J. D.

Bernardin
Research Engineer,
Los Alamos National Laboratory

i. Mudawar
Professor and Director
Boiling and Two-Phase Flow Laboratory,
School of Mechanical Engineering,
Purdue University,
West Lafayette, IN 47907

The Leidenfrost Point:

Experimental Study and
Assessment of Existing Models
This study presents a detailed and thorough parametric study of the Leidenfrost point
(LFP), which serves as the temperature boundary between the transition and film boiling
regimes. Sessile drop evaporation experiments were conducted with acetone, benzene,
FC-72, and water on heated aluminum surfaces with either polished, particle blasted, or
rough sanded finishes to observe the influential effects of fluid properties, surface
roughness, and surface contamination on the LFP. A weak relationship between surface
energies and the LFP was observed by performing droplet evaporation experiments with
water on polished copper, nickel, and silver surfaces. Additional parameters which were
investigated andfound to have negligible influence on the LFP included liquid subcooling,
liquid degassing, surface roughness on the polished level, and the presence of polishing
paste residues. The accumulated LFP data of this study was used to assess several existing
models which attempt to identify the mechanisms which govern the LFP. The disagreement between the experimental LFP values and those predicted by the various models
suggests that an accurate and robust theoretical model which effectively captures the LFP
mechanisms is currently unavailable.

1 Introduction
Recent demands for superior material properties and more efficient use of materials and production time are forcing manufacturers to develop intelligent processing techniques for enhanced
process control in order to better dictate the end product. In the
heat treatment and processing of metallic alloys, the desire to
obtain parts of enhanced and uniform mechanical properties is
requiring increased control over heat removal rates and enhanced
temperature control. In particular, spray quenching has been
shown (Bernardin and Mudawar, 1995) to be an effective means to
control and enhance the cooling rates of heat treatable aluminum
alloys. Rapid quenching is required to obtain high material
strength, while uniform temperature control is necessary to reduce
warping and deformation. In addition, the quench rate and material
properties of aluminum alloys following solution heat treatment
are dictated mainly by low heat flux, high-temperature film boiling
spray heat transfer, and the Leidenfrost point (LFP) which forms
the lower temperature limit of the film boiling regime (Bernardin,
1993). Thus, when quenching most aluminum alloys, it is desirable
to traverse through the film boiling temperature range and get
below the LFP as quickly as possible. Consequently, accurate
knowledge of the Leidenfrost temperature is necessary if accurate
and enhanced control of the quenching process and resulting
material properties is desired.
A common technique used for determining the Leidenfrost
temperature requires measuring evaporation times of liquid sessile
droplets of a given initial volume over a range of surface temperatures to produce a droplet evaporation curve as shown in Fig.
1(b). The curve displays droplet evaporation lifetime versus surface temperature and exhibits the four distinct heat transfer regimes shown on the traditional pool boiling curve of Fig. 1(a). In
the single-phase regime, characterized by long evaporation times,
heat from the surface is conducted through the liquid film and is
dissipated by evaporation at the liquid-gas interface. In the nucleate boiling regime, vapor bubble production and the corresponding

heat flux increase dramatically, thus decreasing the droplet lifetime. The upper limit of the nucleate boiling regime, known as
critical heat flux (CHF), corresponds to a maximum heat flux and
minimum drop lifetime. In the transition regime, a noncontinuous,
insulating vapor layer develops beneath portions of the droplet,
leading to reduced evaporation rates and increased drop lifetime.
At the upper end of the transition boiling regime, referred to as the
LFP, the vapor layer grows substantially to prevent any significant
contact between the drop and surface and the droplet evaporation
time reaches a maximum. At surface temperatures above the LFP,
the droplet remains separated from the surface by a thin vapor
layer through which heat is conducted.
Literature Review and Focus of Current Study. Table 1
displays the large variations in the Leidenfrost temperature for
water which have been reported in the literature. The discrepancies
in these reported values arise from differences in size of the liquid
mass, method of liquid deposition, amount of liquid subcooling,
solid thermal properties, surface material and finish, pressure, and
presence of impurities. These parameters and their observed effects on the LFP are summarized in Table 2 along with the
corresponding references.
While many of the LFP investigations have been qualitative in
nature, several studies have reported various correlations for predicting the Leidenfrost temperature. One of the correlations most
frequently referred to is a semi-empirical expression developed by
Baumeister and Simon (1973). Adapting the superheat limit model
of Spiegler et al. (1963), Baumeister and Simon included corrections to account for the thermal properties of the heated surface
and wetting characteristics of the liquid-solid system, and arrived
at the following semi-empirical expression:
T
1

= T
leid.meas

-* /

0.844TJ 1 - exp - 0 . 0 1 6
'f
Contributed by the Heat Transfer Division for publication in the JOURNAL OF HEAT
TRANSFER. Manuscript received by the Heat Transfer Division, Feb. 2, 1998; revision
received, Mar. 18, 1999. Keywords: Boiling, Droplet, Evaporation, Film, Heat Transfer, Two-Phase. Associate Technical Editor: D. Kaminski.

894 / Vol. 121, NOVEMBER 1999

exp(3.066 X 106j3) erfc (1758 J/3)

(1)

where

Copyright 1999 by ASME

Transactions of the ASME

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

o
Vapor
o / Blanke
Single Nucleate Transition
Boiling
Phase Boiling
Regime Regime Regime
Isolated
3ubbles

YT_

Vapor
(^~~~\/ Blanket

Film
Boiling
Regime

Single Nucleate
Phase Boiling
Regime Regime

Film
Boiling
Regime

Jets and
Columns

^Critical Heat Flux

Incipience of
Nucleate Boiling

E
1=

Minimum Heat Flux

or Leidenfrost Point

.Leidenfrost Point

I
- Critical Heat Flux

log ATsi
Wall Superheat ATS.

Wall Superheat ATsa[

(a)

(b)

Fig. 1 (a) Boiling curve for a hot surface in a stagnant bath of liquid at saturation temperature and (b)
sessile drop evaporation curve

The first objective of this study is to present previously developed models that attempt to describe the governing LFP mechakspscPiS'
nisms. Next, experimental LFP data for several different liquidThe temperature generally measured and reported as the LFP solid systems from the current study will be used to assess these
corresponds to that of the solid in the near vicinity of the surface. models to display their weaknesses. Based upon lack of experiTo be more precise, it is better practice to associate the LFP with mental validation and sound scientific arguments, a need for a
the temperature of the liquid-solid interface, which is often several correct and robust theoretical model that correctly captures the
degrees less than that measured within the solid. It is commonly LFP mechanisms will be identified.
accepted that during the initial stages of droplet-surface contact,
the interface temperature between the liquid and solid is dictated 2 Previous LFP Models
by the thermal properties of the liquid and solid as well as by their
This section discusses several of the most commonly proposed
initial temperatures. This interface temperature, T,, is given by the
solution to the one-dimensional energy equation with semi-infinite mechanisms for the LFP for droplets and the minimum film boiling
point for pools of liquid. Table 3 contains a pictorial summary and
body boundary conditions (Eckert and Drake, 1972)
corresponding correlations associated with the various models.
|3 =

(kPcXX,+(kpcp)f%,,

=
Ti

(2)

'

(kpc,,) + (kpc,,)0/

(3)
'

Hydrodynamic Instability Hypotheses. Several authors (Zuber, 1958; Berenson, 1961; Hosier and Westwater, 1962; Yao and

Nomenclature
At
cp
d
g
h
h'fs
J
k
kb
M
m
N
Na
P

= atomic weight of surface material

= specific heat with constant pressure
= droplet diameter
= gravitational constant
= enthalpy
= modified latent heat of vaporization = cp(Tf Tm) + hlg
= vapor embryo formation rate per
unit volume of liquid
= thermal conductivity
= Boltzmann constant
= molecular weight, constant
= mass of a single molecule
= number of liquid molecules per
unit volume
= Avogadro's number
= pressure

