Professional Documents
Culture Documents
MPC Journal
Copy of e-mail Notification
does not include the final edits. Such preprints may be posted as electronic files on the Authors own website for
personal or professional use, or on the Authors internal university or corporate networks/intranet, or secure
external website at the Authors institution, but not for commercial sale or for any systematic external distribution by
a third party (eg: a listserver or database connected to a public access server). Prior to publication, the Author
must include the following notice on the preprint: This is a preprint of an article accepted for publication in (Journal
Title Copyright @ (year) (copyright ASTM International, West Conshohocken, PA)10.1520/MPC20140003. After
publication of the Work by ASTM International, the preprint notice should be amended to read as follows: This is a
preprint of an article published in (include the complete citation information for the final version of the Work as
published in the print edition of the Journal) and should provide an electronic link to the Journals WWW site,
located at the following ASTM URL: http://www.astm.org/10.1520/MPC20140003. The author agrees not to update
the preprint or replace it with the published version of the Work.
Sincerely,
On behalf of ASTM International
Denise Stanley, Journal Production Manager
E-mail: Denise.Stanley@cenveo.com
R e p r i n t s O r d e r F or m
For reprints of articles and/or papers from ASTM Journals,
Selected Technical Papers (STPs), book chapters and
Standardization News.
Note: Authors receive a 25% discount
Ship To:
Name: ___________________________________________________________________________________
Organization: _____________________________________________________________________________
Address: _________________________________________________________________________________
City: _________________________________________________State: _________________ Zip: _________
Country: _____________________________________________
Bill To:
Name: ___________________________________________________________________________________
Organization: _____________________________________________________________________________
Address: _________________________________________________________________________________
City: __________________________________________________State: ________________ Zip: _________
Country: ______________________________________________
Date:
Phone:
email:
Fax:
(check one)
100
200
500
1000
2000
$1,102
$1,323
$1,543
$1,764
$1,984
3-4
$2,116
$2,221
$2,326
$2,425
$2,824
5-8
$2,168
$2,278
$2,394
$2,499
$3,024
9-12
$3,307
$3,470
$3,638
$3,801
$4,126
Pages
25
50
100
200
300
500
100 0
2000
1-4
$136
$152
$168
$184
$199
$226
$310
$472
5-8
$195
$226
$257
$278
$299
$352
$499
$709
9-12
$225
$257
$289
$336
$357
$425
$577
$787
13-16
$263
$294
$325
$394
$430
$520
$751
$1,118
17-20
$278
$336
$394
$472
$514
$635
$887
$1,307
21-24
$301
$373
$446
$535
$583
$724
$1,003
$1,475
25-28
$331
$404
$478
$583
$646
$808
$1,165
$1,795
29-32
$383
$441
$499
$651
$724
$913
$1,323
$2,058
Cover (optional)
$100
$115
$136
$157
$178
$220
$373
$598
Pages
ASTM International Customer Relations department will finalize your reprint quote including
freight charges. ASTM International will contact you with the quote price.
10/02/12
Article Number:
MPC20140003
Dear Author,
Below are the queries associated with your article; please answer all of these queries before sending the proof
back to Cenveo. Author please indicate the correct color processing option from the list below:
1. Author, please confirm Figure number(s) that should appear as color in print. Please know that any
associated mandatory fees will apply for figures printed in color.
2. Author, please confirm Figure number(s) that should appear as color online only, there will be no fees
applied.
3. Author, your paper currently does not include any color figures for online or print. If color is needed please
indicate which figures it should be applied to and whether it is color in print or online.
Location in
article
AQ1
AQ2
DOI: 10.1520/MPC20140003
Reference
QuenchProbeAn
Alternative Probe Design
for In-Situ Estimation of
Cooling Rates, Heat Flux,
and Hardenability During
Immersion Quenching of
Hardenable Steels
VOL. 3 / NO. 4 / MONTH 2014
Stage:
Page: 1
Total Pages: 26
doi:10.1520/MPC20140003
Vol. 3
No. 4
Month 2014
Reference QuenchProbeAn
Alternative Probe Design for In-Situ
Estimation of Cooling Rates, Heat Flux,
and Hardenability During Immersion
Quenching of Hardenable Steels
Reference
Kumar, T. S. Prasanna, Hernandez-Morales, B., and Totten, G. E., Reference
QuenchProbeAn Alternative Probe Design for In-Situ Estimation of Cooling Rates, Heat
Flux, and Hardenability During Immersion Quenching of Hardenable Steels, Materials
Performance and Characterization, Vol. 3, No. 4, 2014, pp. 124, doi:10.1520/
MPC20140003. ISSN 2165-3992
ABSTRACT
Manuscript received January 9,
2014; accepted for publication
March 7, 2014; published online xx
xx xxxx.
