Professional Documents
Culture Documents
• For online submission please insert your corrections in the online correction form. Always
indicate the line number to which the correction refers.
• You can also insert your corrections in the proof PDF and email the annotated PDF.
• For fax submission, please ensure that your corrections are clearly legible. Use a fine black
pen and write the correction in the margin, not too close to the edge of the page.
• Remember to note the journal title, article number, and your name when sending your
response via e-mail or fax.
• Check the metadata sheet to make sure that the header information, especially author names
and the corresponding affiliations are correctly shown.
• Check the questions that may have arisen during copy editing and insert your answers/
corrections.
• Check that the text is complete and that all figures, tables and their legends are included. Also
check the accuracy of special characters, equations, and electronic supplementary material if
applicable. If necessary refer to the Edited manuscript.
• The publication of inaccurate data such as dosages and units can have serious consequences.
Please take particular care that all such details are correct.
• Please do not make changes that involve only matters of style. We have generally introduced
forms that follow the journal’s style.
Substantial changes in content, e.g., new results, corrected values, title and authorship are not
allowed without the approval of the responsible editor. In such a case, please contact the
Editorial Office and return his/her consent together with the proof.
• If we do not receive your corrections within 48 hours, we will send you a reminder.
• Your article will be published Online First approximately one week after receipt of your
corrected proofs. This is the official first publication citable with the DOI. Further changes
are, therefore, not possible.
Please note
After online publication, subscribers (personal/institutional) to this journal will have access to the
complete article via the DOI using the URL: http://dx.doi.org/[DOI].
If you would like to know when your article has been published online, take advantage of our free
alert service. For registration and further information go to: http://www.link.springer.com.
Due to the electronic nature of the procedure, the manuscript and the original figures will only be
returned to you on special request. When you return your corrections, please inform us if you would
like to have these documents returned.
Metadata of the article that will be visualized in
OnlineFirst
ArticleTitle Nucleation and Growth of Graphite in Eutectic Spheroidal Cast Iron: Modeling and Testing
Article Sub-Title
Article CopyRight The Minerals, Metals & Materials Society and ASM International
(This will be the copyright line in the final PDF)
Journal Name Metallurgical and Materials Transactions A
Corresponding Author Family Name Carazo
Particle
Given Name Fernando D.
Suffix
Division Instituto de Mecánica Aplicada
Organization Universidad Nacional de San Juan
Address Av. Libertador Gral. San Martín 1109 (Oeste), C.P. J5400ARL, San Juan,
Argentina
Division
Organization CONICET
Address Godoy Cruz 2290, CABA, C1425FQB, Argentina
Email fcarazo@unsj.edu.ar
Received
Schedule Revised
Accepted
Abstract A new model of graphite growth during the continuous cooling of eutectic spheroidal cast iron is presented
in this paper. The model considers the nucleation and growth of graphite from pouring to room
temperature. The microstructural model of solidification accounts for the eutectic as divorced and graphite
growth rate as a function of carbon gradient at the liquid in contact with the graphite. In the solid state, the
microstructural model takes into account three stages for graphite growth, namely (1) from the end of
solidification to the upper bound of stable eutectoid intercritical, (2) during the stable eutectoid
intercritical, and (3) from the lower bound of stable eutectoid intercritical to room temperature. The micro-
and macrostructural models are coupled using a sequential multiscale approach. Numerical results for
graphite fraction and size distribution are compared with experimental results obtained from a cylindrical
cup, in which the graphite volumetric fraction and size distribution were obtained using the Schwartz–
Saltykov approach. The agreements between the experimental and numerical results for the fraction of
graphite and the size distribution of spheroids reveal the importance of numerical models in the prediction
of the main aspects of graphite in spheroidal cast iron.
Footnote Information Manuscript submitted June 9, 2015.
Journal: 11661
Article: 3430
Please ensure you fill out your response to the queries raised below
and return this form along with your corrections
Dear Author
During the process of typesetting your article, the following queries have arisen. Please
check your typeset proof carefully against the queries listed below and mark the
necessary changes either directly on the proof/online grid or in the ‘Author’s response’
area provided below
F
11 eutectic as divorced and graphite growth rate as a function of carbon gradient at the liquid in
12 contact with the graphite. In the solid state, the microstructural model takes into account three
O
13 stages for graphite growth, namely (1) from the end of solidification to the upper bound of
14 stable eutectoid intercritical, (2) during the stable eutectoid intercritical, and (3) from the lower
15 bound of stable eutectoid intercritical to room temperature. The micro- and macrostructural
O
16 models are coupled using a sequential multiscale approach. Numerical results for graphite
Author Proof
17 fraction and size distribution are compared with experimental results obtained from a
18 cylindrical cup, in which the graphite volumetric fraction and size distribution were obtained
PR
19 using the Schwartz–Saltykov approach. The agreements between the experimental and
20 numerical results for the fraction of graphite and the size distribution of spheroids reveal the
21 importance of numerical models in the prediction of the main aspects of graphite in spheroidal
22 cast iron.
23
D
24 DOI: 10.1007/s11661-016-3430-x
26
25 Ó The Minerals, Metals & Materials Society and ASM International 2016
27
28
I. INTRODUCTION
NODULAR cast irons are alloys in which the main
TE
to monitor the evolution of these three characteristics
during and at the end of the cooling process.
Studies dealing with the thermo-metallurgical evolu-
40
41
42
EC
29 components are Fe, C, and Si. From a technological tion of the complete cooling process of spheroidal 43
30 perspective, the quality and mechanical properties of eutectic cast iron[2–6] have limited microstructural capa- 44
31 nodular cast irons depend on the type and characteris- bilities to model phase changes. During solidification, 45
32 tics of (1) graphite spheroids and (2) metallic they consider a non-divorced eutectic and do not take 46
33 constituents.
RR
FERNANDO D. CARAZOM, Adjoint Professor and Researcher, graphite spheroids is limited to carbon diffusion through
is with the Instituto de Mecánica Aplicada, Universidad Nacional de the shell of solid solution of carbon in Fea (BCC iron), 58
San Juan, Av. Libertador Gral. San Martı́n 1109 (Oeste), C.P. namely ferrite, towards the spheroids. 59
J5400ARL, San Juan, Argentina, is also with the CONICET,
Venugopalan studied the growth of graphite spher- 60
C1425FQB Godoy Cruz 2290, CABA Argentina. Contact e-mail:
fcarazo@unsj.edu.ar PATRICIA M. DARDATI, Professor, is with oids during the continuous cooling of spheroidal cast 61
the GIDMA and Departamento de Ingenierı́a Mecánica, Universidad iron.[7] The same author modeled the growth of graphite 62
Tecnológica Nacional Facultad Regional Córdoba, Maestro M. López spheroids in an isothermal process.[8] In both works, 63
esq. Cruz Roja Argentina s/n, Ciudad Universitaria, Córdoba, there is a simplified representation of the graphite 64
Argentina. DIEGO J. CELENTANO, Associate Professor, is with
the Departamento de Ingenierı́a Mecánica y Metalúrgica, Research growth. 65
Center for Nanotechnology and Advanced Materials (CIEN-UC), The simulation of the complete cooling of an eutectic 66
Pontificia Universidad Católica de Chile, Av. Vicuña Mackenna 4860, spheroidal cast iron was presented by Wessen and 67
Macul, Santiago, Chile. LUIS A. GODOY, Professor and Researcher, Svensson.[9] In this work, the authors take into account 68
is with the Institute for Advanced Studies in Engineering and
Technology, FCEFyN, National University of Cordoba and CON-
the growth of spheroids from the end of solidification to 69
ICET, Av. Vélez Sarsfield 1601, Córdoba, Argentina, is also with the the beginning of the stable eutectoid transformation by 70
CONICET. means of CALPHADÒ. 71
Manuscript submitted June 9, 2015.
F
83 graphite growth with microstructural solidification this research is based on two different but related scales. 139
84 models. This is a severe limitation because in any cast At a macrostructural level, the equation of energy 140
O
85 alloy the phase changes in the solid state depend on the conservation is solved to obtain the temperature (T) and 141
86 microstructural characteristics at the end of the cooling rate (T)_ fields. These variables are subsequently 142
87 solidification. employed as data at the microscale model, in which 143
O
88 This short literature review attempts to highlight some phase changes during solidification and in the solid state 144
89 aspects of the current state of the art: are formulated and solved at a representative volume 145
Author Proof
PR
90 Conservation of carbon mass at a microstructural
91 level has not been included in existing models. This is along the cooling process are as follows: (1) the 147
92 a crucial aspect to evaluate carbon gradient and temperature is constant at each time step and its value 148
93 consequently diffusion towards both spheroids and is given from the macroscale problem solution; (2) the 149
94 austenite located far from the interphase with carbon mass conservation is satisfied; and (3) there is 150
95 graphite. equilibrium at the interphases. 151
With T and Fe-C-Si and Fe-Fe3C-Si systems (see 152
D
96 Growth of spheroids in contact with liquid has not
97 been considered for equiaxial dendritic morpholo- Appendix A), it is possible to obtain the gradients of 153
98 gies which are typical in austenite for any spheroidal carbon concentration at different interphases, the vari-
TE 154
99 cast iron compositions. Furthermore, microsegrega- ables required to evaluate the growth rate of phases and 155
100 tions in the liquid state and their effect on the microconstituents, their volume fractions, and the 156
101 thermodynamics and kinetics of the solid-state phase energy released due to the latent heat during the phase 157
102 transformations are not currently modeled. These changes. This energy released is part of the macrostruc- 158
tural formulation for the solution of the energy 159
EC
108 current models account for neither the carbon theories based on thermodynamics and physical metal- 164
109 diffusion from austenite at the interphase with lurgy. Their results are taken to the macrostructural 165
110 graphite towards the graphite spheroids, nor the level in terms of rule of mixtures,[14] written as a 166
111 stable eutectoid intercritical. For temperatures lower function of the volume fraction of microconstituents 167
112 than the lower bound of the stable eutectoid inter- and their respective latent heat values. 168
CO
113 critical, the present models only represent the carbon In Carazo et al.,[15] there may be seen a scheme of the 169
114 diffusion towards the spheroids through the ferrite relation between the thermal and microstructural fields 170
115 envelope. in a phase-change problem solved by the finite element 171
method, whereas the coupling scheme of the 172
116 This paper presents a new model for nucleation and thermo-metallurgical problem is written in terms of the 173
UN
117 growth of graphite spheroids in an eutectic cast iron. rate of the element phase-change vector: 174
118 The model aims to Z
ðeÞ
119 model growth of spheroidal graphite spheroids from L_ PC ¼ NT qLf_PC dX; ½1
120 the end of solidification to the end of austenite Xe
121 transformation (solid solution of C in Fec or FCC
122 iron), where NT is the transpose of the shape function matrix, 176
123 couple the growth of graphite spheroids in solid state q is the density, L is the phase-change latent heat, Xe is 177
124 with a microstructural model of solidification (in- the integration domain (corresponding to a finite 178
125 cluding nucleation and growth of graphite spheroids element), and f_PC is the time derivative of the 179
126 during solidification), and phase-change function which is obtained from ad hoc 180
127 provide a phenomenological description of the phase transformation models. In the macroscopic met- 181
128 growth process of graphite spheroids in solid state, allurgical phase-change models, fPC is an explicit func- 182
129 together with its implications for austenite transfor- tion of T. On the contrary, in the metallurgical 183
130 mations. phase-change models defined from a microstructural 184
@ ðAami ami Þ _
Z
ðeÞ
L_ PC ¼ NT qL TdX; B. Microstructural Model 249
F
@T
Xe Most metallurgical research at micro level dealing 250
O
with solidification of spheroidal cast irons considers a 251
195 where Aami and ami are the set of state variables and the non-divorced eutectic with cooperative growth. The 252
196 corresponding parameters of the phase-change models underlying assumption is that the spheroids that nucle- 253
197 (e.g., solidification and transformations in solid state).
O
ate are instantly surrounded by austenite.[2–4,18,20,25–43] 254
198 Thus, the phase-change effects described at the micro- However, according to the process described in Sec- 255
Author Proof
199 scopic level can be taken into account at the macroscopic tion III–A, the thermodynamics and kinetics of the 256
PR
200 level in the energy balance as a variable source term.[13] eutectic transformation would not be related to a 257
non-divorced eutectic. In agreement with this line of 258
thought, the phase-change model should account for 259
201 III. SOLIDIFICATION OF EUTECTIC independent nucleation of austenite and graphite in the 260
202 SPHEROIDAL CAST IRON liquid and, in addition, independent equiaxial dendritic 261
growth of austenite and spherical growth of graphite in 262
D
203 A. Phenomenological Approach at the Microstructural the liquid and in contact with the austenite. 263
204 Level The main features of the microstructural model for 264
TE
205 In spheroidal cast irons, the eutectic transformation solidification of an eutectic spheroidal cast iron adopted 265
206 develops according to a divorced eutectic, also known as in this work are presented below. In this model, 266
207 anomalous.[16,17] This is a difference with flake cast irons originally proposed by Dardati,[11] a divorced eutectic 267
208 in which the eutectic transformation occurs as a regular is considered. A scheme of a two-dimensional grain of 268
209 eutectic which is characteristic of a faceted–faceted austenite, together with graphite spheroids and carbon 269
EC
210 morphology, i.e., flake graphite–austenite. concentration profile, is shown in Figure 1. 270
211 In spheroidal cast irons, eutectic crystallization is
212 produced from the independent nucleation of graphite
213 spheroids and equiaxed austenite dendrites at different
214 points, undercooling, and time stages.[18–22] There have
RR
F
282 changes in which nucleation of austenite and graphite 320
283 are as follows: (1) austenite and graphite volume spheroids takes place allows incorporating the charac- 321
O
284 fractions, (2) number and size distribution of graphite teristic incubation time of phase changes out of equilib- 322
285 spheroids; (3) alloy composition of the first and last zone rium. In this case, the carbon concentrations are 323
286 that become solid; and (4) grain size of austenite. obtained by means of a procedure proposed by Heine[45] 324
O
287 According to Figure 2, which shows a schematic part which is based, for temperatures lower than the eutectic 325
288 of a pseudo-binary Fe-C-Si system, the composition of one, on a metastable extrapolation reported by 326
Author Proof
289 the alloy considered in this study varies along the three Hultgren.[46] 327
PR
290 zones z1, z2, and z3 in Figure 1. The main features of the
291 nucleation and growth models of graphite spheroids
292 during an eutectic solidification are presented in Sec-
293 tions III–C and III–D. Details of such models may be IV. GROWTH OF GRAPHITE SPHEROIDS 328
294 found in Dardati et al.[44] FROM THE END OF SOLIDIFICATION UP TO 329
THE UPPER BOUND OF THE STABLE 330
D
EUTECTOID INTERCRITICAL 331
295 C. Nucleation of Graphite Spheroids
A. General Considerations 332
296
297
As mentioned above, the nucleation of graphite
spheroids is modeled as a continuous process. The
TE
For a temperature T lower than the eutectic temper- 333
298 process starts when the alloy temperature reaches the ature (TE0 ) and higher than the upper bound of the 334
299 eutectic temperature (point C¢ in Figure 2), it stops stable eutectoid intercritical (TaaT ), the equilibrium car- 335
300 under recalescence and, finally, it restarts if the temper- bon concentration in austenite at the interphase with 336
EC
301 ature is lower than the lowest temperature at which the c=g
graphite (CC ) is shown in Figure 2. The carbon 337
302 process had previously stopped, before the end of
concentration profile for such temperature range and a 338
303 solidification (z2 and z3 in Figure 1). In this context,
quarter of a graphite spheroid surrounded by austenite 339
304 nucleation is assumed to occur in the interdendritic and
are both depicted in Figure 3. If the temperature 340
305 intergranular zones (zones z2 and z3, respectively).
RR
Fig. 2—Schematic of an isopleth Fe-C section of the Fe-C-Si equilib- Fig. 3—Schematic representations of one quarter of a graphite
rium phase diagram with the composition of interest for a tempera- spheroid surrounded by austenite and the associated carbon profile
ture TaaT T TE0 . for a temperature between TE0 and TaaT .
F
The growth of graphite spheroids during the stable eu- 399
@CcC
1 c
/C ¼ DC qc ; ½2 tectoid phase change can be considered in two steps that 400
O
@r r¼Rg are separately described below. 401
356 where DcC is the diffusion coefficient of carbon in
O
357 austenite, qc is the density of austenite, and A. Growth of Graphite Spheroids During the 402
c Stable Eutectoid Intercritical 403
Author Proof
@CC
358 @r r¼Rg is the gradient of carbon in austenite at
PR
Figure 4 shows a schematic Fe-C section of the 404
359 the interphase with graphite. Carbon that diffuses from Fe-C-Si equilibrium phase diagram for a temperature 405
360 austenite to the graphite spheroids is incorporated into between the upper and lower bounds of the stable eu- 406
361 the latter as tectoid intercritical. 407
When the temperature is between the upper and lower 408
c=g
/1C ¼ qg Cg qc CC R_ g ; ½3
bounds of the stable eutectoid intercritical (TaaT and TaA1 , 409
D
363 where Cg and qg are the carbon concentration and respectively), the growth of graphite spheroids is due to 410
the carbon diffusion caused by the difference between 411
364 density of graphite, respectively, and R_ g is the rate of c=g c=g
the values of CcC and CC , as long as CC CcC <0 (see
TE
365 change of graphite spheroids radius (i.e., the rate at 412
366 which the graphite/austenite interphase advances Figure 3). If the temperature reaches TaaT , the ferrite 413
367 towards austenite). From Eqs. [2] and [3], it results grains can nucleate on the surface of the spheroids in 414
such a way that the graphite is covered by ferrite grains 415
@Cc
1 depending on its size and on the number and size of 416
R_ g ¼ DcC qc C
EC
½4
qg Cg qc CC
c=g @r r¼Rg ferrite nuclei that nucleate on the surface of each 417
spheroid. 418
@CcC
Unlike what was discussed in Section IV, carbon 419
369
371 where @r r¼Rg is the only unknown in the right-
370 diffusion could take place now through ferrite and 420
austenite. The growth rate of graphite spheroids in the 421
RR
c=g
@CcC
CcC CC
¼ : ½5
@r r¼Rg Rg
383
384 Substituting Eq. [5] into Eq. [4], the radius of graphite
385 spheroids rate from the end of solidification until the
386 upper bound of stable eutectoid intercritical is given by
c=g
D c CcC CC
R_ g ¼ C : ½6
Rg q =q Cg Cc=g
g c C Fig. 4—Schematic of an isopleth Fe-C section of the Fe-C-Si equilib-
rium phase diagram with the indication of stable and
388 metastable equilibrium carbon concentrations (solid and broken
389 The growth of spheroids occurs at the expense of lines, respectively) for a temperature between the upper and lower
390 carbon diffusion from the austenite, so that the carbon bounds of the stable eutectoid intercritical.
metastable equilibrium carbon concentrations (solid and broken the carbon distribution during graphite growth for a temperature
PR
lines, respectively) for a temperature lower than the lower bound of lower than the lower bound of the stable eutectoid intercritical (see
the stable eutectoid intercritical. Fig. 5).
a=c
424 contact with austenite (CC ) is lower than the equilibrium grow as a function of carbon diffusion to the graphite 457
425 carbon concentration in ferrite in contact with graphite through ferrite if Ac=g ¼ 0 (flow /4C in Figure 6) and 458
a=g
426 (CC ). This may be seen in Figure 5 which shows a through ferrite and austenite if Ac=g 6¼ 0 (flows /4C and /1C 459
D
427 schematic Fe-C section of the Fe-C-Si equilibrium phase shown in Figures 3 and 6, respectively). 460
428 diagram for a temperature lower than the lower bound of TE
429 the stable eutectoid intercritical, by extrapolation of the 1. Growth rate of graphite spheroids completely 461
a=g covered by ferrite 462
430 line of CC at temperatures higher than the lower bound
431 of the stable eutectoid intercritical. In a graphite spheroid that has been completely 463
432 Thus, the radius of graphite spheroids rate is given by covered by ferrite grains that nucleated on its surface, 464
the flow of carbon allowing for the growth of graphite 465
EC
c CcC CC
c=g spheroids is given by 466
D
R_ g ¼ C Ac=g ½7 @CaC
Rg q =q Cg Cc=g 4 a
/C ¼ DC qa ; ½8
g c C
@r r¼Rg
RR
434 where the coefficient Ac=g takes into account the fraction
435 of the spheroid surface in contact with austenite, i.e., where qa is the density of ferrite, DaC is the diffusion 468
a
436 0 Ac=g 1, such that Ac=g ¼ 1 for the start of the coefficient of carbon in ferrite, and @C C @r is the
r¼Rg 469
437 transformation, while Ac=g ¼ 0 is achieved when the
gradient of carbon concentration in ferrite in contact 470
438 spheroid is completely covered by ferrite grains, and with graphite (straight line at r = Rg in Figure 6). 471
CO
439 Ac=g is calculated as proposed in Appendix B. In Eq. [7], Under the same assumptions of Section IV–B, in this 472
440 a spheroid that is fully covered by ferrite (Ac=g ¼ 0) a=g a=c
case the gradient of Eq. [8] depends on CC and CC 473
441 means that it is not in contact with austenite and this
442 (see Figure 6) and, therefore, the growth rate of the 474
implies that there is no carbon diffusion from such phase
c=g graphite spheroid radius results in 475
443 to the graphite spheroids. The value of CC in Eq. [7] is
UN
444 obtained by extrapolation of the line of maximum a=c a=g
qc CC CC R
445 solubility of carbon in austenite for temperatures lower R_ g ¼ DaC a : ½9
446 than TaaT , depending on the values of silicon concentra- qa Cg Ca=g Rg Ra Rg
C
447 tion in austenite in contact with graphite CSi or in the
448 first zone of solidification (see Appendix A). The carbon 477
449 profile considered in this growth step is shown in
450 Figure 6. 2. Growth rate of graphite spheroids partially covered 478
by ferrite grains 479
If a graphite spheroid has not been completely 480
451 B. Growth of Graphite Spheroids at Temperatures Lower covered by ferrite grains that nucleated on its surface, 481
452 than the Lower Bound of the Stable Eutectoid the flow of carbon to the spheroid occurs through ferrite 482
453 Intercritical and austenite. In this case, the growth rate of the 483
454 As the ferrite grains have nucleated on the spheroids, graphite spheroid radius may be written as 484
455 when the temperature is lower than the lower bound of the
R_ g ¼ R_ g c Ac=g þ R_ g a Aa=g ;
456 stable eutectoid intercritical (see Figure 5), the spheroids ½10
F
five zones shown in Figure 7(c) were analyzed and 533
497 PROCEDURES processed using the software ImageJ.[49,50] Three impor- 534
O
498 A. Experiments tant aspects were considered during these stages: 535
499 The alloy employed in the tests was molten in a 1. The results obtained were limited to objects for 536
which the aspect ratio was larger than 0.9. 537
O
500 high-frequency induction furnace with 1500 kg capacity.
501 Its composition was approximately 23.3 pct of SAE 2. The minimum diameter to consider an object as a 538
Author Proof
502 1010 scraps, 23.3 pct of retrieved nodular cast iron, graphite spheroid was defined as 6 lm. 539
PR
503 6.6 pct of pig iron, and 41.8 pct of charcoal. To obtain 3. The size of each pixel was 0.3478 lm. 540
504 the adequate carbon content, 1.6 pct of carbon (with a For the five zones shown in Figure 7(c), the following 541
505 90 pct of performance), 2 pct of steel plates, and procedure was considered: 542
506 0.15 pct of SiCa and thick Fe-Si (with 75 pct of Si)
507 were added to adjust the Si composition of the base 1. Micrographs without chemical etching were used to 543
508 metal in the molten furnace. The base metal was analyze and characterize graphite spheroids. 544
D
509 overheated to 1923 K (1650 °C) during approximately 2. Micrographs with chemical etching were used to 545
510 20 minutes where 1.5 pct Fe-Si-Mg-Ce was used as a quantify the microconstituents corresponding to the 546
metal matrix. 547
TE
511 nodularizing agent.
512 Inoculation and nodularization were done by the The characterization of the graphite spheroids 548
513 sandwich method. In the reaction ladle, 0.7 pct of fine included the following: 549
514 FeSi (with 75 pct of Si) was added. The cast metal was
515 poured into the cast ladle to fill the cups shown in 1. Quantification of the graphite fraction per unit 550
EC
516 Figures 7(a) and (b). Five cups typically used to area. 551
517 determine the carbon equivalent were also employed in 2. Measurement of the area of each object. This 552
518 this study. Table I shows the main components of the allowed estimating the diameter of each sphere as 553
519 alloy used in the experiments (in weight percentage). equivalent to the identified object by application of 554
520 The thermal history was recorded by means of cooling the criterion to determine the equivalent diameter of 555
RR
521 and cooling rate curves at the central zone of the part a sphere of the same projected area. 556
Table I. Average Chemical Composition (Main Elements) of the Molten Alloy (Values in Wt. Pct)
CO
Element C Si Mn S P Cr Cu Sn Mg CE
Wt. Pct 3.67 2.8 0.21 0.01 0.038 0.025 0.01 0.0009 0.052 4.61
UN
Fig. 7—Cylindrical cup used to determine carbon equivalent and specimen poured used in the laboratory experiments. (a) Top view in which the
bifilar ceramic and the thermocouple covered by refractory cement can be observed. (b) Longitudinal midplane section view. (c) Longitudinal
midplane section view of the specimen with its dimensions (in mm) and the indication of five microstructurally characterized zones.
F
567 All the values mentioned were obtained from the size simulate the heat flow between the part and the cup, 604
568 distribution of spheroids per unit area and volume (with whereas surface elements were considered to deal with 605
O
569 the exception of the graphite fraction, which only is the heat extraction through convection in the external 606
570 obtained from the size distribution of spheroids per unit surface of the part and the mold in contact with the 607
571 volume). ambient temperature. 608
O
572 A brief description of the methodology used to obtain All the thermo-physical properties and material 609
573 the number of spheroids per unit volume as a function parameters used in the numerical simulations are given 610
Author Proof
PR
575 For each micrography, once the largest diameter of an
576 object (dmax ) was identified and the desired number of
577 classes to investigate the size distribution was selected VII. RESULTS AND DISCUSSION 612
578 (nclasses ), the interval D between one diameter di 1 and
the next di is D ¼ dmax=nclasses , obtaining Table II. A. Numerical Results 613
579
D
580 Known D, the diameter of spheroids, the number of Numerical cooling and cooling rate curves computed 614
581 classes, and the area of the micrographs, the size at the nodes indicated in Figure 8(b) are, respectively, 615
582 distribution of spheroids per unit area of the corre- plotted in Figures 9 and 10. Both plots also show an
TE 616
583 sponding micrography area was calculated (NgrAi ). This enlargement of the areas of interest. Table III summa- 617
584 allowed the calculation of the number of graphite rizes the solidification time values. From Figure 9, there 618
585 spheroids per unit volume (NgrVi ) according to is a relation between the local time of solidification (tf ) 619
indicated in Table III and the plateau of the character- 620
586 Schwartz–Saltykov approach.[51]
EC
F
633 show the highest recalescence values in both cases are seems to be justified because a higher cooling rate 659
634 those placed in the central region of the specimen (i.e., increases the undercooling and, thus, the driving force 660
O
635 nodes identified as HC, HM, and HE) which is, in fact, for graphite nucleation. The evacuation of the energy 661
636 the hot spot as the cooling process progresses. There- released during the phase change (due to both latent 662
637 fore, the cooling rate (see Figure 10), and consequently heats of solidification of graphite and austenite) is 663
O
638 the latent heat released during the phase changes, higher in regions of the part with higher cooling rate, 664
639 becomes slower in the central zone of the casting. with the consequence that the criterion established to 665
Author Proof
640 The above thermal consideration is relevant because end the nucleation (presence of recalescence) is less 666
PR
641 of the influence of temperature on diffusive phase likely to occur in these regions than in the others with 667
642 changes. The driving force for nucleation increases with lower cooling rate. The plateau which is observed at the 668
643 over-cooling, which increases with the thermal cooling
644 rate, and stops at recalescence. The growth rate of
645 spheroids increases with atomic mobility, i.e., higher
646 temperatures (we consider the influence of Si on the
D
647 thermodynamics of phase changes only, not in its
eutectoid zones.
UN
F
680 is lower than that in a liquid state, and decreases with cation and the start of the stable eutectoid phase change 720
681 decreasing temperature and value of the diffusion in terms of the elapsed time for this stage are presented 721
O
682 coefficient of carbon in austenite. in Figure 14, together with the trend line and associated 722
683 Finally, during the stable and metastable eutectoid mean value as a function of time. 723
684 phase changes in the third step, the growth rate of the It is seen that there is a direct relation between the 724
O
685 graphite spheroids increases with respect to that at the final graphite fraction that has been transformed during 725
686 end of the second step and, as already discussed in this stage and elapsed time of this stage. The mean value 726
Author Proof
PR
688 The graphite fraction, its trend in terms of the cooling
689 rate at the initial instants of solidification, and its mean
690 value (see Table III) for the nodes indicated in
691 Figure 8(b) are shown in Figure 12.
692 The trend (a straight line obtained by a least square
693 approximation) shows that by increasing cooling rates
D
694 the graphite fraction also increases. The value of the
695 correlation coefficient ( 0.87) highlights this fact. The TE
696 horizontal line, on the other hand, is associated with the
697 mean graphite fraction value of 11.73 pct.
698 Figure 13 shows percentage of the final graphite
699 fraction in terms of the cooling rate transformed at the
700 final instants of solidification, from the end of solidifi-
EC
706 solidification up to the upper bound of stable eutectoid Fig. 13—Percentage of the final graphite fraction as a function of
the cooling rate transformed at the final instants of the three stages
707 intercritical, an increasing trend is predicted. From a of graphite growth for the nodes shown in Fig. 8(b).
708 phenomenological point of view, this could be due to the
CO
UN
per unit volume as a function of their sizes for nodes BE, TM, and Fig. 16—Comparison of simulated and experimental cooling curves
HC (according to the classes listed in Table II). at the central zone of the specimen with magnification of solidifica-
PR
tion and eutectoid zones (zone 5 in Fig. 7(c) and node HC in
Fig. 8(b)).
727 is 18.6 pct of the final fraction, which is very close to the
728 value of 19 pct indicated in Figure 13.
729 Discrete and continuous distributions of graphite
730 spheroid quantities as a function of their size (as
D
731 specified in Table II) are shown in Figure 15 for nodes
732 BE, TM, and HC identified in Figure 8(b), which are
733 those exhibiting, as shown in Table III, the highest,
TE
734 mean, and the lowest cooling rates, respectively.
735 There is a similar distribution for the three nodes,
736 with a tendency of peaks to shift towards larger
737 diameters when cooling rate, at the start of solidifica-
738 tion, increases, and consequently graphite volumetric
EC
739 fraction also increases. The peak of the continuous
740 distribution tends to be higher and to shift towards
741 smaller diameters as the cooling rate at the start of
742 solidification is lower. The shift of the peak in the
RR
751 Computed cooling curves and cooling rate curves at volume, as obtained from experiments and simulation at 769
752 the central region of the sample are, respectively, zone 5 in Figure 7(c) and node HC in Figure 8(b), are 770
753 compared in Figures 16 and 17 with those correspond- illustrated in Figure 18 following the definition of 771
754 ing to three experiments. classes and diameters listed in Table II. The three 772
755 The complete curve in Figure 16 shows good agree- discrete distributions and their interpolated continuous 773
756 ment between experiments and computations. As shown forms in Figure 19 show similar trends, but the peak in 774
757 in Figure 17, there is also a good prediction of the local the simulation is shifted towards classes with larger 775
758 and global solidification times, evidenced at the charac- diameters. 776
759 teristic zones in Figure 17. Regarding the stable and Figure 20 shows discrete distributions of graphite 777
760 metastable eutectoid phase changes, the small change in fractions from experiments and simulations at the 778
761 slope in the numerical cooling rate curves shows that the central region, whereas continuous distributions are 779
762 stable eutectoid transformation starts at 468 seconds. plotted in Figure 21. The experiments show three peaks, 780
763 Although there are differences between cooling curves in while the simulations show only two. The peak in the 781
764 the eutectoid region, the starting time of the stable eu- experiments associated with large diameters occurs 782
765 tectoid phase change in the simulation is close to that of because of primary spheroids (a hypereutectic 783
PR
ing to the classes listed in Table II) at the central zone of the speci-
men.
D
TE
EC
RR
F
Regarding the stage of growth of spheroids from the 832
end of solidification up to a temperature TaaT , this new 833
O
model allows understanding the phenomenon that takes 834
place during the phase changes in thermo-metallurgical 835
processes with cooling interruptions, such as continu- 836
O
ous-isothermal-continuous cooling processes or heat 837
Fig. 23—Experimental and simulated continuous accumulated gra-
treatment processes. An interruption during a cooling 838
Author Proof
PR
austenite away from the interphase with graphite, thus 840
promoting the formation of cementite or highly dis- 841
torted austenite by the presence of carbon atoms 842
dissolved in an FCC iron among other problems. A 843
similar analysis may be conducted for isothermal 844
processes in the stable eutectoid intercritical. 845
D
The importance of the characteristics of thermal 846
evolution of a part or of locations of a part having 847
TE
different cooling rates was highlighted. If the cooling 848
rate becomes high, the carbon content of austenite at the 849
interphase with graphite at the start of the stable eutec- 850
toid phase change will be low; this is due to the longer 851
time that carbon has to diffuse towards the spheroids, 852
EC
thus reducing carbon in the solid solution in austenite, 853
decreasing the driving force for carbon diffusion during 854
this stage. Low cooling rate, on the other hand, may 855
cause or promote carbon accumulation in austenite at 856
RR
802 VIII. CONCLUSIONS dependence, and the influence that a phase change has 862
on the subsequent ones. 863
803 A new thermo-metallurgical model has been reported The theoretical developments, as well as the compu- 864
804 in this paper, which is fully coupled from the point of tational results, represent advances with respect to the 865
805 view of micro-modeling and takes into account nucle- present state of the art in models, theoretical formula- 866
806 ation and growth of graphite spheroids along the entire 867
UN
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
TaaT ¼ 1:10 1011 þ 2:74 1035 ðCSi Þ2 3:25 1035 CSi þ 1:46 1034 þ 1:74 1023 CSi 1:02 1023 =8:53 1019
F
O
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
TaA1 ¼ 3:47 1011 þ 1:60 1022 ðCSi Þ2 2:19 1021 CSi þ 5:85 1019 þ 7:4 1016 CSi 15
9:85 10 =7:69 1012
O
Author Proof
PR
888 Carazo and Luis A. Godoy are members of the re- 905
c=a 1
889 search staff of CONICET. Diego J. Celentano thanks CC ¼ 1:29 10 þ 5:13 10 3 CSi
890 CONICYT (Chilean Council of Research and Tech-
891 nology) for the support provided by Project Fondecyt 2:56 10 4 T þ 2:9 10 4 ðCSi Þ2
892 1130404. þ 1:24 10 7 T2 þ 3:27 10 4 CSi T
D
907
893 APPENDIX A: PHASE DIAGRAMS Fe-Fe3C-Si:
TE 908
c=h 3 1
894 The equilibrium carbon concentrations for eutectoid CC ¼ 7:34 10 þ 1:82 10 CSi þ 1:7
895 phase changes together with the lower and upper
896 bounds of the stable and metastable eutectoid intercrit- 10 T þ 2:92 10 2 ðCSi Þ2
6
897 icals of the Fe-C-Si and Fe-Fe3C-Si systems are as þ 1:72 10 8 T2 þ 4:17 10 5 CSi T
EC
898 follows:
910
899 Fe-C-Si: a=h
CC ¼ 2:99 10 3
þ 1:37 10 4 CSi 9:84
c=g 3
CC ¼ 1:60 10 þ 8:13 10 5 CSi 6:46 10 6 T þ 1:10 10 5 ðCSi Þ2 þ 8:23
RR
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
TPaT ¼ 2:16 1010 þ 5:07 1020 ðCSi Þ2 3:94 1020 CSi þ 5:23 1019 þ 3:08 1015 CSi 2:82 1015 =2:32 1012
UN
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
TPA1 ¼ 4:05 10 þ 6:72 10 ðCSi Þ 7:3 10 CSi þ 7:5 10 þ 5:46 10 CSi þ 1:16 10 =2:4 1013 ;
11 21 20 19 16 16
901
a=g 4 where CSi is the Si content in austenite at different 912
CC ¼ 9:53 10 þ 1:02 10 2 CSi þ 1:55 interfaces expressed in weight percentage and T is the 913
10 6 T þ 9:59 10 4 ðCSi Þ2 þ 1 temperature of the alloy in Celsius degrees. 914
10
10 T2 1:76 10 5 CSi T
903
F
921 each graphite spheroid (see Table B-IV) and Rai is the unit volume associated with the j event of nucleation. 945
922 radius of ferrite grains nucleated on the spheroid. Aa=gj As the austenite fraction is transformed into graphite, 946
O
923 could be computed with higher precision by means of a ferrite, and/or pearlite, its value should be computed again as 947
924 surface integral and assuming that each ferrite grain is
925 located on the spheroid; however, the differences with fc ¼ 1 fg fa fP ;
O
926 the results of Eq. [B1] are negligible.
where fa and fP are the volume fractions of ferrite and 949
Author Proof
PR
927 APPENDIX C: VOLUME FRACTION Details of the microstructure models from which fa 951
928 OF GRAPHITE AND AUSTENITE and fP are calculated are given by Carazo.[12] 952
D
932 time t þ Dt (tþDt Rgj ) is AUSTENITE 954
The value of CcC per unit volume of RVE is computed as 955
TE
tþDt
Rgj ¼ t Rgj þ DRgj :
where CXC , qX ; and t UX are carbon concentrations in 956
9
833 (560) 47.1 1073 (800) 750 9 103
973 (700) 43.6 1418 (1145) 820 9 103
1113 (840) 38.1 1428 (1155) 840 9 103
1253 (980) 32.5 1673 (1400) 840 9 103
1393 (1120) 28.8
CO
1673 (1400) 45
Mass density (kg/m3) 7300
Solidification phase-change model[11]
Eutectic latent heat (J/kg) 2 105
Carbon diffusion coefficient in liquid and austenite (m2/s) DlC ¼ 5 1010 and DcC [10]
Graphite nucleation coefficients bM ¼ 1 1013 (grains/m3 Ks) cM ¼ 280 ðÞ (K)
UN
F
373 (100) 0.54 experiments: 293 K (20 °C). 971
573 (300) 0.57 The values of specimen–mold conductance coefficient, 972
O
773 (500) 0.65 specimen–environment and mold–environment convec- 973
973 (700) 0.79 tion heat transfer coefficients, and specimen thermocou- 974
1173 (900) 1.00
O
ple conductance coefficient are shown in Tables E-VI, 975
1373 (1100) 1.26
E-VII, and E-VIII, respectively. 976
Author Proof
PR
Mass density (kg/m3) 1550
Specific heat (J/kg) 1 9 106 REFERENCES 977
1. ASTM A247-10: ASTM International, West Conshohocken, PA, 978
2010. doi:10.1520/A0247-10. 979
Table E-VI. Specimen–Mold Conductance Coefficient 2. D. M. Stefanescu and C. S. Kanetkar: in Computer Simulation of 980
981
D
Microstructural Evolution, D. J. Srolovitz, eds., TMS, Warrendale,
Temperature [K (°C)] Conductance Coefficient (W/m2 K) PA, 1985, pp. 171–88. 982
3. BC Liu, HD Zhao, WY Liu, and DT Wang: Int. J. Cast Met. Res., 983
293 (20) 60 TE 1999, vol. 11 (5), pp. 471–76. 984
773 (500) 70 4. S Chang, D Shangguan, and DM Stefanescu: AFS Trans., 1991, 985
1123 (850) 90 vol. 99, pp. 531–41. 986
1443 (1170) 100 5. S Chang, D Shangguan, and DM Stefanescu: Metall. Trans. A, 987
1673 (1400) 100 1992, vol. 23A, pp. 1333–46. 988
6. A. Almansour, K. Matsugi, T. Hatayama, and O. Yanagisawa: 989
EC
Table E-VIII. Specimen–Thermocouple Conductance y Naturales, Universidad Nacional de Córdoba, 2012. 1002
Coefficient 13. D Celentano, E Oñate, and S Oller: Int. J. Num. Meth. Eng., 1994, 1003
vol. 37, pp. 3441–65. 1004
Interface Heat Transfer Coefficient (W/m2 K) 14. R Hill: J. Mech. Phys. Solids, 1965, vol. 13 (4), pp. 213–22. 1005
15. FD Carazo, PM Dardati, DJ Celentano, and LA Godoy: Metall. 1006
Part-thermocouple 40 Trans. B, 2012, vol. 43B (6), pp. 1579–95, DOI:10.1007/s11663- 1007
UN
012-9710-y. 1008
16. B Lux: AFS Cast Met. Res. J., 1972, vol. 8 (1), pp. 25–38. 1009
17. B Lux: AFS Cast Met. Res. J., 1972, vol. 8 (2), pp. 49–65. 1010
957 weight percentage, density, and carbon quantity in a 18. D. M. Stefanescu and D. K. Bandyopadhyay: in The Physical 1011
Metallurgy of Cast Iron. Proceedings of the Fourth International 1012
958 micro-constituent X, respectively. X may be austenite, Symposium on the Physical Metallurgy of Cast Iron, G. Ohira, T. 1013
959 graphite, ferrite, or pearlite. The derivation of Eq. [D1] Kusakawa, E. Niyama, eds., Materials Research Society, Tokyo, 1014
960 may be seen in Carazo.[12] Japan, 1989, pp. 15–26. 1015
19. J. A. Sikora, G. L. Rivera, and H. Biloni: in F. Weinberg Int. 1016
Symp. on Solidification Processing, J. E. Lait and I. V. Sama- 1017
rasekera, eds., Pergmon Press, Inc., New York, NY, 1990, pp. 1018
961 APPENDIX E: THERMO-PHYSICAL PROPERTIES 280-88. 1019
962 AND MATERIAL PARAMETERS USED IN THE 20. DK Banerjee and DM Stefanescu: AFS Trans., 1991, vol. 99, 1020
pp. 747–59. 1021
963 NUMERICAL SIMULATIONS 21. Z. Jiyang: Acta Metall. Sin., 1989, vol (4), pp. 261–65. 1022
22. DM Stefanescu: J. Mater. Sc. Eng. A, 2005, vols. 413–414, 1023
964 Tables B-IV and E-V show the values of coefficients pp. 322–33. 1024
965 and thermo-physical properties of the alloy and sand 23. R. W. Cahn and P. Haasen (eds.): Physical Metallurgy, vol. 1, 4th 1025
966 used in the numerical simulation. The initial ed., Elsevier, Amsterdan, The Netherlands, 1996, p. 767. 1026
F
1041 Physical Metallurgy of Cast Iron. Proceedings of the Fourth vol. 73 (6), pp. 977–83. 1073
1042 International Symposium on the Physical Metallurgy of Cast Iron, 45. RW Heine: AFS Trans., 1986, vol. 94, pp. 391–402. 1074
O
1043 G. Ohira, T. Kusakawa, E. Niyama, eds., Materials Research 46. A Hultgren: Trans. ASM., 1947, vol. 39, pp. 915–89. 1075
1044 Society, Tokyo, Japan, 1989, pp. 433–40. 47. A. Fick: Ann. Phys. (Berlin, Ger.), 1855, vol. 170(1), pp. 59–86. 1076
1045 30. E Frás, W Kapturkiewicz, and HF Lopez: AFS Trans., 1992, 48. PG Shewmon: Diffusion in Solids, 2nd ed., Wiley, New York, NY, 1077
1046 vol. 100, pp. 583–91. 1989, p. 28. 1078
O
1047 31. Y. Zhang, S. V. Subramanian, and G. R. Purdy: in The Physical 49. M Abramoff, P Magalhaes, and S Ram: Biophotonics Int., 2004, 1079
1048 Metallurgy of Cast Iron, G. Lesoult and J. Lacaze, eds., Trans. vol. 11 (7), pp. 36–42. 1080
Author Proof
1049 Tech. Public, Switzerland, 1994, pp. 461–68. 50. CA Schneider, WS Rasband, and KW Eliceiri: Nat. Methods, 1081
PR
1050 32. A. Almansour, K. Matsugi, T. Hatayama, and O. Yanagisawa: 2012, vol. 9 (7), pp. 671–75, DOI:10.1038/nmeth.2089. 1082
1051 Mater. Trans., JIM, 1995, vol. 36(12), pp. 1487–95. 51. EE Underwood: Quantitative Stereology, 1st ed., Addison-Wesley, 1083
1052 33. L. Wenzhen and L. Baicheng: in Proceedings of the Technical Reading, MA, 1970, p. 119. 1084
1053 Forum, 62nd World Foundry Congress, American Foundrymen’s 52. DM Stefanescu: Science and Engineering of Casting Solidification, 1085
1054 Soc., