5°C, and at pressures up to 800 atm (Table 38),
and of benzene at 95 to 248°C and at pressures up to 650 atm (Table 39).
Figures 52- 54 show the viscosity of some substances as a function of
temperature and pressure.

95
 24'wu,
I-- ~
22.(?"v: C-
2!AWO,, I -~
!§Q.'9-- 1-
~
I
18.'ODD,
~/b:000 1
, V f-- ~
V ~
,. 14IJI],'0 ' / ,, I -~
6f.?'000 '
~r- f-' ~I -
I-
--
--
~ iw.. f- l25ffi l -l -I-- lI---
-
I--I -
0(

';"" lfA'000
I
V
~~800.V
\V f.--" ~ I-- ii1!l f..-- l -
V f-' I
>$
60.'00
\ VV
<; V V I i
40.'00
lJ~--)}. V
2fA'{}O ~
"
I
0 I I
100 2'0'0 300 I;(}O seD 60'0 70'0 80'0
fJ,atm
FIGURE 52. Viscosity of methanol.

28'0'0'0
260'0'0
1 ,/ "
I V
1 V
,,
.--
21;(}fA -
ISl'y
~ 22000 I
/
'~2UOfA'{}' ,, ,
I

: i /
$!t#~:::t-- ---
/ I
,. 18'0'0'0
616000
Ir
'.-- ,.-
-
I
btl
';"" 1.'0'0'0
I
,, i
---r--,
-l--
~-: ~
I-~

~/2DOD , / " l----

>$
1'0'0'0'0
\ / ,.-,"-..;.---::::::t:"- ~ !

8'0'0'0
I \, V,.--:::::r---r I i I '
Qj
5'0'0'0
WO'O
i \/./V,
\(V i -H-1 : i
!
I

2808;~65:JJ/ I
8 rr 1'0'0
I
2'0'0 300 I;(}'O 5'0'0 600 7'0'0 80'0
p,atm
FIGURE 53. Viscosity of ethanol.

15'000
IMflJ
V
13'0'0'0 V
, I ..-,-V::
/
12[)J(J
1180'0
i I /' v-: ~ V ,.-
7uf(}{Q} 100•5~ 3!l.J. --;:; /- V ~
<J)

"i~gOOO V 121 ~ / ' /

V
....-
88'000 V ~ -sI / V V
t 7'OOO
/ ~V V .... /
~6arJ
V VV ~/ I
>$ // V V 25);. ,.- ....' ! i
5000 ./
1(/ /' /, , I
woo V V
3080 '/ i
I V./,V I !
~
28170 ?-
1800 ! I
100 2()'] 3'0'0 1;(}8 500 6'00 700 8'00
p,atm
FIGURE 54. Viscosity of dimethyl ether.

96
4. Viscosi~y of water and steam

Most of the research on viscosity at high pressure was carried out for
water and steam because of the problems involved in designing and operating
steam boilers and stearn engines.
Speyerer /81/ was the first to measure the viscosity of stearn up to 3 atm,
and later up to 10 atm. The author found an appreciable increase in viscosity
with increase in pressure. A laTge number of papers then appeared in which
the viscosity of high-pressure steam was measured by various methods.
The results dHfered widely, but again the capillary method gave results with
the best agreement.
The viscosity of stearn was determined by Shugaev /82/, Shugaev and
Sorokin /30/, Sigwart /22/, and Timrot /31/. Shugaev measured the
viscosity at 400 e at pressures up to 93 atm. His results did not confirm
0

those of Speyerer. Shugaev gives no numerical values expressing the

pressure-dependence of the viscosity, but concludes that this is much less
than indicated by Speyerer, and that at pressures up to 93 atm the viscosity
can be considered as pressure- independent provided we do not require data
more accurate than ± 6%. In a later paper Shugaev and Sorokin attempted
to derive an equation for the pressure-dependence of the viscosity of stearn,
but they did not obtain sufficiently reliable basic data due to the large
scattering of the experimental points.
Sigwart determined the viscosity of water and steam at temperatures
up to 382. goe and pressures up to 270 kg / cm 2 • Sigwart's results confirm
those of Shugaev. Timrot carried out wider research on the viscosity of
water and steam up to 600 e and 300 kg / cm 2 • In general, his results agree
0

with those of Shugaev. Later, Khlopkina /83/ used Timrot's viscosimeter

to measure the viscosity of steam up to 700 e and 500 kg/ cm 2 • Figure 55
0

shows the viscosity of stearn as a function of temperature and pressure /83/.

Schiller determined the viscosity of steam at temperatures up to 282°e
and pressures up to 30 atm. He lneasured the coefficient of discharge of a
normal nozzle (standardized by water) when stearn is discharged from the
nozzle at different rates. The data of Schiller show a very marked increase
in viscosity with pressure.
The theoretical principle of Schiller, namely, that the coefficient of
discharge of a nozzle is a function of the Reynolds number, is not quite
accurate. Moreover, in his calculations Schiller did not take into account
the compressibility of steam. Therefore, his data are not reliable, and
we do not recommend his method.
The falling weight method was used by Hawkins, Solberg and Potter /84/
to determine the viscosity of steam up to 528°e and 238 atm. Their data
are more than double those obtained by the capillary method. Thus, for
steam, as for other gases, the results of the falling weight method are
higher than those of the capillary method.
Timrot /85/ showed that the very high pressure-dependence of the
viscosity of steam obtained by these authors is due to incorrect calculation
Jf the viscosity coefficient. He corrected the theory of the falling weight
method, and recalculated the data of Hawkins, Solberg, and Potter. He
found a satisfaC'.tory agreement with the data of the capillary method.
Later Hawkins et al. /84 a/ determined the viscosity of steam with a
,~apillary-type viscosimeter. The capillary was coiled into a spiral of
diameter 0.61 m and length 30 mm. Steam superheated in a boiler entered

97
the capillary, and after passing through was choked and condensed. The
amount of steam passing through the capillary in unit time was determined
from the amount of condensate. The pressure drop between the inlet and
outlet of the capillary was determined by a differential mercury pressure
gage with metal tubes. The position of the menisci in the tubes was found
by X-ray techniques.

0-200
,-250 "
5 0-300
".
E @>-Saturated
vapor
~ Khlopkina
• -30kg/cm 2
~¥
,
~ '-100 "
0
,-;
'-2iJ0"
·-300 "
~
+-400 "
3 '-500 .

200 300 5170 600

t, ·C
FIGURE 55. Viscosity of steam as a function of temperature
and pressure.

But even with this capillary viscosimeter the authors obtained \ 'scosity
coefficients of steam which were higher than those of the capillary method.
Since the flow pattern of liquid and steam through a curved capillary is
not sufficiently known, Hawkins, Solberg and Potter used a relative method.
They calibrated the capillary with water and nitrogen as standards. Most
of the calibration points were obtained for Re numbers considerably lower
than those at which the steam viscosity was measured. (Therefore it is
quite possible that many experimental points were obtained in the turbulent
region. )
It thus appears that of the existing experi.mental data on the viscosity of
water and steam, those of Sigwart (Table 40) and Timrot (Table 41) are
the most reliable.
Table 42 shows the results of the graphical- analytical processing of
the experimental data by Vargaftik. They refer to water and steam at
temperatures between 0 and 700°C and pressures from 1 to 400 kg/ cm 2 •
Vargaftik's results can be recommended for engineering calculations.

98
TABLE 40. Viscosity (10- 6 kg sec m- 2) of water and steam (Sigwart)
Liquid Steam

kg
t,oC p,--
I cm 2

152.5 23 18 .55 275.8 25.5 1.97 3511 27 2.25 375.4 38.5 2.35
177.3 29 15 .13 275.8 35.5 1.98 350.9 47.5 2.30 375.5 78.5 I 2.46
243.5 85 11 .47 275.5 49 1.95 350.9 55 2.30 375.5 95 2.49
278.0 93 10.95 275.2 59 1.97 3509 105.5 2.35 375.4 122 I 2.55
3100 147 9 .64 276.2 59.5 2.02 350.5 127 2.38 375.8 149 ,I 2.59
324.7 153 9.06 300.5 27.5 2.03 350.9 127.5 2.37 375.4 174 2.55
349.8 195 7.69 300.8 44.5 2.03 350.5 144.5 238 375.0 183.5 2.52
351.9 178,5 7.31 299.8 60.5 2.10 350.9 162.5 2.45 375.2 195 2.68
351.3 212 5.57 301.0 73 2.11 352.2 28 2.31 375.0 209 2.72
352.0 210 5.43 300.0 80.5 2.15 352.0 55 2.33 375.5 218 2.87
370.0 227 5.87 299.8 81 2.08 361.9 82.5 2.38 382.7 125 2.50
370.4 241 5.21 300.0 85 210 351.6 112.5 2.43 382.7 152.5 2.56
370.5 228 5.81 299,8 85.5 2.14 352.2 130 2.41 382.5 179 2.59
371.4 253 5 37 301.0 86 2.13 351.9 150 2.41 382.7 203 2.57
325.4 50 2.20 351.8 151 2.45 382.7 225.5 2.93
325.0 73.5 2.25 362.0 169 2.47 382.9 235 3.04
325.1 95.5 2.31 351.8 175.5 2.55 382.'1 253 3.93
325.1 105.5 2.27 352.0 179.5 I 2.58 382.7 270 5.10
325.5 120 2.26 352.2 I 184.5 .1 2.56 , 382.9 270 6.15

An analysis of the experimental data of Timrot and Sigwart shows that

we cannot agree with some of the conclusions obtained. Thus, all previous
authors found that the relative change in the viscosity of liquids and gases
with pressure at temperatures above the boiling point (at 760 mm Hg)
increases with increase in temperature. Though the behavior of water is
somewhat anomalous, in particular in the temperature range around DoC,
the viscosity decreases with increase in pressure over the given pressure
range; starting from 30°C the pressure dependence of the viscosity is
similar to that of other liquids.
If we take into account the changes in the viscosity of water at different
temperatures when the pressure changes from 100 to 300 kg / cm 2, we
obtain the values given in Table 43 from the data of Timrot and Sigwart.
The viscosity coefficients were taken from the interpolation tables compiled
by Timrot and Sigwart themselves.
We see that if we use the data of these authors, at higher temperatures
the relative change in the viscosity of water with pressure is smaller than
3.t lower temperatures. This is the opposite of the usual phenomenon.
Golubev and Petrov /37/ determined the viscosity of water at temperatures
from 100 to 300°C and pressures up to 800 kg I cm 2 (Table 44). It can be
,3een that at the temperatures reported our data are somewhat lower than
';hose of Timrot and Sigwart, but in contrast to the findings of these authors,
-;he relative change in the viscosity of water with change in Pl'essure
:.ncreases at high temperatures, i.·e., water behaves like other liquids.
The extrapolation of the data of Sigwart on the viscosity of steam to
temperatures above 400°C at high pressures is incorrect. This was
previously proved by Timrot. However, we should note the following on
the data of Timrot. Let us first quote Timrot: "When we analyzed the
experimental data available on the viscosity at 1 atm, we considered the
data close to the most recent of Brawne and Linke as the most reliable. At
temperatures of 500 and 600°C only, the values of 11 adopted by us are

99
somewhat above the curve of Brawne and Linke." We have already shown in
section 1 that the data of Brawne and Linke for gases at atmospheric
pressure are too low at low temperatures and too high at high temperatures.
It can therefore also be assumed that for steam the data of these authors
are too high at these high temperatures. But Timrot obtained even higher
viscosity values at 500 and 600°C than Brawne and Linke.

TABLE 41. Viscosity (10- 6 kg sec m -') of watet and steam (Timrot)
kg pkg
I,OC p,-, l,oC 'i
em ' cm 2

18.9 100 104.4 400.0 146 2.60

62.5 200.7 46.8 446 304.5 3.45
65.0 200 45.5 449 249 3.05
133.4 205 22.4 453 198.7 2.95
133.4 96 22.2 453 151.2 2.87
202.5 356 14.9 453 92.7 2.78
202.6 100 14.2 5003 301 3.40
236 349 12.4 4,8.5 248.5 3.20
235 195.2 12.2 501 199.5 3.18
263 353 11.5 501 151 3.11
261 100 10.8 504.5 101 2.98
298.5 352 10.1 525.5 96.5 3.43
298.5 100 936 526.2 249 3.39
331 93.5 2.25 526.5 201.3 3.26
336 350 9.05 528 303 3.48
335 200.5 8.35 528.3 305 3.49
343 355 8.57 532 145.5 3.15
343 201 7.96 530.6 105 3.11
357.7 325.5 8.20 529 307.5 3.53
357.7 251 7.67 531 249 3.40
357.2 203 7.26 531 198.5 3.26
363.0 300 7.51 530 149.5 3.14
362.7 209 6.72 529 105 3.09
369.0 149.6 2.55 500.5 296.5 3.83
368.0 149.5 2.53 600 f54 3.78
369.0 202 2.84 599 198.5 3.63
377.0 99.4 2.48 601.5 150.1 357
382.0 254 4.94 601.6 97 3.51
385.5 351 7.36 309 100 2.41
384.0 301 6.43 341 150.8 2.67
380.0 251 5.40 364 199.8 3.16
397.5 102 2.51 368.2 210 3.22
399.0 204 2.75 373 217.8 3.40
398.0 300 4.83 374 22,:;.2 4.15
398.5 249.5 3.19

Thus, the absolute values of the data of Timrot and the corresponding
data of Khlopkina, and also Vargaftik's data based on these, appear to be
somewhat too high at 400°C. The relative values of the changes in the
viscosity of steam with pressure appear to be correct.
Figure 56 shows the viscosity of steam at atmospheric pressure
according to different authors. The data of Bonilla, Brooks, and Walker
/86/ over .wide temperature ranges are at present the most reliable. The
values of Timofeev /87/ agree with these. Timofeev used the Sutherland
equation with the constants 'tjo = 81.7 X 10-8 and C = 961. It can be seen from

100
Figure 56 that the points lie on a straight line described by "l/t ="1/0 + 0.37 5t,
where "I/o = 90.OX 10-8 kg sec m- 2 (the viscosity of steam at Ooe); "l/t is
expressed in kg sec m- 2, and t in degrees centigrade.

TABLE 42. Viscosity C.o- 6 kg sec m -') of water and steam

g I On satul'a-
I
p. em' tion Ii lle
I - -- 1 I 20 40 60 I 80 I 100 150 200 250 300 350 1 400
\
I. ·c I
( water ste am I ! 1
! I i I 1

I i I I I
I
I ' I i I
I I
o I! 182.3 [I

0.84 182 p82 1182 i 181

I
1181 181 180
I

179
i1178 176 175 174
1
I I
10 1133.1 0 .87 133 133 ! 133 : 133 I 133 132 132 132 1131 131 130 130
I 1
20 ; 102.41 0 .91 102 1102 ' 102 1102 ! 102 102 102 102 \102 102 102 102
I I
30 I 81.71 0. 95 81.7 81.7 81.8 81.8 I 81.8 81.8 81.9 81.9 82.0 82.1 82.1
I
82.2
I
40 66.61 0.99 66.6 66.6 66.6 66.7 I 66.8 66.8 66.9 67.0 167.1 67.3 67.4 67.5

50 I 56.0 I .02 56.0 56.0 I 56.1 56.1 I 56.2 56.2 56.3 56.5 i 56.6 56.8 56.9 57.0

60 I 47.9 .06 47.9 I 47.0 1

48.0
I
I 48.0 48.1 48.1 48.2 48.4
I
1

41.9 . 42.0
48.6 48.' 148.9
42.2 42.3
49.1
42.5
70 i 41.4 i .10 41.4 41.4 41.5 41.5 41.6 41.6 41.7
80 36.2 .14 36.2 36.2 36.3 36.3 36.4 36.4 36.5 36.7 36.8 37.0 37.1 37.3

90 32.1 .18 32.1 32.1 32.2 32.2 32.3 32.3 32.4 32.6 32.7 32.9 33.0 33.2
-1.22
-
100 28.8 .22 I 28.8 28.9 28.9 29 29.0 29.1 293 29.4 29.6 29.7 29.9

110 26.4 .27 I 1.26 26.4 I 26.5 26.5 26.6 26.6 26.7 26.9 27.0 27.2 27.3 27.5
24.2 .:31 1.30 24.2 24.3 24.3 24.4 24,4 24.5 24.7 24.8 25.0 251 25.3
120
130 22.2 .:35 1.34 222 22.3 I 22.3 22.4 22.4 22.5 22.7 22.8 23.0 23.1 23.3
20.5 .:18 1.38 20.5 20.6 20.6 20.7 20.7 20.8 20.9 21.0 21.2 21.3 21.4
140
19.0 1.42 1.42 19.0 19.1 19.1 19.2 19.2 19.3 19.4 19.5 19.7 19.8 19.9
150
17.7 ,46 1.46 17.7 17.8 17.8 17.9 17.9 18.0 18.1 18.2 18.4 18.S 18.6
160
170 16.6 1.50 I 1.50 16.6 16.7 16.7 16.8 . 168 16.9 17.0 17.1 17.3 17.4 17.5

15.6 1.54 1.54 15.6 15.7 15.7 15.81 15.8 15.9 16.0 i 16.1 16.3 16.4] 16 ..0
180
190 14.7 1.59 1.58 14.7 14.8 14.8 14.9! 14.9. 15.0, 15.1 152 15.4 15.5 I 15.6
!
13.91.63 1.62 13.9 14.0 14.0 14.1' 14.1 14.2', 14.3 14.4 14.6 14.8 I 1~.9
200
210 13.3 1.67 1.66 133 13.4 13.4 13.5 i 13.5 '! 13.6 13.7 13.8 14.0 14.1 i 14.2

220 12.7 1.72 1.70-11.72 12.8 12.8, 12.9 12.9 13.0 13.1 13.2 L'l.4 13..1 i 13.6
I
230 12.2 1.77 1.75 1.76 12.2 12.3! 12.3 12.4 12.5 12.6 12.7 12.9 13.0 13.1
I
240 11.7 1.81 1.79 i 1.80 11.7 11.8 !l.8 11.9 ': 12.0 12.1 I' 12.2 I 12.4 12.5 12.6

250 11.2 1.86 1.831 1.841 1.85 11.3 11.3 11.4 11.51 11.7 11.8112.0 12.1 12.2

260 10.8 1.92 1.871 1.88 1.91 10.9 10.9 11.0 11.1, 11.2 . 11.3 j 11.5 j 11.6 117

270 lOA 1.97 1.911' 1.92 1.95 10.5 i 10.5 10.6 10.7. 10.81 10.9 i 11.1 11.:2 11.3
280 10.0 2.0:1 1.96 1.97 1.99'2.02/' 10.1 102 1O.3! 10.4 I 10.5\10.7 10.8 10.0

290 9.6 2.10 2.00 2.00 2.03 2.06 9.7 I 9.8 9.90 10.0 110.1 10.3 10,4 10.5

300 9.3 2.17 2.04 2.05 2.07 2.1012:l4i 93 i 9.421 960 9.d 9.85! 9.98 10.1
-1-' I 1

101
TABLE 42 (Continued)

20 40 60 so 100 1150
I
I I
I 1 I II
1

310 9.0 2.24 2.08 2.09 2.11 \ 2.14 2.181 2.24 9. 14 1

9.26 I 9.38 1 9.51 i 9.62 I 9.72
320 8.7 2.33 2.13 2.13 2. 15 1 2.18 2.21 I'
2.27 8.81 , 8.92 ! 9.04 : 9.16 ! 9.2719.38
i
330 8.3 2.44 2.17 2.18 19 2.22 2.25 2.:30 8.42 : 8.55 I 8.68 8.82 8.92 I 9.03
7.9 2.57 2.21 2.22 2.
2.24 2.26 2.29 2.:3.3 7.94. 8.14 I 8.30 8.17, 8.57 i 8.70
340
1'
1

350 7.4 2.71 2.25 2.26 2.28

I
2.30 2.33 ! 2.:37 I 2.551 7.64 i 7.88 8.09' 8.22 , 836

~::~ I
360 6.8 2.97 2.30 2.31 2..12 I 2.34 2..11 I 2.56 7.04 j 7.40 7.66; 7.8618.01
370 5.8 3.44 2.341 2.35 2. 37 1 2.38 2.44 I 2.58 I 2.95 I 6.84 7.20: 7.46 , 7.65
380 2.38 i 2.39 2.41 1
2.43 245 1 2.61 ! 2.871 5.61 6. 65 1 7.00 : 7.27
2,48 Ii I
2.50 ,: 2.52 I 2.64 I 2.85 i 3.49 I 5.85 ,6.49 6.88
i
390 2.43 2.44 2.45 2.47
1 • i
400 2.47 2.48 250 ! 2.51 2..54 i 2..57 ! 2.67 I 2.85 ! 3.24 ! 4.60 1 5.77 6.39
410 2.52 2.53 2.54 1 2.56 2. 58 1 261 II"2.70 I 28C 13.15 ! 3.9215.01
I
5.82
420 2.56 2.57 2.58 i 2.60 2.63 2.G5 i 2.7.3 2.88 I 3.12 I 359 I 4.44 5.27
430 2.61 2.61 2.62 i 2.64 2.66 2.69 • 2.77 2.90 13.11 13.47 : 4.08 481
I
440 2.65 2.66 2.67 ' 2.69 2.71 2.73 2.81 2.93 I 3.12 3.40 j 3.88 4.41
450 2.70: 2.70 2.72 2.73 2.75 2.77 85 2.96 3.12 i 3.37 3.74 4.19
2. 1

460 2.74 2.75 2.76 2.77 2.79 2.82 2.89 3.50 3.14 I 3.35 '3.67 4.03
I
470 1- 2.79 2.79 2.80 2.82 2.84 2.86 2.93 I 3.03 3.16, 3.35 ,3.62 3.94
480 2.84 2.84 2.85 2.86 2.88 2.90 2.97 3.06 3.19 3.36 3.59 3.88
490 2.88' 2.88 2.89 2.91 2.92 2.04 3.01 l 3.10 3.22 3.37 3.58 3.83
500 1- 2.93 2.93 2.94 2.93 2.97 2.99 3.05 i 3.14 3.25 3.39 I 3.58 3.80
510 2.97 2.97 ' 2.98 3.00 3.01 3.03 3.09 3.17 3.28 , 3.41 I 3.59 3.79
520 3.01 3.02 3.03 3.Q4 3.06 3.07 3.13 3.21 3.31 : 3.44 : 3.60 3.79
530 ! - 3.05 3.06 1
i 3.07 3.09 3.10 3.12 3.18 3.25 3.35 I 3.47 I 3.62 3.79
540 i_ 3.10 3.11 3.12 3.13 3.15 3.16 3.22 3.29 3.38 I 3.50 3.64 3.80
I
550 3.15 3.16 3.16 3.18 3.19 3.21 3.26 3.33 3.42 I 3.53 3.66 3.81
560 3.20 3.20 3.21 3.22 3.24 3.25 3.31 3.37 346 ! 3.56 3.58 I 3.82
570 3.24 3.25 3.26 3.27 3.28 3.30 3.35 3.41 3.50 3.59 I 3.71 i 3.84
580 3.29 3.29 3.30 3.31 3.33 3.34 3.39 3.46 3.53 3.6.3. 3.74 ! 3.86
590 - 3.33 3.34 3.35 3.36 3.37 3.39 3.43 3.50 3.57 I 3.66 3.77 3.88
600 3.39 3.40 341 3.42 i 3.43 : 3.44 3.48 3.54 3.61 3.70 3.80 3.90
650 3.64 3.65 3.66 3.671 3.68 i 3.69 3.73' 3.79 I 3.86 3.94' 4.02 4.11
700
3.90 3.9_1_3_.92_1_3._93_1 3] :'_3_.9_9_
1
_4_.0_5_i_4._12_i_4_.2_0_~8_4._35_

102
 1

450 /
V
V
./
351i7
V
"'6
~30i7
,~/V
V-
~
~2£i
,7
....
$
20t)
V 0-Bonilla, Brooks, and Walk er

15(,,I ..¥
V ~-Brawne and Linke
x.AU-Union Energ.lnst .
'Timrot)

10C
"
/r t - Calculated from Suther! and's
equation t%.8!110-U'~
- Calculated from equati on
'It''lo+fJJJ5t; 'lo..(J(J·Ifl-6 I
of,A
100 toO 300 400 500 8f}O 1110 800 $(JO fOOO
t."C
FIGl'RE 56. Viscosity of steam at atmospheric pressure and different
temperatures.

TABLE 43. Ratio '1300/'1100 for water

t. ·C 100 150 200 300

According to
Timrot . . . 31.0
/ 29.3
= 1.058120.4 = 1.051
19.4
14.9 = 1.042
14.3 ~:~ = 1.053
According to
Sigwart. .1 35.0 = 1.14 21.8 = 1.10 154 = 1.06 9.80 = 1 04
30.7 19.B 14.4 9.42 .
A ccording to
Go1ubev and 29.70 19.58 14.0 103 9.80 _ 1075
Petrov • . . 29.21 = 1.017 19.f8 = 1.021 13.6= . 9.12- .

TABLE 44. Viscosity (10- 6 kg sec m -2) of water (Go1ubev and Petrov)

p. kg.'
cm 2
20 40 60 80 100 150
I
·c
I. I
100 29.03 29.07 29.12 29.17 i 29.21 29.33
150 19.04 19.07 19.10 19.14 I 19.18 19.28
200 13.42 13.46 13.51 13.56 13.70
I 13.60
250 - - 10.90 10.96 11.03 Il1.S
~
300 - -
I
- - 9.12 I 9.40
----

103
 TABLE 44 (Continued)

~i
"---i--~
I
1 I I
200 250 300
I 400
I 500 600 800

I
I
' 'c
! 1
I I ,
i
I I
100 29.451 2957 29.70 I 29.96 !
30.20 I 30.49 31.10
150 19.38 19.48 19.58 19.80 20.00 20.24 20.75
200 13.80 I 1389 14.00 I 14.20 14.37 14.54 14.92
250 11.25 i 11.35 11.44 11.62 11.82 1 12.00 1236
300 9.56 I 9.69 9.80 . 10.05 10.27 ' 10.50 , 10.90
I I

The ratio of the viscosity coefficient of heavy water D20 to the viscosity
coefficient of ordinary water H20 in the gaseous phase under atmospheric
pressure agrees with the values obtained from the kinetic theory of gases.
It is equal to the square root of the ratio of the molecular weights of the
M
compounds MD,o = 1.05. This ratio does not apply to the liquid phase. It
H,O
increases with decrease in temperature, and is equal to

10 I 15 I 20 i 25 i 30 I 80 1100 1125
I I I ,

~n,o/~H,oI1.309
I I I
1.286 i
1.267 I' 1.249:'I 1.232 1.21511.164
I I
1.154/1.146 I

5. Viscosity of vapor and liquid on the saturation line

There is a scarcity of data on the viscosity of liquids and vapors on the

saturation line because of the considerable experimental difficulties
involved. On the other hand, these data are necessary to calculate the
hydraulic resistance, mass and heat exchange, vaporization, condensation,
rectification, and other processes.
In work with saturated vapors, the vapors may condense on the capillary
walls, and this may appreciably distort the results. For this reason, the
experiments in this range must be carried out with special precautions, and
the given temperature conditions must be adhered to strictly.
The author determined the viscosity of vapors and liquids on the saturation
line with a setup of the fifth design. The viscosimeter was similar to that
shown in Figure 21. The viscosity of the vapor of each compound was
measured while the pressure was gradually increased at each given
(constant) temperature.
But for the liquid phase, on the other hand, the pressure was gradually
reduced. In general, the scattering of the experimental points increases as
the saturation line is approached. We found the greatest scattering near
the critical point.
However, our research was facilitated because the pressures and
temperatures on the saturation lines for the compounds we studied had
already been determined by other authors with a high degree of accuracy.

104
It is also worth mentioning that in our case condensation in the capillary
could not occur, since the pressure in the capillary was slightly lower than
the saturated vapor pressure as the mercury levels in the upper and lower
viscosimeter reservoirs were at different heights.
Strictly speaking, we measured the viscosity of vapors at pressures
only slightly "::Jelow the saturation pressure, and of liquids at pressures
slightly above. This approximation, however, could not have any real effect
on the value of the viscosity coefficient.

1\
50
,!\ \ 1\
\
5{;
;' 1\ 1\ j-Methanol
\ \ 1\ a-Ethanol
o-Isopropanol
1\ 1\ o· Isobutanol

f\, '\ \
'\ \, i~
~
20
'" "" "'~"'''"'' ~
~
~
r----
~N---- "-:: f0, 1"-",--
i'--+ r::i ~ ~
~
to ~

-':i~ ~ I"'--- r--...

~
I .11l), I).
o
.50 100 fJO zoo t,·c
250

FIGURE 57. Viscosity of alcohols on the saturation line.

The averaged and smoothed results of our measurements are shown in

Tables 45- 50 and in Figure 57. The figure shows that with increase in
temperature the liquid viscosity decreases and the vapor viscosity increases
on the saturation line. These changes are particularly marked near the
critical point, at which the viscosities of the liquid and vapor become
identical. Hence, at the critical point the temperature coefficient of
viscosity changes its sign from positive for the vapor state to negative for
the liquid state. This temperature dependence of the viscosity of liquids
and vapor in equilibrium is characteristic of all substances. We found that
the absolute values of the viscosity coefficients at the critical point (Per and
fer) of all the substances we studied differ little.
It should be noted that the permissible error of the measurements that
we calculated, taking into account the scattering of the points, is somewhat
larger (± 5%) near the critical point than that of the other measurements on
the saturation line (± 2%).

105
TABLE 45. Viscosity (10- 7 g cm- l sec-I) of carbor. dioxide on the
saturation line (Golubev)

~s
/,oc Ps' atm ~t
\-- vapor liquid

0 34.25 1365 1550 10280

5 39_0 1384 1620 9040
10 44.2 1408 1720 7938
15 50.0 1432 1830 7022
20 56_25 1456 1975 6116
25 63.1 1480 2180 5160
30 70.6 1504 2670 4130
31 n25 1510 2960 3709
31.35 72.85 1514 3220 .1220

TABLE 46. Viscosity (10- 1 g cm- 1 sec-I) of ammonia on the saturation

line (Golubev)
'1")8
/,oC p,
s atm ~t
vapor liquid

30 11.5 1040 1075 13600

40 15.3 1075 1 115 12480
50 20.0 1 112 1155 11360
60 26.0 1146 1200 10300
70 32.8 1 183 1245 9220
80 41.0 1218 1290 8260
90 50.7 1250 1335 7320
100 62.2 1287 1395 6400
110 75.0 1322 1470 5600
120 89.8 1358 1625 4840
125 98.1 1375 1760 4370
130 106.8 1394 2020 3730
131.5 109.7 1400 2250 3310
132-4 111.5 1403 2610 2610

TABLE 47. Viscosity (10- 7 g cm- l sec-I) of methanol on the saturation

line (Golubev)

t,OC Ps' atm "fit 11-----~_s___ _

_ _ _ _~-----'--- _ _ _-:-- vapor I liquid
100 3.0 1225 1230 22600
110 4.6 1256 1264 20350
120 5.2 1289 1302 18500
130 8.2 1323 I 340 16900
140 10.6 1355 1380 15400
150 13.6 1393 1416 14000
160 17.0 1426 1460 12750
170 21.0 1462 1505 11650
180 25.9 1497 1550 10500
190 31.8 1531 1607 9500
200 38.8 1563 1678 8500
210 466 1597 1766 7520
220 55.6 1630 1887 6600
230 66.0 1662 2100 5670
235 no 1678 2320 5050
238 76.0 1687 2530 4445
240 78.7 1695 2840 2840

106
TABLE 48. Viscosity (10- 7 g em -1 sec -1) of ethanol on the saturation
line (Golubev)
~.
t, ·C Ps' arm 'It
vapor I liquid
100 2.1 1085 1092 31900
110 2.8 1118 1126 2780J
120 4.1 1146 1158 24350
130 5.6 1175 1190 21500
140 7.4 1203 1220 19000
150 9.6 1232 1255 16900
160 12.2 1260 1293 15050
170 15.3 1289 1329 13400
180 18.9 1318 1370 12000
190 23.2 1347 1412 10700
200 28.4 1376 1465 9500
210 34.5 1404 1530 8360
220 41.5 1433 1620 7250
230 50.0 1460 1774 6150
240 59.6 1488 2196 4800
242 61.5 1493 2560 4200
243.1 63.1 1497 2850 2850

TABLE 49. Viscosity (10- 7 g em -1 sec -1) of isopropanol on the salUration

line (Golubev)
-----~----------~-~------------

I ~s
t, ·C PI)' atm ~t
1- vapor liquid

100 1.9 977 980 37600

110
120
2.8
3.8
1004
1032 I 1010
1038
32500
28200
130 4.9 1058 1068 24600
140 6.05 1085 1097 21400
150 8.4 1112 1126 18850
160 10.8 1137 1156 16650
170 14.0 1162 1190 14700
180 17.5 1189 1225 13000
190 21.1l 1214 1270 11600
:200 27.0 1240 1328 10100
210 33.0 1266 1406 8650
:~20 39.7 1292 1525 7250
:~30 48.0 1318 1850 5800
2:33 51.0 1326 2225 4770
235 53.0 1332 I ----
-----
2820 2820

TABLE 50. Viscosity (10- 7 g em -1 sec -1) of isobutanol on the saturation

line (Golubev)
~8
t,oC Ps' atm ~t
vapor liquid

150' 4.0 1060 1075 26000

160 5.6 1083 1100 23100
170 7.4 1106 1124 20700
180 9.6 1128 1152 18600
190 12.4 1152 1180 16650
200 15.4 1174 1208 15000
210 19.1 1197 1244 13400
220 23.4 1220 1280 11 900
230 28.3 1242 1327 10400
240 33.9 1265 1400 8950
250 39.6 1289 1515 7450
260 45.1 1311 1780 5750
262 46.3 1944 5020
263 46.9 2100 4800
2G5 48,0 1323 2770 2770

107
 Of other papers, we shall mention only that of Khalilov /35/ whose data
are summarized in Table 51. The results of Khalilov truly show the change
in the viscosity on the saturation line for both vapor and liquid. However,
if we plot his data, for example, for methanol and ethanol, together with
the literature data for these compounds at 1 atm (see Figure 58), we see that
over the temperature range studied the coefficient of viscosity of the
saturated vapor at the saturation pressure, which is much higher than
1 atm, measured by Khalilov is smaller than at 1 atm. In this respect, our
measurements do not confirm the results of Khalilov for alcohols; this can
be seen in Figure 58.

UOO~-1---+--~--~--+---r-~--~--~

~ 24001------ <Ii c - Golu bev

-Khalilov
0;<
00
~'
522{)
I!
t)!)

';o20a
.-;
I ilt, I
IS
f800

f500

fWD

f200
."-
fOOD
50 tOO f50 ZOO 250
t,OC

FIGURE 58. Viscosity of metharrol (I) arrd etharrol (II)

vapars orr the saruratiorr lirre accordirrg to the data of
differerrt authors.

But if we consider the difficulties involved in an accurate determination

of viscosity coefficients on the saturation lines, the quantitative agreement
between the data of Khalilov and our data is quite satisfactory. The
discrepancies are small, and probably due to inaccuracies in some of the
measurements.
For liquids on the saturation line (compa.re Tables 47, 48 and 51), our
data agree fairly well with those of Khalilov.
Slavyanskii /88/ reports a graphic method for straightening the curves
of the viscosity of liquids plotted against temperature. The viscosity is
plotted on a graph with a specially constructed functional viscosity scale
(4)-scale) on the ordinate, and with the conventional temperature scale on
the abscissa. The 4>- scale is plotted using a selected 1/ standard" liquid.
For all other liquids with a temperature dependence of the viscosity
obeying the same law as the standard liquid, the viscosity coefficients will

108
be represented by straight lines with different slopes. This method is
convenient and sufficiently accurate for interpolating and extrapolating
viscosity coeffiGients to different temperatures.

TABLE 51. Viscosity (10- 1 g cm- I sec-I) of some liquids and their saturated vapors (Khalilov)

I I
Methanol Ethanol Propanol n-Hexane n-Pentane

I I I T,v
I
'iL I r,v I r,L l,oC 'iv 'iL
I.'C T,v
I
'iL
I
l,oC 'iv
I "L I, °C
I I,"C
i I I I ~

I I ,
I II I
I
I
1 I
20 I - 6140J 20 I - 1124,"1 20 I - 223700i
I
20 771 1318J 20 i - 41300
II i I
1

- 91710 1 40 Ii - 30 782 \29300 30 i - 36900

30 1
53500 40 i 139100!

