WORKSHOP PRACTICE SERIE

HARDENING, TEMPERING
AND HEAT TREATMENT
for model engineers

Tubal Cain

ARGUS BOOKS
Argus Books
Argus House
Boundary Way
Hemel Hempstead
Hertfordshire HP2 7ST

(OGArgus Books Ltd 1984

Second impression 1985

Reprinted 1987, 1988, 1990

All rights reserved. No part of this publication may be

reproduced in any form, by print, photography, microfilm
or any other means without written permission from the
publisher.

ISBN 0 85242 837 5

Phototypesetting by Performance Typesetting. Milton Keynes

Printed and bound in Great Britain by BPCC Wheatons Ltd, Exeter

Contents

Chapter 1 Iron & Steel

Chapter 2 Principles of the Hardening Process
Chapter 3 Heating and Quenching in Practice
Chapter 4 Tempering
Chapter 5 Heating Equipment
Chapter 6 Casehardening
Chapter 7 Other Heat Treatment Processes
Chapter 8 The Measurement of Hardness
Chapter 9 Home Construction of Furnaces
Chapter 10 Safety Precautions 102
Appendix 1 Thermocouples and Pyrometers 107
Appendix 2 Carbon Steel Cutting Tools 113
Appendix 3 British Standard Steel Specification Numbers 115
Appendix 4 Hardness Conversions 117
CHAPTER 1

Iron & Steel

LI

If "Know your Enemy" is prudent counsel wholly iron.) The common ores are in the
for the soldier, then equally so must be form of oxides, some, notably Haematite,
"Know your Materials" for the engineer — almost 90% ferric oxide whilst others may
model or otherwise. Important enough contain as little as 2596, the remainder
when the materials are to be cut or being lime or silica based stone. Other
formed, but even more so when we types of ore contain iron carbonates, and a
propose to alter their characteristics. And very abundant source is Iron Pyrites —
there can be few such alterations so basically iron sulphide and seldom used in
extreme as when we convert a relatively the manufacture of iron directly.
Soft and ductile material into one which is Iron is extracted from the ore in a Blast
hard enough to act as a cutting tool. So, I Furnace — Fig.1. Even in early times these
make no excuse for this initial exploration were large structures, but today they are
of the nature of iron and steel. You can huge. The "hearth" at the bottom is 45
"skip" the chapter if you wish — it may be feet across — and the stack 100 feet or
that it tells you nothing new, and could more in height, with another 100 feet of
well seem to be irrelevant if all you want "top hamper” above. A medium sized
to do is to harden a scriber point! But to furnace will contain 50,000 cu. feet of
begin at the beginning is always a sound material and will "make" 9000 tons/
policy, and | hope you will bear with me; it week: Fig 2. The largest furnace in this
won't take long! country is designed for 10,000 tons per
Pure iron, known as FERRITE to the day, and keeping it fed with raw materials
metals specialist, is a relatively soft is a major part of the plant operator's
material, with a tensile strength of about concern!
24 tons/sq.in and quite ductile. It can be These raw materials are iron ore, coke
drawn into fine wires and rolled into thin (the fuel) and limestone to act as a "flux"
plates — about the only uses for it in so that the stony matter in the ore may be
practice these days. It is chemically very sufficiently fluid when melted. These
"active", combining readily with many materials, blended and sintered into a
Other substances, so that metallic iron is uniform size and composition, are fed in at
Seldom found in nature despite the fact the top. Preheated air is blown in at the
that it is the most abundant of earthly bottom, through the "tuyeres", and this air
materials. (The Earth's core is almost forms by far the largest mass of material

7
 fuel gas. By the time the "burden" reaches
the lower part of the furnace the flux and
stone will begin to combine to form a slag
which melts and runs down through the
coke. Slightly lower down the iron also
melts. Both slag and iron collect in the
hearth, with the slag floating on top. At
periodic intervals the slag is tapped off
through one hole and the iron from
another, being collected in railway ladies,
the former being used for road metal or
Typical heights cement making and the iron either cast
30 ft(1820)
Ort 11920)
into "pigs" for the foundry or taken,
1007 (1960) molten, to the steelworks.
During the final part of this process the
molten iron is in close contact with very
hot coke and will absorb up to 496 of
carbon. (This may not sound much until |
tell you that a chunk of iron of 22 inch
cube will contain as much carbon as a
sack of coke!) The important thing to
ALE
» 5, eg | ladle appreciate at this stage is that the iron
and carbon form a solution, just like the
sugar in your tea. After passing through a
DIAGRAM OF BLAST FURNACE pasty stage whilst cooling the iron
Fig.! solidifies at about 1130°C, and we now
have a solid solution of iron (Ferrite) and
carbon, but with some of the carbon now
combining to form an iron carbide.
used. Though nominally "free" it has to be However, as the metal cools further its
compressed to from 30 to 45 Ib/sq.in, and capacity to dissolve carbon diminishes,
heated to around 750°C, using turbo- and some "free" carbon appears at the
compressars of 10,000 HP or more, and grain boundaries as graphite. The final
enormous “stoves” to achieve the state does depend somewhat on how fast
required temperature.
Combustion of the coke at the hearth
results in a high temperature and the A
GRAPHITE
production of carbon dioxide. This latter
reacts with the very hot coke and is
reduced to carbon monoxide. This, in turn,
reacts with the iron oxide in the ore to
remove the oxygen content. When the
gases reach the top of the furnace they
will contain about 2796 carbon monoxide,
a little hydrogen, and the rest is carbon
Fig. 3 Structure of grey cast iron, with Hakes of graphite
dioxide and nitrogen; it is a very useful between the grains of metal,
je metal has cooled, but the "grey" cast to render "iron" more ductile. These led to
on which we use will appear under the the manufacture of WROUGHT IRON,
jicroscope as small agglomerations using a process which removed most of
gains) or iron carbide and Ferrite the carbon from the iron made in the
is, surrounded by flakes of graphite. blast-furnace. Pig-iron was melted in a
*ig.3) If the cooling is rapid then more of coal or coke-fired furnace (Fig.4) lined
18 carbon will remain as carbide and the with material containing a large amount of
yetal will be whiter and harder. Hence the iron oxide. A certain amount of oxide was
ceasional "hard spot" at the corner of a also charged with the pig-iron. About 4
jin casting: the metal has cooled too fast. cwt. was melted at each heat and when
Iron in this form is very useful: it can be molten was, of course, in intimate contact
ast into complex shapes and is very with the oxides. The result was that the
trong in compression. But it is less strong carbon in the iron combined with the
| tension and rather brittle; it cannot be" oxygen in the oxides to form carbon
ent or forged. It is useless for edged tools dioxide — the reaction was quite violent at
nd for many machine parts. Very early in times, with the metal appearing to "boil",
18 history of metals means were sought The process was accelerated by the
ig. 2 Four medium-sized furnaces at the Frodinghar works af the British Steel Corporation, Scunthorpe. These make nearly
2,000 tons of iron per week berween them. The four hot-blast stoves serving the nearest furnace aro on the left. The fumaces are
out 220 feet overall height. (Photo, Courtesy British Steel Corporation).
 The bars from the hammer were cut
into. convenient lengths and bound
together ("faggoted") with iron wire, and
these bundles were then reheated to
welding temperature and reforged into
billets, This process could be repeated
several times; each expelled more of the
remaining slag and what was left was in
the form of thin streaks.
The quality known as "Best" was meta!
which had been faggoted twice. "Bes:
Best" was made from faggoted "Best"
bars, and "Treble Best" from faggoted
"Best Best". "Best" bars would have s
tensile strength along the grain of the slag
of about 23 tons/sq.in, whilst “Treble
Best” might reach 28 ton/sqin. The
strength across the fibres would be abou
15% less — the process cannot completely
eliminate all the slag and this forms &
fibrous structure within the material.
Fig. 4 Section and plan of a 19th century puddiing furnace. Under the microscope the main body o!
The bath "C" wouldbe about Git long and 4ft wide. The coal the metal is almost pure Ferrite, intersper-'
fire would be on the grate on the loft
sed with bands of slag, as seen in Fig. 5.
furnaceman, who "rabbled" or stirred the Wrought iron was too soft for use as a
bath with an iron bar suspended from a cutting tool, and from the earliest times a
chain. new material, having a carbon content
Now, the melting point of iron depends midway between wrought iron and cast
upon its carbon content, with pig melting irons, was made from best quality
at about 1130°C and pure iron at about faggoted bars. This was called STEEL, and
1500°C. As the carbon content fell, any reference to metal of this name
therefore, pasty masses appeared in the previous to about 1855 must be assumed
melt and these clung to the end of the to apply to this, and not to the "mild" stee!
puddier's rabble. When this "ball" was as we know today. The specially selected
large as he could handle he extracted it
from the furnace and immediately set it
under the tup of a mechanical hammer.
(Fig.5a) The hammering drove out most of
the slag in the ball and at the same time
formed it into a bar. The process was
repeated until all the iron had been
extracted from the furnace, after which
the furnace walls were fettled with fresh
oxide on the lining and then recharged.
About six heats could be worked in a Fig. B Structure of wrought iron "along the grain" showing
the slag inclusions.
normal 12-hour shift.
10
jars (3 in. wide x Zin. thick) were heated welded (with some care) to make larger
|a box containing charcoal for periods of pieces, or, for very large objects, several
Pto 10 days at a temperature of about crucibles cast into one (sand) mould. The
(000°C. During this time the iron steel could, of course, be hardened lin the
carbon and as the bars were fashion later to be described — that's what
elatively thin this absorption penetrated this book is about!) but it was frequently
through. Means were provided for used simply as a tough, strong material.
vithdrawing a test-bar from time to time, It may seem odd that we first make a
hus giving a crude form of quality control, material high in carbon, then remove most
Up to 15 tons at a time could be treated) of it, and finally, add carbon to produce
fhe carbon content could be roughly material of the required analysis, but a
yonitrolled by the length of the heat, but if little reflection will show that as each step
oo short the penetration would be also refined the iron, to some extent
ncomplete. Carbon content could never removing impurities, and the final step
ye. below about 0.8%, but could easily be introduced the carbon in the form it was
aised to 1.5% for le.g.) the making of needed, the process is not as illogical às it
azors. seems. The Huntsman crucible process
The process was called the was in use until quite recently, and there
'Cementation Process" and though it was may be a few small plants still operating.
fffective the difficulty was that the carbon But the "cast steel" (or, to be more
iontent varied from bar to bar and even definite, high-carbon steel) of today is
ilong the length of a single bar. To made mainly in electric furnaces, and the
wercome this, Benjamin Huntsman in carbon added directly into the melt, with
(744 found means of melting the bars very sophisticated methods of analysis
ifter cementation. (It will be appreciated used to control the final quality.
hat to make a crucible which would stand The MILD STEEL we use today was
ip to the temperatures needed was very originally regarded as a form of wrought
lifficult at that time), Further, he used iron made by direct removal of carbon, so
y Pig-iron smelted using charcoal as a avoiding the troublesome slag inclusions,
uel, thus eliminating the impurities and the intial patents of Henry Bessemer
| eee and sulphur) arising from were for a "new way of making wrought
fuel. After “cementing” the wrought iron". This he did by blowing air through a
ton in the usual way he melted broken up vesse| charged with molten iron, and so
vars — about 50 Ib. at a time — in special burning out the carbon. Later, the open-
lay crucibles, This took about 5 hours. hearth process, introduced by Thomas and
(he contents of the crucible were then Gilchrist about 1878, effected the
ast into pre-heated, split, cast-iron conversion by the reaction between the
noulds to form bars. Steel made by this molten bath and the furnace lining. Today,
mecess was — and is — known as however, almost all mild or iow carbon
‘Crucible Steel” or “Cast Steel”. By steel is made by blowing oxygen through
farefully selecting the intitial cemented 8 reaction vessel containing perhaps 300
ars — and, as a rule, using pieces from tons of molten iron, brought direct from
lifferent bars made in different heats — the blast furnace. Alloying elements are
tery uniform quality could be achieved, far added (including the essential carbon) as
iuperior to the old so-called “blister required, but directly into the molten bath,
iteel", The cast bars could be forged, rather than as in the cementation process.

11
 SPECIAL STEELS — high alloy, stainless, determines the "nature". Above about
and so on, are made by melting “mild” 1.7% carbon content the metal will be a
steel in an electric arc furnace and, again, "cast iron”. A general distinction may also
adding the alloying elements to the bath. be found between "plain" carbon steel,
Whereas the first model engineers had but which contains as a rule only carbon and a
three ferrous materials available — cast little manganese as alloying elements, and
iron, wrought iron and "steel" — today the "alloy" carbon steels which will include
choice is bewildering, and even cast iron nickel, chromium, and other elements
may be had in dozens of grades and such as vanadium and molybdenum as
analyses. well as carbon. But all steels are basically
For our present purposes, however, we alloys of carbon and Ferrite ("pure iron")
are concerned only with those which are and in many cases the presence of other
ranked as "carbon tool steels". In general alloying elements (especially in "plain"
a “Mild” steel will contain from 0.05% up steels) does no more than modify the
to 0.296 carbon and "Medium" carbon effect of the carbon. Even in the case of
steel up to perhaps 0.6%, These last can the “high speed” tool steels — alloys of
be toughened by heat treatment, but not iron, tungsten, cobalt and carbon — it is
hardened sufficiently for cutting tools. the presence of carbon which provides
"Carbon Tool Steel" will contain from many of their properties. It is a salutary
0.75% up to as much as 1.5% carbon. reflection that our enjoyment of mode!
(Nowadays almost all steel will contain engineering depends entirely on the
alloying elements, manganese especially, material found in the core of your lead
but it is the carbon content which pencil!

Fig. 5A Contemporary (1823) drawing of wrought

fron being "Faggoted" under a water-powered
hammer, with two interested spectators. (Copyright *
— Trustees of the Lonsdale Estate),

12
CHAPTER 2
|

Principles of the Hardening

Process
Hardening high-carbon steel is easy — that really, but we are not concerned with
e have been doing it for thousands of atomic theory, just with hardening steel.
years. Just heat the metal to cherry-red This arrangement is typical of newly
and quench in cold water and there it is — solidified iron, and there will always be
hard! However, like the horse (which is nine atoms, one at each corner of a cube
Mess a noble animal but does not always and one in the centre, at this temperature.
do‘so...”) this does not always work, and Itis called a "Body-centred cubic" crystal.
a little knowledge of what goes on inside Such crystals appear here and there in
the metal will both make failure less likely, the melt, and as new ones form they grow
and also give some idea of what to do on those already there, the little groups
When the occasional awkward job growing larger until they meet their neigh-
pi ars. Those who find that horses never bouring group. Where this meeting occurs
"do so” are always surprised how easy it there will be a discontinuity of the growth,
s when someone who knows about so that when all is solid a look through the
takes over! There is, they will say, microscope will show grey areas with fine
nothing difficult about it — you just need to lines round them — Fig.7. (The crystals
understand the animal. It is the same with themselves are too small to be seen
Steel. Nothing difficult unless you are
actually making the stuff, and that part is á
CUBIC
z^
LATTICE

done for us. You may have a slight

problem with some of the "new words"; €
these | will explain as we go along, but
you will also find a list at the end of the
book to which you can refer if need be.

CRYSTALS AND GRAINS

When pure iron starts to solidify the
atoms arrange themselves in a precise
geometrical pattern, as shown in Fig.6. | CENTRAL
& LJ d
ATOM
n Shown the atoms as little balls for Fig. GA Body-Centred cubic crystal of iron, with ene atom at
elarity — we know that they don't look like each corner and one at the centre of the lattice.

13
 The metal now cools still further, and
when we reach 1392°C an odd thing
happens. The fall in temperature is
arrested for a short while, almost as if
there is a source of heat within the metal.
(Fig.8.) This is, in fact, the case, for there
is a change taking place in the crystal
structure which actually releases heat.
(Metallurgists call this an "arrest point" in
the cooling of the metal). If you had
Fig. 7 Structure of a pure metal. The crystals have “grown”
one upon the other toform a mosaic of “grains”

through an ordinary microscope). These

areas are the "grains". As a rule, the
smaller the grains (within reason) the
€ as /M

better, and part of the job of heat treat-

ment is to "refine" the metal to reduce
grain size. Pure iron (Ferrite) tends to form
rather large grains.

1392
Fig. 9.4 Face-Centred Cubic crystal, with ane atam af wach
corner and one in the centre of each face of the lattice. (Those
atthe back have been left out for clariryl.

suitable equipment you would see the

crystals re-arranging themselves, actually
disintegrating and reforming in a new
910
pattern. |t is still "cubic", but this time
with an atom in each corner and one in
y
the centre of each face, 14 in all — Fig.9.
This is called a "FACE-centred cubic"
crystal. Again, these crystals form
aggregates which meet at the grain boun-
daries. That such a change is possible may
seem surprising, for though the metal is
pretty hot it /s a solid, and you would have
TEMPERATURE
——* to hit it fairly smartly with a hammer to
change its shape. Things would seem very
different if you were one of the atoms!
Even at workshop temperature there
TIME TAKEN FOR A CHANGE would be some distance between you and
OF 10° CENTIGRADE.
the next one and at this high temperature
Fig. B The cooling curve of pure iron, showing the three you would find you had plenty of room.
arrest points" where the rate of temperature change slows
What appears to us as a "solid" is, in fact,

14
 ijostly empty space, even within the at which the arrest points occur is very
crystal; it is the forces which act between slightly different when heating from that
the atoms which give us the sensation of found when cooling.

Not only that. The atoms themselves EFFECT OF CARBON

don't stand still. True, they cannot chase
pout as they do in a liquid or a gas, but The presence of carbon has a marked
do vibrate about their mean position. effect on the behaviour of the metal.
fact, it is this vibration which generates When molten, the carbon is dissolved in
coloured light we associate with hot the iron, as we saw in the last chapter — it
ital). In such circumstances it is not so is the same for molten steel as for cast
prising that some of the atoms can drift iron. One immediate effect is to lower the
bout a little. They do this all the time at point at which solidification starts and, in
h temperatures, and at this "critical" addition, to spread the solidification
temperature they carry out a complete process over a temperature range. The
arrangement. magnitude of this depression and the
| As the metal cools further we meet yet range of solidification temperatures
E e "arrest point", this time at 910?C. depends on the amount of carbon present
This heralds a change back to the original but in the case of (say) a 0.596 carbon
Body-centred cubic shape of Fig.6. This is steel the initial solidification point (the
e "bright red" and about the "liquidus") is lowered to about 1500°C —
temperature we should normally use for about 50*C below that of pure iron — and
forging. Cooling down still further there is the "solidus", when all is solid, is not
found an arrest point again, at 770°C. reached until 1430?C. In between these
there is no change of crystal structure two temperatures the metal is pasty,
e, but this is the point at which the metal crystals and molten metal being
Í can become magnetic. (It is some- present together. (You will have found the
es called the “Curie” point). This same thing with some grades of solder).
ange in magnetic properties can be a Now, the carbon and iron are in solution
| Nioicator of temperature for some when liquid, and this state prevails when
solid as well. |have mentioned this in Ch.|
All of€ changes occur in reverse "in passing" and it does need some
E when the metal is heated, but at the explanation.
points the metal seems to "hang If the iron and carbon were a mixture
" and not get any hotter for a short we should have measurable particles of
While. Again, this can be a useful the two substances uniformly distributed.
aparature indication, It is important, They might, however, separate out, as
how r, to observe that we have been when the pigment in paint settles in the
talking of a gradual temperature change. bottom of a tin. No matter how finely
‘hough the atoms are relatively mobile ground, these particles each contain
when hot they "drift" or "diffuse" rather thousands of millions of atoms. A
than "travel". The changes all take time to "solution" is quite different. Here we have
complete. The presence of other elements the individual atoms forming the disper-
= impurities, or alloying material — can sion. The scale is quite different. It is quite
Sometimes make the metamorphosis feasible — indeed, quite normal — to find
Slower still. Finally, the actual temperature such an atomic dispersion ("solution") in a

15
 taining carbon atoms, as shown in Fig 10.
There is NO arrest point at 1392°C. This
"Solid Soluticn", with face-centred
erystals, is called “AUSTENITE”, after the
metallurgist Sir William Austen, who first
identified it. Grains of crystals are found,
as in the case of pure iron.
Still following our 0.5% carbon steel as
it cools we find an arrest point at about
780°C, where the crystal structure
changes to Bady-centred cubic — you will
see that this occurs about 120*C lower
than in the case of pure iron. The
Fig. 10 Crystal
ofasolid solution
of carbon iniron, the small magnetic change, or "Curie Point" takes
carbon atoms lodged inthe empty spaces in the lattice. place a few degrees lower.
solid — a well-made ice-lolly is an Thus, the immediate effect of no more
example. than 0.596 carbon has been, first, to
The carbon atoms are found actually depress the melting point and to introduce
within the Ferrite crystals (they are much a pasty stage in the solidification, to
smaller than the Ferrite atoms) as shown eliminate altogether one of the arrest
in Fig.10. The number of dissolved atoms points and its associated crystal
which can be accommodated in this way metamorphosis, and to reduce the
does depend on the temperature (as in the temperature at which the others occur.
case of any liquid solution) but it also This may not, at first, seem to be of great
depends on the type of crystal, too. (This importance, but it is the implications of
has an important bearing on the harden- these changes which matter, and they
ing process, as we shall see later). Further, have a profound effect on the heat treat-
although they are situated within the ment. You cannot harden pure iron, but
crystal lattice the carbon atoms can drift even 0.5% carbon steel can be toughened,
or diffuse just as can the atoms of iran, even if not made hard enough to use as a
but this drift can occur at any time, not cutting tool,
just at the arrest temperatures. We shall
see later that the iron and carbon atoms
can also be present in the form of a
“Compound” — iron carbide — but this
need not concern us at the moment.
To return to our newly solidified metal,
the first effect of the carbon has been ta
depress the melting point and to cause a
pasty stage during solidification. Once
solid, the carbon is found in solid solution.
In addition, however, the presence of
carbon inhibits the first crystal form found
in pure iron. The metal solidifies directly
into the Face-centre cubic form of Fig 9,
but with many of the crystal atoms con- Fig. 11 Austenite (shaded) surrounted by grainsofFerrite,

16
IRON-CARBON TRANSFOR-
MATIONS
st Us now follow this 0.5% carbon steel
it cools, but with more attention to
t is happening to the carbon and the
e this time. At solidification the
enite is a homogeneous solution of
substances. However, as soon as the
passes the arrest point at about
80°C and cools further, the composition
of the Austenite slowly changes, until at
about 735°C it contains 0.83% carbon —
he maximum it can hold at this
Fig. 12 The Austenite of Fig. TT has transformed to Pearlite,
emperature. It can only achieve this but the Ferrite remains unchanged. There is some alteration in
chment of carbon by ejecting Ferrite the disposition of the grains.
atoms, and this it does progressively as Austenite from which it came. Under the
thetemperature falls. (The steel only con- microscope these grains have a very
tained 0.5% carbon to start with). This beautiful sheen, resembling that of
Ferrite forms grains or bands surrounding Mother-of-Pearl. For this reason it is
he grains of Austenite, which now form called "PEARLITE". (I am sorry about all
only part of the whole. The overall analysis these names, but there is a glossary at the
the steel will still show 0.5% of carbon, end of the book if you find them difficult to
t it is all concentrated in the Austenite remember). The overall composition of
ins. Fig. 11. Pearlite is 0.83% carbon still, so that as
T Just below this temperature, at 730°C, the steel contains only 0.596 we should
we find another "arrest point", this time expect the Pearlite grains to be surroun-
due to a change in the relationship of the ded by grains of Ferrite, and this is, in fact
and Ferrite in the Austenite. Above the case. See Fig 12.
This. recombination of Austenite into
7. o*C it actually combines with some of Pearlite occurs in a// carbon steel when
the iron to form /ron Carbide. This is FE,C, cooled slowly. Maturally, the amount of
Pearlite will depend on the carbon content
afd indeed and is given the name initially, but it is always there in greater or
CEMENTITE'", because it was first less proportion. Being a combination of
identified in steel made by the Cementa- soft and ductile iron (Ferrite) and very hard
llon process — see page1 1. Cementite, it is very tough.
_ This Cementite contains much fess
Ferrite (iron) than the Austenite from THE IRON-CARBON DIAGRAM
lich it was formed, so that there will be
Some Ferrite left over. In the event, the If we were to look at steel with a different
vementite (or iron carbide) crystallises carbon content we should find similar
nio very thin plates, with similar thin changes taking place, but at different
lates of Ferrite sandwiched between temperatures. For convenience, engineers
t , this assembly of plates forming a and metallurgists assemble the data on a
"Brain" corresponding with the grain of chart, called the "Iron-Carbon Equilibrium

17
18
=
31/8033 (NONU
GNOS)
34IN31Snw

(NOTOS
ryt
|
31!11N3w325
NONI) 30/85v2O
{
NN Er
3L
Bv

IONYHI
DNISSOUD
LY
"cc /Jv2U|H2 H3ddn
w SNISSOND LY 30NYHÓ
—c- -— HaEMOV VJA
H3ADT 38d = lw IVILI

Je BeNiveaenal
"i wOduv2s

514
` WEL perduussar
jo aui UOI, VOJO uimuqiinog
wesdarg sy), DUINA,
sy pue adl aH pitia
JO (wars ASENON
dg IIT
Diagram”, which shows the effects of
oth carbon and temperature. | have
made a simplified version of this in Fig 13:
this deals only with "steel" (the full
diagram includes cast iron as well) and |
have not bothered with the temperatures
up in the melting range. Those sufficiently
interested will find the complete diagram
in most books on the Metallurgy of Steel,
but there is a very clear one in
Encyclopedia Britannica, in the section
E "Iron and Steel". This can be
looked at in most public libraries.
| | have shown the 0.596 stee! we have Fig. 14 Grainy of Pearlite surrounded by borders of fron
just considered as a vertical line. The Carbide (Cementite),

upper arrest point, or critical temperature, through the critical. We now find grains of
isshown by the line ABE, and the lower Pearlite surrounded by areas of
by DBE. You will notice that these Cementite, (Fig. 14) in contrast to the
coincide at "B"; we will have a look at this Pearlite and Ferrite of the lower carbon
) 8 moment, but before doing so, let us steel.
Ee at a typical tool-steel, with a carbon Now look at steel with 0.83% carbon.
content of (say) 1.2%. | have shown a You will notice that the two lines repre-
vertical line here also. We start with senting the upper and lower critical
Austenite as before, but this time it temperatures (the "arrest points") now
Contains more than the 0.8396 which is coincide. It is, after a fashion, a “Eutectic”
the maximum which can be held in solution, similar to that found in solders
solution at the lower critical temperature. and brazing alloys, Both the crystal trans-
(This is about 720°C in this case). As the formation and the metamorphosis of the
Steel cools below the upper critical, which Austenite occur at the same time, and the
is at about 900°C, the Austenite finds Austenite changes to Pearlite directly.
itself too rich in carbon this time, whereas This "all Pearlite" steel is about the
the previous example was rich in Ferrite. toughest that can be had with a plain
So, as the temperature falls carbon is carbon steel. You will note that as a con-
Progressively rejected, but it is rejected in sequence of this "Eutectoid" at 0.8396
the form of iron carbide (Cementite) for, carbon the upper critical temperatures for
unlike the case of cast iron, free carbon the various steels vary considerably, along
Cannot exist under these conditions. We the line ABE. This has an important
find, therefore, that the grains of Austenite bearing on the temperatures needed to
&re now being surrounded by first, streaks, effect the annealing of the steel. The
&nd then grains, of Cementite: there is no lower critical temperature, DBE, does vary
free Ferrite available. By the time the steel a little, but in many published versions of
has reached the lower critical temperature the diagram it is shown as a straight line.
(the line BE) the Austenite has again The difference is only a few degrees.
Settled down with 0.83% dissolved | have shown on the diagram a number
Carbon, as before. And, as before, this of little sketches indicating the grain for-
transforms to Pearlite as the metal cools mation — actual micrographs would be

19
 Fig. 15 The effect of carbon
content
on the mechanical proper-
ties ofannealed
steel,
TENSILE STRENGTH

0.2 0:4 0:6. 0-8. 1:0. 1:2.

