-
Defects and Distortion in Heat-Treated Parts Anil Kumar Sinha,
Bohn Piston Division
MOST OF THE PROBLEMS in heat- treated parts are attributed to
faulty heat- treatment practices (such as overheating and burning,
and nonuniform heating and quench- ing), deficiency in the grade of
steels used, part defect, improper grinding, and/or poor part
design. This article discusses overheat- ing and burning, residual
stresses, quench cracking, and distortion in some detail and offers
some suggestions to combat them.
Most of these conditions result in a charac- teristic appearance
of the treated parts that can be easily recognized by simple
inspec- tion. Some of these factors do not produce any
distinguishing features in the semifin- ished or finished part. In
particular, some of the visual evidence does not recognize the
presence of overheating and burning and the development of residual
stresses leading to distortion, quench cracking, and eventual
failure of the heat-treated parts; metallurgical laboratory
examination is needed to establish these problems that contribute
significantly to the service performance of the part. Tool
designers must also be aware of the problems and difficulties in
manufacture, heat treat- ment, and use.
Overheatin 8 and Burning of Low-Alloy Steels
When low-alloy steels are preheated to high temperature (usually
> 1200 C, or 2200 F), prior to hot mechanical working (such as
forging) for a long period, a deterioration in the room-temperature
mechanical properties (particularly tensile ductility and impact
strength or toughness) can be obtained after the steel has been
given a final heat treatment (comprising reaustenitizing,
quenching, and tempering) (Ref 1-3). Linked with the im- paired
mechanical properties is the appear- ance of intergranular matte
facets on the normal ductile fracture surface of an impact
specimen. This phenomenon is known as overheating and has been a
matter of con- cern, especially in the case of steel forgings.
Overheating has also been noticed in steel castings (due to
variation in pouring temper-
ature and effectiveness of the proprietary grain inoculants
applied to the mold surface), in heavily ground parts, and in
affected zones of welds (Ref 4). The usual practice is to reject
the overheated products as being un- suitable for service.
It has now been established that over- heating is essentially a
reversible process caused by the solution of MnS particles in
austenite during heating or reheating at high temperatures; the
amount increases with temperature, and its subsequent reprecipi-
tation during cooling occurs at intermediate rates as very fine (
-0 .5 i~m) arrays of a-MnS particles on the austenite grain
boundaries. On subsequent heat treatment the intergranular network
of sulfides may provide a preferential, lower-energy frac- ture
path in contrast to a normal transgran- ular fracture path. As a
result, when impact loaded, a ductile intergranular fracture de-
velops due to decohesion of the MnS/matrix interface and progress
of microvoid coales- cence. Figures 1 (a) and (b) show the usual
appearance of the fracture surface at differ- ent magnifications
(Ref 1).
When the low-alloy steel is preheated prior to hot working at
too high a tempera- ture (normally > 1400 C, or 2550 F), local
melting occurs at the austenite grain bound- aries as a result of
the segregation of phos- phorus, sulfur, and carbon (Ref 5). During
cooling, initially dendritic sulfides (proba- bly type II-MnS) form
within the phospho- rus-rich austenite grain boundary, which then
transforms to ferrite. This results in excessively weak boundaries.
Subsequent heat treatment provides a very poor impact strength and
almost completely intergranu- lar fracture surface after impact
failure. This phenomenon is termed burning. Burn- ing thus occurs
at a higher temperature than overheating. If this occurs during
forging, the forging will often break during cooling o r subsequent
heat treatment (Ref 4).
Detection of Overheating There are two basic methods for the
determination of the occurrence of over-
166,6 ~rn
I I 12.5 p, rn
Fracture surface of an impact loaded speci- Fig 1 men. (a)
Appearance of intergranular fracture of 4.25Ni-Cr-Mo steel
containing 0.34% Mn and 0.008% S, in fully heat-treated condit ion
but after cooling from 1400 C (2550 F) at 10 C/min (20 F/rain). (b)
Same specimen as in (a) but at higher magnifica- t ion, showing
ductile dimples nucleated by MnS par- ticles precipitated at
austenite grain boundaries. Courtesy of The Institute of Metals
heating, namely, fracture testing and metal- lography (or etch
testing). Overheating may also be detected by a decrease in
mechani-
ASM Handbook, Volume 4: Heat Treating ASM Handbook Committee, p
601-619
Copyright 1991 ASM International All rights reserved.
www.asminternational.org
-
602 / Process and Quality Control Considerations
Table 1 Etching characteristics of overheated and burned steels
Reagent Method Action on overheated steel Action on burned
steel
2.5% nitric acid in ethyl Swab surface for 30 s May produce
grain contrast, but White boundaries outlining alcohol not
indicative of overheating preexisting austenite grains
Saturated aqueous solution White boundaries outlining Black
boundaries outlining of ammonium nitrate preexisting grains
preexisting austenite grains
Aqueous 10% nitric acid + 10% sulfuric acid
85% orthophosphoric acid (Fine's reagent)
Oberhoffer's reagent
Black boundaries outlining preexisting austenite grains
Does not differentiate between overheated and nonoverheated
steel
Does not differentiate between overheated and nonoverheated
steel
Source: Ref 13
Electrolytic, specimen anode, current density 1.0 A cm -2 (6.5 A
in. -2)
Etch for 30 s, swab surface; repeat three times, then repolish
lightly
Electrolytic, specimen anode, current density 0.15 A cm -2 (1.0
A in.-2), etching time 15 min
Swab surface for 30 s
White boundaries outlining preexisting austenite grains
Attacks inclusions at grain boundaries
Shows phosphorus segregation at grain boundaries
cal properties. But such changes are not very marked unless
overheating tempera- ture is high or overheating is too prolonged
or severe; in some instances the mechanical properties do not
change, even after the observation of extensive faceting. Usually
the two methods mentioned above should be used in conjunction with
some measure of toughness by impact or other testing in order to
get a clear understanding of the degree and severity of overheating
(Ref 2).
Fracture Testing. The direction of fracture testing is important
in steels manufactured by conventional methods. It has been ob-
served by some workers (Ref 6) that the longitudinal fracture test
specimens parallel to the rolling direction do not exhibit face-
ring until the corresponding transverse frac- tures display
extensive faceting. However, the testing direction in
electroslag-refined (ESR) steels has been found to be insignif-
icant (Ref 7).
The scanning electron microscope is con- sidered to be the best
and most convenient tool to detect the facets on the overheated
fracture surfaces. These facets are charac- terized by small,
well-defined, ductile dim- ples; each dimple is usually nucleated,
pre- sumably by fine arrays of inclusion particles: a-MnS particles
(Fig 1) in Mn- bearing steels (Ref 8, 9) or chromium sul- fides in
Mn-free steels (Ref 10, 11).
It is now well recognized that the fracture test specimen should
always be tested in the toughest possible state (for example,
quenched and highly tempered [in the range 600 to 650 C, or I 110
to 1200 F] steels after high-temperature austenitization) because
this condition is most prone to overheating effects. Baker and
Johnson (Ref 5) have suggested that an increased proportion of
facets in the fracture specimens with in- creasing tempering
temperature is attribut- ed to the corresponding increase of the
plastic zone size. In this case a slight amount of weakening will
be sufficient to impart faceting because the grain boundary
strength becomes lower (Ref 2). It should be noted that the
existence of facets in the fractured specimens is not always
associat-
ed with a lowering of impact strength (Ref 12).
Metallography (or Etch Testing). The most widely used etchant
technique uses Austin's reagent (aqueous solution of 10% nitric and
10% sulfuric acids), ammonium persulfate, molten zinc chloride,
saturated solution of picric acid at 60 C (140 F), and an
electrolytic etch based on saturated aqueous ammonium nitrate.
Table 1 shows the etching characteristics of overheated and burned
steels (Ref 13). The etchant procedure with Austin's etchant is as
fol- lows: The sectioned specimen is etched for 30 s in the
etchant, removed, washed off, and repeated three times. If the
steel has been overheated, the original austenite grain boundaries
will be preferentially at- tacked, and a black network of etch pits
will be observed under the microscope (Ref 14). According to Preece
and Nutting (Ref 13), the best results are obtained when ammoni- um
nitrate etch is applied on the sectioned steel specimen in the
fully heat-treated con- dition where this etchant preferentially
at- tacks the matrix (original austenite grains), leaving the grain
boundary unaffected (which appears as a white network). Bodimeade
(Ref 15) concluded that all these etchants did not cope with mildly
overheat- ed low-sulfur steels. Table 2 is a summary of the results
of potentiostatic etching tech- niques carried out by McLeod (Ref
12) using nitric-sulfuric, saturated aqueous pic- ric acid (at 60
C, or 140 F), and ammonium nitrate etchants. He considered that
when the suitable etching conditions were estab- lished, the
potentiostatic etching method rendered more reliable and
reproducible results as compared with the conventional etching
techniques. However, the same problem with mildly overheated
low-sulfur steels still persisted. Hence, the use of etch tests for
low-sulfur low-alloy steels is not recommended for the detection of
mild overheating.
Detection and Effects of Burning Burning is not commonly
encountered.
The two etchants (namely, nitric-sulfuric
acid and ammonium nitrate solution) used for overheating can be
successfully em- ployed for detecting burning. When applied to
burned steels, these etchants react in a manner opposite to that of
overheated steels. Preece and Nutting (Ref 13) found ammonium
nitrate solution to be the ideal reagent to detect this phenomenon.
Other reagents are Stead's and Oberhoffer's re- agents, which may
also be used to check the burning effect. However, these etchants
are unable to differentiate between overheated and nonoverheated
steels.
Factors Affecting Overheating The occurrence and severity of
overheat-
ing depend principally on important factors, notably steel
composition, temperature, cooling rate, and method of
manufacture.
