IFHTSE CONGRESS

A PR IL 1 8 -21, 2016

H Y AT T R E G E N C Y

S AVA NN A H, GEORGI A

Organized by:

IFHTSE_Final Program.indd 1 4/8/2016 10:09:01 AM

International Federation of Heat Treatment and Surface Engineering 2016 Copyright © 2016 ASM International®
Proceedings of the 23rd IFHTSE Congress All rights reserved.
April 18–21, 2016, Savannah, Georgia, USA asminternational.org

Metallurgical Aspects of Distortion and Residual Stresses in Heat Treated Parts

D. Scott MacKenzie, PhD, FASM
Houghton International, Inc., Valley Forge PA 19426
[email protected]

Place author name(s) from company 2 here in Arial bold and italic, 10 point
Place company 2 name, city, state and country here in Arial italic, 10 point
Place email, phone, etc. here (not required) 10 point

Abstract Incoming
Shape Machine
Heat
Finish
Material Treat
While modeling of quenchant behavior and microstructural
• Bar • Forge • Turning • Solution • Straighten
response has matured, the sensitivity to variation of common
• Plate • Form • Mill • Quench • Hone
metallurgical and production practices are often not considered.
• Rod • Hob • Temper • Grind
As an example, the variation in hardenability of the purchased • Plate
material is generally not considered when predicting • Paint
microstructure and residual stress. Machining practices,
including worn tools and use of tooling beyond the expected Figure 2. Typical fabrication sequence or process chain of a
life can have dramatic influence on distortion and residual generic steel or aluminum component.
stress. The quench oil chosen, and the quench part are all
important variables to consider. Finally the presence of adjacent The literature is filled with papers trying to understand and
parts and the racking method can also effect the residual stress model the distortion of components. Modeling has been used
state and distortion. The effects of common metallurgical to understand the heat treatment process and the entire process
variables on the residual stress and distortion of heat treated chain. However, modeling is usually used to model the
parts are reviewed. Factors such as hardenability, segregation, component under perfect conditions – new tooling,
and rack and fixture design are discussed. homogeneous steel, uniform heat treating conditions and
quenchant. Rarely are sensitivity studies performed to look at
Introduction the expected variation of distortion due to manufacturing
variation.
The problem of distortion is universal across industry. A study
It is the intent of this paper to review some of the possible
by Thoben (1) indicated that the 1995 losses from heat
variation due to metallurgical and manufacturing processes.
treatment alone in the German machine, automotive and
transmission industry exceeded 850M € (1 Billion USD). This Literature Review
does not include the rest of Europe, Asia or the Americas. The
problem is truly immense. In the typical manufacturing process to manufacture a steel or
aluminum part, incoming material is purchased; shaped by
Distortion is not limited to heat treatment, but each step in the
forming, forging or casting; machined; heat treated; and finish
manufacturing process contributes to distortion 0. The sources
processed (final machining, plating or painting). This is
of distortion follow the manufacturing process chain. Looking
illustrated in Figure 2.
at distortion, the primary sources are volume changes from
phase transformations and precipitation, or from deformation There are many sources of distortion that have been identified.
from either plastic or elastic deformation (Figure 1). A fishbone chart illustrating some of the common sources of
distortion is shown in Figure 3. It is not the purpose of this
paper to review all the possible sources and permutations of
Distortion distortion, but to highlight some of the important factors.
Prior Microstructure
Deformation
Volume The incoming raw material plays a critical role in the quality of
Changes
a finished part. The prior microstructure and hardenability of
the alloy grade selected, as well as the cleanliness of the
Plastic
Deformation
Elastic
Deformation
Phase
Transformatio Precipitation incoming material affect distortion and the mechanical
n
properties (fatigue and impact properties). This is illustrated in
Figure 4.
Relaxation and Transformatio Thermal
Aging n Stresses Stresses
Residual Stress
Hippenstiel (3) evaluated the effect of prior grain size on
distortion and found that finer grain size contributed to reduced
Figure 1. Global sources of distortion (2). distortion. He also found that by reducing the amount of grain
size variation, further reductions in distortion could be
achieved.
392
Figure 3. Fishbone diagram showing some of the many factors affecting distortion.