‘te. CARBON.

rather confusing and are difficult to "hot rolled bar’ which has (or should
reproduce. However, by examining a have) been normalised when you receive
specimen through the microscope an it from the merchant. Cold drawing (e.g.
experienced metallurgist can easily "Bright Drawn Mild Steel") will not alter
recognise the various constituents of the type of structure (Pearlite-Ferrite, or
Pearlite, Ferrite, and Cementite, and by Pearlite-Cementite) though it may alter
noting the proportions of each can make a the mechanical strength, and especially
fair estimate of the overall carbon content the ductility, because the farming process
of the specimen. has distorted the grains. The structures
To sum up: we would expect to find found when the steel is cooled too quickly
very little Pearlite and a lot of Ferrite in a for the changes to occur are quite diffe-
low carbon (mild) steel, with the propor- rent. So, let us now have a look at the
tion of Pearlite increasing as we approach effect of other than slow cooling.
0.83% carbon. Beyond this the Cementite
would begin to appear, increasing in
proportion as the carbon content rose to EFFECT OF COOLING RATE
the maximum associated with a "steel"
(about 1.796 carbon). We would expect Let us again consider our piece of 1.2%
the "toughness" and strength to increase carbon steel at (say) 800°C, which
as the Pearlite proportion increased, up to temperature has been reached either by
0.83% carbon. Thereafter the steel could slow cooling or slow heating. The
be expected to get harder, with some loss Austenite will have partly transformed,
of ductility, This is, in fact, the case. Fig and there will be Austenite grains sur-
15, But | must emphasise that these con- rounded by some Cementite. Within the
siderations all apply on/y to steel which Austenite most of the crystals will contain
has been both heated and cooled one or more carbon atoms — the "solid
SLOWLY, so that the little atoms have solution" condition. Now let us reduce the
time to find their right places, and the temperature very quickly. There will be no
chemical changes at the lower critical line time for the atoms to rearrange them-
have time to complete. It is typical of a selves, no time for crystal transformation,

20
 instead of finding themselves in a relates to no particular steel, but it is
ttice formation which leaves them typical of most; every steel specification
jJenty of room the carbon atoms are has its own "S-curve" like this. It is impor-
pped inside the crystals with inade- tant to realise that, so far as we are con-
quate space for them. In addition, the cerned, it must be used only to Mustrate
Austenite is compelled to retain far more the effects of time/temperature changes.
Carbon that it can normally hold at the Detail interpretation of the diagram is a
lower temperature. The Austenite crystals matter for experts.
put under enormous internal stress, Looking at Fig. 16 you will see that
this is just the condition which is there are two "S" shaped curves. On the
sociated with hardness. In addition, the left of the line ABC we have a zone where
Cementite which is present is unable to the steel is Austenitic (in a state of solid
Grystallise normally, but instead is con- solution) but it is "unstable" — it has not
Strained into a needle-like structure yet transformed but is ready to do so. On
(known as an "Acicular" formation) which the right of the line DEF we have the steel
is very hard indeed. In fact, it is these in its final, totally transformed state. In
"needles" which help in the formation of a this area, if we were talking about the
C cutting edge. This new structure — 1.296 carbon steel there would be grains
o stressed crystals associated with of Pearlite surrounded by Cementite: if it
acicular crystals of Cementite — is given were the 0.5% steel we looked at earlier
the name MARTENSITE, and is the basis there would be grains of Pearlite surroun-
of all hardened carbon steel. ded by Ferrite. In between the two "S"
_ | have already referred to the need to curves the structure is undergoing the
heat the metal slowly, partly to ensure transformation process and will consist of
iat the various transformations have time a mixture of Cementite (carbides), Pearlite
to occur, but this is also necessary so that and Ferrite and, of course, some of the
we can be sure that the metal is hot right original Austenite — carbon in solid
through. It does take time for the heat to solution, The nearer to the line ABC, the
travel through to the interior. A similar more Austenite, whilst close to the line
consideration applies when cooling. When DEF more of the other constituents.
cooling rapidly ("quenching") what about At the bottom of the "S"s two horizon-
metal in the middle of the workpiece? tal lines are shown dotted. At one, marked
ie outside has cooled fast, it is true, but “Ms” the hard Martensite starts to form,
is clear that (with large specimens and below the line "Mf" this change is
especially) the core may not have kept up complete; the metal is all in the hard con-
With the cooling of the exterior. What is dition. For "silver steel" Ms lies at about
the effect on the metal? We can best 150°C However, Martensite cannot exist
answer this question with another to the right of the line at “F”.
diagram, Fig. 16. This shows temperature Now, consider a piece of high-carbon
On the vertical scale, and time on the steel at the point "Q'. It is at about
horizontal. Because metallurgists and 780°C, If it cools slowly it will follow the
engineers need to examine cooling rates line QP, taking perhaps 24 hours to cool
lich may last hours or days as well as down to 15?C. The line passes through
which take only the odd second or both of the "S" curves, indicating that
$0the time scale is compressed — it is, in total transformation to Pearlite and
E "logarithmic". The diagram shown Cementite has taken place. The material is
|b
bi
21
"annealed" — the steel has had time to go has taken place. This sample is
through all the changes we talked about "hardened". These are examples of the
earlier. Now look at the line QR. The metal two extremes, slow and fast cooling.
has cooled through an identical You will see on the diagram a line
temperature range, but this time in no "QS", which JUST touches the nose of
more than one second. At no point does it the left-hand "S" curve. This is the
cross the. "S" curve, so that no transfor- slowest rate — the "Critical" rate — of
mation to Pearlite is possible. But it HAS cooling which will achieve a fully
crossed both the Ms and Mf lines, indicat- hardened Martensitic condition. (The
ing that total transformation to Martensite location of the point "S" will depend on

Fig. 16.4 Time-Temperature Transformation Diagram” (S«Curve ] for a high-carbon steel Note -this is a typical curve only;
the shape will vary from steel to steel.
A. B

mee
p

at
700,
wo - — -— "c

5 RANSFORMATION ENDS
ARLITE|& CEMENTITE.

|
Se
i o

(Pe e
TEMPERATURE.

O51 2 5 10. 20. 50. 100200 500 1f. 12. 1.

22 Seconds. Hours.
the type of steel; it might typically be 2 or not too good in this respect. But NO thick
3 seconds). The quench shown by the line section of steel can be hardened right
"QU", however, does cross the "S" curve through, even when alloys are added. For
ABC, at "V" and "W", in and out again. "Silver Steel", which contains a little
Between "V" and "Ww" some of the Chromium to help matters, the limit is
Austenite will transform to Pearlite, but as 5/16 inch thick, though the loss of
the cooling curve only lies to the right of hardness at the centre of a piece } inch
the line ABC momentarily the remainder thick is very small indeed. This "Limit to
will carry on down QU to form Martensite. through Hardening” explains the
The final state will, therefore, be mainly difficulties sometimes met with when a
Martensite but with a little Pearlite as large section carbon steel tool is ground
well. It will not be as hard, BUT, the down at the point to bring it to centre-
ence of the Pearlite will make it less height in a small lathe, as in Fig. 17. The
jrittle and a bit tougher.** point of the tool is now located in the
This diagram helps us to understand centre of the section and will not be as
what can happen in the centre of a piece hard as the original. The loss of hardness
of steel during the quench. If it has a small will be quite noticeable if a half-inch tool
cross-section, and is cooled along some is ground down to. 5/16 inch point-height.
line such as "QR", then the centre of the The cure is to anneal and reharden. See
section may follow a line between "OR" p.37.
and "QS", and will be fully transformed
fight through: it will be "through DISTORTION DURING THE
hardened". If, however, it was a very thick QUENCH
piece the centre might very well be cooled
along "OU" even though the outside As well as hardness we are frequently
followed “QR”. The centre of the concerned with the shape of the
Specimen would not be as hard as the workpiece, especially when making
outside — though it may well be tougher. gauges or form tools. A little thought will
One of the purposes for which some of suggest that there is likely to be a dimen-
the alloying elements are added to sional change with all the transformations
modern "carbon" steels is the improve- within the metal and that, in addition, the
ment of the "through hardening" highly stressed crystals will “deflect” just
characteristics; the alloy modifies the "S" as any stressed member will under load.
curve. A plain, unalloyed carbon steel is Not only that — the very act of cooling
causes a contraction and in a “quench”
**Note that it is not possible for any but an this contraction is sudden in the extreme.
expert metallurgist to determine HOW MUCH If the workpiece is not exactly sym-
Pearlite will be formed. metrical there is risk of uneven shrinkage.
FULLY QUENCHED^ 100% MARTENSITE
ORIGINAL. HARD
pe

Ur.
Fig. 17 1 a /argo-section =
Carbon steel! too! is ground |
fo suit e smaller Cathe k
thenew cutting edge willnat
beas hardas the original, PARTIALLY wk SOME PEARLITE
SLIGHTLY SOFTE

23
The higher the cooling rate the greater the properly heat treated, will have reached
risk. the maximum available hardness, but it
In cases where "shape" is important will be very brittle. As | have tried to
we must use a slower cooling rate, even explain, the crystals are under consider-
though this may mean that the quench able internal strain and any shock loading,
line for the centre of the work may cross or even rough handling, may cause
the nose of the "S" curve momentarily. fracture. This may apply occasionally even
Fortunately this is not so important in with oil-quenched steel, In addition, the
articles not intended for metal-cutting and grain size may not be as favourable to
in some cases can be an advantage, clean cutting action as we would like,
expecially for tools subject to shock especially if the initial temperature (point
loading. The risk of cracking is not great "Q" of Fig. 16) was not exactly right. If,
when dealing with (say) a simple '"'D''-bit, however, we now reheat the steel to a
but can be acute in the case of a milling- relatively low temperature these crystals
cutter, with sharp corners at the tooth can be, to some extent, "stress-relieved"
roots. In such a case it is prudent to and if the temperature is held for a little
sacrifice a little hardness in order to be while some degree of grain refinement
sure of getting our tool in one piece, and a may be had as well. For plain cutting
slower method of quenching must be tools, where shock loading is minimal, a
sought. very low tempering temperature will serve
In industry special methods are used, — the usual recommended is about
not necessarily beyond the capabilities of 230°C, though this did refer to turning
the model engineer, but they do need tools one or two inches square — | shall
proper temperature control. The simplest have more to say about this later! It is, in
method for us is to quench in o// instead of addition, prudent to temper the tool-shank
water. The risk of cracking and distortion even more, so that it does not crack under
is considerably reduced yet the loss of the clamping forces of the tool-post.
hardness is not great. If, however, distor- Tempering can serve another purpose,
tion is of the first importance, as when though. We have seen that the quench
making a gap gauge, then it is wiser to produces Martensite — highly stressed
use the special oil-hardening tool-steel — Austenite crystals interspersed with
"Ground Gauge Stock". (Though it can needle-like Cementite. For some
also be obtained in black bar as well). This purposes, however, we need high strength
material is alloyed so that oil-quenching and resilience rather than hardness as
always misses the knee of the "S" curve, Such — the classic case being the spring.
and full transformation is obtained. It is There is no reason at all why a spring
not as hard as a water-hardened straight should be HARD; the engineer expects
carbon steel, but is adequate for the them to have a high yield stress and high
purpose. resilience. Industrially it is possible to
achieve this directly but this does need
TEMPERING special equipment. (And, in any case,
commercial springs are seldom made
Just as the "Good Shepherd" tempers the from straight carbon tool-steel). The only
wind to the shorn lamb, so the prudent way open to most of us is to harden right
tool-maker tempers his hard steel against out (preferably in oil) and then "temper
shock loads. A high-carbon tool steel, back" to the desired condition. We reheat

24
 haps, 300°C - "blue". At this alloying elements. The diagram (Fig.13 on
erature the needles of Martensite page 18.) is drawn for "plain" carbon
; into little "nodules" and, of steel, but each alloy will have a different
the crystals are completely stress characteristic equilibrium diagram, some
The steel will be “hard” relative very complex indeed.
eel, but nowhere near as hard as Almost all carbon tool steels these days
tool. contain about 0.3596 of Manganese, its
up, we can modify the structure main purpose being to counteract the
ardened steel to suit the purpose effects of the impurity Sulphur. At this
in mind by a secondary heat level the effect on fig. 13 and hence on the
nt at comparatively low heat treatment is negligible. When
atures. | shall be covering the present in higher amounts — 196 or more —
needs for various applications later, the steel needs special treatment. A few
pu should bear in mind that tool-steels ("Silver'" Steel is one — it
g' serves three purposes: to contains no silver, by the way; the name
ieve the "tight" crystals and so refers to its appearance!) include about
e the risk of cracking: to improve the 0.5% of Chromium, partly to help in
so that you can get a keen refining the grain size but also to improve
dge; and, in some cases, to trans- "through hardening”. At this level there is
Martensite to a nodular condi- some slight effect on the shape of the
improve resilience. | shall, in a diagrams, including the "S" curve, but no
pter, deal with this in more detail, special treatment is needed. The-
in relation to the special needs quenching temperature is slightly diffe-
turner's. tools, which must be rent, that is all. These steels are
Pup to a very fine polish and sharp essentially "carbon" steels with additives
to improve performance.
As the alloy content rises, however,
LOY STEELS and, more especially, when a combination
of ailoying elements is used, then the
ity-odd years ago, when the equivalent diagrams to Figs. 13 and 15
d Practices" of model engineer- become very complicated indeed.
being established, carbon steel (Chromium, for example. forms a very
t that — an alloy of carbon and hard carbide on its own). When faced with
rue, it would contain à few a steel containing mare than about 0.6%
s, for the days of charcoal- of any of these elements (and they can be
iron were over (though Nickel, Chromium, Cobalt, Vanadium.
el was still using Huntsman Tungsten, Molybdenum, Titanium and
tool-stee| made from imported others) the answer to the heat treatment
a charcoal iron) and typically problem is, for the model engineer,
sulphur and phosphorus was “DON'T”! A nice piece of 3% nickel 0.3%
m l. Today, however, almost all steels carbon steel will (if you are sure it is
ia n other elements as well, We have EN21) make excellent connecting rod
seen how drastically the bolts, but unless you have the exact heat
stics of pure iron are altered by treatment data it is best left in the “as
on of quite small amounts of supplied” state. Even more so if you come
d the same applies to these across any of the higher alloyed steels —

25
and even more so again if it has come loco on an 18-inch tool-room lathe, no
from "Evans the Scrap". Heat treatment matter how accurate it was.
of such, without proper data and equip- The answer is to use a mild or medium
ment, is likely to be disappointing if not carbon steel, of specification appropriate
catastrophic. (However, | shall, in a later to the loading conditions, and then to
chapter, deal with the heat treatment of modify the surface layer of the metal so
some of the medium carbon steels, which that it can be hardened. You wili
can, with care, be treated quite remember thàt the earliest method of
effectively). Finally, don't forget that manufacture of high carbon steel involveg
brazing, and to an even greater extent, its heating in the presence of carbon. This
welding, is a form of heat treatment! took a long time, but if we carry out the
same process for a short time then the
carbon will penetrate only a shoti
SURFACE HARDENING distance. For models we only need a thin
layer — perhaps as small as 0.005in. would
There are many situations, both in full-size serve — and this is what is achieved by
and model engineering, where a hard- "CASEHARDENING". There are various
wearing surface is needed but the compo- ways in which it can be done, some which
nent also must resist shack loads — or don't involved anything more than a blow-
perhaps repeated reversal of load, almost lamp and some special compound, but all
as bad. Cross-head and gudgeon pins are are quite practical for the model engineer
examples. The usual suggestion made by Once this surface layer has been car-
writers of articles in Mode! Engineer and burised the part can be heated and
elsewhere is to use Silver Steel, harden, quenched and, in most cases, does not
and then temper back. This is not a very need any tempering. The result is a core
good solution to the problem. First, the which is strong and tough, and a surface
tempering reduces the hardness of the which is almost glass-hard. Further. the
wearing surface and, second, tempered process is very flexible, and it is possible
Martensite is not the best material for a to harden selectively, and even to carry
load-bearing component. Horses for out machining operations on one part of a
courses! A Bugatti type 37 does not make component after another part has been
the best of town carriages, and you would surface hardened. | give more details o!
be hard put to it to make an “O” gauge the procedure on page 64.

26
 chapter | shall deal with the heat treat it it must be for some fairly
ctical aspects of Heat Treatment, but | important purpose.
aving detail discussion of the actual However, there are cases where metal
g devices till later. For the purpose of more or less known provenance can be
s chapter, therefore, | shall assume recycled, and if you are in a tight spot and
'our source of heat is, like the engine cannot get new steel, then the following
r Rolls-Royce, "adequate for its table may help in suggesting the probable
jse". | am going to leave the con- carbon content of old tools etc. |
of Tempering to the next emphasise the word "old", for these days
too, as this is really quite a even the humble cold chisel may be made
from a non-tempering nickel alloy.

% Carbon Type of Tools

»cessity is to decide what material you Crowbars; Pickaxes; Screwdrivers
fhardening. If you are not sure about 0.8 Large Masonry chisels; Quarry
if it is just a "piece of good stuff" rock-drills; Wood-splitting Wedges
ave picked up from Evans the Scrap 0.9 Cold Chisels; Shear-blades:
¡my advice is "Don't". | have by me Blacksmith's hot setts:
1.0 Smaller Cold Chisels; Old wood-
T be found in a scrap motor-car working machine cutters;
3 days. If typed out on paper 7 inches Handsaws.
the list would be several feet long. 1.1-1.2 Engineer's cutting tools; Drills;
than one fifth are "plain" carbon Reamers etc.
teels — all the rest are more or less highly 1.2-1.3 Files;
byed — and even if you knew that your 1.3-1.4 Razors; Engraver's tools; Saw-files.
tece was carbon steel it had probably 1.4up Wire-drawing dies; Cold Saws.
n heat treated already and would
1 8 proper annealing before reharden- Many of these could be made from
NG. The chances of finding any steel safe high-speed steel, but this can easily be
šat treat on a scrap-heap these days distinguished by using the "Spark Test" —
Slim. Far better to start with a new see the Mode! Engineer's Handbook. In
še of stock; after all, if you are going to any case, however, the material should

27
first be annealed (See page 71.) and a long’? You have both to make sure tha!
small test piece then sawn off and test- the metal is hot right to the centre, AND
hardened with the estimated carbon give those atoms time to drift to their new
content as a guide. spot. Fortunately in this case the answe:
Heating We have already seen the need is quite definite; metallurgists and genera.
for slow heating, and | cannot emphasise tions of blacksmiths over a century or
this need sufficiently. The question is more have established the rule: heat for
"How slow?" If put into a cold muffle ONE HOUR PER INCH Or
furnace and heated up with the muffle, THICKNESS, once the tool is up to the
this will suffice. With the average muffle, hardening temperature, This means that
in fact, the work will heat up slowly dinch thick tool should be heated fo;
enough if put into the furnace when it is about 20 minutes. Yes, | know! You have
already at the hardening temperature. (In never held the heat for as long as this! |
some books you will find this called the am sorry, but that means that you have
"Austenising Temperature", by the way). never achieved the maximum possible
If you are using a molten lead or salt bath hardness, either! Just to persuade you |
then it is necessary to preheat the work a have set up specially and hardened four
little anyway, as | explain later. But with pieces of carbon steel from the same bar
the ordinary blow-lamp or gas torch some holding the temperature for one, five, ten
care IS needed. It is all too easy to and fifteen minutes, the last being “right
overheat thin parts locally. | find | take for the }-inch stock used. The hardness
about five or six minutes to bring a small measured on my "vintage" Shore
tool — about $ inch x 3 inch long — up to Scleroscope came out at 74, 76, 78, and
the arrest point, and it takes a little longer 81, respectively. Those who know their
to climb the remaining 60*C or so. This hardness numbers will realise that these
gives good results, and can be taken as a figures are all low — but equally, they will
guide. Better to be too slow than too fast. realise that the "Shore" is not suitable for
At the arrest point there is a temptation to such small specimens. A subsequent tes!
turn up the gas, as the metal seems to be on the last speciment showed a hardness
making no progress. This is just the wrong of above 900 Vickers, and this is abou!
thing to do! The little atoms are in process right. The reduction in hardness for the
of moving house, and won't welcome shorter heating times is quite evident -
being hurried over the job! Keep heating the "one minute piece" is softer “as
ata steady rate and when the metal starts hardened" than the tool should be when
to colour up further you are very nearly at tempered.
the right temperature. Overall, the short It is quite true that our tools are
answer is that you must use your judge- relatively small compared with those
ment, and try to get as much experience formerly used in industry — or today, for
as you can, if you rely on torch or kitchen that matter. They don't contain muc"
fire heating. For the odd D-bit that will be metal — but their surface area is small, too
used once and then scrapped the heating and the RATE of heat transmission
rate may not be all that critical, but you needed to get even a i-inch too! to full
MUST pay attention if the tool is an temperature at the centre in 15 minutes i5
important one, likely to be used for years. very high. Clearly for the odd scriber-poin!
Once up to temperature you must hold it isn't all that important, but for cutting
the heat for some time. Again — "How tools it does matter. | shall, later, be telling

28
 »fitout the use of a molten salt-bath These temperatures are for quenching
gating, and one of the great advan- in water or brine. For oil-quenching the
ages of this type of furnace is that the metal is best heated perhaps 10-20*C
aking time can be reduced; the rate of higher, but | do not advise going above
it transfer is much better. Setting the 800?C, FILE STEEL, usually 1.2596
cin a cold muffle furnace and letting it carbon with 0.5% chromium, gives
eatqe with the furnace is good practice, maximum hardness from 800°, but it
p, for the muffle heating rate is about must not be heated much above this
e same as that for the metal. You can figure.
alve the heating time in this way. In all cases it is much better to obtain
m Right Temperature. At first sight it the correct quenching temperature from
might be assumed that we need to take the manufacturers or the stockists if the
e metal up to above the upper critical best results are to be obtained. You will
emperature — the line ABC of Fig 13, — not go far wrong with the figures given in
ir‘this is not the case. First (for high the table, but as is always the case
b tool-steels, that is) we do need "Perfection demands more care". If the
"Free Cementite" when we quench, EN No., or the newer numbers from BS
d this does not appear until the metal 970/1972 (which, incidentally,
. fallen below the upper critical incorporate the carbon content in the
erature. Second, experience shows specification number) is known, then
ithe grain size is more favourable if the reference to the British Steel Corporation
| is quenched from a lower may produce the required information.
perature. As a general rule steels (Look in the telephone directory, or try
having
av -1% carbon content or more are BSC, Swinden House, Rotherham $60
juenched from below the upper critical 3AR). If the SAE number (Society of
and those below 0.9% from above it. This Automotive Engineers of America) is
$a "general" rule — there may be special known, then data is given in their
; for departing from it — and one handbook, which should be available in
es is that the quench the local reference library. This gives
erature does not vary much over the almost ALL steels made to U.S. specifica-
e range of carbon content. Note that tions, and they don't differ much from BS
E treating OTHER than for pure har- 970; many have the same number. In the
ng this may not apply— see Ch.VII. case of the oil-hardening gauge steels the
T ə following table is drawn up from a proper ternperatures are given on the
number of "authorities" for steels all of wrapper.
Which contained about 0.35% manganese As a final point of comfort, these
= What today is regarded as a "straight" temperatures are not critical to a few
steel. The exception is the "Silver degrees; most steelmakers give a range, if
*, which contains also about 0.45% only because there is a tolerance on the
-hromium. actual carbon content. If you are within
2arbo , 96
Temperature, °C 10°C you will not be far wrong,
07 ~0.85 790-800 PROVIDED you hold the steel there for
86—1.05 770-780 the proper length of time.
1.06-1 25 770-780 Judging the Temperature. This is what
Above 1.25 760-770 separates the men from the boys! The old-
Silver Stee!" 770-790 timers could estimate almost to a degree

28
just by looking at the steel, but they were (Have a piece of clean steel set in place —
hardening carbon steel all day long. If you don’t rely on the appearance of the
have a gas or electric furnace which is furnace chamber), There is no substitute
fitted with a pyrometer you are in no for this type of experience. However, to
difficulty — provided that you check the help you | have taken some colour
accuracy of the instrument now and photographs of a chunk of steel in my own
again. (If it has no pyrometer | give some muffle, and the printer has done his excel-
hints on making one on page 107.) With lent best to reproduce these on page 33.
gas or fire heating we have to rely on the As a further guide, | find that the larger
colour of the metal. Now, make no boiling ring (one of the spiral type) on our
mistake; this is an EXACT measure of the Creda SOR electric cooker reached and
temperature — in fact, one form of stabilised at 780°C after running full on for
pyrometer makes use of the colour. The 15 mins, with no pan over it; and the
frequency of vibration of the atomic firebed of aCourtier Type BR closed stove.
Structure is a direct function of the burning "Phurnacite" (the egg-shaped
absolute temperature, and this frequency things) lay between 1000 and 1100°C
determines the colour of the light which is after about 15 minutes drawing up the
emitted, The problem lies in how to fire, the firebed being about 7 inches
describe the colour. The birds in my deep.
garden probably would not recognise your There are a few other guides. The lower
cherries as being red at all! And, for that critical temperature is an "arrest point" at
matter, some readers may have blue which the colour will seem to hang back.
blood in their veins! Only experience can This is an indication that you are within
give you a proper judgement. For what it about 50°C of the required temperature.
is worth, generations of blacksmiths over (In passing, provided you are above this
a thousand years or so have adopted the arrest point you will get some hardening
following “names” for the colours as seen effect, but none if you are below it) Very
in a relatively dim light. (But NOT, on any near to this arrest point — very slightly
account, under light from a fluorescent below — lies the "Curie Point" at which
tube, which makes a pig's ear of any the steel ceases to be magnetic. You can
colour judgement). keep checking with a magnet or, as | do.
use a small compass. Be careful, though:
700*C Dull Red neither magnets nor compasses like heat.
750*C Blood Red and there is a risk that the magnet may
800°C Cherry Red pick up a small workpiece and refuse to let
825°C Bright Cherry Red go!
850°C Red There are some "Temperature Indicat-
900°C Bright Red ing Crayons" available, of two types. One
1000*C Yellow-red changes colour when the temperature
marked on the packet is reached; the
The best thing to do is to find someone other is "stroked" over the work and if it
who has a muffle furnace with pyrometer, melts within two seconds the temperature
or arrange a club visit to the local is as indicated; if longer it is too low, and if
Technical College, and have the furnace shorter, too high. Both are pretty accurate
brought up to the various temperatures so provided they haven't been in stock for
that you can see what it really looks like. years, but they do need a little practice in

30
 sg. Mine are made by the "Markall can be reduced by setting a few pieces of
Sompany" — an American firm — but dry charcoal at the mouth of the muffle,
s are available from Bayer Chemicals just against the door. This will consume
id, Richmond, Surrey, or from T.P. any oxygen in the muffle — or most of it —
land & Co. Ltd, Rose Kiln Lane, and reduce that available to oxidise the
ading, Berks. For use in muffle furnaces metal. (Don't use coke or coal).
hout pyrometers the well-known Alternatively there are some "Anti-scale
sager Cones” (Wengers Ltd, Etruria, Paints" available. These are used by
e on Trent) can be used. These curl ceramic enamellers, but are equally effec-
yr at the top when the right tive in protecting steel. It is painted on the
nperature is reached, and are used in work and is effective up to around 800*C.
kilns, where careful temperature | use it very successfully, but also use a
ntrol is equally important. They have the substitute. This is ordinary whitening
advantage that they can be used only (powdered chalk) mixed with water or
once, so that you need them by the methylated spirit and painted on. Some-
times | bind iron wire over the top, and
‘sense of proportion must be main- clart that up with the paste as well. (The
d in all this, as in all aspects of heat wire must, of course, be pulled off before
ent. If you make the odd form tool quenching) This is effective also. Both
nch once or twice a year, then the have the disadvantage that the paint or
t on page 36 showing the heating chalk tends to ""incandesce" and so
rs will meet all your needs, And, for prevents proper colour judgement. No
|portant jobs, provided the matter in a muffle, but fatal if using a
ature lies between the two critical torch. The answer to this is to have an
(i.e. for Silver Steel, between identical piece of scrap steel (it need not
"C and 830°C) you will get SOME be carbon steel) alongside and provided
ardening effect, though not very good at you ensure that both the workpiece and
Bi extremes. If you harden tools fairly fre- the test specimen are equally heated the
ly then you will very soon develop latter can be used to judge the colour.
iecolour sense which will enable you to Horologists coat their small parts with soft
r k by eye; in any case, you have to soap when heating. This is said to avoid
je the colour under your workshop scaling too, but | must confess | have not
tions, not those of an industrial heat used it myself.
ent shop. But, if you make all your
n tools and gauges, and are hardening
mac.shinery parts as well, then you would QUENCHING
ohWell advised to look out for a second-
ind muffle-furnace (or even a new one!) There are three quenching mediums in
orlook at some of the other arrangements common use; plain water, brine, and oil.
detail in a later chapter. Water is cheap and easy to obtain, and
ling. This is of little importance in a will serve almost all the model engineer's
tool, as it must be ground in any needs. Brine gives a somewhat better har-
but if it is a form-tool, which has a dening effect and can give more even
le which cannot be formed on the cooling — the results are more uniform. Oil
grinding wheel, it can be a nuisance, if not is essential when quenching complex
a difficulty. If using a muffle the scaling shapes — the slight loss of hardness is

31
accepted as a fair exchange for the fluids) that water would take about 3j
reduced risk of distortion and cracking. seconds to reduce the temperature of a
The actual cooling process with the standard probe from 800°C to 420*C,
three quenchants is different — it is not and oil about 5 seconds - no great
just a case of difference in degree of difference. But whereas the water-cooled
quench. With plain water the initial effect probe only needed a total of 5 seconds to
is the generation of steam all over the fall to 200?C, oil took 25 seconds. This
workpiece. This causes an intense and means that both fluids bring the stee!
sudden cooling effect, as the latent heat of through the lower critical temperature at
vaporisation of water is very high. about the same rate — oil slight slower
However, this results in the workpiece both are rapid enough to prevent the
being blanketed in a steam jacket and transformation to Pearlite. However, the
steam is a very poor conductor of heat. much slower rate of cooling in the lower
The cooling rate would drop dramatically reaches — below 400°C — with oil means
if this insulating coating were not that the transformation to Martensite
removed, So, the work must be agitated takes place in a much more leisurely
in the bath; if thís is done, then cooling fashion, and this reduces the risk of distor-
proceeds at a fairly rapid and steady rate tion and cracking considerably. The Mar-
until the approach to 100°C, (As we shall tensite will form anyway, even if we air
see later, once the temperature has fallen cooled from (say) 200°C. This test was, of
to about 300°C the rate of cooling is not course, under quite artificial conditions
important). the metal would not be properly hardened
In the case of Brine a similar procedure even with the water — we need cooling
takes place, but the presence of the salt in rates of thousands of degrees/minute
the water seems to retard the formation of through the critical range for that — but it
the steam blanket. For some reason the does give an indication of the different
cooling appears to be more even but it is behaviour of the two fluids. Agitation wil!
faster — about twice as fast as water down reduce the time taken, but the character
to 300?C or thereabouts. Brine is of the cooling curve will be similar.
invariably used for "water" quenching For mode! engineers, therefore, we can
steels in industry, and the figures quoted say that in general water will serve for our
for hardness in commercial specifications needs, but brine is preferable when
all assume that the work was quenched in ultimate hardness is needed. Oil should be
1096 salt/water brine. used for milling cutters and for gauges or
Oil behaves quite differently. hardened jigs — preferably using the
(Assuming a proper "Cold Quenching Oil" proper oil-hardening steel for the latter, as
is used. "Motor Oil" will have quite it is formulated to avoid distortion. The
unpredictable results). It is slower in the brine is made from "Vacuum-dried Salt" —
initial cooling, because oil has à much the coarse sort used by farmers for butter:
lower latent heat than water. Once the cheaper than domestic salt, and without
"vapour blanket" starts to form, however, the additives, some of which may not be
the cooling rate (with the same degree of desirable for a quench-bath. An 8 to 1095
agitation) is very little different from that solution is general — 12-160z/gallon or
of water. Below 400°C, however, the oil 80-100gm/litre. In use, some of the water
cools the work very much slower. Using a will evaporate and this must be made up.
test rig, it was found (with still, unagitated An old test was that a fresh potato would