Composition. Sulfur is the constituent that greatly influences
overheating. For steels with less than 0.002 wt% sulfur, over-
heating does not occur; this is because of the very low volume
fraction of sulfides formed. However, the commercial produc- tion
of such very-low-sulfur steels (for ex- ample, ESR steels) is
expensive. Above this level of sulfur, the overheating onset tem-
perature rises with the increasing amount of sulfur. It has now
been explained that steels with low sulfur content (0.01 to 0.02%)
are more prone to this defect than those with high sulfur content
(>0.3%) because the transgranular strength is high, and
therefore a small amount of grain-boundary sulfide precipitation is
enough to induce intergran- ular failure (Ref 16). The phosphorus
con- tent has been regarded with the most con- cern in connection
with burning. At constant phosphorus level, there is an in- crease
in the overheating temperature with the increase of sulfur content,
whereas the burning onset temperature decreases. Burn- ing
temperature is reduced with the increase in phosphorus content. At
low sulfur con- tents, a wide gap between overheating and burning
temperatures exists. For example, in the case of vacuum remelted
steels, the temperature gap between the onset of over- heating and
burning is -300 to 400 C ( -570
-
Defects and Distortion in Heat-Treated Parts / 603
Table 2 Summary of potentiostatic etching experiments Best
etching conditions
Anodic loop Solution voltage, mV Observed effect Voltage, mV
Observed effect Comments
Saturated aqueous -400 Slight general 2200 (for 2 Classic white
ammonium etching min) boundaries on a nitrate dark background
Aqueous 10% nitric acid + 10% sulfuric acid
Saturated aqueous picfic acid at 60C (140 F)
Source: Ref 12
200 Vigorous None dissolution of specimen; formation of flaky
black film
-250 Milder attack; About -250 large black (for 30 s) pits in
mildly etched matrix
100 No real, None positive indication of overheating
Discontinuous array of grain-boundary pits and some random pits
within grains
Operates best in the transpassive region at >+1500 mV; time
at any potential is important
Underetching: random array of black pits
Overetching: uniform black surface film
Most aggressive etchant of the three examined
Polish lightly after etching to eliminate matrix etching
effects
Anodic loop very weak, necessitating long etching times because
current density is very low; Teepol additions gave no
improvement
to 750 F) and there is a remote possibility of burning occurring
within the forging range, unless the overheating is severe (Ref 2).
However , at high sulfur content the gap becomes narrow.
Temperature. To avoid overheating, care must be exercised in
choosing a correct heating temperature so that uneven heating,
flame impingement, and so forth, do not occur (Ref 3).
Cooling Rates. The cooling rate through the overheating range
affects the size and dispersion of intergranular et-MnS particles.
The intermediate cooling rate generally em- ployed, 10 to 200 C/min
(20 to 360 F/min), gives rise to maximum faceting as well as to the
greatest loss in impact strength. How- ever, slow and rapid cooling
rates will sup- press overheating. At very slow cooling rates, the
sulfide particles become large, small in number, and more widely
dis- persed, and they have no more deleterious effects than the
other inclusions already present. At rapid rates, the sulfide
inclu- sions are too fine to produce any damaging effect (Ref
17).
Methods of Manufacture. Electroslag- remelted steels are less
susceptible than vacuum-remelted steels, presumably due to the
difference in oxygen level. Similarly, nickel steels are more prone
to overheating. Vacuum-remelted steels have a lower over- heating
temperature than some comparable air-melted steels.
Prevention of Overheating and Burning
For preventing overheating of steels, a properly selected
temperature should lie
between a temperature low enough for the metal to be safe and
high enough to be sufficiently plastic. The better the tempera-
ture control, the better the compromise.
Severe overheating can be reduced to mild overheating by soaking
the steel at 1200 C (2200 F); with care, it may be removed
completely. Hot working through the overheating range to a low
finish tem- perature is also reported to remove the effects of
overheating.
The alloying additions with a greater sul- fide-forming
tendency, such as calcium, zir- conium, cerium ( -0 .3% of the
melt), or mixed rare earth metals (in the form of misch metal
containing 52% Ce, 25% La, and 12% Nd), have been shown to increase
significantly both the overheating tempera- ture and mechanical
properties of the steel (for example, ductility and toughness).
Pro- vided that a high Ce/S ratio (>2) existed, a complete
change in sulfide morphology oc- curred in low-alloy steels where
the elon- gated MnS inclusion occurring in the un- treated steel
was totally replaced by small globular type-I rare earth sulfides
and ox- ysulfides of high thermal stability even after
austenitizing at 1400 C (2550 F) (Ref 2). This treatment does not
show intergranular faceting. Burning can also be avoided in the
same way by treating with calcium, zirconi- um, cerium, or mixed
rare earth addition to form refractory, less-soluble sulfides.
Control of Cooling Rates. Control of cool- ing rates is not a
practical method for large forgings because extremely slow cooling
is prohibitively time consuming and causes excessive scaling and
decarburization, and rapid quenching from high temperatures
produces cracking and distortion of the parts (Ref 2).
Reclamation of Overheated Steel Severely overheated steels can
often be
completely restored by any of the following heat treatments:
Repeated normalizing (as many as six) starting at temperatures
50 to 100 C (90 to 180 F) higher than usual, followed by a standard
normalizing treatment (Ref 2)
Repeated oil-hardening and tempering treatments after prolonged
soaking at 950 to 1150 C (1740 to 2100 F) in carburizing
atmosphere. Rehardening more than three times is not advisable
Soaking at 900 to 1150 C (1650 to 2100 F) for several hours.
This causes growth of MnS particles by the Ostwald ripening process
and results in an excessive scale formation and a loss of
dimensional accu- racy of the forgings
Residual Stresses
Heat treatment often causes stress- and strain-related problems
such as residual stress, quench cracks, and deformation and/ or
distortion. The residual stress may be defined as the
self-equilibrating internal or locked-in stress remaining within a
body with no applied (external) force, external con- straint, or
temperature gradient (Ref 18, 19). There are two types of residual
stresses:
Macro- or long-range residual s tress is a first-order stress
that represents an aver- age of body stresses over all the phases
in polyphase materials. Macroresidual stresses act over large
regions as com- pared to the grain size of the material.
Traditionally, engineers consider only this type O f residual
stress when design- ing mechanical parts
Microresidual stress, also t e rmed tesse- lated stress or
short-range stress is a second-order or texture stress, which is
associated with lattice defects (such as vacancies, dislocations,
and pile-up of dislocations) and fine precipitates (for ex- ample,
martensite) (Ref 20-22). Microre- sidual is the average stress
across one grain or part of the grain of the material. This
information is indispensable in studying the essential behavior of
materi- al deformation
These two types of residual stresses may also be classified
further as a tensile or compressive stress located near the surface
or in the body of a material. This section focuses on the effects,
development , con- trol, and measurement of long-range resid- ual
stresses.
Effects of Residual Stress The major effects of residual stress
in-
clude dimensional changes and resistance to
-
604 / Process and Quality Control Considerations
Surface residual stress (root of notch), ks -200 -160 -120 -80
-40 0 40
1100 8645 notch cold rolled I 8645 notch warm rolled 160 0.25
notch radius/~- I 0.25 notch radius
~i~ J / 18645 I 1045 j I - " ~ . ~ / I shot peened I untempered
a~-,,,,L// j 1 4 B 3 5
825 I 1045- ~ , , " t e m p e r ~ - - 120 ~. tempered . . . .
.
te ,~ ,e red \~~ , \ temperecl .~ ._E I 8630-N ~ \ .E_ -~ = 550
I I t empered~ 8630 80 N ~
Specimen X ~ / o i l quenched
6.75 ~ ~ ' ~ 1 8660 oil uenched "' i 8645 - - " ~ ~ . / 275 - I
I tempered [ ~ ' ~ - ~ _ ~ 40
L_ 1.750 in. L1.550 in. diam ~ 8645 oil quenched 60 V-notch
1
diam 0.025 root radiu Compression ~--~-Tension 0 i i I I 0
-1375 -1100 -825 -550 -275 0 275 Surface residual stress (root
of notch), MPa
Effect of surface residual stress on the endurance limit of
selected steel. All samples were water Fig 2 quenched except as
shown, and all specimen dimensions are given in inches. Source: Ref
23, 24
crack initiation. Dimensional changes occur when the residual
stress (or a portion of it) in a body is eliminated. In terms of
crack initiation, residual stresses can be either beneficial or
detrimental, depending on whether the stress is tensile or
compressive.
Compressive Residual Stress. Because re- sidual stresses are
algebraically summed with applied stresses, residual compressive
stresses in the surface layers are generally helpful because the
built-in compressive stresses can reduce the effects of imposed
tensile stresses that may produce cracking or failure. Compressive
stresses therefore contribute to the improvement of fatigue
strength and resistance to stress-corrosion cracking in a part and
an increase in the bending strength of brittle ceramics and glass
(Ref 22).