Variations in phase. An additional difficulty was the calculation after

Hardness tempering was overcome by the inclusion of a coarse secondary
hardening carbide. The validation of the model showed good
Differential correspondence with experiment.
Alloy
response during
Segregation The examination of the prior microstructure on the distortion of
machining
press-quenched gears was conducted by Reardon (5). In this
Differential
Response during study, 464 mm diameter AISI 8620 H steel automotive
Heat Treatment transmission gears were processed in two separate furnace
loads. For the first set of gears, which were designated Series
Changes in 1, the normalizing temperature that was used was 926ºC. These
hardenability
Quality of Steel Uniform grain gears received a single normalizing cycle. This is illustrated in
size Varies by Figure 5.
manufacturer & The second set of gears, labeled Series 2, was normalized twice
Country
in succession. The first normalizing cycle they received was
identical to that used for the Series 1 gears. For the second
Impact Strength normalizing cycle, the temperature was raised to 954°C. Each
Material group of 24 gears was subsequently processed through a gas
Cleanliness
Fatigue carburizing furnace using an endothermic atmosphere at a
Response temperature of 926ºC to generate effective case depths in the
range of 0.0016 to 0.0022 m. The gears were all cooled to room
Figure 4. Quality of steel and impact on distortion and property temperature after carburizing was completed. They were then
response. reheated to the austenitizing temperature of 854ºC, and
individually press quenched.
In another study (4), electro-slag remelted and vacuum melted
The resulting out of round distortion was measured (Figure 5).
and remelted X38CrMoV5-1 hot-work steels were examined
It was found that the double normalized gears exhibited
for distortion, residual stresses, microstructure and properties.
substantially lower distortion, and much reduced distortion
The vacuum heat treatment were varied to develop a model of
scatter than the single normalized gears. The double
phase volume fraction, residual stress and hardness. Using a
normalized gears had a mean out of roundness of 100µm, with
Gleeble, the transformation kinetics and flow curves for the
a standard deviation of 25.4 µm. The single normalized gears
alloys were measured. A hardness after tempering was had a mean out of roundness of 330 microns, with a standard
developed based on the volume fraction and hardness of each
deviation of 154µm.

393
 20 Banding or segregation can produce a wide variation in the local
hardness, even within the carburized layer (Table 1). This
segregation can cause non-uniform response to heat treatment
15 and result in uncontrolled distortion (Figure 7 & Figure 8).
Frequency

These studies are all consistent, in that it shows that the prior
10 thermal processing prior to machining and heat treatment, play
a critical role in controlling the distortion of a heat treated
5 component.

Distortion, Out of Roundness (inches)

Normalized Double Normalized

Figure 5. Effect of normalization cycle on the out of roundness

of two different groups of carburized AISI 8620H gears after
press quenching and normalization.

An examination of the effect of the microstructure prior to heat

treatment was examined by Prinz (6). In this investigation, the
microstructure of bar material was varied by normalizing,
annealing or hardening and tempering. The resultant
microstructures were primarily ferrite and pearlite in the
Figure 7. Segregation and banding in AISIS 4320 carburized
normalized “as-delivered” condition, with some bainite and
gear showing non-uniform effective case depth (Courtesy of L.
martensite due to the high hardenability. The annealed
Ferguson DANTE Solutions).
structure was ferrite and pearlite, while the hardened and
tempered structure was tempered martensite. The extent of the Hardenability
distortion was evaluated by measurements before and after heat
treatment of machined shafts. The effect of hardenability on The hardenability of a steel is influenced by its chemical
the distortion of the heat treated shafts were also examined. composition and grainsize. The chemistry further affects the
martensite start and finish temperatures, Ms. Because of
The results of the study showed a strong influence of the prior variations in chemistry within an alloy specification. Even
microstructure on the resulting distortion (Figure 6). Increasing within a specific heat lot of material, the chemistry and
hardenability reduced distortion. However the authors noted hardenability can vary. Since higher hardenability steels are
that it is likely that the reduced grain size from the prior heat more deeply hardened than a lower hardenability steel within
treatments likely changed the hardenability, so the effects are the same grade, it can be expected that the higher hardenability
compounded with prior heat treatment effects. chemistry will have a higher core hardness and effective case
0.04 depth. This would also indicate that the distortion would
increase as hardenability increases.
Change in Curvature (mm)

Funatani (7) showed that the distortion of SCr.420 steel (a Cr

0.03
case hardening grade) increased with a small variation in the
Jominy End Quench hardenability. The change in over ball
0.02 diameter changed from 20 µm to 80 µm as the hardness at 12.5
mm from the quenched end on the Jominy End Quench Test
changed from 24 to 30 HRC (Figure 9). Hypoid ring gears of
0.01
SCM430H showed that the maximum pitch angle increased as
a function of the hardenability of the steel increased (Figure 10).
0.00 In one study (8), a customer was looking to increase
Normalized Annealed Hardened and performance of an automotive pinion gear by changing alloy
Tempered from AISI 8620 to AISI 4320. the distortion of an AISI 8620
Prior Heat Treatment to Hardening and 4320 automotive pinion gear was compared, using identical
heat treating conditions. The core requirements were the same
Figure 6. Variation of distortion with microstructure prior to (35-40 HRC).
heat treating (6).