1
This article reviews some of the limitations of the standard cooling curve
analysis and the Jominy hardenability test in extending the results to actual
quenching, which can be used by the heat treater in the plant. The specimen is
made of the same grade of steel as the quenched component with section
laboratory tests. The test is carried out in the plant under actual conditions
10
dispensing the need to correlate the standard cooling curve data and end
11
12
13
14
15
Using the cooling data at the point of measurement, the cooling rates,
16
17
C 2014 by ASTM International, 100 Barr Harbor Drive, P.O. Box C700, West Conshohocken, PA 19428-2959
Copyright V
Stage:
Page: 2
Total Pages: 26
18
austenite decomposition models for handling the latent heat liberated during
19
quenching. Several steels ranging from low carbon to medium alloy steels were
20
21
22
23
values. The Reference QuenchProbe is thus shown to generate data needed for
24
25
26
Introduction
27
There are two important techniques for obtaining quantitative information to assist
the heat treater in achieving fully hardened steel during quenching viz., the cooling
curve and the Jominy hardenability test. Both tests have been around for nearly 100
years. A cooling curve is the graph obtained by plotting the measured temperature
of a specimen against time at a known location (usually at the geometric center of
the specimen) when the specimen is quenched in a uid medium [1]. Hardicke [2,3]
reported rst published examples of cooling curve characterization of various
quenching media using Le Chateliers 18 mm diameter by 18 mm cylindrical iron
bar probe with a mass of 37 g and a Pt/Pt-Rh thermocouple inserted to the geometric
center. Data acquisition was performed by connecting the thermocouple to a mirror
galvanometer which intermittently monitored the image of the oscillating pendulum
from the galvanometer on to photographic lm.
There are a number of variants of the specimen shape, size and materials,
though the preferred shape is cylindrical. Many of these have been reviewed previously [46]. In addition to ISO 9950 [7], several standards have been developed
including ASTM D6200-01 [8], ASTM D6482-06 [9], and ASTM D6549-06 [10],
which are all based on a 12.5 mm diameter by 60 mm cylindrical Inconel 600 probe
with a Type K thermocouple inserted at the geometric center. Other standard probes
are constructed using silver with a thermocouple at the geometric center such as
ASTM D7646-10 [11] or with the thermocouple positioned at the surface such as
that described in Method A of JIS K 2242 or method which utilized a silver probe
with a center thermocouple [12]. Currently, there is considerable interest in the
standardization and use of a larger Liscic/Petrofer 50 mm diameter by 200 mm cylindrical Inconel 600 probe with three thermocouples; one is placed at 1 mm below the
surface, another 4 mm below the surface, and the third at the center mid-length of
the probe which is proposed for use to characterize the heat transfer properties steel
materials under plant quenching conditions [13,14]. These probes are used in commercially available quenchant testing equipment that utilize electronic data acquisition systems and numerical analysis to provide an insight into the nature of heat
transfer conditions throughout the quenching process using gas, vaporizable liquids,
and molten salts.
The second tool refers to the hardenability information which is obtained in the
laboratory by end-quenching a cylindrical steel piece of a specic grade as described
Materials Performance and Characterization
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Stage:
Page: 3
Total Pages: 26
in ASTM A255-10 [15]. The hardness along the length of the specimen, which is a
function of the cooling rate at the location due to end quenching, is measured after
the test is completed. The distance from the quenched-end which gives a hardness
corresponding to 50 % martensite is taken as the hardenability of the steel grade.
The specimen size and the equipment are standardized to ensure identical quenching severity. Since the austenite decomposition depends on the cooling rate and the
metallurgical characteristics of the steel, the hardenability will reect the ease with
which martensitic structure could be obtained in practice. The test gives an indication of the comparative ability of different steels to undergo martensitic
transformation.
This article addresses some of the limitations of the standard cooling curve analysis procedures and the Jominy hardenability test in extending the analysis to actual
quenching in industrial setups and how by suitable design of a new system, the limitations can be overcome.
ASTM D6200 covers acquisition and interpretation of cooling curves for oil
quenching [8]. An Inconel probe of standard dimensions is heated to the austenitizing temperature and quenched in a vessel containing a specied amount of the oil
sample. The temperature at the center of the specimen is measured by a thermocouple and parameters such as maximum cooling rate, critical times, etc., computed
from the cooling curve thus obtained. The test parameters serve to compare the performance of a given oil with that of another one, a control sample, or previous performance of the same oil. The test will show the effect of oxidation, the presence of
additives and their concentrations, or contamination on the cooling characteristics
of the quenching oil. It is also known that different grades of steels yield different
hardness values under identical quenching conditions. As such, the cooling curves
obtained from the standard Inconel probe or similar ones using a material different
from the material undergoing heat treatment cannot be used as a guideline for
achieving the desired hardness in the part of interest.
The parameters that affect the cooling rates during quenching include agitation
of the quenchant, thermo-physical-chemical properties of the quenchant such as viscosity, ashpoint, composition, etc [16]. Both the latent heat liberated during phase
transformation and the diffusion of heat in the solid also affect the overall heat
transfer process during quenching. Therefore it becomes imperative to develop techniques for estimating the cooling rates, heat transfer coefcients, and other quality
control parameters under plant conditions specic to the process on hand. Cheng
et al. [17] showed that the heat transfer coefcient had a peak value and its corresponding temperature interval varied with the cross section dimension of the work
piece, quenchant type, and axial position of the specimen being quenched. Sedighi
and McMahon [18] studied the effect of quenchant circulation on the heat transfer
rate during quenching of steels. Maniruzzaman et al. [19] studied the effect of agitation on the quench performance of a mineral oil based quenchant by computing the
effective heat transfer coefcients as a function of surface temperature. The effect of
section size on the surface temperature and heat transfer coefcient during quenching of steel cylinders were studied by Heming et al. [20] and Woodard et al. [21].