11~0
II I
40 47000 60 - I 65420: 60
I
- ' 88760 1 40 792 ! 27000 40 - 33400
I ,
1- I
1

50 j
1140 41300 I 80
I
46600
I
80 993 63450 ! 50 779 124800 50
i 792 30400

60 1180 36300 ! 90 ' 1110 39300 1

90 i 1000 53550 60 784 I 22800 60 i 806 27700
!
70 11200 32100 100 I 1 130 I 33800i 100 1010 45700 70 790 I 20800 70 815 25400

80 'I 1210 28500 110 1140 I, 29200' 110

I
1020 39200 80 794 19000 80 I 800 23300
1

90 j 1240 25300 120 1 160 I 25300! 120 11 020, 33 700 90 797 17300 90 798 21200
, '
1,

100 22500 130 1210 i 22 000: 130 1030 , 28900 100 797 ! 15800 100 793 19500

110 20100 140 1230: 19400 140 11070 24700

1 110 802 14400 110 778 i 17900
120 18000 150 1240 16900 150 1110 21100 120 819 13200 120 801 16500

130 16500 160 1260 14800 160 1120 18300 130 828 12100 130 822 15100

140 14800 170 1280 13100 170 1160 16000 140 835 11 000 140 839 14100

150 1350 13600 180 11 600 I 180 1180 14300 150 851 10 150 845 13000

160 1380 12700 190 1310 10200' 190 1200 12800 160 878 9120 160 847 12000

170 1410 II 900 200 1340 9040 200 1230 11 400 170 881 8320 170 871 11000

180 1440 210 1430 7910 210 1270 10200 I 180 889 7590 180 909 10100

190 1470 220 1450 7090 220 1300 190 893 6890 190 921 9250

200 1510 230 1490 230 1380 200 906 6220 200 926 I 8510
210 1600 240 1770 240 1550 210 935 7840

220 1680 242 1960 250 1780 220 914 7270

230 1960 260 2190 230 878 6720

234 2120 ; 263.5 2380 240 933 6100

I
238 2450 I 250 1150 5380
260 1230 4190

I 265 2990

109
 We know, however, that the temperature dependence of the viscosity of
an arbitrary group of liquids is not governed by the same law in all cases,
and general rules apply to groups of liquids with similar properties.
Shimanskii and Ravikovich /89/ showed that by using the ¢- scale it is
possible to straighten the temperature curves for kinematic and dynamic
viscosities of groups which have a similar molecular structure. The ¢-scale
is applicable to both pure liquids and their binary mixtures or solutions.
Below we shall show how to apply the scale to straightening the curves
indicating the temperature dependence of the viscosities of liquid and vapor
on the saturation line.
" The functional scale is plotted as follows. The conventional temperature
scale is plotted on the abscissa in Cartesian coordinates (Figure 59). From
the origin of the coordinates we draw a straight line at an angle of 45°, and
from tabulated data we plot the viscosity coefficients for the selected
standard substance on the ordinate corresponding to the projection of the
given temperature on this axis. In this way we plot the scale of viscosity
coefficients for liquids on the saturation line up to the critical temperature
(point A). At an arbitrary angle (which should however fit into the scale),
we draw a straight line from point A to the ordinate. On the ordinate, as for
the liquid, we plot the viscosity coefficients for the corresponding tempera-
tures. As a result we obtain a uniform temperature scale and a nonuniform
viscosity scale in Cartesian coordinates.

'l~_
'1~ L

' + ex
, I

'lcr I A

I,

FIGURE 59. Plotting the functional vis- FIGURE 60. Application of the ¢-scale to different sub-
cosity scale (¢-scale). stances.

To construct this functional viscosity scale it is best to plot on the

ordinate the maximum possible number of points over the largest tempera-
ture range possible for the standard.

110
 t075
:~~1J--+---"'''<l--+-+--l---j
'200 980
~ '~'t5 fOlD
19.18 ~
~ njll--4---+--+~~.--+-- ~~
-t
I :g~$
7' '''70 ~ ffj§
~ ~l-,
"U..a
6 ~l% ~ j~~b
u ~~'+--4j~.-+--+-+~~~-j E :~$~
~0fJ 5600 () ~~i~
b.OS800
""0
b?. . . I>
6WO'+--++-----f·---f---Ji'--+---j ';" 1250
'0
.....-1320
1:;;'8260
co B6SO
~0100
1/m~
9~O~-+----f-~~-r--+-~
10J
11800
!.3000
14700
IY 10
m~I+_--+---~~--+_~--J-__j
12M!
!(j{j50
18850 V 1/
13600+__----,I!=--_+__
21400
2't600
28200
1/'1/ 1- Isopropanol
(standard)
32500
37600 V 1/1 'I- Isobutanol
100 125 150 100 150 200 250 SOO
t;C t, "C

FIGURE 61. Application of the ¢-scale to FIGURE 62. Application of the ¢-scale to
ammonia and carbon dioxide. isoalcohols.

Figure 60 shows that for substances of different molecular structure the

scale is inapplicable. On the other hand, we see from Figures 61, 62, and 63
that the scale gives satisfactory results for liquids and vapors of similar
structures on the saturation line.

~
r--.....
:~ffi .........
~ '-". ~
I ()
Q) 1156
(I) IfIJQ
-t ~~~ ...........
'"'-4 ,'a... ~ ,,~~
'5 t{U ~
~~«~ ~~~
~o ~11 ~? )P
~... 8!50
'>-7.?50
8360
9500
k?" /
10700
12000
13400
V~ ~
15050
18900
/ YY
19000
21500 ,YS ~./ I
!
2f350
27800
31900
V Y/ !

't8500
/ 17~ I
/- Methanol
A-Golubev
x-Khalilov
69000
",'
V V V~ !Y iff- Ethanol (standard)
o-Golubev
98500 1/ VIh II/- n -Propanol
!ff-~'Khalilov
IWO(J(J
180000 V IA 50 /00 150
n - Butanol
a Thorpe and Rodger
2Cfl 250 3Cfl
t:o
FIGURE 63, Application of the ¢-scale to normal alcohols.

III
6. Viscosity of liquefied gases

The term liquefied gases denotes the liquid state of substances which
under normal conditions (1 atm, room temperature) are gaseous, and exist
as liquids at very low temperatures only.
The viscosity coefficients of liquefied pure gases and their mixtures are
important in many engineering problems involving separation of gas mixtures
by deep-freeze methods, and for the thermal and hydraulic designs of
separation equipment.
The viscosity of liquefied gases was studied in detail by Rudenko and
Shubnikov /90/, who determined the viscosity oJ liquid nitrogen, oxygen,
carbon monoxide, argon, methane, and ethylene, from the triple point to
the 760 mm Hg boiling point. The authors used the capillary method.

1415R W
t:t t'r iii;. 1"--' mrffi I •

140
'30
-H
J lit I -l-+-\-4-h+ I .:.'
LI.i i i: iT:
1

20-'.--""; ~. ii ililllil.ill
i i iii i ! I
. '-'--:' C,H o 1 'ii i II iT I II II
10 --1---t+---'-+-+--t---+-~-++1-+-+-++++1-+-H
! : :!:! iii i ! L .1 [ i

10

o ~ 0
10 30 50 70 90 /10 130 150 170 180 5010 80 (00 (20 140 (60 (80 200 220 240 260 280 JOO 320
T,'l( T,"/(

FIGURE 64. Viscosity of liquefied H2• N2• CO. Ch4, FIGURE 65. ViSCOSity of liquefied gases O2• C2l-4.
Ar. and air. C 2H 6• C 3H 6• C3HS. C.H 10 . CSH12. C 6H14 (for propane
and propylene the values on the ordinate must be
multiplied by 10).

In a later work Rudenko /58/ measured the viscosity of these gases at

temperatures and pressures close to the critical. He determined the
viscosity coefficients of nitrogen, oxygen, methane, ethylene, and air,
with the instrument described in Chapter II. Gerf and Galkov also studied
in detail the viscosity of liquid gases and of liquid gas mixtures /59, 98/.
The data of these and some other authors are reported in Tables 52- 57,
and plotted on the graphs in Figures 64- 68.
An analysis of the experimental data shows that the viscosity of liquefied
gases decreases with increase in temperature. The viscosity varies in the
same way as for conventional liquids. Helium is the only exception, and its
viscosity becomes very low at very low temperatures /92/.

112
 MO

5~fT*-,)t'-t7'
OIl

, 1:, 3{),1A-~*,.->,j~ri;
~250
"" ""
2o.rk::==~~-,eo-..:J'''''''''od----1

200

150
o -.20 40 60 20 40 60 80 100
Mole % Nz Mole % CH.

FIGURE 36. Viscosity of liquid nitro- FIGURE 67. Viscosity of liquid

geu - oxygen mixtures. methane - ethylene mixtures.

The temperature dependence of the viscosity of liquefied gases is

expressed by an exponential equation of the following type:

(100)

where A and C are constants. The viscosity of a liquefied gas mixture is

not determined by simply adding the viscosities of the components and
taking an average. The equation proposed by Kendall and Monroe /93/

(101 )

where n n2 , ••• and "11' "12' .. , are the mole fractions and the viscosities of
"
the components, can be considered suitable for an approximate calculation
of the viscosities of liquefied gas mixtures. Verkin and Rudenko /61/. and
later Zhdanova /62/. determined the viscosity of liquid and gaseous nitrogen
at various densities (0.38 to 0.861 g/cm 3 ), from 60 to 300 OK. They used
the viscosimeter described in Chapter II. Figure 69 shows the results of
these authors, but we believe that they require checking. It can be seen
that all the 17 vs. P isotherms intersect at approximately the same point
corresponding to 2 Pcr(Pcrfor nitrogen is 0.311 g / cm 3 ), and that at this point
the temperature dependence of the viscosity changes, namely at p < 2Pcr (%i) >
a~)
> 0, and at p > 2per (aT p <0.
p

Bresler and Landerman /94/ used the capillary method to measure the
viscosity of liquid methane and deuteromethane at 90.PK, and obtained the
following results: "ICH. = 210X 10-5 and "lCD, = 234X 10-5 g cm- I sec-I. The ratio
"ICD/"ICH. o~ 1.113 is almost equal to the ratio of the square roots of the
molecular weights:

This confirms the fact predicted by theory /98/ that the viscosity coefficients
of isotopes are in the same relation as the square roots of their molecular

113
weights. This applies to the gaseous and liquid states. It has already been
shown (Table 11) that this rule applies to gaseous hydrogen and deuterium
at different temperatures and at the same densities.

400
350 1'\
'\
0<
u
'"
~~ 250
IS
300

i'-..
"',
I
"'-
N~. i I
~?OO --
1

"
~/50
~I
i . I /00 ~,'-1---;----
!

r---~ I-li-."~ I .
/~: i~
IS
100
r-::~- 130
0
::1-i--.
I-
!-.:-t--
50
o 150' 160' I 167' I I 1

o 10 20 30 40 50 60 70 80 go fOO
Mole '10 CZH6

FIGURE 68. Viscosity of liquid ethane - propylene

Inixtures.

2000~--~----~--~--~----~-+~

,
u
~ 5
~f500r----r----+---~·-----+-----rr.r-~
u
bO
,
~

o
rl

""-
moo~---+----+-----

500r---------~~~--~·----7_--~

a6 a? 0.8_ 3
jl, g cre

FIGURE 69. Viscosity - density isotherms for nitro-

gen. The temperatures Tare: 1 - 273°; 2 - 200°;
3 - 150°; 4 - 100°; 5 - 85or<:; a - a -- equilibrium
vapor pressure curve.

II
Jenkins and Di Paolo /99/ employed the capillary method to determine
the viscosity of liquid ozone, and found that at T = 90.1 OK, 1/ = 1560 X 10-5 g
cm- 1 sec- 1 , and at 77.5°K (supercooled liquid) 1/= 4140Xl0-5 g cm- l sec- 1 •
The authors recommend the Kendall and Monroe equation for determining
the viscosities of liquid ozone- oxygen mixtures.

114
TABLE 52. Viscosity (10- S g cm -, sec-') of liquefied gases. Rudenko
and Shubnikov /90/. The capillary method was used

Nitrogen Carbon monoxide Argon

T, 'K T, OK T, oK

63.9 292 68.55 287 84,2 280

64.3 200 68.6 286 86.25 262
64.8 284 72.0 2H 86.9 256
69,1 231 72.4 237 87.3 252
69.25 228 77.7 192 87.3 252
71.4 209 80.8 171
76.1 165 80.9 170
77.33 158

Oxygen \lethane Ethylene

T, 'K ~
T, OK T, OK

54.4 873 90.1 210 110.5 522

54.5 863 93.5 188 126.0 402
54.6 821 94.6 179 127.2 388
54.9 772 110.6 100 128.1 369
56.4 717 111.2 98 134.1 333
57.1 638 141.2 281
57.4 638 155.0 215
59.7 631 160.0 207
61.7 521 168.3 181
63.5 476 168.7 178
65.4 435
68.9 377
72.3 323
77.4 273
80.0 250
90.1 190

Rudenko /58/. The concentric cylinder method was used

Nitrogen Oxygen Air

T, OK ~ T, OK T. OK

77.4 1.18 90.1 190 90.1 132

90.1 116 111.0 125 107.2 94,5

104.1 85 111.8 124 111.0 90

105.3 84 125.6 108 125.1 82.5

111.2 75.5 138.4 100 1264 80.5

111.7 74 14.5.6 96.5

154.2 91.5

115
TABLE 52 (continued)

Methane Ethylene

T, OK T. OK

111.2 98 169.:3 167

125.9 90 206.1 96

134.2 82 233.9 77.5

:54.6 67 240.9 75

167.1 63.5 273.1 65

168.4 62.5 280.9 62.5

TABLE 53. Viscosity (10- s g cm -1 sec -1) of li.quefied gases. Gerf and
Galkov /91/. The capillary method was used
Nitrogen Oxygen Methane
T. 'K

66.2 247 64.4 541 94.4 187

69.0 217 69.6 399 98.3 162
71.1 201 74.4 343 102.4 144
733 184 79.2 279 108.8 125
75.4 171 84.0 253 111.2 119
76.7 164 91.0 218
77.3 159

Ethane Ethylene'- .
T. 'K ! T. OK T. oK I

101.2 878 115.4 526 105.0 660

103.3 787 121.0 458 105.3 652
105.7 729 126.8 412 108.0 600
108.0 675 127.2 403 110.4 553
111.1 624 141.2 310 129.8 334
111.4 615 249.5 277 138.4 282
112.0 570 150.3 271 148.8 231
113.4 556 15)'8 270 156.8 197
168.2 164

____~---~P~ro+p~an~e~~-----I------~y~le-n-e--,_
T. 'K T._o_K--;-_____ T. OK I 0 I T. OK
88~-1'1~-;1-1-2-3-0-'--13-1-
'I

83.5 1353111.6 210

85.1 1154 I 119.4 149 898 1273 131.2 90
86.6 988 133.2 98 90.1 1240 1416 70
88.1 866. I 141.8 97 94.3 I 784 141.9 67
89.8 738 I 144.4 7-1 98.0 I 537 150.0 55
93.1 609 I 149.8 72 102.6 358 159.8 45
96,8 4;9 I 160.0 56 105.9 273 169.6 34
101.6 3"8 I 169.6 38 111.1 215 173.0 38
106.7 257 175.8 41 119.0 155 1748 37
107.8 249

116
~~ABLE 53 (Continued)
Gerf and Galkov /59/. The concentric cylinJer method was useu

C arbon monoxide /Ylethane

T, OK T, oK

73.2 224 130.6 84

75.2 203 148.6 68
77.8 136 161.4 59
828 165 165.8 56
90.1 146 180.8 51
99.6 116
111.6 100
129.6 66

'·-===---==;h~~~--r--
Ethylene

T, OK I T, oK

172.8 175 183.8 135

186.0 137 204.6 115
201.0 114 226.4 92
215.4 97 252.2 72
230.0 90 273.1 64
231.8 92
243.4 82
247.6 76
270.0 67
273.1 63
288.0 55

TABLE 54. Viscosity (10- 5 gem -1 sec -1) of liquefied gases

Helium Hydrogen
(Keesom and Mac \\food /95/) (Keesom and MacWood /95/)
----- ~~- ------1-------

-----,------ i
1.282 0.154 14.8.'; 21.62
1.304 0.124 15.89 19.32
1.305 O.17Y 16.35 18.4~

1.586 0.233 17.77 15.93

1.76:2 0.356 18.78 14.48
1.906 0.675 20.43 19.00

117
TABLE 54 (Continued)
1.988 0.960
Ammonia
2.086 Ll5 (Shatenshtein el al. /57/)
2.111 1.26 T, OK
2.116 1.35
2.145 L55 188.16 147.9
2,159 1.77 193.16 141,1
2,17l 2.02 198.16 134.5
2.174 1.95
(Monoszon and Pleskov /96/)
2.178 2.30
223.16 317
2.29 1.87
233.16 276
2.315 1.97
237.66 255
2.642 2.35
283,16 162.4
2.934 2.40
288.16 i55.5
3.738 2,7,5
298.16 148
3.81 2.87
3.97 2.88
4.021 2.98

TABLE 55. Viscosity OO-s g cm- l sec-I) of liquefied gases. Stel,

Mac Intosh and Archibald /97/

Hydrogen chloride Hydrogen bromide

T, OK T, OK

1608 566 186.8 874

166.7 546 188.8 865
171.7 508 190.8 854
177.0 493 19.3.7 841
183.2 473 197.3 822
188.2 457 199,4 816

Hydrogen iodide Hydrogen sulfide

T, OK T, "K

223.3 1418 191.0 525

225.6 1394 193.3 506
227.2 1378 198.2 489
229.6 1368 201.2 468
231.5 1345 206.1 451
233.9 1321 209.8 435
236.4 1298

Viscosity at boiling point

Substance T, OK

188.2 450
206.16 800
237.16 1290
212.76 430

118
TABLE 56. Viscosity (10- 5 g crn- 1 sec-I) of liquefied gas mixtures.
Gerf and Galkov /59, 91/

Composition of mixture I Method T. OK

(vol 0/0)
I
68.7 354
70.2 319
40% N~ + 60% 02 I Capillary 72.8 232
76.8 229
I
67.2 252
80.4%N 2 + 19.6%02 I Capiilary 72.6 204
77.7 173
I
! 84.1 214
19.6% N2 + 80.4% CH4 ICapillary 85.0
87.8
206
186
89.8 172
I

96.6 155
103.6 133
24%N2 +76%CH, Concentric cylinder 109.6 118
132.4 75
145.8 66

I
78.5 217
81.4 195
41.2% N2 + 58.8%CH, I Capillary 84.7 178
86.1 171
I
93.4 133
I 96.4 125
I 103.8 113
49.4% N2 + 50.6% CH, i Concentric cylinder 110.8 103
II 138.8 64
I 139.4 64
1468 58
II

68.2 275
70.1 253
71.7 237
60.8% N2 + 39.2% C!-l4 Capillary 75.1 210
78.0 188
81.6 167
84.4 152

96.2 116
II 102.5 103
72.8% N2 + 27.2%CH. I Concenttlc
. cylm
. d er 110.3 93
128.8 68
I
I 137.2 57
146.5 44

119
 TABLE 56 (Continued)

Composition of mixture I Method T, OK ~

(vol '10)

64.8 286
65.3 280
65.7 269
67.6 245
68.2 240
70.0 223
81.2%N2+ 18.8%CH 4 Concentric cylinder
70.2 221
71.7 211
74.3 190
76.7 174
79.1 164
80.3 154
----

76.4 183
83.0 151
25.2%N,+74.8%CO ! Concentric cylinder 90.1 132
I
I
100.8 109
111.6 80
I
1

I 77.2 171
83.0 147
45.1 %N z + 54.9% CO I Concentric cylinder 90.1 127
100.8 108
!
111.6 86

81.0 153
68.7% N, + 31.3% CO IConcentric cylinder 90.1 123
111.6 84
I ------------

I
1 98.9 595
1
1 101.5 526
23.7% CH 4 + 76.3% CzH 4 1
I
Capillary 104.1
106.4
482
451
I 108.4 421
I 111.1 390
I
I
"
I

149.8 193
I 165.0 165
23.8% CH4 + 76.2% C,H 4 ] Concentric cylinder 178.4 161
I
198.8 129
211.6 114
i 238.2 100

96.6 488
I 98.9 438
41.0% Cl1 4 + 59.0% C,H 4 i Concentric cylinder 102.6 382
! 104.9 352
111.2 295

5680 120
T A 3L£ 56 (Continued)

Composition of mixture I Method I T. OK I

(vol "IaJ

1436 163
164.4 131
102
41.3~.,' CII. + 58.7 % C 11.
2 Capillary 183.0
199.0 82
205.4 80
217.8 68
I
93.7 406
95.1 382
97.5 344
60.29';, ell. + 39.8% C H.
2 Capillary 99.5
102.6
327
299
105.2 263
107.2 253
111.2
I 226
I I
92.6 311
94.9 281
99.2 240
:11. + 23 % C H.
2 Capillary 101.1 224
104.1 I 205
108.4 I 182
111.0 170
I I
133.4 I 109
80.'[o;.,CII. + 19.60f0C H.
2 Concentric cylinder 152.4
173.4
89
73
I 184.0
I 62

99.6 115
1052 92
21.'.. OJc,CgH6 + 78.9% C H 2 6 Concentric cylinder 110.6
140.6
76
37
155.8 29
168.4 24

101.4 200
105.2 164
110.8 121
50°1OC3H6 + 50 % C H 2 6 Concentric cylinder
142.0 50
154.4 38
I
I
169.8
I 30
j I

98.6 297
105.0 202
126.0 86
74~,C 3H 6 + 26% C H6 2 Concentric cylinder 143.0 54
159.2 41
169.2
174.4
I 35
31

121
 TABLE 56 (Continued)
Composition of mixture \
Method T. OK
(vol "/0)

102.0 734
104.8 654
107.8 594
109.7 557
111.2 541
145.0 253
154.3 216
156.7 209

102.6 739
104.8 665
107.8 604
109.7 560
: Concentric cylinder 110.0 552
i 111.2 537
146.7 234
152.7 211
1574 196
160.8 185

TABLE 57. Viscosity (10- 5 g em-I sec-I) of liquefied gas mixtures.

Vershaffelt and Nicaise /60/. The rolling ball method was used

Composition of mixture (vol "/0) T. OK

+
36%N2 61 %0 2 3%Ar+ 82.34 186.5
+ +
89%N2 9.5%0 2 1.5% Ar 77.91 159.5
I %N2 + +
96.0%0 2 3%Ar 89.6'2 189.5
Air 80.83 171.8

7. Reference data on the viscosity of liquids

Table 58 gives the viscosity coefficients of various liquids at different

temperatures. The reader should note the strong decrease in the viscosity
coefficient with temperature for all the liquids.
Table 59 shows the data compiled by Bridgman /40/ on the pressure
dependence of the viscosity of liquids. The table gives the logarithm of the
relative viscosity at various pressures and temperatures; the viscosity at
atmospheric pressure and 30°C is taken as unity. From these logarithms we
can find the ratio of the viscosity at 30°C to the corresponding value at
75°C at different pressures. The last column of the table shows the absolute
viscosities at 1 atm and 30°C.
Table 60 shows the relative viscosity of water. It can be seen that the
variation in the viscosity of water at temperatures below 30°C differs from
that of other liquids. This applies in particular to the viscosity at 0 and
10.3°C. It first decreases with increase in pressure, passing through a
minimum at about 1000 kg / cm 2, and then increases with increase in pressure.

122
 The following interpolation (for small temperature ranges) and extrapola-
tion formulas can be used:
The equations of Bachinskii

c (a)
Tj=v-w'

B
"1= (I+a)n . (b)

The equation of Andrade

(c)

The equation of Thorpe and Rodger

(d)

In these equations C. ill. B. a. n. A. b. a. ~. D are constants to be determined by

'~xperiment.
The reader interested in a survey of papers dealing with the theory of
j.iquid viscosity is referred to the monograph of Panchenkov /125/.

TABLE 58. Viscosity (JO .. 2 g em-' see) of organic liquids /100, 100a/
OJ
I
,
, '0
<3 § OJ
'0 "
""Q)"
'0
'u OJ
0
0 <J)
0
(3
::: U '0 oj
<20 '5 oj '0
:2 :2
~*o
c: .,- c:
~~
~
-
g :r;
'2§ :r;q:
<J) 0
0 <J '':: cr)
(,) (,) OJ ~
0 K °10 (l)
c:~
Q oj
0.. oj -

~§~ g~o
(3 Z oj

E
<J)
....
~
;:;
"" 0
v£
;:;:: U U
:2
U U
.J.,
§
0 i'
u ~Q
oj

-
""i;:lu
.;I' ';:I
;:; ,§
«:
:s
U
OJ
~..2u
~ N
U 1
!j
~ ~
CiJ :r:
u
""
.... <J)
'" oj

-100 16.0 - - - - 47,0 - - - -

- 90
- 80
8,8
5.7
-
-
-
-
I -- -
- I 28.3
18.1 I -
-
-
-
-
-
-
-
- 70 4.02 - - I -- - 12.4 I - - - -
- 60
- 50
2,98
2.26
..
..
-
- ! -
-
-
I 8.7
6.4
-
-
-
-
-
-
-
-
- 40 1.75 .. - - - 4,79 - - - -
- 30 1.39 .. - - - 3.65 - - - -
- 20 1.16 -
I - ! '- - 2.83 - - - -
-10 0.970
I
- - - 2.23 - - - -
0 0.817 - 0.70 7.3 0,844 1.78 1.24 0.43 1.14 2.66
10 2.5 0,63 5.0 0.742 - 1.05 0.38 1.00 2.13
20 0.584 1.78 0,57 3.75 0.657 - 0.90 0.345 I 0.88 1.75
30 1.46 0.51 2,94 0.587 - 0.79 0,3i5 0.80 1.48
40 0.450 1.22 0.466 2.43 0.528 - 0,69 0.72 1.28
50 0.396 1.03 0.426 2.04 0.478 0.695 0.62 0.66 1.11
60 0.351 0 .89 0,390 J.71 0.433 - 0.55 0.60 0.97
70 0.311 0 .78 1.42 - - - 0.56 0.85
80
100
0 .68 I 1.17
0.83
0.357
0,326
- 0.453
0.377
0.51 0,75
0 .54 0.441
120 0,63 0,248 0.320 0.383
I I I

123
TABLE 58 (Continued)

I~
o
o 0 IB,
g ' 0
I Q)
g
~...('l
..--... U
'>-.v
g.:3 :'3
>.. ~
c:...
o ~ ."
"'.",.,
l-> <.o! () -8_<0 c::!=-p" o..:31=-
~~ ~=: Q-' ,0-;" 0°--<
-<u lou ~
~U
____ ~
c:~U
____ ~
~.oU
_ _ _ _ _ _L -_ _ _ __

-70
-60
-50
-40
54.6

20.2
66.1
37.6
23.2
i
I
b:~~g:~~ :! ==
-30 9.5 14.9 , 0.57 -
-20 6.9 10.1 ,0.50 -
-10 5.1 6.8 0.442' -
o 3.85 4.60 1.52 1.61 ,I 0.395 ,12100 0.645 0.605 0.436
10 2.89 3.26 1.29 1.33 0.356 (' 3 950 0.575 0.538 0.390

~g "
2.20 2.39 1.10 1.12 0.322 1480 0.381 0.517 0.482 0.352
1 1.72 1.76 0.96 0.96 0.293 600 0.344 0.467 0.435 0.319
40 1.38 1.33 0.84 0.S3 0.268 330 0.312 0.42,) 0.394 0.291
50 0.75 0.73 0.246 180 0284 0.388 0.359
60 0.92 0.80 0.67 0.65 102 0.258 0.356
80 0.63 0.52 0.545 0.52 35 0.217 0.,328
100 0.452 0.430 )3.0 0.182

f:~
120 0.380 0.360 0.154
140 0.322 0.306 0.130
160 0.264 1.00 I'
180 0.45
200 0.22 ,

TABLE 58 (Continued)

I§ I
g" 'E I

;g, 6 i

"
. . U.
<.0
d, ::c
rl
()
,

o~o I
i
-120
-100
I,, gr I I =i 2.31
1.25
- 80
- 70 0.460
0.97
I 0.79
I
I 0.768
0.640
I'

- 60 0.401 0.454 I

I I 0.65 i 0.546 I
- 50 ,I
- 40
0.355
0.314
0.392
0.343 I,
34.7
22.4 51.3 [ 0.55
0.470
I 0.474
0.419
I
- 30 I 0.280 0.301 14.6 29.9 0.410 0375
- 20 I
,'I

0.252 0.267 0.244 10.3 18.4 0.364 0.341 [I

- 10 0.227 0.238 0.214 7.4 12.3 0.328 I 0.311

o
10
0.207 0214 0.188
0.167
5.19
3.87
8.3
5.65
0.296
0.268
I 0578
0.507
087
0.75
0.283
0259
0.555
0.491
20 0.149 2.95 3.95 0.243 I 0.449 0.65 0240 0.439
30 0.13,3 2.28 2.85 0.220 I 0.401 0.58 0.220 0.393
40 0.119 1.78 2.12 0.199 ! 0.360 0.52 0.354
50 0.106 1.41 1.61 0.326 0.468
60 0.095 1.14 1.24 OJ66 I 0.297 0.424
70 0.93 0.97 -- I 0.270 -
SO 0.76 0.78 I 0.140 0.248 I 0350
100
110
0.54
0.46
0.52 I 0.~~8 f
0.210
0.193
i
i
-
-
:

124
TABLE 58 (Continued)

"""
"0
"<l)
..oz
N

o ~
.~ ~
z u

o 11.1 86 0.77 1.33 1.52 1.06 0.76 3.09 10.2

10 7.4 6.1 0.67 1.12 0.76 1.31 091 0.66 2.46 6.5
20 5.1 4.36 0.58 0.95 0.65 1.13 080 0.58 2.01 4.4
30 3.59 3.20 0.51 0.83 0.56 1.00 0.71 0.52 1.69 3.12
40 2.61 2.41 0.46 0.73 0.492 0.89 0.64 0.469 1.44 2.3
50 1.94 1.85 0.41 0.436 0.79 057 0.427 1.24 1.8
60 1.47 1.45 0.368 0.58 0390 0.72 0.52 0.390 1.09 1.5
70 1.15 1.15 0.334 0.350 0.66 0.475 0.357 1.28
80 0.91 0.93 0.304 0.482 0.316 0.60 0435 0.326 0.87 1.10
90 0.74 0.76 0.'143 0.286 0.55 0.400 0.298 0.94
100 061 063 0.250 0.261 0.52 0.370 0.273 0.70 0.80
110 0.51 0.53 0.239 0.480 0.345 0.69
120 0.43 0.45 0.219 0.320 0.231 0.59
130 0.39 0.201 0.420 0.295
140 0.185 0.275 0.198
150 0.170 0.366 0.255
160 0.156 0.240 0.168
170 0.144 0.225
180 0.132 0.210 0.144
190 0.121 0.195

TABLE 58 (Continued)

-90 1.82
-80 1.38
-70 1.10
-60 0892
-50 0.742
-40 0.633
-30 I 0.547
-20 0.479
-10 0.426
o 0.381 0.648 0.883
0.759
10
20
0.343
0.307
0.569
0.505 0.660 0970 I 2.33 · 3.63 I 11.6
30
40
0.278
0.253
0.450
0.404
0.581
0.515
0.822
0.706
1.90
1.60
36.0
19.6 I 2.93
2.41
7.0
4.77
50 0.610 1.36 12.0 204 i 3.43
60 0.538 1.17 7.40 ' 1.74 2.56
70
80
0.474
0.426
i,
1.02
0.890
505
3.51
I 1.50
1.31
2.00
1.59 9.12
90 0.800 2.52 1.16 1.29 5.72 6.35
100 0.720 1.92 , 1.035 1.05 4.20 4.70
120
140
0.582
0.491
I 1.16 !
0.782 I
0.827
0684
0.78
0.69
2.46
1.61
3.02
2.07
150 0.450 0.655 I 0.628 1.33 1.72
160 0.575
170 -I 0544
Data of Golubev anc Frolova.

125
TABLE 58 (Continued)
I
0 , Q)
.s" Q)

I
""
()
.s <])
I '0
.5
:s .5
;'<]); 0u oj
'"" " ~~ &l iG '0 3 .80 z ""
:§z
P< 'P. ~
£@ "~ :£ 0
y :10 iG 0 .82
.s " "I :£ uT: o
S ~~ "
::r:;:I:~ ~ :r::
:l ~ "
>..~,..,.. U .L
o ~
f-<:r:: f-< :£ f-< :r::
~

OJ
f-< 2 I ~U
t)
2:..c: U- w P< U
Q)-I- 0
f-< S 0 U- I S S ~U
I
o U
...
sU- I ...
""U
I
-90
I 3.85 - - i -
I
I
- I -- - -
I --
!
-
-80 I 2.57 I - - - I -
I
-
I - i -
-70 I
I
1.92 - I -
I - -
I
- - - I - -
-60 1.46 - - - - - - - - -
-50 I 1.18 - .- I
i - - I - - - - I -
-40 I 0.965 I - I - - - i - - - - I
-
-30 0.807 - - i - - - - - - -
-20
-10
!
0.689
0.599
1
I -
-
-
-
I
-
-
-
-
I -
-
-
-
-- -
-
I -
-
0 0.526 0.991 I 0.725 I 0.768 - I 95 - 10.2 8.7 -
10
20
i
0.466
0.417
0847
0.731
0.637
0.565
I 0.667
0.586
-
9.8
44
21
-
20.2
6.4
4.35
5.5
3.81
-
-
30
40
0.375
0.339
0.639
0562
i
I
I
0.506
0.456
I 0.522
0.466
6.1
4.3
I
I
10.0
6.2
10.3
6.7
3.20
2.44
II 279
2.14
-
-
50 0.309 - I -
I 0.420 3.2 4A 4.7 1.94 - 1.75
60
70
0.283
0.261
-
-
I -
-
0.381
I 0.348
2.3
-
!
,
3.2
2.5
3 ..5
2.7
1.57
1.30
i 1.40
-
1.45
1.20
-
I
I
- i 0.319 -
I
80
100
0.241
0.209 I - - I 0.271
-
2.1
Ui
-
-
1.11
0.83
1.00
0.77
1.00
0.7S
120 0.231 0.58
140 0.199 0.50
160 0.172
180 0.150

TABLE 58 (Continued)
<]) I
>,
I
0
"
j" "" ..c:
'0 <]) <])
<]) <]) <])

I
I I
<]) <])
~ "
""'"
oJ
oj

:up,. U
"
'0 It<])
>-"
P<p,.
S <])
" "
<l)
"
0 ;';
~g __~12~_~ .0
~ " "?;- :r:: >, ~
,:r::~
-I'" "
"'<])
<])
"
00
[..s -'~
S -~ a) 00
:x: ,3 ;% :c;£
~
N
S Hof 0 H .c.
.sw..c:E0 _S?~-;;, u
I
""' " _ U ro
I
<:J () U ~
I c¢ I 00 00 I 00 I
f-< 2 U U o..U t=: rl,D
- Q)
U o U p,.U
" I V) " ro

-60 2.47 - - - - -. - - - -
-50
-40
1.83
1.43
-
-
-
-
-
-
-
-
-.
-.
-
-
I -
-
-
-
-
-
-30 1.16 - - - - -- - - - -
-20 0.967 - - -- I 1.170 1.447 - - - -
-10 0.829 - - - ! 1.021 1232 - - - I -
0
10
0.714
0.622
1.139
0.972
0.839
0.886
0.772
0.679
0.917
0.800
0.705
I
!
0.873
0.757
0.671
I
1.05 5
0.91 7
0.80 4
1.047
0.879
0.749
I 1.10
0.93
0.81
I 0.80
0.70
0.61
0.74
0.64
-
20 0.546 I
0.596 0.707 0.648 0.71 0.55 0.57
30
40
50
0486
0.435
0.392
0.735
0649
-
I
0.603
0.538
-
i
0.624
0.557
-
i
:
:
0.530
0.482
I
0.62 3
0 ..55 6
0.565
0.502
I 0.62
0.56
I 0.490
0.433
0.51
0.456
60 0.356 - i - I - 0.436 0 ..50 a 0.453 I
i
0.50 0.403 0.414
70
80
90
0.325
0.298
0.275
-
-
-
I -
-
-
II
I
-
-
-
0.397
0.363
-. I
OA5 6
0.41 7
.-
0.413
0.3 7 4
0.349
i
0.453
o. 41 1
0.376
0.369
0.339
0,313
0.377
0.340
0.317
100 0.255 - I - - I -. .- 0.309 0.346 0.285 0.292
110 - - - - - .- 0.289 0.318 0.269 0.270
120 0.220 -
1
- I -
1
-. .-. 0.269 0.294 0.250 0.251
I iI
130
140
-
- I -
- I,
-
-
-
-
I -
-
,
, .-
.-
0.249
0.232
0.273
0.254
0.233 0.233

126
TABLE 58 (Continued)

, <J)
l:!<J) u
0
0::
oj
0.. 0::
S Q)"
o g
<J) ~
Z:G
f-< 3 ~ U

-30 -
I 2.51 I 15.71 7.29
-20 - 1.93 !D.65 5.38
-10 J.]45 1.55 7.56 4.08
0 0.968 1.409 1.535 1.180 1.075 1.27 5.60 3.223
10 0.827 1288 1000 0.915 1.07 4.29 2.580 1.204
20 0.7144 1.184
1.004 I 1.094 , 0.855 0.788 0.907 1.286 3.371 2.122 1.020
30 0.628 0.869 0.946 0.746 0.691 1.089 2.75 1.769 0.880
40
50
0.555
0.495
0.~7 i 0824 I 0655 0611 ! -- I 0.937
0.818
2.232
1.861
1.489
1.278
0.767
0.681
60 0.445 - i 0.720 i 1.583 1.111 0.608
70 0.403 -I 1.359 0.975 0.544
80 0.366 J.]84 0.862 0.492
90 0.335 1.042 0.770
100 0.308 0.917 0.690
110 0.284 0.817 0.624
120 0.264 0.750 0.570
130 0.244 0.682 0.519
140 0.228 0.620 0.475
150 0.213 0.567 0.437
160 - 0.519 0.105
180 - 0.438 1 0.349
I

TABLE 58 (Continue,!)