394
 For 8620, during quenching, the core transforms almost
completely to bainite. While some asymmetry of
transformation occurs, equal amounts of bainite are formed.
There is less volume expansion due to bainitic transformation,
resulting in less distortion.
For AISI 4320, the core of the pinion shaft transforms to
approximately equal portions of martensite and bainite.
However, because of the differential cooling rates, one side of
the pinion shaft transforms to martensite much earlier than the
opposite side. This asymmetry of martensite formation and the
accompanying volume change results in an initial bow of the
pinion. Once the opposite side of the pinion transforms to
martensite, a stress reversal occurs, with the final bow of 0.15
mm.

80

Change in Overall Diameter, µm

Figure 8. Segregation and banding in 4320 carburized gear
showing hardness variation within case (Courtesy L. Ferguson, 70
DANTE Solutions, Inc.). See Table 1 for hardness at locations
1-6. 60

Table 1. Hardness variation noted within case of carburized 50

AISI 4320 gear (Courtesy L. Ferguson, DANTE Solutions, Inc.).
40
Hardness 30
Location
(HRC)
1 56.3 20
22 24 26 28 30 32
2 60.3
Jominy (12.5 mm) HRC
3 33.3
Figure 9. Influence of steel hardenability on the over ball
4 66.3 diameter of Carburized Module 2.8 Spur gears. Gears were
5 60.9 carburized at 920°C and quenched in 90°C oil (7).
6 54.9
Recommendations included increasing the agitation or
modifying the racking to achieve uniform heat transfer around
The results of the modeling, show that the distortion of the AISI the circumference of the shaft.
4320 carburized pinion gear is approximately 0.15 mm –
approximately one order of magnitude greater than the bow of 7
the carburized AISI 8620 gear. This behavior was explained by
Maximum pitch error, µm

6
the differences in the Continuous Cooling Curves of the two
materials. The AISI 8620 material, with a short martensite 5
shelf, has a significantly lower hardenability than AISI 4320. 4
Fast cooling rates are required to achieve a fully martensitic
structure. AISI 8620 also has a significant bainite bay at the 3
expected cooling rates expected during quenching. AISI 4320
2
has a much higher hardenability, with a large martensite shelf.
Slower cooling rates can still achieve a fully martensitic 1
structure.
0
The cooling rates on opposite sides of the pinion were
18 22 26 30 34 38 42
determined to be approximately 800°C/min and 400°C/min
respectively. Jominy (12.5 mm), HRC
For AISI 8620, the CCT diagram predicts that the
microstructure would be predominately bainite with small Figure 10. Influence of steel hardenability on the maximum
portions of martensite and pearlite for both cooling rates. For pitch angle of hypoid ring gears (SCM420H) (7).
AISI 4320, the CCT diagram predicts approximately equal
portions of martensite and bainite. However, for the faster An investigation into the effect of variation of hardenability
cooling rate, martensite would form earlier than for the slower within heats of 16NiCrS4 carburizing steel was conducted by
cooling rate. Stormvinter, et al (9). Three different heats of material (Table
2) were obtained as tubes of 71 mm diameter and a wall

395
thickness of 17 mm. The tubes were sectioned and machined 20
into rings of 70 mm diameter, with a wall thickness of 16 mm
and a height of 19 mm. The calculated ideal diameter, DI in
mm, for Heats A, B and C were 80.47 mm, 91.73 mm and 15

Height, mm
100.82 mm. All heats satisfied the requirements for 16NiCrS4.
10
Table 2. Heat examined by Stormvinter et al (9).
5
Steel heat C Si Mn P S Cr Ni Mo Cu
0
Heat A 0.2 0.07 1.00 0.12 0.41 1.01 1.03 0.13 0.17
69.98 70.00 70.02 70.04
Heat B 0.2 0.07 1.03 0.16 0.41 1.04 1.05 0.17 0.17
Width, mm
Heat C 0.2 0.07 1.02 0.08 0.41 1.03 1.39 0.16 0.18

Heat B Heat C Heat A

The rings were first machined then stress relieved at 600°C for
5 hours. Rings were austenitized at 850°C for 1 hour with no Figure 12. Radial distortion of the outer surface of the heat
carburizing. The rings were then quenched in an accelerated treated rings. The nominal outer diameter is 70.01 mm. The
quench oil at 50°C. The rings were measured for distortion surfaces become concave.
before and after heat treatment. The results of the
measurements at the inner and outer diameters are shown in Four mill heats of steel were examined, and the results show
Figure 11 and Figure 12. The hardness of the rings were that there is a correlation between the measured hardenability
identical within the error of the hardness measurement. of the steel, and the resultant distortion of the wheel. Different
18 heats, with different hardenability, affected the distortion of the
16 crown wheel.
14 10
12
Height, mm