The effect of start temperature on boiling water heat transfer was studied during
spray quenching and a method was proposed to predict the boiling curves under different start temperatures by Li et al. [22].The effect of initial soaking temperature of
Materials Performance and Characterization
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
Stage:
Page: 4
Total Pages: 26
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
Stage:
Page: 5
Total Pages: 26
TABLE 1
A comparison between the ASTM D6200 [8] quench probe and the Reference QuenchProbe.
Sl
Number
Test
Reference QuenchProbe
(This article)
Effect of viscosity,
oil contamination,
oxidation, usage factor etc
Microstructure and
hardness; hardenability
prediction/measurement
Not possible
Effect of agitation
Mathematical Models
The estimation of quench parameters such as heat ux, cooling rate, hardness and
microstructure distribution during quenching of an alloy steel specimen is based on
mathematical models which have been published elsewhere [3133]. The procedure
is based on coupling inverse heat conduction analysis with austenite decomposition
Materials Performance and Characterization
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
Stage:
Page: 6
Total Pages: 26
FIG. 1
The hand held Reference
QuenchProbe showing the
probe attached at one end,
with a thermocouple xed to
the probe for recording the
time temperature data during
quenching(left). Portable
electric resistance furnace for
heating the specimen to the
soaking temperature before
quenching the specimen in the
tank (right).
models. The input to the model is the time-temperature curve measured at a known
position in the specimen. The mathematical model combines the composition
specic TTT diagram of the steel and empirical hardness models based on local
cooling rate.
The heat transfer phenomena on the surface of the component being quenched
in a uid is quite complex, with lm boiling followed by nucleate boiling occurring
at high and medium surface temperatures; at lower surface temperatures, the heat
transfer takes place by pure convection through a boundary layer. A review of methods used for analyzing heat transfer during quenching has been published recently
[34]. Boundary heat ux estimation during any thermal process is an ill-posed
inverse heat conduction problem (IHCP). When a solid body is subjected to heating
or cooling due to a uid, the heat transfer mechanism at the solid/uid boundary
can be extremely complex. Particularly with reference to immersion quenching,
there are different regimes operating at distinct temperature ranges. The IHCP simplies the problem by analyzing the result of any such complex causes in that it
is based on the solution to a specic time-temperature history at a given location
inside the solid body. Prasanna Kumar [35] developed an algorithm for estimating
the unknown boundary heat ux, which was shown to be applicable under diverse
situations [32,33,3537]. A detailed validation of the IHCP algorithm for a 2D
quenching case can be found in Ref. [38].
Estimation of heat transfer coefcients during quenching requires the accurate
measurement of surface temperature also. Due to experimental difculties it is common practice to embed a thermocouple very near the surface. It is quite well known
that the temperature gradients near the surface will be quite steep for high heat
transfer processes like quenching which makes such surface temperature measurements unacceptable. The specialized LiscicNanmac probe also has limitations as
the thermal eld around the point of measurement is inuenced by the construction
of the probe [39]. The inverse algorithm overcomes these difculties mathematically.
Materials Performance and Characterization
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
Stage:
Page: 7
Total Pages: 26
FIG. 2
The solution domain in
cylindrical axi-symmetric
coordinates showing an
example of the FE mesh.
The solution gives both the unknown heat ux at the surface and the thermal eld
as functions of time at all nodes simultaneously. Thus, the transient surface temperature can be accurately computed, even with a single thermocouple measurement
inside the specimen.
During alloy steel quenching the steels undergo phase change at various rates
which requires concurrent solutions to the two phenomena of heat transfer and austenite decomposition. The details of combining the IHCP with the TTT diagram of
the steel of interest can be found in Refs. [35,36].
AN IMPLICIT ENTHALPY BASED INVERSE HEAT CONDUCTION MODEL
201
202
203
204
205
206
207
208
209
The heat transfer in the specimen during immersion quenching was modeled in the
cylindrical coordinate system, assuming symmetry about the longitudinal axis, with
the heat ux at the boundary treated as an unknown quantity. Graded, 4-node isoparametric toroidal elements were used for meshing the middle 10 mm of the specimen, where the thermocouple tip was positioned, at 4 mm depth from the specimen
surface. An example of the FE mesh is shown in Fig. 2.
Initial efforts in modeling the IHCP during quenching were based on treating
the latent heat as a source term [35,36]. It was observed that handling the source
term could lead to instabilities; therefore, the present work incorporates the latent
heat through the implicit non-linear enthalpy model due to Swaminathan and Voller
[40]. Application of a semi-implicit analysis for simulation of freezing of water can
be found in [41], which was extended in this work with a fully implicit formulation
(see the Appendix).