32
 725°
Indications of colour/ temperature using a 3x 4 in. steel plate photographed in
the author's muffle furnace.
A difficulty with printing à colour reference is that the result is reflected light;
the original colour is radiated light.
 840°
For greatest accuracy, these cofour samples should be viewed 2 ft, away
from a GOW pear! bulb indirectly reflected by a silver shade

36
 .ouyine ayy Ag
alam
ave
ABA
gjeng
ainjeradiva)
Guueduia)
snB)y')
Aq
pue
SMAI
sosip
peanpoJd
20S
pjequosy
Agnjajga
paredaid
apduses
wos
sayy
pejaduiay
(15940

315*
295°
275°

285°
265°

305
245°
235°
225°

255
 gst
E
float in it, but you can easily make up you! Local overheating is almost inevit-
E E" test hydrometer if you want to be able if oxy-acetylene is used, and | go into
this in more detail later. This type can be
irreparable, with oxide inclusions forming
bts with the top about E inch above the in the grain boundaries, rendering the
?ace in plain water, It is then calibrated steel useless for anything but as a missile.
y setting it in brine of differing strengths. The commonest difficulty, though, is with
workpieces which are either very thin or
ard let it down when it tops 1096 you will which have a-very uneven cross section.
pewithin the not very important limits, The scriber-point is just such a case; it is
Decarburisation. Occasionally one finds impossible to heat the tapered section
3 tool that has been properly treated, evenly. The remedy here is to form the
averything as it should be, but somehow it taper and the point by grinding after heat
oes not seem to be hard. A file can treatment. For other cases — e.g., where
touch” it. However, any attempt to saw it there is a small protrusion on a larger
intwo results in a sawblade with no teeth mass — the remedy is to fit a "heat sink”.
jnit! The cause is surface decarburisation A piece of scrap steel with a hole in it (not
- the outer skin of the metal has lost some too large a piece; something in proportion
ofits carbon content, and it is not harden- to the main mass of metal) is set over the
ble steel any more, This is the origin of delicate part, and the whole heated up as
the old blacksmith's rule "forge large and one. The slender part is protected from the
direct heat of the flame, and will reach the
proper temperature just the same, both by
would have to be ground off. The problem conduction from the main body and by
isnot acute for the model engineer, as he radiation from the shield. The protecting
Is not heating tools with shanks up to a shield must, of course be removed before
E of inches square, but it can quenching. No problem arises, of course,
with furnace heating.
dis caused by the action of oxygen on Metal which is overheated will almost
he metal, and this is one reason for certainly crack on the quench. The best
avoiding the use of oxy-gas heating procedure is to stop work, anneal the
es. It can also be caused by scale: you piece, and start again. If, however, the
will recall from Ch.l. that Wrought Iron steel has been brought up badly over
Vas made by the reaction of iron oxide temperature — yellowy-red, for example —
ith the carbon in the cast iron. The then it will need reforging before it will be
medy is, of course, to avoid scaling in any use again. If it has got hotter than
the first place, as suggested above. | have this, throw it away; it won't be any use —
Very little trouble with tools around 5/16 not as tool steel, anyway.
i 3/8 inch square, but the few half-inch Whether water, brine, or oil is used,
Ones were all made deliberately about some debris will collect in the bottom of
i inch oversize at the point to permit the tank. In the case of oil this may
ufficient grinding down, include some sludge, which might
Üverheating. This can arise from several interfere with the quench. In industry
causes. The first — sheer forgetfulness, elaborate methods are used to keep the
When the work is left in the muffle too baths clean, but for us, simpler methods
— we can ignore; the cure is up to are suitable. For water, just pour it away

37
and use fresh. Brine may be kept longer larger tin for oil, which once held emulsion
(in plastic containers, not metal) and then paint; this has a lid, but | am sorry to find
flushed down the sink with plenty of that present day paint tins have very poor
water. It should last a long time anyway. tops. Mine holds about 2 gallon, which is
Oil is more precious and costly, but a quite adequate for all the work | do.
simple strainer can be used to remove the The initial temperature of the bath
dirt. Keep a lid on the oil-bath when not in should not be too cold. True, you get a
use. | find that even after two years my better quench the colder it is (| have an
own oil-bath seldom needs cleaning, and | account of tools quenched in freezing cold
use oil more than most model engineers. mercury!) but you do have to compromise
The proper oil should be used if at all against the risk of cracks. Something
possible. There are various grades, and between 20°C and 30°C is about right
each oil company has its own. | use B.P. However, in the summer | just use tap
"Quendila A22", though the "Quendila water which has stood in the workshop
19" is almost as good and cheaper. It is for a while: in winter | draw some from the
unfortunate that these oils are no longer hot tap. Oil should be at about 25°C.
supplied in the old one gallon tins, but a These figures are not critical — no need to
25 litre (about 5 gallon) drum can be go round with precision thermometers to
shared between friends — and a 400 litre measure them! A useful point to note is
drum could be cheaper still for club that brine or water at 70°C will give a
purchase. Failing a special oil, a good similar effect to that obtained from cold oil
"Spindie Oil" of about SAE10 to 20 and if you have problems with a
viscosity will serve; But NOT motor oil, workpiece constantly cracking in the har-
which is blended for quite different service dening and have no oil you could try
and characteristics, and most heating the water/brine to 50°C. A
emphatically NOT "used" motor oil. This quench in water at 90*C will not harden
is downright dangerous; it can contain up at all, but it can toughen the steel a little.
to 2596 of the heavier fractions of petrol The cooling in this case is almost certain
and can flash off into a furious fire if used to cross the "S" curve, with some Pearlite
for quenching hot metal. formation.
The size of the quench bath must Quenching Technique. This does require
accord with the size (and quantity] of the some attention. Far too many people set
work passing through at any one time. the work in the water and swirl it about
Whether water, brine, or oil, the amount violently. The recommendation seen in
should be 1 gallon per Ib of hot metal (10 some books that the water in the bucket
litres/kg) to avoid undue temperature rise. be set in rotary motion and the tool taken
A half-inch tool 5 inches long weighs down vertically at one side is even more
about 0.35Ib. For the odd centre punch unfortunate, Look at Fig.18. The water is
and similar small casual jobs | keep a large passing the work lor the work is passing
"Golden Syrup" tin, as these have good through the water — same effect) with the
lids, but for serious work a plastic result that there is a reduction in water
"builders bucket" does very well. (They pressure on the downstream side. This
can be had with lids). This will hold up to means that steam will form more readily
two gallons and is very robust — it will on that side and it will be under
even withstand the occasional accident quenched. There is plenty of motion to
with hot metal on the plastic! | have a displace steam at the front, and perhaps

38
 dg. 18 Quenching. Iftheflowof
water past
work is fast enough to cause eddies at
sam formation on the downstream
be aggravated.

t the sides, but very little at the back. the middle part of the quench — or the
Distortion is inevitable; indeed, skilled file- vapour blanket in the case of oil. With a
eners use this phenomenon to very large object you may have to move it
ghten a file which has bent in the more, but still gently, to avoid hot coolant,
ing! The preferred method is a but clearly a large piece needs a larger
‘ely SLOW circular motion of the quench tank, and you should not try to get
t, combined with an equally fairly away with an inadequate volume of
- vertical up and down motion, so that coolant. In industry, of course, elaborate
he work travels on a helical path through systems of propellers, underwater jets,
Ihe quenchant. All sides get equal treat- and so on, are used to get a controlled but
ment. The motion should be just enough random movement of the coolant, in
| remove steam bubbles as they form — which the work can simply be lowered.
more. Anything more violent than that Even so, it does need a lot of experiment,
Ses more harm than good. and it is interesting to know that the best
a long and slender object — say a files — even though made by the thousand
h — even this gentle rotary motion — are still hand quenched in brine.
be too much. In which case you must There is just one point about the oil
quench which may be helpful. | find that a
nent; again, fast enough to remove deepish bath is better than one of the
Ebubblesbut no more. You must try same volume but "fatter". This means
for yourself — a little experiment is that the type of workpiece which needs an
thanseveral pages of mine. The oil quench can be moved more in a
it is to remove the steam bubbles in vertical direction than otherwise. In many

39
cases | find that no rotary motion is annealed it, as the forging temperature is
needed at all. about 950°C — | first heat up to between
"Blood" and "Cherry" red and quench just
Stage Quenching The critical part of the the end 6 inches. the point is then briskly
quench is to get the work reduced in rubbed with a broken grindstone, with the
temperature past the knee ofthe "S" curve hot part laid on my small anvil to act as a
(Fig 15, just to remind you!) Once past heat-sink. Then, when the point is clean, |
this point the cooling can be quite a bit hold it to the light until the colours run up
slower — Martensite will still form. This to temper the point. Thus the point is hard
being so there would seem to be no and tempered, but the main body of the
reason why we should not remove the head is relatively soft. | shall be expanding
work from the bath once it has got down on Tempering in the next chapter, but can
to, say, 300 to 350°C, and cool the rest of say now that this procedure is quite
the way in a second bath, This is done in legitimate, though nowadays one would
commercial plants almost universally. It be happier (for important tools) if the
has several advantages. The primary bath temper heat could be held for longer.
does not heat up so much — important if To sum up. We must heat the metal
you have a number of pieces to quench. If slowly, both to ensure that it is hot right
using brine, the second bath can be clean through and so that the transformations
water, which will remove most of the salt, needed have sufficient time to take place
and there will be less evaporation from We must "soak" the metal at the harden-
the brine bath itself. There is no need for ing temperature, choosing the
agitation in this second bath — the metal temperature appropriate to the material
can take its own time. The only care you being used. The “quench” should follow
must take is to avoid dropping the immediately, and this must be done in a
workpiece, especially if it is fragile. It will manner which removes steam bubbies
be pretty brittle at this stage. without causing violent eddies in the
Partial Quenching There is no need to cooling medium. And the quench tank
quench the whole of a large workpiece must be proportioned to the size and
(nor need to heat the whole of it for that quantity of the work in hand. If all this is in
matter) if only part needs hardening. After order you should get optimum hardness
reforging the end of my pickaxe, for every time! It only remains to temper it
example — an act which will have and this we deal with in the next chapter
The need for tempering has been toughness at a relatively low temper. It
scussed in Ch.ll, and to some extent the must be said, too, that the loss of
ge of temper also. But whilst the hardness may be more apparent than real,
ach temperature is fairly easily deter- and is difficult to put a number to. Indeed,
, depending mainly on the carbon the point of a lathe tool may well show a
tent of the steel, the tempering variation of two points on the Rockwell
perature depends almost entirely on scale over the surface. (Hardness testing
the use to which the too! will be put. It is numbers and their determination are con-
pot even sufficient to say “Lathe Tools" sidered in Ch.B.)
for the temper needed (say) for turning There is a further point, and one which I
ENTA will be very different from that believe is often overlooked, Just as quoted
appropriate to a tool to cut chilled cast cutting speeds in Production Engineering
'on. The effect, as we have already seen. textbooks are geared to industrial, not
| to improve the "toughness" - model engineering, conditions, so are the
stance to shock — at the expense of recommended tempering temperatures in
è of the hardness. A tool not subject toolmakers catalogues and metallurgy
D shock, such as a file, may not be textbooks. The temper they suggest for a
pered at all, while a spring will be carbon steel lathe tool, for example,
pered almost to soft. It follows that assumes that it will be in use for a full
siderable personal judgement must be seres of shifts over a 44 hour week.
d, if only to take account of the work (Actually, probably a 56 hour week at the
hormally done in the shop. One reason for time the book was written!) Some of my
using carbon steel tools instead of HSS form tools have had no more than half-an-
apart from the fact that they are harder) is hour's work in twenty years! So —use your
it is very easy to temper to suit the judgement, bear in mind the discussion in
Ch.l, and, above all, experiment. The
e change in toughness and hardness premier model engineering club, is after
already been mentioned, but | show a all, the "Society of Model and Experi-
ier diagram in Fig 19. | must mental Engineers"!
asise that the actual shape of the Temperatures. Having said all that, | will
curves will vary from steel to steel, with try to give some help in the form of
ne, for example, showing a "peak" of guidelines in the table below. | have run

41
 8865

88

TEST.
'C'
ROCKWELL
HARDNESS
-
—
TOUGHNESS
TORSION
IMPACT
FT.LB.
TEST,

200. 300 :
TEMPERING TEMPERATURE — *C.

Fig. 19 Effect oftempering on the mechanical properties of a brine-quenched 1.15% carbon steel.

into quite a difficulty over this, as different to 15*C higher than do European ones.
authorities, different steelmakers, and This may be due to the different steel
different users quote varying tempers for specification, but | suspect it is probably
the same tools. In particular, American because, at the end of the 19th century,
reference books seem to give figures 10 when much of the work was first

TEMPERING TEMPERATURES FOR VARIOUS TOOLS

Arbors 200°C Dies (Drawing) 200°C Pickaxe 255-275"C

Axe, Cutting edge 255 Drill (Very small) 210 Planer Tools 215-225
Brass-turning tools — 170-190 Drill (Small) 220-240 Reamers 230-240
Chasers (Thread) 230 Drill (Large! 245 Scrapers 200
Chilled iron-turning — 180 Engravar's tools 230-250 Taps 210-220
Cold Chisel 280-280 Gauges 220 Screwdrivers 280-290
Counterbore 220-240 Hammer-head 230-250 Shaper Tools 215-225
Centres (Lathe) 215 Lathe Tools** 200-220 Springs 300-310
Dies (Screwing) 215-225 Milling Cutters 210-230 Wood Chisels 215-225

NOTES (1) ** Sae comment In the text on lathe tools

(2) "Silver Steel" may be tempered at slightly lower figures than plain carbon steel.

42
 jblished, American firms worked their tempers we have been advised to use in
achines rather harder than we did. In the the past are far too high. When the
sble | have tried to iron out these accepted practice was first established
erences, but where in doubt | have carbon steel tools were universally used,
towards the lower figures, to take both by mode! engineers and in industry.
ccount of the lighter duty cycle of model Tool sections were large — a 34 inch
engineers’ tools. Britannia was designed for half-inch tools
| These temperatures are, traditionally, — and we used industrial practice where,
iated with "Tempering colours", and as | have already pointed out, the cutting
ny publications simply refer to "straw" edges were subject to an 8 or 10 hour
- "blue" instead of quoting the actual day. :
figure in °C. The following can be used as It is obvious that we must temper
B guide, but again, there are slight down the shank, otherwise the clamping
differences between different publica- forces will cause fracture, but the actual
ons. There is also a problem in cutting edge is seldom subject to any
interpretation: what sort of straw, for shock load — on a lathe tool, at least; the
example! case of planer or shaper tools is different.
E
Yellow 215°C Light Purple 275°C
D
Pale Straw 225°C Dark Purple 285°C

Straw 235°C Dark Blue 295°C itodft

|= colours appears
Dark Straw 245°C Blue 305°C enne se
i
Reddish dark straw 255°C Pale Blue 315°C
IBrown-red 265°C Grey Above about 330°C
- These colours are as viewed in daylight Now, the cutting edge is subject to
or strong tungsten lamp electric light. heating. We can use a higher cutting
| will be quite at fault if judged under a speed with "High Speed Steel" simply
luorescent tube. We shall look at the use because its tempering temperature lies
ofcolour-tempering in more detail later in between 600 and 700°C; some grades
the chapter. will still cut when red-hot. We use lower
Special case of lathe tools. First, let us speeds with carbon steel because if the
think again about what we are doing, The tool point does get at all hot it will
Object of tempering is to increase the "Temper down" and soften. Clearly, the
toughness — resistance to shock and lower the initial tempering temperature
hipping — even though this means a the less risk of damage to the tool point
ght loss of hardness. Clearly, in a from wear. And the harder the tool point is
ntre-punch the risk of damage from to begin with, the longer it will take to
ows is considerable. But in the case of heat it up to danger point when (in effect)
ie tools, WHEN EMPLOYED ON THE it "tempers itself" and loses hardness.
OF WORK DONE BY MODEL For these reasons my practice for some
INEERS, | would suggest that the time has been to temper lathe tools very

43
lightly — just sufficient to effect a little Tempering should take place as soon
grain refinement and to give some stress as is practicable after quenching. The tool
relief to the crystal structure; there is still should, of course, have cooled to not more
some marked increase in toughness. | first than 50°C, and cooling to room
temper the whole tool at between 180°C temperature is safer. The reason for this
and 200°C, and then "let down" the expeditious tempering is that the metal is
shank only at around 280°C whilst under internal stress from the quench and
keeping the tool point cool either in a bath if, as is most usual the workpiece has a
of water or by using the ancient device of "shape" as opposed to being just a block
sticking the business end into a new-dug there is risk of spontaneous cracking
potato. This gives excellent results, both arising from differential contraction as
with "Silver Steel" and with a straight well. There is also a risk of “crazing” —
1.1596 carbon steel. | use an even more time cracks; these can appear if a tool is
refined method for my Ivory-turning tools left too long in the as-quenched condition.
and those for slide-rest work when "orna- This need for "haste" is one reason
mentally turning” in exotic wood, and this why horologists and gunsmiths like the
is dealt with on page 74. "blazing off' procedure — the tempering
takes place as soon as the work comes
Tempering Procedure. To be properly out from the quenching oil. If for any
effective the tempering process must be reason tempering must be delayed then it
given time to work and we are also faced, may help if the work is set in boiling water
as in the case of heating before for a quarter of an hour or so before
quenching, with the need to be sure that setting it aside to await the final treat-
the tool is hot right through. If we have a ment.
tempered surface with a brittle core we
have not really reduced the risk of Heating for Tempering. First, the "easier
cracking under load at all. The rule is the way" mentioned above. Go and look at
same — ONE HOUR PER INCH OF the cooker in the kitchen. You will see that
THICKNESS, though this is not quite so it has a thermostat on the oven if gas or
important in the case of a too! shank, electric fired, or a built-in thermometer in
where any temper between 260 and it if fired by solid fuel. You will, | think, find
290°C will serve. For the tool point, of that the thermostat goes up nearly to
course, the rule applies to the thickness 300°C — certainly to 250°C — or to "Gas
there, not that of the shank. Now, one of Mark 9". (Degrees Centigrade = 140 +
the virtues of the "classical" tempering 11 times the gas mark approximately) So,
method — allowing the heat to creep up you have a good thermostatically con-
the shank to the point — is that this heat trolled, or at least "temperature indicat-
travels, in the main, through the core of ing", tempering furnace on the premises!
the tool; and provided the shank is heated The cookery book will give you the
gently enough, the "soaking time" will temperatures, and if you time things righ!
look after itself. Even so, there is a risk you can set your tools in the oven along-
that the temper may not be complete; yes; side the joint. The table below gives the
| know it works, but it is so easy to make it usual temperatures used in this sort of
"work better" and we should always try heat treatment shop, though the time may
for perfection. We will look at this "easy vary a bit depending on the judgement of
method" in a moment. the oven-shop forewoman.

44
 eC Irish stew 2h "Tempering Oils" which can be used to
> Braised pigeon, 45m: Casseroles, temperatures as high as is normally
24h: last stage of roast duck etc, 3h. needed. (Quenching oil is not suitable),
aC Roast chicken 20m/Ib: last stage However, you do need a substantial con-
of roast beef, 15m/Ib. tainer — a thin one may overheat the oil
; Cottage pie, 1 3h: Mince pies, 25m locally — and if using gas heating, a gentle
Cheese straws, 15m. flame, not the sort you would use for
Yorkshire pudding, 25m: most pies, brazing.
25m: First stage of roast duck or The Salt Bath is the normal method of
turkey, 20-30m. tempering used in industry. Tempering
‘Scones, 10m; Bread, 30-40m., salts are quite innocuous, being a mixture
First stage of roast beef of Sodium Nitrite and Potassium Nitrate,
or pork, 20m. (Nitrite, and Nitrate) which melts at
160?C. Again, you need a substantial
addition, you probably have a deep- container, and | use the outer vessel from
g chip-pan, Chips are cooked at an old-fashioned glue-pot. Fig 20 shows
> for 5 minutes or so, and then the setup; the old boiling ring | picked up
rowned'" at 195°C, but the oil can at a sale for a few pence finally gave up,
| be taken up to about 200-210*C and you see the pot here on a replace-
an electric cooker. (One has to be ment. Ordinary mercury thermometers,
ful on gas, as unless the oil is fresh usually nitrogen filled, are available which
ire will spitting and some vapour). go up to 400°C, and are not expensive —
fm compare the above with the they can be had from laboratory equip-
jle on page 42 you will see that this ment suppliers, or ordered through your
mes ic workshop can temper many of local pharmacist. Alternatively the weli-
E^
E. for you while performing its known "Rototherm" dial thermometers
Drmal office, and it is only necessary to can be used. But for most of our work an
eg tiate mutually acceptable terms with oven thermometer from a scrapped
} proprietor! There is, of course, no cooker will go up high enough — though it
may be marked in Fahrenheit.
The salts are marketed through Edgar
êd "out of cooking hours" and in this Vaughan & Co. Legge Street, Bir
nection it is worth noting that a too! mingham and it may be that by the time
hich ought to be tempered at 200°C is this book is published the usual model
biten better treated if it is "cooked" for its engineers suppliers will market them in
time at 190?C than if merely "heated rather smaller quantities. (Direct supply
from Vaughans may involve a "Club
purchase"). There are, of course, a number
le t. The only caveat | would enter is of brands, but the only sort | have used is
at itwould be prudent to use an oven or the DEGUSSA type TS150/AS140. It
pan thermometer, as the thermostat comes as 8 powder of the consistency of
domestic salt, but must be kept dry, as it
can absorb moisture fairly quickly. It is
non-toxic, and the only serious hazard is
ü"pering oil in the kou and you that, like some weedkillers, if absorbed
“Still be able to obtain the special into woodwork lor your overalls) it

45
 Fig. 20.4 simple tempering salt bath. The outer case gi ,
cast-iron glue:pot is set an on electric bailing ring, Note tha
protective case on the thermometer. Normally à glass fis
insulating blanket is wrapped round the por, bul was re
moved for the photograph.

rise again — there is not, as a rule, any

need to increase the heating rate. Leave
the work in for the prescribed time, having
an eye on the temperature; you may have
to adjust the control switch from time to
time, but if it stays within a few degree:
this will be sufficient, The piece can be
allowed to air-cool after the time is up
there is no need to "quench"; quenching
is only necessary when, as in flame
heating, stored heat in one part of the too!
might overheat the part being tempered
Any salt which adheres can easily be
washed off in warm water afterwards.
Once you have tempered the too!
"supports combustion". | deal with overall in this way the shank can be “ler
"Safety" in heat treatment on page 102, down" with a gas torch, as | have pre-
and only mention this now to reassure you viously explained. With milling cutters, D-
that we are not talking about material bits, taps, dies, or any other complicated
which is poisonous. tools this is not necessary. In the case of
The great advantage of this method long tools like reamers you may have to
(like that using the oven) is that you have use a deep water-bath with some means
complete control of the temperature to of holding the. tool upright in order to le:
within a degree or so. There is no problem down the driving square on the end.
with “time” — you can leave the work in as The salt bath is the answer to almost all
long as you like. And the actual time can the problems met with in tempering. For
be a bit shorter, as the heat transfer is very complex shapes — especially milling
good indeed, The one thing you do have to cutters — it is almost the only way that the
watch is that there is no oil or water tool can be tempered properly, and size |5
trapped in little holes, as this would be no problem either, provided the pot is
dangerous. The procedure is first to bring proportioned to the size of the work — you
the bath up to the working temperature. need about 10lb of salt for one Ib of steel
The workpiece can be set alongside the but it isn't critical. You must, of course
pot to dry if you have any doubts about it remember that the tide will rise when the
being dry. (With oil-quenched work ! work goes in; don't fill the pot too full'
usually wash in carbon tetrachloride or à Finally, a detail point on heating. Wher
similar degreasant), You need some cold, a crust will form on the top of the
means of holding the work — it is no fun salt. Initial heating must be slow, and it i5
groping about in hot molten salt to find it, advisable to apply just a little heat to th?
A piece of wire wound round will serve, outside at the top of the pot to get a little
Immerse the metal slowly: you will notice melt around the crust. A lid should alwav*
that the temperature falls, but will slowly be in place when restarting a salt pot.

46
 r Tempering. This is the method gone before! It does help to have a piece
most model engineers are accustomed to. of cold steel alongside for comparison if
bviously. even if you have a salt furnace the very palest straw is the aim. As soon
just isn't worth firing it up if all that as the colour is "right" the piece must be
eds treating is a rehardened screwdriver quenched — in oil or water — to prevent
(here are many jobs in the average any heat stored in the shank of the tool
shop for which sophisticated from taking it further. The quenching
iods are not worth while. Of course, if serves no purpose so far as the actual
have no salt-bath, and the use of tempering is concerned.
stic oven and chip pan are denied The work can be heated in a blowlamp,
using a VERY soft flame, but whenever
possible | use a spirit lamp. It is quite
any way be disdained as "bad practice": adequate for most sizes of tool, and
method has been used for centuries, though it takes rather longer than a gas
words and pike-heads in the days burner this is all to the good. We are not in
there were any lathe-tools, a hurry! It provides a very clean flame, and
e colours are formed by a film of one which does not interfere with colour
»which is very thin indeed and which formation. Even so, it is necessary to
remove the work from the flame to
ould seem that the colour change is, in observe the colour, as it will not develop
art, due to the light reflected from the properly in the presence of burning fuel.
lace of the metal itself passing through Indeed, it is possible actually to remove
fering thicknesses of the oxide. Cer- the colour with a reducing type of flame.
fainly the exact colour does depend on Naturally, you must keep the work moving
gth the type of metal and the surface about in the flame in order to get even
ish. The difference between the colour heating if you are tempering the whole.
Objects which are of irregular shape
can be difficult. The heat should be
e@tween that of quenched and applied to the heavy part of the section,
Inquenched carbon steel. The effect is to and you must keep a very careful eye on
ive a variation in temperature for the things. Fine corners may "take off" into
Same shade, but this is only the odd the blue before the main body has
gree centigrade and is not important. reached straw. For such work | always use
the salt bath or domestic oven. But if you
ished; a dull but smooth matt surface is must use a flame, go very slowly indeed.
est unless you are blueing the steel for The difficulty with the direct heating
Smetic effects. The colours will, of method is that we are applying the heat
Se, run into each other if the work is from the outside, and there can be no
ed from one end, and they should be guarantee that the internal structure is up
to the temperature; it certainly won't be if
tions. The most usual difficulty is in you heat too quickly.
Beciding when to stop; the straw may If flame heating is the only means
Apear to be too pale, and then, before available, then you can try one of the
you have time to think, becomes too dark. "semi-direct" methods. One which is
"is is perhaps a stronger argument for often recommended is to bring a fairly
"Ow and gentle heating than any that has thick steel plate up to just below dull red

47
and to expose the workpiece to the that the temper is uneven, from poin
radiant heat, turning it about the while. backwards. In the case of a knife-tos
The higher tempering temperatures may there will be marked difference in
need the plate to be red-hot. You can do a hardness along the cutting edge, and a
similar thing with a firebrick oven; just parting tool will get softer as it i
arrange a few firebricks to form an open reground. This is the main reason why |
box, heat them with a torch, and then hold changed to the “two-stage” tempering
the work inside with tongs — again, process already described. Nevertheless
turning it about all the time. Many use a the method is sound, and if care is taken
sand-bath. A tray of dry sand is heated (very gentle heat, with the flame mover
from below and the work either laid up on backwards, away from the point, as the
it or buried, the sand being turned aside to work gets hotter) satisfactory results can
inspect the colours from time to time. (It be assured.
helps to have a thin piece of similar The idea can be extended. Fig 21!
material laid on the surface of the bed in shows a method of tempering a small
that case, to act as an indicator). The sand milling cutter, of the type (used in
must, of course, be kept well stirred whilst horology) which is screwed on to an 8mm
heating, to make sure that it is at an even collet arbor. The copper rod is threaded
temperature. | would not recommend and screwed into the cutter, and heated
sand for temperatures in the "Pale Straw" with a blowlamp. The heat runs down the
region, but for deeper colours it works rod, through the metal of the cutter, to the
well; as soon as you see a pale colour you teeth. Note that the cutter is set on the
know that the others are not far behind. rod as shown, so that the teeth are more
Transmission Tempering This is the or less of uniform distance from the hot
"classical" way for lathe tools, but it can rod. Fig 22 shows a similar device in use
be extended. The tool is heated at the tempering a die. In this case the circular
shank end, and as this turns blue, so the die is wedged into the end of a piece o!
purple — dark straw — straw — pale straw copper tube, which has four slits in the
colours run towards the point. As soon as end. There is a ball of steel wool inside the
the pale straw reaches the region of the tube just below (but not touching) the die
cutting edge the tool is quenched, point to prevent hot air currents from heating
down, in water. The virtue of this method the centre part. The tube transmits the
is, as | have already suggested, that the heat to the die from the outer perimeter
heat travels down inside the metal, and and the whole is quenched when the
we can be fairly sure that the tempering cutting edges reach the right temperature
has reached the centre. The difficulty is Again, the heating must be gentle. The

Fig. 21 Using a “heat pipe " when tempering?

small dovetailmilling cutter.
 122 Tempering
a die, The copper tube acts as a hear pipe.

age here is that the outside of the

ihich acts more or less as a spring in
Vice! is hotter than the cutting edge
so more deeply tempered — as it
id be.
iti g saws these days, but similar
acts can be tempered by gripping
ween two washers on a bolt, held by a
. The heat is applied by heating the
Many other examples will come to
ind. The object is the same in all cases:
rst, to heat from the part of the tool
hich need not be hard, towards the
ing edge; and second, to arrange
$ so that where possible the heat
s through the interior of the tool, of decades if not centuries of trial with
+
ng the cutting edge from within. one oil — from the sperm whale — but with
Blaz Off. This is applicable mainly to the variety available and used today |
ork which has been oil quenched. As would not care to give any instruction at
s00N as it is removed from the quenching all except to "Experiment first". It works
ath the tool or component is held over a very effectively, but does need experience.
I| flame until the residual oil ignites To sum up; the ideal tempering
medium is the salt or oil bath with the
Ork is taken from the torch or lamp domestic oven or chip pan a practicable
ame) tempers the work. For larger pieces alternative. With these methods the exact
ié books" suggest binding some iron temperature of the work is known, and
a round, to hold more oil. The '"theory" conditions can be repeated exactly. The
anced is that the temper temperature use of colours to judge the temperature is
301 lated with the flash point of the safe, but needs a little trial and experiment
to judge the tints accurately; the
Merely an indication of the temperature at temperatures will then be within the
ch (under very artificial conditions) oil "tolerance" for the type of work we do.
ll Start to ignite. The actual temperature The actual degree of tempering needed
should be considered, bearing in mind that
most of the recommendations in Hand-
S well as on the heating value of the oil. books etc are for industrial conditions, not
Or the experienced practitioner, and for those of the model engineer. The temper
Yey small clock parts like click-springs heat should be held for a period — ideally
Ad pallets, the system works well, but for one hour per inch of section — and
hey will have experimented a great deal should be applied slowly. Tempering
ld, most important, will "know their oil". should be carried out as soon as possible
ructions in old books were the result after the initial quench. | will add just one

49
final point; in some cases we do want a at 100°C — just boil the part in water fo;
surface to be of maximum hardness and at half-an-hour or so. This will not reduce the
the same time to have a fine grain — hardness at all, but will achieve some
wearing and rubbing parts are cases. If grain refinement. For "fine grain" cutting
this is so, then it is worth trying a temper tools | suggest a procedure on page 74.