10 5
8
6 0
-10 -5 0 5 10
4
-5
2
0 -10
38.92 38.97 39.02 39.07
Width, mm -15
Heat B Heat C Heat A
Heat H Heat J Heat M Heat N
Figure 11. Radial distortion of the inner surface of the heat
treated rings. The inner surface becomes convex. The inner Figure 13. Score scatter plot from principal component
diameter shrinks for steel heats B and C, but grows for A. The analysis. The first PCA component and the Second PCA
nominal inner diameter is 38.98 mm. component are represented by the x-axis and y-axis
respectively. Each dot represents the relative distortion of a
For ring components, a lower hardenability alloy would tend to crown wheel. The distortions are examined corresponding to
cause the ring to become thinner and have a larger overall different steel heats.
diameter. For higher hardenability alloys, the ring would
become thicker and tend to shrink. Machining
The effect of distortion was also studied (9) on production Prior operations such as machining create residual stresses at
vehicle crown wheels. In this study the crown wheels were the surface. This stress is relieved during heat treatment,
manufactured from forgings of 17NiCrMoS6-4 using a steel resulting in distortion.
from a single steel mill. All crown wheels were machined and During turning or milling operations, the work piece is held
heat treated at the same time to minimize unknown influences. fixed by a chuck or jaws. The method of holding the work piece
The distortion of the crown wheels was then measured using a can result in residual stresses. In one study (10), 100Cr6 (SAE
coordinate measuring machine (Figure 13). 52100) bearing rings were evaluated on the influence of cutting
speed, depth of cut, and feed rate on the roundness. Two
different types of clamping mechanisms were used: a mandrel
clamp and segmented jaws.

396
The study showed (Figure 14) that the residual stresses in the 700
tangential direction of the two rings had a mean residual stress

Tangential Residual Stress, MPa

for the different clamping methods were nearly identical with a 650
mean residual stress of 600 MPa. The segmented jaws show a 600
periodicity of three around the circumference, with the
tangential residual stress varying between 500 and 700 MPa. 550
The mandrel chuck showed a slight periodicity of 6, but varied 500
to a much less greater degree, with the variation between
450
approximately 590 and 620 MPa.
The segmented jaws had a segment of approximately 120°, 400
however, the real contact was limited to a single point in the 0 45 90 135 180 225 270 315 360
middle of each segment. Therefore the clamping forces are Circumference Angle, °
induced at three locations 120° apart. This results in the ring
bulging at these locations. The mandrel supports the bearing
Mandrel Segment Jaws
uniformly around the circumference, resulting in a uniform
residual stress.
Figure 14. Residual stress distribution around a ring
Further work was conducted by Grote (11), comparing the circumference depending on the clamping technique used.
residual stresses generated by different methods of clamping a
work piece during internal turning.
650
The influence of cutting parameters on residual stresses were

Mean Tanjential Residual Stress,

also investigated for rings that were clamped by a mandrel. 600
Four cutting parameters were examined: cutting speed, nose 550
radius, depth of cut and feed rate. The influence of the different 500
parameters is shown in Figure 15. The parameters used for the
DOE is shown in Table 3. 450
Heat Treatment MPa 400
Feed Rate Nose Depth of Cutting
Heat treating and quenching is a complex business. The
(mm) Radius Cut (mm) Speed
configuration of parts is endless, as is the types of furnaces
available for heat-treating. Numerous variables in the (mm) (m/min)
quenching process alone govern the ability of a part to meet Machining Parameter
distortion requirements (Figure 3). Heat-treating is a constant
balancing process. It is important to balance the ability of the Low High
material to achieve properties, while at the same time control
distortion. Figure 15. Results of DOE conducted by Nowag, et al, on
Multiple heat transfer regimes occur during the quenching 100Cr6, using a TiN tool and a 3% emulsion coolant.
process at the same time. These heat transfer mechanisms
change as a function of the quench path. Further changes occur Table 3. Levels used by Nowag et al, examining the effects of
as the quenchant oxidizes, complicating the analysis. Agitation machining parameters on the external turning of 100Cr6 (SAE
and racking interact, changing heat transfer at different 52100).
locations on the part. There is no quantitative method to
understand the interaction of fluid flow and the part, with the Nose Cutting
Feed Depth of
accompanying changes in time dependent heat transfer. Radius Speed
Rate Cut (mm)
Phase transformations result in volume changes and (mm) (m/min)
transformational strains. Diffusional processes (pearlite + 0.3 0.8 0.7 200
ferrite, and bainite) and non-diffusional processes (martensite) Levels
0.6 1.2 1.2 300
are time and temperature dependent.
Sources of stress and strain include thermal strain, creep and
strains from phase transformation further complicate the
business of heat treating (Figure 16).