The heat transfer in the specimen during quenching was modeled using the heat
conduction governing equation written in terms of enthalpy:
@
@T
@
@T
@H
k
(1)
@r
@r
@z
@z
@t
210
225
Tr; z; t Tsoak
at t 0
211
212
213
214
215
216
217
218
219
220
221
222
223
224
Stage:
Page: 8
Total Pages: 26
226
@T
0 on S2
k
@z
t 0 < t tend
where:
Tsoak the soaking temperature,
q(t) the unknown heat ux at the boundary during quenching, and
Tmeas (t) the measured thermal response at the thermocouple position dened
by the coordinates (r0,z0).
The unknown heat ux is vectorized into qi; i 1, 2,, m,, n. For a given calculation, the heat ux is treated as constant over a small time interval, Dt. The
inverse algorithm is to estimate qm assuming qi; i 1,, m 1 are known (already
determined). The mth time step is the current time step. Similar to heat ux components, the continuous thermal histories measured during experiments are discretized
and denoted by Yj(i); j 1, 2,, s and i 1, 2,, n, where s is the number of temperature measurement locations, which in the present case is equal to one. Following
Ref. [21], the objective function for minimization is chosen as
(2)
s X
r
X
227
^
Yj;mi1 T
j;mi1
228
229
230
231
232
233
234
235
236
237
238
239
240
2
j1 i1
241
242
s X
r
X
^j;mi1 /j;i
Yj;mi1 T
(3)
Dqm
j1 i1
s X
r
X
/j;i
2
j1 i1
244
q
qm1
^ Tj;i m;;mr1
T
j;i
q
0
m;;mr1 1 eqm1
^j;i , is dened by
and the temperature corresponding to the base ux, T
(6)
243
245
^j;i Tj;i qm;;mr1 qm1
T
The volumetric enthalpy term, H, was modelled as the sum of the enthalpies of
the mixture of phases present at any time step as follows:
Materials Performance and Characterization
246
247
Stage:
Page: 9
Total Pages: 26
FIG. 3
A typical graph of enthalpies of
the initial (austenite) and a
product phase (ferrite) for an
alloy steel undergoing phase
change [42].
(7)
Hi
X T
T0
where the superscript refers to the transformation product, T0 is any reference temperature and the subscript tr refers to the fraction transformed.
In the context of modeling alloy solidication [40,41], the enthalpy term would
explicitly involve the latent heat term at temperatures up to solidus. Generally, the
solidication range is determined based on composition and the corresponding
phase diagram with the latent heat modeled as a function of fraction solid. With
solid state transformation, as in alloy steel quenching, the transformation beginning
and end can extend over different ranges depending on cooling rates. The transformation fractions can also vary widely.
As the austenite starts decomposing into other phases, the net enthalpy of the
mixture decreases since the enthalpy of any phase resulting from the decomposition
is smaller than that of austenite at that temperature. The enthalpies of different
phases as a function of temperature are shown in Fig. 3, taken from JMatPro [42],
for a typical steel. Starting from fully austenitic phase with the highest enthalpy, the
enthalpy of the mixture always diminishes, thus accounting for latent heat
liberation.
FINITE ELEMENT FORMULATION
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
268
(9)
Hnm1 Hnm
dHnm m1
Tn Tnm
dT
266
267
269
Stage:
Page: 10
Total Pages: 26
10
where:
n the time step, and
m the iteration within the time step.
The global set of equations now take the form for an iteration within the time
step
K m
n
270
271
272
273
274
m
m
1
dH
1
dH
m
m
f
T
g
f
H
g
f
H
g
fT gm1
fF gm
C
C
n
n
n
n
n
Dt
dT n
Dt
dT n
(10)
The matrix [K] and the force vector {F} are computed for each itreation using
the converged values of temperatures at the previous iteration. The iterations were
continued until the end condition dened by the tolerance dened by Eq 11 was
achieved.
!
fH gm
fH gm1
n
n
abs
(11)
< 0:001
fH gm1
n
The heat transfer is coupled with austenite decomposition through the enthalpy
term appearing in Eq 1, which requires the modeling of austenite decomposition. As
a rst step towards this aim, the TTT diagram of the steel was digitized and used in
the software for computation of phase transformation. A schematic of the methodology adopted for tracking the austenite decomposition using well-known formulae
for diffusional [43] and non-diffusional [44] transformations is given in Fig. 4.
Scheils additivity rule was invoked for marking the beginning of measurable
amounts of fraction transformed (0.01), given by
(12)
m
X
Dtj
1
t
Tj
s
j1
FIG. 4
Schematic representation of
the technique applied to obtain
the CCT diagram from the
composition specic TTT
diagram of the steel and
measured/computed cooling
curves.