Journal of Heat Transfer

Q = heat of adsorption
R = particular gas constant, drop, film,
or bubble radius
T = temperature
u = droplet velocity
v = specific volume, velocity
Greek Symbols
)3 = surface thermal parameter (kpcp) ~
T = number of monolayer surface adsorption sites
17 = parameter for embryo formation
rate equation
A = wavelength
p, = dynamic viscosity
d = contact angle
p = density
cr = surface tension

T = molecule residence time on surface

Subscripts
c = critical
/ = liquid
fg = difference between liquid and vapor
g = vapor
;' = interface
leid = Leidenfrost point
mfb = minimum film boiling point
o = initial
r = reduced property
s = surface, wall
sat = saturation
thn = thermodynamic homogeneous nucleation limit

NOVEMBER 1999, Vol. 121 / 895

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

Table 1 Summary of Leidenfrost temperatures for water (P = 101.3 kPa)

as reported in the literature
Reference
Blaszkowska
and Zakrzewka
(1930)
Borishansky
and Kutateladze
(1947)t
Borishansky
(1953)t

Tamura and
Tanasawa
(1959)
Gottfried
(1962)t
Betta(1963)tt
Lee(1965)tf

TtewCC)
157

Surface Material
Silver

310
255

Graphite

Tf= 20 C
T/= 85 C

222
194
250
237

Brass
Brass
Copper
Copper

7/=19C
Tf= 89 C
7/= 20 C
7/= 85 C
^ 0 =4.5 mm

302

Stainless steel

285

Stainless steel

245
280
320

Not given
Not given
Stainless steel

Godleski and
Bell (1966)
Gottfried et al.
(1966)
Kutateladzc and
Borishanski
(1966)
Patel and Bell
(1966)

Baumeister et
al. (1970)

Emmerson
(1975)
Xiong and
Yuen(1991)

Notes

280

Stainless steel

250

Not given

305

Stainless steel

515
305, 325
230, 235
>200
235
155
265
<184
282
316
284
280-310

Tf= 25 C
3.7 < d0 < 4.3
4.6 < d0
7.8 < d0
7W= 264 C for
extended liquid
masses and 161 C for
transient technique

Henry, 1978) have used a hydrodynamic stability theory by Taylor

(1950) to describe the minimum film boiling temperature for pool
boiling. Assuming potential flow and a sinusoidal disturbance
between two fluids of different densities (the more dense on top),
Taylor (1950) used a first-order perturbation analysis to show that
gravity induced interfacial disturbances with wavelengths given by
the following expression will be most likely to grow and disrupt
the smooth horizontal interface:
3 07

X,, = 2-n-

s(Pf~ P)

(4)

Berenson (1961) showed that the bubble spacing in film boiling

was hydrodynamically controlled by a Taylor-type instability and
that the presence of the corresponding vapor layer and bubble
departure supported film boiling by keeping the liquid from contacting the heated surface. Berenson's analytical expression, Eq.
(5), to predict the minimum film pool boiling temperature, Tmfb,
coincides with the point at which vapor is not generated rapidly
enough to sustain the Taylor waves at the liquid-vapor interface.
rmfb = r sat + 0 . 1 2 7

0.05<F<10ml
extended masses
Pyrex (3-4 mis)
d0 = 0.39 mm
Stainless steel (3-4 rms)
d0 = 0.39 & 2.25 mm
Brass (3-4 rms)
do = 0.39 & 2.25 mm
Brass fresh polish (3-4 rms)
d0 = 2.25 mm
Aluminum (3-4 rms)
Alum, fresh pol, (3-4 rms) d0 = 0.39 & 2.25 mm
dQ - 0.39 mm
Aluminum (25 rms)
Gold fresh polish
d0 ~ 2.25 mm
d0 - 2.25 mm
Stainless steel
LFP also given for
pressures of 210, 315,
Monel
420, and 525 kPa
Brass
Stainless steel

1* As referenced from Patel and Bell (1966), t t As referenced from Testa and Nicotra (1986)

Table 2

(5)
Sakurai et al. (1982) and Groenveld (1982) showed that Berenson's model was only fair in predicting their minimum film boiling
temperature data at low pressures and was in extreme error at high
pressures.
Metastable LiquidHomogeneous and Heterogeneous Nucleation Hypotheses. Yao and Henry (1978) and Sakurai et al.
(1982) proposed that spontaneous bubble nucleation at the liquidsolid interface is the mechanism for the minimum pool film boiling
point. Bubble nucleation can be either heterogeneous, in which the
vapor bubbles are produced within cavities at a solid-liquid inter-

Summary of the Influential LFP parameters

Parameter
Observations/References
Size of
LFP independent of liquid mass size (Gottfried et al. 1966 and Patel and Bell, 1966).
liquid mass LPF increased with droplet volume (Nishio and Hirata, 1978).
LFP differed between steady state drop size technique using a pipet and the transient sessile
drop technique (Godleski and Bell, 1966).
Method of LFP increased with droplet velocity (Patel and Bell, 1966, Yao and Cai, 1988; Klinzinger
liquid
al, 1993; and Labeish, 1994).
deposition LFP did not differ between sessile and impinging drops (u0 < 5 m/s) (Bell, 1967 and Nishio
and Hirata, 1978).
Liquid subcooling had little effect on LFP for water on polished aluminum, brass, and
Liquid
stainless steel, but did cause an increased LFP on Pyrex (Baumeister et al 1970).
subcooling Subcooling increased drop lifetime but did not influence the LFP (Hiroyasu et al., 1974).
Subcooling raised the LFP for water and other fluids at high pressures where both sensible
and latent heat exchange are significant (Emmerson and Snoek, 1978).
LFP increases as solid thermal capacitance decreases (Patel and Bell, 1966; Baumeister et
Solid
al, 1970; and Nishio and Harata, 1978).
thermal
Baumeister and Simon (1973) developed a LFP correlation accounts for solid thermal
properties properties.
LFP independent of solid thermal diffusivity (Bell, 1967 and Emmerson, 1975).
Gottfried et al. (1966) estimated that the vapor layer beneath a film boiling sessile water drop
was on the order of 10 urn, which is on the same length scale as surface aspirates on machine
finished surfaces (Bernardin, 1993). Thus, rough surfaces in comparison to polished surfaces
would be expected to require a higher LFP to support a thicker vapor layer to avoid liquidsolid contact for a sessile drop (Bradfield 1966).
Surface
LFP increased as surface roughness and fouling increased (Baumeister et al, 1970;
conditions Baumeister and Simon, 1973; and Nishio and Hirata, 1978). In contrast, Bell
(1967) claimed that surface oxide Films had a negligible effect on the LFP for droplets.
LFP increased with increasing surface porosity (Avedisian and Koplik, 1987).
LFP decreased with increased advancing contact angle in pool boiling (Kovalev, 1966; Unal
et al., 1992; and Labeish, 1994 and Ramilison and Lienhard, 1987).
LFP increased with pressure for various fluids (Nikolayev el al, 1974; Hiroyasu et al, 1974;
and Emmerson, 1975; Emmerson and Snoek, 1978)
(Tleld - Tsat) found to remain constant for various pressures (Hiroyasu et al, Emmerson,
Pressure Nishio and Hirata, 1978, and Testa and Nicotra, 1986).
Rhodes and Bell (1978) observed (T/ejd Tsa{) for Freon-114 to be constant over a reduced
pressure range of 0.125 to 0,350 and found it to decrease with increasing pressure above this
range. Klimenko and Snytin (1990) reported similar findings for four inorganic fluids.