<J)
, 0::
oj
l:! u t>
p.,
"

S
0

ul
""" ;S
<J)
[-<
~
Ei 9"0:: :GU: : ,

-10 2.163 :2.90

0 1.742 :2.264 2.962 12.47
10 1.425 1.827 2.339 2.96 8.33
20 1.182 1.492 1.878 2.322 2.841 3.451 5.88
30 1.010 1.253 1.561 1.889 2.291 2.754 3.291 3.813 4.34
40 0.871 1 064 1.312 1.560 1.873 2.232 2.652 3.060 3.32 4.072
50 0.759 I) 919 1.114 1.323 1.570 1.852 2.169 2.491 2.62 3.259
60 0.671 I) 803 0.969 1.135 1.335 1.560 1.829 2.060 2.12 2.665
70 0.597 ().709 0.849 0.988 1.155 1.338 1.557 1.748 175 2.220
80 0.535 ().632 0.751 0.868 1.010 U61 1.340 1.484 1.49 1.188
90 0.482 ().566 0.668 0.769 0.888 1.014 1.161 1.297 1.26 1.614
100 0.437 (),510 0.598 0.685 0.786 0.892 1.014 1.131 1.090 1.403
120 0.365 0.489 0.716 0.794 1.094
140 0.311 1).408 0.584 0.655 0.876
160 0.268 0.347 0.486 0.•')46 0.717
180 0.233 0.299 0.409 0.460 0.598
200 0.204 0.260 0.349 0.392 0.505
220 - 0.299 0.339 0.432
240 - 0.259 0.296
260 - 0.260
280 -
300 - 0.229
0.203 "

127
TABLE 58 (Continuec!l

~ J.~ "'"
;.l , ,
I-T!-]~~l--~-'
I
roI
;:!
c\j

';:!
I

3 ~
I
'"
X

0 ~ CJ ~ (') 0 () c;;
~"''''IE-<
Q (l)
~
I
{) \ g oj
x ;1 0coroC",,+-'CC'J :::~'.j...JCoo'-<
.sc ~
C0 ~ (.)
.J} ...!pI 'C cc: -!f C ('j...!'" 8 I;:! 0
t ~
E ,;' ~ I =:1 'C ...::1
E-< -
;:; -" (J
~~. "~-I- 01
1 <iJ
~-.
..!;""
i
C"j
~-
-,00
E-<
:c. . .
8'" ();;
(')
E-t 2
'-<
o
c: " NI
() ~
-~
u I U b slI '" 9 8 uj
<Il 00

~
C
8:J I
P-.
~
C
8
l()

u\b 8
...:.... c: ~ r
c3 ~ 8 U ~
I
q M

I I I I I I
- I
- II - I - I -
50
60
3.68
2.99 3.11
-
4.25
-
5.110 -
-
-- - I -
-
- I -
I
I
-
1 i
70 2.48 2.60 3.47 4.170 5.02 6.03 I - - I -
80 2.07 2.19 2.89 3.451 4.12 I 4.85 I 6.71 I 7.24 - -
90 1.77 - 2.43 2.884 3.42 3.97 5.50
I 5.89 8.60 i -
100 1.53
-
-
- I 2.08 2.432 2.87 I 3.30 1.57 487 7.07 -
110
120 - -
1.801
1.573
-
-
2.46
2.13 I 2.82
2.38
3.84
3.28 I
I,
4.13
3.52
5.85
4.97
14.30
11.85
130 - - 1.385 - 1.86 I 2.07 2.84 3.04 4.26 9.92
140
150
-
-
- 1.230
1.101
- 1.62
1.452
- 2.47 2.64 3.68 838
- - - 2.17 2.308 3.20 7.12
160 - -
I,
0.991 - 1.305 - 1.93 2.053 2.81 6.20
170 - - 0.897 - 1.186 - 1.73 1.832 2A9 5.42
180 -- - I 0.817 - 1.065 - 1.55 1.643 2.21 I 4.76
190 - - 0.747 - 0.967 - lAO 1.479 1.988 I 421
200 - - 0.684 - 0.883 - 1.27 1.339 1.795 3.74
220 - - I 0.579 - 0.745 - 1.06 1.116 1.483 3.03
240 - - I 0.496 - 0.636 - 0.897 0.944 1.242 2.53
260 - - 0.428 - 0.547 - I
0.768 0.807 1.053 : 2.09
280 - - 0.372 - 0.475 - 0.662 0.696 0.903 1.77
300 - - I 0.325 - 0.416 - 0..575 0.607 0.778 1.52
I I I I

TABLE 59. Viscosity of liquids as a function of pressure /40/

Pressure, kg/ cnl

Compound Ig -'Land~ 'lI300.
'fjo 'fj7~
1000 4000 8000 12000 gcm-1sec- 1

Ig - {370°5°
"I]
0.000 0.167 0.471 0.750 0.998

Methanol
'1]0
I 1.769 1.933 0208 OA48 0.655 0.00520

I 1.702 1.714 , 1.832 2.004 2.203

"I]
Ig-- {30° 0.000
I
0.200 0.617 1.023 1.390

Ethanol "0 75° 1.657 1.873 0.289 0.634 0.919

0.01003
I
! 2.203
I 2.123 2.128 2.449 2.958

'1 {30° 0.000 0.283 0836 1.402 1.915

Ig ~ 750 1.598 1.880 0.368 0.827 1.223
n-Propanol 0.01775
'1?0 2.523 2.529 2.938 3.758 4.920
1Jn

0000 0.321 0.934 1.609 2.208

I g'1- {30°
n-Butanol "1]0 75° 1.548 1.867 0.312 0.941 1.396
0.02237
1ho 2.845 2.858 3.343 4.679 6.518
'In

128
TABLE 59 (Continued)

Pressure, kg/em2
Compound Ig4nd~
7)1) 1j75 I· 8000 12000
'lj:jO'"
gem -1 see- 1
I 1000 4000

I "I {30° 0.000 0.341 1.060 1.811 2.495

g -:;;; 75° 1.540 1.871 0466 1.049 1.562
Amyl alcohol
"1'0 2.884 2.951 3.926 5.781 8.570
lJ7!')

"I {30° 0.000 0.343 0.982 1.640 2.311

Ig --:;; 1'5° 1.505 1851 0.425 0.957 1.424
Isopropanol 0.01757
'YJ30 3.141 3120 3624 4.844 7.748
"lJ75

"I {30° 0.000 0.388 1203 2.075

I 2.898 I
I Ig -:;;; 750
Isobutanol
"130
1.444 1.824
I 0.488 1.158 1.747
0.02864
3.597 3.644 5.188 2.260 14.16
I '175
I
II 0.000
I I I 2.952
0386 1.185 2.069
Ig To
'1 {30°
750 1.424 1.780
I
L787 0.492 1.168
Isoamyl alcohol
'130 3.805
I 4.012 4.970
II 8.042
I
15.76
I
'175
I ! .-
I -----
-- 1

'1 {300
Ig --:;; 750 I 0.000 0.315 0.847 1.360 1.846
I
n-Pentane I 1130
1.811 0.163 0.676
II 1.119 1.493
i 0.00220

- I 'In
I 1.545 1419 1483
I
1.742
I
2.254
I
I

'1 {30° 0.000 0.332 0.914 1.514

n-Hexane
Ig --:;; 750
I 1.803 I
I
0.171 0.701 l.l98 1.646 0.00296
"Iso I 1.574 1.449 1.633 2.070
TJ75
I I
Ig
'1 {30°
r;750 I 0.000
1.810
0.327
0.153
1.088
0.763 1.363 0.00483
n-Oclane
'lao II 1.549 1.493 2.113 I
I
'175
I i I
! I I,
I

To
'1 {30° i 0.000 I 0.344 0.894 1.431 1.947

Isopentane
Ig 750 1.821
I
0.193 0.715 1.179 I 1.586 0.00198
'130
'175 II 1.510 1.416
I
1.510 1.786 2.296

'1 {30° 0.000 0.242 0.649 1.008 1.323

Ig --:;; 750 1.850 0.131 0.514 0.834 ,
1 1.111
Ethyl chloride
I I
'130 1.413 1.291 1.365 1.493 I 1.633
'lj75
I I !

129
TABLE 59 (Continued)
.~------~--------~----------------------------------------------

Compound

'1 {30°
1-110.000
1000

0.222
Pressure, kg/em'

4000

0.631
I 8000

1.043
12000

1.400
'1')30°,
g cm -1 sec-I

Ig'iO" 750 1.806 0.072 0.472 0.816 1.123 0.00368

Ethyl bromide
'130
I
1.567 1.413 1.442 1.687 1.892
'175 ; i
I
0.000 0.269 0.816 1.408 2.018

n-Butyl bromide
Ig
'1
~
{30°
750 I 1.832 0.090 0.564 1.040
I 1.484 I
I
0.00537
'130 1.786 2.333 3.420
1.472 1.510
'~75
I I I
I I \

'1 {300 0.000 0.226 0.605 0.987

Ig -r;; 750 1.895 0.113 0.445 0762 1.031 0.00285
Acetone
1)30
1.274 1.297 1.445 1.679
:l7{>
I I

1){300 0.000 0.258 0.818 1.393 1.974

Ig --;:; 750 1.836 0.081 0.517 0.992 1.416 0.0039
Ethyl acetate
1)30
, 1.459 1.503 2.00 2.518 3.614
'176
i I
'1{300 0.000 0.260 0.936 1.741

Glycerol
Ig --;:; 750 2.810 2.023 1.529 I 0.094 I 0.628 3.8
'130 17.26 25.53 44.36
15.49
'175
i \
'1 {300
Ig --
I 0.000 0.351 II
Carbon tetrachloride '10 75°
'~30
I
1.760
I 0.100 0.542
0.00845
1.738 1.782 , ,
1)75 !
I

'1 {300 0.000 0.160 0.509 0840 1.189

Ig --;:; 750 I I
1.875 0051 0.372 0.671 0.946 0.00352
Carbon disulfide
'130 1.334 1.285 i 1.371 I 1.476 1.750

I
'175
I I
I
I, "I] {300 0.000 0.211 0.660
I 19 ~ 750
Chloroform
1.858
I 0.094 0.480 0.914
I 0.00519
'130 1.387 1.309 1.514
\
"il75
I I
"I] {300 0.000 0.324 0.792 1.261 i 1.670

Ethanol
19 ~ 750 I 1.878 0.149 0.601 0.986 I 1.311 0.00212
'130 i
1.324 I 1.496 1.552 1.884 2.286
'175
I I I

130
TABLE 59 (Continued)

Pressure, kg/cm 2
"fj300,
Compound
1000 4000 8000 12000 g cm- 1 sec- 1

I
~ {300 0.000 0.401 1.230 2.091 I
n-Amyl alcohol 19 ~ 7.1° I.736 0.107 0.776 1.437 2.007
j~Q 1.837 1.968 2.844 4.508
I 1)7:-, !

I -"'- {300 0.000 0.347

g 1)0 75° f.765 0.081 0.498
Benzene 0.00566
~ 1.718 1.845
"'17:1

I g~- {30°
0.008 0.253 0.867
1Jo 75° 1.814 0.053 0.563 1.146 000711
Chlorobenzene
~30
1.533 1.585 2.014
'17;')

~ {30° 0.000 0.262 0.897

Ig --;;;;- 750 1.801 0044 0.558 1.029 0.00985
Bromobenzene
"1/~o
1.581 1.652 2.183
'tJ73

I 'tJ {ZOO 0.000 0.376

g --;;;;- 75° f.551 1.847 0..160
Aniline 0.0319
~RO
2.812 3.381
'fJ7f)

I
~ {300 0.000
I I
Cyclohexane
I
I
19 ~ 750 I
I 1.723 0.169
I I 0.00828
I ~'O
I 1.892
I ~75
I I I I

I 19
~ {30°
~ 750
0.000
1.747
0.388
0.154 I
1.274
0.900
I
2.318
0.756
I
2.582
I
I
0.00639
Methylcyclohexane
~no
I 1.791 1.714 2.366 3.648
I
I
1)7;)
I I I I
I I
~ {300 I 0.000 0.274 I 0.897 1.699 I
19 ~ 750 ! 1.796 0065 0.597 1.186 1.832 I 0.00523
Toluene I
11'0 I I
i 1.600 1.618 1.995 3.258
1]75
: i
I
I I I
i I
{30°
11 I 0.000 0.311 I
I i
I Ig -:;; 750 f.767 0.057 0689 I 0.00709
a-Xylene
I ~30
1175
I 1.710 1.795 !
I
I "
I I

131
TABLE 59 (Continued)

Pressure, kg/em'
Compound Ig --'l and 17 30 1730°'
170 17 75 g em -I sec-I
1 1000 4000 8000 12000

m-Xylene
Ig ~
~ e oo
750
0.000
\.799
0290
0.079
0.967
0.637 1.333 0.00552
_'b> 1.589 1626 2.638
I "fJn i i
I "I {30°
I --
0.000
g ~o 75° 1.797 0.092
p-Xylene 0.00568
I "'lao 1.596 i
'~7;) i I
I I I

"I {30° 0.000 0.333 1.194

Ig ~~ 750
p-Cymene 1.800 0.087 0749
I 1.612
'130
1.585 1.762 2.786
717:)
I I I
I ..21.... poo 0.000 0.300 0.930 1.590 2.180
g "10 175° 0.560 1.060 1.490
Petroleum ether
'1:0
I I
2.34 3.390 4.900

,
"175
I I
Ig -"I- {30°
0.000 I 0.460 1.710 I,
;
Kerosine
"fjo 75° 0.910 1.880 2.800
~oo I 4.300
1175
I
! I i

TAGLE 60. Relative viscosity of water /40/

t,oC
Pressurep,
kg/em' 10 ..3 20 75

1.00J 0.779 0.488 0.222

500 0.938 0.775 0.500 0.230
1000 0.921 0.743 0.514 0.239
1500 0.932 0745 0.530 0.247
2000 0.957 0.154 0.550 0.258
30110 1.024 0.791 0.599 0.278
4000 1.111 0.842 0.658 0.302
flOOD 1.218 0.908 0.740 0.333
COOO 1.347 0.981 0.786 0.367
7000 1.064 0.834 0.404
8000 1.152 O.9n 0.44.')
9000 Solidified 0986 0.494
10000 1.058
1100:) I 1.126
.-~~~----- . _ - - - - - _.._---

132
Chapter IV

THE VISCOSI7'Y OF GASES AT ATMOSPHERIC PRESSURE

1. The viscosity of gases at atmospheric pressure as a

function of temperature

We have alr'ead:" shown that according to equation (5) the gas viscosity
is proportiona:. to the square root of the Kelvin temperature. However, the
experimental data do not quantitatively confirm this relationship, Sutherland
/101/ proposed that not only the elastic collision of molecules but also their
attraction should be considered. Let us imagine two molecules moving
near one another. If no attractive forces are present the molecules do not
collide. On the other hand, the attractive forces alter the path of the
molecules so that the molecules may collide. Thus, the number of
collisions of each molecule per unit time will increase. The free path of
each molecule will decrease because of these forces, and the decrement
will increase with decrease in temperature. At very high temperatures
the molecules move as if no attractive forces exist. In this case the
kinetic energy of the molecules is much higher than the potential energy
of the attractive forces. Hence, the mean free path of the molecules is a
function of the temperature, and is analytically expressed as follows:

[_ 1
(102)
- V2 N1td 2 ( 1 + i-) .
By combining (5) and (102), at T oK we obtain

kpc
'1JT=
V2N1td 2 1+
(C ) . (103 )
7
Thus, at O°C we have

(104)

and from (103) and (104)

C
~rr 1+y-
'1JT='1JO V Tn' ___0 , (105 )
To 1+-.f.
T
where C is a constant characteristic of every gas. However, the experi-
mental studies indicate that C is not constant but a function of temperature.

133
It increases with increase in the attractive forces, so that C is dependent
on the constant a of the Van der Waals equation, and the heat of vaporization.
In the majority of cases C increases with increase in the boiling point. To
determine C from experimental data, the Sutherland equation is used in the
form

T~2 = d~ (T+C), (106 )

AM2

where A = 2.7'10- 21 is the constant of the classical kinetic theory. If we

plot T'1.f"IJ against temperature, we obtain a straight line. The intercept
of the line on the temperature axis is equal to C. Thus, C values obtained
in this way are very dependent on the error in the determination of 1'). A
1 % error in the determination of 1') leads to an error in the determination
of C of up to 100/0. Consequently, we find differi.ng C values for the same
gas. In spite of the fact that C can be considered constant over a limited
temperature range, (105) is widely applied and it is a sufficiently accurate
interpolation equation for certain temperature ranges.
Table 61 shows the C values for numerous gases. It also indicates the
temperature range over which these values can be used.
Attempts have been made to derive an analytical expression for C of the
Sutherland equation. Rankine /102/ reports that C is approximately
proportional to the critical temperature

(107 )

Vogel /103/ gives the equation

(108 )

according to which C is proportional to the boiling point of the substance

at 760mm Hg.
Though these rules are empirical and approximate, they can be used
for an orientative calculation of C.
It follows from (106) that C can be calculated with the degree of
accuracy with which d and 1') are known; d can be approximately determined
from
1
d= 10- 8 • v 3 , (109)

where v is the molar volume at the boiling point (760 mm Hg). The law of
corresponding states theoretically justifies this equation (according to this
law the molecules of all substances occupy the same part of the overall
volume of the liquid when the reduced temperatures and pressures are
the same).

134
fABLE 61. COLstant C of the Sutherland equation for different gases

No. Gas C Temperature range, °c

Helium • . • • • • • • • • • • • . • • • • • • • • • • • . • . 83 100 - 200

95 200- 250
173 682 - 815
2 Neon ••••••••••••.•••••••••••••••• 61 20 -100
70 100 - 200
82 200 - 250
128 686 - 827
3 Argon ••••••••••••••••••••••••••••• 142 20-827
4 Krypton ••••••••••••••••••••••••••• 188 0-100
5 Xenon ••••••••••••••••••••••••••••• 252 0-100
6 Hydrogen •••••••••••••••••••••••••• n 20 -100
86 100 - 200
105 200- 250
234 713 - 822
Nitrogen ••••••••••••••••••••••••••• 103.9 25-280
104.7 20 - 825
Oxygen •••••••••••••••••••••••••••• 126.6 20 - 280
125 15 - 630
9 Air •••••••••••••••••••••••••••••• 106.8 20- 280
III 16- 825
10 Carbon monoxide ••••••••••••••••••••• 101.2 22- 277
11 Nitric oxide .............................................. .. 128 20 - 250
12 Carbon dioxide •••••••••••••••••••••• 254 25- 280
218 300 - 824
13 Steam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673 100 - 350
14 Sulfur dioxide •••••••••••••••••••••••• 306 300- 825
15 Nitrous oxide .............................................. .. 260 25- 280
16 Ammonia •••••••••••••••••••••••••• 503 20- 300
17 Methane •••••••••••••.••••••••••••• 164 20- 250
162 20- 500
18 Ethane •••••••••••••••••••••••••••• 252 20- 250
19 Ethylene •.••.•..••••••••••••••••••• 225 20- 250
20 Chlorine ..••••.•••••••••••••.••••.• 351 20- 500
21 Bromine ..•••••••••••••••••••••••••• 533 190 - 600
22 Iodine .. " ................................................. .. 568 106 - 523
23 Hydrogen chloride •••••••••••••••••••• 360 0- 250
24 Hydrogen bromid e .•••••••••••••••••••• 375 0-100
25 Hydrogen iodide •••••.•••••••••••••••• 390 0-100
26 Phosphine •••••••••••••••••••••••••• 290 0-100
27 Chloroform ••••••••••••••••••••••••• 373 121- 308
28 Methyl chloride •••••••••••••••••••••• 441 20- 308
29 Methyl bromide •••••••••••••••••••••• 379 20-120
30 Methylene chloride •••••••••••••••••••• 425 22- 309
31 Carbon tetrachloride ••••••••••••••••••• 335 50 - 250
365.4 128-315
32 Hydrogen ,;ulfide •••••••••••••••••••••• 331 0-100
33 Carbon disulfide •••••••••••••••••••••• 499.5 114- 310
34 Thiophen •••••••••••••••••••••••••• 407 20-245
35 Methylthiophen •••••••••••••••••••••• 400 50 - 250
36 Arsine ••••••••••••••••••••••• , •••• 300 0-100
37 Hydrogen cyanide ••••••••••••••••••••• 901 20-330
38 Cyanogen •••••••••••••••••••••••••• 330 0-100
39 Pyridine 320 98- 265
40 Mercury ••••••••••••••••••••••••••• 942 220- 610

135
TABLE 61 (Continued)

No. Gas C Temperature range, 'c

41 Mercuric chloride .' •••••••••••••••••••• 947 255 - 587

42 lv:ercuric bromide .................................... .. 657 219 - 582
43 Mercuric iodide ••••••••••••••••••••• 717 282 - 512
44 Zinc ••••••••••••••••••••••••••••• 876 600 - 680
45 Tetrachlorozincates ••••••••••••••••••• 432 107 - 592
46 T etrabromozincates ••••••••••••••••••• 525 105 - 578
47 Cadmium •.. '" •••••••••••••••••••• 105.~ 506 - 627
48 Propane ••••••••••••••••••••••••••• 278 20 - 250
290 25- 280
49 n-BUlane •.••••••••••••••••••••••••• 377.1 20-120
50 Isobutane •••••••.•••••••••••••••••• 368 20-120
51 n-Pentane •••••••••••••••••••••••.•• 382.8 122- 306
52 n -Hexane •••.•••••••••••••••••••••• 436.1 121 - 307
53 n-HeptaEe •••••••••••••••••••••••••• 445 100-252
54 n-Octene •••••.••••.••••••••••••••• 337 100 - 250
55 n-Nonane •••••••••••.•••.•••••••••• 276 100 - 250
56 Propylene ••..•.••••.••..••••.•••.•• 321.6 20-120
57 ex - Butylene •••••••••••••••••.••••••. 328,9 20 -120
58 B - Butylene ••••••••••••••••••••••••• 362,1 20 -120
59 Isobutylene ••••••••••••••••••••••.•• 339 20 -120
60 Isoamylene ••••.•••••••••••••••••••• 368 20 -120
61 Acetylene ••••••••••••••••••••••.•• 198.2 20 -120
62 Cyclopropane ••••••••.•.•••••.•.••.•• 372 20 -120
63 Cyclohexane •••••••••••••••••••••••• 350,9 122 - 306
64 Benzene ••••••••••••••••••••••••••• 447.5 130 - 313
65 Toluene ••..••••.•••••••••••••••.•• 370 60 - 250
66 Mesitylene ••••••••••••••••••••••••• 136,3 100 - 200
67 Diphenylmethane ••••••••••••••••••••• 387 166 - .360
68 Methanol •••••••••••••••••••••••••• 486,9 111 - 312
69 Ethanol •••••••••••••••••••••••••••• 407.3 120 - 309
70 n-Propanol ••••••••••••••••••••••••• 515.6 122 - 273
71 Isopropanol ••••••••••••••••••••••••• 459.9 119- 308
72 Acetone ••••••••••••••••••••••••••• 541,5 119- 306
73 Methyl acetate. , • , , • , • , , ••• , •• , , ••• , • 501,8 143 - 307
74 Ethyl acetate ,., ••• " . , . " ••• " , •• ' , 504 128 - 314
75 ,Dimethyl ether, •• , •• , •• , •• , ••• , •• ,. , • 345 20 -120
76 Diethyl ether •• " ••• " , •• " , ••• , •• ,. 404 122 - 309
77 Diphenyl ether, •••••• , , ••• , • , • , •••• , • 400 176 - 362
78 Acetic acid ••• , •• , . , • , , ••• , • , • , ••• , , 725 93- 250

By eliminating d from (106) and (109), and solving the equation for C, we
obtain

1 3

C=2.7. 10-5M2~2 -T. (110)

~V3

Rebganum /3/':< has another theory for the decrease in the mean free
path of molecules with decrease in temperature. According to him, the
decrease is due to the formation of "swarms" of molecules. Under equal

• [Reinganum, M. - k Jager G. Forschritte der Kinetischen Gas Theorie, p.85. Braunschweig. 1919.]

136
conditions the molecular density is higher at sites of lower potential energy,
i. e., at sites where the attractive force acts. The density increases
,
according to e -kT where (- e) is the positive work of the attractive forces.
If we imagine a stationary molecule, in its vicinity the potential energy of
any molecules attracted by it will be lower, the density will be greater,
and "swarms" of molecules will be formed. With increase in temperature,
the formation of swarms will be decreased, since with increase in
temperature the thermal fluctuation motion destroys, the swarm and equal-
izes the density (with increase in temperature e-TI' tends to unity). Accord-
o
1
ing to Reinganum, 1= n2N1td should be used instead of N in NeT, where C
2

is a constant characteristic of every gas. Thus, for the mean free path
we have

1= -----,0". (111 )
l'2N1td'eT

For the viscosity at T oK

''IT= 0 ' (112 )

nN1td 2eT

and at DOC

kpoCo
"lo = ---'-"-"--co" • (113)
v'2 N1t d2/T,
From the laGt two equations we obtain

(114)

In scientific literature we frequently find the following expression for

the temperature dependence of the viscosity at atmospheric pressure:

(115 )

where "lo is the viscosity of the gas at DOC, and m is an experimentally

determined constant. Vargaftik /83/ reports m values for many gases
(Table 62).

137
TABLE 62, Values of 1)0 and m for some gases

Substance

Acetone .................................................... .. 0,70 1,03

Air"",,""""""""""'" , 1. 75 0,683
Ammonia, , , , , , , , , , , , , , , , " , , , • , , , , 0,954 1,06
Amyl alcohol. , , , , • , , , , , , , , , , , , , , , , , , 0,632 0,96
Argon , " , " ' " . , ' , ' , " " " , " " " 2,16 0,72
Benzene .................................................. .. 0,712 1,00
Butane, , , , • , , , , , , , , , , , , , , , , , , , , , , , 0.697 0,97
Butanol, , , , . , , , , , , , , , , , , , , , , , , , , , , , 0,673 0,98
Carbon dioxide, , , , , , , , , , , , , , , , , , , , , , , 1.43 0,82
Carbon monoxide , , , , , , , , , , , , , • , , , , , , , 1,69 0.695
Carbon tetrachloride, . , , , , •, •, , , , • , , , , , 0,942 0,92
Chloroform , , , , , , , , • , •• , , • , , , , , , , • , , O,n81 0,94
Cyclohexane , , , , •• , , , , , , •, , , , , , , , , , , O,G51 0,907
Ethane, . , , , , , , , , , , , , •• , •• , , , , , , , • , 0,877 0.90
Ethanol, , • , , , , , , , , , , , , , •, . , , , , , • , , , 0,800 1,02
Ethyl acetate, , , , , , . , , , •, , , , , , , , , , , , , 0.705 1,01
Ethyl ether ," , , , , , , , , •••• , , , •• , , , • , 0,699 0,97
Helium."""." •• , . , .• " " , ••• , , . 1,880 0,68
Heptane, , , , , , , • , , •••••. , •• , , .••••• 0,1535 1,05
Hexane. , , , • , , , , ••• , , , , , , , , , , , , , • , , 0.602 1,03
Hydrogen ,," , •• , •• , ••• , • , , , , , , • , , , 0.852 0,678
Krypton, , , , , , , , • , , , , , , , . , , • , , , , , , , , 2.39 0,83
Methane, , • , , •• , •.• , • , • , , , , , ••• , , , , 1.06 0,76
Methanol, ••.. , , • , , .•• , •• , , , • , , , , , , 0.901 1,04
Methyl bromide , , •• , , , • , , • , •• , , , , , , , , 1. 25 1.05
Methyl chloride, , , , , • , , , , , , , , , , •.• , , , 1.00 1,02
Neon""" " • ,,' • "., " , . , . , .,'" 3.03 0.65
Nitrogen. , , , , , • , , , , , , , , , , , • , , , •, , •, 1.70 0,68
Octane ••• , , , , , , , , , , , , , •, , • , , , . , , , , 0,493 1,02
Oxygen, . , , , , , , , •, , , , , , . , •• , , •. , , • , 1.98 0,693
Pentane " , " ' , ' , . , " , " ' . " , •• , ' , ' 0,648 0,99
Propane ................................................... .. 0.765 0,92
Propanol, • , , • , , , • , , , , , , • , • , , , , • , , , , 0,731 1,00
Steam " ' , " ' , " ' . , ' " . , , , . , •• ,.,' 0,84 1,20
Sulfur d iox ide , ' . , . , , " , ' • , • , • , •• , , , , 1,23 0,912
Toluene. , , , , , , , , , , , •••••• , , , , , , , , , 0,674 0,89
Xenon ................................ ~ .................... .. 2,15 0,89

The equations for the viscosity coefficients derived from the kinetic
theory of gases are not quite accurate because they are based on the mean
velocity distributions of the molecules of gases instead of on the true
distributions, The best general equation of the kinetic theory for the
viscosity (and also for other phenomena of molecular transport, viz"
thermal conductivity and diffusion) and its temperature dependence can
be obtained if we consider the velocity distribution of the gas molecules
to be the result of collisions, and of the action of outside forces. The
distribution function is then expressed by an equation first formulated by
Boltzmann /104/:

of +~ of +~ of +1- of +u of +v of +W OF =
ox m ou m ov m ow ox oy oz
(116)
=-/ JJ(FF,-F'F~)dOl,d2Crcos&sin&d&d'f'
138
where X. Y. Z are the components of the outside force acting on the
molecule m; Cr is the relative veloc:.ity; and d is the diameter of the mole-
cule. The magnitudes with subscript 1 refer to molecules which collide
with the molecules under consideration (without subscript). and the
magnitudes with a prime refer to the same magnitudes after collision. The
equation is integrated over all the possible velocities of the colliding
molecules and all the possible velocity orientations. However. the inte-
gration is very difficult. and up to the present it has been carried out for
some simplified molecular models only /105. 106. 107/. The integral on
the right-hand side of (116) describes the change in state of the particles
induced by collision (" impact term" or "collision integral"). The equation
can be made to correspond more to actual cases if we consider molecule
interaction over large distances besides the elastic collisions. We do not
yet known any law which accurately defines molecule interaction. but a
number of equations have been proposed which give results agreeing with
the experimental data. The best molecular model proposed is the
Lennard -Jones potential well model (the Lennard -Jones potential is
expressed as P=xr-n-x'r- n', where the first term characterizes the
attractive forces. and the second the repulsive forces). The parameters
of this potential are usually selected so that the potential energy of the
interaction of the two molecules E (r) separated by a distance r is given by

E(r)= 48 W: r -( ~rJ· (117 )

where s is the minimum value of the potential energy, and ro the distance
between molecules when the energy of interaction is equal to zero.
Hirschfelder. Bird, and Spotz /108. 109/ used this model to calculate
the coefficients of molecular transport of gases and gas mixtures. All the
calculated magnitudes were tabulated and can be used in practice.
The authors give the following equation to calculate the viscosity
coefficient:

(118)

where "I is the viscosity coefficient in 10- 7 gcm- 1 sec- 1 ; T is the Kelvin
temperature; :Q(2.2) is the reduced collision cross -section integral; T* is the
reduced temperature given by

T'=!!..< T. (119 )

In (119) k is the Bolzmann constant, and f~ is a slowly varying function

of T*. The values of f~ are shown in Table 63. It can be seen that f~ devi-
ates from unity by less than 0.8% over a wide range of T*. To determine the
viscosity coefficient from (118). we must known i and rD. These values
were found by Hirschfelder et al. from experimental viscosity data. They
had to solve (118) for 7i and fo at two temperatures at which the viscosity
::oefficients are known. These values found as above for some substances

139
are shown in Table 64. The reader will see that for some substances
different i and ro values are reported. This is because the starting data
for "I differed.

TABLE 63. Values of f .. for different values of T"

T* I,. T* f"

0.3 1.0014 3.0 1.0034

0.5 1.0002 4.0 1.0049
0.75 1.0000 5.0 1.0058
1.0 1.0000 10.0 1.0075
1.25 1.0001 50.0 1.0079
1.5 1.0004 100.0 1.0080
2.0 1.0014 400.0 1.0080
2.5 1.0025

The following equations can be used for an approximate calculation

of the constants:

(120)

(121 )
/1
ro = 0.833 (ve~S , (122 )

where Tb is the boiling point (in oK) at 760mm Hg; Ter and verare the
critical temperature (in oK) and critical volume (in gj cm 3 ), respectively.
The error involved in the determination of i leads to a much smaller
error in the determination of the viscosity coefficient. Thus, an error of
100/0 in the determination of i- gives a 1.50/0 error in the determination of "I,
and the absolute magnitude of the viscosity changes by a factor of two.
Furthermore, to determine "l from (118), we must know Q(2.2). The
collision integrals were calculated for different values of T* by Hirschfelder
et aI., and are given in Table 65.
Example of calculation. Let us determine the viscosity coefficient
of methane at 450 o K. From Table 64 we find for methane

c
,,=137°K °
andro=3.882A.

The reduced temperature is equal to r = ~~~ = 3.28. For this temper-

ature we find from Table 65 Q(2.2)(r) = 1.016, and from Table 63 f~ = 1.004.
Thus, the viscosity coefficient is given by 1] =266.93· ({~:2~2\~~~6·1.004 =
= 1487'10- 7 gcm sec.

140
TABLE 64. Constants • and ro calculated from 1} data
Ii

o
Gas Gas Til' A

Ie 10.22 2.576 F. 112 3.653

} I. 33.3 2.968 Cl. 357 4.115
} I. 38.0 2.915 Br. 520 4.268
[
N'e
'. 39.3
35.7
2.948
2,789
I.
HgI.
550
698
4.982
5.625
l\ e 27.5 2.858 HgBr. 530 5.414
A.r 124 3.418 Hg 851 2.898
Ar 116 3.465 SnBr. 465 6.666
Kr 190 3.61 SnCl. 1550 4.540
Xe 229 4.055 C 2N. 339 4.38
A ir 97.0 3.617 COS 335 4.13
A ir 84.0 3.689 C.H. 185 4.221
NT
• 91.5 3.681 C2 H. 205 4.232

o••
N 79.8 3.749 C.Hs 230 4.418
113 3.433 CaHs 254 5.061
o2 88.0 3.541 n-C.Hl0 410 4.997
CO. 190 3.996 iso -C.Hl0 313 5.341
C O. 213 3.897 n-C"H 1• 345 5.769
N,,0 II 220 3.879 n-CeH14 413 5.909
N.,0 237 3.811 n-C SH 1S 320 7.451
C H. 137 3.882 n-CgH. o 240 8.448
CIi. 144 3.796 C 6Hs 440 5.270
CCl. I 327 5.881 Cyclohexane 324 6.093
SO. 252 4.290 C.HsOH 391 4.455
I

TABLE 65. Values of the reduced collision cross-section integral

L:(2,2) for different values of T·

£(2,2) @(2,2) 2(2,2) r· 2(2.2)

T*
II
T*
I T*

o.ao 2.785 1.35 1.375 2.8 1.058 4.9 0.9305

1.048 5.0 0.9269
0.35
0.40
2.628
2.492 I 1.45
1.40 1.353
1.333
2.9
3.0 1.039 6.0 0.8963
0.45 2.368 1.50 1.314 3.1 1.030 7.0 0.8727
0.50 2.257 1.55 1.296 3.2 1.022 8.0 0.8538
0.55 2.156 1.60 1.279 3.3 1.014 9.0 0.8379
0.60 2.065 1.65 1.264 3.4 1.007 10.0 0.8242
0.65 1.982 1.70 1.248 3.5 0.9999 20.0 0.7432
0.70 1.908 1.75 1.234 3.6 0.9932 30.0 0.7005
0.75 1.841 1.80 1.221 3.7 0.9870 400 0.6718
0.80 1.780 1.85 1.209 3.8 0.9811 50.0 0.6504
0.85 1.725 1.90 1.197 3.9 0.9755 60.0 0.6335
0.90 1.675 1.95 i 1.186 4.0 0.9700 70.0 0.6194
0.95 1.629 2.0 1.175 4.1 0.9649 80.0 0.6076
1.00 1.587 2.1 1.156 ,I 4.2 0.9600 90.0 0.5973
1.05 1.549 2.2 1.138 , 4.3 0.9553 100.0 0.5882
1.10 1.514 2.3 1.122 4.4 0.9507 200.0 0.5320
1.15 1.482 2.4 1.107 4.5 0.9464 300.0 0.5016
1.20 1.452 2.5 1.093 4.6 0.9422 400.0 0.4811
1.25 1.424 2.6 1.081 4.7 0.9382
1.30 1.399 2.7 1.069 4.8 0.9343

141
 Hirschfelder et al. /109/ calculated the coefficients of viscosity,
thermal conductivity, and diffusion for polar gases also. They found the
following formula for the potential energy of the molecule interaction:

E(r) = 40[( ~)12 - (!f )6J __ 2~, (123)

where I-" is the dipole moment.