50
 firebed, but where you can see it and
observe the colour. It must be turned
and perhaps more who have inherited about at intervals so that no one side is
hem from an earlier generation: it seems uppermost for any length of time. Take
have been an essential piece of equip- care that it is not exposed to the direct
nt in the earlier times. | do have a small draught from the fan. Slow, even, heating
is the order of the day, avoiding the
ad mainly for smithy work, and then impingement either of direct, cool,
nly rarely, | have hardened large objects draught or a draught which has passed
Vit. The last job was a pickaxe point, and through any extra hot part of the fire,
must confess that winding the handle Overheating must be guarded against like
8 bit wearisome! The forge was the the plague, and you must bear in mind
ural thing to use — apart from saving that the metal will look cooler than it is
sive gas — as, of course, the point until you get used to observing the colour
ded reforging as well. in contrast to the hot coke. You can bring
a forge fire up to 1500°C and metal at
780°C will look dead cold in comparison,
As. an alternative the egg-shaped A useful expedient is to use the semi-
janufactured fuel ("Phurnacite") is indirect method. Set a piece of (say) 2
indy, as the eggs are more or less of inch steal pipe in the midst of the fire. Get
ir size. Housecoal is useless, and this hot — a little above the temperature
hould not be used even to start the fire, you need — and heat the work inside it.
ind steam coal little better. For hardening You are then able to judge the
u need a fairly deep fire, hot all through. temperature more directly and, further,
© achieve this the fuel must be stirred will be protecting the workpiece both from
Bout in the earlier stages. Once you have cold draughts and from local overheating.
nice glowing firebed the fan must be It is slower, but, as we have seen, fast
Sed with discretion — you need a draught heating is NOT the prime requirement.
ather than a "blast". The work is laid on Domestic Fires. This is a practical
‘9p ofthe fire at first — in fact, it can be left alternative to the smith's hearth for
re, preheating, whilst you bring the fire smaller pieces. After all, Nasmyth made
ip.7 e piece is then set in the heart of the the castings for his first steam engine in

51
 Fig. 23A domestic stove of t)
type can be used to heat work |,
hardening.

his bedroom fireplace! The ordinary open 1300°C! The main problem | find is tha:
grate will get hot enough only for very unlike the smith's hearth you have ne
small work, but any closed stove with a positive control; the response to adjust-
proper air control.can be used for anything ments of the air valve is slow, and you
that will pass through the front firebars. have to anticipate changes in the rate ol
(Fig 23) As with the forge, coke or burning.
"Phurnacite" can be used. | have had no Naturally, you must take account o!
experience using anthracite. your surroundings — it IS your hearthrua
The grate must first be cleared of ash after all! But the main objection | have to
and clinker and the fire then built up by the use of the domestic stove for harden
adding small quantities of fuel at a time ing is that tha front grate-bars prevent the
until the firebed is as deep as can be use of a piece of pipe. However, with the
accommodated. The aim must be to get a closed stove there is little problem in
uniform rate of burning right through. “Soaking”: once the fire has reached its
Once this is achieved the front doors lor temperature it tends to "stay put" fo!
one of them) must be opened and the air quite long enough for the size of work
valve in the ashpan adjusted to get the fire normally done. It is very useful for annea!
above, but not too much above, the ing, as the piece can be allowed to coc!
temperature needed. The tool should be with the fire overnight. Even if the fire is
set in the upper part of the fire and, as in not "out" in the morning this is not impor
the case of the smith's hearth, should be tant |t will have cooled through the
turned about at intervals. You need have critical range very slowly and can be set in
no fears that it won't get hot enough; the the ashpan to finish off.
stove in Fig. 23 can be brought up to Some books suggest that there may be
1000°C very quickly indeed, and if you a tendency towards both scaling and
find any clinker in your grate, this is an decarburisation when using the smith 5
indication that the grate temperature has, hearth, but this need not cause us too
at some time, been of the order of much concern, It is true that a hearth with

52
 23A The author's hand fan
pht forge. Situated in an
'ho for safety, as the fire can
ppraw eut sporks to some height.

blast can decarburise, but we are not deal with the various types in turn. | most
dealing with the size of work normally strongly recommend that you NEVER,
except in the most dire necessity, use
iO take care if the heating. time is half an Oxy-Acetylene (or Oxy-anything else) for
hour or more, but it is only a case of any form of hardening operation unless
avoiding direct impingement of the blast. the use of such equipment is part of your
ime scale will be formed, of course, but daily work. Even then it is risky. First, if the
flame gets too close to the work, even for
a second, that point will be overheated,
decarburised, and almost certainly suffer
Mtrogen — there is much less oxygen from oxide penetration at the grain boun-
Present than in a fan-driven fire. It is good daries. Second, it is very difficult to get
practice, of course, to remove any scale even heating as the heat source is so con-
w ihich may be present from previous work, centrated. If the flame is adjusted to avoid
including the "mill scale" on black bar. scaling there is then risk of excess car-
lore scale will be formed in the long heat burisation which would result in cracking
Ssociated with fire-annealing, and if the on quenching. Finally, it is a very expen-
Voidance of scale is important the work sive method of heating. Few people seem
should be protected. For overnight anneal- to appreciate that though the flame is very
ing a good way is to encase it in fireclay hot indeed the actual heat output is
WOT fire-cement) reinforced with iron extremely low. In fact, the largest nozzle in
B. Or put it in a tin box packed with the outfit | used to use (for welding 4 inch
e with a little charcoal mixed with it. steel plate) gave out rather less heat than
Slowlamps and torches. Most model did my smallest paraffin blowlamp. It is a
Engineers will have some form of paraffin case of horses for courses; for its purpose
OF gas torch for use when brazing and it is oxy-acetylene torch does its work very
Natural that these should be the first well indeed, but that work is not heat
fesort when heating for hardening. | will treatment of steel unless you own a

53
 e r a Fig. 24 Typi
cgo RM = temperatures rs ache,

——— ——.
within the fame from a seii
blowngattorch
«2009c, 650°C. *1300°C. #800
Q L^
À 8009c
p- ———

"Shorterising’ surface hardening plan:

and that is unlikely! So, let us turn ou:
attention to the gas-air type
The Torch Flame. Though | shall be refer
ring to "gas", the same remarks can be
applied to the paraffin blowlamp flame
The difference in the fuel is not important
- the heating value per pound mass of fuel
(liquid or gas) is very nearly the same, and
the paraffin (kerosene in USA)
vaporised before burning, so that the
flame is in all essentials a gas flame. Suct
flames will appear as shown in Fig. 24
with a central cone surrounded by a more
or less diffuse outer curtain, though its
actual shape will depend on the fuel used
the type of air supply, and the air/fue
ratio, Primary combustion occurs withir
the inner cone and the heat release
causes decomposition of the rest of th«
fuel gas. These fuel constituents burn in
the outer envelope, using any excess ai!
present within the flame and air drawn i:
from outside as well. The inner con:
temperature will seldom exceed 1100°C
but just beyond the tip of the inner con:
the figure will be in excess of 1600°C
(The theoretical maximum is higher st
but is not reached because no burner i5
perfect and in any case there is conside!
able radiation of heat from the flame
itself). Beyond this point the flam?
temperature diminishes again, as showr
in the sketch, It is not always appreciated
that a similar temperature gradient occurs
Fig. 25 Gas torch with pressure air-supply, showing t^t
rangt of Hare formation possible. (Courtest Flarmefast Ltd)

54
 526 Two types of seif-blown
orth. The lower burner
has the
holes remote from the flare
Ern and can be used in confined
aces. That shown above needs a
upply of air rourid the flame.
vt Ltd.)

ACROSS the flame, but this is the case. If do need careful attention both to gas
there is a long inner cone it may be that pressure and to burner head selection.
heating will be more efficient if the flame And they do need clear air round the
is laid alongside the work rather than be burner head except for the special ones
directed straight at it. which have extension tubes beyond the
The general form of the flame will be air entrainment holes. When used in a
he same for all fuels. perhaps some firebrick "cave" the flame must have room
fference in colour and some fuels will to develop outside the entrance or it will
" cone than others. but "blow itself out". The paraffin (kerosene)
is considerable difference in blowlamp is rather more flexible, largely
because the fuel vapour is preheated, but
hich come from a "self-blown" burner. also because the inflammability limit of
such as those normally used on Propane paraffin is much wider than that of
Or Butane, and those which have a Butane, Propane and Methane (N.Sea
eparately controlled air supply derived Gas) so that the air/fuel ratio is much less
Tom a fan or blower. There is some critical.
feasonably wide control of the size of the All these flames will, if the burner size
e with the self-blown type, but any is right, exceed the temperature needed
ge in character (eg. from “bushy” to for heat treatment of steel. There is no
edie”) requires a change in burner problem here. However, we are concemed
d. With the air-gas torch there is with even heating and, in most cases, with
plete control of the flame, both as to "economy". It is desirable that the flame
Size (within reason) and character solely be short and broad rather than one which,
BY Manipulating the air and gas control though equally "powerful" (in terms of
ives. There is no doubt that the burners heat release) is long and needle-like. With
th separate air and gas supply and the air/gas type this requires only the
control are by far the most efficient and manipulation of the controls but with the
enient to use. The self-blown self-blown torch the burner head must be
selected accordingly. Those described as
"broad blowlamp" are usually the most
appropriate.
Applying the flame. Heat transfer from

55
 Fig. 27 A crude but very effective heat conserving deview
The hot gases escape between the back and top bricks. Nore
the little ceramic trivets for supporting the work.

will be heated more evenly. And, finally, i!

we have several pieces to harden then we
can save a considerable amount of gas
The "cave" is really a little furnace.
The intelligent use of firebrick will bot
save gas and ensure more even heating. |
use three types. First, ordinary firebrick
from 1 to 3 inches in thickness. These are
refractory rather than heat insulators
they will withstand very high
temperatures. Second, lightweight
“Insulating Bricks", of which "Fossalcil'
and "Folsain" are typical. These look and
feel like cork and are intended for heat
insulation behind walls of firebrick in
furnaces. They can be at 1000°C at one
end and dead cold at the other, and are
available in the same sizes as firebrick
Both firebrick and insulating bricks can be
flame to workpiece is very poor indeed. had from most builders merchants. The
On test, a piece of steel about the size of a third type are "Hot Face" bricks — more
5/16 inch sq. lathe tool, laid flat on refractory than the insulating type (which
firebrick, took five minutes to bring up to are friable and melt at around 1250°C)
the hardening temperature — something and with much better heat insulation than
like 1500 units of heat from the burner the ordinary firebrick. These seem to be
transferring only 10 heat units to the available only from specialist suppliers
steel. Not a very good score! By setting up (eg. Messrs Flamefast Ltd, Pendlebury
a simple "cave" like that in Fig. 27 the Trading Estate, Manchester M27 1FJ, or
heating time was reduced to 105 MPK Insulating Ltd. Hythe Works
seconds; about 500 heat units, Not good, Colchester CO2 8JU). They can be cu!
but better than before. You will, of course, with a saw and drilled, and will stand
quote my own words and say that the five reasonable loads. An accessory which !
minutes of the first example is rather too find most useful both upon open firebrick
fast for the heating-up time, and you and in furnaces is the little ceramic
would be quite right. But we have to "trivet" as used by enamellers — Fig. 28. |
HOLD the temperature for about another use these to support work so that flame
quarter of an hour and if this is to be done can get beneath and, in furnace work, to
on the open firebrick the burner must be ensure that the underside of the piece is
kept at full throttle the whole time. Using radiation heated and not by conduction
the "cave" it can be throttled down. from the muffle.
Moreover, as the work will be heated Furnaces. The obvious next step from the
largely by radiation from hot firebrick it cave is a proper furnace, Those who do a

56
 2B Method
of holding a screwing die forquenching, with
on which itis supported during heating on the right,

of brazing may well have one of the old

"Utile" gas-fired forges which lin the days
when | had access to town gas) served me
brazing hearth, heat treatment shop,
melting furnace. They are still avail-
e, modified for use on North Sea gas,
from Alcosa Ltd (trading as William Allday
& Co. Ltd, Stourport on Severn, DY13
QAP) but rather costly, I'm afraid. But for
the serious practitioner, who needs heat
for all the purposes mentioned above,
1 are, | believe, the best small general
purpose heating unit available. The only
drawback is the lack of temperature
indication, but if you use heat sufficient to
justify the cost of a Utile you will be pretty small gas-fired kiln as used by enamellers,
expert at colours anyway. marketed by the Flamefast Co. as their
There are two small furnaces on the kiln type LN1000. This is rated to reach
market which | know of which may be 850°C in 10 minutes or so, and will run
worth looking at. The first is simplicity up to 1000?C. It is fitted with a pyrometer
— the Alcosa "Export" Portable and manually controlled firing rate, the
Forge. Fig. 29. Intended for use by maximum being equivalent to 34kw. The
itinerate farriers abroad, it can reach up to chamber is about 7 in x 5 in x 73 in and it
1150°C and is very rapid in heating up. can be fired on N.Sea gas, Propane, or
Once up to temperature the supply can be Butane. Fig. 30.
reduced — the full heat output is equiva- There are others, of course, in varying
lent to about 16kw. There is no pyrometry degrees of sophistication and size. Anyone
and temperature must be judged by who practises the art of enamelling (or
colour. Very simple — almost crude — but some types of pottery, for that matter) will
Cheap and effective. The second is the have such a furnace which can usually be

Fig. 29The "Alcosa" portable propane Hirod furnace. The

‘chamber
is93inwide x
x 5} inhigh x 113 in deep and can
teach 12007 cwith ease. (Courtesy William Adr Ltd)
 Fig. 30 The FLAMEFAST LN100Q kiln, This is intended for
use by enamellers. but is more than adequate for heat treat
ment up to 1200°C, It isfitted with apyrometer and ig fired
from either bottled or mains supply gas. (Courtesy Fiametas:
Ltdi.

than interesting at times! But see Ch.IX.

Tuba Furnace Most of my own work is
done either in an old electric muffle or a
salt bath, but | do make frequent use of a
rigged up gas-fired tube furnace, shown in
Fig. 31. The tube itself is a piece of 14
inch exhaust pipe, relic of the days when |
used "proper" (ie. "vintage") motorcars,
about 6 inches long. The tube is supported
in 1 inch thick Folsain bricks with a hole
cut using the ordinary tank hole cutter.
One of the cores from the drilling is used
to plug the back and as it has a hole from
used quite as well for heat treatment. the pilot drill this enables me to set a
There is also the implication that for home-made thermocouple inside if need
model engineers, being the sort of be. The other core is used as a "front
ingenious people they are, it would not be door". | give more details of the construc-
difficult to "make your own". This is not as tion on page 90, but it IS very useful for
easy as it seems. There have been more those odd jobs which crop up which are
fortunes lost (and eyebrows too!) over the neither large enough to warrant firing up
combustion of gas than in most branches the muffle, nor trivial enough to trust to
of engineering — it is not just a case of direct heating.
providing a jet and lighting the gas! It Electric Muffles. There are many small
does need a lot of experiment, and | would electric muffle furnaces on the market,
suggest that it be done outdoors at first: some for laboratory work, others for
delayed ignition associated with a enamelling, and these provided the most
chamber of a cubic foot or so can be more economical heating device where any

Fig. 31 A very simple gas-fired

tube furnace.

58
 Fig. 32 The author's electric mutfie, capable of working up to
FOOO"C. lt is fitted with à pyrometer, but has "Energy
Regulator’ controlofthe temperature
"soaking" is needed. They do take a fair
time to heat up, but once at the operating
temperature use very little power indeed.
All are fitted with pyrometers these days,
so that accurate temperature control is
possible even if there is no thermostat.
Heating is radiant into the work and it is
easy to avoid scaling. Last but not least,
they can be left to look after themselves
with safety, making no noise and, if not
thermostatically controlled there is usually
a "thermal fuse" which breaks the circuit
if the safe maximum is exceeded. My own
{Fig. 32) was obtained second hand for
the. proverbial song some time ago, but
even new ones are not prohibitively standard ranges of muffles) and these
expensive — about the same cost as a gas- need no more than encasing in insulating
fired one. Mine has a chamber 5 in wide x materials and a metal or "Syndnyo'" (or its
4 in x Gin deep and is rated at 1000°C equivalent) case. Heating elements are
maximum, though most today go up to also available, either in the form of "Hot
1200?C. Rods" or "Heating Bricks" — refractory
Unlike the gas-fired type, it is relatively bricks with elements embedded in the
easy to make your own electric furnace so face — or as refractory tubes with an
long as you take the normal precautions element wound round. Home construction
when dealing with lethal mains voltages. is dealt with in Chapter IX and the only
Ready-wound muffles are available from point | would make is that you should
such firms as Gallenkamps, Griffin & seek the advice of the element suppliers
George, and other laboratory equipment early in the project. Fig. 33 shows a
Suppliers (they are spare parts for their home-made furnace which is supplied in

Fig. 33 The Kanthal Electroheat electric fornace

This is supplied ws a kit for home construction Note
that an earthed shield has yet to be fitted over the ter-
minais. One of the heating elements is seen in front af
the furnace. (Courtesy Kanthal Ltd]

59
"kit" form by Kanthal Electroheat, former — but for the model engineer the
Inveralmond, Perth PH1 3EE, though as vertical electrical tube furnace heating à
shown it does need guards fitting over the pot made from steel tube is the most con-
electrical connections. venient. Gas heating presents certain
Salt Baths. | have already referred to the problems (other than those already
salt-bath for tempering. Neutral and non- mentioned) and an electric heater is easier
toxic salts are also available for the higher to control. The size of the pot depends on
temperatures needed before quenching. the work you want to do, of course, but
They must not be confused with the that shown in Fig. 34 is quite adequate.
ACTIVE salts used in cyanide hardening — The actual pot is about 3 in, internal
the constituents of the neutral salts are no diameter and is 10 in. deep, with the salt
more than common salt (sodium chloride) bath itself about 74 in. deep. (You must
mixed with a certain amount of potassium allow for the "tide to rise" when the work
chloride to give the required thermal is put in!). It is, in fact, a small home-made
characteristics. The type which | use laboratory furnace, and | give details of
(Degussa GS660/WS720) melts at the construction later. It is, perhaps, larger
670°C and can be used up to about than is really needed for model work but
1000°C; it is obtainable from the same you need room for the essential pyrometer
sources as the tempering salts. (p.45]. as well as the work.
The salt is melted in a welded steel For the initial melt the salt, which
‘pot’. In industry these may be large and comes as a fine powder, is put in to about
heated by gas or electricity — usually the 4 inches below the top of the pot and, as
this melts, more is very carefully added as
Soe ee zm
the level sinks, When re-melting subse-
quently a fairly heavy iron lid should be set
over the mouth, as a hard crust may form
when the salt sets on cooling and there is
just a risk of spitting. However, so long as
the top end of the heating element lies
above the salt level the crust should melt
from the outside inwards with no such
trouble.
The workpiece should be preheated a
little if it is likely to have water or oil in any
holes, and if any holes are deep or large
some consideration should be given to
these, as there is a risk of spurting of hot
salt as the air (or. worse, vapour) trapped
in the hole expands. The piece can then be
set in the bath, with a wire "handle" if
need be. (A job like a file or scraper can, of
course, have the tang protruding from the
surface]. The immersion of the work will
Fig. 34 The author's laboratory type Austenising salt-pot
turnece, Construction fx described on page 98. Note that the
"POISON" label! refers to previous use with a cyanide
hardening pot

60
 orarily reduce the bath temperature, constructed and be solid enough and
& is no need to adjust the controls heavy enough not to tip over, for the hot
ould recover very quickly as there is salt will set the floor on fire just as would
erable store of heat in the furnace. hot metal.
there the workpiece can be left for Furnace Control, There is no method of
escribed time which, for a salt bath, control of a gas fired furnace which is
e HALF THAT NEEDED FOR A reasonable in cost other than "manual". It
FFLE FURNACE; ie. 30 minutes per is possible to obtain modulating flame
of thickness. This is one of the great burners with thermostats but the cost and
ages of the salt bath — the other complication would be difficult to justify
that the heating is bound to be for the model engineer. Electric furnaces
orm and to the correct temperature. can, however, be controlled automatically,
fhe work can be quenched in water, in two ways. The easiest and cheapest is
ine,or oil. Most of the adhering salt will the "Energy Controller" — as is used on
all off’ on quenching, in the first two, the hotplates of most electric cookers.
| the rest will dissolve, This will do no This is, in effect, a simple time-switch.
Sait will, however, collect in the With the dial set at "O" the current is off
ottom of an oil bath and this must be all the time; set at "10" it is on all the
Ved out at intervals. As | have said, the time, In between the current is switched
on and off in proportion — eg. at "6" the
current flows for 6096 of the time, at "4"
| Same respect as that accorded to for 4096, and so on. This is very effective,
giten metal — see Ch, 10.) and debris can and spare cooker controls (or even one
disposed of down the sink with a good from a scrap cooker) are usually rated at
sh of water. In which connection, the about 3kw, which is quite adequate in
iterial DOES absorb water easily, and most cases. The one essential is a pilot
"salt powder should be kept in plastic lamp, for it is all too easy to "forget" and
js in an airtight container. For the same leave it switched on all night. Again,
son, a lid should be kept on the pot cooker controllers have a set of contacts
ennot in use. specially for such a lamp.
Naturally such a pot should not be used The proper "Thermostat" is far better,
Jarge tools. The salt level should not be of course, as the temperature can then be
is than three-quarters full nor should it controlled within close limits. The cost is
e more than an inch above the top of greater, of course, as some form of relay
| heating element — and certainly or "solid state" switch is needed as well
lould not approach the top of the pot. In as the thermocouple. This latter CAN be
Bsense a larger diameter and shorter the indicating thermocouple, but is usually
pth would help in this respect, but the a separate element. The cost and com-
iL of the tube and heating element then plication does depend, of course, on the
comes rather high. Safety precautions degree of control needed. For the sort of
re dealt with in Ch. 10, but | will heat treatment we are concerned with
)phasise one point now, The liquid is one which will hold the temperature to
as hot as molten brass, and you within +/— 8°C would serve and +/— 5°C
is as close as is necessary: all steelmakers
tw,ear, overalls and a cap. DON'T try to quote a temperature range for their
ig up” a salt pot; it must be properly products, usually 20?C wide. Quite

61
frankly, | have found that the only problem the end of a 3/32 in. diameter rod up ;,
with my energy controller type is quenching temperature quite happily ,
“forgetfulness”! | know pretty well where have already referred to the use of a Spir,
the dial has to be set for any particular lamp for tempering. For scriber points an,
temperature, but | have, on more than one even tiny boring tools the whole proces,
occasion, been called away from the shop can be carried out with this little fellow
and forgotten to check the temperature Handling. Naturally, you have to be ab),
before leaving. to hold the hot metal. | have already made
Spirit Lamps. From the sublime to the
passing reference to this from time tg
ridiculous? By no means! The humble time. In many articles | have read th;
spirit lamp is quite invaluable and, | would author has mentioned holding the wos
say, essential when dealing with very with “pliers”. This does NOT refer to th;
normal mechanic's tool which, being [if ci
small and delicate parts. It is cheap. it
provides a very clean flame, and it needs reasonable quality) hardened will soc:
lose their temper if so treated. The
no heavy gas bottles or trailing wires to
reference should be to “Blacksmith:
feed it. My own is one of the little glass
affairs with a round wick about 1 inch pliers” — the layman would call them
“Tongs”, perhaps! Fig 35 shows such :
diameter, but | do use one from a spirit
fired model which has a larger, flat, wick pair, and you will see that the nose ha:
from time to time. The glass one will bring almost decorative curves, This is to give
some slight spring in the holding. A ring
Fig. 35 A pat of
blacksmiths — "ptiers", like the link of a chain can be slipped ove
about 18 inches long the handles to hold the nose closed on the
work. Such pliers or tongs come in many
shapes, but that shown is the most useful
for heat treatment.
For smaller work the “laboratory
tongs” shown in Fig 36 serve the same
purpose. Usually made of stainless steel
they can be used to grip either with the
ends or within the "ring". It is; however
important to obtain only the best quality
even these do occasionally tend to “cross
their legs" and the cheaper variety (often
Fig. 36 Laboratory tongs.
about 8 inches long, far
sold for handling bacon in the frying par!
smälter work are useless. They can be obtained throug!
 reputable chemist, though they are programme is “occasional. The same
tely to be held in stock. In all cases the heat sources should cover the needs of
» of the tongs should be warmed annealing, and brazing torches are
gre handling the metal, and on no adequate for “open hearth” caseharden-
sount should they be wet, with either ing referred to in Chapter 6. Those who
ter or oil. make a lot of special tools, whose work
-situations where the use of “pliers” involves the hardening of jigs and gauges,
| possible then iron wire can be or who make a practice of hardening
d round to make a handle or wearing parts on their models (as we all
eaded through a hole if there is one. ought to do) will find that the simple
peFig. 28.) Remember that when the electric or gas-fired muffle will help a
1 gets hot it both expands, which may great deal, and the salt-bath even more
isen the hold, and loses strength. Take so. For tempering all but the simplest
& that it is secure, and proportioned to parts the low-temperature salt-bath is, as
‘size of the job. At the same time, use | have already indicated, far easier and
fe that is the thinnest gauge that will be more reliable than any other method.
quate, so that it will not interfere with In using brazing torches, however, it is
q enching action. There remains the essential to set brazing practice and
> of the part which is so small that brazing experience aside. You must "think
n thin wire would cause trouble. The of the atoms" as it were, all the time. The
swer here is to heat the work in a temperature is higher than is needed for
Jat" — just a piece of tube sawn in half— silver-soldering and is more important.
d touse a quench tank of such depth Further, you have to hold an even
It. no stirring is needed: the part will fall temperature for quite a long time — just
he bottom at a rate which strips it of the opposite to brazing requirements,
jam or vapour. The heat content of where long heating can cause problems
ise tiny parts is so small that the and where we deliberately get one part
ch is almost instantaneous anyway. hotter than the rest so that the alloy will
"follow the heat". Naturally, it is imposs-
ible to cover every eventuality or every
expedient in a chapter like this, but if you
nclusion. Domestic fires and the keep the "principles" outlined earlier
irmal brazing torch equipment will cover always in mind you should have no
S of the heating arrangements needed difficulty in devising a satisfactory scheme
e model engineer whose hardening for anything out of the ordinary.