896 / Vol. 121, NOVEMBER 1999

Transactions of the ASME

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

Table 3 Summary of proposed LFP models

Pictorial Description

Model

Relevant Correlations
Most dangerous wavelength:

Hydrodynamic
Instability

-H

^d

JKJ

K-

X j

=2K

TTT^TTTTTTTT^
Isotherm
"

Saturation

Mechanical stabiltiy condition:

'dP)

Metastable
liquidmechanical
stability

=0

dv
[unstable ^Vapor
region
spinodal

Spinodal or liquid superheat limit:

(using Van der Waals eqn.)

Liquid Spinodal
Homogeneous nucleation limit:

f3aV'5

Metastable
liquid kinetic
stability

-16,1 a3

\M, Tfbp..,(Tf)-pf]2
r\ = cxp'

RT,

Homogeneous nucleation
Implicit energy balance for LFP:

Film
boiling
Thermomechanical
effect

LFP^
Transition
boiling
CHF
Nucleate
boiling
Incipience

" < &

oO

"Liquid
flow

[pMidypm,(Te)]
Contact angle temperature dependence

Wettability contact
angle

Wettabiltiy surface
adsorption

*.(r.>*/0L)=ft5^(r.>v/(r)]

c(e)=i+c(r,~rp
Continuous
Monolayer

Discontinuous
Monolayer

Monolayer molecular surface coverage

temperature dependence:

a
r-

^jaur)f9s
Nafh

face as a result of the imperfect wetting of the liquid, or homogeneous, where the bubble nuclei are formed completely within the
liquid due to density fluctuations over a duration of 10 to 1(T8 s
(Skripov et al., 1980).
In the discussion that follows, the metastable state and related
physics of homogeneous and heterogeneous nucleation are briefly
presented. A more detailed and lengthier discussion of the subject
can be found in Skripov (1974) and Carey (1992).
In classical thermodynamics, phase transitions for simple compressible substances are treated as quasi-equilibrium events at
conditions corresponding to the saturation state. Between the saturated liquid and saturated vapor states exists a two-phase region
where liquid and vapor coexist. Within this region, the temperature
and pressure of the two phases must be constant, and the Gibbs
function, chemical potential, and fugacity of each phase must be
equal. In real-phase transformations, deviations from classical
thermodynamics occur under nonequilibrium conditions, such as
the superheating of a liquid above its boiling point. These nonequilibrium or metastable states are of practical interest and are
important in determining limits or boundaries of real systems.
Shown on the pressure-volume diagram in the pictorial of Table
3 are the superheated liquid and supercooled vapor regions separated by an unstable region. The lines separating these regions are
referred to as the liquid and vapor spinodals, which represent the
maximum superheating and supercooling limits.
Journal of Heat Transfer

RT,

Two different approaches have been used in the literature to

predict the superheat limit. The first, based on a mechanical
stability condition described by Eberhart and Schnyders (1973)
and Carey (1992) for a closed system containing a pure substance
which is not in thermodynamic equilibrium, is given as
DP
dv

<0.

(6)

Along the portion of the isotherm between the spinodal lines of

Fig. 2, the inequality dp/dv > 0 violates the mechanical stability
criterion given by Eq. (6). For this reason, this area is referred to
as the unstable region. In the metastable and stable regions, where
dp/dv < 0, the liquid or vapor may remain in its form indefinitely.
The spinodal limit, at which dp/dv = 0, represents the onset of
instability.
Cubic equations of state such as Van der Waals (Spiegler et al.,
1963), Himpan (Lienhard and Karimi, 1981), and Berthelot
(Blander and Katz, 1975) posses the type of behavior within the
vapor dome as discussed above and thus can be used to predict the
spinodal limit. Van der Walls equation in terms of the reduced
variables P, = PIP,, T, = TITe, and v, = vlvc, which have been
nondimensionalized with the corresponding critical point variables, can be written as
NOVEMBER 1999, Vol. 121 / 897

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

uously with temperature. However, because the exponential term

has such a strong dependence on the liquid equilibrium temperature, Tf, there exists a small temperature range over which the
embryo formation rate begins to increase in a drastic manner. It is
within this temperature range that the critical embryo formation
rate required to initiate homogeneous nucleation is defined with a
corresponding value of Tf equal to the maximum superheat or
kinetic homogeneous nucleation temperature. From experimental
superheat data for a large variety of fluids at atmospheric pressure,
Blander and Katz (1975) obtained a threshold value of 1012 m~3
s -1 . Using this value for J, Eq. (10) can be solved iteratively for
the maximum superheat temperature of a given liquid.
Carey (1992) showed how the development of Eq. (10) can be
modified to account for the liquid contact angle, 8, and thus
describe the heterogeneous nucleation rate of a liquid at a perfectly
smooth surface:

Syringe

Heated Cylinder
with Dished
Test Surface

Temperature
Controller

J=

Cartridge
Heater

iV?'3(l + c o s 6) (2Fa
IF

-16TTFO- 3

exp 3kT [i}P (T )

f
m f

PfV\
(12)

Fig. 2

Schematic diagram of sessile drop experimental apparatus

where

Pr =

8rr

(7)

3vr 1

Using this form of Van der Walls equation of state, the condition
of mechanical stability given by Eq. (6), and the fact that P, < 1
for most fluids at atmospheric conditions, the thermodynamic
homogeneous nucleation temperature limit, r,h, can be derived as
(Spiegler et al., 1963)
r,hn = 0.8447;

(8)

where absolute temperature quantities are used. Modified forms of

Eq. (8) using other equations of state and the success of these
models in predicting the superheat limits of liquids are discussed in
Carey (1992).
For fluids at higher pressures up to the critical point, Lienhard
(1976) offered the following maximum superheat correlation:
0.905 + 0 . 0 9 5 ( ^ ?

(9)

where absolute temperatures are implied.

The second approach to describing the maximum liquid superheat temperature is referred to as the kinetic homogeneous nucleation theory, which bases the temperature and pressure dependence
of bubble nucleation on molecular fluctuation probability. At and
above saturation conditions, molecular fluctuations occur in such a
way to cause a localized decrease in the liquid density, leading to
the formation of vapor embryos. The fluctuation probability increases with temperature, and at the superheat temperature limit,
the probability of a high bubble embryo formation rate is sufficient
to transform the liquid to vapor.
By using conventional bubble nucleation theory, Carey showed
how Eq. (10) could be derived to describe the rate of critical-size
embryo formation, J, for a superheated liquid
J = N,

3a

-16TT(T3

exp'

MTfcqPjTf) ~ PfY

(10)

where

"j[Pf-Pm(Tfm
T)

= exp

2 + 3 cos 9 - cos36
F =-

RT,

(ID

Slightly different assumptions have led to minor variations of Eq.

(10) by several authors (Skripov, 1974; Blander and Katz, 1975;
and Lienhard and Karimi, 1981).
The embryo formation rate given by Eq. (10) increases contin898 / Vol. 121, NOVEMBER 1999

(13)

The principle factor which is not accounted for in the homogeneous and heterogeneous nucleation models is the influence on the
molecular interactions caused by the presence of the solid-liquid
interface. Surface energies become influential and continuum fluid
theories are not necessarily valid within 50 A of the interface.
Gerwick and Yadigaroglu (1992) recognized that liquid molecular
interactions at an interface will be quite different from the bulk
liquid. Using statistical mechanics, they developed a modified
equation of state for the liquid which was a function of the distance
from the solid surface. This equation of state was used to predict
the superheat limit of the liquid and thus the rewetting or Leidenfrost temperature of the surface.
Thermomechanical Effect Hypothesis. Schroeder-Richter
and Bartsch (1990) refuted the superheated metastable hypothesis
of Spiegler et al. (1963) and proposed that the liquid and vapor
near the solid surface are in saturated states at different pressures.
The authors used, a nonequilibrium flow boiling model with conservation equations and appropriate boundary conditions across
the liquid-vapor interface, along with assumptions that the liquid
immediately in front of the interface is at the Leidenfrost temperature, and that the change in enthalpy during the evaporation is
supplied solely by the mechanical energy of the depressurizing
liquid to establish the following implicit equation for the Leidenfrost temperature:
hg(Tg) -

hf(Tkii)
= 0.5[vg(Tg)

- vf{T*JIpm(TlJ

- pJ.Tt)].