Thus, the viscosity coefficient for polar gases is given by

V MT (124 )
"i = 266.93 20(2,2) * *
ro" To

where =21 3"'~ 'Y) is in 10- 7 g cm- I sec-I.

o· Era
The constants '0' i,
f1 and 0' for some polar gases are reported in
Table 66. The values g(2.2) ( T*) (0*) for different'values of T* and 0' were
tablulated by the authors.
Example of calculation. Let us determine "'I for NHs at 800 o K.
From Table 66 we find

i= 146.8°K; ro=3.441oK; 0*= 1.2499;

T* = 181~8 = 5.45.

From Table 67 g(2.~)( T*) (0*)=0.9388. The viscosity coefficient is

therefore

. 107 = 266.93V17.082. 800 = 2803 g em-I sec-I.

"'I (3.441 )2(0.9388)

We calculated "I for many gases, and some of the results are given
below. We concluded that the viscosity coefficients calculated from (118)
and (124) agree well with the experimental data. The agreement is better
than that obtained with the Sutherland equation .

TABLE 66. Values of the constants • ro,

Ii' l'- and 6' for some polar

gases

Gas ~Debye +,01\ Tu' A 6*

NO 0.0672 112.4 3.508 0.003327

CO 0.1172 109.9 3.585 0.00983
CHCla 1.05 415.2 5.117 0.07177
H2 S 0.931 221.1 3.733 0.2739
HCI 1.034 218.0 3.506 0.4123
S02 1.611 191.4 4.341 0.5995
NOCI 1.83 205.2 4.332 0.7272
CH,CI 1.861 243.6 4.076 0.7606
CH,OH 1.680 194.9 4.082 0.7715
CHgCOCH, 2.85 158.4 5.485 1.1253
NH, 1.437 146.8 3.441 1.2499
H 2O 1.831 230.9 2.824 2.333
----,

142
TABLE 67. The reduced collision cross-section integral g(2.2) of polar gases for different values of r' and S'

~*IT* '-_
0.00 0.25 0.50 0.75 1.()() 1.25 1.50 1.75 2.00

1.0
1.2
I 1.5938
1.456;,
1.7578
1.5872
1.9312
1.7302
2.0806
1.8661
2.1963
1.9881
2.3025
2.1035
2.4920
2.2749
2.5916
2.3782
2.6199
2.4257

1.4 1.355'7 1.4622 15812 1.7026 1.8224 1.9395 2.0945 2.2051 2.2713

1.6 1.2800 1.3688 1.4689 J..5756 1.6878 1.8019 1.9430 2.0599 2.1413

1.8 1.2216 1.2965 1.3813 1.4739 1.5759 1.6843 1.8128 1.9339 2.0263

2.0 1.1751 1.2390 1.31l:2 1.3907 1.4817 1.583'2 1.6999 1.8224 1.9217

2.2 1.1377 1.1923 1.2539 1.3220 1.4026 1.4957 1.6015 1.7232 1.8258

2.4 1.1066 1.1537 1.2063 1.2643 1.3.349 1.4198 1.51.5S 1.6343 1.7373

2.6 1.0803 1.1212 1.1662 1.2155 1.2769 1.3537 I.HOI 1.5544 1.6559

2.8 1.0579 1.0935 1.1321 1.1737 1.2270 1.2961 1.3738 1.4827 1.5812

3.0 1.0385 1.0696 1.1026 1.1377 1.1837 1.2456 1.3153 1.4182 1.5126

3.2 1.0214 1.0488 1.0769 1.1065 I.I 460 1.2012 1.2636 1.3601 1.4499

3.4 1.0063 1.0304 1.0544 .1.0792 !.l129 1.1621 1.2178 1.3078 1.3926

3.6 0.9928 1.0141 1.0345 1.0552 1.0838 1.1274 1.1770 1.2606 1.3401

3.8 0.9807 , 09996 1.0168 1.0339 1.0581 1.096.1 I 1.1406 1.2179 1.2922

4.0 0.9696 0.986 4 1.0009 1.0149 1.0351 1.0690 1.1081 !.l794 1.2485

5 0.9265 0.935 9 0.9411 I 0.9446 0.9510 0.9674 0.9874 1.0334 1.0797

6 0.8960 0.901 0 0.9014 0.8994 0.8980 09037 0.9118 0.9401 0.9694
7 0.8725 0.8750 0.8727 0.8677 0.8618 0.8609 0.8614 0.8777 0.8946
8 0.8536 0.8545 0.8507 0.8440 0.8356 0.8304 0.8260 0.8342 0.8422

9 0.8378 0.8376 0.8330 0.8254 0.8154 0.8077 0.8000 0.8028 0.8043

10 0.8242 0.823 2 0.8183 0.8102 0.7994 0.7901 0.7802 0.7793 0.7760

11 0.8123 0.810 8 08058 0.7975 0.7863 0.7760 0.7648 0.7613 0.7544

12 0.8017 0.799 9 0.7948 0.7865 0.7752 0.7644 0.7523 0.7471 0.7376
13 07922 0.790 2 0.7851 0.7769 0.7656 0.7546 0.7421 0.7356 0.7243
14 0.7836 0.781 4 0.7765 0.7684 0.7572 0.7462 0.7334 0.7262 0.7134

15 0.7756 0.773 4 0.7686 0.7606 0.7497 0.7388 0.7260 0.7182 0.7045

16 0.7883 0.766 0 0.7613 0.7536 0.7429 0.7322 0.7195 0.7114 0.6971
32 0.6939 0.6924 0.6897 0.6846 0.6782 0.6717 0.6637 0.6569 0.6462
64 0.6262 0.625 9 0.6251 0.6213 0.6181 0.6151 0.6114 0.6078 0.6033
128 0.5634 0.563 8 0.5640 0.5612 0.5593 0.5579 0.5563 0.5547 0.5526
256 0.5056 0.506 1 0.5065 I 0.5047 0.5035 0.5028 0.5021 0.5014 0.5006
512 0.4528 0.453 3 0.4539 I 0.4527 0.4518 0.4515 0.4512 0.4508 0.4505

:2. Application of the law of corresponding states to the

c.etermination of the temperature dependence of the gas
viscosity at atmospheric pressure

Golubev and Petrov /63/ attempted to find a simple general relationship

for the temperature dependence of the gas viscosity. They used the law of

143
corresponding states, and assumed that at the same reduced temperatures
't ~,
= Lcr the reduced viscosity 'Ylt = "IT must be the same. If we plot 't on one
~

coordinate and 'Ylt on the other, the values for all gases must lie on a
single curve.

FIGURE 70. The viscosity of sa tura ted hydrocarbons.

2..5
2. 1-
I -
2..3 I--~-
-
--
2. 2
2.I
2..0. ... 1
1.9 I
~I. 8 ,

.iii.7(}
'il
';ff..5
I
/:
- i
4-
<t1 , I

.
.3
1.2 ---- General curie for
d>' .L-Pt? -~ated hydrocar ODS
. ;
I
I.I
I.0. £f' f--~O -Ethylene
, 0

:i=~m
J'
a9 ki" f-+ J "'oo£y',=
x - utrlene
0.'.8
f---
(t-

a7 I
-+-;--
I
!Ai"
__ 0 -
." +:-j --;, "~ =:_cyclo
Acety ene
a.(J '!.it I:1etliylacetylene
a.5.9 ,
·---F -lli ropanef--f--
o _L
a3
0..2 I
1-,
. --cyclohexane
Benz ene
, 1 -i • -Toluene
I
0

f--f--
I--+-
-l-

.I
8. 00. 10. f.2 16 [8 20. 22 24 26 ZB 3D
r" 1/Tcr
FIGURE 71. The viscosity of unsaturated hydrocarbons.

The law of corresponding states accurately reproduces the experimental

data for any group of components. Therefore, we first applied the law of
corresponding states to saturated and unsaturated hydrocarbons, and then to
all other gases. The results are plotted in Figures 70,71,72, and 73. The

144
data of Table 4 were used to plot the curves. The critical temperatures T cr ,
the viscosity coefficients Dcr at this temperature and at 760mm Hg, are
shown in Table (i8. The values for Dcr were determined from experimental
data by graphical and analytical interpolation. In some cases we also used
extrapolation.

TABLE 68. Boiling point and critical constants /64, 100/ of different substances
-
Mole- Critical Critical
Bp at 1)cr experi- 1) cr calculated
cular temper- pressure,
rmula 760· mm mental /63/, from (129),
weigh! ature ICr. intern. 10-' gcm-1sec-1 10-' g cm-I sec -I
M
Hg
'Cft °C atm.

I I I
!
Helium He 4.002 1-268.93 - 267.9 2.26 116 88
Neon Ne 20.183' -246.1 -228.7 46.86 735 750
Argon Ar 39.944 -185.9 -122.4 5KO 1243 1267
Krypton Kr I 83.7 -153.2 -63.0 54.2 1850 1885
Xenon Xe 1131.3 -108.8 14.7 57.2 2225 2310
Hydrogen Hz I 2.0156: - 252.78 -239.9 12.8 173 152
Nitrogen N i
28.016 ! - 195.81 -147.1 33.5 860 859
Oxygen Oz I 32.0 - 182.97 -118.8 49.7 1160 1155
Air - I 28.96 : - 193.0 -140.7 37.2 922 931
Carbon monoxide CO 28.01 ,-191.5 . - 140.2 34.5 874 870
Nitric oxid e NO i 30.008 -152.0 ' -94.0 64.0 1240 1290
Carbon dioxide CO 2 i
44.01 - I 31.1 ,
73.0 1520
2260
1550
Steam IzO 18.0156 100
1
374.2 217.5 1825
Sulfur dioxide S02 64.06 -10.0 157.3 77.8 1880 1860
Nitrous oxide ~1 0
2 44.016 -88.7 36.5 71.6 - 1540
Ammonia ~ IHs 17.031 -33.4 132.4 111.5 - 1225
Methane C H. 16.04 , - 161.4 -82.5 45.8 743 748
Ethane C2 H6 I 30.07 -88.6 32.3 48.2 950 980
Ethylene CzH. I 28.05 -103.5 9.7 50.9 975 994
Chlorine CI 2 170.914 -33.95 144.0 76.1 1857 1930
Bromine Br2 1159.832 58.78 302.0 - 2875 -
Iodine I. 1253.864
I
184.35 514.0 - 3580 -

Hydrogen chloride HCI 36.465 -85.0 52.3 8.6 - 1565

Hydrogen bromide l-IBr 80.924 -67.0 91.0 84.0 2305 2260

Hydrogen iodide HI 127.94 -36.0 150.8 82.0 2630 2730

Phosphine p Hs 34.04 -87.5 51.0 64.0 1265 1246

Chloroform C HCI. 119.39 61.2 260.0 54.9 1787 I 1870
Methyl chloride C HsCI I 50.49 -24.0 143.1 65.9 1520 1485
Methyl bromide C Hsllr
1
94 94
.
4.6 194.0 - 2125 II -
,
Carbon tetrachloride CCI. 1153.84 76.75 283.1 45.0 1770 1910
Hydrogen sulfide 1-1 2S 34.08 -60.4 , 100.4 89.0 1610 1520
Carbon disulfide CS. i 76.12 46.26 273.0 72.9 1845 1864

~~:~~
Carbon oxysulfide COS I
-48.0 105.0 65.1 1575 1630
Propane CHH8 ! I
-42.6 96.8 42.0 999 1050
n-Butane C4H IO 58.12 +0.5 153.0 36.0 1080 1060
Isobutane C 4H l0 I 58.12 ! -10.2 i 134.0 37.0 1030 i 1088
n-Pemane CoHl2 72.14 36.1 I
197.2 33.0 1077 1096
I
n-Hexane C6HJ,! , 86.17 68.73 i 234.8 29.6 1117 1100
n-Heptane C,H I6 ,100.19 98.4 266.85 26.8 1054 1100
n-Octane CSHl8 '114.22 125.6 296.2 24.7 1100
n-Nonane C 9H20 128.16 150.7 322.8 22.8 1100
Propylene C,H 6 ' 42.08 -47.0 92.0 45.3 1053 1080

145
TABLE 68 (Continued)

Mole-
Bp at
Itemper-
Critical Critical
1]cr eXperi- 1] cr calculated
cular pressure
Gas Formula 760mm I mental /63/, from (129),
weight H ature Per' in- 10-7 gcm-l seC l 10-7 gcm-l sec-l
M g fer,oC tern .. atm.

a -Butylene C 4 HS 56.10 -6.32 160.0 42.1 1120 1150

6 -Butylene C4HS 56.10 1.0 155.0 42.1 1090 1155
Methylacetylene C aH4 40.03 -23.23 128.0 52.8 1150 1148
Acetylene C 2H 2 26.04 -83.6 36.1 62.0 1060 1075
Cyclopropane CSH6 42.05 -32.9 124.6 54.2 1194 1197
Cyclohexane C6 H12 84.09 80.8 281.1 40.4 1260 1265
Benzene C6 B 6 78.05 80.12 288.6 47.7 1425 1415
Toluene C 7HS 92.06 110.65 320.6 41.6 1380 1390
Methanol CHsOH 32.D4 64.51 240.0 78.7 1695 1285
Ethanol C2 HsOH 46.07 78.3 243.1 63.1 1496 1339

n-Propanol CaH70 H 60.06 97.19 263.7 49.95 1410 1290

Isopropanol C SH 70H 60.06 82.42 235.0 53.0 1330 1355

Acetone C SH60 58.08 56.1 I 235.0 47.0 1300 1230
I
Methyl acetate C SH60 74.05 57.2 233.7 46.3 1395 1376
Ethyl acetate C 4Hs02 88.06 77.1.5 i 250.1 37.8 1338 1300
Dimethyl ether (CH S)20 46.05 -24.9 i 126.9 52.0 I
1248 1220
74.08 35.5 1167
Diethy1 ether C4H lO O 34.54 1 193.8 1160
Acetic acid C2H 40 2 60.03 1182 321.6 57.2 .- 1386
I

Figure 70 shows the temperature dependence of the viscosity of saturated

hydrocarbons in reduced temperature and viscosity coordinates. For all
the gases except methane, the data over the whole temperature range are
shown in Table 4, i.e., all the experimental points are plotted. However,.
intermediate values for some gases are omitted on the figures because the
experimental points are too close to one another.
Figure 71 shows the temperature dependence of the viscosity of un-
saturated and aromatic hydrocarbons. The dotted curve represents the
general curve for saturated hydrocarbons taken from Figure 70. It can
be seen that this general curve corresponds to unsaturated hydrocarbons
also.
The lh vs T curves for different gases are shown in Figures 72 and 73.
Here the dotted line corresponds to the general curve for all the gases.
We will term it the "reduced viscosity curve of gases."
It can be seen from the figures that the experimental data can be
correlated to the reduced viscosity curve of gases. The small deviations
are due to a) experimental errors; b) errors in the determination of the
critical temperature; and c) errors in the determination of the viscosity
at the critical temperature.
Figure 73 shows the TIT vs T curves for a wider temperature range.
The general curve of this graph can be used to calculate the viscosity
of any gas provided that Tcr and Tlcr, or T cr and the gas viscosity at any
temperature, are known.

146
 2. I
2.1 .'"
2,

1.9
,,"
1.8 ~
f.7 r
1.6
,,/' I
1.5
p.- I
,,-
1.4 -
1.3 J -
1.2 kFf ____ General curve for -
1.1 I l# saturated and unsat- -
1.0 L I
# urated hydrocarbons -
I
I
k¢'" 0-Chlorine
0.9
0.8 t ~
e -Chloroform
()_Iodine
-
-

t.,.
0.7 WI I o-Hydrogen iodide r-
I % ; ~ I , tI-Sulfur dioxide
0.6 , x-Phosphine r-
I I
0.5 -9-.Carbon disulfide r-
art "It/' I p-Hydrogen sulfide
I1-Carbon oxysulfide
r-
0.3 of-Carbon tetrachloride r-
Oo ;
#-Bromine r-
0.1
0.0
0.0 0.2 0.6 0.8 10 1.2 1.4 1.8 1.8 2.0

FIGURE 72. The vi,;cosity of various gases.

!J0
8.0

z0 ,e--lI
-- f-'"
,~

6.
........ -""
~
5.0
~.r-
.....",. --
.~
0
1;- l+.H" ____ - General curve for
0 all gases
~4P' +I-Helium
0 .~ t-Hydrogen

/. 0 .L a-Neon
A_Argon
.,-Nitrogen
/ a-Oxygen

0.5
/ I I,
m-Air
8-Carbon dioxide

i/ I I i
I
I
+-Methane
t-Ethylene
E1-Sulfur dioxide
a0
0.0 0.5 1.0 U M W M M W M M ~ ~ ~ m M 6 e " •
Z'= TITcr

fIGURE n. The viscosity of various gases.

147
TABLE 69. A comparison of the experimental viscosity coefficients with the calculated for hydrogen and
nitrogen

Hydrogen Nitrogen
t,Oc
experi- from from from (125) experi- from from from (125)
mental ( 105) (118) and (126) mental (l05) (l1S) and (126)
I
!, I I
-260 70 - 61.5 I
I
70.7 I
.- 250 122 - 104.5 122 !
I I
. 240 173 - 151 173 i I
. - 230 218 - 196 221 ! I
, i
.- 220 259 - 240 , 254
I
·-210 297 - 281 302 I, ,
! J

, I
- 200 333 - 320 336 ,
-190 367 - 358 368 ,i i
-180 399 - 393 400 i 650 632 660 643
i i
I
-170 430 - 426 429 ! 717 705 i 730 708
-160 460 - 458 458 783 772 795 I 773
-150 488 - 489 484 846 835 859
i 840
-140 516 - 518 512 907 900 924 i 904
-130 542.5 - 546 538 967 963 I 982 969
-120 568 - 574 562 1028 1025 1040 1033
-110 593 599 586 1082 1082 1100 1095
-100 618 596 626 611 1143 1141 1157 1151
-75 677 662 689 669 1285 1283 1292 1290
-50 733
I 1419
725 745 727 I 1419 1415 1426
-25 788 788 801 780 1542 1540 1550 1539
±O 840 840 855 832 1665 1665 1660 1655
20 880 884 895 872 17.55 175.'; 1752 1741
25 890 893 905 881 1778 1780 1775 1763
50 938 944 954 930 1883 1895 1875 1866
,
75 985.5 995 1000 976 1986 2000 1982 1965
100 1033 1038 1050 1024 2086 2105 2083 2064
150 1123 1135 1130 1116 2278 2300 2275 2262
200 1213 1230 1215 1204 2464 2490 2455 2434
250 1299 1340 1299 1290 2639 2670 2620 2610
300 1382 1425 1373 1371 2800 2830 2780 2773
400 1538 1528 1529 3118 3130 3100 3092
500 1686 1682 1678 3403 3410 3400 3397
600 1828 1820 1828 3665 3670 3665 3689
700 1965 2380 1948 1966 3916 3930 3920 3973
800 2103 2540 2080 2104 4160 4150 4160 4231
900 2230 2710 2200 2237
1000 2355 2335 2367

148
TABLE 70. A comparison of the experimental viscosity coefficients with the calculated for methane dnd
carbon dioxide

~
Methane Carbon dioxide

eXperi- from from from (125) experi-I from from Ifrom (125)
mental (105) (118) and (126) mental ( 105) (118) and (126)
!
-200 29.5 263 283 294
-190 336 302 320 334
I
-180 375 349 360 373 I
-170 414 377 399 411
-160 453 433 439 450
-150 490 475 478 488
-140 528 515 517 526
-130 566 554 556 564
-120 603 595 595 600
-110 640 633 634 638
-100 677 672 672 675 886 884
-75 769 765 764 767 1007 998 1005
-50 860 855 855 861 1127 1120 1127
-25 946 940 939 951 1247 2250 1246
+0 1026 1026 1024 1028 1367 1367 1370 1366
+20 1092 1090 1089 1092 1463 1465 1460 1464
25 1108 1108 1108 1111 1486 1485 1480 1488
50 1185 1185 1178 1184 1603 1610 1599 1606
75 1260 1261 1257 1256 1716 1730 1700 1724
100 [332 1332 1329 1300 1827 1830 1800 1842
150 1472 1470 1463 1452 2045 2050 2000 2055
200 1604 1610 1596 1574 2254 2265 2205 2265
250 1725 1735 1720 1690 2456 2465 2435 2455
300 1850 1850 1840 1805 2646 2635 2620 2637
400 2070 2065 2060 2015 2994 2895 2960 2965
500 22G8 2270 2270 2222 3309 3200 3290 3260
600 24[,5 2470 2464 2407 3605 3490 3590 3570
700 2650 2650 2650 2588 3876 3780 3880 3846
800 2820 2805 2825 2764 4140 4020 4150 4119
900 2980 2980 3005 2934 4400 4270 4420 4385
1000 3135 3130 3185 3106 4658 4500 4700 4636

149
 The next step was to find an analytical expression for the reduced vis-
cosity curve of gases. We obtained the following formula, which can be
used in engineering calculations:
For temperatures below the critical (T< 1)

or (125 )

and for temperatures above the critical (T from 1 to 25)

O.29T er
0.71 + 0.29 T )O.~l + - T -
'YJ't;='t' 't or "Ij ='Yl ( - (126)
T er, Ter

Trautz /110/ showed that 1'Jer can be represented in terms of the

molecular weight M, critical pressure Per' cri.tical temperature Ter and
the gas constant R. The following equation was formulated by Trautz:

(127)

"Ij M _
=k _ er
3p 4]"
1

[ (128)
er Ter

where k is a dimensional constant (including R).

The application of (128) to experimental viscosity data shows that the
mean value of k is 3.5'10- 6 , so that

'fAT 'f
"Ijer
=3.5.1O-6~
T'f, (129)
er

Here 1Jer is expressed in poise, Per in atmospheres, and Ter in Kelvin

degrees. The values of "Ijer calculated from (129) are shown in Table 68.
lt can be seen from the table that the calculated and experimental results
agree well for most substances.
If no experimental data on the critical parameters are available, the
parameters can be approximately calculated from the equations shown
in /111, 112, 113/.
Tables 69 and 70 give a comparison of the experimental viscosity
coefficients of some gases with those calculated from (105), (118), (125)
and (126). The agreement is good.
The same applies to other substances also.

150
Chapter V

THE VISCOSITY OF COMPRESSED GASES

1. The pressure dependence of the gas viscosity at

different temperatures

The theory of the gaseous state of substances at high densities (or

pressures) and different temperatures is far from complete. We have at
present no equation of state which will quantitatively and accurately
represent the behavior of real gases over wide ranges of temperature and
pressure. Of the numerous equations of state proposed, the Van del' Waals
equation is mOBt widely applied. However, at high pressures it gives only
a qualitative correlation between p. v and T of gases, and a qualitative
representation of variation in physical properties. Nevertheless, the
qualitative picture of the actual properties of gases is so correct that the
equations can be used to predict the behavior of the gases under experi-
mental conditions. For very rarefied gases only. an equation of state
satisfying the experimental conditions. that is the Boyle -Mariotte
equation, was found [in Russian text, the Clapeyron -Mendeleevl . (pv =RT)
The theory of compressed gases has been developed even less.
Numerous authors have proposed equations to give the pressure depend-
ence of the gas viscosity. Some will be discussed below.
The idea of Van der Waals on the influence of the geometric dimensions
of the molecules and the molecular interaction on the state and physical
properties of liquids and gases is the theoretical basis of most equations.
A theory on the viscosity of compressed gases was first suggested by
Enskog /4/. It applies to moderately high densities. According to this
author, the momentum is transferred in the gas from one layer to another
not only by the molecules whose centers pass through the plane separating these
layers (interface), but also by molecules whose centers after collision are
on both sides of this imaginary interfacial plane. Enskog proved that with
increase in pressure the ratio of the viscosity to density of the gases
passes through a minimum. given by

where p is the gas density; and d the diameter of the molecule. The
viscosity 17 is given by

-1P = 1
2.5 45 (-1)
Pmm
. (A
P
+ 0.800 + o. 7614bpk). (131)

151
and bpk is determined from the equation of state also proposed by Enskog

(132 )

Here a is a constant; b the quadruple volume of the molecules per g; and

k is a factor characterizing the probability that one molecule is found near
another. If the densities are not too high, k is determined from the
equation of Clausius and Boltzmann

(133)

Enskog (for CO 2 ) and Michels andGibson /32/ (for nitrogen) used a more
general equation of state

(134)

from which it is possible to determine bpk at dIfferent pressures and

temperatures from experimental pvT data. The equation used is

(135 )

It is derived directly from (132) and (134).

It is easy to see that all the magnitudes necessary for calculating the
viscosity coefficients from (131) can be obtained from (130) and (135). The
diameter of the molecules can be calculated from experimental viscosity
data, using the equation of the kinetic theory of gases (see equation (151)
below).
Tables 71 and 72 illustrate the application of the Enskog theory to
carbon dioxide and nitrogen. The experimental and calculated data agree
well.
Predvoditelev /114/ considered cyclic moHon of liquid and gas molecules
as the basis of his viscosity theory, and derived the following equation:

=A RT _v_ (136)
"fip,T Va V-b '

where u is temperature dependent only. Since, for gases at 1 atm ~-

v-b
= 1'

we have

RT
A Vu = "fiT (137 )

and hence

v
"fi p ,T='1]T v-b'

5680 152
 TABLE 71. Comparison of 1) values calculated from the Enskog equa-
ticn and values observed for carbon dioxide at 40.3·C
I I '1",
- em
lO-fg -1 sec _1
S ' 10-7 cm2 sec '"'1 ~
p, atm f,g/cm \ (f) exp I
I

'
I
I 9KCne p11M·1 calc.

45.3 0.100 18000 I 1800 1910

I
64,3 0.170 11500 1960 1990
75.9 0.240 9080 2180 2160
82.7 0.310 7840 I 2430 2400
86.8 0.380 7240 2750 2730
89.2 0.450 7020 3160 3150
91.7 0.520 7040 3660 3670
94.9 0.590 7220 4260 4270
101.6 0.6(0 7560 4990 4980
114,6 0.730 79.50 5780

TABLE 72. Comparison of n values calculated from the Enskog equa-

tio;) and values observed for nitrogen at 50·C
10- 7 -1 -1
!, Y
g cm sec
2 11"
p, atm S ,10- 7 em sec- I
1,g/cm \ (f)exp
expo calc.
I
15.37 0.01623 1179JO 1913 1810
57.50 0.06049 32074 1981 1900
104.5 0.1083 19280 2088 2050
212.4 0.2067 11480 2373 2240
320.4 0.2875 9520 2737 2650
430.2 0.3528 8870 3129 3080
541.7 0.4053 8660 3509 3480
6304 0.4409 8590 3786 3800
742.1 0.4786 8700 4163 4180
8.')4.1 0.5117 8890 4550 4550
965.8 0.5404 9090 4913 4920

Let us transform (137) into

"iT b
--= 1-- (138 )
"i p , T v

If we plot the values 'ITirJp,T on one coordinate, and the values of 11v on the
other, we shall obtain a straight line if (138) holds. The magnitude b is
taken as constant. An example of such a graph plotted for nitrogen (SOne)
aLd CO 2 (40°C) from our data is shown in Figure 74. The volume changes
shown in this graph correspond to the pressure changes from 1 to 800 atm
fo r' both gases. It can be seen that there are marked deviations from a
st r'aight line. Hence (138) does not accurately represent the pressure
dependence of ga~;es at b = const. We could not find an equation which
adequately expressed b = f (p, T),

153
 0.1 f""'"
lJ,/lJ~
a2
V
0,3
/
~
air ./
0.5
/ -- o-CO, -
/;x !
x -nitrogen
0.6

0.7
) 'f i

0.8 ~
0.9
/ I
1.0. ..1'''[
0.0 at 0.2 0.3 O,'r 0.5 0.6 0.7 0.8 0.9 to It
VY
FIGURE 74. Viscosity of nit,ogen (50°C) and of CO,(40°C)
in ~-~ coordinates (v in ems),
V 1Jp , T

Shirokov /115/ started from the ideas of the kinetic theory of gases, and
concluded that the Maxwell equation Tj=(l/3)pcl can be applied to gases under
high pressure and liquids, provided that a correction is introduced similar
to that used in the Boyle -Mariotte equation. Let us quote the author /116/:
"My theory refers to gas viscosity expressed in terms of the Van der
Waals equation. Its Simplified principle is as follows. We shall, as in
the derivation of the Van der Waals equation, introduce the appropriate
corrections into the viscosity equation for gases

(A)

In this equation Is is the mean free path of the molecule; Cs the mean
speed of the random thermal motion; and p the gas density. Because of
the average jump of 2/3d when the molecules collide (d Is the diameter of
the molecules), Is is corrected to

V
1=18 v-b' (B)

Instead of es we shall simply use e, i. e" the propagation velocity of the

momentum which is (v~o) times es • Hence,

e=e sv _V o ' (c)

The cohesion is not considered since the cohesive action of the individual
molecules is balanced. Its indirect effect, however, is included in the
'effective' diameter (Sutherland's correction).
In analogy to the Van der Waals equation, in accordance with (B) and (C),
in place of (A) we obtain

'IJ = CI. (p +~)

v'2 (-~-)
v-b'
(D)

154
where C1. is a temperature -dependent function. At high rarefaction

"1/= C1.pv=C1.RT= AT, (E)

in other words, we have Maxwell's equation for the viscosity of gases as

a function of temperature only. From equation (E) it is possible to calcu-
late C1. from the viscosity of the substance in the gaseous state. At sm all
values of p, from (D) 71= V~b ' which corresponds to the Bachinskii
C '1 f C =C1.a. "
f ormu l a "I/=v:::(;;
Thus, in aceordance with the considerations of Shirokov, from (E) we
obtain C1.= Ri, where 71 is the viscosity of the rarefied gas (viscosity at
760mm Hg). In the above discussion this was designated 'IT. From
Van der Waals equation, (P+ ~)= 'U~lJ ' so that (D) can be written as

(139 )

It can also be represented as

V "lJT
'1p, T
=1-!....
v
(140)

If we plot on one coordinate V "IT

"Ip,T
and on the other l/v (or the density),

we should obtain a straight line if Shirokov's equation holds.

a81---+----f,......,~r--_I

O'7~---+----~--~~--~

O'6~--~----+----+--~

O'5~-~---±~~~~~
o 50. 100 150. 20.0
'p,ID -. mole/ em"

FIGURE 76. Viscosity of me-

thane in .. ~_ p co-
V%:T
ordinates from 10 to 800 atm.

155
 Figures 75 and 76 show the variation in the viscosity of nitrogen and
methane as a function of the pressure (the pressure is varied from 1 to
800 atm) and at various temperatures plotted in these coordinates. It can
be seen that the experimental points for nitrogen lie on a curve rather than
on a straight line at the different temperatures. Let us note that the
constant b of the Shirokov equation was derived in the same way as the
constant b of the Van der Waals equation, but its absolute value is far from
identical with that of the b of Van der Waals.
The same applies to other gases also.
A more detailed analysis of the derivation of the Shirokov equation shows
that the author introduced the same correction (v ~ b) twice.
If the molecules have geometric dimensions (e.g., they are spheres)
then if other conditions are equal each molecule travels a smaller path
between two successive collisions than imaginary molecules with no
geometric dimensions (points).
Let us imagine a molecule inside the gas moving towards the wall of
the vessel. If this molecule collides with a molecule moving in the
opposite direction, there will be a change in velocity. Instead of the first
molecule, the second will move towards the wall. However, at the instant
of collision, the momentum to be transferred jumps, as it were, over
some part of the path, since the impact is aimed at the surface of spheres
(molecules), while the mass is considered to be concentrated at the center.
Hence, the momentum transfer over the distance between the point of
impact on the surface of the spheres and its center occurred not because
of shifting of the spheres but because of an elastic wave (if the spheres
are elastic). Therefore, the momentum is transferred almost instantane-
ously. This means that molecules which move at thermal velocities C. and
are three-dimensional transfer the given momentum over the distance
between the centers of the molecules in a shorter time interval than point
molecules.
Therefore, the correction introduced by Shirokov to the mean free path
actually refers to the "propagation" velocity of the momentum. The
distance over which the momentum is transferred, L e., the distance
between the centers of the molecules, will be same whether the molecules
are points or spheres. For this reason it is more correct to introduce
the correction V~b only once, that is, to the rate of momentum transfer,
which in this case is not equal to the velocity of thermal motion of the molecules

butislargerbY(V~b)' Le., c=c8v~b'

For point molecules let us note the number of collisions by Zm = ~s, and
for spherical molecules by zp=+. Then, according to Shirokov we ~btain

or zp = Zm' and according to the Van der Waals theory from which the

156
Shirokov t~leory was derived we have

U
Zp=Zrn u-b'

This contradietion shows that in the equation of Shirokov the correction

u~b·was introduced twice, which is incorrect. Dubief /117/ even earlier
introduced a correction into Maxwell's equation, which takes into consider-
ation the change in the propagation velocity of the momentum. He derived
the following viscof'ity equation:

u
7Jp , T = TJ T U - b ' (141 )

Using this equation the author calculated the viscosity of CO 2 , and

compared the results with those of Phillips. The equation satisfactorily
represents experimental data over narrow ranges of temperature and
pressure. Dubief used the following empirical equation for b:

(142 )

where bo is a constant. The equation holds for v<; ~bo.

'It

Of the other equations for the pressure dependence of the viscosity,

we shall mention the empirical equation proposed by Borovik / lIS/:

wM
7J = k ---,:2 , (143)

where k =4.62'10- 26 ; M is the molecular weight; r is the radius of the

molecule; .and w is velocity of sound given by

,_ , CP(iJP )
w - - v C; iiV T'

In this equation v is the specific volume, and Cv and Cp are the specific
heats at constant volume and constant pressure, respectively. Borovik
proved that this equation gives a qualitative relationship for the
pressure dependence of the viscosity of, for example, nitrogen.

TABLE 73. Constants Band m in Vargaftik's equation

Substance B m p expressed in TJ expressed in

3
~itrogen • , •• , • , , 0,0068 1.63 glcrn gcm- 1 sec-1
0,00525 1.70 . .
'I.
<=arbon dioxide .••• I

0,0477 1.50 I . .
\1ethane , , , , • , , I
:;team . . . . . . . . . 4,31'10. 10 1.48 kg/rn 3 kgsec m-'
I I

Gonikberg /119/ proposed the following empirical equation for the

157
viscosity of carbon dioxide at different temperatures and pressures:

(144)

where k is a constant; and p is the density. This equation holds for our
experimental results on the viscosity of e0 2 at 20,40,60, and lOOoe, and
at pressures between 100 and 800 atm.
Vargaftik /120/ proposed an empirical equation expressing the variation
in the viscosity as a function of the density, temperature, and pressure:

(145 )

where Band m are constants, experimentally determined for every gas.

The agreement between the theoretical and experimental viscosity coeffici-
ents was satisfactory at p ~ 1. 5 per. The constants Band m for some of the
gases studied by Vargaftik are shown in Table 73.

J I'

(I
3. 51---
J
~f 0
4~ /#
/ V' 2..5 //
V' II
I l 1~
l/:,V 2.
/;f1'
2.0
'''~
h.,. V o -O'C
~ -50' t---- 5
l'lf'"
(,/I'? .
o-O'C
S
/-' x - I{l}'

~I
~-25'
1.5 f -200' f---- 0-50'
~~
0-100'
0 ! =§i5b: f----
l' )
1.0 a5
05 1.0 1.5 2.0 25 WI 1.0
t;(,P'fO,)

FIGURE 77. Viscosity of nitrogen in 19 iI.'l - 19 P FIGURE 78, Viscosity of methane in 19 A'1-1g P
coordinates at pressures from 20 to 800 atm (p in coord ina tes at pressures from 20 to 800 atm ( p in
mole/em 3), mole/em3).

Vargaftik's equation is interesting because it is based on a rule obtained

by experiments, namely, the "fjp, T-"fjT values as a function of density for a
given substance fit a general curve at all temperatures.
Vargaftik's equation was checked for nitrogen (Figure 77) and methane
(Figure 78). If we plot on one coordinate axis Ig ('1l1" T-'1lT) = 19 <l'1l and on the
other 19 p, the experimental values should lie on a straight line if the
Vargaftik equation holds. But this is not the case, as can be seen in the

158
figures. The same applies to other gases. On the other hand, the
Vargaftik equation can be used over some narrow density ranges, provided
that constants Band m are correctly selected, since the experimental
points at different temperatures fit a curve which deviates only slightly
from a straight line.
Stolyarov /121/ derived equations for the viscosity of various gases
as a function of the reduced temperature and pressure. The general form
is

"'Ip, T = "'IT (I + ,~ ), (146 )

where 't= TT , a=/(rc), b=/(rc) , and 'Ir=L. The values a and bare
cr Pcr
determined separately for each gas. Stolyarov processed our experimental
data and obtained the following equations for H2 and CO:
for hydrogen: "'Ip, T= "'IT + "'IT'lr°.92't- S,-0.os,
for carbon monoxide: "'Ip, T= "'IT+O.61jT'lr0.93't-s.4 -0.11.

These equations are fairly satisfactory for the temperature and pressure
ranges for which the constants were selected, and they can therefore be
used for interpolation.