53
CHAPTER 6

Casehardening
So far we have been concerned with work the carbon. Industrially the time is-suct
which needs to be hardened right through that a casing of appreciable thickness -
— tools, punches, chisels and the like — but 1/16 inch or more — will allow for subse.
there are many applications where quent grinding. For model-making such
SURFACE HARDNESS only is needed, depths are unnecessary and we seldom
the classic case being a crosshead pin. resort to grinding anyway, though we ma,
Such a component could, of course, be lap the surface to improve the finish. With
made from "Silver" steel, hardened, and thick cases there can be a dimensions
then tempered to reduce the brittleness: change, but there is no need to worry
However, the tempering would reduce the about this for the depth of carburisation
surface hardness and so the wear normal for model-making.
resistance — and we have had to use an Once the part has been surface car
expensive material. In "Casehardening" burised it is heated again up to the har-
we take a low or medium carbon steel (in dening temperature and quenched. We
industry. perhaps an alloy steel) and CAN temper it if we wish, but this is no'
modify the structure of the outer skin so necessary and even if it is thought desir-
that this can be hardened without unduly able to refine the grain a little, 100-140°C
affecting the core. The main body of our will suffice. However, we can modify the
pin will then have all the characteristics of process and so derive one of the grea!
the “mild” steel originally used but the advantages of casehardening ove!
surface is as hard as an untempered through hardening. We can machine the
hardened high-carbon steel. piece after carburising but befort
To bring this about we adopt a process quenching. This will remove the case fror
known as "Carburising". You will recall the machined parts. Quenching will have
the Chapter | that to make "Cast Steel" no effect on the machined surfaces, anc
from wrought iron the metal was heated these will be soft. This is quite norms
for a long period in the presence of practice, and many engine and machin:
carbon, so that the carbon content of the parts are designed to be treated in this
whole mass was increased. For casehard- fashion. Fig. 37 is a case in point — a pi^
ening the process is similar, but the with a hardened bearing surface and two
heating time is much less, with the result screw threads. If made from silver stec!
that the skin only of the metal takes up these threads would be dead hard unless

64
 Fig. 37a Piece of M.S, prepared for carburising. 37b; After
corburising, ane end screwed and the other plain turned. 376,
The same piece after reheating and quenching, 37d; The plain
end has been screwcut and the whole cleaned up. The bearing
surface is dead hard, but both screwed ands are soft.

pered, yet if they were tempered to

safe hardness the wear resistance of
pin will be reduced, perhaps unaccep-
ly. The pin shown is, in fact, made of
'asehardened mild steel. At "a" the pin is
ough machined at the ends, about 1/16
3 oversize, but the bearing surface is
inished. At "b" it has been carburised all
Wer and then the case turned away from
ends, down to finish size. One thread
| been screwcut. At "c" the pin is seen
r reheating and quenching — anti-scale
nt was used. Finally, at "d" you see the
hread cut on the other end (after
quenching) and the whole cleaned up and

heB tct that one was cut before and the

ler after quenching, yet the bearing
Surface is dead hard, and took the teeth
off the file. The hard skin was subsequen-

ere are other situations where

Basehardening can be used to advantage.
"or example, a complex form-tool can be
nade from mild steel and carburised.
er hardening it will have a surface
rder than tempered carbon steel. It will
not stand regrinding, but the tool life will
je slightly longer and if it is to be used
y the once we have saved a lot of
ensive material. Made from carbon
teel a pivot pin or a lathe centre will need
Io be tempered; casehardened — there is
"O need, and the wear resistance will be

65
greater. In ANY situation where the part is bonate) is the most easily obtainable. Ac,
subject to shock loading or where about 1% by weight to the charcoal ang
toughness is needed for other reasons, but mix it thoroughly. Alternatively you can
at the same time resistance to wear is add some of the “Open Hearth”
required, casehardening is the answer; compound, “Kasenit No. 1", which | shal|
you have the load bearing core in a state be mentioning later. About 595 will suffice
to take the stress, and the surface con- Other materials can be used — | have
ditioned to resist wear, successfully casehardened using olg
Carburising. There are three main leather bootlaces wrapped round the
methods in use, one of which, the active metal. for example! Presumably the
salt-bath, is not suitable for model tanning process provided an activator!
engineer's uses as it involves molten The work must be completely surroun
cyanide salts, which are lethally ded by the compound, which should be
poisonous. In industry the next most packed down hard. The lid must be sealed
common method is BOX CARBURISING. against ingress of air. It is then heated to
This involves the setting of the work in a between 880°C and 920°C for à length
heavy welded steel box filled with a car- of time which is determined by the depth
burising material. The lids are sealed or of case required. The first hour will
“luted” with fireclay. The compound used penetrate about 0.008 inch and a case
is specially formulated; largely charcoal, depth of 1/16 inch may be reached afte:
but with additives to increase its seven hours. (This assumes we are casing
effectiveness. Heating times are of the mild steel). For model engineer's work a
order of six hours or so. For our small depth of 0.008 inch should be adequate
parts nothing so elaborate is needed. | and a one-hour heat will be long enough:
have used stout tin boxes — they last for any greater depth of case may enforce a
two or three heats — or a piece of steel subsequent grinding operation. Note tha!
pipe (not galvanised) with one end fitted these times are times AT the carburising
with a screwed plug and the other sealed temperature, and you must allow some
with fireclay. (Note, fireCLAY, not firece- extra for bringing up the temperature.
ment, which will set hard; obtainable from The box is allowed to cool, and this can
most builder's merchants), The carburis- be outside the furnace if no subsequent
ing element is charcoal, but this, by itself, machining is required. The parts can then
is slow acting. An “Activator must be be taken out — and here it is worth
added and Sodium Carbonate (not Bicar- suggesting the inclusion of a blank test

Fig. 38 This case-hardened p”

has had a fat ground on the vc
amd has then been etchedto sho?
the hardaned case.

66
 ce. This can be reheated and quenched Frankly, this is totally unnecessary for
eck that sufficient hardness has been cases of the thickness we need — and
ched. | use a piece of round bar and seldom for thicker ones — but it is found
nd off the side to form a flat after sometimes that surface crazing appears
enching. This is etched with 2556 nitric during the grinding operation, and normal
idin water and, as seen in Fig. 38, the practice is to temper at about 150°C if
of case is quite visible. This is, of experience shows that this is likely.
purse, Magnified by the “circularity” of Open-hearth Carburising. Closed box
e specimen; the actual thickness CAN carburising is seldom needed for model
p calculated if you like, but | am, as a work, and even less often used, if only
le, content if the test shows the case to because it requires the use of a muffle.
B adequate!" Any "part-machining" can Though it can be done by setting the box
w be done. in an open fire; | have done this, and it is
We next have to harden the case, but quite effective, as the temperatures at this
casionally a further operation may be stage are not all that critical. However,
ssirable. We have held the metal well surface hardening on the smith's hearth
e its critical temperature for quite a has been used for well over a century, and
= longer than we might for heat is more than adequate for thin cases. In
ing a carbon steel. This may have this process the metal is brought up to
about 800°C (between 780 and 850) —
usually referred to as "bright red". The
' less, but if longer it can be worth atten- part is then covered with a special
; ng to. If the box was heated for 3 hours compound, either by dipping or by rolling
f more | would always refine the grain it about, until the surface requiring har-
e before hardening. This refining is dening is completely covered. The
lone by heating to a temperature which compound in contact with the metal
ipends on the carbon content of the melts, and carbon is absorbed. The
inal steel as under: process can be repeated several times if a
thicker case is needed. (A single "dip"
; Ca 01 02 03 04 provides perhaps 0.002 to 0.003 inch).
imperature°C 910 890 870 860 The compound | have used for 50-odd
years is Kasenit No. 1, and it is readily
Heat for the regular "one hour per inch obtainable from model suppliers and even
bf thickness" and then quench in oil at from good ironmongers.
bout 30°C. This is NOT a hardening Some practitioners quench
Mocess, but solely directed to reducing immediately if no intermediate machining
he grain size of the core. In fact, with very is needed, but | prefer to reheat to
between 760 and 800°C. | always use
vill suffice and there is no need to plain water, though brine is permissible.
luench. No grain refining is needed. The resulting
case will be very hard — some would say it
/50—780*C. The heating time is short, as is harder than with box carburising, but it
Ve are concerned only with the case; a is very thin. | seldom use less than two
W minutes is enough, and the work is applications and often three or four, but |
water-quenched. | have already have yet to find a case which was not
arked on the need for tempering. thick enough for the purpose intended.

67
 Naturally, if intermediate machining is be protected from the action of the
required it is necessary to allow the work compound by wrapping wire or shim stes!
to cool. This can be allowed in air, but ! round it. Tapped holes can be plugger
usually clap a piece of firebrick on top to with a screw, and plain holes by inserting
reduce the rate slightly. It is not critical at plugs. And, of course, if all you need to
all; indeed, there is little that is critical harden is the face of a tappet or the end c
about the whole process though naturally à tool it is easy enough to devise means to
the more carefully the temperatures are restrict the application of compound only
observed the better the results. An to that particular spot.
occasional problem is scaling. However, Steels for casehardening. Any “straight
we are dealing with a MILD lor at most, carbon steel can be carburised — you can
medium carbon) steel and there is no if circumstances require it, carburise silver
need in this case to heat slowly. When steel! As a rule, however, the process is
carburising it is only the skin that is normally applied to the low-carbon “mild
affected anyway. It is not possible to use steels or, at most, to those with carbon
anti-scale paint, nor a salt-bath, but for- contents below 0.3%, Free-cutting steels
tunately the compound itself does have can occasionally present problems — the
anti-scale properties. It is, of course, only inclusions which provide the free-cutting
prudent to remove any existing scale from qualities may cause surface crazing. |
black bar lif one needs to harden such) would not advise box carburising these
before starting. There is, of course, no materials, but have successfully surface
reason why you should not heat the metal hardened them by the open-hearth
in a little box surrounded by the method, Steels specifically designed for
compound. | have done this occasionally, (box) casehardening are available, These
when | have a number of very small parts usually contain about 196 of Nickel, the
to deal with at once. It saves time. You do object being to reduce the grain growth
have to be sure that the contents are up to during the prolonged heating. Straight
temperature, of course, and you should carbon steels intended specifically lo:
use a "thin" (shallow) box for that reason. casehardening by either process are Er
But such a procedure does rather spoil the 32C (080M15 in BS970/1972) which is
innate simplicity of open-hearth carburis- 0.15% carbon, 0.896 manganese, and En
ing, and you are unlikely to get better 21 (130M15) which is 0.1596. carbon
results by using a box, It is, occasionally, 1.396 manganese: somewhat stronger i!
more convenient, that is all. 40 ton/sq.in tensile strength.
Now and then there may be difficulties Alloy steels should not be casehardened
in applying the compound — on a ratchet without detailed heat treatment specifica-
wheel, for example — to be sure that it tions from the makers. All such steels are
works on the desired faces. The "heat treatable" and you may well set up
compound can be wetted with water and adverse conditions during the carburisina
applied as a paste. It is then heated very process if you do not get the temperature
gently to dry out, after which the proce- right. Never try to caseharden scrap stee!
dure is as described above. There are or offcuts from the machine-shop skip
situations where secondary machining They may be anything, from low grade
might be needed with box carburising but steel for making decorative handrails t°
can be avoided with the open hearth chrome-nickel-molybdenum-titaniur
method. A screw thread, for example, can alloy! On the other hand, wrought iro?

68
 sily identifiable by the spark test) pleted and parted off, after which the
sehardens beautifully, and you CAN flanged end was finished. It was then
gharden cast iron! Obviously you must reheated and quenched. There was no
è regard to the shape of the casting — difficulty at all in riveting the spigot on
mplex outlines mày crack in the quench, assembly.
t hardening the end of a plunger or even Top right is a different situation, This is
alve-rocker presents no problems. a tap needed for finishing the internal
«amples. Fig. 39 shows a few examples. threads of the boxwood chucks used on
ip left is a ball-joint made to replace my Holtzapffel lathe. The cost of a length
A| carburettor control on an ancient of silver steel this size would be prohibi-
tive for the amount of work it has to do.
sec and finish machined completely Further, there would have been some risk
€ sre hardening. The thread was of cracking even if great care were taken,
‘otected with thin iron wire and whiten- as the threads have sharp roots and a fine
g. and only the ball end was dipped in angle of 50°. It was made of black mild
psenit No. 1. It was quenched direct steel (ie. hot rolled, not bright drawn) so
er the second application. Centre top is that there was no need to normalise it
first. It was given a single treatment with
- Williamson") with à hardened Kasenit No. 1 and quenched directly. The
prankpin. In this case the bearing surface cutting edges were very lightly stoned
yasfinished to dimension, but the stem afterwards. It has served its purpose with
hich fits into the web was left 1/32 inch no difficulties.
oversize. The workpiece was not parted Bottom left is a drilling jig for some
ff from the stock at this stage. The cylinder covers and the associated cylin-
Baring was given the Kasenit treatment ders, the number going through the shop
at , being dipped into the compound justifying the use of jigs. It was made from
nly sufficiently to cover it. After cooling Bright Drawn steel which was first nor-
owly the rest of the machining was com- malised before machining. It was
carburised for 30 minutes in a tobacco tin ventional tool steel. And contrary to
using a mixture of charcoal and Kasenit supposition, such tools have a harder time
No. 1 compound. This was done to ensure turning these exotic woods than when
that the inner surface of the holes was turning most metals, for the only route for
hardened. Bottom centre is the grooved the escape of the heat is through the tool
jaw for a drilling vice, which was — and, of course, no coolant can be used.
casehardened "open hearth", the part Conclusion. Casehardening is not a sub-
being laid, face down, in a tray of stitute for through hardening with a high
compound. Three doses were applied. carbon steel, but a process which earns its
There was a little scaling on the back, and place from its very nature. We can design
just a very little distortion. It would have a component in material which provides
been wiser to normalise it before machin- the strength, fatigue resistance, or other
ing (BDMS was used) but the fault was property needed and then surface harden
not sufficient to warrant the making of it to give the required wear resistance. The
another. nature of the process is such that selective
Bottom right is an example of a special hardening can readily be applied, and the
form tool. This was made from mild steel risk of cracking is very much reduced. Use
for no other reason than that | had no of the open hearth method gives quite
high-carbon of the right size and only a adequate thickness of case for ou:
limited stock of a size which could be purposes and is much quicker than
machined down. This was given doses of through hardening, with the need for sub-
Kasenit No. 1, allowed to cool, and then sequent tempering adding to the time
reheated and quenched. The compound taken. At the same time, casehardening
was applied only to the tool tip. The result can be used for such things as lathe tools
was outstanding. The case was thick etc. if circumstances demand. One final
enough to permit light stoning of the top point is worth making. There is often no
face to give a sharp edge, and the perfor- real need to harden a rubbing surface, bur
mance {ornamentally turning gritty hard- if it IS hardened and polished the friction
woods) surprising. As no tempering was loss will be reduced. The almost glass
needed the cutting edge was harder than hard surface produced by casehardening
could have been obtained with a con- is a real advantage in all such applications.

70
 ZHAPTER 7

Other Heat Treatment

rocesses

(1) Annealing, Normalising, and Stress In the case of low carbon and medium
Relieving. These three processes are often carbon steels (below 0.85%) this will be
confused. They are carried out in similar about 30°C above the upper critical, but
ways, but the three purposes are different. for higher carbon content a temperature
Annealing is the softening of steel pre- somewhat above the /ower critical will
suffice. The recommended figures are
ment or by cold working. Normalising is a given in the table below. (See also Fig. 41),
process where the internal structure of the In each case the higher temperature
steel is restored to "normal" after some relates to the lower carbon content. None
previous operation — say forging. Stress is critical to within + 10°C.
The metal must be heated slowly and
of stresses which may have been locked then “soaked” at the annealing
Ip inthe steel, either by heat treatment or temperature. for one hour per inch of
by cold forming. "Tempering" is a special thickness — in the case of rectangular
orm of stress relieving, We will deal with sections the smaller dimension is taken. It
these processes in turn. should then be cooled as slowly as can be
Annealing. The steel will be Martensitic contrived. If heated in a muffle with a
in structure and it is necessary to change good heat insulation to the chamber it
this to one containing pearlite and either may be allowed to cool down in the
ferrite or cementite, as it would have been furnace, Otherwise it must be set in hot
had the steel been cooled slowly from the ashes or some similar material which will
jitial austenitic state, This means that we retain the heat. Thin, flat, sections can be
must first bring it to a temperature high cooled between two pre-heated insulating
ough to ensure that the steel is above bricks. The higher the carbon content the
he transformation temperature, and then more important is the slow rate of cooling.
pol it at a rate that will allow the trans- For "mild" steel the piece may be cooled
ormation to pearlite. in air once it has dropped to around
"Silver
arbon,% Below 0,1 0.1-0.3 0.3-0.5 0.5-0.7 0.7-0.9 Above 0.9 Steel"
emp.*C 920-900 870-830 850-800 820-780 800-780 790-760 770°C

71
 These must be taken right up into the ful|
austenitic region. Fortunately tool steels
seldom require normalising, which is
needed chiefly to improve the
machineability of medium and low carbon
(and alloy) steels, The following table
shows recommended temperatures.

Carbon % 0.1 02 O03 04 O6

Temp, *C 920 900 880 860 840
OB 10 11 12
820 830 900 925

Fig. 40 Elongation of the grains of steel caused by 10°C up or down from these
mechanical
forming processes.
temperatures will serve. The temperature
400°C, but it is inadvisable to quench it must be maintained, as in the case of
even from a low temperature. Note that annealing, for one hour per inch of
work must be fully preheated before section, and the cooling must, again, be
immersing in a salt-bath for annealing. slow.
Annealing high speed stee/ is a chancy Fig. 41 shows the annealing and nor-
business for the amateur, especially as malising range on the iron-carbon diagram
any subsequent hardening may well be Stress Relief. Certain manufacturing
beyond the capabilities of the equipment processes induce considerable stresses
available. However, if such an operation within the metal. in particular cold
must be carried out the piece should be drawing and welding. The former sets the
heated VERY slowly to 850°C, heid there surface skin under compression and there
for one hour per HALF-inch of section, and may be a locked-up tensile stress in the
cooled very slowly indeed; a muffle core. The result is that when any part of
furnace is almost essential, and the the surface is removed symmetry is lost
cooling can be best effected by and the shape will distort. In welding, part
progressively reducing the power input of the metal is brought up to very high
until it has fallen to around 550°C before temperatures indeed and differential con-
turning it off completely. traction between this and the cold or
Normalising. After forging or rolling a relative cold parts can induce stresses -
metal the natural shape ofthe "grains' will often sufficient to distort during the actual
be distorted — see Fig. 40. The situation welding. The degree of stress relief
may be even worse if some of the forging needed is different in the two cases. Cold
or leg.) bending has been done at below drawn or cold rolled steel may be relieved
the proper forging temperatures or when sufficiently to prevent distortion during
the steel is bright-drawn. To reform and machining by heating to between 400*C
refine the grain structure it is again and 500°C and allowing the part to coo!
necessary to reheat to above the critical in air. It is important that you ensure even
temperatures but to a rather higher heating. otherwise the last state may be
degree than is needed for annealing, worse than the first due to thermal expan-
especially for the high carbon steels. sion differences. For welded components

72
 ig. 41 Approximate annealing and normaliising temperature ranges superimposedon the irón-carbon equilibrium diagram. This
an be used inconjunctionwith fig.13 ifnodefinite data are avalable.

(l am referring to the small parts used in binding with wire and encasing in whiting,
models, not to home-made bridges or or by the use of charcoal in the muffle if
radio masts!) it is best to give a full anneal the latter is used. However, for small parts
as the condition in the vicinity of the weld an ingenious method is described in some
is quite unknown. early books. The parts to be heated are set
. Parts which have been heat treated in a box filled with slaked lime (“builder's
will, of course, be under considerable lime") mixed with up to 596 of charcoal.
intercrystalline strain — this is part of the The latter absorbs any oxygen present, but
hardening process. The subsequent is not sufficient to cause any carburising.
lempering process is a form of stress Another such recipe suggests the use of
relief, even though in this case we also fine cast-iron turnings as the packing
Bim at some controlled transformation as medium. The former method does work,
well. but the heat insulating properties of the
(1) Scaling. In all the above processes lime enforce along heating time.
Some surface scaling is to be expected, (2) Rehardening. In the course of time
ind this can be a nuisance. It can be pre- tools may be reground so far that the tip is
vented by the use of anti-scale paint, brought back into the more deeply

73
tempered region. This applies especially to His problem is twofold. First, there is NO
woodworking tools, but cold chisels and means whereby the incised decoration
scrapers suffer in the same way and it can can be polished. The surface must be
apply to sliderest tools as well. When this reflective with "tool finish". This surface
happens there is no option but to finish can never be better than the finish
reharden. It might be thought that it on the tool, and for this reason the
would be necessary only to reheat, sharpening of the tool is taken to what
quench, and temper again. This is well might appear to be extreme limits. After
enough with "coarse" tools made from forming with normal grinding methods
relatively low carbon steel; | treat my and a fine India stone the surface is
pickaxe, wrecking bar, and similar tools in "rough polished" using hard Arkansas
that way. But if you try to reharden a high- stone, the tool being held in a special jig
carbon steel tool in this fashion the results (the "Goniostat") which ensures that both
will, at best, be disappointing and you the cutting angle and the "pattern angle”
may find that the tool cracks. We must are accurately formed. It is then lapped
restore the metal to the condition it was using the finest oilstone dust on a brass
before the initial hardening process was lap, and finally polished (still using the
carried out, and this means that it should goniostat) on an iron lap with jewellers
be fully annealed first. | have already dealt rouge. The problem is that modern tool
with the process and there is no need to steels will not give the refined grain size
repeat it. However, | can add a refine- which was normal with the steel used
ment. If the tool of some importance is to when the art was being developed, and
be rehardened it will pay to re-temper to the common complaint is that "the steel
about 300°C before commencing the won't take the polish".
annealing process. This will effect an The second difficulty is that such a too!
almost complete stress relief to the may have to make three or four hundred
structure and | have never had a tool crack "cuts" each at a different setting of the
when this procedure has been followed. | apparatus. It is quite impossible to remove
recommend it strongly for those tools it for sharpening once the process has
where there is considerable difference started. As previously mentioned, the only
between the temper in the shank and in route by which the heat of cutting can
the point. It is only necessary to heat up escape is through the tool; the workpieces
very gently to "blue", allow it to cool just a are very good heat insulators. This means
little, and then to proceed with the full that the tool must remain sharp and.
annealing heat. hence, hard over a long period — and any
(3) Hardening Ornamental Tools. The loss of sharpness can be fatal to the
"Ornamental" turner has a problem not effectiveness of the work. These two
met with in most other types of turning. requirements, finest grain size and
The art comprises the cutting of fairly maximum hardness, cannot be achieved
deep patterns on the surface of work pre- with the normal carbon tool steels (and
viously turned by normal methods to àn even less so with high-speed steel). It |S
acceptable shape. The final effect depends possible to achieve the hardness required
entirely on the reflection of light from the by using a file steel. After annealing at
multitude of facets so formed — and with about 770°C (and on no account above
well executed work almost dazzling reflec- 790°C) the tool can be hardened from
tions can be seen on a wood like Ebony. 760°C/780°C and quenched in brine. The

74
Subsequent tempering should be as mild cooled) conditions consists of grains of
as possible at the point — | do no more pearlite surrounded by ferrite; a tough
than boil in water for half an hour for materia! set in a matrix of soft material.
Cutting frame tools, but sliderest tools The resulting steel is fairly tough and
must be tempered at least to straw in the strong — "better" (for some purposes)
shank. than mild steel. Suppose, now that we
| Alternatively, with "Silver Steel" a were to quench this steel from a high
grain refining process can be carried out. temperature? There is insufficient carbon
and this will give quite satisfactory results present for total transformation into the
— even if the purists will say "Still not as form of iron carbide we know as Marten-
good as a genuine Holtzapffel"! site, so that the structure will, in effect, be
_ After machining the piece to the that of hard Martensite "diluted" with
required section and rough forming the ferrite, It will possess some of the
necked end lif any) heat to between character of a high-carbon steel (hardness
820°C and 850°C for one hour per inch and toughness) and some of those
of section. In the case of tools with the associated with "mild" or low-carbon
9/16 inch square shanks the "ruling steel; ductility and softness. Further, we
Section" is the inch or so behind the point. can temper the steel. This has a consider-
Quench in oil at about 20-25°C. The able effect, as in the case of high carbon
working end may now be ground — steel, and in this case the selection of the
carefully — almost to finish size, but in the best temperature is the most important
case of the heavy sliderest tools 1/32 inch part of the treatment. The mixture of
should be left on. Reheat slowly again, but tempered Martensite and ferrite crystals
to 760-770°C only, and quench in brine can be adjusted to give higher strength,
at 15-20°C. Temper immediately to not high yield point, better impact resistance,
more than 150°C for at least 30 minutes, and so on. Fig. 41 shows the effect on a
and then "let down" the shank of sliderest typical 0.4% carbon steel, and you will
tools with the actual cutting point kept notice that there is a dip in the Impact
cool in water, At this temper very little, if resistance (Izod Impact Test) in the middle
any, of the hardness will be lost, but the tempering range. This need not worry you
hardening stresses will have been too much, as it is still considerably higher
relieved. The metal will have a very fine than that of the unheat-treated steel.
grain and be as hard as possible. To give an example, let us consider our
(4) Heat Treatment of Medium Carbon old friend EN8. (0B80M4O in the BS 970
Steel. So far we have been considering 1972 nomenclature) In the normalised
the “hardening” of steel to make cutting condition typical properties would be 38
tools and the like, but there are many T/sq.in. UTS; 27 T/sq.in Yield; Elongation
other cases where some form of heat (ductility) 28% and an lzod impact
treatment will improve the quality of the resistance figure of 15-17 ft.lb. The metal
material in other directions. Indeed, by far is Austenised by heating to 830-860°C
the greatest amount of heat treatment in for 1 hour/inch, and then quenched in oil.
industry is directed to this end. The tempering temperature must then be
If you refer back to Fig. 13 (and, selected according to the properties we
perhaps, refresh your memory from the need. For maximum toughness and the
associated text!) you will see that (say) most arduous fatigue and shock loading
0.4 carbon steel under normal (ie. slowly conditions this would be between 550

75
and 650°C and then air cooled. (Note that malised steel.
tempering salt will not operate at this For completion, the normalising
figure and the work must be done in a temperature is 840-860*C, but if you
muffle). At 600°C the UTS would be have EN8 which is bright drawn from the
around 42 T/sq.in, Yield point about 28 T/ Oil Hardened and Tempered condition
sq.in, Elongation perhaps 2996 — not a (O.H.&T. in the stockholder's list) it is, in
great change. But the impact resistance fact, in about its toughest condition, with
has gone up to 80 ftlb — around four a yield point of around 46T/sa.in. and an
times as great as before. The fatigue Izod figure of about 60. You are not likely
resistance also will be improved. If we to better this. However, most ENS is
tempered at 400°C we would raise the supplied to model engineers in the bright
UTS to about 50 T/sq.in, but both ductility drawn state with no heat treatment, and
and shock resistance would be less — normalising may be prudent before
though still better than that of the nor- working on it.
Fig. 42 The intivence af tempering temperatures on the properties of a rypical 0.4% carbon steel after oilquenching from about
850*C.
IZOD IMPACT
A TEST FT.LB.

d |

N we
d 4 o | ULTIMATE
TENSILE
! STRENGTH
TONS /SQ. IN.
s
ELONGATION- 9j.
L 30. |
(DUCTILITY).
z
a

AL |
> 20. —— e

500.
400. 600.
76 TEMPERING TEMPERATURE- °C.
 ilar treatment can be carried out on Handbook" and in App.3, and | would
nost all steel with a carbon content strongly recommend that you start using
e 0.2%. But it is MOST important these new numbers — not least because
steel is just not being made to some of the
v eating and what treatment to give it. old EN numbers any more. (This includes
fortunately most firms now send their a favourite of mine — 3% Nickel steel
otk out to specialists and published data EN21!) The "new" British Specification
hard to come by. If the steel can be has been in use now for over 10 years,
entified with an “SAE number" (Society after all.
F Automotive Engineers, USA) then (5) Coil Springs. We make the majority of
n the SAE "Handbook" or our coil springs from wire which is already
hardened and tempered. The act of coiling
them round the mandrel stresses these
3neer is in selecting the required beyond the elastic limit (the springmaker's
mpering temperature. It is on this, rather term for this is "scragging"!) and this
an the initial quench, that the final actually improves the strength of the
operties depend, and if a range of material. However, there are occasions
when this is just not possible — the gauge
@sents a puzzle. Perhaps you might of wire is too thick for the diameter of the
ink it safe to work at the midpoint of the mandrel, In these circumstances it is
nge, but in the absence of specific data necessary to soften lanneal) the wire,
wind the spring, and then harden and
& improving. So, for any heat treatment temper again. Unfortunately we seldom
'medium carbon or alloy steel it is best have details of the steel of which the wire
her to do nothing, or to go to the is made. It may lie between 1.0% carbon,
untain head. The makers of the steel can 0.496 manganese, and 0.796 carbon, 0.896
ve you chapter and verse, and if you give manganese. It should be annealed at
é British Steel Corporation the about 780?C; this is not critical, but the
ecification number and the application small mass of the wire does mean that
iu have in mind they should be able to some care must be taken over the cooling.
avide you with detailed figures both for This wire can air-cool fast enough to re-
tial heat and for tempering. It is for- harden. The spring can then be wound,
and | suggest that you allow either for
3d high tensile or high impact values — subsequent stretching or compression to
el hope that the data | have given for "scrag" the spring after heat treatment.
Close-coil the ends and either flatten them
ling better than “BDMS” is needed. on the grinder or form the loop, as
d should, perhaps, add a final note, The required, at this stage.
d BS970/1955, with its "EN"* Direct hardening with a flame is most
imbers, was replaced in 1972 with a unwise. There is risk of getting some part
Ww, and far more logical, specification too hot and it is impossible to be sure of
ference. (The number itself, in most even heating. Set the spring or springs
ises, tells us the carbon content). Details inside a steel or copper tube and heat this
given in "Model Engineer's to 780°C from the outside as evenly as
you can. The springs can then be tipped
, means Engineering Number) into the oil-bath — there is no need to

77
 CORE QUENCH LINE.

SKIN
UE NCH
UNE “TEMPERED
MARTENSITE"

MS.
MF.
MARTENSITE.
Fig. 43 /dealised disgrarm of the norma “Hot Quench and Temper” hardening process. There is considerable thermal shock 23
the metal passes through the "MS" to "ME" range when Martensite is being formed,

agitate for so small a mass of metal and spirit flame and allow the oil to burn away
the spring falling to the bottom will be suf- Quench in oil when the flame goes out
ficient. You must now be very careful, for and then repeat if necessary, as it almost
at this stage the spring will be very brittle certainly will be for heavy gauge wire. It is
indeed — | have known fine ones even almost essential to make a trial spring it
break under their own weight. this procedure is to be adopted.
The tempering does present difficulties. The salt bath is the ideal method, oí
It is almost impossible to use the "colour" course. This gives absolute control and if a
indication, as any attempt to polish the little wire "spray" — like an umbrella
spring may result in breakage. The upside down which has lost its cover — is
required temperature is quite high — 300- made a number of springs can be
310°C — well above the limit either for the tempered at once. There is no need to
domestic oven or for boiling in oil. quench from the salt bath, but once the
"Blazing off', referred to earlier, is fre- spring has cooled somewhat it can be
quently resorted to. With experience (or immersed in hot water to remove the salt
prior experiment) it is effective, but let me adhering to it.
repeat that the flash-point (or even the fire An alternative is the sand-bath, 4
point) of the oil has nothing to do with the process which can be very effective fo!
tempering temperature; this depends leaf springs. The bed should be fairly deep
entirely on how much ail is retained on the and kept turned about as it is heated. 4
spring. If there is not enough to bring the test piece of similar bright steel is set on
wire up to 300°C then the process must the sand (a piece of the same sprind
be repeated, and repeated immediately. material is best) and as soon as the
Only a trial can tell whether it is temper colour appears on this the springs
necessary. However, it does work, and the can be arranged in the bed and hot sand
procedure is to extract the spring from the heaped over, the heating lamp or torch
quenching oil and then hold it in the flame being removed at the same time. Sand is
of the spirit lamp until the oil ignites; as not à very good conductor, so that ample
soon as it does so move it further frorn the time must be allowed.