Figure 2 shows that the endurance limit fatigue strength of
selected steels increases
with the surface residual compressive stress developed by
specific heat treatment and surface processing. It is also apparent
that, in the presence of high compressive stress, a poor
microstructure in steel samples has a small influence on good
endurance limit fatigue strength (Ref 23-25). These fatigue
improvements are of great significance in components, particularly
where stress rais- ers, such as notches, keyways, oil holes, and so
forth, are highly desirable in the design of components (for
example, crank- shafts, half-shafts, and so on) (Ref 26). Many
fabrication methods have been devel- oped to exploit this
phenomenon. Pre- stressed parts (including shrink-fits, pre-
stressed concrete, interference fits, bolted parts, coined holes,
wire-wound concrete pipe), mechanical surface working pro- cesses
(such as shot peening, surface roil- ing, lapping, and so on) of
hardened ferrous
Table 3 Summary of compressive and tensile residual stresses at
the surface of the parts created by the common manufacturing
processes Compression at the surface Tension at the surface
Surface working: shot peening, surface rolling, lapping, and so
on
Rod or wire drawing with shallow penetration(a) Rolling with
shallow penetration(a) Swaging with shallow penetration(a) Tube
sinking of the inner surface Coining around holes Plastic bending
of the stretched side Grinding under gentle conditions Hammer
peening Quenching without phase transformation Direct-hardening
steel (not through-hardened) Case-hardening steel Induction and
flame hardening Prestressing Ion exchange
Rod or wire drawing with deep penetration Rolling with deep
penetration Swaging with deep penetration Tube sinking of the outer
surface Plastic bending of the shortened side Grinding: normal
practice and abusive conditions Direct-hardening steel
(through-hardened)(b) Decarburization of steel surface Weldment
(last portion to reach room temperature) Machining: turning,
milling Built-up surface of shaft Electrical discharge machining
Flame cutting
(a) Shallow penetration refers to ~
-
Defects and Distortion in Heat-Treated Parts / 605
Table 4 Changes in volume during the transformation of austenite
into different phases
Change in volume, %, as a function of carbon
Transformation content ( % C)
Spheroidized pearlite - 4 . 6 4 + 2.21 (% C) ---, austenite
Austenite ~ 4.64 - 0.53 x (% C) martensite
Spheroidized pearlite 1.68 x (% C) martensite
Austenite ~ lower 4.64 - 1.43 (% C) bainite
Spheroidized pearlite 0.78 x (% C) lower bainite
Austenite ~ upper 4.64 - 2.21 x (% C) bainite
Spheroidized pearlite 0 upper bainite
Source: Ref 4
the transformation of austenite into mar- tensite or other
transformation products (Ref 27). Table 4 lists the changes in vol-
ume during the transformation of austenite into different
structural constituents (Ref 28).
Thermal Contraction. The relation be- tween the thermal stress
~th during cooling and the corresponding temperature gradient in
the component is given by:
tYth = E - AT" ct (Eq
where E is the modulus of elasticity, and tx is the thermal
coefficient of expansion of the material. It is thus apparent that
thermal stresses are greatest for materials with high elastic
modulus and coefficient of thermal expansion. Temperature gradient
is also a function of thermal conductivity. Hence, it is quite
unlikely to develop high-tempera-
1000 ? c w
~ 500 ~. u
~- 0 1
Water quenched 100 mm (4 in.)
specimen
1700 ou-
1100 ~
600 E
100 ~- 10 103 Time, s
_
e~ e~
E E o o
Deve lopmen t of thermal and residual stresses F i g 3 in the
longitudinal direct ion in a 100 mm (4 in.) d iameter steel bar on
wate r q u e n c h i n g f rom the aus- tenitizing t empe ra tu r e
, 850 C (1560 F). Transforma- tion stresses are not taken into cons
idera t ion . Source: Ref 30
Table 5 Relevant physical properties in the development of
thermal stresses Coefficient of
Modulus of elasticity expansion Thermal conductivity
Metal GPa psi x 10 6 1 0 - 6 / K 1 0 - 6 p F W m -1 k - l Btu
in./ft 2 h F
Pure iron (ferrite) 206 30 12 7 80 555 Typical austenitic steel
200 29 18 10 15 100 Aluminum 71 10 23 13 201 1400 Copper 117 17 17
9 385 2670 Titanium 125 18 9 5 23 160
Source: Ref 29
ture gradients in good thermal conductors (for example, copper
and aluminum), but it is much more likely in steel and titanium
(Ref 29). Another term involving thermal conductivity, called
thermal diffusivity (Dth), is sometimes used in context with
temperature gradient. It is defined a s D t h = k/pc, where k is
the thermal conductivity, p is the density, and c is the specific
heat. It is clear that low Oth (or k) promotes large temperature
gradient or thermal contrac- tion. It should be emphasized that
large size of the part and high heating or cooling rates (severity)
of quenching medium also aug- ment temperature gradients leading to
large thermal contraction.
Table 5 lists some of the relevant material properties that
affect thermal and residual stresses (Ref 29).
Residual Stress Pattern Due to Thermal 1)Contraction. Residual
stress is developed
during quenching of a hot solid part that involves thermal
volume changes without solid-state phase transformation. This situ-
ation also exists when a steel part is cooled from a tempering
temperature below the A t. Figure 3 shows the development of
longitu- dinal thermal and residual stresses in a 100 mm (4 in.)
diam steel bar on water quench- ing from the austenitizing
temperature, 850 C (1560 F) (Ref 30). At the start of cooling, the
surface temperature S falls drastically as compared to the center
temperature C (top left sketch of Fig 3). At time w, the temper-
ature difference between the surface and core is at a maximum of
about 550 C (1020 F), corresponding to a thermal stress of 1200 MPa
(80 tons/in. E) due to linear differ- ential contraction of about
0.6%, if relax- ation does not take place. Under these conditions,
tensile stresses are developed in the case with a maximum value of
a (lower diagram), corresponding to time w in the upper diagram,
and the core will contract, producing compressive stresses with a
max- imum of b. The combined effect of tensile and compressive
stresses on the surface and core, respectively, will result in
residual stresses as indicated by curve C, where a complete
neutralization of stress will occur at some lower temperature u.
Further de- crease in temperature, therefore, produces
longitudinal, compressive residual stresses at the surface and the
tensile stresses at the core, as shown in the lower right-hand
diagram of Fig 3. Figure 4(a) is a schematic
illustration of the distribution of residual stress over the
diameter of a quenched bar due solely to thermal contraction in the
longitudinal, tangential, and radial direc- tions (Ref 19).
The maximum residual stress attained on quenching increases as
the quenching tem- perature and quenching power of the cool- ant
are increased. Tempered glass is made by utilizing quenching
techniques in which glass is heated uniformly to the annealing
temperature and then surface cooled rapidly by cold air blasts.
This produces compres- sive surface stresses to counteract any ten-
sile bending stress, if developed during loading of the glass,
thereby increasing its load-carrying capacity (Ref 31).
Residual Stress Pattern Due to Thermal and Transformational
Volume Changes (Ref 32). During quench hardening of a steel (or
other hardenable alloy) part, hard martens- ite forms at the
surface layers, associated with the volume expansion, whereas the
remainder of the part is still hot and ductile austenite. Later,
the remainder austenite transforms to martensite, but its
volumetric expansion is restricted by the hardened surface layer.
This restraint causes the cen- tral portion to be under compression
with the outer surface under tension. Figure 4(c) illustrates the
residual stress distribution over the diameter of a quenched bar
show- ing volume expansion associated with phase transformation in
the longitudinal, tangen- tial, and radial directions (Ref 19). At
the same time during the final cooling of the interior, its
contraction is hindered by the hardened surface layers. This
restraint in contraction produces tensile stresses in the interior
and compressive stresses at the outer surface. However, the
situation as shown in Fig 4(c) prevails, provided that the net
volumetric expansion in the interior, after the surface has
hardened, is larger than the remaining thermal contraction. In some
particular conditions, these volumet- ric changes can produce
sufficiently large residual stresses that can cause plastic de-
formation on cooling, leading to warping or distortion of the steel
part. While plastic deformation appears to reduce the severity of
quenching stresses, in most severe quenching the quenching stresses
are so high that they do not get sufficiently re- leased by plastic
deformation. Consequent- ly, the large residual stress remaining
may
-
606 / Process and Quality Control Considerations
+~
Longitudinal
I Tangential
I
I Radial
(a) (b)
.~_ T j
L = l ong i t ud ina l T = tangen t ia l R = radial
I
I i
Longitudinal
I I
i
I Tang?ntial
I
t-'--->' Rad al
(el
Schematic illustration of the distribution of residual stress
over the diameter of a quenched bar in the F ig 4 longitudinal,
tangential, and radial directions due to (a) thermal contraction
and (c) both thermal and transformational volume changes. (b)
Schematic illustration of orientation of directions. Source: Ref
19
reach or even exceed fracture stress of steel. This localized
rupture or fracture is called quench cracking (Ref 32, 33).
It should be emphasized again that for a given grade of steel,
both large size of the part and higher quenching speed contribute
to the larger value of thermal contraction, as compared to the
volumetric expansion, of martensite. In contrast , when the parts
are thin and the quenching rate is not high, thermal contraction of
the part subsequent to the hardening of the surface will be smaller
than the volumetric expansion of martensite. Similarly, for a given
quenching rate, the temperature gradients decrease with decreasing
section thickness, and con- sequently the thermal component of the
residual stress is also decreased (Ref 24).
Figure 5(a) shows the continuous cooling transformation diagram
of DIN 22CrMo44 low-alloy steel exhibiting austenitic decom-
position with the superimposed cooling curves of the surface and
center in round bars of varying dimensions. If the large- diameter
(100 mm, or 4 in.) bar is water quenched (that is, for slack
quenching), martensitic transformation occurs at the surface, and
pearlitic + bainitic transforma- tions occur at the center,
resulting in a residual stress pattern (top of Fig 5) similar to
that due solely to thermal stress (Fig 4a).
During the rapid quenching of the medium- size (30 mm, or 1.2
in.) bar diameter, the start of bainite transformation at the
center coincides approximately with the transfor- mation of
martensite on the surface. This results in compressive stresses at
both the surface and center, with tensile stresses in the
intermediate region (middle of Fig 5). When the smaller-diameter
(10 mm, or 0.4 in.) bar is drastically quenched (for exam- ple, in
brine), the entire bar transforms to martensite. This is associated
with very little temperature variation between the sur- face and
the center of the part. In this situation, tensile residual stress
is devel- oped at the surface and compressive stress at the center
of the bar (bottom, Fig 5) (Ref 34, 35).