2. Derivation of equation for the viscosity of compressed

gases /122/

In this case we shall use only the concepts of the kinetic theory of
gases, since a more rigid treatment of the viscosity of real gases and
liquids with a simultaneous dependence of the pressure and temperature
would lead to very complex equations. Moreover, these would not
r'epresent the experimental data with greater accuracy.
In general, the equation for the viscosity of compressed gases should
be identical with the Maxwell equation if the gas density is low. On the
other hand, if this equation holds for gases at high densities (comparable
to the density of liquids), the equation should also hold for liquids (at
1east near the critical temperature) in accordance with the theorem of the
continuity of the gaseous and liquid states. As early as 1913, Bachinskii
assumed that such a general equation must exist /123/.
When deriving an equation expressing the pressure dependence of the
viscosity of real gases, we must consider the two factors responsible
f::>r the deviation in the behavior of real gases from the perfect gas laws.
These are: forces of molecular interaction, and the geometric dimensions
of the molecule. Van der Waals considered these factors when deriving
his equation of state for real gases. We shall assume that molecular
interaction has no direct effect on the internal friction of gases, since when
these forces are applied to certain molecules they are balanced, and their
resultant is zero. However, an indirect effect of these forces must be
c)nsidered because:
a) they influence the parameters of state of a real gas, i. e., the number
0:' molecules (number of transferrers of momentum) per unit volume;

159
 b) as shown in Chapter IV, the molecular forces of attraction increase,
as it were, the effective diameter of the molecules. For this reason, the
diameter of the molecule must be considered as a function of temperature.
Thus, we shall consider in our discussion the diameter of the molecules
as temperature-dependent, and use experimental p. v. T data in our
calculations. Thus, we consider the effect of molecular interaction on the
viscosity of a real gas.
To calculate the effect of the molecule dimension, let us imagine the
same gas of volume v, in one case consisting of points of matter (perfect
gas), and in the other as spherical molecules with a certain volume (real
gas).
Let us suppose that an infinitely expanded gas layer moving along the
axis will have a velocity gradient on the Z axis (i. e., in a direction per-
pendicular to the motion). Moreover, let us assume that all the gas
molecules have the same (average) absolute speed C, and that the
distribution of directions of their thermal velocity is random (in other
words, the average distribution of the velocities over the X, Y, and Z axes
is uniform). Consequently, from the overall number of molecules present
per unit volume, one third will move in a direction normal to some
imaginary interface inside the gas. Half of these (Nj 6) pass through this
plane in one direction (from right to left or f]~om top to bottom), and the
other half in the opposite direction. Thus, through unit surface (1 cm 2 )
(1 j 6) Nc molecules pass from left to right per second, and the same number
from right to left. This consideration holds, however, for a perfect gas
only, in which the dimensions of the molecule can be neglected.
Let us now determine how the number of molecules passing per 1 cm 2
through the interface changes for a real gas consisting of spherical mole-
cules.
Let one of the moiecules of the gas layer A (Figure 79) moving along
the X axis, after its last collision in layer A fly to a molecule of the
adjacent layer B (parallel to A) along a line drawn between the centers of
the molecules. Let us suppose that on its way the molecule will not collide
with other molecules, but travel a distance equal to its mean free path.
It is easy to see that the distance over which the momentum is transferred
during one collision (distance between the centers of the molecules) is given
by l~ +d. The momentum is transferred over distance l;, with the mean
velocity of the thermal motion of the molecules c, and "instantaneously"
over distance d. It should be noted that the ratio of the time necessary for
transferring momentum over distance d by impact to the time which would
be needed for the transfer at the velocity of molecular motion is equal to
the ratio of the velocity of the molecules to the velocity of the elastic wave
(since we consider the molecules to be solid, absolutely elastic bodies).
This ratio is vanishingly small because of the much higher velocity of the
elastic wave, so that we can consider that the transfer of momentum from
the "periphery" to the center occurs simultaneously on impact. The
instantaneous transfer of momentum over di.stance d can take place only if
we consider central collision of the molecules (Figure 79).

160
 r
FIGURE 79. Central collision of FIGURE 80. Collision of
molecules. molecules at an angle.

In reality, however, there is no equiprobability of the events that the

molecules of layer A moving along z will collide with the molecules of
layer B at arbitrary points of the semisurface of the sphere aa (central
collision). Let us, for example, assume that the molecules collide so
that the straight line connecting their centers forms a certain angle {i with
the direction of the relative velocity c
(Figure 80). In this case, the
distance by which the mean free path is reduced is given by d cos {i (d = AB).
The number of collisions possible at angle if (when 0 <: if <: -'f) is pro-
portional to sin if dif. For this reason, the average distance by which the
mean free path is reduced is given by

~/2

f
o
dcos&sin&d&={=r.

Thus, the influence of the geometric dimensions of the molecules on the

layer-to -layer momentum transfer inside the gas can be characterized as
follows: the momentum is transferred over distance il with a velocity of c
and over distance r instantaneously. Hence, a molecule of the real gas
travels the distance between two collisions in a shorter time interval than
a molecule of a perfect gas.
The time decrement is given by

where I is the mean free path of the molecules of the perfect gas; II is the
mean free path of the molecules of the real gas; and r is the radius of the
molecules of the real gas.
However, in reality the situation is rather different. In a real gas very
different types of collision may occur, and not only the case when molecule
pairs collide. It can be assumed with a fair degree of certainty that not
only pairs of molecules, but three, four, five or more molecules may
simultaneously collide /124/.
The momentary state of a gas can be imagined as a set of molecule
groups of different complexity, namely a certain number of molecules move
singly, some molecules collide in pairs, or three, four, or more molecules
::ollide simultaneously.

161
 ~ '2 To each equilibrium state of a gas, there
corresponds a fully determinate distribution.
In the above explanation of molecule

r groups, we used only the simplified scheme

of collision in pairs. For this reason the
equations obtained are only approximate.
Let us now place a division plane S be-
tween layers A and B (Figure 81), and we
shall assume that an arbitrary molecule
as a result of collision in one second
completes its movements between the two
FIGURE 81. Scheme for determining
the length of th-o free path taking
layers only. Because of the time decrement
into account pair collisions .. between two successive collisions, a
molecule of real gas will pass through the
plane (1 + ~ ) times more than the perfect
gas molecule. If we apply this consideration to the number of molecules
per unit volum e, Nt spherical molecules will pass through a square
centimeter of the division plane S. This number is equal to

Let us now determine the momentum transferred through unit area of the
division plane inside the gas.
First, we shall note that the molecules arrive at the division plane from
both sides, from a distance equal to their mean free path. The distance
over which the momentum is transferred will be the same for perfect
and real gases. Of course, we consider that molecule a is in the plane S
(Figure 81) only when its center is in this plane, i. e., when it is in
position a'.
Let us denote the velocity gradient in the gas by ~~. This means that
in direction z the velocity decreases by ~~ on each cm along axis Z. If
at the division plane the velocity of the gas is w, at a distance I on the left-
hand side it will be larger by ~~ I and at the same distance on the right-
hand side it will be smaller by ~~ I. A molecule passing from the left-hand
side transfers momentum m (w + ~~ I) to plane S, and a molecule passing
from the right-hand side the momentum m (w - ~~ I). Here m is the
mass of the molecule. Thus,

of such spherical molecules pass from the left to the right per cm 2 per
sec, and the same number from the right to the left. Thus, the momentum

162
transferred from the left-hand side to the right-hand side is

and from the :right-hand side to the left-hand side

6
r) (
1 - ( 1+- m w -dW)
-Nc
£1
-l.
dz

The excess momentum per second is

Since this excess momentum determines the force acting upon unit area of
the division plane, we can write

1 - (
F=3 Nmcl 1 +~ liZ'
r) dw

dw
On the other hand, F = "l dZ' By comparing the expressions for F, we
obtain

"l=}NmCl(1 + ~), (147)

1 - 1 - r
"l =3 Nmcl+ 3 Nmclt; . (148)

The first term on the right-hand side of (148) indicates the viscosity of
the perfect (ideal) gas "l id, and the second term the correction to the
viscosity of the perfect gas due to the geometric dimensions of the real gas
Nm - Nm - )
molecules. Sinee -3-cl=7/id and P=-3-C2, (148 can be written as

(149)

Let us determine the ratio llll' We start from identical densities (equal
number of molecules per unit volume) and temperatures for perfect and real
gases:

By putting

where v is the molar volume of the gas and b the quadruple volume of all
the molecules per mole of the gas, we obtain

163
By substituting 1/11 into (149), we obtain

'Y/='Y/id +:...c P _v_.

v-b

If the above considerations refer to perfect and real gases at the same
temperature and pressure, we can write the last equation as

since an equal number of molecules of a perfect gas and real gas occupy
different volumes under identical temperature and pressure conditions.
From the equation of state for a perfect gas, PVicl=RT, and from
Van der Waals equation for a real gas we obtain

(v-b)=-.
RT
Pt

and therefore the viscosity equation is

(150 )

Pt is the so-called thermal pressure, which is determined by the thermal

motion of the molecules. Quantitatively, it is the pulse obtained by unit
interface per second from the molecules impacting the interface.
If the pressure is low (atmospheric pressure), Pt is small, and the
magnitude :... Pt is small in comparison with 'f)id so that it can be neglected.
c
Then, (150) becomes simplified to the Maxwell equation. At very high
pressures, when the gas density becomes commensurable with the density
of a liquid, and Pt may be as high as several thousand atmospheres,
~ Pt becomes so large that 'f)id can be neglected, and we get the equation for
c
the viscosity of liquids from (150).
Furthermore, for a real gas at constant temperature and two differing
pressures, we can write

'f)p, T = '1Jid + :c. . P~'

'Y/T = 'Y/id +:...c p;.
By subtracting the second equation from the first equation and putting P~ = Pt ,
P;= 1, we obtain

At high gas pressures, when Pt is much larger than unity, we obtain

(151 )

164
 Equation (l!j 1) does not contain any magnitude s characterizing a perfect
gas. Thus, the concept of a hypothetical perfect gas was only used to
derive this equation.
If we assume that the gas molecules
£ + are absolutely incompressible spheres, r will
be constant at all temperatures and
pressures. However, such assumptions
are only approximate, and do not conform
to modern theory and facts. It is more
correct to consider the molecules as a
system over which electric charges are
distributed. At considerable distances
between the molecules attractive forces
act, and it is only when the distances are
FIGURE 82. Potential energy of inter-
very small that the rapidly increasing
action of two molecules as a function repulsive forces predominate, and prevent
of the distance between them. complete approach of the molecules.
The potential energy E of interaction
of two molecules as a function of the
distance between these two molecules is shown in Figure 82. This function
has a minimum (potential well).
By means of this potential energy curve, we can better explain the
11 diameter of the molecule." It can be seen in Figure 82 that if the distance
1 between the molecules is very large, the energy of interaction is close
to zero.
With decrease in I, E decreases, and reaches a minimum at point
1 = druiD' Further decrease in 1 leads to' an increase in E. At do, E = 0, or
in other words in this case the energy of interaction is numerically equal
to the energy of molecules infinitely distant from one another. With
further decrease in I, E rapidly increases.
The value dmin corresponds approximately to the distance between the
:molecules in the crystal lattice. It is also equal to the distance between
gas molecules when the temperature is OOK under vacuum conditions. The
'value of druin can be determined from the molecular volume of a liquid with
';he densest pac!{ing, i. e., from Vo at OOK. The following equation is used:

~1t (d'2111) = 0,74 ~ .

The constant O. '74 is the coefficient for the densest spherical packing. For
1he majority of gases; the value of dmin is such that the corresponding
volume of all the molecules is about half of the critical volume Vcr. The
values of Vo and Vcr can be determined from the Cailletet-Mathias relation
(the law of rectilinear diameters). According to this rule, the arithmetic
mean of the densities of a liquid and vapor in equilibrium, PL tP v
, plotted
c,gainst temperature gives a straight line. Thus, the density at the critical
temperature and absolute zero can be determined by extrapolation, i. e.,
Vcr and Vo can be determined.

165
 do is the smallest distance to which two infinitely remote molecules with
no kinetic energy can approach one another. The following approximate
relationship holds between do and d min :

Since the molecules always have a certain kinetic energy, they approach
one another on collision regardless of the repulsive forces. The smallest
distance between colliding molecules when they approach one another is
termed the gas -kinetic diameter of impact, and termed dT.
Thus, three differing diameters may correspond to each molecule,
do, d ulln , and d T , and dT < do < dmin. These three diameters correspond to
different initial conditions of interaction.
In our viscosity equation, the molecular diameter dT = 2 r. Hence, in
our case 'T should be considered as a fUl;lction of temperature (over the
pressure range up to 1000 atm, , can be considered as pressure
independent). It is easy to see that with increase in the kinetic energy of
the molecules, i. e., with increase in gas temperature and consequently
increase in the mean velocity of the molecules, the molecu12s penetrate
deeper into the sphere of action of repulsive forces on collision, and hence
rT will decrease. The type of the functional dependence of the diameter
of the molecule on temperature is not yet clear. We have only approximate
equations, giving more or less correct values for some limiting conditions,
namely, for the molecule diameter dco at infinitely high temperatures.
This can be calculated from the viscosity coefficients and their temperature
dependence according to /3/

(152 )

Since for the majority of molecules the potential of repulsion abruptly

changes at high temperatures (of the order of several hundreds of degrees),
dco will change slightly, so that the values obtained from (152) may be
considered as the gas -kinetic diameter of impact at high temperatures.
It is worth noting that the theory of Sutherland /101/, leading to the
equation

(153 )

cannot correctly represent the temperature dependence of the molecule

diameter, since in Sutherland's model of the molecule the repulsive forces
are not taken into account. If these forces are considered, the dT - T curve
will be rather steeper than the curve obtained from Sutherland's equation.
Thus, for the temperature dependence of dT, we derived the following
empirical equation:

A
dT=n' (154 )

where A is a constant, characteristic of every gas.

166
 This equation holds over a wide temperature range, and gives a greater
i emperature dependence of the viscosity than Sutherland's equation.

By substituting r = rT from (154) in (151), and using c= y 3~T, we

obtain

(155 )

where a is a constant, characteristic of every gas.

~. Application of the equation for the viscosity of

compressed gas to experimental results

It follows from "Ip,T-"fJT=a jt that if "Ip,T-"IT is plotted on one coordinate,

and P.j on the other, experimental viscosities of a gas at all temperatures
and pressures must fit a straight line, provided that the equation is correct.
Figures 83 and 84 show the results of such processing of data for nitrogen
and carbon dioxide, At the temperatures indicated in the figures, the
experimental data were taken at pressures between 1 and 800 atm, It can
be seen that the experimental points do not lie on a straight line but on a
c I1rve, which has the same shape at all temperatures, The straight line
drawn between the origin of the coordinates and the point on the curve at
4l = 1 lies slight1y above the curve for values of P.j from a to unity, and is
bI~low the curve when P.j is greater than unity. The same applies to the
other gases that we studied.
Golubev and Petrov analyzed the experimental data and found that the
deviations of the viscosity coefficients calculated from (155) from the
e][perimental values have the same character. Thus, an equation of
th e following type:

(156)

corresponds bettcr to the experimental data, In (156), a and n are

constants characteristic of every gas.
To apply (156) to experimental data, it is better to transform the equation
to the logarithmk form

If on one axis we plot the 19 h p, T-"IT) values, and on the other the 19 jt values,
WE'should obtain a straight line for all the experimental points if the
equation is correct. This line intercepts a section equal to 19 a on the
axis 19 ("fJ p, T - "IT) for the value 19 jt = 0( jt = 0). The slope of this line with
the axis 19 jt is eq ual to n.

167
 5000
,, V
,
250,
, V
I~ /
// !

,/ '/
1500 '/
,l/ -
! I

1000
,tj
~v
- , -O'C I-- -
0-50
/ "-100
500 Il-lfj[] I-- t"-
V
0,0 V 2 3
-L
" pIT 5

FI GURE 83. ..
ViSCOSIty 0
fmtrogen
' . (
III '11>, T -
.,,)
"T - TPt
coordinates at pressures from 20 to 800 atm •

1400,'0 .--
f'
1301J
12000

11000
/
1000,'{j
V
'j:90oo
" l'
lo'{

/
V I
I
I
I
8000
'jO!/,i /
7000
§-:
5000 , /
5000
0/ /
4000 '!>'''/
3000
0'
Y
2000
.J/ • -20'C
o-itO - r--
1000 / ·-50
' -100 '---- -
0,0 / 2 J 5 6 8
, -ISO
9 10 II 12
P,/T

FIGURE 84. Viscosity of carbon dioxide in ('1Jp , T-'1JT)-

- Ff coordinates at pressures from 20 to 800 atm.

168
 5
V

2..4
.I
V
..0 f--- 0-40
CO, 0 -6{)
I-- '-100
rm
• -ISO
~ lA,
lI:
/cO,

2..2 3.5
.r
V V [7
2.
~ / V
.8

V / r/ i
6

V 2. 5 / 7 I! 1

!
/ ~I
i I
.4 r I i !

1/
r"1
I

.2
2.0 // I i
.0
/ o -25·C
V/, ! I N ~ =fSo
I 0 -200
l>"\
8/ • -100
x -200 i.5
-10
i
-0.5 0.0
I
1.0 1.5
19 If-
!'IGURE85. Viscosity of hydrogen inlg ('Jp , T-'JT)' FIGURE 86. Viscosity of carbon dioxide and nitrogen in
- Ig P; coord inates at pressures from 50 to Ig ('J p , T - "IT) - 19 If coordinates at pressures from
800 atm. 50 to 800 atm.

0
5 /
V ~t'
3.
/ 5
V
/
2. 5 / 0
/
~
V 1//
0
Lf& 0- -15'C 5
/
0- O'C
./
5
VS~ .-25·C -
~ -50'C
i' o-150·C
; -175(interpolated)

:V·
A-IOO'C __I 2.0
x-200 I
&-1507: /1)' O-"25.(interpollt ed)
-¢--200·C \
1>-250
i
:
~ I
0. I 1.5
o 0.5 1.0 1.5 2.0 2.5 3.0 -/'O -0.5 0.0 0.5 10
If! ;rIO' lflt;.

FIGURE 87. Viscosity of methane in Ig('1p , T- '1 T)- FIGURE 88. Viscosity of ammonia in 19 ("I p, T - "IT)
- 19 P,j coordinates 2.t pressures from 10 to 800 atm. -lg -f! coordinates at pressures from 40 to 800 aun.

169
 Figures 85 - 88 show that the equation is applicable to hydrogen,
nitrogen, carbon dioxide, ammonia, and methane over a wide range of
temperature and at pressures up to 800 atm. It can be seen that all the
experimental points for all these gases fit a straight line. The differences
between the theoretical and experimental values are not more than the
errors permitted in the determination of the viscosity coefficients, or
involved in the calculation of Pt. Table 74 gives constants a and n for
some gases. It can be seen that n is between 1.10 and 1.12. Thus the
mean value of n for all gases is n = 1.115. We have not yet found any
relationship between a and the gas type. This constant must therefore be
determined from the viscosity coefficients at high pressures.

TABLE 74. Constants a and n for some gases

Gas a

Hydrogen .................... 73 1.12

Nitrogen .................... 567 1.12
Carbon dioxide ••••••• 930 1.117
Ammonia .................... 550 1.115
Methane .................... 540 1.102
Ethane ...................... 880 1.118
Propane ...................... 1475 1.12
Ethylene ••••••••••• 1000 1.11

It can be seen from the example of carbon dioxide that (156) can be
applied to the liquid state also. It should be noted that the theory of liquids
is still not sufficiently advanced, and the interpretation of data in this
range is rather uncertain. This is because the liquid state is intermediate
between the gaseous and solid states. This intermediate position of
liquids makes it possible to explain its phenomena by approaching this
state from the solid or gas phase. These two opposite approaches to an
understanding of the liquid state can also be seen in attempts to solve the
problem of the viscosity of liquids /125/. Many arbitrary constants in the
equations derived from the different theories define certain temperature
and pressure ranges over which the experimental data of the viscosities
can be represented by these equations. It is, however, impossible to give
preference to any of the equations.
Equation (156) very probably describes the pressure and temperature
dependence of the viscosity for the gaseous state over wide temperature
and pressure ranges, including the critical point. For the liquid state it
may hold as long as the liquid is considered to be a strongly compressed
gas.

4. The determination of the thermal pressure

The thermal pressure is defined by

(157)

170
 But since

we have

(158)

The thermal pressure can be calculated from the equation of state of real
gases, or directl.y from p, v, T data graphically representing the variables
)f (157) and (158).
To calculate Pt we used the most popular equations of state, namely the
Van der Waals and the Beattie -Bridgman equations /126/.
Let uS write the Van der Waals equation as follows:

RT
P=v-b-v2 ·
a
:3y differentiating it with respect to T for v = const, we obtain

Thus

Pt = T(g~)v =v~b=P +:' . (159 )

Let us now write the Beattie - Bridgman equation:

p= RT(l~vYs)[V+Bo(l_f)]_~g(l_%).
By differentiating with respect to T, we obtain

(160 )

Analogously, the thermal pressure can be determined from any other

equation of state. The equation of state must be used over temperature
and pressure ranges for which the equation adequately represents the
}, v, T data.
The constants of the Van der Waals and Beattie - Bridgman equations
for some gases are given in Tables 75 and 76.
Graphical differentiation of experimental data will give more accurate
thermal pressures. The p vs. T curve is plotted while the volum e is
kept constant. For the majority of gases the isochores obtained are

171
linear over wide temperature and pressure ranges. Therefore, for a given
isochore, ~~ = (~~)v = k, which considerably simplifies the calculation.

The differential is the slope of the tangent at any point of the curve.

TABLE 75. Constants of the Van der Waals equation calculated for Pcr> vcr> and T cr> when '/I is expressed
in ens and P in atmospheres (if v is expressed in molar volumes, a and b must be multiplied by 5.03.108
and by 2.24.104 .respectively)

Gas a b Gas a b
I
Helium •••••••••••• 0.000068 0.001058 Cyclohexane •••••••• 0.04347 0.006359
Neon ................. "" .... 0.000422 0.000763 Benzene ...................... I 0.03588 0.005150
Argon ••••••••••••• 0.00268 0.001437 Toluene ••••••••••• 0,04795 0.006533
Krypton •••••••••••• 0.00462 0.001776 Methanol •••••••••• 0.01898 0.002992
Xenon •••.••••••••• 0.00818 0.002307 Ethane ........................ 0.01071 0.002848
Hydrogen ................ . 0.00386 0.000977 Ethylene .................... 0.00891 0.002551
Nitrogen ••••••••••• 0.00268 0.001719 Chlorine .................... i 0.01294 , 0.002510

Oxygen .•.••••••••• 0.00271 0.01421 Hydrogen chloride •••• 0.00731 0.001822

Carbon monoxide ••••• 0.00296 0.001779 Hydrogen bromide ••••
i 0.00887 0.001978
!
Nitric oxide .................. 0.00267 0.001245 Phosphine ••••••.••. 0.00923 0.002302
Carbon dioxide ••••••• 0.00716 0.001909 Chloroform ••••••••• 0.03023 0.004562
Water vapor ••••••••• 0.01089 0.001362 Methyl chloride •••••• 0.01489 0.002894
i 0.003862
Sulfur dioxide .............. 0.1338 0.002510 Ethyl chloride ••••••• 0.02174
Nitrous oxide •••••••• 0.00754 0.001971 Ethanol ••••••••••• 0.02395 i 0.003753

Ammonia ................... 0.00831 0.00166 n -Propanol ••••••••• 0.02974 0.004548

Methane ••••••••••• 0.00449 0.001910 Isopropanol ••••••••• 0.02747 ! 0.004377
Hydrogen sulfide •••••• 0.00883 0.001914 Acetone ••••••••••• 0.02774 iI 0.004437
,
Carbon disulfide •••••• 0.02316 0.003431 Carbon tetrachloride ••• 0.03892 0.005661
!
Thiophen •••••••••• 0.04130 0.005670 Ethyl ether •.••••••• 0.03464 0.006002
Propane ..................... 0.001727 0.003770 Aniline ••••••••••• 0.05282 0.006113
n-Butane .................... 0.02884 0.005472 Acetic acid ••••••••• I 0.03505 0.004767
Isobutane .......... 0.02564 0.005098 Mercury ••••••••••• 0.01613 I 0.000757
I
n-Pentane •••••.•••• 0.03788 I 0.006516 i
n-Heptane •••••••••• 0.06280 0.011850 I
I I
n-Hexane ••••.•••.• 0.04928 0.007850
II
n-Octane .......... 0.07440 0.010570
1
i
Propylene •••••••••• 0.01670 0.003693
Acetylene •••••••••• 0.00875 0.002293
I I

Deming and Shupe /127/ used magnitudes characterizing the deviation

of real gases from the Clapeyron equation

Il=VU~ -1). (A)

RT
CI.=--V. (B)
P

The values of Il and CI. calculated from experimental p, v, T data were

plotted against T or p, and the derivatives (gnp' (~!)T or (;;)p' (g;)T found.
This method is convenient, because the curves representing the dependence

172
of !1 or a on T or p have an appreciable curvature. Thus, it is easier to
construct the tangents, and hencc the derivative can be more accurately
determined.

TABLE 76. Constants of the Beattie-Bridgman equation (units: atm, liter, ·K; R = 0.08206)

Temperature
I I
Maximum Minimum
range,OC
Gas A, a B b C.1O- 4 pressure, volume,
from I to atm ml/mole

0.02)[i
0.212."
I 0.05984
0.02196
0.01400
0.02060
I 0.0
0.0
0.0040
0.101
400
400
-252
-217
102
106
100
118
1.290"( 0.02328 0.03931 i 0.0 5.99 400 -150 114 167
O.1(I7c -0.00506 0.02096 -0.04356 0.0504 200 -244 103 100
I.2b1"7 0.01866 0.04603 -0.02587 6.16 400 -70 1000 84
1.491 1 0.02562 0.04624 0.00420S i 4.S0 100 -117 103 III
1.3012 0.01931 0.04611 -0.01101 4.34 200 -145 177 125
5.0065 0.07132 0.10476 0.07235 I 66.0 100 0 III IS2
2.2769 0.01855 0.05587 I -0.01587 12.83 200 0 243 167
31.278 0.12426 0.45446/ 0.11954 33.33 325 150 90 370
6.1~20 0.04964 0.12156 0.03597 22.68 200 0 286 125
2.3£30 0.17031 0.03415 , 0.19112 476.8'1 325 -35 130 340
23.7 0.305 0.59 0.622 0 126 30 18.5 1430
5.8800 0.05861 0.09400 0.01915 90.00 25 250 193 200
11.9200 0.07321 0.18100 0.4293 120.00 97 275 305 100
17.7940 0.12161 0.24620 0.09423 350.00 150 300 118 280
54.520 0.20066 0.70816 0.19179 400.00 275 350 315 200
16.960 010860 0.24200 0.08750 250.00 150 250 250 111
33.309 0.09246 , 0.60362 0.09929 52.031 - - - -
i

TABLE 77. Comparison of the P t values (in atm) at 0 and 100·C determined from the equations of state and
by graphical differentiation

t ~O·C I ~ 10O·C
f---
Ptfrom the Pt from the
P, ann Pt from the P t according Pt from the PI according
Beattie- Beattie-
Van der Waals to Deming Van der Waals to Deming
Bridgman Bridgman
equation and Shupe equation and Shupe
equation equation

1 1 1 1 1 1 1
60 I 70 - 70.8 65 - 64.8
100 I 127.6 129 129.2 113.4 113.2 113.0
200 300.5 304.5 301.5 248.4 246.0 I 244.4
300 487.5 486 489 395.2 - I 393
400 I 672 659 665 547 528 543
500
r
848 811 845 700 - 687
600 1015 950 1010 852 804 835
800 J.330 1195 1312 1147 1049 lll8
I I I

173
 The partial differentials (gnp and (~;)T were calculated from the follow-
ing equations: .

1)

2)

3)

4)

These equations were obtained by differentiating (A) and (B), with

successive transformations. It can be seen that (g~)p and (~~)T can be
determined from two equations, and hence from two curves corresponding
to b. = f (T) and IX = f (T), It is thus possible to correct errors in the deter-
mination of one partial differential from one curve by determining the same
differential from the other curve. This method is at present the most
accurate, and the (~~)p and (~;)T values calculated for several gases by
this method are completely reliable.

TABLE 78. Thermal pressure of hydrogen (in atm)

~\ -75 \ -50 I -25 I I ~ I I ~ I BIB \

w a I a

1 1 1 1 1 1 1 1 1 1 1 1

50 51.8 51.2 50.9 50.7 50.6 50.5 50.4 50.3 50.2 50.1 50.0
100 106.2 104.5 103.3 102.8 102.3 102.0 101.4 100.9 100.6 100.3 100.2

200 220.0 215.0 212.0 209.5 208.0 206.3 204.7 203.0 202.0 201.5 201.0
300 337.0 328.5 321.0 317.0 314.4 312.0 308.0 304.6 303.4 302.5 301.5

400 455.0 441.0 431.0 425.0 420.0 416.0 412.0 406.0 404.0 402.0 401.0
500 573.0 554.0 541.0 529.0 526.0 520.5 514.0 507.0 504.0 501.0 500.0

600 681.0 662.0 646.0 637.0 629.0 623.0 615.0 607.0 602.0 599.0 599.0
800 891.0 868.0 851.0 838.0 829.0 820.5 811.0 802.0 796.0 793.0 790.0

1000 1090.0 1068.0 1043.0 10280 1017.0 1008.0 995.0 986.0 983.0 982.0 981.0

I
I i

174
TABLE 79. Thermal pressure of nitrogen (in atm)

~I -70 I -so I I -25 I • I so I ~ I BIB I G I a I B

1 1 1 1 1 1 1 1 1 1 1

20 22.1 21.9 21.5 21.2 21.0 20.8 20.6 20.4 20.2 20.1 20.07 20.05
60 85.5 79.1 74.0 70.8 68.3 67.0 61.8 63 61.6 60.9 60.7 60.5

100 179.5 156 140 129.2 122.6 118 113 107 104 102.3 101.9 101.7
200 436 377 328 301.5 282.5 272 244 224 213 207 204.5 203

300 678 588 526 489 460 435 393 348 326 315 309 308
400 910 794 712 665 624 593 543 476 442 426 420 415

500 1125 006 914 845 790 749 687 606 564 540 528 522
I
600 1310 170 1080 1010 948 I 905 835 736 685 654 641 628
I !
I
800
1000
1630
1870
477
730
1378
1640
! 1312
1 1570
1260
1510
1204
1460
1118
1375
992
1240
I 1160
925 I 1100
876
830
1080
I
I
850
1060
!
II i
1 I I i I

TABLE 80. Thermal pressure of carbon dioxide (in atm)

'"
P'~
t, 'C I 20
150

1 1 1 1 1 1 1 1
20 26.5 - - 23.9 - 22.5 -
50 - 102.6 92.7 79.6 71 67.8 61.9
iiO 1200 162 129.3 106.5 - - -
70 1400 - 187 139.8 120 - -
80 1565 1033 197.5 183.3 - 131.2 -
90 1655 1192 550 240.5 - - -
100 1725 1328 897.5 316.5 210 187.7 149.8
120 1855 1524 1190 539 300 261.6 -
150 2015 1750 1460 888 470 403 272
200 2265 2010 1775 1295 780 684 430
250 2470 2225 2025 1595 1080 955 615
300 2635 2410 2230 1841 1340 1200 810
400 2925 2730 2560 2227 1770 1635 1170
500 3180 3000 2850 2548 2110 1990 i500
600 3410 3240 3100 2825 2410 2280 1800
700 3615 3460 3310 3060 2675 2545 2085
800 3800 3655 3510 3265 2890 2775 2335
1000 4175 4030 3885 3650 3330 3220 2785

Table 77 gives the Pt values for nitrogen at 0 and 100°C calculated by thE
Van der Waals and Beattie -Bridgman equations, and those determined by
the Deming and Shupe method, The table shows the satisfactory agreement
between the Pt values obtained from the equation of state and by graphical
differentiation, Similar results are also found for other gases, and

175
therefore the above equations can be used for an approximate calculation
of Pt, and hence the pressure dependence of "i.
Tables 78-82 give as an example the thermal pressures of some gases
obtained by graphical differentiation over wide ranges of temperature and
pressure. The partial differentials (~7t and (~)T were used. These had
been previously determined by graphical differentiation of the experimental
P,v, T data for nitrogen and hydrogen /127/, methane /128/, and ammonia
/129/. 'I'

TABLE 81. Thermal pressure of methane (in atm)

~,oc
p,atm ~
1 - 25 1
I ~ 1 w 1 ~ 1 ~ 1 ~
i
1 1 1 1 1 I
20 22.8 22.0 2;.51 21.2 20.9 20.8 20.7
50 72.0 65.5 63.0 I 60.5 56.2 54.6 53.6
80 148 127 114 109 98.5 93.0 89.0
100 213 181 156 142 128 120 113
120 299 235 202 180 161 147 138
160 471 360 305 268 223 204 196
200 639 i 500 424 372 306 273 I 254
250 811 650 560 493 406 : 355 I 324
300 942 790 I 690 620 515 i 440 [ 395
400 1140 1020 ! 920 840 709 615 , 539
500 1310 1200 1110 1030, 895 780 . 681
600 1460 I 1360 1280 i 1200 I 1060 930 : 820
800 I 1700 1610 I 1535 1460 '[1320 1190 'I' 1070
1000 i 1910 II 1830 ! 1750 11680 I 1550 1430 1310

TABLE 82. Thermal pressure of ammonia (in atm)

~1150 175 200 225 250 300

1 1 1 1 1 1 1
20 22.9 22.7 22.2 21.8 21.5 21.2
40 54.5 51.8 49.5 47.5 46.7 44.3
60 98.0 88.3 81.8 76.5 74.3 70.0
80 156 135 121 113 106 97.2
100 209 186 163 152 142 128
125 350 263 230 210 191 173
150 775 397 314 276 243 222
175 1400 651 441 353 316 277
200 1740 919 605 447 395 338
250 2220 1490 1040 740 601 475
300 2520 1900 1400 1060 ! 840 638
400 3010 2440 1960 1580 1310 990
500 3380 2880 2430 2050 1740 1350
600 I 3730 3270 2830 2460 2140 I 1680
700 - 3580 3210 2860
i 2480 2020
I
800 3900 3480 3120 2830 2340
1000 4070 3720 3420 2920

* [No data given for CO,.J

176
5. The density dependence of the viscosity of gases

A typical 11 vs p curve for gases is shown in Figure 89. The data for
nitrogen at three temperatures (ooe, 1 to 600 atm; 100 and 200°C,
1 to 300 atm) G.re shown on this figure. It can be seen that a separate
curve exists for each temperature. The curves lie higher at higher
temperatures. However, if we plot 'Ip, T -'IT on the ordinate and densities
on the abscissa, the experimental points for the gaseous state fit a
single straight line. This is shown for some gases in Figures 90-94.
Figures 90-93 show the viscosity data for nitrogen, hydrogen, methane,
carbon dioxide, and ammonia, at pressures up to 800 atm (the viscosity
coefficients were taken from Tables 8,13,20,21, and 22). The densities
were calculatE,d from the most reliable p. v. T data found in the literature.
In Figure 94 the same coordinates are used (but the density is given in
g/cm 3 and not in mole/em 3 ), and the data of Michels, Schipper and
Rintoul /79/ for hydrogen and deuterium are plotted up to 2000 atm
(see Tables 9 and 10).

,{

, II /
350. l/ II
I
V20%
VI
~
"s 300. / roo I
~ Vo-c
';-
()
bI)
'/ / j
""~25o.'O~ / V V
""
/'
/'
I
/ /
2000

,..cr
V
1500
25 50 75 100 125 f50 175 2m
1',10 -, mole! em"

FIGURE 89. Density dependence of the viscosity of

nitrogen.

It follows that for every gas the same values of lip, T-'IT correspond to
the same values of p at all temperatures. Thus, if we determine the
viscosity of a gi.ven gas at one given temperature only over a wide range of
densities, the viscosity at all the other temperatures can be read from the
graph, pro~ided. that we know the gas density at the given temperature.

177
 250,'0 t;c
)

2000
J
,
V
()

~
's f500
o-O"C
x-50 j
~ -100

If
()
co 8-200
':'-
!O;?
~
Nz /
':'"1000

V * -25"C
500
1/ --
0-50
:It-loO
V @-200

V ..-, """~ I~
o -""'
o 50 fOO 150 200
p. 10-4 mole/emS

l
FIGURE 90. Dependence of "Ilp • T - ''IT) on the density of
nitrogen and hydrogen.

woo

3500
/
j
3000
,
()
~7
}~
, 2500
8
()

if'2000
)
7
j
~5Oo
~ ff.
v* x~O'r:

o~5o
~/OOo --
J;£ q,-IOO
1/
,
!;~200
500
I~
o ~

o 50 foO 1.50 200

pJO-< mole/ ems

FIGURE 91. Dependence of ("'IP. T - "'IT) on the density of

methane.