275
276
277
278
279
280
281
282
283
284
285
286
287
Stage:
Page: 11
Total Pages: 26
11
where m gave the starting point for the computation of real fraction of transformed
product starting from the mth time step, the normalized fraction product was
computed as the ratio of the real fraction transformed until that time step and the
equilibrium fraction of that product at that temperature.
i
Xnorm
Tm
(13)
(14)
291
292
293
294
1
i
ln 1 Xnorm
Tm nTm1
bTm1
i
Xnorm
Tm1
1 exp bTm1
i
teq
Dt
295
296
nTm1
From which, the real fraction of the ith phase forming at the temperature step is
calculated using
(16)
290
The new normalized fraction at the end of the present time step is then calculated by
(15)
289
i
Xreal
Tm1
i
Xeqbrm
T m
i
teq
288
297
298
i
i
i
Tm1 Xnorm
Tm1 Xeqbrm
Tm1
Xreal
CRITICAL TEMPERATURES
299
For computing austenite decomposition into its product phases by the aforementioned approach, two important pieces of information are necessary: (i) the critical
temperatures Ae3, Ae1 associated with the TTT diagram, and (ii) those associated
with the equilibrium diagram. It is observed that the critical temperatures obtained
from these two approaches never match. However, for estimating austenite decomposition, it is important to ensure consistency in the critical temperatures obtained
from the TTT diagram and the FeC equilibrium diagram, as the kinetic parameters
in the JMAK equations are linked to the equilibrium phase quantities read from the
FeC equilibrium diagram. To ensure this, the procedure adopted by Prasanna
Kumar [23] for dening the phase boundaries in the FeC equilibrium diagram was
used.
300
MATERIAL PROPERTIES
311
The true fractions of all the phases were summed and the remaining austenite fraction was computed for each element at the end of each time step. The thermal conductivity and enthalpy properties were computed as sums of the contributions of the
individual fractions of phases present in the element at the time step. The properties
of individual phases were taken from the commercial software JMatPro as explained
elsewhere [36].
312
301
302
303
304
305
306
307
308
309
310
313
314
315
316
317
Stage:
Page: 12
Total Pages: 26
12
HARDNESS MODELING
318
In general, a mixture rule based on the amounts of the phases formed and their
hardness is used for predicting the macro hardness of the quenched steel. Maynier
et al. [45] developed regression equations for the hardness of various phases as function of the chemistry and cooling rates at 700 C. However, their equations depended
only on the nal amounts of the phases. In this work, the hardness w of the phases
that are forming isothermally at each time step during transformation of austenite
was taken from the TTT diagrams for the respective steel grades. A step-wise linear
model was developed from the data as a function of temperature. The hardness due
to continuous cooling was written based on the rule of mixtures as
319
(17)
we
XX
j
320
321
322
323
324
325
326
327
wij DXji
where the hardness of the fractional ith phase at the jth time step was summed continuously to get the macro hardness of the eth element.
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
AND CORE
346
The Reference QuenchProbe is used to obtain the cooling curves at the surface and
the core of the specimen which can reveal the differential cooling rates across the
section as well as the surface phenomena. Figure 6 shows the cooling curves obtained
by measurement as well as by computation in a 41Cr4 steel quenched in a polymer
solution under agitation.
The surface temperature brings out clearly the latent heat effects due to martensitic transformation at around 300 C, at which temperature the heat removal rates
of the uid is quite low resulting in slower cooling rates. The cooling rate as a function of the local temperature is shown in Fig. 7. The onset of the convective regime
in the quenchant starts at 300 C.
Materials Performance and Characterization
347
348
349
350
351
352
353
354
355
356
Stage:
Page: 13
Total Pages: 26
13
FIG. 5
Cooling curves measured at
4 mm depth from the surface of
a cylindrical specimen of
different grades of steel
quenched in different media.
The surface heat ux computed during quenching of three different grades of steel
samples in four types of quenchants shown in Fig. 8 reveals the very complex nature
of heat transfer in the uid media. The surface heat ux during quenching of alloy
steels in agitated tanks is affected by many factors which cannot be duplicated in the
laboratory. It depends on thermal diffusivity of the steel, the thermal dissipation
mechanism on the uid side, the austenite decomposition rates, intensity of agitation, etc. It is therefore obvious that the heat ux is specic to the steel, quenchant,
and plant conditions. This is well brought out in Figs. 8(a)8(c). It must be noted,
however, that each case refers to different plants.
Figure 8(a) shows the heat ux obtained during quenching of C45 and 41Cr4
grades of steel in similar quenchants (13 % aqueous solution of a polymer) in different plants. The maximum heat ux is around 1 MW/m2 in both the cases, but the
occurrence of peak heat ux at different temperatures brings out the differences in
FIG. 6
A typical set of cooling curves
obtained during quenching of a
41Cr4 grade steel in a 6 %
aqueous polymer solution
during an industrial trial.
357
358
359
360
361
362
363
364
365
366
367
368
369
370
Stage:
Page: 14
Total Pages: 26
14
FIG. 7
Computed cooling rates at the
surface and core corresponding
to the data shown in Fig. 6.
the steel being quenched. In the case of 52 100 and SUP 9 grades quenched in oil
(Fig. 8(b)), the peak heat ux varies from 0.8 to 1.4 MW/m2, while the peak occurs at
about the same temperature. The inuence of lower concentrations of polymer on
peak heat ux is evident in the case of 41Cr4 grade of steel (Fig. 8(c)).