(14)

Using saturation tables and an iterative procedure, Eq. (14) can be

solved for the LFP.
Wettability Hypotheses. It has been speculated by several
researchers that the temperature dependence of the contact angle is
influential in controlling the Leidenfrost phenomenon. In a fundamental study by Adamson (1972), a theoretical model was developed that related the molecular surface adsorption of a solid to the
liquid-solid contact angle:
= 1 + C(TC0 - T) bHa-b)

(15)

where Tco represents a pseudo-critical temperature, or the temperature at which the contact angle goes to zero, C is an integration
constant, and b and a are temperature-independent coefficients
from a molecular force balance expression given by Adamson. It
Transactions of the ASME

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

is evident from Eq. (15) that the contact angle decreases with
increasing temperature, a trend consistent with experimental findings.
Based upon the work of Adamson, Olek et al. (1988) presented
a semi-theoretical analysis which suggests that the rewetting temperature or LFP corresponds to a zero contact angle or perfect
wetting. The authors suggested that at the temperature, TL., where
the contact angle goes to zero, the liquid drop spreads into a
sufficiently thin film such that enough vapor can be generated to
disjoin the film from the surface. Olek et al. were only able to
provide experimental data for two water-nonmetallic solid systems
with which to evaluate their model. Their comparison showed fair
agreement between the predicted and measured temperaturedependent contact angle trends. However, they failed to provide
Leidenfrost temperature data for the two surfaces.
Segev and Bankoff (1980) offered a more plausible explanation
of the Leidenfrost phenomenon based on wetting characteristics.
They proposed that wetting of a hot solid surface by a liquid is
controlled by a microscopic precursor film which advances in front
of the much thicker spreading liquid film. The presence of the thin
film, which is required for the advancing and wetting of the
remainder of the liquid, is controlled by the temperature-dependent
surface adsorption characteristics. The precursor film thickness
decreases with increasing temperature and drops off sharply as the
temperature threshold (the LFP) is reached. Above this temperature, adsorption of the liquid molecules beyond a monolayer is no
longer possible, and surface wetting cannot occur.
Segev and Bankoff based their model on the Langmuir adsorption isotherm

exp

r
r0

/(2irMRTysr0\

(^)
(QA

(16)

which describes the fraction of total monolayer surface adsorption

sites, r, occupied by foreign molecules in terms of the liquidsolid interface temperature, T ; , heat of adsorption, Q,n and residence time of a molecule in the adsorbed state, T. Segev and
Bankoff claimed that the LFP corresponds to a surface monolayer
coverage fraction of 0.9, and by using F0 = 1 0 " molecules/mz
and T0 = 10~' 3 s, Eq. (16) can be solved explicitly for the surface
temperature if the heat of adsorption of the fluid's vapor on the
solid is known.
3

Experimental Apparatus and Procedure

The sessile drop apparatus shown in Fig. 2 was used to study
the evaporation characteristics of droplets on a heated surface.
In particular, the liquid/solid interface temperature corresponding to the Leidenfrost point was determined from droplet evaporation curves for a variety of operating conditions. The sessile
drop facility consisted of an instrumented test heater module,
temperature controller, and a syringe. The various working
fluids included acetone, benzene, FC-72, an inert fluorocarbon
produced by the 3M corporation, and distilled water. Several
test heater modules were fabricated from either a solid aluminum or copper cylinder with a shallow concave surface designed to contain the liquid droplets during states of transition
and film boiling. To investigate surface material effects on the
LFP, several copper heater modules were also electroplated
with either silver or nickel to a thickness of 0.025 mm. The
heater module was mounted in an insulating shell formed from
G-7 phenolic, which is capable of withstanding surface temperatures of 300C for short durations. An Ogden Type 33 temperature controller, a Watlow 150 Watt cartridge heater, and a
calibrated Chromel-Alumel (type K) thermocouple (calibrated
accuracy = 0.2C) located 2.5 mm beneath the center of the
test surface were used to monitor and control the surface
temperature. A finite element analysis and several thermocouple
Journal of Heat Transfer

measurements near the edge of the module were used to verify

that the temperature distribution across the plane just beneath
the surface was uniform and representative of the surface temperature. Three different surface finishes including polished,
particle blasted, and rough sanded, with arithmetic average
surface roughness values of 97, 970, and 2960 nm, respectively,
were used in the study. A glass syringe with a 24-gauge hypodermic needle having a 0.58-mm (0.023-in.) outer diameter,
was used to slowly dispense droplets of uniform diameter onto
the test heater. A static force balance between gravity and
surface tension dictated the nearly consistent droplet diameter
for a given fluid. A high-speed Ektapro motion analyzer was
used to verify that the slow droplet generation technique produced uniformly sized droplets within an error band of ten
percent. Preliminary tests, performed with water and different
diameter needles, revealed no dependence of the LFP on initial
droplet size. This is consistent with findings reported by Gaottfried et al. (1966) and Patel and Bell (1966). Consequently,
only one initial droplet diameter (fluid dependent) was used in
this study.
For each test, single droplet evaporation times were recorded
versus surface temperature over a temperature range encompassing
the entire boiling spectrum for each particular fluid. The experiments began by dispensing a single drop from a syringe onto the
center of the test surface at a temperature well within the film
boiling regime from an approximate height of 1 cm. A manual
digital stopwatch was used to record the time to the nearest tenth
of a second for complete visual evaporation of the drop. To
minimize timer (0.1 s) and initial droplet size (10 percent)
errors, five evaporation times were recorded for each temperature
increment and then averaged together. This procedure was performed for ten-degree centigrade surface temperature increments
from a temperature within the fluid's film boiling regime down to
the boiling incipience temperature, with finer two degree centigrade increments being made around the LFP. Each set of droplet
evaporation data was used to generate a droplet evaporation curve,
similar to the one displayed in Fig. 1(b), from which the LFP was
identified by interpolation. The Leidenfrost temperature, or
droplet/solid interface temperature corresponding to the LFP, was
then determined with Eq. (3), using the measured surface temperature corresponding to the LFP.
The sources of experimental error in determining the Leidenfrost temperature included uncertainties in initial droplet size (10
percent), droplet evaporation time (0.1 s), and surface temperature measurement (0.2C). An additional error was imposed by
the graphical LFP interpolation uncertainty caused by the 2C gap
between data points near the LFP on the droplet evaporation plots.
The uncertainty in droplet evaporation time was deemed minimal
since the accuracy of the timer was nearly two orders of magnitude
smaller than typical droplet evaporation times near the LFP. The
uncertainty in droplet size was minimized by taking the average
evaporation time of five droplets at each data point. The temperature measurement uncertainty combined with the graphical LFP
interpolation error created by the 2C gap between data points,
resulted in a total experimental uncertainty of 4.4C. This was
found to be consistent with reproducibility tests that revealed the
LFP measurements were repeatable within 5C.
An extensive database was required for identifying key influential parameters and to assess several analytical and theoretical
models. Consequently, the experimental procedure was performed
for four different test fluids with and without degassing, various
degrees of liquid subcooling, four different surface materials, a
variety of surface finishes, and different forms of surface contamination. To investigate the effect of surface impurities left behind
from previous drops, two different tests were performed. In one
case, the surface was wiped clean with a fine tissue between
successive drops, and in the other case, the surface was left as is.
More detailed operating conditions for the various tests are discussed with the experimental results.
NOVEMBER 1999, Vol. 121 / 899