397
 Thermodynamic 10

Change in Vector Curvature Length, µm

f(T, t, h, Cp, k)
8

2
Metallurgical Mechanical 0
f(Φd, Φnd) f(εe, εp, σe, σp, T)
-2
Figure 16. Interrelationships of primary variables on a heat Carburizing Surface Carbon Process Carburizing
treated part and resulting residual stress and distortion. Temperature Content Sequence Depth
Process Variable
Thermal stresses result from thermal gradients: surface to
center and surface to surface. Surface to surface thermal Figure 17. Effects carburizing parameters on curvature vector
gradients result from differential heat transfer around the length of an AISI 5120 carburized shaft 14 mm in diameter.
surface of the part. This could be from inadequate agitation
during quenching, or from thinner sections (gear tooth) attached
0.02
to a thicker or heavier body. Surface to center thermal gradients
0.00
are due to the surface cooling faster than the interior, limited by

Change in Diameter, mm
-0.02
the thermal diffusivity of the part. As a general rule, the faster -0.04
the quench, the greater the thermal gradient. -0.06
Transformation stresses are complex in heat treating. This is a -0.08
-0.10
coupled stress related to phase changes that occur during heat
-0.12
treatment. As a part is heated, the volume expands due to -0.14
thermal expansion. As the temperature increases thru the first -0.16
critical temperature, the volume decreases as ferrite transforms -0.18
to austenite. The volume then expands due to thermal Carburizing Carburizing Process Surface
expansion of austenite until the steady state temperature is Depth Temperature Sequence Carbon
reached. Content
Process Variable
During quenching, there is a volumetric contraction as the work
piece cools. As austenite transforms, there is a volume
expansion. Martensite expands the most, as a function of the Figure 18. Effects of carburizing parameters on the change in
carbon content. Other phases can form, such as bainite or diameter of a AISI 5120 carburized 14 mm shaft.
pearlite, which have much smaller volumetric expansion.
Generally outer fibers transform first, then inner fibers. This is An investigation into the distortion behavior at elevated
further complicated by the effect of martensitic transformation carburizing temperatures was conducted by Kleff et al (13).
start temperatures. Depending on the relative Ms temperatures Heavy duty transmission components with deep case
and the thickness of a part, it is often possible that the core will requirements require long furnaces times to satisfy case depth
transform first due to a much higher Ms temperature. As the requirements. The use of elevated carburizing temperatures
remaining part transforms, a stress reversal occurs. would significantly reduce the time required for carburizing,
An investigation into the influence of carburizing parameter on and increase production rates. Using a selection of various
the distortion behavior of stepped shafts was conducted by parts, the distortion was measured after carburizing at 930°C,
Bahnsen et al (12). In this study, the influence of prior 980°C and 1050°C. Alternative heating and cooling profiles
microstructure, carburizing depth, surface carbon content, were also investigated.
process sequence, carburizing temperature and the secondary Two different grades of steel were used in this study: either
hardening temperature on distortion were evaluated in a DOE. 18CrNiMo7-6 or 20MnCrB5. Both grades were processed as a
The results are shown in Figure 17 and Figure 18. conventional heat with Al or N added. A modified grade with
The largest dominating effect was found to be the carburizing additional Ti (0.002%) and Nb (0.03-0.04%) for grain size
depth. All geometrical parameters exhibited a significant control. Case hardening was performed by gas carburizing at
sensitivity to the depth of carburizing. Other parameters the carburizing temperature. The parts were cooled to the
showed an effect, but not necessarily statistically significant for hardening temperature and oil quenched. An iso-thermal step
this investigation. at 800°C was added to achieve more homogeneous temperature
distribution in the part.
Examination of the radial run-out of the helical bevel gears, as
well as the outer and inner diameters was conducted. The radial
run-out was found to be small for all temperatures processed.

398
The inner and out diameters showed a slight increase after 0.30
carburizing at 940°C and 980°C (≤ 200 µm), or stayed nearly

Change in Distortion Hard-Soft

constant when carburizing at 1050°C (Figure 19). 0.20
0.10
0.30 0.00
Change in Distortion Hard-Soft

0.25

(mm)
-0.10
0.20
-0.20
0.15
-0.30
0.10
(mm)

920 970 1020 1070

0.05 Carburizing Temperature, °C
0.00
-0.05 Clutch Face Opposite Face
-0.10
Figure 20. Influence of cooling down from the carburizing
900 950 1000 1050 1100 temperature to the hardening temperature of 850C on the
Carburizing Temperature, °C distortion of AISI 5120 helical gears.

Figure 19. Results of carburizing temperature on the bore 0.30

diameter of a 191.85mm diameter H6 helical bevel gear.