Although the shallower hardening steels exhibit higher surface
compressive stresses, deep hardening steels may develop moder-
ately high surface compressive stresses with severe water
quenching. When these deep hardening steels are through-hardened in
a less efficient quenchant, they may exhibit surface tensile
stresses (Ref 24, 31). Rose has pointed out the importance of
transformations of core and surface before and after the stress
reversal. According to him the tensile surface residual stress oc-
curs when the core transforms after, and the
surface transforms before, the stress rever- sal (Fig 4c and bot
tom of Fig 5), whereas compressive surface residual stress takes
place when the core transforms before, and the surface transforms
after, the stress re- versal (top of Fig 5). His analysis is
capable of explaining complex stress patterns for various
combinations of part sizes, quench- ing rate, and steel
hardenability (Ref 21). However , the residual stress pattern in
the hardened steels can be modified either with different
transformation characteristics or during the tempering and
finish-machining (after hardening) operations.
Residual Stress Pattern after Surface Hard- ening. In general,
thermochemical and ther- mal surface-hardening treatments produce
beneficial compressive residual stresses at the surface.
Carburized and Quenched Steels. When low-carbon steels are
carburiZed and quenched, first the core transforms at high
temperature (600 to 700 C, or l l00 to 1300 F) to ferrite and
pearlite with the attendant relaxation of any transformation
stresses. Later , the high-carbon case transforms to martensite at
much lower temperature (less than 300 C, or 570 F), accompanied by
volume expansion and under conditions of no (or minimum) stress
relaxation. As a result, residual compressive stress is devel- oped
in the case with a maximum at the surface.
Large differences in carbon level between the case and the core
determine the se- quence of phase transformation on cooling after
carburizing and the resultant develop- ment of compressive residual
stress in the case. Likewise, compressive residual stress in the
case increases as the core carbon content decreases. Increasing
case depth reduces the contribution from the low-car- bon core in
the development of compressive stress in the case, thereby
adversely affect- ing the fatigue propert ies (Ref 36).
In actual practice, a maximum compres- sive stress develops at
some distance away from the surface (Fig 6 and 7). This effect
occurs because of the presence of retained austenite, the extent of
which depends on steel composition, carbon content of the case,
quenching temperature, and severity of quench. According to
Koistinen (Ref 38) and Salonen (Ref 39) the peak compressive stress
takes place at 50 to 60% of the total case depth corresponding to
about 0.5 to 0.6% carbon level, which produces a low retained
austenite content and martensite hardness around the maximum.
Another factor that might influence this compressive residual
stress profile is that the martensite formed in the lower-carbon
regions of the case is of the lath type, which also affects the
retained austenite content (Ref 20). The reversal sign of residual
stress takes place at or near the case/core interface. Later, when
Kois t inen 's theory was applied to the mea- sured data, it
appeared that the position of
-
Defects and Distortion in Heat-Treated Parts / 607
1000 1830
800 , ~ 1470
600 ~ ~ 1110
400 \ ~ . _ 750
200 " ~ ~ 390 Surface Center
0 1000 1830
800 ~ ~ 1470
400 - 750 X E E
200 390 er
0 Surfacel
1000 1830
800 ( 1470
600 ~ . . _ . . . . 1110 /
400 " ' ~-~,........._. 750
nter 0
1 10 100 103 Time, s
(a) (b)
+20
-20
Distribution of residual stresses
+3
-3
m -40 -6
~ Center Surface ~ ~ +20 +3 "~
"o "o .- "~ o o
30 mm -20 diam -3
Center Surface +20 [ +3
0 0
I ~ / 0 mm -20 ~ diam -3
Center Surface + = Tensile stresses - = Compressive stresses
(a) Continuous cooling transformation diagrams of DIN 22CrMo44
steel showing austenitic decompo- Fig 5 sition with the
superimposed cooling curves of the surface and center during water
quenching of round bars of varying dimensions. (b) The
corresponding residual stress pattern developed because of thermal
and transformational volume changes. Source: Ref 34, 35
maximum compressive stress depends on severity of quenching,
total case depth, steel hardenability, and so forth (Ref 21, 40).
Figure 7 shows the details of generation of axial stress
distribution of a carburized gear (made from deeper hardening
steel) during quenching. In the early stages, the contour lines of
equal stress were largely unaffected by the surface profile. Later
a zone of high compressive stress distribution occurred in the
central portion of the teeth, which remained until the end of the
quench (Ref 37).
In nitriding, like carburizing, a compres- sive residual stress
is set up in the surface layers. High-temperature nitriding
produces a little relaxation of stresses, whereas low- temperature
nitriding imparts a maximum residual stress. In nitrocarburizing,
im- provement in residual surface compressive stress and fatigue
strength depends on the hardness and depth of diffusion zone. These
properties, in turn, decrease with increasing carbon and alloy
content (that is, increased hardenability). During quenching, after
ni-
trocarburizing, a (macro-) compressive re- sidual stress is
produced in the compound layer and gamma prime phase (Ref 41). When
nitrocarburized parts are rapidly quenched, the above properties
are further enhanced (Ref 42).
In borided steel processed at 900 C (1650 F), a high compressive
residual stress is developed at the surface layers (Fig 8), which
consists of FeB and Fe2B phases (Ref 43); this is attributed to the
lower thermal expansion coefficient and the larger specific volume
in a borided layer compared to that in a ferrite matrix (Ref 18,
43).
In an induction-hardened steel part, a compressive surface
residual stress is pro- duced when wear-resistant hard martensite
(with slightly lower density) is formed on the surface of a section
concurrently with volume expansion while nonhardened core remains
essentially unchanged (Fig 9) (Ref 44, 45). The magnitude of the
compressive stress, which is affected by both thermal contraction
and martensite formation, may be a considerable fraction of the
yield
1.0
c" o 0.5
8
-o , c.-~_
~ o
m
.9_ tt-
Tensile
Compressive
Distance from the surface
Relationship between carbon content, re- Fig 6 tained austenite,
and residual stress pattern. It shows the development of peak
compressive stress some distance away from the surface. Source: Ref
20
strength, which permits the application of significantly higher
stresses than could nor- mally be possible in fatigue loading. As
in the carburizing practice, the surface com- pressive residual
stresses are usually found to increase, with depth below the
surface (Ref 45) (Fig 9, Ref 44). A fairly sharp transition to a
tensile state takes place near the hardness drop-off between the
case and unhardened surrounding material. With an increase in
distance from the steep transi- tion, the tensile condition
gradually fades away toward zero stress (Ref 44). In induc- tion
hardening, an increase in hardenability changes the depth at which
transition from compressive to tensile stress occurs. The increase
in the rate of heating produces an increase in the maximum
compressive and tensile residual stresses without affecting the
mode of stress distribution (Ref 46).
Residual Stress in Other Processing Steps. As welding
progresses, the temperature dis- tribution in the weldment becomes
nonuni- form and varying as a result of localized heating of the
weldment by the welding heat source. During the welding cycle,
compris- ing heating and cooling, complex strains develop in the
weld metal and adjacent areas. As a result, appreciable residual
stresses remain after the completion of welding. Since the weld
metal and heat- affected zone contract on cooling (Fig 10a), they
are restrained by the cool adjacent part. This produces tensile
residual stress in
-
608 / Process and Quality Control Considerations
- 900
Gz = 200MPa -600
100 ~ -300
( ~100
t= 3 s 30 s
300
3 O O
Carburized SNC815
0300 -600 60~ j
l i 300 60 s
-900 -600
0 500
Distance from surface, in. 0.002 0.004 0.006
0 0 z~ ~ 0 #_
- 5 0 0
-1000 ~ ~ o '~ -1500
300 -2000 0 0 0
-2500 600 0 0.05
Fig 7 Axial stress distribution (given in MPa) in carburized
gear during quenching process. Source: Ref 37
the weldment region and compressive resid- ual stress in the
surrounding base metal region (Fig 10b).
In general, a steep residual stress gradient is developed
because of the steep tendency of the thermal gradient. This may, in
turn, lead to hot cracking (between columnar grains) or severe
center line cracking in the weld area (Ref 48). Catastrophic
failures of welded bridges and all-welded ships are mostly
attributed to the existence of large and dangerous tensile residual
stress in them (Ref 49).
The grinding step in manufacturing is important, since it is
always utilized to produce the finished surface. It has been shown
that gentle surface grinding, using a soft sharp wheel and slow
downfeed, pro- duces compressive residual stress at the surface,
whereas conventional (normal practice) and abrasive grinding result
in surface tensile stresses of very high magni- tude (Fig l l) (Ref
22, 50). However, the
400 (58)
Distance from surface, in. 0.08 0.16 0.24 0.32 0.40 500 gm Knoop
test 3 mm case I
I I Is ss
/ Hardness
2 4 6 8 10 Distance from surface, mm
8O
o r~ 60 -r-
40 ~
20 ,~
u,l
0 12
A
200 (29) #_
0
-200 (-29) tr
gentle grinding method is expensive from the viewpoint of
operating time and wear of the wheel.
As a result of temperature gradient during cooling, castings
develop compressive stress- es at the surface and tensile stresses
in the interior (Ref 22). However, transient temper- ature gradient
and phase transformation oc- curring during the early stages of
solidifica- tion and cooling of continuous steel castings in the
mold may give rise to the development of harmful residual stresses
leading to the formation of cracks (Ref 51).
Chemical processes such as electroplat- ing, scale formation,
and corrosion of met- als can produce residual stresses due to
coherency strains arising from the matching tendency of crystal
structures of the outer surface product with the crystal structure
of the adjacent layer (Ref 22). Residual stress- es are also
introduced when heat-treated parts are subjected to successive
heating and cooling cycles during service condi- tions.