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

Table 4

Leidenfrost temperatures for various fluids and aluminum surface conditions

T,eiAQ
Surface Finish
Fluid

Acetone (wiped)

Acetone (unwiped)

Polished

Particle Blasted

Rough Sanded

135

155

160

[130,140,135]

[160, 150]

[160, 160]

185

200

178

[185, 185]

[195,205]

[180,175]

Benzene (wiped)

175

220

218

Benzene (unwiped)

180

215

215

FC-72 (wiped)

90

110

120

FC-72 (unwiped)

115

110

120

Water (wiped)

Water (unwiped)

171

250

263

[l75, 180,160, 17o]

[250, 250]

[260, 265]

225

280

263

[220, 230]

[280, 280]

[260, 265]

4 Experimental Results and Discussion

In the discussions that follow, the reported empirical Leidenfrost
temperatures correspond to measured surface temperatures at the
LFP. However, in the evaluations of the LFP models (Table 6),
both the empirical Leidenfrost temperatures and adjusted LFP
values (using Eq. (3) to account for the liquid/solid interface) are
presented.
Table 4 presents the LFP data for acetone, benzene, FC-72, and
distilled water on three different aluminum surface finishes for
both wiped and unwiped conditions between successive drops. The
average LFP values are displayed with large text in Table 4 while
the small test in brackets indicates Leidenfrost temperatures from
individual runs when more than one test was performed for a
single set of operating conditions. The focus of this experimental
data was to study the effects of fluid properties, surface roughness,
and surface contamination on the LFP.
The Leidenfrost temperature data of Table 4 indicate the following general trends:
Effect of Surface Roughness: For all test fluids, polished
surfaces had significantly lower Leidenfrost temperatures than
particle blasted and rough sanded surfaces. The surface roughness
dependence of the Leidenfrost temperature is speculated to be
related to intermittent liquid-solid contact caused by surface aspirates poking through the thin vapor layer, which, as reported by
Labeish (1994), is on the order of 1 /im. As the surface roughness
increases, a thicker vapor layer, and hence a higher surface temperature, is required to keep the liquid separated from the solid
surface. This effect would be expected to taper off as surface
roughness increases, which is observed in the similar Leidenfrost
temperatures for the particle blasted and rough sanded surfaces.
Effect of Surface Contamination: A wiped surface generally
had a considerably lower Leidenfrost temperature than an unwiped
surface. This was most evident for the polished surface and to a
lesser degree for the particle blasted and rough sanded surfaces.
900 / Vol. 121, NOVEMBER 1999

The surface deposits left from previous drops tended to serve as

vapor bubble nucleation sources when making contact with newly
deposited drops, much in the same way as the surface aspirates
acted on the rougher surfaces. With deposits present, a higher
surface temperature was required to sustain film boiling. This
finding is consistent with those of Baumeister et al. (1970) who
found that the Leidenfrost temperature for water on a freshly
polished aluminum surface was 155C, 70C less than that of a
conventional contaminated surface. It is intuitively obvious that
surface contamination from previous drops will act to increase the
roughness on a polished surface to a much larger degree than for
an initially much rougher surface. This explains why the Leidenfrost temperature for a polished surface is highly influenced by
deposits while the rougher surfaces are not.
Table 5 presents Leidenfrost temperature data for water and a
variety of polished surface materials. The numbers in large text
indicate average LFP temperature values while the numbers in
small text and brackets indicate single experimental data points.
The accuracy and sensitivity of the measurements resulted in a
15C band around the average Leidenfrost temperatures tabulated herein. The focus of this portion of the study was to investigate the influences of surface material, surface contamination
from polishing pastes, surface roughness on the polished level,
liquid subcooling, and liquid degassing on the LFP.
Effect of Surface Material and Polishing Paste Residue:
Leidenfrost temperature values were obtained for water on polished aluminum, silver, nickel, and copper. The average Leidenfrost temperature is nearly identical for the aluminum, silver, and
nickel surfaces but is significantly higher for the copper surface.
The higher LFP value of the copper surface is speculated to be the
result of surface roughening which accompanied large amounts of
surface oxidation during heating. Jeschar et al. (1984) also reported a higher Leidenfrost temperature for copper compared to
nickel and, as in this study, attributed this to roughening of the
copper test piece by heavy oxidation. Labeish (1994) reported
Transactions of the ASME

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

Table 5
faces

Measured Leidenfrost temperatures for water on polished sur-

Surface
Aluminum

Silver

Nickel

Copper

17oM
[185 17(1

176 [m M

173 i1 n
[|85

203

198 H

Nickel
181
Copper

Aluminum
Aluminum

Notes

Tkhi (C)

180 I9(
175 19(
170

193 M
175

170 17.'
180
[190 18(1
16^

18,[l90

Aluminum

185 M

Aluminum

ml 1 8 0 "i

Aluminum

178ll63

Aluminum

175

Aluminum

170

Aluminum

160

[175 16(3

Study: material effect

Surface: polished with 45. 30. 15. 9. 6. & 3
micron diamond paste and chemically cleaned
Study: material effect
Surface: polished with 45. 30. 15. 9. 6, & 3
micron diamond paste, silver plated, polished with
Simichrome, wiped with acetone, oxidized upon
heating
Study: material effect
Surface: polished with 45, 30,15, 9. 6, & 3
micron diamond paste, nickel plated, wiped with
acetone, no apparent oxidation
Study: material effect
Surface: polished with 45. 30. 15.9. 6, & 3
micron diamond paste and chemically cleaned,
heavy oxidation upon heating surface
Study: material effect
Surface: polished with Simichrome. nickel plated.
wiped with acetone, no apparent oxidation
Study: material effect
Surface: polished with Simichrome paste, heavy
oxiation upon heating surface
Study: roughness effect
Surface Prep.: polished with 45 micron paste
Study: roughness effect
Surface: polished with 45. 30. & 15 micron
diamond paste
Study: roughness effect
Surface: polished with 45. 30. 15. 9. 6. & 3
micron diamond paste
Study: roughness effect
* Surface: polished w / 4 5 , 30, 15, 9, 6, & 3 micron
diamond paste then with Simichrome paste
Study: degassing effects (water degassed)
Surface: polished with 9,6, & 3 micron diamond
paste and chemically cleaned
Study: subcooling effect (Tf= 90 C)
Surface: polished with 45, 30, 15, 9, 6, & 3
micron diamond paste and chemically cleaned
Study: subcooling effect (7y= 60 C)
Surface: polished with 45. 30. 15. 9. 6. & 3
micron diamond paste and chemically cleaned
Study: polishing paste effect
Surface: polished with Simichrome, then soaked
& wiped with acetone

theoretical rewetting wall temperatures for smooth surfaces of

different materials wetted by water drops. Accounting for surface
thermal properties and neglecting surface effects, nearly identical
rewetting temperatures of 270, 282, and 292C were predicted for
copper, nickel, and carbon steel, respectively. These predictions,
while higher in absolute value than those reported in this study,
indicate a relative insensitivity of the LFP to surface chemistry
effects.
As the data of Table 5 indicates, no significant difference was
observed in the Leidenfrost temperatures of polished aluminum

samples with the following surface finish preparations: polished

with Simichrome paste; polished with Simichrome paste followed
by soaking and wiping with acetone to remove the paste residue;
and, polished with an array of diamond compounds followed by an
acid bath chemical cleaning. The lack of variability in the LFP
values for these three surfaces suggests that the polishing paste
residue has little influence on the LFP.
Effect of Surface Roughness on the Polished Level: Average Leidenfrost temperatures for water on aluminum surfaces
polished with different grades of diamond polishing compound all
fell within a 15C band, thus indicating no significant dependence
of the LFP on surface roughness on the polished level.
Effect of Liquid Subcooling: For identical surface conditions, water liquid subcoolings of 10, 40, and 80C resulted in
Leidenfrost temperatures of 170, 170, and 175C, respectively.
The lack of sensitivity of the LFP on liquid subcooling results
because the small amount of liquid contained in a single droplet,
regardless of initial temperature, is rapidly heated to near saturated
conditions when placed on the surface. This finding was also
reported by Hiroyasu et al. (1974) and Grissom and Wierum
(1981).
Effect of Liquid Degassing: Table 5 lists average Leidenfrost
temperatures of 170C and 178CC for nondegassed and degassed
water, respectively, on a polished aluminum surface. Negligible differences of less than five percent were observed between nondegassed
and degassed Leidenfrost temperatures for acetone and FC-72 on
polished aluminum as well. Clearly, the effect of air and other noncondensible gases within the liquid on the LFP is minimal.