Change in Distortion Hard-Soft

0.20
No negative influence on higher carburization temperatures 0.10
were observed, provided that an intermediate heating step to
insure homogeneous temperature distribution is included. 0.00
Since the primary focus was to investigate the use of higher (mm) -0.10
temperatures to shorten production time, quenching parts -0.20
directly from the carburizing temperature instead of the
hardening temperature of 850°C was investigated. Helical -0.30
gears were carburized at 930C, 980C and 1050C and quenched 920 970 1020 1070
directly from the carburizing temperature. The results of bore Carburizing Temperature, °C
diameter distortion quenched from the hardening temperature
of 850°C is shown in Figure 20. Bore diameter distortion
Clutch Face Opposite Face
quenched directly from the carburizing temperature is shown in
Figure 21. Figure 21. Influence of quenching directly from the carburizing
The tooth sections and the bore of the gear showed significantly temperature on the distortion of AISI 5120 helical gears.
more scatter of distortion critical dimensions when quenched
directly from the carburizing temperature. The tooth flank The authors found significant reduction in distortion using gas-
deviation increased significantly so that more stock would have pressure quenching than the traditional oil quenching.
to be removed. However, this increases the cost of grinding. However, they found that the use of nitrogen gas as a quenchant
The results showed that cooling down to the hardening lead to higher out-of-flatness than oil quenching. They also
temperature was necessary to achieve good distortion control observed a greater scatter in the results with gas quenching.
and process capability. This was attributed to inhomogeneity of gas flow with in the
Quenching quenching chamber.
The literature of quenching and control of distortion is as Braun (15) reviewed the effects of racking and condition of the
immense as the problem of distortion in industry. The author oil on the resultant distortion of steel components. He found
will try and cover the high points. that as an oil aged, greater distortion resulted.
A comparison of the distortion resulting from different In another study (16), the viscosity and cooling curve was
quenching methods (traditional oil and high-pressure gas evaluated. A single viscosity base oil, and different quantities
quenching) to the same hardness, and the racking method was of a high viscosity (450 cSt at 40°C) were mixed. A standard
conducted by Jurci (14). Helical gears and mating pinions made speed improver quantity was added to the mixtures. Cooling
from low carbon structural steels were gas or low pressure curves were measured using the JIS K2242 silver probe. They
carburized to an effective case of 0.7 mm. The dimensional showed that the speed increased from 90°C/s to 100°C as the
distortion (shrinkage & growth) and the shape distortion viscosity increased to approximately 100 cSt. However, they
(roundness, flatness and tooth deviations) were measured. found that the surface hardness of S45C decreased from HRC
They also measured the distortion as a function of position in 55 to HRC 29 as the kinematic viscosity increased. They
the rack. attributed the improved quenching performance at low

399
viscosities to the composition of the base oils having a low increased (Figure 22). As the temperature of the quenchant was
boiling point fraction. increased, hardness and distortion also decreased (Figure 23).
A comparative study comparing the residual stresses and Racking and Fixturing.
distortion in cylindrical components of AISI 5160H spring steel The impact of racking and fixturing on 120 mm diameter
quenched at various temperatures of 25% poly-alkylene glycol 20MnCr5 (SAE 5120) gear blanks was investigated by Clausen
(PAG) and compared to an accelerated quench oil was (22). In this study, gear blanks were carburized and quenched.
conducted by Sarmiento (17). During carburizing, the disks supported by 2-point loading or
The results showed that no residual stresses were measured in three point loading. Parts were preheated at 850°C and
cylinders quenched in either PAG or quench oil for diameters carburized at 940°C to obtain a carburized depth of 0.8 mm.
of 13.5 and 20 mm. Quenching was accomplished high pressure nitrogen.
In a sensitivity study of quenching variables (Orientation, Measurements were carried out before and after heat treatment.
Slenderness ratio, Ms temperature, Quench Path and Immersion It was determined that the method of racking was the most
Rate), MacKenzie (18) examined the effect of the primary critical factor affecting gear blank distortion (Figure 24 and
quenching variables on the hoop and axial stresses in simple Figure 25).
cylinders. This study found that the orientation (horizontal or 45 200
vertical) and the Ms temperature were the primary main effects
for increasing radial and hoop stresses. 40
150