Residual Stress in the Heat-Treated Non- ferrous Alloys. In
nonferrous alloys, notably age-hardenable aluminum alloys, copper-
beryllium alloys, certain nickel-base super-
(a)
Residual stress Compression Tension
i i i i i i i i i i I i i ~
Yl
(b)
Fig 10 (a) The transverse shrinkage occurring in butt weldments.
(b) Longitudinal residual
stress patterns in the weldment and surrounding re- gions. This
also shows longitudinal shrinkage in a butt weld. Source: Ref
47
-400 (-58)
F ig 9 A typical hardriess and residual stress profile in
induction-hardened (to 3 ram, or 0.12 in.,
case depth) and tempered (at 260 C, or 500 F) 1045 steel.
Source: Ref 44
-- 50 /xA z, /x A 0
- -50
- -100
-150 ~
--200 "~ (b
FeB - -250 rr Fe2B /x Ferrite - -300
0.10 0.15
Distance from surface, mm
Residual stress distribution of FeB and Fe2B Fig 8 layers in
borided steel processed at 900 C (1650 F). Source: Ref 18, 43
alloys, and so on, a significant amount of thermal stress is
generated during quench- ing prior to precipitation hardening. The
quenching process in this condition does not invariably involve a
phase change; rath- er, this is confined to the postquenching aging
treatment. In other nonferrous alloys such as uranium and titanium
alloys, the final structural condition is not obtained by a slow
cool.
When high-strength titanium alloy is quenched from a solution
annealing temper- ature of 850 to 1000 C (1560 to 1830 F), it
develops large residual stress caused by poor thermal conductivity
of titanium lead- ing to high-temperature gradient. This prob- lem
can, however, be avoided by stress- relief annealing at 650 to 700
C (1200 to 1290 F), which produces a slight reduction in mechanical
properties. When a high- strength aluminum age-hardening alloy is
rapidly quenched from the solution temper-
Depth below surface, mil 3.15 6.3 9.45 12.6
800 120
600 A
g_g ~: 400
"~ 200
r,~ o
~. -200 E o
-400 0
/\.
\ \
- 9O
- 60 -- Abrasive
% 30
Conventional
0 -%
- Gentle ,-, . . . . - -30
-60 80 160 240 320
Depth below surface, pm
L~
Residual stress distribution after gentle, con- F i g 11
ventional, and abrasive grinding of hard- ened 4340 steel. Source:
Ref 22
-
Defects and Distortion in Heat-Treated Parts / 609
Table 6 A compiled summary of the maximum residual stresses in
surface heat-treated steels
Residual stress (longitudinal)
Steel Heat treatment MPa ksi
832M13 (type) Carburized at 970 C (1780 F) to 1 mm (0.04 in.)
case with 0.8% surface carbon
Direct-quenched 280 Direct-quenched, - 80 C ( - 110 F) subzero
treatment 340 Direct-quenched, - 90 C ( -130 F) subzero treatment,
200
tempered 805A20 Carburized and quenched 240-340(a) 805A20
Carburized to 1.1-1.5 mm (0.043-0.06 in.) case at 920 C 190-230
(1690 F), direct oil quench, no temper 805A ! 7 400 805A17
Carburized to 1.1-1.5 mm (0.043-0.06 in.) case at 920 C 150-200
(1690 F), direct oil quench, tempered 150 C (300 F) 897M39
Nitrided to case depth of about 0.5 mm (0.02 in.) 400--600 905M39
800-1000 Cold-rolled steel Induction hardened, untempered 1000
Induction hardened, tempered 200 C (390 F) 650 Induction
hardened, tempered 300 C (570 F) 350 Induction hardened, tempered
400 C (750 F) 170
(a) Immediately subsurface, that is. 0.05 mm (0.002 in.).
Source: Ref 29
40.5 49.0 29.0
35.0--49.0 27.5-33.5
58 22-29
58.0-87.0 116.0-145.0
145.0 94.0 51 24.5
ature, high thermal and residual stresses are induced due to
high coefficient of expansion of aluminum. Uphill quenching from
liquid nitrogen temperature ( - 196 C, or - 320 F) in a steam blast
alleviates this problem. This induces stresses opposite in sign to
those developed on water quenching from the solutionizing and
cancels out their effect. This is followed by aging of the alloy in
the conventional manner (Ref 29).
Fast polyalkylene glycol (PAG) quench- ing of solution-treated
aluminum alloys tends to reduce residual stress levels be- cause of
its more uniform heat extraction rate (thermal shock is smaller,
and thereby machining is less likely to produce further
distortion), thereby helping solve major and long-standing
distortion problems among aluminum workpieces (Ref 52).
Control of Residual Stresses in Heat-Treated Parts
Table 6 lists some typical values of max- imum residual stresses
developed in the surface-hardened steels that have been re- ported
in the literature (Ref 29). It is worth noting that there is a
marked influence of tempering on the residual stress level. Tem-
pering must be accomplished at about 150 C (300 F) to maintain 50
to 60% retention of the residual stress level obtained after
quenching because a higher tempering tem- perature greatly reduces
surface compres- sive stresses. However, a higher stress- relief
temperature ( -600 C, or 1110 F) is used for mechanically deformed
compo- nents (for example, hot-rolled bars) or com- ponents with
tensile surface residual stress- es. Alternatively, serious
residual tensile stresses may be avoided effectively by gen- tle
grinding of the surface.
Measurement of Residual Stresses There are two methods of
measuring re-
sidual stresses: the destructive method, also
called the dissection method, and the non- destructive methods
comprising mainly x-ray diffraction, neutron diffraction, ultra-
sonic, and magnetic methods.
Destructive (or Dissection) Method. This method is old but
reasonably accurate, practically nondestructive, uses well-estab-
lished methods, and can be employed in confined situations at site
(Ref 53). Howev- er, it is tedious, time consuming, and expen- sive
(Ref 54). The other drawbacks are the destructive, or at best
semidestructive na- ture of the method, and its ability to mea-
sure only the macroresidual stresses. The hole-drilling method is
used extensively for measuring residual stresses, which depends on
the dissection approach. It consists of the mounting of strain
gages or a three- element strain-gage rosette on the surface and
measurement of strains. Then a rigidly guided milling cutter is
used to drill a small, straight, circular, perpendicular, and fiat-
bottomed hole not exceeding 3.2 mm (0.125 in.) at the center of the
rosette and into the surface of the component being analyzed.
Strain redistribution occurring at the sur- face in the surrounding
area of the hole (resulting from the residual stress relief) is
then measured with the previously installed strain gages. The
residual stress is calculat- ed at a large number of points in a
surface from the strain measurements using the well-established
method (Ref 22, 28). To minimize the introduction of spurious
strains by the grinding operation, the rate of metal removal should
be less than 3.125 x 10 -4 m/s (1.23 10 -2 in./s), and readings are
recorded after 15 min of the end of the grinding process to ensure
that any heat generated has been dissipated (Ref 55).
Nondestructive Methods. The main diffi- culty with the
nondestructive methods is that measurements of crystallographic
lattice pa- rameters, ultrasonic velocities, or magnetiza- tion
changes are made that are indirectly
related to the residual stress. The above quantities are usually
dependent on the stress and material parameters (such as
metallurgi- cal textures), which are difficult to quantify (Ref 54,
56).
The x-ray diffraction method is the well- established technique
for measuring both macro- and microresidual stress nondestruc-
tively. In most instances, the x-ray diffraction method has been
employed to provide quan- titative values for residual stress
profiles in surface or fully hardened components (Ref 57). This
technique depends on the determi- nation of lattice strains and the
stress-induced differences in the lattice spacing. Macroresid- ual
strain is measured from the shift of dif- fraction lines in the
peak position using the so-called nonlinear SinZC method from which
residual stress is calculated (Ref 57). For the measurement of
microstrain the Voigt single- line method is applied (Ref 58).
Precision in lattice strain measurement of the order of 0.2% is
possible.
Portable x-ray diffraction equipment is now commercially
available in various forms that allow stress measurement to be made
very quickly (ranging from 4 to 30 s). The main drawbacks are that
it cannot be applied to noncrystalline materials such as plastics,
and it is only capable of measuring residual stresses of materials
very close to the surface under examination. That is, the
measurement is purely surface related (a depth of 0.01 mm, or 0.4
mil, is commonly quoted) (Ref 59).
Neutron radiography or diffraction, used for polycrystalline
materials, has a much deeper penetration than x-rays, but has major
safety problems and the disadvantage of being nonportable.
Ultrasonic method for evaluating residual stress involves
ultrasonic stress birefrin- gence or sonoelasticity; this depends
upon the linear variation of the velocities of sound in a body
(that is, ultrasonic waves) with the stress. This method has the
poten- tial for greater capability, versatility, and usefulness in
the future (Ref 53, 56). How- ever, this has the disadvantage, in
common with the magnetic methods, that it requires transducers
shaped to match the surface being inspected (Ref 60).
The magnetic method is based on the stress dependence of the
Barkhausen noise amplitude. Each time an alternating mag- netic
field induced in a ferromagnetic mate- rial is reversed, it
generates a burst of Barkhausen noise. The peak amplitude of the
burst, as determined with an inductive coil near the surface of the
component material, varies with the surface stress lev- el. Since
Barkhausen noise depends on composition, texture, and work
hardening, it is necessary in each application to use calibrated
standard (reference) samples with the same processing history and
com- position as the component being analyzed. This method is used
to measure residual
-
610 / Process and Quality Control Considerations
stresses well below the yield strength of the ferromagnetic
materials. This method is rapid, and the measurements are made with
the commercially available portable equip- ment. However, this
method is limited to only ferromagnetic materials (Ref 56).
Thermal evaluation for residual stress anal- ysis (TERSA) is a
new nondestructive meth- od that is in an experimental stage. It
has the advantage that it is completely independent, remote, and
noncontacting. It consists of merely directing a controlled amount
of ener- gy from a laser energy source into the volume of the
material being inspected and then mak- ing a precise determination
of changes in the resulting temperature rise by infrared radiom-
etry. However, the working instrument will also require some form
of display to enable visual examination to be made of any high-
stressed regions (Ref 60).
Quench Cracking Anything that produces excessive
quenching stress is the basic cause of crack- ing. Quench
cracking is mostly intergranu- lar, and its formation may be
related to some of the same factors that cause inter- granular
fracture in overheated and burned steels. The main reasons for
cracking in heat treatment are: part design, steel grades, part
defects, heat-treating practice, and tempering practice (Ref
61).
Part Design. Features such as sharp cor- ners, the number,
location, and size of holes, deep keyways, splines, and abrupt
changes in section thickness within a part (that is, badly
unbalanced section) enhance the crack for- mation because while the
one (thin) area is cooling quickly in the quenchant, the other
(thick) area immediately adjacent to it is cool- ing very slowly.
One solution to this problem is to change the material so that a
less drastic quenchant (for example, oil) can be em- ployed. An
alternate solution is to prequench, that is, to cool it prior to
the rest of the part. This will produce an interior of the hole or
keyway that is residually stressed in compres- sion, which is
always desirable for better fatigue properties (Ref 61). The third
solution is a design change, and the fourth is to use a milder
quenchant.
Steel Grades. Sometimes this can be checked by means of a spark
test, whereas at other times a chemical analysis must be made. In
general, the carbon content of steel should not exceed the required
level; other- wise, the risk of cracking will increase. The
suggested average carbon contents for water, brine, and caustic
quenching are given below:
Method Shape Carbon, %
Induction hardening Complex 0.33 Simple 0.50
Furnace hardening Complex 0.30 Simple 0.35 Very simple, such
as bar 0.40
A decrease in carbon content from 0.72 to 0.61% has been shown
to slightly increase the thermal crack resistance of rim- quenched
railroad wheels (Ref 62).
Because of segregation of carbon and alloying elements, some
steels are more prone than others to quench cracking. Among these
steels, 4140H, 4145H, 4150H, and 1345H appear to be the worst. A
good option is to replace the 4100 series with the 8600 series. An
additional disadvantage with the use of 1345H steel is the manga-
nese floating effect, which leads to very high manganese content in
the steel rolled from the last ingot in the same heat. Simi- larly,
dirty steels (that is, steels with more than 0.05% S, for example,
AISI 1141 and 1144) are more susceptible to cracking than the
low-sulfur grades. The reasons for this are that they are more
segregated in alloying elements, the surface of this hot-rolled
high- sulfur steel has a greater tendency to form seams, which act
as stress raisers during quenching, and they are usually coarse
grained (for better machinability), which increases brittleness and
therefore pro- motes cracking. If these high-sulfur grades are
replaced by calcium-treated steels or cold-finished leaded steels,
this problem can be obviated (Ref 61).
Part Defects. Surface defect or weakness in the material may
also cause cracking, for example, deep surface seams or nonmetallic
stringers in both hot-rolled and cold-fin- ished bars. Other
defects are inclusions, stamp marks, and so forth. For large-seam
depths, it is advisable to use turned bars or even magnetic
particle inspection. The forg- ing defects in small forgings, such
as seams, laps, flash line, or shearing crack, as well as in heavy
forgings, such as hydrogen flakes and internal ruptures, aggravate
cracking. Similarly, some casting defects, for exam- ple, in
water-cooled castings, promote cracking (Ref 50).
Heat-Treating Practice. Higher austenitiz- ing temperatures
increase the tendency toward quench cracking. Similarly, steels
with coarser grain size are more prone to cracks than fine-grain
steels because the latter possess more grain-boundary area to stop
the movements of cracks, and grain boundaries help to absorb and
redistribute residual stresses. An outstanding contribu- tor to
severe cracking is improper heat- treating practice, for example,
nonuniform heating and nonuniform cooling of the com- ponent
involved in the heat-treatment cy- cle. It is a good heat-treating
practice to anneal alloy steels prior to the hardening treatment
(or any other high-temperature treatment, for example, forging,
welding, and so forth) because this produces grain- refined
microstructure and relieves stresses (Ref 63).
Water-Hardening Steel. The water-hard- ening steels are most
susceptible to cracks if they are not handled properly. Soft
spots
( Typical appearance of thumbnail check as
Fig 1 2 soft spot on chipping chisel. Source: Ref 64
are most likely to occur in the water-hard- ening steels,
especially where the tool is grabbed with tongs for quenching.
Normal- ly the cleaned surface shows adequate hard- ening and the
scaled surface insufficient hardening, which can be examined with a
file. Soft spots may occur from the use of fresh water, or water
contaminated with oil or soap. Most large tools emerging from
hardening operations contain some soft spots. However, accidental
soft spots in the wrong place should be investigated, and steps
must be taken to eliminate them.
Figure 12 shows the typical appearance of a thumbnail check as
soft spot on chipping chisels, which occurs on the bit near the
cutting edge. The cracks enclosing the soft spots should be avoided
by switching to brine quench (Ref 64).
Air-Hardening Steel. Similarly, when air hardening steels are
improperly handled, they are likely to crack. For example,
avoidance of tempering treatment or use of oil quenching in
air-hardening steel can lead to cracking. However, the common
practice in the treatment of air-hardening steels is initially to
quench in oil until "black" (about 540 C, or 1000 F), followed by
air cooling to 65 C (150 F) prior to tempering. As compared to air
cooling right from the quenching temperature, this practice is to-
tally safe and minimizes the formation of scale.
Polymer quenchants have found well-es- tablished use in the
quenching of solution- treated aluminum alloys, hardening of plain
carbon steels with less than 0.6% C, spring steels, boron steels,
hardenable stainless steels, and all carburizing and alloy steels
with section thickness greater than about 50 mm (2 in.),
through-hardening and carbur- izing steel parts, and induction and
flame- hardening treatments because of their nu- merous beneficial
effects, including elimination of soft spots, distortion, and
cracking problems associated with trace
-
Defects and Distortion in Heat-Treated Parts / 611
Fig 13 Microcracking in a Ni-Cr steel. Source: Ref 67
water contamination in quenching oils (Ref 65).
Agitation is an important parameter in polymer quenching
applications both to en- sure a uniform polymer film around the
quench part and to provide a uniform heat extraction from the hot
part to the adjacent area of quenchant by preventing a buildup of
heat in the quench region.
Salt bath cooling of induction-hardened complex-shaped cast iron
parts reduces danger of cracking, which is usually expe- rienced
when air cooling followed by hot- water quenching is used (Ref
66).
Decarburized Steel. Decarburization usu- ally arises from
insufficient protection as a result of plant failure (for example,
defec- tive furnace or container seals, defective valves), poor
process control (for example, insufficient atmosphere-monitoring
equip- ment, poor supervision), or the existence of decarburizing
agents in the furnace atmo- sphere (for example, CO2, water vapor,
and Hz in the Endogas (Ref 61, 67).
A partially decarburized surface on the part occurring during
tool hardening also contributes to cracking because martensite
transformation is completed therein well before the formation of
martensite in the core. Decarburized surface on the tools has
reduced hardness, which will lead to prema- ture wear and scuffing.
Partial decarburiza-
4.7 ~m
tion must be avoided, especially on all deep- hardening steels,
either by providing some type of protective atmosphere during the
heating operation, stock removal by grind- ing, or carbon
restoration process. In addi- tion to protective atmosphere, salt
baths, inert packs, or vacuum furnaces may be used to obtain the
desired surface chemistry on the tools or dies. The fact that the
better and more consistent performance of the tools is observed
after regrinding reveals the existence of partial decarburization
re- maining.
Carburized Alloy Steel. Two types of peculiar cracking phenomena
prevail in the carburized and hardened case of the car- burized
alloy steels: microcracking and tip cracking. Microcracking of
quenched steels are small cracks appearing across or alongside
martensite plate (Fig 13) (Ref 67) and the prior austenite grain
boundaries (Ref 68). They form mostly on those quenched steel parts
that contain chromi- um and/or molybdenum as the major alloy- ing
elements with or without nickel con- tent and where the hardening
is done by direct quenching.
Microcracks are observed mostly in coarse-grained structures,
such as large martensite plates. This is presumably be- cause of
more impingements of the larger plates of martensite by other large
plates.
Another cause of microcracking is the in- creased carbon content
of martensite (that is, increased hardenability), which is a func-
tion of austenitizing temperature and/or time (Ref 67). This
finding was established for 8620H steel, which has a higher austen-
itizing temperature prior to quenching where there is a greater
tendency to micro- crack (Ref 69). This problem can be avoided by
selecting a steel with less hardenability (that is, with less
austenitizing tempera- ture). Another solution is to change the
heat-treating cycle to carburizing, slow cooling to black
temperature, reheating to, for example, 815 or 845 C (1500 or 1550
F), and quenching (Ref 61). Microcracking in case-hardened surfaces
may be aggravated by the existence of hydrogen, which tends to
absorb during carburizing. However, this hydrogen-enhanced
microcracking can be eliminated by tempering the carburized parts
at 150 C (300 F) immediately after quenching. Tempering exhibits an
addition- al beneficial effect in that it has the ability to heal
the microcracks due to the volume changes and associated plastic
flow that develop during the first stage of tempering (Ref 70). No
adverse report on the influence of microcracks on the mechanical
proper- ties has been noted; however, the control- ling factors
should be varied so as to keep the incidence of microcracks to a
minimum (Ref 67).
Tip cracking refers to the cracking that appears in the teeth of
carburized and quenched gears and runs partly or fully to the ends
of the teeth in a direction parallel to the axis of the part. Many
heat treaters have solved this problem to a great extent by
decreasing the carbon content and case depth to the minimum
acceptable design level or by copper plating the outer diame- ter
of the gear blank prior to hobbing (Ref 66).