5 Assessment of Models
As mentioned previously, the temperature generally measured and
reported as the LFP corresponds to that of the solid in the near vicinity
of the surface. However, boiling is an interfacial phenomenon, and
thus it is better practice to associate the LFP with the temperature of
the liquid-solid interface. In the model assessments that follow, both
the empirical Leidenfrost temperatures measured within the solid, and
adjusted LFP values (using Eq. (3) to account for the liquid/solid
interface) are presented in Table 6 for comparison.
Evaluation of Instability Models. To investigate whether or
not a Taylor-type instability could control the Leidenfrost phenomenon, a length scale comparison can be made between the droplet
diameter and the Taylor most dangerous interfacial wavelength,
Arf. For Benzene, FC-72, and water the corresponding values of A,,
are 17.7, 8.4, and 27.3 mm, respectively. These wavelengths are of
the same order or larger than typical droplet diameters, which
indicates that the Taylor interfacial instability, while possibly
suitable for pool boiling analysis, does not lend itself to isolated

Table 6 Comparison of various Leidenfrost temperature (C) models to experimental data

for a polished aluminum surface
Measured
Leidenfrost
temperature
(C)

Corrected
liquid/solid
interface
Leidenfrost
temperature
(eqn. (3))

Bcrenson
(1961)
hydrodynamic
model

Thermodynamic
liomogen.
nucleation
limit
temperature

Kinetic
liomogen.
nucleation
limit
temperature

Baumcister
and Simon
(1973)
correlation

Acetone

134

132

152

156

198

130

Benzene

175

172

140

201

239

171

180

Water

170

162

152

273

310

156

221

FC-72

90

89

106

144

102

116

Fluid

SchroederRichter and
Bartsch
(1990)
thermomechanical
model

% Fluid properties unavailable to evaluate model.

Journal of Heat Transfer

NOVEMBER 1999, Vol. 121 / 901

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

boiling drops. Table 6 compares predictions for Tle[i, using Berenson's (1961) model for Tm{b (Eq. (5)) to experimentally measured
sessile drop Leidenfrost temperatures for several of the fluids used
in this study. The predictions show significant error for acetone
and benzene and give only satisfactory results for water.
Evaluation of Metastable Liquid Models. Two theoretical
models, the thermodynamic or mechanical stability model and the
kinetic homogeneous nucleation model, have been developed using entirely different approaches to predict the maximum superheat temperature of liquids. However, attempting to use these
models to predict the Leidenfrost temperature for sessile drops has
not met reasonable success.
The Leidenfrost point correlation of Baumeister and Simon
(1973) contains two sources of concern. First, in developing a
conduction model to account for a decrease in the surface temperature at liquid-solid contact, the authors fail to explain how they
arrived at the chosen value of an average heat transfer coefficient.
Second and most importantly, Baumeister and Simon introduce a
surface energy correction factor to the superheat model of Spiegler
et al. While this factor leads to a correlation which successfully fits
the data, the results may be deceiving in that they suggests that
homogeneous nucleation, around which the correlation is constructed, is the mechanism governing the Leidenfrost phenomenon, when in fact, it may not be.
Experimental Leidenfrost temperatures for various liquids on a
polished aluminum surface from the current study are compared to
thermodynamic and kinetic superheat limits as well as the correlation
of Baumeister and Simon (1973) in Table 6. All predictions were
made with absolute temperature quantities and then converted to
degrees Celsius. For the theoretical metastable liquid models, the
superheat limits are considerably higher than the measured Leidenfrost temperatures for all fluids tested, consistent with the results of
Spiegler et al. (1963). The semi-empirical correlation by Baumeister
and Simon agrees quite well with the experimental data of the present
study, but as previously mentioned, it fails to accurately model the
physics governing the process. Obviously, superheat criteria alone do
not accurately describe the Leidenfrost phenomenon for sessile drops
on a heated surface.
While elegant, the modified equation of state and homogeneous
nucleation model of Gerwick and Yadigaroglu (1992) involved
several assumptions which severely limit its applicability and
accuracy. First, a simple hard-sphere potential interaction model
using London dispersion forces was used to describe the molecular
interactions. This limits the model's applicability to nonpolar
liquids, since liquids such as water, with highly polar hydrogen
bonding forces, would not lend themselves to such modeling with
any high degree of accuracy. Second, a parameter describing the
strength of the wall-fluid interactions was stated to be unknown for
most practical applications. Consequently, a simplified model
which related this parameter to the contact angle was employed.
The major argument against this simplification is that the contact
angle is typically measured over a distance which is at least six
orders of magnitude larger than the thickness of the fluid layer
influenced by the presence of the solid surface. In fact, Adamson
(1982) has hypothesized that the microscopic contact angle at the
leading edge of the liquid film, which is on the order of several
molecular diameters in thickness, is significantly smaller than the
macroscopic contact angle commonly reported. In addition, the
contact angle is highly influenced by surface roughness and impurities (Miller and Neogi, 1985; Bernardin et al., 1997), making
it a highly undefined variable.
Evaluation of Nonequilibrium Model. Table 6 compares
Leidenfrost temperatures predicted by Eq. (14) to experimentally
measured values for several different fluids. The prediction for
Benzene is quite good, while that for FC-72 is satisfactory, and the
estimate for water is extremely poor.
Several problems exist in the development of Eq. (14) and its
application to predicting the Leidenfrost temperature for droplets.
First, the original model was constructed to emulate a vertical
902 / Vol. 121, NOVEMBER 1999

dry-out flow boiling situation, a condition far from that of a sessile

or impinging droplet. Next, and more importantly, the concept of
saturated states at different pressures for the liquid and vapor
rather than metastable superheating of the liquid at constant pressure is unsupported. Metastable states for fluids have been frequently observed (Avedisian, 1982; Shepherd and Sturtevant,
1982; McCann et al., 1989) and the physics of such nonequilibrium states have been well documented (Eberhart and Schnyders,
1973; Skripov, 1974; Lienhard and Karimi, 1978; Carey, 1992). In
fact, liquid superheating forms the entire well established basis for
bubble nucleation theory in boiling (Han and Griffith, 1965;
Blander et al., 1971).
Evaluation of Wettability Models. The reasoning behind the
contact angle model of Olek et al. (1988) appears unrealistic. In
addition, the implicit equation for the LFP is difficult to verify
since the required coefficients are only available for a few liquidsolid systems for which no Leidenfrost temperature data exists.
The temperature-dependent contact angle measurements found by
Bernardin and Mudawar (1997) for water on aluminum show little
indication of a zero contact angle condition acting as the Leidenfrost point mechanism. Also in contrast to the model of Olek et al.,
nearly identical Leidenfrost temperatures were obtained in this
study for two identically polished aluminum surfaces, one of
which was left with a polishing paste residue, and the other which
was chemically cleaned. Also, nearly identical Leidenfrost temperatures were obtained for aluminum, silver, and nickel surfaces,
all of which have different wetting characteristics. The contact
angle depends to such a large extent on the surface conditions
(roughness, contamination, adsorption), as well as on liquid velocity and direction, it is a difficult parameter to characterize and
effectively utilize. Thus it can be concluded that while surface
wetting, as measured by the contact angle, may play a role in
boiling heat transfer, it is not the controlling LFP mechanism.
The surface adsorption hypothesis of Segev and Bankoff (1980) is
very difficult to verify for a liquid-surface combination because it
requires the corresponding heat of adsorption of the fluid's vapor on
the solid surface. Correct knowledge of the chemical makeup of a
solid surface is very difficult to obtain. The presence of oxide layers
or adsorbed layers of grease and other impurities changes the surface
chemistry considerably. In addition, the experimental data of this
study tends to disprove the hypothesis proposed by Segev and
Bankoff. Using heat of adsorption for water vapor on aluminum oxide
(McCormick and Westwater, 1965) and nickel oxide (Matsuda et al.,
1992), Eq. (16) predicts Leidenfrost temperatures of 162 and 425C
for saturated water on aluminum and nickel, respectively. The predicted LFP value for the aluminum surface agrees reasonably well
with the corresponding experimental value of 170C, however, the
model fails miserably for the nickel surface which had an experimental Leidenfrost temperature of 175C. Segev and Bankoff s model
suggests that the LFP for an aluminum surface possessing a polishing
paste residue would be significantly different from an identically
polished surface without the residue, a trend not observed in the
experimental data of this study.