Hardness, HRC

Distortion, µm
Racking in a vertical fashion increased the hoop stress, as did
decreasing the martensite start temperature. Radial stresses 35
were decreased by racking vertically, while decreasing the 100
30
martensite start temperature increased radial stresses. The
immersion rate was shown to have little effect on the residual 50
25
stresses as the part cooled. The quench path, specifically the
stability of the vapor phase, caused increased residual stresses. 20 0
A more stable vapor phase resulted in reduced residual radial JIS 2-2 JIS 2-1 JIS 1-1 JIS 1-2
and hoop stresses. This was thought to be due to decreased
Oil Type
thermal gradients present. This result was unexpected, and was
thought to be associated with the alloy selected (41XX). This
Hardness Elliptical Distortion Cylindrical Distortion
alloy has a large hardenability, so decreased initial thermal
gradients would show less resultant residual stresses.
Figure 22. Comparison of quench oil selection for distortion
Arimoto (19) reviewed the status of quenching Fe-Ni alloys, control. SCM420 bearings (80 mm in diameter by 44 mm
and modeling residual stresses. His work showed that thick), quenched into oils of increasing speed.
transformation plastic strain is generally similar to plastic strain
during heat treatment. Plastic strain is mainly produced during Base grids used to support work are expensive, and must be
rapid cooling during quenching, while transformation plastic replaced at routine intervals. To save money, maintenance of
strain is due to martensitic transformation. Both the strain these grids is delayed or neglected. As these grids are used,
distributions show a tendency of the mirror symmetry distortion of the grid occurs. Takahashi (23) investigated the
especially in circumferential and axial components. distortion of SCM420H helical gears with controlled
Liscic (20) indicated that there are three simultaneous process hardenability. He found that flatness of the gears decreased as
occurring during the heat treatment process. He also proposed a function of time of use of the base grid (Figure 26).
a new quenching technology of utilizing both a volume of gas, In another study, MacKenzie (24) used CFD and FEA to
and dispersed liquid nitrogen, to specifically control heat minimize the distortion of SAE 8620 pinion gears. The gears
extraction and heat transfer to minimize distortion and residual were originally rack in a pyramid on a grid and quenched in a
stresses. This method would be amenable to automated control. cold, medium speed oil. Examination of the distortion showed
Narazaki (21) predicted the heat transfer coefficients and that the pinions would bow along the shaft up to 2 mm. An
applied them to a steel helical gear. Using the Japanese silver alternative method of racking and quenching in mar-tempering
probe, an estimate of the heat transfer coefficient was oil was proposed and modeled. A CFD model was created for
determined by the lumped method. They found that the the entire quenching system and velocity vectors established.
difference in material between the silver probe and the steel These velocity vectors were applied as to the FEA model of
(SCr420H) gear affects the minimum temperature of the vapor each pinion gear as a boundary condition. A look-up table of
blanket phase. The high conductivity of the silver rod probe heat transfer coefficient as a function of velocity and surface
resulted in stabilizing the vapor phase, with the result of temperature was applied to the FEA model.
lowering the minimum temperature of vapor blanket.
Funatani (7) reviewed the effects of quenchant and quenchant
temperature on the distortion of SCM420 bearings 80 mm in
diameter by 44 mm tall. As the heat extraction speed increased,
hardness increased, but elliptical and cylindrical distortion also

400
 65 200 The steel making process and forming processes, with the
variables of grain size, grain flow, prior microstructure and
segregation can vastly effect ease of machining, as well as
60 150
Hardness, HRC

Distortion, µm
response to heat treatment.
While a great deal of attention has been made to control of
55 100 residual stress and distortion in heat treating and quenching,
these can be undermined by improper attention to the details of
50 50 grid and fixture life or the changes in quenching due to
oxidation of the quenchant.
45 0 It is an understanding of the entire process, and specific
60 100 150 200 operations and complexities that control residual stress and
Oil Temperature, °C distortion. These specific variables can change from machine
to machine, or from vendor to vendor. Only by extreme
Hardness Elliptical Distortion Cylindrical Distortion
attention to detail and attendant process variables can true
distortion control be achieved.
Figure 23. Comparison of quench oil temperature for distortion
5
control. SCM420 bearings (80 mm in diameter by 44 mm
thick), quenched into oils at different temperature. 4
3
0.25 2
1

Re (µm)
0
-5 -4 -3 -2 -1-1 0 1 2 3 4 5
-2
Re (µm)

0 -3
-0.25 0 0.25
-4
-5
Im (µm)

Two Point Racking Three Point Racking

-0.25
Im (µm)
Figure 25. Change in flatness (amplitude of dish slopes)
dependent on the type of racking support.
Two Point Racking Three Point Racking
120
Figure 24. Change in the amplitude of the outer radius
dependent on the method of loading. 100
Ring Gear Flatness, µm

The predicted bowing of the pinion was predicted to be less than 80

0.20 mm. Validation of the model showed that the model over-
predicted the distortion, with the actual bowing of the pinion 60
shaft averaging 0.12 mm. This reduction in distortion on this 40
single part resulted in a documented $17,000,000 savings in
rework, scrap, straightening and warranty issues. 20
Conclusions 0
The control of distortion and residual stress is a complex 0 2 4 6 8
undertaking. Great strides have been made in the modeling and Jig Useage, Months
simulation of machining, heat treatment and carburizing.
However the best modeling efforts can be undermined by Figure 26. Effect of time-in-use of base grids on the flatness of
variations in process. Dull tools, feeds and speeds and even SCM420H helical gears. Increased distortion and scatter
clamping methods result in changes and variations in residual increased as the in-use grid time increased.
stress and distortion.