78
 | Austempering and Martempering. constant temperature, as it passes
e are fairly advanced methods of heat through the two borders of the "S", After
ent, much used in industry. Mar- a time which depends on the period A-B,
ring is of little use to the model and the size of the work it is removed from
neer, but Austempering can help, the hot tempering bath and allowed to
ecially with the case we have just been cool naturally in air. For silver steel the
scussing — the heat treatment of small time for "spring temper" would be 16
springs. The process can best be under- minutes at around 320°C and 25 minutes
stood by reference to a few diagrams. Fig. at 300°C. The time increases markedly, at
43 shows the normal process of lower temperatures — 45 minutes 'at
uenching and tempering, displayed on an 275°C and 2 hours at 250°C. No harm
"curve. The tempering period is not, of will be caused by slightly longer periods.
ourse, in any way connected with the The end result is not tempered Martensite,
ime scale of the "S", because the trans- but a new structure (called Bainite, but
ormations which occur are not the same don't let that worry you — it doesn't matter
is those within the "S". The diagram at all) which has very similar properties.
hows a high temperature temper, as The great advantage of this method is that
night be used for a spring, and the there is no thermal shock during transfor-
nateria! will end up as tempered Marten- mation to Martensite (or Bainite in this
strong, tough, but not as hard as is case) as there is with conventional
e@eded for cutting tools. quenching. Almost all commercially made
In Fig. 44 the steel is quenched from springs — and such things as roll-pins and
‘similar high temperature, but in a hot the like — are Austempered in this way.
alt bath, somewhere between 280°C The Martempering process is shown in
ind 330°C, not cold water. The Fig. 45, just for explanation, as its main
wenching temperature is higher than that use is when heat treating very thick or
it which the Martensite forms, and the massive components. The piece is
ransformation occurs along a line of quenched from the high temperature in a

"Fig. 44 “Austempering” or constant temperature transformation. Thermal shock is eliminated and no tempering fs needed.

CORE.

TRANSFORMATION
AT CONSTANT
TEMPERATURE.
SALT OR LEAD
-—— BATH
—— — —— —— TEMPERATURE.
"BAINITE"

79
salt bath as before, but instead of remain- large crankpin in ENB. In this latter Casa,
ing there for transformation it is taken out however, do remember that the maximum
as soon as it has reached the temper bath working temperature for a tempering salt
temperature and allowed to cool in air. It bath is 500°C, and unfortunately you
is subsequently tempered. The advantage cannot quench the piece by setting it in a
over the conventional method is a reduc- muffle furnace! However, most steels wil|
tion in the risk of distortion. There are no give good heat treated conditions at this
advantages to be gained for model temper.
engineers, as even the heaviest of our A few practical notes. First, this is a
workpieces would be regarded as “tiddly Heat Treating process, not a tool.
bits" in industry, and would be hardening one, At Austempering
Austempered. temperatures below about 280°C the
The procedure (say as applied to time required for the transformation
springs) for Austempering to heat to a increases quite quickly, and might be as
somewhat higher figure initially, say long as an hour at 250°C. For Austemper-
900°C. The salt bath is prepared and ing pieces requiring a higher temperature
brought up to the same temperature as than can be achieved with the salt a lead
would be needed for normal tempering — bath may be used (lead melts at about
300°C for a spring. As soon as the 330°C). The high temperature salt bath
workpiece has been “soaked” for the can be used for initial heating if desired
necessary length of time it is very quickly and no harm will result from the carry-
— with the absolute minimum of delay- over of this salt into the tempering bath
transferred to the hot salt bath. It is left However, it is most important that no
there for a length of time which depends LOW temperature salt ever gets into the
both on the type of steel and its thickness; high temperature bath; as detailed in
20 minutes will suffice for even the "Safety Precautions" a rather serious
heaviest spring, and 30 minutes is all that reaction can occur. Naturally, the high-
would be needed of heat treating even a temperature salt will solidify at the
Fig. 45 "Martempering" This process is used for large or complex work where freedom trom distortion is important, The work
needs subsequent tempering.

CORE.

SKIN.

TEMPERED
QUENCH MARTENSITE
TEMP:

80
 mpering temperature and will collect in degreasing should be followed by washing
ne bottom of the bath. This should be with hot detergent and water, and then air
moved from time to time, but for the dried, after which it should be handled
mount we use it is, perhaps, best to only with clean tissue paper; the "acid
n the tempering bath altogether free" grade used by jewellers is best. A
ien so much H.T.salt has set in the bath spirit lamp is best for small parts: larger
bottom as to interfere with heating. may be heated in a sand bed or in one of
As | remarked at the beginning, these the improvised muffles already referred to.
fe specialist processes, and do need a Direct heating with a torch flame is
Ot quenching bath. Model engineers unwise, as it is difficult to get even
ave managed without them for almost heating.
OO years, so that you need not worry More decorative parts, as for screws in
bout deciding to ignore them! However, clocks and parts of instruments, or even
you HAVE got the equipment and the clock hands, must be highly polished first.
pad or salt, then it is worth a few experi- fine emery followed by polishing paper,
ients, especially if you make springs (leaf until the finish is satisfactory. In this case
if coil) in any quantity. The process is there are differences of opinion; some
deal for small, complex parts made from hold that the final strokes of the polisher
medium carbon steel, which might distort should be "in line" (or, for screwheads,
‘dealt with in the conventional manner. done in the lathe) others hold that a
7) Cosmetic Heat Treatment. Not all random effect is more pleasing. You must
leat treatment is done for strength, try both and see which suits you best. It IS
lardness, or other mechanical properties. a matter of taste.
can be applied both to help resist corro- In all cases after heating to blue the
iion, and — especially in instruments and parts are quenched in oil; not for any
ks — for sheer decoration. In model metallurgical reason, but because this
vork we may need to simulate the “as provides quite an effective rust-proofing.
ge condition, especially when the The combination of the oxide layer and
prototype is an early one, the oil seems to repel corrosive fumes. As
Heat Blueing of steel lor "Browning" as to colour, this has to be watched fairly
the gunsmiths call it) is relatively easy — it carefully, as the "tints" of blue follow very
5 only necessary to heat the component closely. However, | find that for workshop
to just over 300°C. However, the results tackle (which | "blue" for preservation
Will be less than satisfactory unless the rather than appearance) a shade of deep
Ireparatory work is well done. In the case brown is not unpleasing, whilst many of
Where the part is to simulate “fine my very early 19th century tools are straw
orgework" on the prototype it is a coloured. They are not carbon steel
nistake to work for a polished finish. The tempered, but look as if they have been so
iurface should be fine matt, as produced treated.
by a rotary motion of an abrasive, the Blacksmith's Oxide is very difficult to
grade depending on the size of the part. simulate. This hard, corrosion resistant
fhe important point is to ensure that the surface is the result (in full size) of some
Surface, once prepared, is free from oil and time at forging heat (1000*C) coupled
rea se and especially fingerprints. The with forge working. If a piece of mild steel
atter will show up distressingly clearly is held at this temperature for any length
fter the heat treatment. Chemical of time some grain growth is inevitable. It

81
doesn't matter if the piece js being forged, 1050°C, and forging must stop when j;
as this will correct matters, but it is falls to about 900°C. These temperatures
difficult to do this with mode! parts. If it is can be used right up to about 1,0%.c. Fo,
practicable, the following procedure will 1.196 carbon and above, including silver
give tolerable results. Arrange the steel, the temperature should be som;
manufacturing programme so that any 50°C lower, but the higher the carbon
parts which matter (eg. the journals and content the more care must be taken to
crankpins of a crankshaft) can be avoid overheating. My experience is tha:
machined after the heat blacking is done. with $ in. square tools | can just get two
Bring the "forged" part to size and heat to blows in with the hammer before reheat-
bright red, giving a few light taps with a ing is needed. | always give a full annea|
flat hammer to "distress" the surface. Dip before hardening after any but the ver,
in oil (don't quench right out) and again lightest forging operation. Ordinary mild
reheat, just for a few moments. Repeat steel can be forged much hotter — up to
the process, going up to bright red each 1250°C according to some authorities ~
time. Once satisfied with the "finish", nor- and is less subject to damage if foraed
malise the steel as described on page 72 cool, though | would prefer to normalise if
and then carry on with the machining. much “black forging” has been done. In all
taking care not to damage the finish. cases overheating can ruin the metal, and
Naturally if you can effect any actual carbon tool steel is especially at risk
forging, so much the better, There is no way in which a piece of over-
(8) Forging. !t is, perhaps, apposite to heated steel can be redeemed except by
deal with this now, having just remarked melting it down.
on the difficulty of forging model compo- For detailed instructions in the use of
nents! Most books on “Tempering & Har- blacksmiths’ tools the reader is referred to
dening” published in the past lay great one of the many books on “Wrought
stress on the forging of lathe and shaper Ironwork” (though “iron” is seldom used
tools. This is very seldom done — or these days) — it is a specialist craft in its
needed — nowadays. Carbon steel tools own right. About the only use | make of
are used only when High Speed tools are forging tool steel is to reduce square
not available. Further, the size of tool we section down to rectangular — less
now use is much smaller than obtained wasteful than machining it — and this does
even 50 years ago, when half-inch square not call for anything more than keeping an
shanks were common. Even a à in. square eye on the temperature. Forging down the
piece of stock will barely hold the heat end of a pickaxe or crowbar (about 0.8%
long enough to give more than a couple of carbon) is easier, using à 5 Ib. hammer
hammer blows, and with 5/16 in., the though my anvil, at 56 Ib., is rather light
more usual size these days, proper forging for such work.
is almost impossible, However, there is (9) High Speed Steel. The hardening of
the odd occasion when a tool must be high-speed tool-steel is really beyond the
slightly cranked, or a fly-cutter needs capacity of the majority of mode!
bending. engineers. To begin with, there are con-
High carbon tool steel must be held to siderable variations in alloy conten!
fairly close temperature limits whilst ranging from 0.5% to 1.096 carbon, 0.5%
forging. At 0.796 carbon, as might be used to 4.096 Chromium, O to 1896 Tungsten
for a special screwdriver, the maximum is and some with 5% Cobalt — and they may
82
 clude varying proportions of Molyb- is done before the tool has cooled right
num and Vanadium as well. Each down — from perhaps 150°C. The typical
mands its proper heat treatment and an tempering temperature is about 550*C,
yproximation is not good enough — either but in some cases, especially for tools
g tool will crack or it will not be hard requiring considerable toughness (for
ough. roughing work on castings etc) a second
Second, the temperatures are high, with tempering is carried out at (typically)
pw initial heating to the range 780- 325°C;
380°C followed by a fairly quick heating The occasions when a high-speed tool
pto 1200 or 1300°C depending on the requires re-hardening are rare, and unless
nalysis. Most can then be cooled in air, the exact analysis is known, and the
_ some require a partial oil-quench workshop has precise temperature control
own to 500°C followed by air cooling. of the furnace, this work is best left to the
ampering must follow immediately, and specialist. Even he will probably decline
gain there are wide variations in the the work unless the source, type, and
quirements — in some cases tempering grade of metal is exactly known.

83
CHAPTER 8

The Measurement of
Hardness

It is unlikely that many mode! engineers The majority of hardness testing done
will be concerned with the actual today depends on the resistance to
measurement of hardness, but it is helpful penetration of the metal. In the Brinell test
to know HOW it is done when faced with (Fig. 46) a hard steel ball, 10mm
one or other of the several "hardness diameter, is impressed on the specimen of
numbers" used. The main difficulty is in metal under a load of 3000 kilograms
deciding exactly what to measure, for (about 3 tons) The diameter of the
"Hardness" is not easy to put on a scale. impression so formed is measured using à
It is fair enough to say that steel which microscope with an internal scale. From
breaks under 50 ton/sq.in. is "twice as this the surface area of the impression is
strong" as that which fails at 25 ton/sq.in., calculated, and the surface stress
but we cannot be so definite with obtained by dividing 3000kg by the ares.
hardness numbers, The result is the "Brinell Number", which
Mineralogists define hardness as "the is, of course, a "stress" indicated in
ability to scratch" and use a scale called kilograms/sq.mm. This test is limited to
"Moh's scale" on which Diamond, the about 500 Brinell number, as above this
hardest, is No:10 and Talc ("French figure the ball itself distorts. For harder
Chalk”) is softest at No.1. On this scale materials a tungsten carbide ball is used
Mild Steel would appear at No.6, and while for soft materials the standard ball is
hardened tool-steel at No.7 — not a very used but the load is reduced — usually to
big difference! A comparable test for 500 kg. — and a corresponding adjustment
metals uses a standard "scratcher" made made to the formula. The Brinell numbers
of diamond carrying a definite load which obtained agree within the standard limits
is drawn across the surface of the metal. The Rockwell hardness tester (Fig. 47!
Load is adjusted to make a "standard also uses a spherical impressor, but made
width" of scratch, and this load is taken as from industrial diamond. However, the
a measure of the hardness. This test is method and principle of operation are
now obsolete, but simple portable scratch quite different. The indenter, with a radius
test outfits are still used for casual "on of 0.2mm, is first impressed into the
site" hardness testing. specimen under a load of 1Okg.

84
 . 464 Brinet hardnes testing machine. (Courtesy Avery
ison Ltd)

This ensures that it is firmly seated in

specimen and that the specimen is
Secure on the supporting anvil. The dial
Indicating the depth of the impression is is
eseve
£
then zeroed. The "major" load, of 150kg
for the most usual tests, is then applied at 3
@ steady rate controlled by a mechanism [|
within the machine. After a set time —
about 5 seconds — the major load is with-
drawn. The dial then records the DEPTH
-
B
(|
of the impression caused by the major
load and by this alone; there is no error
3
due to "spring" in the machine frame or
due to "settling down" of the specimen.
Unlike the Brinell, the Rockwell hardness
is a “number”, not a surface stress,
though the magnitude of this number can
be correlated with such a stress.
The Rockwell machine can be used
With several types of indenter and with
Several magnitudes of "major" load. For
hard steel, and unhardened high carbon or
alloy steels, the "Rockwell C" scale is
used, but "Rockwell B", with a 1/16 in.
Steel ball, is used for "mild" steels and
some non-ferrous alloys. The "A" scale is
used for extremely hard specimens and for
thin, hard sheet material.
The advantage of the Rockwell test
bver the Brinell is that there is no need to
Measure the impression — the hardness
tan be read directly; it is, therefore, pre-
ferred for "Production" testing. Further, as
the indenter can penetrate even hardened
Steel it can be used for materials beyond

Fig. 47 Hardness testing using a Rockwell typemachine. The

depth of the impression is shown on the large dial indicator.
(Courtesy Avery-Denison Ltd

85
 Fig. 48 A Vickers type hardness tester, This machine has a
projectar microscope which throws an enlarged view of ri
impression on a screen, [Courtesy Avery-Denison Ltd)

tunately, the VDP numbers lie very close

to the Brinell hardness number, so that
"visualisation" is easier than with the
Rockwell scales. The main disadvantage
of the Vickers machine is that the impres-
sion is very small, so that special surface
preparation is required for really accurate
work. The Rockwell machine requires only
8 "good machine finish". On the other
hand, the Vickers does not disfigure the
surface so much — the indentations are
hardly noticeable. All three tests require
the specimen to be brought to the
machine. Hardness figures quoted in
specifications and textbooks will all have
been made on fairly robust specimens, to
avoid any error due to specimen distortion
under load, and all are subject to a small
tolerance — generally a range of 1 or 2
the range of the Brinell. The disadvantage digits on the Rockwell scale.
is that it is not "Absolute" — it is, really, a The Shore Scleroscope (Fig. 49) relies
hardness comparator and must be on a different principle. It is found that if
recalibrated from time to time. specimens of reasonable size are used 8
Both the above make a visible impres- ball-ended hammer will bounce to a
sion in the material. The Vickers Diamond degree which bears an almost linear
Pyramid (VDP) testing machine Fig. 48 relationship to the hardness. The
uses a square based diamond pyramid as Scleroscope is a portable hardness tester
the indenter, with an included angle of in which a small diamond-faced steel
136°. The applied load is usually 50kg, plunger is drawn up to a height of 10
but can be altered to suit special condi- inches within a glass tube. It is held there
tions (eg. thin specimens or extra hard by a catch and, when released, falls on the
material). If this is done the magnitude of specimen and rebounds up the tube. The
the load must be stated. The "across amount of the rebound can be read from a
corners" dimensions of the impression are scale. The succession of operations is
measured and tabulated values of "VDP automatically controlled by air-pressure
Hardness” (in ka/sq.mm.) derived from the generated from a rubber ball. This is a very
mean of the two. In the Vickers machine useful instrument, and especially when
the microscope is "built in" and the tests must be made in the works — it is
specimen automatically registers for extensively used in steelworks for measur-
observation of the impression. The test is ing the hardness of rolls in the mill. !t
regarded as very accurate and, for- leaves almost no impression at all, but

B6
fare does have to be taken to avoid distort and the Vickers diamond will not
uncing the hammer on the same place results in a divergence of the results. The
more than once, for the previous test will "Shore" hardness readings are compar-
E work-hardened the surface. able with other tests only when the test
_ The co-relation between Scleroscope specimen has a mass of at least one
numbers and VDP or Rockwell and Brinell pound (0.5kg), but its indications are
is fairly good, provided the Scleroscope reliable when making comparisons — eg.
specimen was reasonably massive; the an increase from Shore 56 to 60 will
“standards” provided with the instrument imply a rise of approximately 100 points
are about one inch square x 24 inch long. on the Vickers scale.
On very small pieces the specimen itself The table of comparisons given at the
may bounce a little on the support anvil end of the book is drawn from a set
and this will cause a false reading. | have prepared jointly by the American Society
one of these (as seen in Fig. 49) picked up for Testing Materials, The American
at an auction sale for a really silly price, Metals Society, and the Society of
and find that it gives excellent Automotive Engineers, but British
comparative figures when specimens are
all the same size, but until the tool is over
= in. square the results tend to be low
when compared with tests done on the
Same piece with a Vickers instrument.
Comparison of Hardness Numbers. The
choice of which test to use is often made
from considerations other than
“comparability”. The Brinell machine is
most often used for the softer grades of
Steel, and has the advantage that there is
(for use on steel) a close relationship with
the tensile strength: a Brinell test may
often save the trouble of making a tensile
test specimen. As already remarked the
Rockwell test is quick, and much used in
production workshops. The Vickers
machine leaves very little evidence of its
use, and because it is very accurate at
high hardnesses it is naturally chosen for
tool steels etc. Both the Vickers and the
Brinell give the answer in kg/sq.mm. of
impression, and agree closely at low
hardness — below 220. With hard
materials the fact that the Brinell ball may

Fig. 49 The author's Shore Scleroscopa. This js a very early

model, and differs in detail from thosa currently used,

87
Standard No.860 — obtainable from most to 43 Rackwel! could be machined a:
libraries — also gives comparative tables, about 80 ft/min, and a Nickel-Chrome-
and these go down to somewhat lower Molybdenum alloy steel, hardened and
hardness. Great care must be taken in tempered to about 48 Rockwell would
comparing the hardness of different need the speed reduced to around 55-60
classes of materials; the fact that (eg.) a ft/min for reasonable tool life.
steel and a heat treated aluminium alloy The File Test. | have left this one till last!
have the same hardness number will It is, of course, the "test of the first
reflect the different characteristics of the resort"; after quenching the piece is
two metals, and one may well be much "shown to" a file, and if the latter slips off
harder than the other. without cutting we know that it has
The hardnesses which may be hardened successfully. If, however, the file
expected from various classes of steel offers to cut the material, but doesn't
depend very much on the precision of quite manage it, then something has gone
temperature control, the effectiveness of wrong: the hardness is not as high as it
the quench and, in some cases, the mass should be — for a high carbon steel, that is
of the test specimen. The figures given in Fine files will cut blue, or even purple-
manufacturers specifications are always tempered carbon steel, but should not
obtained from standard test specimens, make an impression on straw-coloured
not working lathe tools. However, the tool points. If the workpiece takes the
following may give a guide to the "as edge off the teeth of the file you have
quenched” state. done very well indeed.

1.1-1.2C, 0.35Mn, O.45Cr (Silver Steel) 66-67 Rockwell "C"

Chrome Manganese "non-shrink" die steel 64-65
Non-tempering chisel steel 52-54
2296 Tungsten 12% cobalt high speed steel 65-66
1896 Tungsten High speed steel 64-65
File Steel, 1.2596 Carbon 68-69

The above figures will, of course, be con- Conclusion. The difficulty with hardness
siderably moderated by tempering; for numbers is that though we know, sub-
carbon steel lathe tools, tempering to jectively, what hardness is, and what it
150° will drop the hardness by no more does, it is very difficult to devise a means
than one point on the scale, and 200°C by of measuring this quality. By convention,
twice as much. Carbon steel tempered to the surface stress needed to cause an
“blue” (310°C) is just machineable, at a indentation under a known load is used.
Rockwell "C" figure in the region of 56- but this clearly has limitations; a material
58. At the other end of the scale, free- which was very brittle would not accep!
cutting mild steel, machineable at 200ft/ an indenter without splintering. However
min, will have- a hardness equivalent to 11 by custom the Brinell, Rockwell, and
on the "C" scale (which is not appropriate Vickers tests and, within its special
to metal as soft as this; the Brinell figure application, the Shore, are accepted, and
would be 180-185). A 1.896 Nickel 1.196 can be used for comparing different steels
Chrome steel, oil quenched and tempered or even different materials if due caution

B8
is used. The model engineer should not,
er, be too concerned about an odd
legree difference on the Rockwell or
Minell scales, or the equivalent on the
kers, for it is quite normal to find a
difference of one point on Rockwell "C",
or 30 divisions on the Vickers between
different tests over the cutting surface of
the same tool. These differences are partly
due to "experimental error" in reading the
microscope or dial, partly to differences in
surface finish at the point of measure-
ment, and partly to actual hardness
differences. So, if you are offered a steel
which hardens to 67, and the sort you
normally use "only" gives 66, don't worry
too much!

89
 CHAPTER 9

Home Construction of
Furnaces

Gas Firing. | have already referred briefly gives the details. The tube is a piece of 14
to the difficulties which arise with gas in. bore steel exhaust pipe, about 16 gauge.
firing. The design of a combustion For heating with a single propane burner it
chamber does require specialist should not be more than 6 in. long,
knowledge, and of a burner even more, otherwise the heating will not be uniform
and a lot of experiment. The ordinary The end-supports are pieces of 1 in. thick
"torch" — air or self-blown — is intended to Fossalcil bricks, with a hole bored through
produce a flame which can be directed at with a tank-cutter to suit the tube. You will
the work, whereas for heating a muffle notice that this hole is nearer the top of the
you need one which distributes the heat brick than the bottom, to give space for the
sideways rather than vertically or "at a flame to develop properly below the tube
point". The "D'-shaped muffle referred to One of the cores from the holes is used as a
on page 57 is normally used on the Utile permanent plug at the rear end and a home-
brazing hearth, which has a fixed side- made thermocouple lor even a commercial
burner as well as the two blowtorches. one) can be inserted through a hole in the
The firebricks supplied with the hearth are centre of it. The other core is slightly
sized so that they can be arranged around tapered and used to close the front end,
the muffle itself to direct the flame holding it with tongs, of course.
properly. It is most effective, and when | The burner used is a Sievert No.2952.
was "on Town gas" | made considerable which is rather too large, but can be throt-
use of it. If you are considering a muffle of tled back once the tube is up to
this sort, built into a home-made hearth, temperature. It is of the type which used
then a letter to William Allday & Co., to be known as a "Neck-tube" burner.
Alcosa Works, Stourport on Severn, with the air entrainment holes at the gas-
Worcs. DY13 9AP, should bring you feed end of the tube. Current types are
details of the arrangement used, together known as "Cyclone" burners. The flame is
with a quotation for the muffle. directed to heat up the firebricks (or
A small “mufflette”. | mentioned on page “Hotface” bricks) and NOT directed at the
58 and in Fig. 31 a small tube furnace tube itself. which is heated by radiation
which | rig up from time to time. Fig. 50 from the hot brickwork.

90
 1/8* to 1/4* €
le ADJUST TO | d
SUIT FLAME. |

| |

AAs

ALL BRICKS 'FOLSAIN' OR HOT FACE' TYPE

A. Sx 4tfatx3
B Sx 41/2*xt"-
SEE DETAIL
c Ox A1[2"x'* CUT IN HALF
D or 3* THICK IF NEEDED

Fig. 50 Outline details ofa small gas-fired tube furnace. The dimensions can be adjusted ro auit the materials available.

In the photograph, Fig. 51, you can see — Fig. 81 Thefurnace

intig50inaction
that the burner is directed between two
broken half-bricks on the left-hand side,
but as a rule | use a full brick with a
"mouse's hole" cut in the bottom. Hot
gases escape from a gap between the top
brick and this front one — a space of about
$ in. is sufficient. This little furnace heats
up in about 5 minutes, and is not as
expensive in gas as the burner size might
suggest. Once the bricks are red hot the
gas can be turned down considerably.
There is no reason why such a furnace
should not be made permanent, with the
bricks cemented together with firecement,
_ but | am a bit short of space in this depart-
ment, so simply keep the two endplates
and the tube and build up the rest when it

91
is needed. | have used a 1-pint paraffin the work. The material from which wire-
blowlamp in the same way, but this wound muffles or heating tubes are made
needed twice as long to get up to is not a perfect insulator.
temperature. It would have been better, Wound Muffles. Furnace muffles ready
too, had all the bricks been of the Fossalcil wound are available from Messrs
type. But it did the work required at Gallenkamps, Griffin & George Ltd, and
around 780°C. other laboratory equipment suppliers.
Naturally, a smaller diameter tube can They are, in fact "spares" for furnaces
be used, but if it is at all longer than the 6 similar to that seen in Fig. 32. Special
in. (about 43 in. "hot length") you may insulating material will be needed
need two smaller burners, or keep moving adjacent to the muffle itself, and this, too,
the larger one from side to side. | have can be had from the same source. Outer
been meaning for a long time to devise a insulation can be the glass- or slag-wool
fan-shaped head for the burner, rather like insulation used in (e.g) AGA cookers
those used on paint-stripping torches, but around the firebox. A minimum of 3 in. of
have not had the time! Such a head would good heat insulation is required, otherwise
distribute the heat sideways much better. the heat loss to the case will prevent the
Furnace contro! would be easier, too. In furnace from reaching its proper
this connection, you must not be tempted temperature — the element rating is only
to hurry things up by heating the tube about 2 KW. The outer case itself is — or
directly, for it will only grow cold again. was — usually Syndanyo sheet, about $ in.
Once the firebrick is hot it is only a case of thick, joined with self-tapping screws, but
minor adjustments to the burner to keep à as this contains asbestos it is now
steady temperature, and once the bricks replaced by similar material based on
are hot all over the muffle tube keeps Calcium Silicate. This may be available
pretty steady. from builder's supply merchants but if in
Electric Heating. This is rather easier for difficulty Messrs Cape Insulation Ltd,
the home constructor, but the furnace will Washington, Tyne & Wear NE38 8JL, lor
be markedly slower in reaching the at their local offices) can give the address
operating temperature. You will of the nearest retailer.
appreciate that a burner such as that just The actual arrangement of the furnaces
mentioned has a maximum output of is a matter for your own taste and
something like 18 kW, and an electric ingenuity! Don't forget that the door
element of that capacity would strain your needs insulating (a piece of hot-face brick
power supplies somewhat! On the other will serve) and, if it has a metal handle,
hand, electric heating is clean, silent, and this needs earthing. You can add a
easy to control — automatically if need be. pyrometer — indeed, | would most strongly
There is little risk from fire. The one thing advise this, as the virtues of muffle
that must be borne in mind, however, is heating are considerably diminished
that you are dealing with lethal voltages otherwise. These can, of course, be
and must pay really strict attention to obtained from the same source as the
safety rules as well as to the I.E.E. regula- muffle, complete with flange for screwing
tions. All metal parts must be earthed, to the case and a bracket to hold the
proper insulation materials applied and — indicator. Commercial furnaces are fitted
rule No. 1 — the supply should ALWAYS he with a "thermal fuse"; this is no more than
Switched off when charging or discharging a pair of rods projecting into the chamber,

92
 . 52 The Kanthal electric
e built from a kit. (Courtesy
thal Electroheat Ltd)

ulated of course, united with a wire of drawing supplied by Kanthal, but the
Suitable melting point (silver for 1000*C, chamber base is a special refractory plate
example) and wired in series with the which will stand more wear and tear than
inding. If the furnace over-runs and gets will the surface of the bricks. Fig. 53
hot the fuse melts and cuts off the shows sectional views of the furnace — the
Supply. size is, of course, determined by the
A Radiant Element Furnace Kit. Fig. 52 number of elements used. Thermostat
is a photo of a furnace designed by control is available, but rather expensive,
Kanthal Electroheat Ltd, Inveralmond, and a simple energy controller is really
Perth PH1 3EE specifically for home con- quite adequate. A thermocouple/
Btruction. (The photo was taken, by the pyrometer is an essential if the best is to
way, before the essential earthed quard be got out of the equipment. This, once
ad been fitted over the terminals). It can built up, is a professional furnace without
be built either as an 8 KW unit rated at the normal casing and hinged or sliding
350°C or with a 4 KW loading for door, and the chamber size is more than
200°C, The latter is probably adequate adequate for most mode! engineers.
all model engineering purposes. The Electric Tube Furnace. Over the years a
eating elements in this case are Silicon number of home-made electric tube
sarbide rods, the ends of which can be furnaces have been described, using
n in the photo, and these are set in the ordinary electric fire elements and home
p of the chamber. Heating is entirely fired ceramic tubes. There are several
diant, but | would advise users to bring hidden problems. The first is that such
e furnace up to temperature before elements are rated to work in free air — i.e.
tting in any work, otherwise the upper losing heat as fast as they can. Once they
rface is liable to be heated faster than are encased in insulating material they
e lower. The furnace body is made will rapidly burn out. They must be
ntirely of Hot Face Insulating Bricks derated for furnace work. Second, it is not
ich are sawn and drilled to the detail easy to make a fire-cement tube without

93
 4off -1200"C
SILLIMANITE
FLOOR

EXTERNAL VIEW
(TRIMETRIC PROJECTION)
Fig. 53 Outline details of the furnace shown inFig 52. The bricks con easily be cut without special tools

blisters and spalling. Third, such a tube is

usually very fragile, and it cannot be
strengthened by moulding it round a stee!
tube as some makers have suggested; the
expansion of the tube will burst the clay.
However, for those willing to experi-
ment, and face one or two failures with
equanimity, Fig. 54 shows the stages in
making the tube. The "mould" is a piece
of steel tube of the desired internal
diameter, furnished with two thickness
washers. Waxed paper is wound round
the tube — three layers — and the
endplates or spacers are greased. The
firecement (that seen is "Kos'"] is made
into "worms" of thickness just over halt
the desired thickness of the tube. (This

Fig. 54 The tube former, which should be covered wë

waxed paper, winding& "worm (the paper is not shownin Lo"
Photo) and the tube after firing. The wire binding acts as ^
anchor far the elemant at each end.