Nitrided Steels. The nitrided cases are very brittle.
Consequently, cracking may occur in service prior to realizing any
im- proved wear and galling resistance. This can be avoided by a
proper tool design, for example, incorporating all section .changes
with a minimum radius of 3 mm (0.125 in.).
Tempering Practice. The longer the time the steel is kept at a
temperature between room temperature and 100 C (212 F) after the
complete transformation of martensite in the core, the more likely
the occurrence of quench cracking. This arises from the volumetric
expansion caused by isothermal transformation of retained austenite
into martensite.
There are two tempering practices that lead to cracking
problems: tempering too soon after quenching, that is, before the
steel parts have transformed to martensite in hardening, and skin
tempering, usually observed in heavy sections (=>50 mm, or 2
in., thick in plates and >75 mm, or 3 in., in diameter in round
bars).
-
612 / Process and Quality Control Considerations
It is the normal practice to temper imme- diately after the
quenching operations. In this case, some restraint must be
exercised, especially for large sections (>75 mm, or 3 in.) in
deep-hardening alloy steels. The rea- son is that the core has not
yet completed its transformation to martensite with the ex-
pansion, whereas the surface and/or projec- tions, such as flanges,
begin to temper with shrinkage. This simultaneous volume change
produces radial cracks. This prob- lem can become severe if rapid
heating practice (for example, induction, flame, lead, or molten
salt bath) is used for tem- pering. Therefore, very large and very
intri- cate tool steel parts should be removed from the quenching
medium, and tempering should be started while they are slightly
warm to hold comfortably in the bare hands ( -60 C, or 140 F).
Skin tempering occurs in heavy section parts when the final
hardness is >360 HB. This is due to insufficient tempering time
and is usually determined when the surface hardness falls by 5 or
more HRC points from the core hardness. This cracking often occurs
several hours after the component has cooled from the tempering
temperature and often runs through the entire cross section. This
problem can be removed by retempering for 3 h at the original
tempering temperature, which is associated with a change in
hardness of 2 HRC points maxi- mum (Ref 61).
Distortion in Heat Treatment
Distortion can be defined as an irrevers- ible and usually
unpredictable dimensional change in the component during processing
from heat treatment and from temperature variations and loading in
service. The term dimensional change is used to denote changes in
both size and shape (Ref71). The heat-treatment distortion is
therefore a term often used by engineers to describe an un-
controlled movement that has occurred in a component as a result of
heat-treatment operation (Ref 72). Although it is recog- nized as
one of the most difficult and trou- blesome problems confronting
the heat treater and the heat-treatment industries on a daily
basis, it is only in the simplest thermal heat-treatment methods
that the mechanism of distortion is understood. Changes in size and
shape of tool-steel parts may be either reversible or irreversible.
Reversible changes, which are produced by applying stress in the
elastic range or by temperature variation, neither induce stresses
above the elastic limit nor cause changes in the metallurgical
structure. In this situation, the initial dimensional values can be
restored to their original state of stress or temperature.
Irreversible changes in size and shape of tool-steel parts are
those that are caused by stresses in excess of the elastic limit or
by
changes in the metallurgical structure (for example, phase
changes). These dimension- al changes sometimes can be corrected by
mechanical processing to remove extra and unwanted material or to
redistribute resid- ual stresses or by heat treatment (annealing,
tempering, or cold treatment).
When heat-treated parts suffer from dis- tortion beyond the
permissible limits, it may lead to scrapping of the article,
rendering it useless for the service for which it was intended, or
it may require necessary cor- rection. Allowable distortion limits
vary to a large extent, depending on service appli- cations; in
cases where very little distortion can be tolerated, specially
desired tool steels are used. These steels possess metal- lurgical
characteristics that minimize distor- tion.
Types of Distortion Distortion is a general term that
involves
all irreversible dimensional change pro- duced during
heat-treatment operations. This can be classified into two
categories: size distortion, which is the net change in specific
volume between the parent and transformation product produced by
phase transformation without a change in geomet- rical form, and
shape distortion or warpage, which is a change in geometrical form
or shape and is revealed by changes of curva- ture or curving,
bending, twisting, and/or nonsymmetrical dimensional change with-
out any volume change (Ref 72, 73). Usually both types of
distortion occur during a heat-treatment cycle.
Dimensional Changes Caused by Changes in Metallurgical Structure
during Heat Treat- ment. Various dimensional changes pro- duced by
a change in metallurgical structure during the heat-treatment cycle
of tool steels are described below (Ref 74).
Heating (Austenitizing). When annealed steel is heated from room
temperature, ther- mal expansion occurs continuously up to Ac~,
where the steel contracts as it trans- forms from body-centered
cubic (bcc) fer- rite to face-centered cubic (fcc) austenite. The
extent of decrease in volumetric con- traction is related to the
increased carbon content in the steel composition (Table 4).
Further heating expands the newly formed austenite.
Hardening. When austenite is cooled quickly, martensite forms;
at intermediate cooling rates, bainite forms; and at slow cooling
rates, pearlite precipitates. In all these transformation
sequences, the magni- tude of expansion increases with the de-
crease in carbon content in the austenite (Table 4). The volume
increase is maximum when austenite transforms to martensite,
intermediate with lower bainite, and is least with upper bainite
and pearlite (Table 4). The volume increases associated with the
transformation of austenite to martensite in 1 and 1.5% carbon
steels are 4.1 and 3.84%,
respectively; the volume increases involved in the
transformation of austenite to pearlite in the same steels are 2.4
and 1.33%, re- spectively. Such volume increases are less in alloy
steels and least in 2C-12Cr and A10 tool steels. It should be noted
that plastic deformation (or strain) occurs during such
transformations at stresses that are lower than the yield stress
for the phases present (Ref 75). The occurrence of this plastic
deformation, called the transformation plas- ticity effect,
influences the development of stresses during the hardening of
steel parts (Ref 76). During quenching from the austen- ite range,
the steel contracts until the M~ temperature is reached, then
expands dur- ing martensitic transformation; finally, ther- mal
contraction occurs on further cooling to room temperature. As the
hardening tem- perature increases, a greater amount of car- bide
goes into solution; consequently, both the grain size and the
amount of retained austenite are increased. This also increases the
hardenability of steel.
More trouble with distortion comes from the quenching or
hardening operation than during heating for hardening, in which the
faster the cooling rate (that is, the more severe the quenching),
the greater the dan- ger of distortion. When the milder quen-
chants are used, the extent of distortion is lessened. The severity
of quenching thus influences the distortion of components.
The dependence of volume increase, par- ticularly in tools of
different dimensions, on grain size (or hardenability) is another
im- portant factor. Variations in volume during quenching of a
fine-grained shallow-harden- ing steel in all but small sections is
less than a coarse-grained deep-hardening steel of the same
composition.
Tempering. There is a certain correlation between the tempering
temperature and volume change. Tempering reduces the vol- ume of
martensite but not adequately enough to equalize completely the
prior volume increase as a result of martensitic transformation
unless the components are completely softened. In low-alloy and
plain (medium- and high-) carbon steels, during the first and third
stages of tempering, a decrease in volume occurs that is associated
with the decomposition of: high-carbon martensite into low-carbon
martensite plus ~-carbide in the former stage, and aggregate of
low-carbon martensite and t-carbide into ferrite plus cementite in
the latter stage. In the second stage, however, an increase in
volume takes place (due to the decomposi- tion of retained
austenite into bainite) that tends to compensate for the early
volume reduction. As the tempering temperature is increased further
toward the A~, more pro- nounced volume reduction occurs. In some
highly alloyed tool-steel compositions, the volume changes during
martensite forma- tion are less striking because of the large
proportion of retained austenite and the
-
Defects and
Table 7 Typical volume percentages of microconstituents existing
in four different tool steels after their standard hardening
treatments
Retained Undissolved As-quenched Martensite, austenite,
carbides,
Steel Hardening treatment hardness, HRC vol% vol% vol%
W1 790 C (1450 F), 30 rain; WQ 67.0 L3 845 C (1550 F), 30 min;
OQ 66.5 M2 1225 C (2235 F), 6 rain; OQ 64 D2 1040 C (1900 F), 30
rain; AC 62
Note: WQ, water quenched; OQ, oil quenched; AC, air cooled.
88.5 9 2.5 90 7 3.0 71.5 20 8.5 45 40 15
resistance to tempering of alloy-rich mar- tensite. These
hardened steels show sharp increases both in hardness and volume
be- tween 500 and 600 C (930 and 1110 F) owing to the precipitation
of very finely dispersed alloy carbides from the retained
austenite. This produces a depleted matrix in alloy content,
raising the M~ temperature of retained austenite. During cooling
down from the tempering temperature, further transformation of
retained austenite into martensite will occur with an additional
increase in volume.
Size Distortion. Table 7 shows the typical volume percentages of
microconstituents present in four different tool steels after their
standard hardening treatments. Typi- cal dimensional changes during
hardening and tempering of several tool steels are given in Table
8. It is apparent here that some steels such as M3 and M41
high-speed steels show appreciable increase in size of about 0.2%
after hardening and tempering between 540 and 595 C (1000 and 1100
F) to produce complete secondary hardening. Other types, such as
A10, expand very little when hardened and tempered over the en-
tire temperature range up to 595 C (1100 F). Excessive size changes
in oil-hardening nonshrinkable tool steel is usually caused by lack
of stress relief (when necessary), and hardening and/or tempering
at the in- correct temperature. The golden rule is to learn to be
suspicious of tools that are seriously off size in only one
dimension. It is further noted that alloying addition in steels
brings about a change in the specific volume of many
microconstituents, but to a
lesser extent than carbon (Ref 77). This table provides
comparative data on size distortion in a variety of steels;
however, this information cannot be used alone to predict shape
distortion factor.