SURFACE COOLING RATES
The surface cooling rates computed simultaneously with the heat ux are shown in
Figs. 9(a)9(c) for the same cases as shown in Figs. 8(a) and 8(c). Similar observations can be made with reference to the cooling rates, as with heat ux. The maximum cooling rate was around 150 C/s for 41Cr4 quenched in 6 % polymer solution,
while it was around 60 C/s when quenched in 13 % polymer solution. While oil
quenching showed a clear convective regime below 420 C, with low cooling rates,
such distinction could not be made with other quenchants.
The occurrence of peak cooling rate during quenching will affect the products
of transformation and is connected with the TTT diagram, which is brought out in
Fig. 9(a), while in the case of C45 steel, the peak occurs at a higher temperature
(780 C compared to 580 C in the case of 41Cr4 steel) thus promoting ferrite/pearlite. In the case of oil quenching of 52 100 and SUP9 grades, the trends were similar,
with the peak ranging from 58 to 98 C/s. The different regimes of heat transfer were
unaffected by the steel grades studied. Similar observations can be made in the case
of 41 Cr4 steel quenched in two different concentrations of polymer solutions, with
the lower concentration quenchant giving higher cooling rate of 150 compared to
60 C/s with higher concentration quenchant.
HARDNESS ESTIMATION
Apart from the immersion quench tests with several steels and quenchant combinations, three grades of steels, En19, C45, and 8627 H, were end quenched to introduce
a large range of cooling rates in the same specimen. The temperature history at a
point 5 mm from the quenched end was recorded during end quenching and used as
input to the model. Using the mathematical models described in this paper, the heat
Materials Performance and Characterization
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
Stage:
Page: 15
Total Pages: 26
FIG. 8
(a).Computed heat ux at the
surface during quenching of
C45 and 41Cr4 grades of steel
quenched in 13 % polymer
solutions. (b).Computed heat
ux at the surface during
quenching of 52 100 and SUP 9
grades of steel quenched in oil.
(c).Computed heat ux at the
surface during quenching of
41Cr4 grades of steel quenched
in aqueous polymer solutions of
different concentrations.
15
Stage:
Page: 16
Total Pages: 26
FIG. 9
(a).Computed cooling rates at
the surface during quenching
of C45 and 41Cr4 grades of
steel quenched in 13 % polymer
solutions. (b).Computed
cooling rates at the surface
during quenching of 52 100 and
SUP 9 grades of steel
quenched in oil. (c).Computed
cooling rates at the surface
during quenching of 41Cr4
grades of steel quenched in
aqueous polymer solutions of
different concentrations.
16
Stage:
Page: 17
Total Pages: 26
17
FIG. 10
Measured and estimated
hardness of several grades of
steel after end quenching
(EndQ) and immersion
quenching in several media.
ux due to water impingement at the end of the cylindrical specimen was computed
along with the cooling rates at several locations along the length of the specimen
and the hardness. The results of the computed and measured hardness in both
the immersion quenched and end quenched samples are shown in Fig. 10. The
computed hardness along the length of the end quenched specimens and across the
section of immersion quenched specimen were shown to be in good agreement with
the measured values which validates the procedure.
ERROR ESTIMATE
Like any numerical technique based on the nite-element method, the accuracy of
the calculation depends both on the domain discretization and time step size. An
estimate of the error in the computation during quenching of a C45 steel specimen
in mineral oil is given in Fig. 11 for the discretization scheme shown in Fig. 2 and two
time steps, viz., 0.5 and 0.1 s. The maximum percentage error in the estimated
temperature at the thermocouple location coincided with the occurrence of the
FIG. 11
Error in the estimation of
recorded temperatures at the
thermocouple location for
different time step values
during quenching of a C45
specimen in mineral oil.
399
400
401
402
403
404
405
406
407
408
409
410
411
412
Stage:
Page: 18
Total Pages: 26
18
maximum heat ux in all the cases. It was seen that the maximum error could be
brought down to less than 0.5 % with a time step of 0.1 s, while the value was nearly
2.25 % for a time step of 0.5 s. Considering that the calculations were CPU intensive,
a time step of 0.5 s was chosen for all cases.
Summary
Appendix
415
416
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
The phase change during quenching of hardenable steels involves latent heat liberation, which will have to be considered for quenching simulation. The implicit
enthalpy method, mentioned briey in the section entitled An Implicit Enthalpy
Based Inverse Heat Conduction Model is a very powerful algorithm for taking into
account the liberation of latent heat during quenching. The enthalpy formulation
considers the temperature dependency of all the properties including thermal conductivity. The nite element formulation of the scheme is therefore elaborated
here.
The partial differential equation describing heat transfer during alloy steel
quenching is written in cylindrical axi-symmetric coordinates as
(A1)
414
417
A portable Reference QuenchProbe using specimens of the same grade of steel as the
component of interest has been developed for in situ testing of cooling rates and
hardenability during immersion quenching of hardenable grades of steels. Using the
time-temperature data recorded during quenching of the sample, the surface cooling
rates at the surface and the surface heat ux density during quenching were calculated by combining Inverse Heat Transfer and austenite decomposition models.