178
 16mIj

14000

1200,'0 J
I'

.. ~ 1000.V -
x -20'C
~-JO
0-40
6 -100
I
E
()

~oo8000
~
o -150
o
@
-200
-Tc .Per f
W
;
woo I
rr

2000 /
V~
o ~
o 50 100 150 200 250 JOO
,P, 10-+ mole/ ems

FIGURE 92, Dependence of ("lp, T -''IT) on the den-

sity of carbon dioxide,

1500.'0

lwa
~
x -30'C
-80
~
J
6 -100
a -120
r'
()

·t'" /000.
o -140
0-150
* -200 1
E
()
co
1, 8000
'"
+ -250
<!J -TepPer

t
I
~
,6000
...0.-
J
~
,tOOO
V
2000
?'
~ V
~
o ~
o 50 tOO 150 200 250 300 350 WO
jJ, 10-· mole/ems

FIGURE 93, Dependence of ("lp, T-"lT) on the density of ammonia,

179
 1100
tOo.'0 /
900
I
I'
I'M'0
u
OJ
f</i 700
0-25T
x-50 V
8
U600
·-75
H88 I /
OJ)

~500
0-125
II Ii
/
·-25"C
~ ! ,/
. 40'0
I/Hz
c_ '-50 -
~ ~." 0-75
3'0'0
V --'~ !-+
=tt-
200
p/
18'0
;iF'
.--
o k¢ !.'-' ",--
I(JO lOu 300 408 500 600 100 800 90'0 1000 1100 1200 f!!{}'O 140'0 15m
.0,10-' g/cm 3

FIGURE 94. Dependence of ("Ip , T - ~T) on the density of deuterium

and hydrogen according to Michels, Schipper, and Rintoul.

We see from the CO 2 and NB3 diagrams that the data for the liquid state
fit into a straight line rather well over the ranges of temperature and
pressure that we studied (for the liquid state this applies at temperatures
close to the critical only).

TABLE 83. Critical viscosities and densities of some gases

p (STP), pcr,10-
4
~(Tler,10-7 ~(p, Tl cp10- 7 ~(p, T)cr
Gas
10-3 g/cm 3 mole/crr;3 gem -1 see -1 gem -1 sec-
' ~(Tkr
-----------------4---------+--------r--------- ---------+--------f-------
Hydrogen . . . . . . . . . . 0.08987 0.0301 149.5 173 294 1. 70
Deuterium ... . 0.1800 0.0670 166.0 262 484 1.845
Nitrogen .. , . . . . . . . . 1.2505 0.311 111.0 860 1810 2.105
Oxygen. . . . . . . . . . . 1.42895 0.430 134.3 1160 2470 2.13
Ai, . . . . . . . . . . . . . . 1. 2926 0.335 115.5 922 1960 2.125
Argon . . . . . . . . . . . . . 1. 7839 0.531 133.0 1243 3065 2,46
Carbon monoxide 1.2500 0.311 1l1.0 874 1835 2.10
Carbon dioxide . . . . . . . 1. 9768 0.460 104.5 1520 3220 2.12
Ammonia. . . . . . . . . . 0.7714 0.235 138.0 II 1405 2560 1.82
Methane . . . , . . . . . . I 0.7168 0.162 101.0 I 743 1580 2.125
Ethane . . . . . . . . . . . '11.3560 0.203 67.5 I 950 2130 2.24
Propane. . . .. . . . . . .
Ethylene. . . . . . . . . . .
I I
2.0037
1.2605
Propylene . . . . . . . . . . I 1.9150
________________________________ I
l__ ._ ___________ _
0.220
0.216
0.233
50.0
77.0
55.4
I
999
975
1053
J I 2400
2255
2700
,
2,40
2.31
2.56

The experimentally determined viscosities at the critical points for CO 2

and NB3 lie on the (1i p , T--'lJT)--P curves for these gases with a fair degree
of accuracy (Figures 92 and 93). Thus, for CO 2 values of "'l(p, T)c/ound
experimentally and graphically coincide (in both cases they are equal

180
to 3220.10- 7 g cm -1 sec- 1 ). For ammonia the experimental value of lI(p. T)cr
is equal to 2610.10- 7 g cm -1 sec -1, and the value determined from the graph
is 2560.10- 7 gcm- 1 sec- 1 (the discrepancy is less than 2%). This indicates
that we can determine the viscosity at the critical point for other substances
also. Table 83 shows '1J(P. T)cr values for some substances determined from the
(7)P. T-7)T)-P graphs according to our experimental data (for deuterium the
data of Michels, Schipper, and Rintoul were used).

6. The application of the law of corresponding states to

the determination of the gas viscosity at different
temperatures .and pressures

If in the Van der Waals equation we use the reduced pressure, volume,
P v T
and temperature, 71: = - , l' = - . ~ =-T instead of p, v and T, we obtain the
P cr vcr .::r
following dimensionless equation:

(161 )

This reduced Van der Waals equation can be universally applied, and is the
general equation of state for all gases. It follows from (161) that if two
substances have equal values for any two of the reduced variables (7:. '~. 7),
the third variable is also the same (this follows from the law of correspond-
ing states).
RT
At the critical point the ratio ~ is called the critical coefficient kcl and
Perver
it should have the same value for all substances. Since in the Van der Waals
equation

we obtain

In practice, the critical coefficients of imperfect gases are not equal to

this value, and considerable discrepancies are found /130/. For the
majority of substances kcr is between 3.2 and 5.0.
The states of substances with the same reduced variables are termed
corresponding states.
The law of corresponding states says that if two or several substances
satisfying the same reduced equation of state have the same two of the
three variables 71:. 'P. ~, the third reduced variable will also be the same,
or in other words the substances are in corresponding states.

[This is Russian terminology. The reciprocal of kcris termed the compressibility factor Z.J

181
 A.ny equation of state which has only two constants characterizing the
nature of a given substance, and which is applicable to all substances, or
to at least a group, can be transformed into 11 reduced equation.
The law of corresponding states refers not only to the parameters 71:, '1', 'tbut
can be expanded to some other physical properties of substances. Thus,
there is a general equation expressing the state and physical properties
of all gases (and also liquids):

f(71:, 't, x)=O, (162 )

where x is a magnitude characterizing a certain physical property of the

substance.
When we apply the law of corresponding states, we must remember that
it is only approximate, since the behavior of different substances in the
corresponding states is similar, but not identical, especially for substances
with widely differing natures. Nevertheless, this law gives the most
general method for forecasting variations in the physical properties of
gases with temperature and pressure. This is of great importance, since
when establishing the function f(r., 't, x) = 0 for a11 gases, it is necessary to
determine it for one gas accurately.
For any other gas it is sufficient to know the critical parameters Per and
Ter •
Some authors /27,64,131/ used the law of corresponding states to
determine the viscosity of gases at different temperatures. The following
equation was used:

(163 )

Thus, it is assumed that the ratio of the viscosity at P and T to the viscosity
at 1 atm and T should be the same for all substances in the corresponding
states.
This assumption can be justified by the theorem of the mechanical
similarity of molecular systems. This theorem was first stated by
Kamerlingh Onnes /132/, and in principle says that for corresponding
"I T T'/,
states the expression ~=b is the same for all the substances.
M 2pC~
From (129), we obtain

T~F a -6
~=-, where a=3.5· 10 .
M· P2 'ler

It was shown in Chapter IV, Section 2 that "fJT ='"fJC,!f ('t) ; at the same time
ff ('t) is the same for all substances. From these expressions: 'lp, T = ~ l' ('t) ,
"IT a
p
i. e. the "fJ , T values at the same reduced temperatures and pressures
"IT
should be identical for all substances. Therefore, if we plot on a graph the
p
corresponding isotherms or isobars, i. e., the dependence of "fJ , T on
"fJT
't at r. = const, or on r. at 't = const, then "Ip, T should fit the same curve for
"IT

all substances if the law of corresponding states holds.

182
 Figure 95 shows the isotherm of "'ip, T plotted against 'It for several
"'iT
substances. The data were taken from /158/. The experimental data
were first interpolated to the same reduced temperatures, and then the
isotherms were plotted. For all the substances the reduced pressures
were over the range 1 to 800 atm, except for diethyl ether, when the range
was 1- 600 atm. It can be seen that the law of corresponding states gives
a correct qualitative description of the variation '~1', T in viscosity with
'iT
temperature and pressure, although the values for all substances given at the same
reduced temperature do not fit into a single curve, but spread over a
certain region i.ndicated by the dotted line.

B.or·---.---'r---~---r---.---'----~---r---'-~~~~

18.0

--
~

78910 f! 1213/415 16 1~18 1920

7( =p;.!

FIGURE H5. Dependence of "'ip, T for different gases on the reduced pressure at var-
"'iT
ious reduced temperatures.

There is no quantitative agreement; the discrepancies for the substances

reported are sometimes as high as 30%. On the other hand, for substances
for which the absolute values of the other physical parameters (molecular
weight, Per , Ter , etc.) are close to one another, the quantitative agreement
p
between the values 'rJ , T for identical 'It and 't are completely satisfactory.
"T

183
Thus, for example, the discrepancies between the "Ip, T values for propane
"IT
and propylene at the reduced temperatures shown on the figure are only 30/0,
and are even lower for ethane and ethylene.

10.

9
/
8
J
7 II
~
),
I
CO,0-
6- Nz
!i lI--Air
5 0- CO
!1
0-02
f- NH,
x_ Ar
,;
f--. 0

1
:/~
3
I
2 I f--I f--

I
;? I
.#) ,
~~ I
0. ~ ~ I
o 0.5 iD - 20. 25
,DtpC[

FIGURE 96. Dependence of the viscosity of some gases on the density

in coordinates "Ip, T-1)T P
"I(p, Ti cr "i(T)cr Pcr

Thus, the law of corresponding states represented by (163) can be used

only for an approximate, qualitative evaluation of the viscosity coefficients
of different substances at different temperatures and pressures. It is
interesting to note that with increase in ~ the discrepancies decrease
appreci.ably. Therefore, at temperatures above the critical, the law of
corresponding states will be more correct.
It is also quite possible to introduce some arbitrary supplementary
factors and functional relationships, which will Detter represent thc
experimental viscosity data in terms of the law of corresponding states.
Consequently, we could then morc accurately calculate the viscosity
coefficients of substances whose viscosity has not been determined, or
for which we have data over a narrow temperature or pressure range
only.

5680 184
 12 I
j
t
If ,
10
j
9

8 J
7
I I I
!6

.I
I
3
i
/
2 /
~ 0- CHip
,- C,H,
I
.".,-
~ a-
6-
C,HB
Ct H4
:
~ V 0- C,H,
o 0.5 1.5 2.0 25
jJ/):;,cr

FIGURE 97. Dependence of the viscosity of some hydrocarbons on the

density in coordinates _~YJp,-",-T_-_YJ_T_
YJ(p, T)cr-YJ(T)cr Pcr

5 I I
It
I
0-Hz
x-D z , V
!
!
'/
~/<
:
i /V
o
o
y/ l
0.4 0.8 f.2 f.6 Z.O 2.4
)'ko·cr

FIGURE 98. Dependence of the viscosity of hydro-

gen and deuterium on the density in coordinates

~p, T-Y1T

185
 Thus, if we start from the experimentally determined regularity that
the '1p, T-'1 T values as a function of the density (over a wide range of p)
fit the same curve at all temperatures for given substances, then in
accordance with the law of corresponding states, the values ~7" T - ~T as
~(p, T)cr T (Tl cr
as function of plP cr will fit the same curve for all the substances,
Figures 96,97, and 98 show the experimental data for several substances
in the above coordinates. Figure 96 shows the data for N2 , CO 2 , 02' air,
CO, Ar, and NH 3 , over all the temperature and pressure ranges that we
studied (see Tables 13,14,15,17,18,20, and 21). The data for CH 4 , C 2H 6 ,
CaRs' C 2H 4 , CaR6 (data from Tables 22,24,215,28, and 30) and the data of
Michels, Schipper and Rintoul for H2 and D2 (Tables 9 and 10), are plotted
in Figures 97 and 98, respectively. The values ,ucr''1(p, T)cr and "'i(T)cr were
taken from Table 83, compiled by us, and the values of p at the required
temperatures and pressures were calculated from the most reliable
data.

8 /
/
6
I-N 2 ,C0 2 ,o"crJ,
/ II
I ~~
air, Ar, NH 3 ·
I/-CH,.C"H"C,H,-,-
C,H4 ,C,H,
IJf- Hz ,],

J if
2 ~
/
o ~~
III
o as 1.0 l5

FIGURE 99. Dependence of the viscosity of different gases on the

d'ensity in coordinates ~p, T-1JT ..L
'lip, Ticr'l(T)cr Pcr .

These figures to help to confirm the law of corresponding states for

individual groups of substances. The deviations of the separate experi-
mental points from the most probable curve connecting the experimental
points are within the limits of experimental error or inaccuracies of the
calculation method. If we plot the most probable curves from Figures 96,
97, and 98 for the individual substances into a single graph (Figure 99), we
see appreciable quantitative differences in their shape.
For other gases whose viscosity we studied, the law of corresponding
states was not checked, because no p, v, T data were available for the re-
quired temperature and pressure ranges. The same applies to liquefied
gases, since from the experimental data available we do not accurately know
at which density the viscosity was measured.

186
Chapter VI

THE VISCOSITY OF GAS MIXTURES

1. The viscosity of gas mixtures at atmospheric pressure

There are a large number of papers on the viscosity of gas mixtures at

atmospheric pressure. The viscosity was usually found by the capillary
method. The viscosity coefficients of gas mixtures are given in reference
books (see for example /164/). Below we give an appreciable part of the
experimental data published.

2500

/~ !.---
,
2000
/'
V-I ./
V
~
"
V ~~ ../' .....
5'500
~
bI)
!~V yz~ V
.,.t/ / Hz-CHi
/
~
1000
j ../'
~ :: I
Hz-CzH6 _
f - H -C3 H"
-I z1
500
1(10 90 110 10 60 50 W SO ZO 100/ 0
"Hz
FIGURE 100_ Viscosity of gas mixtures at 20°C and 1 atm.

2500
I
~O'-...;; F-
l I---
,
"
1;l - J
CO-~12

V
'e 2000 N2O-COZ ./"
V
~bI)" Hz-C02___ j.--I V
~
~/500 V ~ -",V
V ------
~ ~
I--
.-"
1/-V ].A,H,-H
/000
-I ~
o 10 ZO SO m 50 m 10 M mw
0/0 of second component
FIGURE 101. Viscosity of gas mixtures at 126.9°C and
1 atm.

187
 2200
2/00k~l-=f::::t==:p::,hd:--\-M
2000
EOO~~4--r-+-,--~-~4-~~

T ~ro~~~+-~1==r~.
(j

f700r----r--i---C::::j;1£.==!=
::l
~1&70~~~~--T~~=4~
(J
~600~t-~-r-=~=~'~-~~tH
'0./WOb---F""'=-'---r---'-
""" '~---:-......
300~~~~--'-~--r----~~WM

200~L-+-4-~-+~-­

~OO~+-~-+--~+--r­

moo~+-4-~---+--r­

WO~~~~~~~~c~~~
o 70 80 gO 100
XHz
FIGURE 102. Viscosity of an ti2-S02 mix-
ture at 1 atm and different rem peratures.

TABLE 84. The viscosity (10- 1 g cm- 1 sec- 1) of gas Ulixtures according
to Trautz and his co-workers

tiydrogen- Nitrogen

~%H'I
t,OC~
100 80.77 79.51

-78 676 896

-38 7.54 1116
+ 19 874 1305
100 1030 - 1553
200 1212 - 1827
250 1297 - i 1952

~I 50.81 I 50.53 I 49.37 I 20.21

I I
-78 - - 1183 1246 - 1263
-38
+19
-
-
I -
1598
I 1361
- II
1440
1703
-
-
1464
1739
100 - 1900 I! - -- 2050 2084
200 2230 - i -- -- 2412 2461
250 2385 I
, -
I \
- -- 2575 I
I
2629

Hydrogen - Carbon monoxide

~%H'I
t,OC~
100 69.87 69.47 51.21 50.82

-78
-38
I 676
754
-
-
1081
1242
-
-
1179
1361
+19 874 - 1449 - I -
100 1030 (1672) - - -
200 1212 2016 - 2230
II -
250 1297 (2144) - I 2382 -

~lt,OC
47.55 40.96 37.37 21.84 19.27

-78 1219 1250 1264

-38 1410 1455 1476
+19 1610 1651 1717 1745
100 1920 1971 2038 2085
200 2330 2405 2466
!
250 2489 2574 2636

188
TABLE 84 (Continued)

Hydrogen - Neon

''''~% H'I
t. 0C ~ 100.0 88.95 I 77.15 I 46.09 I 25.20 I 0.0

I
20 875 1301 1684 2427 2782 3092
100 1029 1529 1981 2845 3269 3623
200 1211 1795 2319 3327 3887 4220
250 1296 1917 2476 3540 4054 4501

Hydrogen - Helium

'-",~ H'I
t. "C ~ 100.0
69.18 60.69 55.20 0.0

20 875 1166 1252 1317 1974

100 1029 1383 1478 1551 2320
:200 1211 1619 1728 1817 2715
250 1296 1732 1852 1939 2903
1

Hydrogen - Argon

~~~:I 100.0 I I I I 65.15 62.62 44.57 29.42 0.0

20 875 1857 1895 2056 2140 2211

100 1029 2238 2275 2488 2586 2684
200 1211 2636 2697 2948 3070 3208
250 1296 2826 2894 3164 3310 3448

Hydrogen - Ammonia

':~\ 100.0 \89.18\77.61 \70.25 \ 48.23\29.131 9.95 \ 0.0

20 875 1011 1072 1087 1080 1047 1004 982

00 1029 1204 1299 1329 1354 1333 1299 1279
00 1211 1432 1560 11610 1676 1680 1660 1646
50 1296 - 167811737 1823 1837 1825 1813
I

189
TABLE 84 (Continued)

Hydrogen - Methane

'X
%H'I/, ·C
100.00
I 92.23 I 60.22 I 48.55 I 28.08 I 0.00

20 876 955 1086 1098 1099 1087

100 1032.5 1132 1306 1328 1337 1331
200 1212.5 1337.5 1551 1586.5 1602 1603
250 1296 1423 1661.5 1699 1718 I 1724.5
i I
Hydrogen - Ethane

'XI /, ·c
100.00
I 85.68
I 85·15
I
50.90 45.00 I 0.00

~JJ'
995 993 988 977.5 909
100 1032.5 1194 1189 1216 1208 1142
200 1212,5 1417.5 1412 1469 1466.5 1408.5
250 1296 1517 1511 1586 1588 1526

Hydrogen - Ethylene

~I
/, ·c
100.0 81.07
I 80.82 80,43 12.01 70.33

-78 718 731

-40 818 839
-1 943 959
+20 1012 1039 1053
55 1122 1154 1164
100 1264 1291 1298
150 1408 1432 1441
200 1547 1568 1574
250 1681 1694 1699

~I t,Oc 61.44 I I
51.97 51.73
I 25.01
I
21.60
I 21.14 I 0.0

-
-78
-40
754
852 -
I -- 772
866
-
--
-
-
670
740
-1 985 -- I - 996 - - 830
I
+20 - - 1067 - 1060 - 873
55 - - I 1173 -- 1156 - 943
100 \ -_. - 1311 - I - 1278 1030
150 - 1463 I - - - 1409 I 1123
i
200 ' - 1588 -- I - I - 1529 I 1211
! -- I
250 - 1708 - -- 1627 1294
i
I I II I ._-
.- I

190
TABLE 84 (Continued)

Hydro gen - Propylene

:~~! 0.0 110.7117.2125.11 M.a!46.5164.7176.aI8a.219z71 100.0

20 876 982 994 9951980 959 9101 887 875 858 844
100 103111731205121111201 1181 114511122 1109 1090 1076
200 1210 1395143814531453 1440 1410 1388 1373 1353 1339
250 1296149611541115661158011568/152911509[149411477 1467

Hyd rog en - Propane

':~! 0.0 118.21 I 37.04/ 5818 178.82/ 87.50 192.25 1 100

26.9 817 836 874 924 985 987 970 891
126.9 1070 1091 1130 1172 1233 1221 1194 1081
226.9 1308 1324 1366 1417 1459 1433 1392 1256
276.9 1422 i 14381152911566 1566 1536 1485 1347
l

Hydrogen - S -Butylene

' . . . ~%c.H81
t. "'C
0.0 1125118.2120.4136.1 151.3164.2177.3192.81100.0

20 876 945 944 939 893 8441 810 780 754 747
100 1031 1135 1145 1142 1102 1054 1020 991 958 944
200 1210 1348 1372 1376 1351 1308 1275 1243 1207 1192
250 1296 1446 14771 1480 147211422 1389 1357 1318 1301
I 1

Helium - Argon

'~I
t •. "'c
100.0 49.06 38.20 34.05

20 1973 2296 2291 2278 2211

100 2320 2750 2745 2736 2684
200 2715 3250 3208
250 2903 3488 3448

191
 TABLE 84 (Continued)

Helium - Neon

'--~I
t, 'C !
100.0 76.21 56.24 26.59 0.0

20 1941 2429 2702 2971 3092

100 2281 2846 3171 3479 3623
200 2672 3327 3702 4056 4220
250 2853 3555 4310 4501

Neon- Argon

~I
~ ,I
100.0 74.20 60.91 26.80 0·0

20 2213 2401 2504 2808 3092

100 2693 2885 2990 3313 3623
200 3222 3413 3529 3890 4220
250 3460 3658 3793 4150 4501

Hydrogen - Carbon dioxide

~I 0.00 111.79119.931 41.291 59.461 78.50 1 88.881100.00

--------~--~--~----7_--~
I I
1
I
26.9 1493 1502 1501 I 1506 1478 1370 1232 891
126.9 1944 1951 1945 1933 1878 1713 1526 1081
226.9 2353 2360 1 2358 2321 2239 202611783 I 1256
276.9 i 2556 i 2554 I 2542 2506 2471 2173 1904 I 1341
_ _ _I. 1 I I I I
Nitrogen - Oxygen

~It, 'C
18.64 1 24.08 58.93
I
59.20 78.22 100

I
26.9 I 2057 2008 1995 1894
I 1893 1843 1781
126.9 2568 2489 2480 2345 2342 2275 2190
226.9
I 3017 2920 2909 2741 I 2741 I 2658 2560
276.9 3220 3109 2932 2928 2840 2727
1
I
-l-

5680 192
 TABLE 84 (Continued)

Hydrogen - Oxygen

·~x I 0.0 1 18 .35 1 39.451 60.30 / 78. 08 1 86.331 95.861100.00

,
26.9 2057 20191192511784 1494 1314 1053 889
126.9 2568 2507 2381 2192 1858 1602 1282 1087
226.9 3017 2950 2790 255612158 1867 1488 .1259
276.9 3220 3147 2978 2733 2288 1991 1589 1381

Ethylene - Oxygen

~"~C'H'I
I."C ~ 0.0 4.21 / 13.06! 41.451 60.81 / 7).031 100.0

20 2019 1967 1854 1529 1341 1198 1010

50 2181 2125 2004 1658 1456 1308 1107
100 2433 , 2376 2243 I, 1865
I
1645 1479 1262
! I
!
I

i
!
I

Ammonia - Oxygen

':~~I 0.0 13.51 29.86 4•. 86 70.79

I 87.55 100.0

20 2023 1924 1783 1604 1350 1143 982

100 2440 2326 2170 1972 1689 1459 1279
200 2902 2773 2604 2390 2085 1840 1646

----.~-

Carbon dioxide - Ethylene

18.28/ 26.32143.54165.57 1 80.64 1 87 .20 100

26.9 1776 1616 1553 1402 1253 1135 1107 1033

126.9 2183 2008 1943 1763 1581 1460 1420 1342
226.9 2548 2363 2279 2098 1891 1760 1716 1622
276.9 2714 251612433 2240 2039 1900 1851 1753

193
TABLE 84 (Continued)

Nitrogen - Ethylene

~I ° I B.OO I 24. 05 1 41. 94 1 5695 1 76.21 1 85.93 1 100

26.9 1781 1715 1574 1417 1308 1169 \110 1033

126.9 2190 2108 1956 1785 1655 1491 1430 1348
226.9 2560 2464 2292 2100 1963 1786 i 1714 1622
2769 2727 2636 2453 2254 I 2108 1921 I 1848 1753

----- I I i

Ethylene - Ammonia

~C'H'I ° 111.33119.291 30.391 48.281 70.071 89,041 100·0

_t.OC~

20 982 1001 1013 102211030 1027 lOIS 1008

100 1279 1294 1301 1304 1303 1291 1269 1257
200 1646 1647 1648 1639 1622 1595 1561 1541
250 1813 I 1809 i 1805 [1791 1764 1729 1689 1666
I !

Nitrogen - Carbon monoxide

x l 0.0 118.46122.891 3970 1 6S.68 1 77-80 1 83.71 I 100.0

26.9 1781 1782 1781 1781 1775 1778
1774 1776
126.9 2190 2186 2193 2183 2191 2184
2184 2183
226.9 2560 2560 2555 2558 2549 2551
2551 2548
I
276.9 • 2727 I 2721 2722 I 272212721 1 2719 2714
I
i 2719
,
I I ,

Oxygen - Carbon monoxide

~ll00.o I 82. 28 1 76.63 1 57.991 ,0.73\ 22.6711B.061 0.0

26.9 1776 1824 1841 1900 1948 1998 2012 2057

126.9 2183 2250 2268 2343 2407 2482 2501 2568
226.9 2548 2626 2650 2741 2820 2908 2928 3017

194
TABLE 84 (Continued)

Ethylene - Carbon monoxide

'-~ 1100.0 I 81. 72 1 73. 68 1 56.461 34.43119.38112.80 1 0.0

26.9 1776 1616 1553 1402 1253 1135 1107 1033

126.9 2183 2008 1943 1763 1581 1460 1420 1342
226.9 2548 2363 2279 2098 1891 1760 1710 1622
276.9 2714 2514 2433 2240 2039 1900 1851 1753

Carbon dioxide - Oxygen

117.721 23.371 42.01 I 59. 27 1 77.331 81.94 1 100

26.9 1776 1824 1841 1900 1948 1998 20121 2057

126.9 2183 2250 2268 2343 2407 2482 2501 I 2568
226.9 2548 2626 2650 2741 2820 2908 2928 3017

Nitrogen - AmlTIOnia

''--~ NH'/ 11.11 I 29.20 I 56.38 I 71.47 I 88.33 I 100.0

t.oC~

20 1745 1690 1585 1383 1254 1092 982

100 2085 2031 1920 1710 1569 1398 1279
200 2462 2408 2296 2085 1946 1768 1646
250 2627 2572 2460 2250 2112 1939 1813
i

Nitric oxide - Nitrogen

~~NO/ 100.0 69.48 58.37 26.74 0,0

t,oC~

20 1882 1833 1827 1778 1747

100 2272 2222 2209 2132 2084

195
TABLE 84 (Continued)

Ni trous oxid e - Propane

~It.·C ,
0.0 20.16 40.16
I 58.29
I
79,82
I 100.0

26.9 I 817 926 1046 1167 1326 1488

126.9 1070 1213 1372 1525 1725 1943
226.9 1308 1478 1667 1854 2083 2355
276,9 1422 1610 1807 2012 2271 2556

Nitrous oxide - Carbon dioxide

~'CO'I
I. ·C
0.0 I 19.97 I 40.24 60.33 I 80.97 I 89.13 I 100.0

I
26.9
126.9 1943
1488
1942
- 1491
1950
1495
1950
1490
1941 I
1495
1945
1493
1944
226.9 2355 2357 2365 2365 2358 2358 2353
276.9 I 2555 2555 2562 2564 2551 2355 2556
I I
Nitrous oxide - Hydrogen

~I 0.0 39.89 I 59.61 I 78.57 I 91.10 I 100.0

I
26.9
126.9 I
1488
1943
1481
1907
1451
1849
1348
1684
I 1201
1484
891
1081
226.9 2355 2292 2206 1990 1704 1256
276.9 2555 2477 2376 2137 1863 1341
!

Carbon dioxide - Propane

~I 0.0 18.93 I 40.25 I 57.76 78.83 I 100.0

26.9 817 926.4 1058 1174 1326 1493

126.9 1070 1213 1383 1533 1730 1944
226.9 1308 1461 1670 1856 2093 2353
276.9 I 1422 1601 1815 I 2010 2267 2556
! I I I

196
TABLE 84 (Continued)

Methane - Ethane

~~IIOO.O I 81. 16 1 56.70 I I 48.74 20.45 1 19 . 03 1 0·00

20 1087 1045.5 999 986 938.5 937.5 909

100 1331 1288 1239 1226 1173.5 1174 1142

200 1603 1562 1511 I 1496 1441.5 1442 1408.5

250 1724.5 1682 1630 1614 1560.5 1559.5 1526

.-

Methane - Propane

'~I t, ·C
100.00 63.16 36.17 16.59 0.00

20 1087 948 878 831 801

100 1331 1182 1100.5 1042 1008

200 1603 1440.5 1355 1291 1255

250 1724.5 1553 1465 1403 1362.5

Ethane - Propane

~I t,OC
100.00
I 43.27
I 25.63
I
15.26 II 0.00

20 909 841 828 814.5 801

100 1142 1058 1039 1025 1008
200 1408.5 1313 1298 1272 1253
250 1526 1425 1401 1382 1362.5
,,

Propane - Propylene

~I 0.0
26.08 50.97 78.69 100.0

40
I 904
I
896 886 878 873
60 960 952 946 940 932
80 1016 1009 1001 996 988
100 1070 1062 1056 1046 1040
I
\97
TABLE 84 (Continued)

Hydrogen - Nitrogen
(Van Ittetbeek, Van Paemel and Van Lierde /144/)

~1100.0 1 84.0 1 84.9 1 55.9 I 38.0 I 24.1 1 134 I 0.0

291.1 877 1251 - 1560 11660 1677 17421 1752

90.2 392 523 604 620 I 629 640 645 651
82.2 362
I
473 509
I
519 533 537
-I 544

TABLE 85. Viscosity of vapor mixtures of methanol- water and ethanol

- water (Silgardo and Storroy /145/)

Methanol- Water

Mole "/0 of Condensation

methanol temperature,OC at condensation I
______________T-__________~,--te-m~p~e-r.a-t-u-re~!I-----------

0.0 1000 125;) 1255

15.0 96.39 1248 1261
36.5 89.73 1225 1265
52.5 83.89 1207 1267
69.0 76.73 1180 1268
85.0 70.1 1146 1249
98.0 65.22 1112 1228
100.0 64.61 1097 1223

Ethanol - Water

7/, 10-7 g em -1 sec- 1

Mole "/0 of Condensation
ethanol temperature,oC at condensation
at 1000 C
temperature

0.0 100.0 1255 1255

9.0 97.71 1251 1258
17.4 95.17 1233 1252
25.7 92.5 12'20 1238
37.4 88.62 1184 1226
54.0 82.72 1149 1197
69.8 79.51 1116 1183
83.4 78.51 1088 1163
91.9 78.0 1077 1139
96.8 78.0 1070 1114
100.0 78.0 1016 1080

198
TABLE 86. Viscosity ofternary mixtures (Trautz and Kipphan /146/)

Com position, vol 0/0

t, °C 1),10- 7 g cm- 1 sec- 1
Ne Ar He

55.76 26.70 17.54 -80 2740

I 100
200
3237
3790
31.93 32.13 35.94 -80 2569
100 3044
200 3574
21.66 5851 19.83 -80 2411
100 2886
200 3415
21.89 23.82 54.29 -80 2504
I 100 2957
I
I
200 3470

TABLE 87. Viscosity of multicomponent mixtures at 20°C

(Herning and Zipperer /143/)
-
Com position, vol 0/0
7 1 1
heavy 17,10- cm- sec-
CO 2 O2 CO H2 CH. N2 hydro-
carbons
-
8.6 2.3 89.1 1756
13.3 3.9 82.8 1749
6.2 10.7 83.l 1793
10,4 28.5 1.6 59.5 1738
10.6 29.8 3.9 0.3 55.4 1743
8.9 30.7 3.3 0.4 56.7 1747
8.7 32.8 1.5 0.2 56.8 1749
1.7 0.9 6.0 57.5 24.0 7.8 2.1 1262
2.1 0.9 5.7 53.0 24.3 11.7 2.3 1304
2.0 1.4 4.6 54.9 23.5 11.6 2.0 1310
3.3
2.2
0.6
0.6
3.8
4.1
51.3
53.1
29.6
29.5 I
10.0
9.2
I 1.4
1.3
1332
1306
2.2 1.0 4.0 52.3 1.2 1307
29.91 9.41
2.5 0.8 14.9 53.0 I 18.1 9.1 1.6 1355
I
4.8 0.3 26.4 17.21 2.6 48.2 I 0.5 1714
3.5
3.1
0.3
0.5
27.3
28.6 I
14.4
17.71
3.71 50.0
4.2 45.0
I 0.8 .
0.9
1712
1715
I

199
 TABLE 88. Viscosity of multicomponent mixtures at different
tem peratures (Schmidt /147/)

Composition, vol 0/0

t, °C 1],10- 7 g cm- 1 sec -1
CO, 0, CO I H, CH4 N,

10.8 2.0
I 2.2 I 85.0 27.5 1827

I 251.5
700.0
2715
4117
1006.0 I 4856
6.7 0.1
'8:I 2.2 83.2
I
34.5
246.0
1842
2655
702.0 4048
I 1012.0 4808
6.4 3.0 07 896 41.0 1904
0.31
245.0
701.5
I 2706
4113
I 1014,0 4895
6.0 0.1 25.71 1l.5 56.7 1823
29.0 1
25:,.0 2686
4041
I 703.0
1010.0 I 4777
3.7 0.3 27.1 9.5 1.6 57.8 27.5 ! 1815
292.5 I 2819
708.0 I 4045
1009.0 4792

The experimental data show that the viscosity of gas mixtures cannot be
calculated simply by the additivity rule.
This rule can be applied to mixtures of gases with rather similar
physical properties, such as oxygen - nitrogen or propane - propylene.
A typical dependence of the viscosity of binary gas mixtures on the
composition is shown in Figures 100, 101, and 102. The viscosity
coefficients for some gas mixtures are shown in Tables 84- 89; the
composition is expressed in volume percent.
There have been no systematic measurements of multicomponent gas
mixtures, so that very few literature data are available. We should mention
the exhaustive measurements of Trautz and his co-workers (capillary method)
/133- 139/, and the papers of Adsumi /140/, Jung and Schmick /141/,
Buddenberg /142/, and Herning and Zipperer /143/.

TABLE 89. Viscosity of some gases at high temperatures (Kenney,

Sarjant, and Th"ing /148/)

Cornposition of gaseous mixture

44 1844
288 2723

10.62% CO 2 + 88.65% N2 + 0.73% O 2 414 3121

540.5 3536
754.5 4184
887.S 4516

200
 T ABLE 89 (Continued)
Composition of gaseous mixture

32 1763
275.5 2733
464 3220
631,5 3805
766 4111
861 4413

- "_.- ------

18.2 1703
130 2169
223 2489
358 2901
19.82% CO 2 + 79.54% N2 + 0.64% O 2 538 3477
659 3823
774 4187
873 4374

2. The calculation of the viscosity of gas mixtures at

atmospheric pressure

In gas mixtures the momentum is transferred by the molecules of all its

components. The total momentum transferred from layer to layer is equal
to the sum of the momenta transferred by each component, so that

(164)

where nt, n2 , • • • are the numbers of molecules of the components present

per unit volume: m" m2 . ••. are the masses of the molecules of the compo-
nents; and 1;. 1;. . ..are the mean free paths of the molecules of the individual
components in the mixture. For a gas consisting of two components, 1~ and
t;are determined from (24) and (25), as follows:
for the molecules of the first component

(165)

and for the molecules of the second component

,; = 1 • (166)
l r-2"2
d2 n , 2 ~rM,+M"
2T"S n 1 V ~

If we multiply and divide the right-hand side of (165) by (Jf2 ..d;nl) and of
(166) by (V2.-:d:;n,), taking into account (164) and the equalities

201
 kmtC, km2C2
1),= V 21tdi' 1j2=V21td~' after simple transformations we obtain for the
viscosity of binary mixtures

(167)

where s = d ~ do By combining all the coefficients (d" d 2 , M, s) of each term

t

into one coefficient, from (167) we obtain

( 168)

This form was first derived by Sutherland /149/, and later by Thiesen
/150/.
Many authors have proposed different types of equations for determining
the viscosity of binary mixtures, and also some modifications of (168).
Maxwell /105/ derived an equation which can be represented as

( 169)

Enskog /4/ proposed the following equation:

(170)

This equation differs from the Maxwell equation in the term 4~ in the
"h~e

denominator. For elastic spheres the constants have the following values,
if no attractive forces act:

Equation (168) is applied in the papers of Jung and Schmick /141/ and
Schroer /151/. These authors use the following values for A, and A 2 :

M, ,
A, = (..2..)2 1- ~ 6 ~r Mt + M, 112 (171)
d, 0.797 V 2M2 Ii'

A _ (..2..)2 1- M1 ; Me 62 ~ r M, + M2 ~ (172)
z-
---
dz
I '----
V 2M, 12 •
0.797
II - - - -
III

202
 In these equations three factors are taken into account: (I), differences
in the diameters of the molecules; (II), differences in the masses of the
molecules; and (III), the forces of interaction of the molecules. The following
symbols are used:

where CI2 = YC;C;, CI , C2 and CI2 are the Sutherland constants for the first and
second components, and for the mixture, respectively;

d~= 0.7063 VMIO- 2 (173)

'IT~T VT
where d is given in A. The constants AI and A2 in (168) calculated from (171),
(172), and (173) for some binary mixtures are shown in Table 90.
An empirieal equation like (174) is widely used in engineering to calculate
the viscosity of multicomponent mixtures /143/:

(174)

where Ien. Tcr"' ••• are the critical temperatures of the components. This
equation can be used only for approximate calculations of the viscosity
coefficients of gas mixtures. The so- called reciprocal equation of Mann
/152/ is also an approximate formula for the determination of the viscosity
of multicomponent mixtures

(175)

where ul • u2• • • • are the dynamic viscosity coefficients of the components.

Schudel /153/ simplified the Sutherland- Thiesen equation (168) as
follows. The author proved that

(176)

so that (168) can be represented as

(177)

Thus, (177) contains only one constant B. It was sh~wn in /154/ that B
can be expressed as a function of the coefficient of diffusion DI2 of gas pairs

(178)

203
 A,

0.931
1.2038
0.370
0.380
0.2494
0.3526
0.416
0.256
0.802
0.210
0.1956
0.6475
0.8609
0.6669
1.021
0.7499
0.7301
1.061
1.1018
08699
0.8277
0.815
0.n9
0.937
1.045
0.7636
0.765
1.00

It follows from (177) and (178) that

(179)

where P1 and P2 are the densities of the components.

A more general f@rm of (179) /155/ is

(180)

where? is a constant for the given gas mixture.

From (179), in self-diffusion (500/0 of component 1 + 500/0 of component 2)
we obtain

204
 ~ = ~1 = D~~Pl • (18l )

~ = ~2 = D~~Z. (182)

where Dll and D22are the coefficients of self-diffusion of the components of

the mixture. The relationship between the coefficient of self-diffusion of a
binary gas mixture D12 and the coefficie)l\ts of self-diffusion of the gas
components, Dll and D 22 • is expressed by

( 183)

From (181), (182), and (183), we obtain

(184)

By taking ~'=~1=~2 and combining (180) and (184) to eliminate the

diffusion coefficient, we obtain

[(~1:)]~ CM1)~ r+ + ~{I+[(~:)]~ (M )+ r'

"112= (185)
!:!.!{ 1+ 2

1 + nl ~Z PI ~Z 1 nz ~1 P2 ~1
_4
V2
(I + Ml)2
M2
_4
V2
(I + Ml
M2)2
For a gas at atmospheric pressure, (185) can be expressed as

"I/l~: == (186)

If we represent

(187)

and

Cf21 (188)

205
then from (186) /157/, we obtain

(189)

For multicomponent mixtures (189) has the following form:

(190)

where

'fij =
[1 + (~i.)+ (~l
~J , 1
r (191)
_4 (1+Mi)2
V2 Mj

Equation (190) can be considered as a summation of the partial viscosities

of each component, i. e.,

i='Z

"l",=~['1d· (192)
i=l

The partial viscosity is given by

(193)

Equations (189) and (190) can be used in engineering, since to calculate

the viscosity of a gas mixture of any composition, we must know only the
molecular weights and viscosity coefficients of the pure components. These
4 .
equations do not contain any empirical numerical constants, since V2 IS

derived from the kinetic theory.

To facilitate the use of (189)' and (190), we plotted 'f'ij as a function of the
ratio of the molecular weights and the ratio of the viscosity coefficients of
the components (Figures 103 and 104).
The viscosity coefficients of binary and multicomponent mixtures
calculated from these equations agree well with the experimental results.
Below we give an example of the calculation of the viscosity of a multi-
component mixture at 283°K. This illustrates the use of the equations and
graphs. Table 91 shows the composition of this mixture, and the procedure
is shown in Table 92.

206
 60 0.0.60.0.801 02 03 04 06 08 I 23.681

02 .03 04 08.08 J 2 3 •

FIGURE 103. Dependence of 'P" on ~i for different :'ll values Cindicated

by the numbers on the lines). 'J j 'fjj

~~ ":',1 I,

3 ""- . ~,,~' T~'I-IT!-t-t-t-Ht~+-+-t-+++-+++--t+t--+--!

2 0.L' ."":~ -+-t-i-+lctt--I--I-++++--I-+-I-++-I--+---l
!)-;' ~~ !

t?

;:-.; ::K~~ ,
......
~~~-1
:-.... ~~. I
I