94
 g. 55 Winding the element in
lathe. Note the notch in the
guide.

should not be more than 1 in. and 3/16 is You will need two elements, from the
better). These worms are then wound local electric shop, of 1 KW rating.
found over the paper liner and well Carefully unwind one by threading a
kneaded together. The first layer is length of 3/16 in. rod through and pulling
allowed to air dry, and then a second laid off the wire. Avoid kinks. Measure the
On top, this one being smoothed and length, and then wind on to a bobbin,
kneaded down using the two end washers Unwind part of the second element, and
as guides. This too, is air-dried for several cut off one fifth of its length. Double this
days on a hot radiator, after which the and twist, to make a twin wire. This will
tube is withdrawn. If it sticks, a blowlamp then be cut in two to form the connecting
can be directed down the inside to fire the wires, and also serves to reduce the
paper, but on no account must the firece- applied voltage to the actual heating
ment tube be made really hot at this element. You will need two small con-
Stage. With the steel tube extracted the nectors made by drilling 4 in. steel rod 4
Cement must be further air-dried, and the in. and tapping for two small set-screws:
longer this lasts the better — up to a week,
h must then be "fired", and this process
p" also be very slow. (The problem is
that moisture in the interior may form
Steam pockets which lift the surface). Fire
m the inside with a brazing torch, very
wly raising the temperature to red heat.
If any spalls farm, these can be picked off
and “patched” with firecement.

|. 56 The wound element, with one of the wooden plugs

MC the left-hand end. and the element after coating with
sment tQ retain the coils — before firing,

95
 Fig.57 A por furnace suitable fortemperaturesupto 950° C

re-wind with a finer pitch. If the winding

does not reach the end, use your
judgment as to whether to tolerate a shor:
unheated part of the tube or to coarsen
the pitch. This part of the job takes very
little time. Fig. 56 shows the wound tube
The winding must now be covered with
a very thin layer of cement — the wire
should just show. This, once set, will form
a binding to prevent adjacent coils
touching when the wire expands with
heat. Before going any further, test the
arrangement. Attach the two twin feeder
voltage dropper lengths, one at each end,
and a piece of mains cable to these with
the normal twin porcelain jointer, Make
the standard brass connectors will not do, sure that there is no risk of any part of the
as the temperature is too high. bare wire touching anything, and set all on
Make two wooden plugs to fit inside 8 piece of firebrick. Use a plug with 13
the cement tube, one with a centre-hole in amp fuse and switch on. DON'T TOUCH
it. Wind iron wire round each end of the ANYTHING. After a short while the tube
tube, about $ in. from the end, and twist should start to get hot, and you must note
tightly; these are the anchor points for the the final temperature. If this is no better
ends of the element. This will be wound than black, pull out the plug. cut about
on using the lathe. one fifth off each of the twisted end
From the known length of the element lengths and try again. If it gets to dull red
and the diameter of the tube you can and stays there, all should be well; once
calculate the number of turns. Select an insulated it should reach the desired
easy screw-cutting pitch which will permit temperatures. If, however, it burns out
this number of turns on the bobbin — after half an hour, you need more on the
something between 12 and 14 tpi will twisted end lengths!
probably serve. Set a piece of wood in the Once satisfied it is worth doing the res!
toolholder with a notch to guide the wire of the work. Insulate the feeder wires with
(Fig. 55) and attach one end of the ceramic beads, and then make up a casing
element to the tailstock end anchor, of insulating bricks carved aut, feeding the
leaving about three inches free. Keep a wires outwards between two bricks. Make
fair tension on the wire and with the lathe a terminal block from 1 in. Syndanyo and
running at about 50 rpm, wind on the fix this ta one of the bricks. (If you use
wire. At the other end, anchor the wire bolts, then these must be earthed, as mus!
again. Note: if you have more than a foot the protective cover over the terminals i!
left over but less than a yard, don't worry, of metal), How you arrange this is up to
but if there is more than that you should you, but you need at /east three inches of

96
 A net cusce.
MORGAN 'TRITON'
INSULATING WOOL.
| TUBE 85mm. BORE (KAOWOOL
x5mm. THICK. —— HIGH DUTY.)
- LENGTH TO SUIT
MORGAN TRIANGLE
SOE. HIGH ALUMINA, 21/2 kw. 240 volt
WINDING.
KANTHAL 'A'or DSD:
OR SIMILAR.
TO SUIT,
; |SAY 14"

1x6. STEEL SELF-

TAPPERS AT 2"PITCH

=-s-
e
w
—

SYNDANYO OR
CALCIUM SILICATE.-

3[4* min. ALL. ROUND.-

Fig. 58 Outline details of a furnace similar to Fig 57. Detail changes have been made to reduce
heat losses.

insulating brick all round. Any gaps can be months of service, and several give way at
filled with ground-up fragments. Close the once. Others have built them and had
back of the tube with a brick plug, and (for them in service for years. A great deal
what it is worth) just lean a firebrick depends on (a) the homogeneity of the
against the other end to act as a "door". cement: it must be well kneaded before
| give these instructions “without and during application, and (b) the quality
warranty"! | have made such a furnace of the element. You can, of course, buy
and had satisfactory results, but have also the proper resistance wire from Kanthal,
had three burn out on me after some specified for this service, and this will not

97
 Fig.59 Wuolded steel! pot. Note the earth terminal.

3} in. bore Morgan 90E High Alumina

wound with a Kanthal element for 240
volts. If buying a wound tube it is impor-
tant to specify the power input — 23 to 23
KW in this case — the applied voltage, ang
the operating temperature of the ceramic
tube. This will be higher than that in the
salt pot itself, so that a sketch of the
arrangement, or a copy of Fig. 58, should
be sent with the enquiry, The outer casing
is of 4 in. Syndanyo, though today one of
the Calcium Silicate substitutes from
Cape Insulation Ltd would be used. The
immediate insulation material round the
element is "Triton" Kaowool ceramic
fibre, supplied by Morgan Ceramic Fibres
Ltd, Bromborough, Wirral L62. 3PH. The
standard grade is good for 1250?C, but |
suggest the "High Duty" rated at 1400*C.
This is packaged into the space between
the element itself and the heat insulating
bricks — which are cut roughly to leave
about 1 inch gap between their surface
D
and that of the element. The bottom of the
need derating and be less likely to "blow" tube is packed with the wool also, but this
at the operating temperature. (The wire must not lie above the first coil of the
may get up to 1200°C with the tube at element.
800°C). You can, of course, buy a tube, Fig. 59 shows the actual pot. This is
too, from Morgan Refractories Ltd; but if made from drawn steel tube, with the
you are going to do that you might as well bottom welded on with full thickness
buy a ready-wound tube, either from penetration. The O.D. of the pot must be
Kanthal Electroheat or from Morgans and about à in. less than the bore of the heater
have done with it! tube, so that it may be necessary to
High Temperature Salt Furnace. Fig. 57 machine the steel tube to a smaller
shows my 23 in. x 10 in. neutral salt thickness. This is not difficult and a rough
furnace for temperatures up to about finish will do. The support flange |s
950°C, rated at just under 2.8 KW. This recessed as shown. This prevents any
is, perhaps, rather larger than might be splashes of salt from getting between the
needed, and a 23 in, x 8 in, would do for pot and the heater. The handle shown is
most applications, but it was originally not intended for carrying the pot around
made (not by me) as a small laboratory when hot, but simply to make it easier to
furnace. Fig. 58 is the drawing. The tube is lift. A lid of some sort is needed, both !o

98
 g. 8O The furnace af Fig. 87 with the lid removed. That in
58 has more and different insulation,

p out dirt when not in use, and to

er the pot when reheating. There is
t a slight risk that the lower layers of
It may melt and expand before the crust
on top melts, and if the latter splits open
ünder the pressure there could be a
Spitting out of hot liquid. Fig. 60 shows
the furnace with the top cover removed.
This furnace has rather less insulation
than is desirable.
None of the dimensions is critical — you
tan make the size of the heater tube or
melting pot to suit your needs and adjust Syndanyo or Calcium Silicate, though
all to suit. However, it /s important to porcelain terminal blocks are sometimes
allow for plenty of insulation — the more available on the "surplus" market. The
the better as this will speed up melting regulation of the furnace is best done with
time and reduce current consumption an "Energy Controller" — a thermostatic
once up to temperature. Reducing the control would be very expensive. Those on
diameter of the pot will not reduce the electric cookers will handle 3 KW. I do not
overall size of the case very much. As to advise building it into the furnace casing,
depth, you have to allow sufficient for the as these controllers incorporate a thermal
tide to rise as you immerse the workpiece, switch to effect the time-intervals, and the
and the top level of molten salt must heat from the furnace would upset them.
always be below the top of the winding. “Heater Bricks”. A relatively new
The loading of the heating element development is in the form of a heat
depends almost directly on the volume of insulating brick with a heater element
the pot, but | don't advise going below embedded in the face. At first sight these
1000 watts (1 KW) as the heat losses are would seem to be the ideal module for the
by no means negligible. construction of a small furnace. There is
Note that the steel pot and the terminal also a variation, with the "brick" in the
box (if of metal) MUST both be earthed, as form of a hollow tube, making the basis of
the insulation resistance of the ceramic a tube furnace. There are two problems =
tube is not perfect. (The current should, of the first is that the units are wound for no
course, be switched off when manipulat- more than 60 volts, so that either a hefty
ing the work). There is no real need to transformer is needed, or one must use
insulate the leads to and from the heating four units. The second is that the hot face
element, but the little ceramic beads used of the bricks will not carry any load: you
on cooker circuits and electric fires are would need a supporting and "heat tran-
quite adequate. The actual leads are a sparent'' floor, or a metal tube if using the
continuation of the heater winding. but tubular type. Both floor and tube would
doubled and twisted, so that they will not have to be of heat resisting steel or of
get very hot. The terminal block is of considerable thickness if any reasonable

99
life was to be expected. However, these Stockport Cheshire. Such element wire ic
are not difficult problems to get over and available either as the normal round wire
readers who are interested in this or as strip, the latter for the higher KW
development should write to Messrs W.J. ratings. The material must be selected fo,
Furse & Co. Ltd, Wilford Road, Not- the operating temperature of the wire
tingham NG2 1EB. The bricks are known which will be higher than that of the work
as “Watlow Ceramic Fibre Heaters”, by quite a bit, and this does mean that the
Resistance Wire. This is obtainable from advice of the manufacturers, with their
the Kanthal firm already referred to, from vast experience, is needed. To "play safe’
London Electric Wire & Smiths Co. and select the highest rated element wire
(LEWCOS) Church Road, London E10, as can be rather expensive.
well as other manufacturers, and in small The problem is to effect a compromise
quantities from K.R. Whiston, New Mills, between three variables. The overai!

DIMENSIONS TO SUIT FURNACE.

Fig. 59A Construction

details of the pot shown in Fig.59.
resistance needed to limit the current to output if wound to a resistance of 57.6
the desired power — eg. for 1 KW at 240V ohm. The 19 gauge element will need a
the resistance at the working temperature winding length of about 7 inch, and this is
must be such as to limit the current to reasonable for 4 inches diameter — or it
1000/240 — 4.17 amp, say 58 ohms. could be wound at 12 TPI over an 8-inch
Next, the surface area of the wire must be length if desired.
such that it can dissipate this power. A not The resistance per foot and surface
untypical rating is 6.5 watts/sq. inch of area per foot lor in metric units) is quoted
surface area. So long as the length of wire by the suppliers, as is the safe dissipation
needed for this criterion is /ess than that in watts per unit area of wire. In some
needed to provide the required ohmic cases it may be necessary to change to
resistance the element will be safe, but if flat strip heaters, in order to match the
the opposite condition holds, then it will resistance per foot to the required surface
probably burn out. Finally, there must be area. The wire manufacturers will be able
enough room to wind this calculated to supply tables which help considerably,
length of wire on the ceramic tube, It is but do not expect them to act as "Con-
not safe to wind at closer pitch than about sultants" for the price of a few pounds
14 T.P.I., as there is then risk of adjacent worth of wire. (The 19 gauge element
turns touching each other. tabled would need about £9 worth).
The following calculation table shows Conclusion. This chapter has necessarily
how four gauges of wire have been been brief, for furnace-making is not a
assessed to meet the required criteria for very common occupation for model
a 1 KW 240 volt heater element. Note engineers, and there is not much
that these figures, though typical, should published information. However, | hope
not be used in an actual furnace; the table that what | have written will help yau to
is just an exercise to show how to go avoid some of the more difficult problems.
about the job. | would, however, suggest that before
Required current — 1000/240 — 4.17 embarking on the manufacture of any but
amp. Required resistance at the operating the simplest muffle you explore the
temperature of 1200°C, 240/4,17 = 57.6 second-hand market. "Exchange & Mart"
ohm. Power capacity of the wire 6.67 may bring forth something, and even if the
watts/sq.in. surface area, hence A = 150 furnace requires a new muffle, or perhaps
sq.ins. Outside diameter of heater tube is the thermocouple of the pyrometer, the
4 in. = 1.05 ft circumference. cost of repair may well be less than the
Wire Length Read, Ft. Turns on Winding Length
Gauge For Power For Resistance 4 in. dia inches.
16 62.2 259 246 17
18 82.9 146 139 10
19 99.3 104 95 6.8
20 110.3 82.4 78 5.6
This table shows that the 20 gauge cost of materials for a home-made affair.
wire requires a greater length to dissipate The little gas-fired tube furnace is, of
the power than is needed to offer the course, à very simple matter, as you must
required resistance; it is, therefore, unsuit- have firebricks around any heat-treatment
able. All the others will handle the power hearth anyway.

101
CHAPTER 10

Safety Precautions

Safety in the workshop is, most of the The answer to these is twofold, First,
time, no more than "applied common the heat-treatment spot should be as
sense", and | fear that many of the Parlia- fireproof as possible. If working on a4
mentary Enactments on the subject seem bench, then the top should be covered
to assume that most people working in with two layers of bricks, with staggered
industry are half-wits. However, common joints, so that flame cannot pass through.
sense is the more easily applied if some- Better, use a purpose-built steel frame
thing is known of the hazards, and my with the bricks set in. | know that space
object in this chapter is to present these to often compels the use of wooden top
you, no more. To those readers who have bench, but if any great amount of heat
been sucking this sort of egg all their lives treatment is done then the bench should
| can only say that many others have not; be made accordingly. The brazing spot
and if YOU tried to use their tools you should be so equipped, and this is, surely
might well do yourself an injury straight the right place to do other "hot work" as
away. well.
Fire. This is, perhaps, the most obvious Second, have fire-extinguishing gear
danger. At 780°C the workpiece will, if available. Fire can get out of hand very
left, burn a hole in the floor — or through quickly indeed, and prompt action may
your boot. Worse, it can burn through your save you a lot of misery — and money. The
gas-hose in seconds, and though the hose quench bucket is the first resource, of
failure valve (dare | assume that your course (but note later the remarks on sal!
equipment has one?) will shut off the baths) sand is the second, and a proper
supply, the initial burst of flame can set fire extinguisher the third. | use the "Dry
fire to other things. The second hazard is Powder" type, and have a CO,
from the heating flame itself. The practice extinguisher as a back-up. (We are 12
of moving the torch aside to observe the miles from the fire station), These types
metal is normal, but how many are both safe for use on electrical fires
practitioners think on as to where the and my workshop, like yours, has a lot of
flame is now pointing? | have myself set wiring in it. A very sensible arrangement i5
fire to a carelessly placed piece of cotton for the whole of the bench (brazing or heat
waste this way. treatment) to stand in a shallow tray con-

102
taining an inch or so of DRY sand. This jumping off a tool in the quench and
will catch any droppings safely, and from lodging down the open neck of a shirt!
it a quick shovelful of sand can contain à The use of gloves is debatable. If these
small fire whilst reaching for the tend to make you clumsy and increase the
extinguisher. risk of dropping workpieces then perhaps
The other hazard from hot metal is better to do without them. But if you have
burning yourse/f. The first point to note is a pair of reasonably supple chrome leather
that cold water won't hurt a burn, and to ones, these should be worn when working
put any sort of dressing over a dirty paw is with hot high-temperature salt lor molten
asking for trouble. So, sluice on cold water lead if you use that). There are some so-
and get someone carefully to wash the called "heat proof" gloves on the market,
area around the burn. Then if it is a bad but they are not suitable for this class of
one — red-hot metal on the back of the work — maybe at tempering temperatures,
hand, for example — or of any size, do NOT but they won't stand red-hot metal. The
apply anything sticky. A dry burn dressing object of the glove is, of course, to protect
will do little harm, but even a clean pad your hand from accident, not to enable
made from a clean handkerchief (NOT lint) you to pick up workpieces! One important
and loose bandage will keep the lesion point — the gloves should be such that you
clean on the way to "Casualty" at the can whip them off smartly: they will keep
local hospital. Don't neglect this — a burn the heat from the hot metal from your
attended to promptly by experts will give paw, but the glove is then hot, and this
little trouble, but leave it till next day and heat will travel through soon afterwards.
you may well be in trouble. For minor Finally, wear a cap. Quench a piece of
burns, then the acriflavene based burn steel with a hole in it, and the odds are
dressings can be used, but even here, seek that a lump of superheated water will leap
the advice of your doctor if in doubt. Keep up, to land on your bald patch. (So you
these dressings in the workshop or have hair? Good, but it would go through
nearby, but don't forget — it is the dirt that, too!). There is absolutely no need to
round the burn which will cause infection, dress up as if you were melting 100 tons
so see that your paws are clean. of steel when you are about to harden a 3
Clothing. What is up must come down, in. form tool, but that tool [S hot, and you
and your feet are vulnerable. Carpet should, as | said at the beginning, use your
slippers are not the best things to wear common sense.
with metal at 800°C a couple of feet Splashes. You may well have dumped a
higher up. Wear leather shoes, Cotton few jobs into the pickle-bath in the past,
twill overalls are safe — or as safe as can so you know what happens. Don't lean
be — but those of man-made fibre can over the quench tank, and if the work is of
literally melt if overheated. Splashes from any size at all, wear proper goggles. These
quench-tanks can be hot, and those from are dirt cheap, but MUST be to British
salt baths ARE hot. Protect your throat Standard 2092/2. These are heat-proof
and neck. On the steelworks we used to and will not shatter from heat as glass
have sweat rags which could be held in will. They have little side-pieces that stop
our teeth. (Though nowadays | have no things from coming that way, and can be
doubt that special hot-metal clothing is de worn over spectacles. They also have a
rigueur!) But an old scarf is useful — little "top shelf" which prevents stuff from
nothing is worse than a hot piece of scale falling into the space between specs and

103
the eyes. If you do get anything in your required amount is in the pot. Allow plenty
eye, wash well with cool water and go to of space for the tide to rise when work is
the doctor; it is just not worth taking any put in. In the case of the electrically
risk here. heated high-temperature furnace the salt
Splashes from the oil-quench tank level should be below the upper end of the
need be no more serious than from water element.
— they should not be too hot — but if at all There will be no frothing on remelting
painful treat as for a burn; wash round the 50 long as you have had a lid on so that it
spot and apply a dry burn dressing until cannot absorb any moisture, and even
you can get medical attention. The brine then it will be relatively slight. However, it
bath differs only from the water quench in is worth again repeating that the salt may
that the salt is an irritant. But don't form a crust, and when re-melted the
neglect any injury from hot splashes, from expansion of the liquid below may cause
water, brine, or oil. The point is not so this to crack and send out a spurt of hot
much that they are hot as that they are salt. So, keep a substantial lid on the pot
DIRTY, and it is from this that subsequent until the whole is melted.
infection can arise. Never stick your head over the top of a
Salt Baths: General. ALL hot liquid baths salt pot in service, always observe at an
are dangerous if they can overturn. The angle, always wear BS2092 grade 2
first rule, therefore, is that the furnace, or "Impact" goggles and a cap. and gloves if
if used, the hotplate for tempering salt, you can possibly manage with them.
should be secure in this respect. A heavy All work put into the bath must be
pot on top of a flimsy stove is just not water- and oil-free, and should be
acceptable and if you are unable to immersed slowly. Particular care must be
arrange for really solid support, don't use taken when oil-quenched work is to be
the system at all. Second, the salts — tempered at above about 220°C. as any
sodium nitrite and potassium nitrate for residual oil may "flash off". Prior degreas-
tempering, and sodium and potassium ing is recommended if the tempering bath
chloride for Austenising, are, though is above 245°C. Slight surface oil con-
strictly non-toxic, irritating: just like tamination should cause no trouble in the
kitchen salt. The cold salt should not be Austenising bath, but if there are any
handled with bare hands, and any powder holes in the workpiece these should be
spilled should be dealt with using a brush cleared of oil before immersion, No
and shovel. fixtures or other devices for use in salt
All salts can absorb a little water, and baths should be made from tube, Great
all contain "water of crystallisation’ until care should be taken to ensure that no
melted. At first melt, therefore, there will foreign matter falls into the bath, whether
be quite a bit of frothing as this water is cold or hot, It is imperative that no water
expelled. Start the melt with the pot only should fall into a salt bath, and water
half-full, and apply heat very slowly. Let should not be used to extinguish any fire
the initial froth die down — don't increase adjacent to one.
the temperature — and then add a little Avoid overheating any salt bath. They
more salt. Use a metal scoop to put it in; should not be left “on heat” when leaving
don't "tip it out of the bag”. Let the the workshop for more than the odd
moisture evaporate and again add more minute or two, even if thermostatic
salt, in small quantities at a time, until the contro! is fitted. Overheating can be

104
 rious with tempering salt if washing down the sink with a good flood
emperatures go above about 550°C. of water.
IN NO ACCOUNT MUST TEMPERING Within the temperature range and
SALT GET INTO THE AUSTENISING POT. quantities used by model engineers this
'empering Salt. This is a mixture of salt can be regarded as non-hazardous
50 dium Nitrite and Potassium Nitrate. provided reasonable care is taken.
These can irritate the skin, but are classed High-temperature (Austenising) Salt.
as “non toxic". Soap and water can be This is a mixture of Sodium Chloride
used to wash off the skin, and an eye- (common salt) and Potassium Chloride. It
wash solution if any powder gets in the is virtually non-harmful and can be treated
eyes. The salt looks like coloured sugar, so as if it were household salt if it gets in the
at it SHOULD BE KEPT AWAY FROM eyes, in a cut, or is taken by mouth. In the
CHILDREN, but if any is taken by mouth event of burns from molten salt these
give plenty of water to drink and call the should be treated as for the tempering
doctor. In the event of a burn from molten salt, but the burn may be severe and
salt — (a) Flood the affected area with immediate resort to the casualty depart-
ater or Sodium Bicarbonate solution (b) ment of the hospital is recommended.
Remove solidified salt and wash again (c) Spillage of the cold salt powder is
treat as a thermal burn; apply a medicated unimportant, though it will absorb water
(dry) burn dressing and see the doctor or, and may damage (rust) metal. Spillage of
if extensive, take the patient to the hot salt can be contained with dry sand
hospital. dams, and the fire, if any, tackled with CO,
The salt is not inflammable in itself, but extinguishers; water should not be used in
a strong supporter of combustion; wood the presence of molten salt.
or cloth contaminated with the salt will Material immersed in the bath must be
burn vigorously. Some organic com- dry, and the work should be set in slowly
pounds can react vigorously with the salt to allow any air trapped in holes to escape
when heated. Care should be taken that slowly — à rapid immersion may cause
splashes or spillages of powder on wood spurting. Care must be taken when
or rubber are cleaned up promptly. remelting, with a substantial lid, in case a
Clothing should be washed, NOT dry- crust forms on the top of the bath, and
cleaned, as there is a reaction with dry- initial melting must be slow: this salt
cleaning solutions. melts at about 670?C, so that frothing àt
If heated above 550°C, oxides of this stage should be reduced to a
nitrogen may form, and these can be minimum with the depth of the bath being
absorbed by the skin, and if inhaled can be increased only slowly.
dangerous. Symptoms are blueness of the This salt is inherently safe, but the
lips and face, and perhaps shortness of hazard from very hot liquid is present. If
breath. The condition requires medical treated as a molten metal, with the same
treatment. precautions, dangerous situations are
The salt is supplied in plastic sacks, and unlikely to arise, (though | repeat my
Bais be ladled from these into the pot previous warning: on no account allow
With a metal spoon or ladle. Storage con- any tempering salt, or metal contaminated
tainers should be moisture-proof: | keep therewith, to get into the high-
mine in plastic bags inside large Nescafe temperature salt, cold or hot.) Careful
tins. Used salt can be disposed of by attention to protective clothing at all times

| 105
is the order of the day. Storage and which must be earthed. Fusing is best
disposal conditions are the same as for done at the 3-pin plug-top, but if fused
Tempering salt. elsewhere (or if circuit breakers are fitted
Electrical. Electrical hazards in the heat- to the control box) these must be arranged
treatment shop are exactly the same as to break the "Line" circuit only; on NO
those elsewhere, but rather more care account must any fuse be set in the
must be taken over trailing leads, Metal at "Neutral" conductor. Failure of a fuse hera
quenching temperature will burn through leaves the apparatus live to earth.
insulation and cause a short-circuit very Finally, because no furnace muffle is 5
quickly, and will itself set the insulation on perfect insulator, it is common prudence
fire. Wires within furnace casings should to switch off the circuit when handling the
be insulated with ceramic beads, obtain- work. This applies especially to muffle
able from the local Electricity Service furnaces; a salt pot with an earthed po:
depots. Glass-fibre insulating sleeving is should be quite safe so long as the earth
available, but this is normally rated only connection is sound.
for the temperatures found in domestic
cookers. ALL metal parts of a furnace Conclusion. The careful practitioner will
must be properly earthed, the earth wire meet few dangers in heat treatment, but
being not less in diameter than the main no amount of "precautions" will save the
leads and preferably more. These connec- careless from injury. Heat treatment
tions, especially when earthing a hot point practice is probably less dangerous than
(e.g. the pot in a salt furnace) must be brazing, as though temperatures may be
inspected and cleaned reguarly, as oxide higher the amount of heat present is less
formation may set up an unacceptable If, in this chapter, | have been "teaching
earth leakage resistance. my betters how to suck eggs" | console
Cables to the supply socket must be myself with the thought that a hot egg is
routed away from risk of contact with hard-boiled, and such cannot be blown at
flame, hot metal, or salt and, if permanent, all. Any material much above 70*C needs
covered by metal shield or conduit — to be treated with respect!

106
 PPENDIX 1

Thermocouples and
Pyrometers
It is possible to obtain mercury-in-glass EMF generated does not vary exactly with
thermometers which will safely register the temperature; plotting EMF against
up to 360°C quite cheaply, and nitrogen temperature difference shows a slight
filled mercury thermometers in special curve to the graph. This means that the
glass can be had which read to 600°C. true temperature may not be exactly that
They are rather fragile, but can be used for found by adding the hot and cold junction
tempering — though | make a steel temperatures. We will deal with this
protecting sheath for mine. The well- problem later, but it is not serious for heat
known "'Rototherm' temperature treatment purposes.
indicators, based on a bi-metallic coil, can The normal arrangement is that shown
be bought to read up to 300°C (they are in Fig. 62. "A" is the hot junction and BB
used as oven thermometers) and are more the wires within the immersed length —
robust. The mercury-in-steel dial ther- the materials being the "thermoelectric"
mometers will operate safely up to alloy. CC are connecting wires leading to
650°C, but they are both expensive and the meter "M" and “R” is a series
bulky. None of these, however, will serve resistance. If CC are plain copper con-
for temperature indication when heating ductors, then the effective cold junction
(Austenising) for quench hardening. The will be at DD. This may be close to the
thermo-electric pyrometer is the only furnace, and DD could be quite hot. The
practicable answer. effective cold junction can be moved to
Fig. 61 shows the principle. If the the meter either by using special alloy
junction between the two wires at "A" is connecting cable (called a “compensating
heated, and that at "B" is kept cool, then lead") or more simply by continuing the
the meter "M" will show that a voltage is thermocouple element wire from A right
developed which depends on the through to M and R. This is the method I
temperature difference at the two junc- use; the meter need not be all that far
tions. Note that it is the temperature away from the furnace, and one has to
DIFFERENCE which matters; if the meter buy à yard or so of thermocouple wire
shows a voltage corresponding to 700°C, anyway. The function of the resistance
and "B" is at 80, then the temperature at “A is to reduce the effects of changes of
"A" will be about 780°C. Further, the resistance in the circuit which may occur

107
 Fig. 61 The principle of the thermocoupi

A. B.

as the couple element gets hot. If the alloys, notably those based on Platinum,
resistance of BB and CC were, say 0.5 but none likely to be of interest to us. Al)
ohm and that of the meter 1.5 ohm, then these thermocouple alloys are now mace
any change in the former would be to a British Standard Specification, and
appreciable. By inserting this "ballast the following table, taken from "The
resistance” say 100 ohm, then the change Model Engineers Handbook" shows the
in resistance of AA and CC with EMF developed in millivolts when the cold
temperature will have negligible effect. It junction stands at O°C. This assumes that
serves a further purpose, too. It can be the wire is to the relevant B.S., but the
used to match the indicator to the par- difference between this and others (eg. to
ticular thermocouple used. We shall make DIN std) is negligible.
use of this when "adapting" instruments The EMF in Millivoits developed by
for home-made indicators. couplesis given in the table.

A=Copper-Constantan to BS1828
THERMOCOUPLES B-lron-Constantan to BS1829
C=NiChrome-NiAluminium BS1827
A number of different alloys are used for
the thermocouple elements. The cheapest re F” A B C
is copper against "Constantan" (60%
Copper 4096 Nickel) but this can operate 051 p 0 o
safely only up to about 500°C. Chromel- 50 76 2.02 2-58 2:02
Eureka (90% Ni 10% Cr vs 40% Ni 60% 100 21/2 4.24 5-27 4-10
Cu) can be worked intermittently up to 150 27€ 6-63 8:01 6.13
850°C. or 700°C continuously, and has 200 5-6 9.18 10.78 8.13
the merit that it delivers the highest EMF/ 250 "9611.86 13.56 10-16
°C of any. lron-Constantan is the next
best in this respect. and can work con- 300 $*^ 14.67 16.33 1221
tinuously at 750°C and intermittently up 350 63% 17.58 19.09 14.29
to 1000°C, It is cheap, and well suited to 400 726 20.59 21.85 16.40
our work, provided you recalibrate if it is 500 15* 26.10 27-39 20-65
used for long periods (days) above about
800°C. The commonest for high- 600 12 Pe — 33-11 24-91
temperature work is Chromel-Alumel 700 /366 — 39.15 29.14
(90% Ni 10% Cr vs 94% Ni, 2% Al + Si 800 /"*b — 45.53 33.30
and Mn) which can work all day at 900 (626 — — 37-4
1200*C. It does not develop quite such a 1000 (FOG — = 41.3
high EMF as the previous types, but suf- 1200 — = 48.9
ficient for our purpose. There are other 1400 — — 55.9

108
Fig. 62 The vlements ofa practica!
pyrometer,

G. R.
Now, let us suppose that we are using reaching the temperature and that
an lron-Constantan couple with the hot indicated. The ends can easily be welded
junction at 700*C and the cold at 100*C. together if you have a little Microflame
The indicated voltage will be 39:15 — oxy-butane torch or, indeed, any welding
5.27 = 33.88 mV. This corresponds to a equipment. The gauge of wire is unimpor-
temperature of 613?C. Add on the cold tant except as regards cost, and 26 gauge
junction temperature and we find the sum is quite satisfactory, though | would prefer
to be 713°C — 13°C higher than it to see this in a protective sheath. That
actually is. So, the cold junction must, seen in Fig. 63 is 18 gauge, and this can
first, be kept as cool as possible (in really be used in a furnace or even a salt pot
accurate work it is set in ice in a thermos with no sheath, though one is preferable
flask) and second, we should keep a ther- in the latter case. Twin thermocouple
mometer nearby to check the pairs, about 30 gauge, insulated with fine
temperature. If the cold junction is kept heat-resistant sleeving, are available. For
within +/— 2°C of 15°C, the error will not the knockabout work of heat treatment |
exceed two or three degrees if the cold would tend to favour the 18 gauge wire,
junction temperature is simply added to but it IS only a question of fragility; there
that indicated on the meter. If the meter is no difference in the effectiveness.
and thermocouple are CALIBRATED with The couple wires must, of course, be
the cold junction at 15*C the error will be insulated behind the actual junction, and
negligible. either single or twin bore ceramic or silica
The construction of the thermocouple insulators can be had for this purpose
is simplicity itself. It can, indeed, work from scientific instrument suppliers.
quite satisfactorily if the wires are tightly (Griffin & George, Gallenkamp etc) or
twisted together, This will read correctly, from Morgan Refractories, already
but there will be a time-lag between mentioned in connection with muffles.
Some of these are seen in Fig. 63. These that it is otherwise, and resistances of
may, in turn, be protected by setting the around 750 ohm are common. Let us
whole inside a steel sheath. The end can suppose it is shown as 735 ohm. This is
be closed either by folding over and ham- where the resistance “R” of Fig. 62 comes
mering, or by welding in a plug. The in. If we make this 1000 — 735 = 265
sheath should be a fairly close fit on the ohm, then we have a total resistance of
insulators. This will protect the couple 1000, and the meter will read direct. Such
from knocks, and also (in the case of the a resistance would be "built up" with one
salt pot) from a gradual eroding away of at 200, one at 50, and one at 15 ohm:
the wires. On the other hand, the sheath again, from the radio service shop, "High
will cause a time-lag in the indications — Stability 196 Tolerance" type.
your indicator will tend to read the However, suppose the meter has à
temperature as it was 3 or 4 minutes resistance HIGHER than 1000 ohm, what
earlier, depending on how heavy the tube then? | have one here, 0-100 wA, 1250
is. There is no need, for our purposes, to ohm. Apply Ohm's law. C — E/R, or, more
use stainless steel. Once clear of the conveniently, E = C x R. 100 uA x 1250
actual furnace the wires may be insulated f) = 125,000 microvolts, or 125
with ordinary sleeving, or you can use the millivolts. So, for a meter with a resistance
glass-fibre type if they are likely to get hot. “R” exceeding 1000 ohm, the conversion
Such sleeving can be obtained lor is "Actual Mv — Scale Mv x R/1000",
ordered) through radio service shops. where R is the meter resistance. It is as
The Indicator. The maximum reading easy as that! Fig. 64 shows three meters,
using iron-constantan will be about 50 one circular "Ex Government" one, which
millivolts, and for BS1827 NiCr-NiAI, is still as good as new; a rather small
about 35 mV. If you have a "multimeter" Japanese meter, really too small a scale
with a millivolt scale, well and good: you length: and a large 0-100 meter which is,
can use this. Otherwise you must either in fact, the one | would use despite its
buy one or make one. There was a time resistance of 1250 ohm.
when "surplus" micro-ammeters were Calibration. This is simply a check on the
readily available, and these are of excel- table, and ought to be done just in case
lent quality. Those reading up to 50 or the meter resistance is not as stated (they
100 uA would be the most suitable. But all are made to a tolerance] and after a
such meters are not expensive new and, period, in case there has been a change in
again, can be ordered through radio the characteristics of the couple elements.
service shops. A 0-50 meter will do nicely, The lower end is done at 100°C. The ther-
but you must know its coil resistance, and mocouple is set in the steam above
if this is not stated in the catalogue, obtain boiling water — NOT in the water itself.
the information from the makers before See Fig. 65, though it can be done in an
buying it. electric kettle just as easily, Make sure
Now, let us suppose that we have the that the cold junction (i.e. the meter, if
meter, O-50uA, and its resistance is used as in Fig. 62) is at 15°C. Let the
exactly 1000 ohm. We are home and dry, junction soak for about ten minutes and
for one millivolt will give a deflection of 1 observe the reading. Let it cool a little, and
micro-amp, and the scale can be used repeat. Do this three times and take the
directly. (The external resistance of the mean. For an intermediate temperature
couple won't matter). Suppose, however, you can, if you have a tempering salt bath

110
Fig. 64 Micro-ammieters. Left, 14
in. Square O-SOpA, centre,
“Surplus” O= f00uA, right 44 in.
0-100pA. normally used by the 7 7
author. t p

and a mercury thermometer, use this.

Strap the end of the couple to the ther-
mometer bulb with care and a bit of thin
wire and raise the temperature to sóme
figure about the limit of the thermometer.
Make sure the heater is stable, and take (4
readings, three, with the temperature
falling about 30°C and then raised again
between each, and note the mean of the — VENT.
three.
For higher temperatures the following
can be used. Lead, melting at 327°C, Zinc
at 419°C (but be careful of the fumes)
and pure aluminium (from an old
saucepan) 658°C. In the case of the lead THERMOCOUPLE.
you will need a layer of resin on the
surface to prevent oxidation, and may
need to skim off oxide as well. The drill in
this case is different. You bring the metal
up above melting paint and immerse the
couple. Take away the heat, and start
noting the temperature indicator every 30
seconds or so. As it cools you will find an
"arrest point" developing (Fig. 66) and
then as the metal is solid, the temperature
falls again. Now reheat. and again note
the temperatures — or the millivolts. Again
Fig. 65 Ono method of calibrating a thermocouple in steam.
Any sort of container will do.

111
there will be an arrest in the INCREASE in The final alternative is expensive, but
temperature. Repeat this a couple of definitely "in the mood". You can buy lor
times, always removing any scum which some of you may be able to make, for all |
forms on the surface of the metal, The know) a solid state digital indicator, which
mean of the arrest points is the actual will read out the temperature on light-
melting point of the metal. You can do emitting diodes. If you have several ther-
another at 801°C, with pure salt. Note, mocouples, high temperatures and low.
NOT the household variety. Ask at the you can build in to this a correcting circuit
chemists for "Sodium Chloride BPC" — which will allow for the use of copper-
this is common salt, but chemically pure. constantan at one end of the temperature
Once all this has been done you can range and Platinum Rhodio-platinum at
either make up a table or make a new the other: a universal pyrometer indicator.
scale out of paper to stick on the dial. | This is not a book on electronics, so that |
use a table, as | have other uses for micro- must leave the detail for you to sort out.
ammeters and don't want to tie any of Sufficient for most of the electronically
them up on one job. minded to know that it can be done!

112
APPENDIX 2

Carbon Steel Cutting Tools

Many readers may never have made a The only disadvantage which afflicts
cutting tool in their lives, apart from the the material is that it “tempers” at quite a
odd D-bit, and would regard anything low figure. HSS is, initially, softer than
except high-speed steel lor even tungsten carbon steel, (Fig, 66) but whereas the
carbide) as being a very inferior tool latter would start to lose some of its
indeed. cutting hardness at about 270°C at the
This is a mistake. The ONLY reason for tool point the former can be operated so
using HSS is that it can operate at a hot that the cutting oil will smoke and,
higher temperature, and can, therefore, indeed, | have seen special grades run for
cut FASTER. HSS is much more expen- demonstration purposes cutting at a dull
sive; it is not as hard as properly treated red heat. The same is true of Tungsten
carbon tool-steel; at the proper cutting Carbide: its virtue is not its hardness — it is
speed carbon steel has a longer tool life, intended as a FAST CUTTING TOOL,
and for finishing cuts below 8BOft/min. perhaps at three times the speed of HSS.
lasts much longer. And (again, when This is the crux of the matter. You can
properly treated) carbon steel will give a buy a couple of feet of carbon tool steel
better finish. Nor is it true that "you can't for the price of one short HSS toolbit: you
take heavy cuts", This is nonsense. | have can file it, bend it, forge it, and harden it to
by me a table of "speeds and feeds" for cut tough alloy steel, even hard cast iron;
carbon steel lathe tools which lists cuts up and you can get superb tool finish on the
to one inch deep at feeds of one eighth of work. But you MUST keep the cutting
an inch per rev. Finally, carbon steel will speed down and keep the tool point cool.
tackle "the really hard stuff". Until quite For roughing cuts the speed should be
recently it was used for machining the about one third that used for HSS (though
chilled cast iron rolls used in steelworks if the coolant supply is good you can try
(the surface deliberately made as hard as faster) and for normal finishing cuts the
the hard spots you can't machine on your machine should be run at about two thirds
castings!) and watchmakers habitually cut that used for HSS. In fact, for brass or
hardened and tempered steel pivots with free-cutting mild steel, taking model
à carbon steel graver. engineers' finishing cuts, there may be no

113
 1200.

cof
cT o E SPEED

ER
N*
HARDNESS
VICKERS
ue 200 400 600. BOO.
TEMPERATURE ~ °C.
Fig. 66 The effect of working temperature on the hardness of typical carbon and high-speed tool-stee!, Once the tool reaches a
particular temperature the former hardness will not be restored on cooling, Note that carbon Stew! has the advantage ifthe tip
temperature can be kept below about 250°C

need to reduce speed at all. (| am, of “tool finish" for the decorative reflective
course, assuming that you use the surfaces which are so much admired, and
"proper' HSS cutting speeds; many even go to the length of lapping the
model engineers run at carbon steel cutting edge with rouge on an iron lap!
speeds all the time!) Carbon steel will cut With the ever-increasing cost of all
anything that can be cut by HSS, and alloy steels (and HSS in particular) it does
many that we often use carbide for (and make some sense to reconsider the place
doesn't chip at the edges either) but it of the "old fashioned" carbon steel. We
does take longer. But what is the hurry, are not all that concerned with optimum
anyway? floor-to-floor times, and few of us would
On the question of "quality of finish" it ever consider working our tools as hard as
is not always appreciated that the surface they do in industry. (The "economic tool
of the workpiece is, when taking very fine life" may be no more than 42 minutes on
cuts, a reflection of the surface of the tool. centre-lathe work!). | would not, of
If the too! point has grooves from the course, suggest that we abandon HSS
grinder it must leave grooves or ribs on wholesale; that would be foolish. But for
the work. Carbon steel has a finer grain what it is worth | will conclude this note
than most HSS, and if grain refined, even by observing that ALL the crankshafts |
more so. Properly hardened, and have made, and that is quite a lot, have
tempered only so far as is essential, it been finished, and most of them roughed,
takes up a better finish from the oilstone, using carbon steel tools on the crankpins,
thus giving an even better finish to the and most have been turned on the
work. In “Ornamental Turning" we have journals that way as well. Why not give it
to go further still; we rely entirely on the a trial?

114
APPENDIX 3

British Standard Steel

Specification Numbers
The Old British Standard 970/1955, sulphur content, again, » 100. 224 will be
characterised by the "EN" numbers which a 0.2496 sulphur steel.
many of us still use, was replaced by the The second group of two digits
new BS970 in 1972. It really is time that indicates the carbon content multiplied by
we started to use it! |t has one great 100. Thus 45 indicates a 0.4596 carbon
advantage, in that the BS number itself steel. The letter shows the basis of the
gives us the carbon content for many specification. "M" means that the steel
steels. must meet the mechanical properties of
The number comprises three digits, the material; "A" means that it must be to
followed by a letter, followed by two more the specified chemical analysis, other
digits. The first three digits serve two properties being secondary; “H” means
purposes. They give a broad indication of that it must meet the requirements of Har-
the TYPE of steel; thus — denability. Most of the steel we use is to
000 to 199 are all "plain" carbon steel the "M" specification.
with some manganese content There is, of course, a tolerance on the
200 to 240 are "Free-cutting" versions of 96 figures for carbon and other materials;
the above thus 0.2% carbon may lie between 0.16
300 to 499 are all "Stainless" or "Heat and 0.24% — though the user can
Resisting” steels (provided he pays) specify closer limits.
500 to 999 are all "Alloy" steels and As a general rule, therefore,
within the number the main alloying approximate heat treatment procedures
elements can be identified. can be deduced from the specification
We are concerned (for heat treatment) number. Thus O80M30 is a 0.3% carbon
only with the plain carbon steels and, 0.8% manganese steel, which may be
perhaps, their free-cutting equivalent. The heat treated by quenching from about
first digit can be ignored; O and 1 mean 875°C and tempering to suit the applica-
"plain" and 2 means "free-cutting". The tion. We can also deduce from the iron-
second two digits, (preceded by O or 1) carbon diagram on page 18 that it should
give the manganesa content multiplied by be normalised at around 880°C.
100; thus 145 means 0.45% manganese. The specification does not cover "Tool
If preceded by 2, these digits indicate the Steels”, with more than 1% carbon.

115
 APPENDIX4

Approximate Conversion of
Hardness Scales
(ASTM Conversion Tables)

(For Steels)
SHORE VICKERS ROCKWELL UTS
No VDP "e" 1000 Lbf/Sq.in.
99 1030 70
98 970 69 —
97 940 68 -
95 900 67 -
93 880 66.4 -
91 840 65.3 -
90 820 64.7 =
88 800 64 —
86 760 62.5 -
84 740 61.6 -
82 710 60.5 -
80 680 59.2 330
78 650 57.8 314
76 630 56.8 304
74 600 55.2 289
72 580 54.1 280
70 550 52.3 264
68 530 51.1 254
66 500 49.1 240
64 480 47.7 230
62 460 46.1 220
 (For Steels)
SHORE VICKERS ROCKWELL UTS
No VDP C 1000 Lbf/Sq.in.
60 446 45 213
58 430 43.6 204
56 410 41.8 195
54 396 40.2 194
52 380 38.8 180
50 360 36.6 170
48 354 34.9 163
46 330 33.3 156
44 314 31.4 148
42 300 29.8 141
40 280 27.1 131
38 270 25.6 126
36 250 22.2 116
34 240 20.3 111 Rockwell "C" is not
32 220 15.7(95B) 101 applicable below this
30 210 13.4(93.4B) 97 range. Rockwell "B"
28 190 8.5(89.5B) 88 in brackets
26 180 6.0(87.1B) 84
24 160 O (81.78) 75
22 150 (78.7B) 71
20 130 (71.2B) 62
18 122 (67.68) 58

These conversions must be used with

discretion, especially as regards the
Ultimate Tensile Strength column. They
relate only to Carbon and low alloy steel,
and exclude “Stainless” steels.

117
Glossary of Terms
ALS American Iron and Steel Institute.
A.S.M.E. American Society of Mechanical Engineers
A.S.T.M. American Society for Testing Materials
Acicular Crystals forming a needle-like structure in place of the
more usual grains.
Alloy Steel Steel containing elements other than iron and carbon,
though up to 0.3596 Manganese is found in all steels.
Annealing A heat treatment which softens an already hardened
steel.
Arrest point Points at which the temperature ceases to rise for a short
period when the steel is heated slowly, and similarly
when cooling.
Austempering A constant temperature transformation process, mainly
used for small parts, springs, etc. Fig. 44.
Austenite A solid solution of carbon in iron.

B.S.. British Standards Institution.

Bainite A structure obtained when Austenite is transformed at a
constant low temperature, less hard, but tougher than
Martensite. Typically the result of "Austempering .
Billet A square or rectangular bar, usually the raw material for
later hot forming processes.
Blister Steel The name given to the steel after the completion of the
cementation process. The surface of the metal is at this
point covered in scale blisters.
Body Centred Applied to crystals, one with an atom at each corner and
one in the geometric centre of the lattice.
Brine Concentrated solution of common salt (NaCl) in water.

Carburise Any process which causes absorption of carbon,

Caseharden A surface hardening process which retains the normal
structure of the steel in the core of the part. Normally
achieved by surface carburising, heating, and quenching.
Cast Steel The name given to high-carbon steel which has been
melted after the cementation process is complete, At one
time synonymous with tool-steel it is now liable to be
confused with castings made from "mild" steel. The
terms “Crucible Steel" or "High Carbon Steel” are to be
preferred.

118
Cementation The process by which Wrought Iron is caused to absorb
carbon through the full thickness of the metal. to produce
“Blister” or “High Carbon” steel.
Cementite Metallurgical name for Iron Carbide, Fe,C. So named
because it was first identified in steel made by cementa-
tion.
Critical Temperature The temperature at which structural changes occur in the
metal. Usually, but not always, coincident with the
"Arrest Points".
Crystal An orderly arrangement of atoms, almost always in
precise geometric shapes. Iron crystals are always of
cubic form.
Curie Point The temperature above which steel loses its magnetic
properties,

Dannemorra An ironworks in Sweden, which was reputed to produce

an iron remarkably free from impurities.
Decarburise The reduction of carbon content, usually on the surface of
the metal. Caused as a rule by the action of oxygen or
iron oxides at high temperatures.

EN No. The “Engineering Number’, the identification number of

steels used in B.S.970/1955, now superseded.
Equilibrium Diagram A chart plotting the critical temperatures of steel against
the carbon content. Fig. 13.
Eutectoid The composition of steel at which the upper critica!
temperature coincides with the lower. "Eutectoid" steel
is entirely pearlitic.

Face Centred Applied to crystals, one with an atom at each corner and
one in the centre of each of the faces of the lattice.
Faggoting The act of welding together bars of Wrought Iron to make
billets.
Ferrite Metallurgical name for the almost pure iron observed in
the microstructure of steel. (Not to be confused with the
"sintered ferrite" used for coil cores etc in electronic
equipment.)

Gauge Stock An oil-hardening carbon-chromium alloy steel compoun-

ded to reduce distortion on quenching to a minimum,
Sometimes known as "ground flat stock".
Grain Aggregates of crystals. These form when the metal
solidifies, but may change their shape and disposition as
it cools.
Grain Refinement Any process which either reduces or makes more uniform
the size of the grains.

119
High Carbon Steel Steel with a carbon content above 0.85%. More
accurately, steel which shows free Cementite in the
microstructure.
High-speed steel Usually a high-Chromium high-Tungsten alloy with
enhanced hot-hardness, but may not be a steel at all.
"Stellite" is an alloy of Cobalt, Chromium and Tungsten,
with no Iron content.

IZOD number The energy in foot-pounds required to break a standard

specimen under a prescribed impact load. A test for
shock resistance.

Lattice A framework of imaginary lines joining the location of

atoms in a crystal.
Liquidus The temperature above which the metal is entirely liquid.

Martempering Heat treatment process designed to reduce distortion on

large or complex components. Fig. 45.
Martensite The structure formed when a combination of Austenite
and Cementite is cooled too quickly to allow the normal
transformation to Pearlite.
Melting Range The range of temperature between the Solidus and the
Liquidus, where the metal is in a pasty state, with both
solid and liquid present.
Muffle A furnace in which the work is protected from the direct
impingement of the heating flame. It is now applied also
to most electric resistance furnaces.

Neutral Salt A salt which is used as a heating medium but which

neither carburises nor decarburises the metal.
Nodular A condition where the Iron Carbide grains are in the form
of nodules rather than the normal needle-like shape.
Normalise Heat treatment intended to restore distorted grains to
their normal shape. A form of grain refinement.

O.H. & T. Oil hardened and Tempered.

Pearlite The normal transformation product when Austenite is

slowly cooled through the lower critical temperature. It
consists of sub-microscopic "plates" or laminae of Ferrite
and Cementite.
Puddling The process of converting cast iron into wrought iron.

RNic} Rockwell Hardness test number, on the "c" scale, the one
normally used for hard materials.
Recalescence The name sometimes used for the critical temperatures.

120
S.A.E. Society for Automotive Engineers (USA).
"S" Curve Curve showing transformation rates against temperature
and time. (Fig. 16),
Scale Iron Oxide forming on the surface of hot steel.
Solid Solution The condition where the atoms of one solid substance are
disposed within the crystal lattice of another. Analagous
to the more usual liquid solution.
Solidus Temperature below which the metal is entirely solid.
Steel Iron-carbon alloy with a carbon content between 0.01%
and 1.796. Before about 1855 the term was universally
applied only to "High Carbon Steel" as we know it.
Silver Steel High Carbon Steel to BS.1407. So-called because it is
usually ground and so presents a “silvery” appearance, in
contrast to the normal "black" tool steel of earlier days.
Typically 1.1-1.2% Carbon, 0.3-0.496 Managanese, 0.4-
0.5% Chromium, 0.1-0.25% Silicon, and maxima of
0.3596 Sulphur & Phosphorus.

Tempering A reheating process which modifies the structure of a

heat-treated steel. (Note that in early books the
"Temper" of a steel often referred to its carbon content).
T.T.T. Curve "Time-Temperature Transformation Curve - the "S"
Curve of Fig. 16. A chart which shows the time taken for
constant temperature transformation of steel,
Transformation Any change of the microscopic structure or chemical
composition of steel brought about by heat treatment.

U.T.S. Ultimate Tensile Strength. The stress at the breaking

point, measured under prescribed conditions on a
standard test piece.

V.D.P. Vickers Diamond Pyramid. The VDP number is a measure

of the hardness of the material.

Wrought Iron The predecessor to steel. Made by decarburising cast iron

in the puddling furnace.

Yield Point The stress beyond which metal will not return to its
original length when the load is removed,

121
 INDEX
See also the glossary on page 118 for definitions of terms used.

Acicular grains Dannemorra

Air/gas torches Decarburising
Die quenching
Die tempering
Distortion
Domestic fire heating

16
Electrical safety
Austenising temperature 28, 29. 30, 33 EMF of thermocouples
"E.N" number
Bainite 118
Energy controller
Bench, heat treatment 102
Equilibrium diagram
Bessemer Steel 11
Eutectoid steel 19,120
"Best" wroughtiron 10
Blacksmiths hearth 51
Blastfurance i Face centred crystal
Blazing off 49 Faggoting
Blister Stee! 118 Farrite
Blowlamps 53 File steel
Blowlamps flame 54 File test, hardness
"Blueing steel Fire extinguisher
Body centred crystal Fire risks
Bricks Forging
Bricks, heating Frothing lof salt bath)
Brine. proportions FURNACES.
Brinell test B4,116 Blast
British Standard Steel specification Electric
Burns, frorn hot salt Gas fired
treatment Mutle, gas
Muffle, home made
Carburising 64, 67Case hardening Puddling
Cast Steel Radiant
Cementation Salt. high temperature
Cementite 17, 1319
Clothing. protective
Gas furnace 57,90
Coil springs
Gas torches 54
Cold junction
Gauge stock 24,120
Cooling rates
Goggles 103, 104
Constantan
Grains 14,118
Control systems, furnace
Grain refining (carburising) 74
Cosmetic heat treatment
Ground Gauge Stock 24
Critical temperature
steal
Crystals Hardness comparison, chart
Curie point Hardness comparison HSS and toolsteel
Cutting speed, carbon steel Heat testing

122
Heat pipe Rototherm thermometer 107
Heat treatment for toughness
Heating time S.A.E, 121
High-speed steel 72,82,120 S-curve 21
Home-made muffles 90 Salt bath, Austenising 60,98
Hot junction (thermocouple) 107 Safety 104
Huntsman progress 11 45
21,73
Insulating bricks 56 85
Insulating wool 98 Seegar Cones 31
Insulators. thermocouple 109 Shore hardness test B5, 116
tron 7 Silver steal 25
-carbide 17 Annealing 71
Pig Austenising temperature 29
Wrought Specification 121
Iron-carbon equilibrium diagram 42
IZOD number 8
Solid Solution 15
Kasenit process Solidus 321
Kiln, gas-fired Spirit lamp 47,62
Springs 24,77,79
Lathe tools, carbon steel Steel, Alloy 25
tempering Carbon 12
Liquidus Casehardening 58
Cast oc Crucible 11
Martempering File 29
Martensite Medium carbon 75
Millivoltmeter Hardness and Toughness 20
MOH's hardness scale Mild 11
Muffte — Electric Stress relieving 72
Gas fired Surface hardening 64
Home made
Neutral saits Temperature
Nicherome wire Annealing
Normalising Austenising
Colours (plates)
Open fire heating Holding times 28.44
“Omamental Turning" tools Indicator crayons
Overheating Judging
Oxy-acetylene Normalising
Stress relieving
Paraffin blowlamp Tempering see below,
Pearlite Tempering 24 41,50
Propane torch Colours 36,43,47
"Puddlina Lathe tools
Pyrometers Tools, various

Quenching, in brine
in oil
in water Through hardening
Quenching techniques Time-Temperture transformation
Tongs
Radiant element furnace Tool materials
Recalescence Torches
Re-hardening Torch flame temperature
Resistance wire Transformation diagrams
Rockwell hardness test Tube furnace
UTS.

Vickers Hardness Test (VOP)

Wire, electric heating

thermocouple

Yield Point
LIST OF TABLES

Carbon context of tool steels

Colour & Temperature.
700-1000°C
215-330°C |
Hardness. Comparison of scales
Typical values
Temperatures.
Annealing
Austenising
Grain refining
Normalising
Oven, domestic
Quenching
Tempering
Thermocouple wire, EMF

124
WORKSHOP PRACTICE SERIES irom Argus Books
1. Hardening, Tempering and 8. Sheet Metal Work 15. Workholding in the Lathe
Heat Treatment R.E. Wakeford Tubal Cain
Tubal Cain 9. Soldering and Brazing 16. Electric Motors
2. Vertical Milling in the Home Tubal Cain V.J. Cox
Workshop Pup 10. Saws and Sawing *, 17. Gears and Gear Cutting
Atnold Throp lan Bradley l. Law
3. Screwcutting in the Lathe —* 11. Electroplati 18. Basic Benchwork
Martin Cleeve 4 rui n Les Oldridge
4. Foundrywork for the Amateur a d ! 19. Spring Design and Manufacture
B.T. Aspin 12. Drills, Taps and Dies balean
5. Milling Operations in the Lathe Tubal Cain 20. Metalwork and Machining
Tubal Cain 13. Workshop Drawing Hints and Tips
6. Measuring and Marking Metals Tubal Cain lan Bradley
Ivan Law 14. Making Small Workshop
7. The Art of Welding Tools
W.A. Vause S. Bray

1. Hardening, Tempering and Heat Treatment

In this valuable book Tubal Cain takes the reader beyond the superficial or
the simply practical with explanations of the composition of steel, its
additives, and the effects of different temperatures on its constituents. With
a grasp of what changes are actually taking place in the metal the care
needed in following the practical processes described becomes under-
standable and will lead to better and more consistent results. Flame, salt
bath and furnace heating are detailed, with information on accurate
measurement or recognition of temperature levels.
For the average small workshop operative or model engineer the
discourses on tool material, hardening and tempering will be of most use,
and in this connection this book replaces the earlier “Hardening and
Tempering Engineers' Tools", providing a broader-based, more detailed
and up to date examination of the subject.

t ISBN 0-ġ5242-837-5

9 TE