401
 Acknowledgments 13. High Temperature Carurization - Influences on Distortion
Behaviour of Heavy-Duty Transmission Components. Kleff, J.,
I would like to thank the management of Houghton et al. Bremen, Germany : s.n., 2005. IDE 2005. pp. 227-234.
International, Inc., for allowing me to present this modest 14. Investigation of Distortion Behaviour of Machine
review at the 2016 ASM IFHTSE Advanced Thermal Components due to Carburizing and Quenching. Jurci, P., et
Processing World Congress. al. Berlin, Germany : s.n., 2007. IFHTSE 5th International
Quenching and Control of Distortion Conference. p. 39.
References 15. Influence of the Quenching Process on the Microstructure
1. Distortion Engineering. Thoben, K.-D., et al. 4, 2002, HTM, and Distortion of Heat Treated Parts with Particular Emphasis
Vol. 57, pp. 276-282. on Quench Oils. Braun, R. Bremen, Germany : s.n., 2007.
2. From Single Production Step to Entire Process Chain - the IFHTSE 5th International Conference on Quenching and
Global Approach of Distortion Engineering. Zoch, H-W. Control of Distortion. p. 47.
Bremen, Germany : s.n., 2005. IDE 2005. p. 3. 16. The Influence of Cooling Characteristics and Viscosity on
3. Metallurgical and production-related protocols to reduce the Hardening Behaviour of Steels. Tomita, Y., et al. Berlin,
heat-treatment distortion in the manufacture of gear Germany : s.n., 2007. IFHTSE 5th International Conference on
components. Hippenstiel, F. [ed.] H-W. Zoch. Bremen, QUenching and Control of Distortion. p. 54.
Germany : s.n., 2005. IDE 2005. pp. 57-64. 17. Modeling Residual Stresses in Spring Steel Quenching.
4. Comparison of different remelted hot-work tool steels and Sarmiento, S., et al. Indianapolis, IN : ASM International,
their distortion, residual stresses and other properties. 2001. Heat Treating 2001 - Including Quenching and
Experiments and Simulations. Schützenhöfer, W., et al. Distortion.
Bremen, Germany : s.n., 2005. IDE 2005. pp. 65-72. 18. Effect of Quenching Variables on Distortion and Residual
5. Controlling Distortion in Heat Treatment Through Press Stresses. MacKenzie, D. S. and Lambert, D. Indianapolis,
Quenching. Reardon, A. April 2015, Thermal Processing, pp. IN : ASM International, 2003. Proceedings of the 22nd Heat
24-29. Treating Society Conference and the 2nd International Surface
6. Metallurgical influence on distortion of the case-hardening Engineering Congress.
steel 20MnCr5. Prinz, C., et al. [ed.] T. Lubben H-W. Zoch. 19. Historical Review of Resdual Stress in Quenched Fe-Ni
Bremen, Germany : s.n., 2005. IDE 2005. pp. 75-82. Alloy Cylinders and Explaination of Its Origin Using Computer
7. Distortion control via optimization of cooling process and Simulation. Arimoto, K. and Funatani, K. Detroit, MI : ASM
improvement of quench oils. Funatani, K. Bremen, Germany : International, 2007. Proceedings of the 24th ASM Heat
s.n., 2008. IDE 2008. pp. 265-274. Treating Society Conference.
8. Effect of Alloy on the Distortion of Oil Quenched Automotive 20. History and Perspective of Controllable Heat Extration
Pinion Gears. MacKenzie, D. S. and Ferguson, B. L. During Quenching. Liscic, B. Indianapolis, IN : ASM
Shanghai, China : s.n., 2010. International Conference on International, 2001. ASM 2001 Heat Treating Society
Thermal Process Modeling and Computer Simulation. Conference.
9. Effect of Hardenability and Quenching on Distortion of Steel 21. Validation of Estimated Heat Transfer Coefficients during
Components. Stormvinter, A., et al. Bremen, Germany : s.n., Quenching of Steel Gear. Narazaki, M., et al. Berlin,
2015. IDE 2015. pp. 39-46. Germany : s.n., 2007. IFHTSE 5th International Quenching and
10. Effect of Machining Parameters and Clamping Technique Control of Distortion Conference.
on the Residual Stresses and Distortion of Bearing Rings. 22. Identification of Process Parameters affecting Distortion of
Nowag, L., et al. Bremen, Germany : s.n., 2005. IDE 2005. Disks for Gear Manufacture - Part II: Heating, Carburizing
11. Distortion Engineering in Turning Processes with Standard and Quenching. Clausen, B., et al. Bremen, Germany : s.n.,
Clamping Systems. Grote, C., Brinksmeier, E. and 2008. IDE 2008. pp. 41-50.
Garbrecht, M. Bremen, Germany : s.n., 2008. IDE 2008. 23. Takahashi, A., Morishima, T. and Yamada, H. 6, 1990,
12. Influence of Carburizing on Distortion Behaviour. Journal of the Japanses Heat Treating Society, Vol. 30, p. 301.
Bahnsen, C., et al. Bremen, Germany : s.n., 2005. IDE 2005. 24. Control of Residual Stress and Distortion. MacKenzie, D.
pp. 235-242. S. s.l. : ASM International, July 2007, Heat Treating Progress,
p. 47.

402