6 Conclusions
Sessile drop evaporation experiments were performed for a wide
variety of operating conditions to establish a large LFP data base
for identifying key influential parameters and assessing existing
LFP models. From the experimental results, several key conclusions concerning the influential LFP parameters can be drawn.

Liquid subcooling, the presence of dissolved gasses, and surface roughness on the polished level do not significantly influence the Leidenfrost temperature.
Surface thermal properties will act to control the interface and
hence Leidenfrost temperature. However, aside from thermal
properties, the LFP is relatively insensitive to surface material as
far as surface energies and wetting characteristics are concerned.
Surface roughness, beyond that on the polished level, appears to
Transactions of the ASME

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

be a dominant parameter in controlling the Leidenfrost behavior.

The data indicate, that for a given fluid, a polished surface possesses a relatively low Leidenfrost temperature in comparison to a
particle blasted or rough sanded surface. In addition, surface
impurities or deposits act to increase the relative surface roughness
and the corresponding Leidenfrost temperature.

Han, C. Y and Griffith, P., 1965, "The Mechanism of Heat Transfer in Nucleate
Pool BoilingPart I," Int. J. Heat Mass Transfer, Vol. 8, pp. 887-904.
Hiroyasu, H., Kadota, T., and Senda, T 1974, "Droplet Evaporation on a Hot
Surface in Pressurized and Heated Ambient Gas," Bulletin of the JSME, Vol. 17, pp.
1081-1087.
Hosier, E. R., and Westwater, J. W., 1962, "Film Boiling on a Horizontal Plate,"
ARSJ., Vol. 32, pp. 553-558.
Jeschar, R., Scholz, R., and Reiners, U., 1984, "Warmeubergang bei der zweiphasiSound arguments supported by experimental data were used to gen spritzwasserkuhlung," Gas-Warme Int., Vol. 33, p. 6.
Klimenko, V. V., and Snytin, S. Y 1990, "Film Boiling Crisis on a Submerged
assess several hypothetical models of the LFP mechanism. These
Surface," Exp. Thermal Fluid Sci., Vol. 3, pp. 467-479.
models were shown to lack robustness and were ineffective in Heating
Klinzing, W. P., Rozzi, J. C , and Mudawar, I., 1992, "Film and Transition Boiling
predicting the Leidenfrost temperature. A model which success- Correlations for Quenching of Hot Surfaces with Water Sprays," J. Heat Treating,
fully captures the Leidenfrost mechanism is currently being devel- Vol. 9, pp. 91-103.
Kovalev, S. A., 1966, "An Investigation of Minimum Heat Fluxes in Pool Boiling
oped to account for several parameters which were found to
actively influence the LFP in both previous investigations and the of Water," Int. J. Heat Mass Transfer, Vol. 9, pp. 1219-1226.
Labeish, V. G 1994, "Thermohydrodynamic Study of a Drop Impact Against a
current study. These parameters include thermal properties of the Heated
Surface," Exp. Thermal Fluid Sci., Vol. 8, pp. 181-194.
solid, thermal and thermodynamic properties of the liquid, solid
Lienhard, J. H., 1976, "Correlation for the Limiting Liquid Superheat," Chem. Eng.
surface structure, pressure, and droplet impact velocity.
Sci., Vol. 31, pp. 847-849.
Lienhard, J. H., and Karimi, A. H 1978, "Corresponding States Correlations of the
Extreme Liquid Superheat and Vapor Subcooling," ASME JOURNAL OF HEAT TRANSReferences
FER, Vol. 100, pp. 492-495.
Adamson, A. W., 1972, "Potential Distortion Model for Contact Angle and SpreadMatsuda, T., Taguchi, H and Nagao, M 1992, "Energetic Properties of NiO
ing II. Temperature Dependent Effects," J. Colloid Interface Sci., Vol. 44, pp. Surface Examined by Heat-of-Adsorption Measurement," J. Thermal Analysis, Vol.
273-281.
38, pp. 1835-1845.
Adamson, A. W., 1982, Physical Chemistry of Surfaces, John Wiley and Sons, Inc.,
McCann, H Clarke, L. J., and Masters, A. P., 1989, "An Experimental Study of
New York.
Vapor Growth at the Superheat Limit Temperature," Int. J. Heat Mass Transfer, Vol.
Avedisian, C. T., 1982, "Effect of Pressure on Bubble Growth Within Liquid
32, pp. 1077-1093.
Droplets at the Superheat Limit," ASME JOURNAL OF HEAT TRANSFER, Vol. 104, pp.
McCormick, J. L., and Westwater, J. W 1965, "Nucleation Sites for Dropwise
750-757.
Condensation," Chem. Eng. Sci., Vol. 20, pp. 1021-1036.
Avedisian, C. T., and Koplik, J., 1987, "Leidenfrost Boiling of Methanol Droplets
Miller, C. A., and Neogi, P., 1985, Inteifacial Phenomena, Marcel Dekker, New York.
on Hot Porous Ceramic Surfaces," Int. J. Heat Mass Transfer, Vol. 30, pp. 379-393.
Nikolayev, G. P., Bychenkov, V. V and Skripov, V. P., 1974, "Saturated Heat
Baumeister, K. J., Henry, R. E., and Simon, F. F., 1970, "Role of the Surface in the
Transfer to Evaporating Droplets from a Hot Wall at Different Pressures," Heat
Measurement of the Leidenfrost Temperature," Augmentation ofConvective Heat and
TransferSoviet Research, Vol. 6, pp. 128-132.
Mass Transfer, A. E. Bergles and R. L. Webb, eds., ASME, New York, pp. 91-101.
Nishio,. S and Hirata, M 1978, "Direct Contact Phenomenon between a Liquid
Baumeister, K. J., and Simon, F. F 1973, "Leidenfrost TemperatureIts CorreDroplet and High Temperature Solid Surface," Sixth International Heat Transfer
lation for Liquid Metals, Cryogens, Hydrocarbons, and Water," ASME JOURNAL OF
Conference, Vol. 1, Toronto, Canada, Hemisphere, New York, pp. 245-250.
HEAT TRANSFER, Vol. 95, pp. 166-173.
Olek, S Zvirin, Y., and Elias, E 1988, "The Relation between the Rewetting
Bell, K. J., 1967, "The Leidenfrost Phenomenon: A Survey," Chem. Eng. Prog.
Temperature and the Liquid-Solid Contact Angle," Int. J. Heat Mass Transfer, Vol.
Symposium Series, Vol. 63, AIChE, New York, pp. 73-82.
31, pp. 898-902.
Berenson, P. J., 1961, "Film Boiling Heat Transfer from a Horizontal Surface,"
Patel, B. M and Bell, K. J., 1966, "The Leidenfrost Phenomenon for Extended