Concurrently with the surface conditions, the cooling rate, microstructure, and hardness variations form the surface to the core were calculated. Deploying the Reference
QuenchProbe in industrial heat treatment plants, tests done for several alloy steels
during quenching in several industrial quenchants, it was shown that the Reference
QuenchProbe could be effectively used for:
413
@
@T
@
@T
@T
Q qc
k
k
in X
@r
@r
@z
@z
@t
443
444
445
446
447
448
449
450
451
452
Stage:
Page: 19
Total Pages: 26
19
453
T Ti r; z at t 0
and the boundary conditions
@T
@T
k
nr k
nz qr; z
@r
@z
454
where k, q, and c are the thermal conductivity, density, and specic heat of the material, which are functions of the phase and temperature T. The term Q represents the
rate of heat generation due to latent heat release. In what can be considered as the
most general and straight forward treatment of latent heat, the source term in Eq A1
is written as
Q qL
(A2)
457
458
459
460
461
462
463
@
@T
@
@T
@ftr @T
k
k
q cL
@T @t
@r
@r
@z
@z
456
@ftr
@ftr @T
qL
@t
@T @t
where:
L the latent heat of transformation of austenite, and
ftr the fraction of austenite transformed into different phases.
We can now write the energy balance equation as:
@
@T
@
@T
@ftr @T
@T
qL
(A3)
k
k
qc
@T @t
@r
@r
@z
@z
@t
(A4)
455
Hi
X T
T0
464
where:
superscript i the transformation product,
subscript fr the transformed amount, and
T0 any reference temperature.
Differentiating the above equation with respect to T, we obtain
@H
@ftr
(A6)
q cL
@T
@T
465
@x
@x
@y
@y
@t
470
471
The nite element formulation of the above equation (Eq A7) can be conveniently
obtained from the Galerkin formulation. The distribution of the dependent variable,
472
466
467
468
469
473
Stage:
Page: 20
Total Pages: 26
20
s
X
474
475
wi r; zTi t
i1
where:
s the number of nodes assigned to element e, and
Ti the nodal temperatures.
wi are the interpolation or shape functions, the form of which is governed by
the order of the element. With the assumed distribution of the temperature over the
element and following the standard Galerkin procedure, the discretized element
equation is obtained in the form
Ce H_ K e fT ge fF ge
(A8)
476
483
Cij
Kij
Xe
Xe
481
482
484
qwi dC
The set of rst order ordinary differential equations in Eq A8 is rendered into a set
of linear algebraic equations by applying weighted average approximations to the
differentiated quantity. The interpolation is linear, dened by
485
486
487
fH gn1 fH gn
h H_ n1 1 h H_ n
Dt
where the subscript n refers to the time step number. The choice of the value of h
leads us to the various difference schemes. With h 0 and observing that [C] is
independent of temperature, we get the implicit scheme, i.e.,
488
489
490
K n1 fT gen1 C H_ n1 fF gn1
491
492
fH gn1 fH gn
fF gn1
Dt
The unknown enthalpy term in Eq A11 is dened by its Taylor series expansion
with iterations within the time step:
(A12)
480
wi wj dX
C1
(A11)
479
@wi @wj
@wi @wj
k
k
dX
@r @r
@z @z
Fi
(A10)
478
and
(A9)
477
Hnm1 Hnm
dHnm m1
Tn Tnm
dT
493
494
Stage:
Page: 21
Total Pages: 26
21
Hnm1 Hnm
(A13)
dHnm m1 dHnm m
T
T
dT n
dT n
With this substitution in Eq A10, the global set of equations now take the form for
iteration within the time step:
K m
n
495
496
m
m
1
dH
1
dH
m
m
fF gm
H
H
C
C
fT gm1
f
g
f
g
f
g
n
n
n
n
n
Dt
dT n
Dt
dT n
(A14)
The matrix [K] and the force vector {F} are computed for each iteration using the
converged values of temperatures at the previous iteration. The iterations are continued until the end condition dened by the tolerance.
(A15)
497
498
499
fH gm1
fH gm
n
n
abs
< e
fH gm
n
500
References
501
[1] Totten, G. E., Dossett, J. L., and Kobasko, N. I., Quenching of Steel, ASM
Handbook, Vol. 4A, Steel Heat Treating Fundamentals and Processes, J. L.
Dossett and G. E. Totten, Eds., ASM International, Materials Park, OH, 2013,
pp. 91157.
[2] Le Chatelier, H., Etudes sur la Trempe de Lacier, Rev. Metall., Vol. 1, No. 9,
1904, pp. 473492.
[3] Haedicke, S., Le Chateliers Harteversuche, Stahl Eisen, Vol. 24, No. 21, 1904,
pp. 12391244.
[4] Totten, G. E., Bates, C. E., and Clinton, N. A., Chapter 3Cooling Curve
Analysis, Handbook of Quenchants and Quenching Technology, ASM International, Materials Park, OH, 1993, pp. 69128.
[5] Canale, L., Luo, X., Yao, X., and Totten, G., Quenchant Characterization by
Cooling Curve Analysis, J. ASTM Int., Vol. 6, No. 2, 2009, 101981.
[6] Hernandez-Morales, B., Vergara-Hernandez, H. J., Solorio-Daz, G., and
Totten, G., Experimental and Computational Study of Heat Transfer During
Quenching of Metallic Probes, Evaporatation, Condensation and Heat Transfer, A. Ahsan, Ed., InTech, Rijeka, Croatia, 2011, pp. 4972.
[7] ISO 9950: Industrial Quenching OilsDetermination of Cooling
CharacteristicsNickel-Alloy Probe Test Method, International Organization
for Standardization, Geneva, Switzerland, 1995.
[8] ASTM D6200-01: Standard Test Method for Determination of Cooling Characteristics of Quench Oils by Cooling Curve Analysis, Annual Book of ASTM
Standards, ASTM International, West Conshohocken, PA, 2012.
[9] ASTM D6482-06: Standard Test Method for Determination of Cooling
Characteristics of Aqueous Polymer Quenchants by Cooling Curve Analysis
with Agitation (Tensi Method), Annual Book of ASTM Standards, ASTM International, West Conshohocken, PA, 2012., 2011.
Materials Performance and Characterization
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
Stage:
Page: 22
Total Pages: 26
[10] ASTM D6549-06: Standard Test Method for Determination of Cooling Characteristics of Quenchants by Cooling Curve Analysis with Agitation (Drayton
Unit), Annual Book of ASTM Standards, ASTM International, West Conshohocken, PA, 2012, 2011.
[11] ASTM D7646-10: Standard Test Method for Determination of Cooling Characteristics of Aqueous Polymer Quenchants for Aluminum Alloys by Cooling
Curve Analysis, Annual Book of ASTM Standards, ASTM International, West
Conshohocken, PA, 2010.
[12] JIS K2242: Heat Treat Fluids, Japanese Standards Association, 4-1-24, Akasaka,
Minato-ku, Tokyo 107-8440, Japan, 2012.
[13] Liscic, B. and Filetin, T., Measurement of Quenching Intensity, Calculation of
Heat Transfer Coefcient and Global Database of Liquid Quenchants, MEMI
J., Vol. 19, 2012, pp. 5263.
[14] Liscic, B. and Singer, S., Large Probes for Characterization of Industrial
Quenching Processes, ASM Handbook, Vol. 4A, Steel Heat Treating
Fundamentals and Processes, J. Dossett and G. Totten, Eds., ASM International,
Materials Park, OH, 2013, pp. 176191.
[15] ASTM A255-10: Standard Test Methods for Determining Hardenability of
Steel, Annual Book of ASTM Standards, ASTM International, West Conshohocken, PA, 2010.
[16] ASTM D6710-02: Standard Guide for Evaluation of Hydrocarbon-Based
Quench Oil, Annual Book of ASTM Standards, ASTM International, West
Conshohocken, PA, 2012.
[17] Cheng, M., Cheng, J., Yuan, S. Q., and Zhao, F., Studies on the Heat Transfer
Behaviour of the Interface between Workpiece and Media in the Quenching
Process, Acta Metall. Sinica, Vol. 10, 1997, pp. 479484.
[18] Sedighi, M. and McMahon, C. A., The Inuence of Quenchant Agitation
on the Heat Transfer Coefcient and Residual Stress Development in the
Quenching of Steels, Proc. IMechE Part B, Vol. 214, 2000, pp. 555567.
[19] Maniruzzaman, M., Stratton, M. B., Rogers, T. A., Barber, L. P., and Sisson,
R. D., Jr., Effect of Agitation on the Quenching Performance of a Quench Oil
using CHTE Quench Probe System, Proceedings of the MS&T 2003 Conference, Chicago, IL, Nov. 912, 2003, pp. 297312.
[20] Heming, C., Xieqing, H., and Jianbin, X., Comparison of Surface Heat-transfer
Coefcients Between Various Diameter Cylinders During Rapid Cooling,
J. Mater. Process. Technol., Vol. 138, 2003, pp. 399402.
[21] Woodard, P. R., Chandrasekar, S., and Yang, H. T. Y., Analysis of Temperature and Microstructure in the Quenching of Steel Cylinders, Metall. Mater.
Trans., Vol. 30B, 1999, pp. 815822.
[22] Li, D., Wells, M. A., Cockcroft, S. L., and Caron, E., Effect of Sample Start
Temperature During Transient Boiling Water Heat Transfer, Metall. Mater.
Trans., Vol. 38B, 2007, pp. 901910.
[23] Babu, K. and Prasanna Kumar, T. S., Mathematical Modeling of Surface Heat
Flux During Quenching,Metall. Mater. Trans., Vol 41B, 2010, pp. 214224.
[24] Huiping, L., Guoqun, Z., Shanting, N., and Chuanzhen, H., FEM Simulation
of Quenching Process and Experimental Verication of Simulation Results,
Mater. Sci. Eng. A, Vol. 452453, 2007, pp. 705714.
Materials Performance and Characterization
22
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
Stage:
Page: 23
Total Pages: 26
23
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
Stage:
Page: 24
Total Pages: 26
[39]
[40]
[41]
[42]
[43]
[44]
[45]
24
623
624
AQ1
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
AQ2