Shape Distortion or Warpage. This is sometimes called
straightness or angularity change. It is found particularly in
nonsym- metrical components during heat treatment. From the
practical viewpoints, warpage in water- or oil-hardening steels is
normally of greater magnitude than is size distortion and is more
of a problem because it is usually not predictable. This is caused
by the sum effect of more than one of these factors:
Rapid heating (or overheating), drastic (or careless) quenching,
or nonuniform heating and cooling causes severe shape distortion.
Slow heating as well as pre- heating of the parts prior to heating
to the austenitizing temperature yields the most satisfactory
result. Rapid quench- ing produces thermal and mechanical stresses
associated with the martensitic transformation. In the case of low-
and high-hardenability steels, respectively, this problem becomes
severe or very small
Residual stresses present in the compo- nent before heat
treating. These arise from machining, grinding, straightening,
welding, casting, spinning, forging, and rolling operations, which
will also furnish a marked contribution to the shape change (Ref
78)
Applied stress causing plastic deforma- tion. Sagging and creep
of the compo-
Distortion in Heat-Treated Parts / 613
nents occur during heat treatment as a result of improper
support of components or warped hearth in the hardening fur- nace.
Hence, large, long, and complex- shaped parts must be properly
supported at critical positions to avoid sagging or preferably are
hung with the long axis on the vertical
Nonuniform agitation/quenching or non- uniform circulation of
quenchant around a part results in an assortment of cooling rates
that creates shape distortion (Ref 79). Uneven hardening, with the
forma- tion of soft spots, increases warpage. Similarly, an
increase in case depth, par- ticularly uneven case depths in
case-hard- ening steels, increases warpage on quenching (Ref
80)
Tight (that is, thin and highly adherent) scale and
decarburization, at least in cer- tain areas. Tight scale is
usually a prob- lem encountered in forgings hardened from
direct-fired gas furnaces having high-pressure burners. Quenching
in ar- eas with tight scale is extremely retarded compared to the
areas where the scale comes off. This produces soft spots, and, in
some cases, severe unpredicted distor- tion. Some heat treaters
coat the compo- nents with a scale-loosening chemical pri- or to
their entry into the furnace (Ref 79). Similarly, the areas beneath
the decarbur- ized surface do not harden as completely as the areas
below the nondecarburized surface. The decarburized layer also var-
ies in depth and produces an inconsistent softer region as compared
to the region with full carbon. All these factors can cause a
condition of unbalanced stresses with resultant distortion (Ref
79)
Long parts with small cross sections (>L = 5d for water
quenching, > L = 8d for oil quenching, and > L = 10d for
austemper- ing, where L is the length of the part, and d is its
diameter or thickness)
Thin parts with larger areas (>A = 50t, where A is the area
of the part, and t is its thickness)
Unevenness of, or greater variation in, section
Table 8 Typical dimensional changes during hardening and
tempering of several tool steels Hardening treatment
Tool Temperature Quenching steel ~U- 1 medium
Total change in linear dimensions
after quenching, %
Total change in linear dimensions~ %, after tempering at
150 *C 205 *C 260 *C 315 *C 370 *C 300 *F 400 *F 500 *F 600 *F
700 *F
425 *C 480 *C 510 *C 540 *C 565 *C 595 *C 800 *F 900 *F 950 oF
1000 oF 1050 *F 1100 *F
Ol 815 1500 Oil 0.22 OI 790 1450 Oil 0.18 06 790 1450 Oil 0.12
A2 955 1750 Air 0.09 A10 790 1450 Air 0.04 D2 11)10 1850 Air 0.06
D3 955 1750 Oil 0.07 D4 1040 1900 Air 0.07 D5 1010 1850 Air 0.07 H
I I 1010 1850 Air 0.11 HI3 1010 1850 Air -0.01 M2 1210 2210 Oil -0
.02 M41 1210 2210 Oil -0 .16
0.17 0.16 0.18 0.09 0.12 0.13 - 0.07 0.10 0.14 0.10 0.06 0.06
0.08 0.07 0.00 0.00 0.08 0.08 0.03 0.03 0.02 0.00 0.04 0.02 0.01 -0
.02 0.03 0.01 -0.01 -0 .03 0.03 0.02 0.01 0.00 0.06 0.07 0.08
0.08
0.00 -0 .05 -0 .06 0.05 0.04 0.01 0.01 0.02 . . . . 0.01 -0
.02
- 0 . 4 -0.03 ' 0.3 0.03 0.3 0.01 . . . . . 0.00
-0 .06 -0 .17
-0 .07 0.06 0.01 0.06
0.05 0.05 0.12 0.06 0.10 0.08
0.14 0,21
0.02
0.16 0.23
-
614 / Process and Quality Control Considerations
Examples of Distortion
Ring Die. Quenching of ring die through the bore produces the
reduction in bore diameter as a result of formation of martens-
ire, assoc ia ted with the increased volume. In other words, metal
in the bore is upset by shrinkage of the surrounding metal and is
short when it cools (Ref 24). However , allover quenching causes
the outside diam- eter to increase and the bore diameter to
increase or decrease, depending upon pre- cise dimensions of the
part. When the out- side diameter of the steel part is induction-
or f lame-hardened (with water quench), it causes the part to
shrink in outer diameter (Ref 63). These are the examples of the
effect of mode of quenching on distortion (Ref 81).
Thin die (with respect to wall thickness) is likely to increase
in bore diameter, decrease in outside diameter, and decrease in
thick- ness when the faces are hardened. If the die has a very
small hole, insufficient quench- ing of the bore may enlarge the
hole diam- eter because the body of die moves with the outside
hardened portion.
Bore of Finished Gear. Similarly, the bore of a finished gear
might turn oval or change to such an extent that the shaft cannot
be fitted by the allowances that have been provided. Even a simple
shape such as a diaphragm or orifice plate may, after heat
treatment, lose its flatness in such a way that it may become
unusable.
Production of Long Pins. In the case of the production of long
pins (250 mm long x 6 mm diameter, or 10 V4 in.) made from
medium-alloy steel, it was found, after con- ventional hardening,
that when mounted between centers, the maximum swing was over 5 mm
(0.20 in.). However , the camber could be reduced to within
acceptable limits by martempering, intense or press quench-
ing.
Hardening and Annealing of Long Bar. When a 1% carbon steel bar,
300 mm long (or more) 25 mm diameter (12 in. long, or more, 1 in.
diameter), is water quenched vertically from 780 C (1435 F), the
bar increases both in diameter and volume but decreases in length.
When such bars are annealed or austenitized, they will sag badly
between the widely spaced supports. Hence, they should be supported
along their entire length in order to avoid distor- tion.
Hardening of Half-Round Files. Files are usually made from
hypereutectoid steel containing 0.5% chromium. Files are heated to
760 C (1400 F) in an electric furnace after being surface coated
with powdered wheat, charcoal , and ferrocyanide to pre- vent
decarburization. They are then quenched vertically in a water tank.
On their removal from the tank, the files appear like the
proverbial dog 's tail. The flat side has curved down, the camber
becomes ex-
cessive, and the files can no longer be used in service. One
practical solution is to give the files a reverse camber prior to
quench- ing. The dead fiat files could, however, be made possible,
and the judgment with re- gard to the actual camber needed depends
upon the length and the slenderness of the recur files (Ref
82).
Similarly, when a long slender shear knife is heat treated, it
tends to curve like a dog 's tail, unless special precautions are
taken.
Hardening of Chisels (Ref 63). Chisels about 460 mm (18 in.)
long and made from 13 mm (0.5 in.) AISI 6150 bar steel are
austenitized at 900 C (1650 F) for 1.5 h and quenched in oil at 180
C (360 F) by stand- ing in the vertical position with chisel point
down in special baskets that allow stacking of two 13 mm (0.5 in.)
round chisels per 650 mm 2 (1 infl) hole. Subsequently, hardened
chisels are tempered between 205 and 215 C (400 and 420 F) for 1.5
h. These heat- treated parts show 55 to 57 HRC hardness but are
warped. The reasons for this distor- tion are:
The portion of the bar that touches the basket cools slowly,
producing uneven contraction and thermal stress
The martensite formation is delayed on the inner or abutting
side of the bar, causing unequal expansion during trans- formation.
This distortion can be elimi- nated or minimized by loading the
parts in the screen-basket in such a way that stacking arrangement
permits sufficient space between each part and by slightly
decreasing the austenitizing temperature (Ref 62). Distortion can
also be mini- mized by austempering the part, provided that the
carbon content is on the high side of specification to produce the
lower bai- nitic structure of 55 to 57 HRC. If higher yield stress
is not warranted, only chisel ends need hardening and subsequent
tem- pering (Ref 63)
Hardening of a Two-Pounder Shot. The hardness of a two-pounder
shot was specified at 60 HRC on the nose and 35 HRC at the base. A
differential hardening technique was performed on the shot made of
a Ni-Cr-Mo steel. This technique consisted of quenching the shot in
the ice-cold water by its immersion in a tank up to the shoulder,
followed by drawing out the water from the tank at a stipulated
rate until the water line reached the base of the nose. The final
step involved withdrawing the shot from the tank when completely
cold. The back end was then softened by heating in a lead bath
after initial tempering. The first few shots hardened in this way
were observed to split vertically across the nose. The failure was,
however, avoided by withdrawal of the shot before attaining
ice-cold temperature and its subse- quent immersion in warm water
(Ref 82).
Hardening of a Burnishing Wheel. In the manufacture of railway
axles, the gearing
surface on which the axle rests in the hous- ing has to be given
a high burnishing polish employing a circular pressure tool that is
made of !.2C-1.5Cr steel. Fo r satisfactory results, the hardness
of the tool surface should be about 60 HRC. It has been found that
the tool usually cracks before its with- drawal from the cold-water
quenching bath. T