fK"~C"
i"'- ~~f:S: ~~

~~~
,
I i::'~!§:
I ~~ ~~~ ./ ~~~ ~
o/mE~mn;
0.08
006 :
ft§SDl.~OI
~ ::>::;::006
0.08

!?o, ! I 'I:::":: '§$;;; 0.04

r/o.J I I! I , 0.03
06 2 6 10. 20. 40. 60 100 M, 200
711
M·
FIGURE 104. Dependence of 'Pij on M ~ for different 2!i values
J ~j
(indicated by the numbers on the lines).

207
 TABLE 91

V iscosi ty of pare
Component Vol "/0 components,
'q, 10- 6 g cm-!sec-!
--------------~--------~
I
CO2 4.8 147.2
O2 0.3 201.9
CO 26.4 174.9
H, 17.2 87.55
CHi 2.6 108.7
N2 48.~ 178.1
Higher hydrocarbons 0.5 90.9
(C,H 6)

TABLE 92
. I
Component
2 Mi frolul
I'-Pij
ni n. n/~'ij
Mj !Fig,103]
"'4j
iand 104,
J
I
CO 2 0, 0.731 1.375 i 0.72 I 0.048 0.003 I 0.0022
CO 2 CO 0.884 1.571 I 0.73 0.048 0.264 0.1930
I
CO 2 H, 1.682 21.83 0.18 0.048 0.172 0.0310
CO2 CH. 1.358 2.744 0.67 0.048 0.025 0.0174
CO 2 Nz 0.828 1.571 0.72 0.048 0.482 0.3470
CO 2 I c, H6 1620 1.464 1.03 0.048 0.00.5 0.00.')2

~ njT;j = 0.5958

Analogously we obtain:
CO ~ n,'fij = 0.6235
O2 ~ nj?iJ = 0.8963
H2 :E n j'f ij = 1.6099
CHi ~ nj?ij = 0.8954
~ njT;} = 0.4188
(2H6 ~ njT;j = 0.6379
----------~--------~~~ ------------
Component i

CO 2 12.42 11.0) . 10- 6

O2 29.88 0.65.10- 6
CO 2.36 52.0.10- 6
H2 9.36 8.45. 10- 6
CHi 34.50 3.06.10- 6
N2 0.87 95.30. 10 -s
C?H6 127.6 , 0.71 . 10- 8
I~ [');1 = 171.2-viscosityofmixture'
, The experimental value of the viscosity of this mixture is 171.4 •
'10- 6 g cm-! sec'! /143/, We refer the reader to /108!, which
describes other principles for calculating the viscosity of gas m'x-
tures at atmospheric pressure, and gives satisfactory results*

3, The viscosity of gas mixtures at high pressures

Of the papers published on the viscosity of gas mixtures at high pressure

we shall mention the one on the viscosity of a mixture of 64.1% H2 +35.9% N2
at 30 and 70°C and at pressures between 50 and 195 atm, carried out by

5680
208
Boyd /70/ by the capillary method, and the one on the viscosity of a
mixture of methane + propane carried out by Bicher and Katz /45/ by the
rolling ball method. The data of Boyd are rather inaccurate, but the results
fDr the methane-- propane mixture will be discussed in detail.
The author of t.his book together with other workers /122, 158/ used the
f::>urth and fifth modifications of the setup for the capillary method, and
neasured the vis(~osity of numerous binary and multi component gas mixtures
c,ver wide ranges of temperature and pressure. Most of the data are shown
in Tables 93- 111.
To show the functional relationship between the viscosity of gas mixtures
and the temperature, pressure, and composition, we plotted some of the
data graphically.
In the tables and graphs the compositions of the gas mixtures are
expressed in volume or mole percent. The reader should remember that for
all gases the volume and mole percent are almost the same, since the molar
volumes of all gases are the same. Figures 105- 111 s'how the dependence
of the viscosity of some gas mixtures on the temperature and viscosity. It
can be seen that variations in the viscosity with temperature and pressure
are qualitatively the same as for simple gases, i. e" the viscosity of a
mixture increases with increase in temperature, and the increment is
larger at lower temperatures.

TAil LE 93. Viscosity (10- 7 g em -1 see - 1) of 75,,/, H, + 25,,/, N, mixture

(Golubev and Petrov)

~"t,CC I 15 30 100 130 200 250

p, an:i~

I
1 1370 1473 1630 1780 1925 2065
0 1395 1495 1550 1795 II 1935 2075
100 1425 1520 1670 1810 1950 2085
20 0 ; 1490 1575 1710 1845 1980 2110
300 1565 1640 1760 1885 I
2010 2140
40 0 1645 1710 1815
I
1930 I 2050 2170
50 0 1730 1785 1875 1975 I 2090 2200
60 0 11115 1860 1935 2025
I 2130 2235
800 1995 2010 2065 I 2130 : 2210 I 2300

r IIBLE 94. Viscosity (10- 7 g cm- 1 sec-I) of 46"70 H,+54% N,

mixture (Golubev and Petrov)

~I
p, ann
15 30 100

1610 1745 1930

50 1650 1790 1965
100 1725 1850 2010
200 1890 1980 2110
300 2075 2130 2230
400 2270 2290 2395

5680 209
TABLE 95. Viscosity (10- 7 g cm- l sec-I) of 65.8"/0 H2+ 34.2"/0 CO
mixture (Golubev and Petrov)

~t,cCI
I

15 25 50
p, ann~1 I I I 100 150 200 250

1 I 1465 [ 1500 1585 1750 1905 2055 2200

50 1500 1525 1605 176.5 1925 2070 2215
I
100
200
II
1585
1620 I
1560 1625 1785 1945 2085 2230
I 1640 1700 1845 1990 2125 2265
300 1715 1730 1780 1905 2040 2175 2305
400 I
I
1815 1825
I 1925 1870 1975 2100 2225 2355
500 1915 1960 2050 2165 2280 2400
I
600 2025
I 2030 2055 2135 2235 2340 2450
800 I 2245 2260 2310 2380 2465 2555
I I

TABLE 96. ViscosilY (10- 7 g cm -1 sec -1) of 33.30"/0 H2 + 66.7"10 CO

mixture (Golubev and Petrov)

'~l
p, atm' I 15 25
I
50
I
100
I
150
I
200
I 250

1 1665 1710 1810 2000 2175 2340 2505

50 1730 - 1865 I 2045 2220 I 2370 I 2525
100 1810 1845 1930
I 2100 2270 2410 2555
200 2010 2030 2095 2370
300 2250 2260
2225
2290 \ 2375 2495 I
2500
2590
I 2620
2696
400 2500 i 2490 2490 2540 2625 I 2700
500 2740
I

2720 I 2700 2710 2755 2800

II 2775
2855
I 2956
600
800
i
!
2990
3460 3410 I
2900
3310
2880
3220
2890
3160 I
I
I
2910
3130 \ 3095
2930

TABLE 97. Viscosity (10- 7 g cm- l sec-I) of 91"10 CO+ 9"10 H2 mixture
(Golubev and Petrov)

~ocl 23 50 100
p, ann ---........\

1655 1760 1865 2075

50
I 1725 1820 1920 2115

100 \ 1860 1930 2010 2180

200 2220 2220 2260 2365

300 2605 2550 2535 2600

400 3025 2875 2800 2845

500 3200 3060 3085

600 3525 3315

800 4175 3775

210
TABLE 98. Viscosity (10- 7 g em -1 sec -1) of 33.5% CH4 +66.50/0 H2
mixture (Golubev aad petrov)

I
P'~
"" I 'C 9.4 52.4 100.8

1055 1200 1340

20 1075 1205 1345
60 1105 1215 1350
100 1135 1245 1360
200 1235 1325 1415
300 1345 1425 1485

TABLE 99. Viscosity (10- 7 gcm- I sec-I) of 800/0 CH4 +20% N2 mixture
(Golubev and Petrov)

I
P"~
"" I 'C 19.8 50.5 100.1

1175 1280 1410

20 1205 1305 1460
60 1300 1390 1515
100 1435 1510 1595
200 1940 1875 1860
300 2520 2310 2175

TABLE 100. Viscosity (10- 7 g cm- I sec-I) of H2 +C02 mixtures at 13°C

(Golu.bev and Petrov)

"--.." " co,

p,al~
I 10 15 20 25

I
1 862 1060 1185 1255 i 1305 I 1350 I
50
100
871
881
1065
1075
1190
1200
I 1270
1285
1325 I 1.365
1350 1410
150 892 1090 1220
I
1315 1390 1455
I

TABLE 101. Viscosity (10- 7 g em -I sec -1) of 930/0 H2 + 70/0 CO2

mixture at 50°C (Golubev and Petrov)

p, arm ~

1 1230
50 1235
100 1245
150 1265

211
TABLE 102. Viscosity (10- 7 g cm- 1 SeC- )
'
of 62"/0 N, + 38"/0 CO, mixture
(Golubev and Petrov)

~ocl 16 50.S 100.3

p, atln ~I

1640 1812 2055

20 1670 1840 2080
60 1795 1935 2160
100 2005 2090 2265
120 2125 2175 2320

TABLE 103. Viscosity (10- 7 g cm- 1 sec- 1) of 80.7% N,+19.3"/0 CO,

mixture (Golubev and Petrov)
' .
• ~~ t,"C
].3 R 52.4 100.0
p,aon~,i

1680 1860 2100

20 1710 1890 2125
60 1815 1980 2200
100 1965 2100 2300
150 2200 2275 2445
200 2480 2485 2610
240 2715 2660 2750

TABLE 104. Viscosity (10- 7 g cm- 1 sec- 1) of 45.35"/0 N,+37.85,,/0 CI1, +

+ 16.80"/0 H, mixture (Colubev and Petrov)
......
~t,OCl 21.3 20.5 100.4
p, ann~1

1530 16..')0 1865

20 1550 1670 1880
60 1630 1720 1920
100 1715 1800 1975
200 2005 2040 2165
300 2340 2320 2380

TABLE 105. Viscosity (10 -7 g em -1 see -1) of 55.5"/0 N, + 24.7"/0 H, +

+ 19.8"/0 eH, mixture (Golubev and Petrov)

~I
p, atln
-----
12.6 54.0 100.8

1525 1680 1840

20 1550 1695 1850

60 1620 1755 1905

100 1710 1825 1955

200 1975 2035 2115

300 2265 2270 2n'i

212
 TABLE 106. Viscosity (l0-7 g cm- 1 sec- 1) of 62.5% H,+20.5% N,+
+ 170/0 CH., mixture CGolubev and Petrov)
-, tOe I
..l:0~ I I
25 50 100 130 200

I i 1350 1425 1580 1730 1885

100 1415 II 1490 1630 1775 1915
200 1495 1560 1695 1825 1~00

300 1595
I 1645 1760 1875 2005
400 1710 1745 1830 1940 2050
500 1815 1835 1910 1995 2100
600 1920 1935 1985 2060 2155
800 2145 2130 2145 2195 2265
TABLE 107. Viscosity (10- 7 g em -1 sec -1) of methane + propane mixtures (lvieshcheryakov and Golubev)

% Propane I - 46.4 I 75.6

~oc
p. atm
. 75 I I 100 1202r 15D I 200 I 250 I 100 I 150 I 200 \ 250 \100 \150 I 200 I 250

1
1169 12 41 1310 1377 1504 1620 1152 1284 140711518 1064 1193 1315 1427
20 1185 12133 1332 1395 1522 1635 1209 1320 1435 1540 1135 1245 1350 1455
40 1217 t31)0 1362 1417 1547 1652 1270 1375 1470 1565 1245 1315 1400 1500
60 1265 13:l8 1396 1430 1575 1672 1370 1434 1507 1602 1470 1418 1475 1556
80 1308 13114 1440 1486 1609 1695 1465 1505 1565 1634 2040 1580 1562 1632
100 1363 143 4 1485 1525 1634 1720 1590 1593 1623 1688 2605 1830 1670 1710
120 1437 1495 1531 1567 1673 1745 1736 1692 1695 1746 3055 2110 1815 1825
140 1520 157 I 1601 1625 1715 1777 1907 1790 1773 1804 3433 2420 2010 1960
160 1617 165 0 1674 1685 1761 1817 2085 1915 1845 1872 3787 2715 2230 2105
180 1717 1733 1737 1750 1811 1856 2265 2018 1922 1940 4108 2980 2470 2255
200 1840 1S40 1830 1823 1867 1891 2425 2123 2009 2018 4385 3230 2693 2400
300 2410 2.32 '7 2260 2217 2157 2147 3186 2735 2473 2401 5554 4307 3647 3130
400 2901 278.)
,. 2670 2600 2480 2426 3845 3315 2975 2813 6515 5186 "03 3815
500 3357 320
3075 2980 2815 2701 4460 3854 3432 3211 7313 5937 5165 4456
9 1 2983 5025 4385 3883 3608 8080 6640 5830 5100
600 37771360Ci 3472 3350 3153
1
~-ABLE 108. Viscosity ,10- 7 g cm- 1 sec- 1) of ethane + ethylene mixtures ClvIeshcheryakov and Golubev)
TABLE 109. Viscosity (10 -1 g cm -, sec -1) of ethane + propylene mixtures (Meshcheryakov and Golubev)

I I
~
% prOPYtl,e:ce 30.45 S2.6 7,;"(6

100 150 200 100 150 200 100 150 200

p. ann
------~----~----~----~----~----~----~----~---

20
1119
1135
1254
1265
1385
1395
1103
1124
1239
1275
1369
1400
1089
1](1'1 :~~~l :~~~ .
40 1165 1295 1420 1180 1325 1440 1198 1280 13%
60 1290 1365 1475 1295 1410 1500 1750 1410 1473
80 1615 1490 1555 1900 1555 1595 2870 1730 1592
100 2390 1705 1660 2910 1860 1740 3890 2330 1810
120 3015 2035 1805 3480 2240 1925 4410 2880 2070
140 3455 2355 1960 3950 2620 2125 48HO 333.') 2370
160 3820 2670 2130 4350 3010 I 2335 524.1 3740 2670
180 4135 2950 2315 4695 3375 I 2550 5590 4100 2950
200 4415 3220 2500 5000 3700 I' 2775 5900 4430 3217

:: ~:~~ :~~~ !~;~ ~~~~ :~~~ I !~~~ ~~~~ ~~~~ :::::

500 7190 5980 4940 8135 6755 i 5565 9270 7700 G340
78_9_0--,,-__6_66_4_L.-_5_61_2_,--_8_87_0_-,--_7_5~_1~630~
_ _6_00_ _,__ ~~"O{)~120

TABLE 110. Viscosity (10 -7 g cm -1 sec -1) of 31.65"/0 ethane + 38.40"/0

ethylene + 29.95"/0 propylene mixture (Meshcheryakov and Golubev)

i !I~
50
_-:-_---"
10J i 150 'I ~'
" p. atm ~
I'C'
I' 50 I 100 i 1·50

1 1190 I' 1 3195 1102611830

1050 1 1320 '.1 110
10 1060 1194 1325 i 140 3620 I 2320 1990
20 1066 1203 1331., 160 3965 I 2618 216·3
30 1082 I 1216 1341 I' 180 4269 i 2910 I 2365
40 1123, 1245 1370 i 200 4535 i 3180 , 2545
.30 1200 I 1280 1405 il' 250 5115 i :0772 I' 3000
(iO 1330 I 1338 1450 I, 300 5600 I 4270 . 3415
70 1525! 1412 1500 I 350 6015 ; 4710 i 380.5
80 1805 i 1505 1560: 400 6390 I 5115 4180
100 2586 I 174.)' 1690 II 450 67.35 • 5495 4538

T ABLE Ill. Viscosity (10 -7 g cm -, sec -1) of the natural gas from the
Saratov deposit (Meshcheryakov and Golubev)

~i
atm
p. ,
2') 31 100 13.) :))0 2·30

1 1007 1069 1169 131:; 1450 1590 1720

20 i 1093 1109 1192 1342 1476 1605 1735
40 1201 1170 1235 1380 1508 1626 1750
60 1322 1243 1291 1423 1544 1651 176.5
80 1458 \.327 1360 1-l73 1584 1680 1786
100 1602 1426 I 1438 1526 1625 1711 1809
120 1762 1537 1523 i 1583 1665 17-15 1833
140 1930 1657 1616 1645 1708 1784 1862
160 2105 1783 1710 1707 1756 1824 1892
180 2278 1914 1804 1777 1808 1866 1925
:200 2445 2046 1900 1850 1859 1904 1957
251 2813 2361 2150 2036 2001 2004 2041
300 3137 2650 2402 2221 2140 2108 2125
350 3430 19:25 2645 2411 2282 2215 2213
400 3723 3192 2887 2600 2420 2320 2293
450 3939 3456 3112 2779 2565 2432 2387
The viscosity of the natural gas of the Saratov deposit at pressures
close to 1 anTI is 1 - 2"/0 lower than the viscosity of methane. At higher
pressures it is a few percent higher; the difference increases with decrease
in temperature. The natural gas was not subjected to any additional
scrubbing, and contained the following components: CH., 91.5"/0; C2 HS,
1.8"70; C 3H" 0.8"/0; C.H,o, 0.6"/0; N" 5"/0, and C s or higher paraffins, 0.3"/0.

214
 2300 b=1-----r
r~ _I--~~~ _1--1--....-
~2fOO~ 0
i~ a ~ _d}:!..:--- f--- ....-"'---
~;~~~- ~~- -~~~~
01) l-..- _ _ r-- lIDO ..--- )...--:~
r' 1800 I .....
<> i I---I--I-! 50 i..--~ :..---; !
'-</700 '_ ' CCi /
i---1i5' I I
~/600~: ~ ~
!,

J-r i '
1400
J..-I-I
'--tV
15001-c1-*'""y'+, Ii

r
I
"""",,"""'1---+,--ii-_~:~:~:~~;~~+-!f- +_----'c-_-'_i~~ -_ I

1300 L.I--------L..-l..--'_'----'---l.:~._ ....-,-I-----'.!---,,--":-I__'.,--'---,..,"'

:!

o 50 m 200 300 WO 500 600 700 800

,o,atm

F[GURE 105. Viscosity of nitrogen + hydrogen mixture containing

750;0 H2 + 250;0 N2 as a function of pressure at different temperatures.

2300
@9!l. V ~

----..-- - a;
2200
V rJJO ....-::: 8::: ;;a
"~2fOO
,.~2COO
, I--: ........ ~ ~ ~ ~ ~
51900
,,0lJ1800
o
rl 1700
~600 t.-
-
:.-- I---
I-"""'" I-"
v V V v:::: V
--
~V
~
~ ;::::: ~
v. ~ ~
V; ~ V
~

I..........
1500 ~~
?of r<!
fWO
I~ ~ :atm
1300 I/"'
o 20 40 60 80 If)] 120 !W 160 /80 200 220 240 2m
t, 'C

FIGURE 106. Viscosity of nitrogen + hydrogen mixture

containing 75% H2 + 25% N2 as a function of temperature
at different pressures.

--- IOO}C_
--- - f.--- ~

--
~
,
E
sl£ ~ l.----
~~

--
()

j
b.C
1200
~ !!£. V
o V
- ~
rl

~
I---
1000 ,
50 100 /50 200 250 300
jJ,atm

fIGURE 107. ViSCOSity of 33.50/0 CH4+ 66.5% H2 mixture as a function of

pressure at different temperatures.

215
 2600

~2200~----r---~-----+-----~--~~~~~

E
"~moo~---+----~----+-~~·~---+----~
~
,
o
rl

~roL-
j-
__~__~~
fOO
__150
~__~±~_~~~~~
200 250 JOO
p,anll
FIGURE 108. Viscosity of 8(yJjo CH4 + 2(yJjo N2 mixture as a func-
tion of pressure at different temperatures.

2300.--------r---,---,-----r----r-~

_, n@~--+---~---~----+----~~~

"~2mOr-----_r------r_----~--~~.~~-~~r_----~
,
f,2000r_--+---+-
~
~noo~--t~~~-~~~·-+--4--~
o
rl 1800f-----+--::7""'+---7q---.-f---+----~
~
1700 f--."..-<'=--t--:A"---~f

100 150 200 250 300

p, atm
FIGURE 109. Viscosity of 24.7'10 H2 + 19.8"10 CH4+ 55.5"10 N2
mixture as a function of pressure at different temperatures.

MOO

-;'
"
-;'~" 3000~-+---+----+
E
"
,
o
rl

~
2@0r---+--~~-+--·~---r-~

mw~ __~__~~~~__~~~~~~
o 100 200 300 400 500 600
p, atm
FIGURE 110. Viscosity of 8(yJjo methane+2(yJjo propane
mixture as a function of pressure at different tem peratures.

216
 Figures ll2-- 117 show the dependence of the viscosity of binary gas
mixtures on the composition of the mixture. It can be seen that in all cases
and at all temperatures the points deviate from a straight line, i. e., the
additivity rule does not apply. The curves for the nitrogen + hydrogen
mixtures are convex, but they become straighter at higher pressures. For
the methane + pl:'opane mixtures the curves are concave. This confirms
the complex dependence of the viscosity of gas mixtures on the composition,
pressure, and temperature.

~
~4aw~~--~~~~~r---~~~~--~
<:>
bO

~ 3000po.,o-----+--=:;:
rl

I$-
2WO~----~~~v==~~~~~==~

moot=::::~'/~a~n~n~'~~====~~====~
50 100 150 200 250
t:C
FIGURE 111. Viscosity of 53.60/0 methane +46.40/0
propane mixture as a function of temperature at differ.
ent pressures.

In Table ll:l our data are compared with those of Bicher and Katz. The
results of the American authors were graphically interpolated to our
ten:peratures and compositions. It can be seen that the discrepancies at
some points are higher than 20%. However, the character of these
discrepancies is similar to that for single gases whose viscosity was
measured by Bicher and Katz (and also other authors) by the rolling ball
method.
Figures 118 and 119 show the dependence of the viscosity of hydrogen +
+ nitrogen and methane + propane gas mixtures on pressure at constant
temperature. The change in viscosity is represented as an isothermal
surface (Figure ll8).
Figures 120 and 121 show the dependence of the viscosities of the same
mixtures at constant pressures. The change in viscosity is represented
as isobaric surfaces.
As in the case of single gases, the 'Y/ vs P isotherms for gas mixtures
intersect on an isothermal surface. For this reason in Figures 118 and 119
,)nly one such surface is shown, i. e., at only one temperature. On the other
Jand, the isobaric surfaces do not intersect, therefore in Figures 120 and
121 we represented the data for the whole pressure range studied.
This three-dimensional representation gives a clearer picture of the
dependence of the viscosity of binary gas mixtures on the temperature,
pressure, and Gomposition, than diagrams in Cartesian coordinates.

217
TABLE 112. Comparison of the viscosity data of Bicher and Katz (10- 7 g em-I sec-I) of methane + propane
mixtures with those of Meshcheryakov and Golubev

p, annl r~::::r:ne 1 2~:5

75 1 100 1 1 150 1 ~o:mr::~url' ~5: 1 200 100 1751:0 1 200
MeshCheryakOV; GOlubevl 11951 127611344/1400115301 1230 11340 11445/1170 1126511365
27.2 Bicher amI Katz 1250 1300 1350 1409 1535 1230 1325 1432 1232 1290 1370
/ Discrepancy, % + 4.6 +1.88 +0.41 +0.64 +0.39 ±O.OO -1.12:-0.9 , +53 1+1.07+0.36
1
IMeshcheryakov; GOIUbevl1279 11357/' 1414114621158511405 1462 153011675 1 1475 i 1505
.
I
68.0 Bicher and Katz
Discrepancy, %
II'

1420, 1450 1490 1518 1596 1462 I 1480 ,1520 1782 11610 j 1560
+ 11.09 1 +686 +5.38 +3.83 +0.69 +4.061 +1.11 ;-0.65 +6.4 [+9.2 ;+3.65
I I I I

MeshcheryakOv; G01Ubevi 137411440 11490 115271164011600 11602116351 26551 J 85511685

102.1 Bicher and Katz I 1610 1600 1600 1607 1668 1720 1657 15951 2978 2020 1835
I
Discrepancy, "/0 [+17.171 +11.11 +7.4 -1-5.24 +1.71 +75 +3.441+3.68 +12.2 +8.9 +8.9
1

Meshcheryakov; GOIUbev1186511847!1840 1184011872/' 2455121551202514440 1328512740

204.1 Bicher and Katz . 2220 2092 2020 1984 1957 2727 2350 2257 4880 3455 2805
1
Discrepancy, % I
+19.3 +13.26 +9.78 +7.83 +L53 +11.07 +9.041+11.5 +9.9 +5.17 +2.37
1

340.2 Bicher and Katz

I
Meshcheryakov ; GOIUbevl' 2610 1261512435123701 2:290 346,512965126751 5950 1466513935
2880 2660 2530 2452 2386 3775 3175 2940 6550 5000 4040
Discrepancy, % 1+1034 +5.76 +3.9 +3.46 +4.19! +8.94 +7.08 +9.951+10.08 +7.50 +1.38
1

5000

4000
-;< '00)
()
~
-;< 600 500
53000
be ;/4D.0
, ~'
?J)O '
,....
'"
''----- zoo
""2000 100
latin

FIGURE 112. Viscosity of Hz + N, mixtures at 15°C and

different pressures.

218
 woo

8ilQ.--- ~ f-----:
==
~ f-60JI- I~
v ~
~ ~ ~ i=-::::=:: ~
e-=
30
a EI
4 ~~~
100 latm

v-
~
~

~
~
fWO

500
102030W5060 70 80 !J(j ffJJ
tN,
FlGURE 113. Viscosity of H2 + N2 mixtures at 200°C and different
pressures.

3fJJO
I---
i---"" I---

,
2500
Vn·
~ V- V j...---
/ zQ.Q- ~ V-
~ j...---
I------ ~
-
~ v ---
7 2fJJO 1/~V!j)' ~ J---::: f:::::= V
E
u
bl)
~~V j::/
~
,
~/500
/, ~ ~ t;(.C
\5

~~
V
"""
1000 ~

500

FIGURE 114. Viscosity of H2 + N2 mixtures at

200 ann and different temperatures.

4000
,
~
, 80 0
§Jooo 6111 500
bl)
,
~

0
c~

1$
2000

FlGURE 115. Viscosity of H2+CO mixtures at 100°C and

different pressures.

219
,
~90Q?r----r----r
,
§ 7000~-+---+-
,
55o,'17\--~--+----+-----j
00
~

';=.30001==-::~~~---'S~.cr

80 100
~Oooe:::--~~~~~~~~~~
o 40 20 60 80 100
propane 7.
FIGURE 116. Viscosity of Cil. + C,lls FIGURE 117. Viscosity of CI-I. + C,l-ls
mixtures at 100°C and different mixtures at 250° C and different pressures.
pressures.

el,
,~

OtH ,, , ~J
-; 5000 I
:
~ , 1 '",,-:--
25 .~ A,-
,
~1t
50'jK ,'
-r,

I! : t--I

+rIB ,~ - 7,',t'H', ~- "

" , -" i1_1..,/';"r, t ~-1 -~,*~ ~+'-1ri--,t-
I ' ... I ... I... I : ...

/ '
75
100 200 300 700 800 !flO
400 500 600
p. ann
FIGURE 118. Viscosity of H2+ N2 mixtures in ~vs.p coordinates. compo-
sition at 15°C. a, b, c, d - isothermal surface.

~I
"'-+-i--
, 10000 ___ :
,I ,
()

7')), 800ofl:
B
()

220
 sc;
M

'~3000
<$

o v
15 50 100 150 200 250
t;c
FlGJRE 120. Viscosity of H, + N, mixtures in "l)vs. t coordinates. Composition
at different pressures.

,
"~
.. 10000
S
U
bD
'0 8000
o
rl

""6000

9000

2000 '{('200

o ~;oO-;-:-
fatm :
fOO !50 ;00
t,'C

FIGURE 121. Viscosity of CH4+C3HS mixtures in-,)vs.t coordinates.

Composition at different pressures.

221
 Let us illustrate the use of such three-dimensional diagrams by an
example. Let us determine the viscosity of the 300/0 H2 + 700/0 N2 mixture at
15°C and 600 atm from Figure 118. On the "composition" axis we find the
point corresponding to the composition of the given mixture (point AI). We
project this point first onto the projection of the 600 atm isobar (point A2)
and then onto the isobar itself (point A3). Then we measure with a ruler or
compass the distance A2Ag. This distance is plotted onto the "viscosity"
axis, and the 11 value for the above conditions is found.
In our case 11 = 3100 g cm- l sec-I. The same procedure for the 70% H2 +
+ 300/0 N2 mixture at 15°C and 200 atm gives ll'= 1400 g cm- l sec- 1 (see
points BI, Bz., and Bs).

4. The calculation of the viscosity of gas mixtures at

high pressures

In the literature we could not find any analytical equation for the
dependence of the viscosity of gas mixtures on the temperature, pressure,
and composition, and therefore we attempted to find such an equation. We
checked (156) to see whether it is applicable to gas mixtures or not. The
checking procedure was the same as for single gases.

V
2.8 L
2.6
J'V
2. 4
i" I

22
V-
0 ~
8 /
6
"Js
,~
* V 0--f5·0
[2 l a--50·
x --fOO· r--
V m --200 0
[0
W 1.2 1.4 1.6 1.8 ao a2 OJ, 0.6
llJ-P-
FIGURE 122. Viscosity of 750/0 H2 +25% N2 mixture in

19 d1J -lg 7- coordinates at pressures between 50 and

800 atm.

Figure 122 represents the viscosity of the nitrogen + hydrogen mixture

(3H2 +N2) in 19 ~1)- 19 jt
coordinates, at pressures between 50 and 800 atm

222
and temperatures of 15 to 200°C. The viscosity of the following methane +
propane mixtures is represented in the same coordinate systems:
1) 24.4% CH4 + 75.6% C3Ha at 60 to 600 atm and 100 to 250°C (Figure 123);
2) 53.6% CH4+ 46.4% C3Ha at 60 to 500 atm and 100 to 250°C (Figure 124);
3) 80% CH4 + 20% C3Ha at 80 to 500 atm and 100 to 250°C (Figure 125).

3,5r------.- 4.0.-----.----.-----.--~-,

IfdlJ

J.5~----+-----~--~~----~
J.01----+--

J.0r---_r--~r_--_r--~

2.5
o -IOO'f}
l> -125' Z.5~----~~--~----_+----~
a -150'
<> -200'
• -250' l>-IOO'C
2.0 0-150'
2.0~--~---+-- 0-200'
0-250'

!'5~ ______-:,-::_ 1.5~ ____~~__~~____~____~

o 0.5 1.0. 1.5 o 05 1.0 p 1.5 2.0
19 :, fO l§ -f'/O

FIGURE 123. Viscosity of 24.40/0 CH4 + 75.60/rP/o C3H8 FIGURE 124. Viscosity of 53.60/0 CH4 + 46.40/0 C3H8
mixture in Ig ll.'l-lg ~!- coordinates at pressures mixture in Ig ~'l-Ig ~coordinates at pressures
between 60 and 600 atm. between 60 and 500 atm.

3.0
w
/
2.5
/
/' e -100'C

/
0-/50'
• -200'
0-250'
20
/
1.5
o 0.5 1.0 P. ~5
l!l-r IO

FIGURE 125. Viscosity of 80.0% CH4 + 20.0% C3H8

mixture in Ig "'1- Ig If coordinates at pressures

between 80 and 500 atm.

223
 It can be seen from the figures that the experimental points fit well into a
straight line. This indicates that (156) can be applied to gas mixtures. The
discrepancies between the calculated viscosity coefficients lying on the
straight line and the observed do not exceed 2- 30/0. The constants a and n
for these mixtures were calculated from the most reliable viscosity data
and are reported in Table 113.

TABLE 113. Constants a and n of equation (156) for some mixtures

l
~""oib
58 ~ '<:j

Composition of 111ixture n '§ ~ ~ ~ C) Q) cti cti

.g e ~ :
\ ~ ~ e .~

75% Hz + 25% N2 1.121 167 157

244% CH 4 + 75.6% CaHs II
1.11 1125 )]23
53.6 % CH 4 + 46.4% CsHs I 815 807
80.()0/0 CH 4 + 20.0% CaHa i
i
1.11
1.11 603 619

By processing the experimental data /1 ~)8/ it was found that the constant
a for gas mixtures can be approximately calculated from the equation

(194)

where a1 and a2 are the constants for the first and second components.
Table 113 also gives the values of a calculated from (194). The a values of
the pure gases were taken from Table 74. It follows from Table 113 that the
a values for the gas mixtures calculated from (194) or determined by
experiment are in good agreement.
The thermal pressures for the given gas mixtures are reported in
Tables 114 and 115. They were calculated"by graphical differentiation of
the experimental p, v, T data /159, 160/.

TABLE 114. Thermal pressure (in atm) of 75"/0 H2+ 25"/0 N2 mixture

~t,OCI 15 50 100 200

p,at~

50 51.6 5\,2 5\.0 50.8 -

100 106 105 103.5 102.3 101.4
200 218 217 214 210 206.6
400 450 446 437 428.5 420.2
600 686.5 678 662 648 628
800 910 900 879 860 I 834
I I

The thermal pressure of gas mixtures can be calculated from the equation
of state. By analytical combination of the constants of the equation of state,
an equation of state for gas mixtures can be found, and thus the thermal
pressure determined. For the combination of the constants of the Van der

224
Waals equation for gas mixtures from the constants of the individual
components, the following equations can be used:

Correspondingly, for the Beattie- Bridgman equation, we obtain

Ao,m=Nl~+N2~'
am=Nlal+N2a2'
B o, m = N 1Bo, 1 + N 2Bo, 2'

bm=Nlbl+N2b2'
Cm=NIC1+N2C2'

where N, and N2 are the mole fractions:

N - nl .
1- 1l1+ n2 '

r
TABLE 115. Thermal pressure (in ann) of methane + propane mixture 0/0158,
80.0
% Met:a,:e 1 100 I 150 241.4 200 I 250 1100 1
150 200 1 250 1100 1 125
p. atm
----'7--
60 160 106 90 78 100 83 77 73 - - - - -
80 348 176 140 122 152 123 110 103 108 101 98 95 83
100 520 260 194 164 214 166 147 136 151 139 130 124 119

120 654 360 254 212 286 216 187 171 196 179 167 157 148

140 762 458 324 262 366 274 230 208 244 223 206 190 178
160 858 546 396 320 450 334 276 247 293 269 247 227 210
i
180 940 632 466 374 535 394 325 288 347 317 291 262 242
200 1018 714 530 432 615 1. 457 373 330 401 367 333 300 275
I
i
300 1342 1064 840 I 705 9331 751 I 629 550 664 607 552 494 444
I ,
400 1598 1332 1
1116 960 1188 I, 1005 871 ': 769 890 826 770 695 630
,
i 818
500 1818 1562 1196 - 1236 1037 895
1
1360 I
I 1:0 I 975
I 1104 976
600 1768 1496 1414 I - - I - - - - - -

225
APPENDIX

T ABLE I. Conversion of pressure

Pounds! I
r-
Bar mm Hg
square inchl

1 atln 1.033228 14.6959 1.013250 760

1 kg/ cm z 0.967841 1 14.2233 0.980665 735.559
1 pounds
0--.- 0.68046 0.70307 10 0.689476 517.149
sq.m. !
1 bar ! 0.986923 1.019716 14.5038 750.062
1 nUll Hg ~ 1.31579 1.35951 19.3368 1.333221 1000

TABLE [[

Temperature scales
oK _ Kelvin scale (absolute temperature); °c - Centigrade scale;
of _ Fahrenheit scale; QOC = 273.l6"K = 32°F.

Conversion formula of temperature

Y"K = 273.16 + XOC; yoC = X"K - 273.16 = S/9 ( XOF - 32); YOF =

= ~5 XOC+32
•

Conversion of degrees Fahrenheit (oF) to degrees Centigrade (oC).

Temperatures below QOF

10 20 30 40

0 -' 17.78 23.33 28.89 34.44 40.00

-100 -73.33 78.89 84.44 90.00 95.56
-200 --128.9 134.4 140.0 145.6 151.1
-300 -184.4 190.0 195.6 201.1 206.7
-400 -240.0 245.6 251.1 255.7 262.2
-
50 60 70 80 .90

0 45.56 51.11 56,67 62.22 67.78

-100 101.1 106.7 112,2 117.8 123.3
-200 156.7 162.2 167,8 173.3 178.9
-300 212.2 217.8 223,3 228.9 234.4
-400 267.8 - - - -

226
 T ABLE II (Continued)
Temperatures above O°F

Of 10 20 30 40

0 -17.78 -12.22 -6.67 -1.11 4.44

100 37.78 43.33 48.89 54.44 60.00
200 93.33 98.89 104.44 110.00 115.56
300 148.89 154.44 160.00 165.56 171.11
400 204.44 210.00 215.56 221.11 226.67
500 260.00 265.56 271.11 276.67 282.22
600 315.56 321.11 326.67 332.22 337.78
700 371.11 376.67 382.22 387.78 393.33
800 426.67 432.22 437.78 443.33 448.89
900 482.22 487.78 493.33 498.89 I 504.44
000 537.78 543.33 548.89 554.44 I 560.00
100 593.33 598.89 604.44 610.00 615.56
200 648.89 654.44 660.00 665.56 671.11
I
Of 50 60 70 80 90

0 10.0 15.56 21.11 26.67 32.22

100 65.56 71.11 76.67 82.22 87.78
200 121.11 126.67 132.22 137.78 143.33
182.22 187.78 1 193.33 198.80
300 176.67
400 232.22 237.78 243.33 "
248.89 254.44
500 287.78 293.33 298.89 304.44 310.00
600 343.33 348.89 354.44 360.0 365.56
,'00 398.89 404.44 410.00 I 415.56 421.11
800 454.44 460.00 465.56 471.11 476.67
900 510.00 515.56 521.11 I 526.67 532.22
000 565.56 571.11 576.67 582.22 587.78
100 621.11 626.67 632.22 637.78 643.33
200 676.67 682.22 687.78 I 693.33
I
698.89
I

For interpolation

3 4 5 6
'f 1 1 1
1 1 1 1 1 81

°C / 0.56/ 1.11 11.671 2.221 2. 78 13.341 3.89 I 4.441 5.00

T ABLE III. Conversion factors for pounds per square inch to kg/cm'

Pounds
sq. in.

0 0.000 0.0 70 0.141 0.211 0.281 0.352 0.422 0.492 0.563 0.633

10 0.703 0.7 73 0.844 0.914 0.984 1.055 1.125 1.195 1.266 1.336

20 1.406 1.477 1.547 1.617 1687 I 1.758 1.828 1.898 1.969 2.039

30 2.109 2.1 80 2.250 2.320 2.391 2.461 2.531 2.602 2.672 2.742

40 2.812 2.8 83 2.953 3.023 3.094

I 3.164 3.234 3.·385 3.375 3.445

50 3.515 3.586 3.656 3.726 3.797 3.867 3.937 4.008 4.078 4.148

60 4.216 4.289 4.359 4.430 4.500 4.570 4640 4.711 4.781 4.851

70 4.922 4.992 I 5062 5.133 5203 5.273 5.344 5.414 5.484 5.555

80 5.625 5.6 ~;J 5.765 5.836 5.906

1
5.976 6.047 6.117 6.187 6.258

227
T ABLE III (Continued)

Pounds 7 8 9
0 2 3 4
sq. in.
90 ! 6.328 6398 6.469 6.539 6.6112 6.679 6.750 6.820 6890 6.961

100 7.031 7.101 7.172 i 7.242 ~.312 7.383 7.453 7.523 7.59:l 7.664
,
----
For pounds per square inch values between 100 and 1000, the decimal point must be shifted one place
to the right; from 1000 to 10,000, 2 places, etc.

T ABLE IV. Vapor pressure of mercury

t, "C I p,mm Hglll,"C I p,mm Hgil I."C Ip,mmHg!1 I, "C 1 p,mmHg

I
1
0 \ 0.0 3185 i\ 105 210 315 339.60
I 0.354411
!
23.723 :1

5 0.033041 110 0.4572 Ii 215 27.654 :1 320 I 376.23

1
10 I 0.0,490 115 0.5857 '1 220 32.133 !I 325 416.31

IS I
'i
i
!
0.03773 120 0.7457 il 225 37.'222 330 459.74

20 0.00120111 125 0.9434 il 230 42.989 335 506.85

25 0.001840 i 130 1.186 235 49,509 3(10 55790

"1
30 I 0.0027771 135 1.484 !
11
240 56.855 345 613.08
,
35 0.004135 140 1.845 245 65.113 350 672.69
I

40 145 2.282 250 74.375 355 736.98

0.006079[1
45 0.008827, 150 2.807 255 84.732 360 806.23

50 0.01267 il 155 3.438 260 96.296 il 365 880.68

I'
55 0.01797 ''1I 160 4.189 265 109.17 370 960.66
:i
60 0.02524 165 5.080 270 123.47 375 1046.5

65 0.03507 170 6.128 275 139.34 380 1138.4

\i
70 :i 385 1236.8
0.04825 175 7.356 280 156.87
II
75 0.06580 180 8.796 285 176.24 390 1341.9
III
80 0.08880 .1 10.472 290 197.57 395 1454.3
185

85 0.1191 190 12.423 295 221.01 400 1574.1

II
I.
90 i 0.1582 195 14.681 300 246.80

95 , 0.2086 II'I 200 17287 305 275.02

100 !I 02729
.
:! 205 20.28:; 310 305.89

228
TABLE V. p, 'V, T data for mercury on the saturation curve at high
tem pera lUres

Temperature of Specific volume, em'

Pressure,
2 saturated
kg/cm liquid vapor
vapors,oC

2.0 395.8 0.0791 141.4

2.2 401.7 0.0792 129.6
2,4 407.4 0.0793 119.8
2.6 412.5 0.0794 111.4
2.8 417.5 0.0794 104.3
3.0 422.4 0.0795 97.98
3.5 432.8 0.0797 85.24
4.0 442.4 0.0798 75.58
4.5 451.0 0.0799 68.01
50 458.9 0.0801 61.87
5.5 466.3 0.0802 56.82
6.0 472.8 0.0803 52.54
6.5 479.1 0.0804 48.91
7.0 485.1 0.0805 45.78
8.0 496.3 0.0806 40.65
9.0 506.3 0.0808 36.60
10 515.5 0.0809 33.33
12 532.3 0.0812 28.37
14 546.7 0.0814 24.75
16 559.8 0.0816 22.00
18 571.4 0.0818 19.83
20 582,4 0.0819 18.08
25 606.5 0.0823 14.87
30 627.1 0.0827 12.68
35 645.5 0.0830 11.09
40 661.8 0.0832 9.88
45 677.0 0.0835 8.92
50 690.9 0.0837 8.15

TABLE VI. Specific gravity and specific volume of mercury as a function

of tem pera ture

T~'mpe- Spccific Specific Tempe Specific Specific

ra.UJ.re, gravity, volume, rature gravity, volume,
°C g/cm' cm' °C g/cm' cm'

-10 13.62017 0.0734205 31 13.51927 0.0739686

--9 61177 4338 32 51672 9820
--8 61522 4472 33 51437 9953
-7 61275
II
4606 II 34 51193 0.0740087
-6 61027 4739 35 13.50949 0.0740221
-5 13.60780 0.0734873 36 50704 0354
-4
-3 I 60533
60286
5006
5140
37
38
50460
50216
0488
0622
-2
-1
! 60039
59792
5273
5407
39
40
49972
13.49728
0756
0.0740891
0 1359546 0.0735540 50 47290 2229
1 Ii 59289 5674 60 44859 3569
2 59051 5808 70 42433 4910
3 I 58804 5941 :j 80 40012 6252
4 58558 6075 90 13.37595 0.0747594
5 I 1358311 I 0.0736209 I! 100 35180 8939

229
T ABLE VI (ContillUed)

Tempe- Specific Specific empe- Specific Specific

rature, gravity, VOlUllle, rature, gravity, Volurne,
3
°c g/cm' em' °c g/cm' em
6 58065 6342 JlO 32830 0.0750285
7 57818 6476 i 120 30440 1633
8 57,572 6610 130 280,50 2982
9 57326 6744 140 13.25670 0.0754334
10 13.57079 0.0736877 150 23300 5688
11 56833 7011 160 20930 7044
12 56587 7145 170 18560 8402
13 56341 7278 180 16240 9764
14 56095 7412 190 13840 0.0761128
15 l3.,55849 0.0737546 200 13.11480 0.0762495
16 55604 7680 210 09130 3865
17 55358 7813 I' 220 06780 5239
! 55113
18 7947 I, 230 04430 6616
19 54867 8081 ;1 240 02090 7996
20 1374622 0.0738215 I' 2-50 12.99750 0.0769381
21 54376 8348 'I,I 260 97410 0.0770769
22 54131 8482 [, 270 95070 2161
23 53886 8616 :1 280 92730 3558
24
25
I 53641
13.53396
8750
! 0.0738883
,i
290
300
90390
12.88060
4958
0.0776364
26 53151 9017 Ii 310 85720 7774
27 52906
I 9151 Ii 320 83390 9189
28 52661 9285 330 81050 0.0780609
29 52416 9419 !I 340 78720 2033
30 I 13.52171 ,2 Ii
0.07 3955 3 50 ! 12.76380 I 3464
360 74050 [ 4900

1
T ABLE VII. CompressibIlity ( v
C,V)
--;;p 6
'10 of mercury

~I 30 100 150 ~_~ __

300_
------~----7_--~----~-

1-50 3.873 4.259 4.518 4.879 5.551 6.970

1-100 3.859 4.231 4.480 4.835 5.461 6.413
1-150 3.850 4.208 4.447 4.798 5.375 6.004
1-200 '3.843 4.188 4.417 4.762 5.292 5.757
1-250 3.834 4.165 4.384 4.724 5.206 5.608
1-300 3.823 4.140 4.348 +.683 5.117 5.520
1-350 3.809 4.112 4.311 4.641 5.026 5.422

T ABLE VIII. Viscosity of mercury

I, (Ie ~, poise t, ·C 7j, poise

-20 0.01855 170 0.01096

-10 0.01764 180 0.01080
0 0.01685 190 0.01066
10 0.01615 200 0.01052
20 0.01554 210 0.01039
30 0.01499 220 0.01027
40 0.01450 230 0.01016
50 0.01407 240 0.01005
60 0.01367 250 0.00995
70 0.01331 260 0.00985
80 0.01298 270 0.00975
90 0.01268 280 0.00967
100 0.01240 290 0.00958
110 0.01214 300 0.00950

230
 T ABLE VIII (Continued)

t, °C 11, poise II t, °C 7), poise

120 0,01191 310 0.00942
130 0,01169 II 320 0.00935
140 0,01149 330 0.00928
150 0,01130 340 0.00921
160 0.01112 II

T ABLE IX. Some properties of mercury under high pressure

Viscosity, poise
Pressure, Melting point,
kg/em 2 cc 30°C 75°C

1 -38.2 0.01516 0.01340

4000 -18.4 0.01663 0.01463

8000 + 1.8 0.01822 0.01598

12000 + 21.9 0.02007 0.01759

TABLE X. Coefficient of linear expansion of different glasses and platinum

It = 10 (I +
cct)

Substance

Soda glass •• 85.0

Lead glass •• 95.0
Pyrex glass • 32.0
lena glass 16~" • 78.0
lena glass 59'" • 57.0
fused quartz • 5.4
PlatinUlu ... 90.0
Platinum - iridium alloy, 90"/0 Pt+ 10"/0 Ir 87.0

t1v
T ABLE XI. Compressibility of glasses -
Vo
= - (ap + bp2) (p in kg/cm2)
7 b'10"
a'10
Ty pe of glass
30°C 70°C 30°C 70°C

Sodium -calc:iu m with boron

trioxide • ..
· . .... 23.29 23.92 -6.1 -10.1
Sodium-potass ium 24.69 25.44 -22.0 -21.2
Pyrex · . . .. . . .
Lead glass with high lead content.
30.12 29.72 + 6.1 +6.7
30.54 30.60 -24.3 -20.6
Lead -sodiun:. borosilicate with
fluorides. · . . ..
Zinc -sodiUlT. borosilica te ...
27.78
25.97
27.71
26.30
+2.5
+4.0
+3.8
+0.2

231
Bibliography

1. Newton,I.-Principia. Lib.II. Sect. IX. 1689.

2. Pet r 0 v , N. P. Trenie v mashinakh i vliyanie na nego smazyvayushchei
zhidkosti (Friction in Engines, and the Effect of the Lubricating
Liquid on It). - Inzhenernyi Zhurnal. 1883.
Vol a r 0 vic h , M. P. N. P. Petrov -osnovopolozhnik gidrodinamicheskoi
teorii smazki mashin (N. P. Petrov - Founder of the Hydrodynamic
Theory of Engine Lubrication).- Trudy 2-i Vsesoyuznoi Konferentsii
po Treniyu i Iznosu, 3, 5. 1949.
3a. Tim i ry a z e v ,A. K. Kineticheskaya teoriya materii (The Kinetic
Theory of Matter). GTTI. 1933.
b.H e r z f e ld, K. F. The Kinetic Theory of Matter. [Russian translation
of the author's articles. 1935.1
Maxwell,J.Cl.-Phil.Mag., 19(4), 28.1860.
J e a n s , J. H. The Dynamical Theory of Gases. Cam bridge. 1925.
c .. K r y 1 0 v , K. I. Fizicheskie osnovy elektrovakuumnoi tekhniki (The
Physical Basis of Electrovacuum Techniques). - Gosenergoizdat.
1949.
d.Stuart,H.A. Molekiilstruktur.-Berlin. Springer Verlag. 1934.
e.Osnovateli kineticheskoi teorii materii (The Founders of the Kinetic
Theory of Matter), edited by A. K. Timiryazev. - ONTI. 1937.
4. Ens k 0 g, D. -Klngl. Svenska Vetenkapsakademens Handlingar 68,
No.4. 1922.
5. C hap man, S. and Cow 1 i n g . The Mathematical Theory of Nonuniform
Gases. - Cambridge. 1939.
6. Knudsen,M.-Ann.d.Phys. 28,75.1909; 31,205,633.1910; 32,
809. 1910; 33, 1435. 1910; 34; 823. 1911; 35, 389. 1911.
7. Kundt,A. andE.Warburg.-Pogg.Ann. 155,337,525.1875.
Warburg,E.-Pogg. Ann. 59, 399.1876.
8. Breitenbach,P.-Ann.Phys. 5, 166. 1901.
9. Meyer,O.E.-Pogg. Ann. 127, 253, 353.1866.
10. Fischer,J.-Phys. Rev. 29, 147. 190H; 30,269.1910.
11. Trautz,M. and W.Weizel.-Ann. Phys. 78 (4).305.1925;
78 (4), 365. 1925.
12. Trautz,M. andA.Narath.-Ann.Phys. 79,637.1926.
13. Hagen, G.-Pogg. Ann. Phys. Chern. 16 (2), 423. 1839.
14. Hagenbach,Ed.-Pogg. Ann. 109 (4),385.1860.
15. Gruneisein.- Wiss. Abhandl. Phys. Techn. Reichsanstalt, 155.1905.
16. Erk,S.-Forsch. Arb. a.d. Geb. lng. Wes. No.288, 9. 1927.
17. Boussinesk,J.-Compt. rend. 110, 1160, 1238, 1292.1890; 113,
9, 49. 1891.

232
18. Knibbs--J. Roy. Soc. New South Wales. 29, 77. 1895; 30, 190.1895.
19. Erk,S.--Z. techno Phys. 10, 452. 1929; Handbuchd.Experimental-
physik, IV, Teil 4. 1932.
20. Herschel.-Bureau of Stand. Technol., Pap. No. 125. 1919.
21. Schulze ,H.-Ann. Phys. 5 (4), 140. 1901.
22. Sigwart,K.-Forsch.Arb. a.d.Geb. 7, H. 33,125.1936.
23. Schm idt, E.- Z. VDI 80, No.21. 1936.
24. Rankine, A.O.- Proc. Roy. Soc. A83, 265. 1910; Journ. Scient.
Instr. 1, 105. 1924.
25. Ku en en, J. P. and S. W. Vi sse r. - Comm. Phys. Lab. Leiden,
No. 138. 1913.
26. Rankine,A.O. andC.J.Smith.-Phil. Mag. 42 (6), '601. 1921.
27. Com ings, E. W., B. J. Mayland, and R. S. E gly. The Viscosity of
Gases at High Pressures.-Univ. Illinois Eng. Expt. Stand. Bull.
No. 35'~. 1944.
28. Phillips,P.-Proc. Roy. Soc. A87, 48. 1912.
29. Nasini,A. andG.Pastonesi.-Gazz.Chim.Ital, 63,821. 1933.
30. Shugaev,V.S. andS.I.Sorokin.-ZhTF, 9, No.10. 1939.
31. Timrot,D.L.-Izv. VTI, No.3, 16.1940.
32. Michels,A. andR.Gibson.~Proc.Roy.Soc. A134, 288. 1931.
33. Rankine,A.O.-Proc. Roy. Soc. A88, 575. 1913.
34. Fisher,J.-Phys.Rev. 32,433.1911.
35. Khalilov,Kh.M.-ZhTF, 8, Nos. 13-14, 1249. 1939; 9, No.8. 719.
1939; ZhETF, 9, No.3, 335. 1939.
36. Khalilov,Kh.M.-Izv.ANAzSSR, No.6, 3. 1952.
37. Golubev,I.F. andV.A.Petrov.-TrudyGIAP, No.2, 5.1953.
38. Lawacz€Gk,F.-Z. VDI. 63.677.1919.
39. Stackelbeck,H.-Z. Ges. KlUte-Ind. 40, H. 3. 1933.
40. B rid g mall, P. W. The Physics of High Pressures. - London.
G. Bell and Co. 1949.
41. Sage,B.B. andW.N.Lacey.-Ind.Eng.Chem.lnd.Ed. 30, No.7,
829. 1938; 31, No.2, 223. 1939; Trans. Am. Inst. Mining. Met. Eng.
127, 118. 1938.
42. Flowers .-Proc. Amer. Soc. Test. Mat. 14, 11, 565. 1914.
43. Her s e y, J. - Washington, Acad. Sci. 6, 525. 1916.
44. Smith,A.S. and G.Brown.-Ind. Eng. Chern. Ind. Ed. 35, No.6,
705. 1B43.
45. Bicher,L.B. andD.L.Katz.-Ind.Eng.Chem.Ind.Ed. 35, No.7,
754. 1B43.
46. Vogel,H.-Ann. Phys. 43, 1235. 1914.
47. Grossman,L.-Wied. Ann. 16,619.1882; 18,119,704.1883;
19, 54~L 1883.
48. Schumann.O.-Wied. Ann. 23, 353. 1884; 27, 91. 1886.
49. Van Itterbeek,A. and W.H.Keesom.-Physica 2,97.1935;
5, 257. 1938.
50. Keesom,W.H. andG.E.Macwood.-Physica5, 745, 757. 1938.
51. Van Itterbeek,A. andO.Van Paemel.-Physica 7,273.1940;
8, 133. 1941.
52. Johnston,H.L. andK.E.McKloskey.-J. Phys. Chern. 44, No.9.
1038. 1940.

233
53. Mason,J.G. andO.Maass.-Canad.J.Res. 18, 125.1940.
54. Andronikashvili,E.L.-ZhETF, 18, 5, 429. 1948.
55. Macwood,G.E.-Physica 5,374,763.1938.
56. Gonikberg,M.G. andL.F.Vereshchagin.-DAN, 55, No.9,
813. 1947.
57. S hat ens h t e in, A. I., E. A. I z r ail e vic h, and N .1. Lad y z h n i k 0 va.-
ZhFKh, 23, No.4, 497. 1949.
58. Rudenko,N.S.-ZhETF, 9, No.9, 1078. 1939.
59. G e r f , S. F. and G. I. Gal k 0 v . - Zh TF, 9, 9, 801. 1941.
60. VerschaffeIt,G.E. andCh.Nicaise.-Heid. Comm. No. 149b.
61. Verkin,B.L and N.S.Rudenko.-ZhETF, 20, 523. 1950.
62. Zhdanova,N.F.-ZhETF, 31, No. 1(7), 14.1956.
63. Golubev,I.F. andV.A.Petrov.-TrudyGIAP, No.1, 7. 1953.
64. Fro s t, A. V. Fiziko-khimicheskie svoistva individuaI'nykh
uglevodorodov (The Physicochemical Properties of Individual
Hydrocarbons), No.2. - GNTI neftyanoi i gorno-toplivnoi literatury.
1947; Landolt-Bornstein, Physikaliseh-·chemische Tabellen,
Vol.L 1927; Vol. II. 1931; Vol. III. 1935.
65. Brawne,H. andR.Linke.-Z. Phys. Chern. 17,148,195.1930.
B r awne, H. R. Bach, and W. Wen tz e 1. - Z. Phys. Chern. A137, 176,
447. 1928.
66. Weber,S.-Comm. Leiden. Suppl., No.71b.
67. Van Itterbeek,A. andH.W.Keesom.-Comm.Leiden, No. 252a;
Physica 5, 257. 1938.
68. Van Itterbeek,A. andO.Van Paemel.-Physica 7,273.1940.
69. Weber,S.-Physica 6, 551. 1939.
70. Boyd, J.-Phys. Rev. 35, 1284. 1930.
71. Ruckes .-Ann. Phys. 25 (4),983. 1908.
72. Wildhagen.- Z. Agnew. Math. Mech. a, 181. 1923.
73. Golubev,LF.-ZhTF, 8, 1932. 1938.
74. Kuss,F.-Z. Agnew. Phys. 4, H.6. 1952.
75. Sage,B., W.Vale, and W.Lacey.-Ind. Eng. Chern. 51,223.1939.
76. Faust,O.-Z.Phys.Chem. 86, 479. 1914.
77. Hat s c h e k, E. The Viscosity of Liquids. - London. G. Bell and Sons.
1928.
78. Schroer andBecker.-Z.Phys.Chem. Al73, 178.1935.
79. M ich e 1 s, A., A. Sch ipp e r, and W. Rintoul.- Physica 19, No. 10,
1011-1028. 1953.
80. Ke st in, J. and K. P i Ia r c zyk. - Transactions of the ASME 76, No.6,
967, 987. 1954.
81. Speyerer,H.-Z. VDI 273.1925.
82. Shugaev,V.S.-ZhETF, 3, No.3, 247. 1933.
83. Teplofizicheskie svoistva veshchestv (spTavochnik) (Thermophysical
Properties of Substances. Reference Book), edited by N. B. Vargaftik.
- Gosenergoizdat. 1956.
Khlopkina,A.V.-Dissertation. MEL 1954.
84. Haw kin s, Sol be r g, and Pot t e r . - Trans. Am. Soc. Mech. Engrs.
57, 395. 1935; 58, 258. 1936.
34a.Ha wkin s, So Ib erg, and Pott e r. - Trans. Am. Soc. Mech. Engrs.
62, 677. 1940.

234
 85. Timrot,D.L.-Teploenergetika, No.1, 27. 1955; The article in:
II Par vysokogo davleniya v energetike." Gosenergoizdat. 1950.
86. Bonilla, C. F., R. D. Brooks, and P. L. Walker. - Proceedings
of the General Discussion of Heat Transfer, The IME and the ASME,
Section 11, Sept., 167. 1951.
87. Vu k a 1 0 vic h , M. P. Termodinamicheskie svoistva vody i vodyanogo
para (TIle Thermodynamic Properties of Water and Steam). -
Mashgiz. 1951.
88. Slavyanskii,V.T.-ZhFKh, 27, 12, 1776. 1953.
89. ShimanElkii,Yu.I. andS.D.Ravikovich.-ZhFKh, 29, No.1, 48,
1955.
90. Rudenko,N.S.-Sov.Phys. 6,470.1934; 8,179.1935.
91. Gerf,S.F. andG.I.Galkov.-ZhTF, 10, 725. 1940; 11,613.1941.
92. Kapitsa,P.L.-DAN SSSR, No.1, 28. 1937; see also Conference on
the Viscosity of Liquids and Colloidal Solutions, Vol. 2. Izd. AN
SSSR. 1944.
93. Kendall,J. andR.P.Monroe.-J. Am. Chern. Soc. 39,1787-1802.
1917.
94. Bresler,S.E. andA.Landerman.-ZhETF, 10,2,250.1940.
95. Keesoro,W.H. andG.E.MacWood.-Physica 8,745.1938.
96. Monoszon,A.M. and V.A.Pleskov.-ZhFKh, 3, 242. 1932.
97. Land 0 1t - B 6 r n s t e in. Physikalisch-Chemische Tabellen, Bd,
1. 1927.
98. Far k as, A. Orthohydrogen, Parahydrogen and Heavy Hydrogen.-
Cambridge University Press. 1935.
99. Jenkins ,A. C. and F. S.DiPaolo.- J. Chern. Phys. 25, No.2, 296.
1956.
100. Henning,F. Warmetechnische Richtwerte.-Berlin. 1938.
100a.Fiziko-khimicheskie svoistva individual'nykh uglevodorodov (Physico-
chemical Properties of Individual Hydrocarbons), Vol. 6, edited by
M. D. Tilicheev. - GNTI neftyanoi i gorno-toplivnoi literatury. 1957.
101. Sutherla.nd,B.-Phil. Mag. 36, 507. 1893.
102. Rankine.A.-Phil. Mag. 21, 45.1911.
103. Voge1,l~.-Ann. Phys. 43, 1235.1914.
104. B 0 1 t zm ann, L. Vorlesungen uber Gastheorie. - Leipzig. Barth.
1896.
105. Maxwel.1,J.-Phil. Mag. 35 (4),129.1868; Phil. trans. 170,
231. 1B80.
106. Enskog.D.-Phys. Z. 12, 56, 533. 1911.
107. Chapman.S.-Phil. trans. 211,433.1912; 216,279.1916; 217,
115. l!H7.
108. Hirschfelder,J.O., R.B.Bird, andE.L.Spotz.-J. Chern.
Phys. 16, 968. 1948; Ind. Eng. Chern. 43, 1641. 1951; Chern. Rev.
44, 20[;. 1949.
109. Hirs chfe 1 der, J. 0., C. F. Curtis, and R. B. B ir d. Molecular
Theory of Gases and Liquids. - New York. Wiley. 1954.
110. Trautz,M.-Ann. Phys. 11, 190. 193!.
1l0a. Li c h t and St ec hert. - J. Phys. Chern. 48, No.1, 23. 1944.
111. K a rap e t I ya n t s, M. Kh. Khimicheskaya termodinamika (Chemical
Thermodynamics). - Goskhimizdat. 1949.
Meissner, H. and Redding. - Ind. Eng. Chern., 34, 121. 1942.

235
112. Meissner,H.- Chem. Eng. Progr. 45, 149. 1949.
113. Riedel, L.-Chem. Ingenieur-Technik ~:7, H.8/9, 475. 1950.
114. Predvoditelev,A.S.-ZhTF, 3, No.3, 280. 1933.
115. Shirokov,M.F.-ZhFKh, 3, Nos. 2-3,175.1932; ZhETF, 3,
No.3, 237. 1933.
116. Soveshchanie po vyazkosti zhidkostei i kolloidnykh rastvorov (Confer-
ence on the Viscosity of Liquids and Colloid Solutions), Vol. 1.-
Izd. AN SSSR. 1941.
117. Dubief,M.J.-J. Phys. Radium 7,402. 1926.
118. Borovik,E.-ZhETF, 18, No.1, 48. 1948.
119. Gonikbe rg, M.G.- ZhFKh, 21, No.7, 811. 1947.
120. Vargaftik,N.B.-Izv. VTI, No.1, 13.1952.
121. Stolyarov,E.A.-ZhFKh, No.7. 1950.
122. Go lu bev, 1. F. - Dissertation. MEL 1946.
123. Bachinskii,A.L-Z.Phys.Chem., 84,6. 1913.
124. Vukalovich,M.P. andG.I.Novikov. Uravnenie sostoyaniya
real'nogo gaza (The Equation of State of the Real Gas). - Gosener-
goizdat. 1948.
125a.Frenkel', Ya.I. Kineticheskaya teoriya zhidkosti (The Kinetic
Theory of Liquids).- Izd. SSSR. 1946.
b.P a n c hen k 0 v , G. M. Teoriya vyazkosti zhidkostei (The Theory of
the. Viscosity of Liquids). Gostoptekhizdat. 1947.
126. Beattie, J.A. and O. C.Bridgman .--Proc. Am. Acad. Arts. Sci.
53, 229. 1928.
127. Deming, W.E. and L.E.Shupe .-Phys. Rev. 37, 638. 1931; 38,
2245, 1931; 40, 848. 1932.
128. Levchenko, G. T. - ZhFKh, 18, No. 10, 453. 1944.
129. Kazarnovskii,Ya.S. andM.Kh.Karapet'yants.-ZhFKh, 17,
172. 1943.
130. Ch'u,P.L. andS.S.Ch'in.-ZhFKh, 25, No.1, 110.1951.
131. Uyehara,O.A. andK.M.Watson.-Nat. Petroleum News,36
764. 1944.
132. Kamerlingh Onnes. -Arch.NeerL, 30,134. 1897.
133. Trautz, M. and J.Husseini,-Ann. Phys. 20, 12l. 1934.
134. Trautz,M. andF.Kurz.-Ann.Phys. 9, 981. 1931.
135. T r aut z , M. and K. So r g . - Ann. Phys. 10, 81. 1931.
136. Trautz,M. andF.Stauz.-Ann.Phys. 2,737.1929.
137. Trautz,M. and H.Benkele .-Ann. Phys. 5, 561. 1930.
138. Trautz,M. and K.Kipphan.-Ann. Phys. 2, 743. 1929.
139. Trautz,M. andA.Melster.-Ann. Phys. 7, 409.1930.
140. Adsumi,H.-Bull. Chem. Soc. Japan, 12, No.5, 199. 1937.
141. Jung,C.. andH.Schmick.-Z. Phys. Chem. (B), 7, 130, 147.1930.
142. Bud d e nb erg, J. W. - Thesis University of California, 1948.
143. Herning,F. and L.Zipperer.-Gas.u.Wasserfach.Ing. 79,45,
69. 1936.
144. Van Itterbeek,A., O.Van Paemel, and J.Van Lierde.-
Physica 13, No.1-3, 88. 1947.
145. Silgardo,R.B. andJ.A.Storroy.-J. Soc. Chem. Ind. 69, No.8,
261. 1950.
146. Trautz,M. andK.Kipphan.-Ann. Phys. 2,746.1929.

236
147. Schmid,F.-Gas u.Wasserfach. lng. 85, 92. 1942.
148. Kenney ,M.J., R.J.Sarjant, and M. W. Thring.-J. Appl. Phys.
7, No.9, 324. 1956.
149. Sutherland,B.-Phil. Mag. 40, 421. 1895.
150. Th ie sen, M. - Verhandl. deut. physik. Ges. 4, 348. 1902.
151. Schroer,E. - Z.Phys.Chem. B.34. H.3/4. 161. 1936.
152. Mann, V. - Gas u. Wasserfach. lng. 73, 570. 1930.
153. Schudel, W.- Schweiz. Ver. Gas.u. Wasserfach. Monaths. Bull.
22, 11:2, 131. 1942.
154. Buddenberg,J.W. andC.R.Wielke.-lnd.Eng. Chern. 41,
1345. 1949.
155. Wielke,C.R.-J. Chern. Phys. 18, 517. 1950.
156. Ke nna r d, E. H. Kinetic Theory of Gases. - New York. McGraw Hill.
1938.
157. Bromley,L.A. andC.R.Wielke.-lnd.Eng. Chern. 43, No.7,
1641. 1951.
158. Spravochnaya knizhka azotchika (Reference Book for the Worker in
the Nitrogen Compound Industry).- Goskhimizdat. 1944.
Go 1 u b e v, 1. F. and M. F. Sin y a v ski i. Spravochnik azotchika
(Reference Book for the Worker in the Nitrogen Compound Industry).-
MKhP GIAP. 1947.
Me shch e ryakov, N. V. and I. F. Go 1 ub e v.- Trudy GIAP, Nos. 3-4.
1954.
G 0 1 u b e v , I. F. and N. V. M e s h c her yak 0 v .- Trudy G IAP, No.6.
1956.
159. Bartlett,E.P., H.L.Cupples, and Tremearne .-J. Am. Chern.
Soc. 50, 1274. 1928.
Wiebe,R. andV.L.Gaddy.-J. Am. Chern. Soc. 60,2300.1938.
160. Reamer,H., B.Sage, and W.Lacey.-Ind. Eng. Chern. 42, 534.
1950; W.Lacey.-Petr. World. 31, No.3, 41. 1934.
161. Ross,J. andG.Brown.-Ind.Eng.Chem. 49,2026.1957.

237
 EXPLANATORY LIST OF ABBREVIATIONS APPEARING IN TEXT

Abbreviation Full name (transliterated) Translation

AN AzSSR Akademiya Nauk Azerbai- Academy of Sciences of the

dzhanskoi SSR Azerbaidzhan SSR

DAN Doklady Akademii Nauk Proceedings of the Academy

of Sciences

GIAP Gosudarstvennyi Nauchno- State Scientific Research

Issledovatel'skii i and Design Institute of the
Proektnyi Institut Azotnoi Nitrogen Industry
Promyshlennosti

GITTL Gosudarstvennoe Izdatel'stvo State Publishing House of

Tekhniko-Teoreticheskoi Technical and Theoretical
Literatury Literature

GNTI Gosudarstvennoe Nauchno- State Publishing House of

Tekhnicheskoe Izdatel'stvo Science and Technology

GTTI see GITTL

Izd. Izdatel'stvo Publishing House

Izv. Izvestiya Bulletin

MEl Moskovskii Energeticheskii Moscow Institute of Power

Institut Engineering

MKhP Ministerstvo Khimicheskoi Ministry of the Chemical

Promyshlennosti Industry

ONTI Otdel Nauchno-Tekhniches- Department of Scientific

koi Informatsii and Technical Information

VTI Vsesoyuznyi Teplotekhni- Dzerzhinskii All-Union

. cheskii Institut im. F. E. Thermotechnical Institute
Dzerzhinskogo

ZhETF Zhurnal Eksperimental' noi Journal of Experimental and

i Teoreticheskoi Fiziki Theoretical Physics

ZhFKh Zhurnal Fizicheskoi Khimii Journal of Physical Chemis-

try

ZhTF Zhurnal Tekhnicheskoi Journal of Technical

Fiziki Physics

238