ASME JOURNAL OF HEAT TRANSFER, Vol. 83, pp. 351-358.
Liquid Masses," Chem. Eng. Progress Symposium Series, Vol. 62, pp. 62-71.
Bernardin, J. D., 1993, Intelligent Heat Treatment of Aluminum Alloys: Material,
Ramilison, J. M and Lienhard, J. H., 1987, "Transition Boiling Heat Transfer and
Surface Roughness, and Droplet-Surface Interaction Characteristics, Masters thesis,
the Film Transition Regime," ASME JOURNAL OF HEAT TRANSFER, Vol. 109, pp.
Purdue University, West Lafayette, Indiana. IN.
746-752.
Bernardin, J. D., and Mudawar, I., 1995, "Validation of the Quench Factor
Rhodes, T. R., and Bell, K. J., 1978, "The Leidenfrost Phenomenon at Pressures up
Technique in Predicting Hardness in Heat Treatable Aluminum Alloys," Int. J. Heat
to the Critical," Sixth International Heat Transfer Conference, Vol. 1, Toronto,
Mass Transfer, Vol. 38, pp. 863-873.
Canada, Hemisphere, New York, pp. 251-255.
Bernardin, J. D., Mudawar, I., and Franses, E. I., 1997, "Contact Angle TemperSakurai, A., Shiotsu, M., and Hata, K 1982, "Steady and Unsteady Film Boiling
ature Dependence for Water Droplets on Practical Aluminum Surfaces," Int. J. Heat
Heat Transfer at Subatmospheric and Elevated Pressures," Heat Transfer in Nuclear
Mass Transfer, Vol. 40, pp. 1017-1033.
Reactor Safety, S. G. Bankoff and N. H. Afgan eds., Hemisphere New York, pp.
Blander, M., Hengstenberg, D., and Katz, J. L., 1971, "Bubble Nucleation in
301-312.
w-Pentane, n-Hexane, n-Hentane + Hexadecane Mixtures, and Water," J. Phys.
Schroeder-Richter, D., and Bartsch, G., 1990, "The Leidenfrost Phenomenon
Chem., Vol. 75, pp. 3613-3619.
caused by a Thermo-Mechanical effect of Transition Boiling: A Revisited Problem of
Blander, M., and Katz, J. L., 1975, "Bubble Nucleation in Liquids," Amer. Inst.
Non-Equilibrium Thermodynamics," Fundamentals of Phase Change: Boiling and
Chem. Eng. J., Vol. 21, pp. 833-848.
Condensation, ASME, New York, pp. 13-20.
Blaszkowska-Zakrzewska, H., 1930, "Rate of Evaporation of Liquids from a
Segev, A., and Bankoff, S. G., 1980, "The Role of Adsorption in Determining the
Heated Metallic Surface," Bulletin International de VAcademie Polonaise, Vol.
Minimum Film Boiling Temperature," Int. J. Heat Mass Transfer, Vol. 23, pp.
4a-5a, pp. 188-190.
Bradfield, W. S., 1966, "Liquid-Solid Contact in Stable Film Boiling," I & E C 637-642.
Shepherd, J. E., and Sturtevant, B., 1982, "Rapid Evaporation at the Superheat
Fundamentals, Vol. 5, pp. 200-204.
Limit,"/ Fluid Mechanics, Vol. 121, pp. 379-402.
Carey, V. P., 1992, Liquid-Vapor Phase Change Phenomena: An Introduction to
Skripov, V. P., 1974, Metastable Liquids, John Wiley and Sons, New York.
the Thermophysics of Vaporization and Condensation Processes in Heat Transfer
Skripov, V. P., Sinitsyn, E. N., and Pavlov, P. A., 1980, Thermal and Physical
Equipment, Hemisphere, New York.
Properties of Liquids in the Metastable State, Atomizdat, Moscow.
Eberhart, J. G., and Schnyders, H. C , 1973, "Application of the Mechanical
Spiegler, P., Hopenfeld, J., Silberberg, M., Bumpus, Jr., C. F and Norman, A
Stability Condition to the Prediction of the Limit of Superheat for Normal Alkanes,
1963, "Onset of Stable Film Boiling and the Foam Limit," Int. J. Heat Mass Transfer,
Ether, and Water," J. Phys. Chem., Vol. 77, pp. 2730-2736.
Vol. 6, pp. 987-994.
Eckert, E. R. G., and Drake, Jr., R. M., 1972, Analysis of Heat and Mass Transfer,
Taylor, G. I., 1950, "The Instability of Liquid Surfaces when Accelerated in a Direction
McGraw-Hill, New York.
Emmerson, G. S., 1975, "The Effect of Pressure and Surface Material on the Perpendicular to their Plane, I," Proc. Royal Society of London, Vol. A201, p. 192.
Testa, P., and Nicotra, L 1986, "Influence of Pressure on the Leidenfrost TemLeidenfrost Point of Discrete Drops of Water," Int. J. Heat Mass Transfer, Vol. 18,
perature and on Extracted Heat Fluxes in the Transient Mode and Low Pressure,
pp. 381-386.
Emmerson, G. S., and Snoek, C. W., 1978, "The Effect of Pressure on the Transactions of the ASME, Vol. 108, pp. 916-921.
Unal, C , Daw, V., and Nelson, R. A., 1992, "Unifying the Controlling Mechanisms
Leidenfrost Point of Discrete Drops of Water and Freon on a Brass Surface," Int.
for the Critical Heat Flux and Quenching: The Ability of Liquid to Contact the Hot
J. Heat Mass Transfer, Vol. 21, pp. 1081-1086.
Surface," ASME JOURNAL OF HEAT TRANSFER, Vol. 114, pp. 972-982.
Gerweck, V., and Yadigaroglu, G., 1992, "A Local Liquation of State for a Fluid in
Xiong, T. Y., and Yeun, M. C , 1990, "Evaporation of a Liquid Droplet on a Hot
the Presence of a Wall and its Application to Rewetting," Int. J. Heat Mass Transfer,
Plate," Int. J. Heat Mass Trans., Vol. 34, pp. 1881-1894.
Vol. 35, pp. 1823-1832.
Yao, S. C, and Henry, R. E., 1978, "An Investigation of the Minimum Film Boiling
Godleski, E. S., and Bell, K. J., 1966, "The Leidenfrost Phenomenon for Binary
Temperature on Horizontal Surfaces," Transactions of the ASME, Vol. 100, pp.
Liquid Solutions," Third International Heat Transfer Conference, Vol. 4, Chicago,
IL, AIChE, New York, pp. 51-58.
263-266.
Gottfried, B. S Lee, C. J., and Bell, K. J., 1966, "The Leidenfrost Phenomenon:
Yao, S. C , and Cai, K. Y., 1988, "The Dynamics and Leidenfrost Temperature of
Film Boiling of Liquid Droplets on a Flat Plate," Int. J. Heat Mass Transfer, Vol. 9,
Drops Impacting on a Hot Surface at Small Angles," Exp. Thermal Fluid Sci., Vol. 1,
pp. 1167-1187.
pp. 363-371.
Grissom, W. M and Wierum, F. A., 1981, "Liquid Spray Cooling of a Heated
Zuber, N 1958, "On the Stability of Boiling Heat Transfer," Transactions of the
Surface," Int. J. Heat Mass Transfer, Vol. 24, pp. 261-271.
ASME, pp. 711-720.

Journal of Heat Transfer

NOVEMBER 1999, Vol. 121 / 903

Downloaded 29 Jul 2010 to 128.211.163.127. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm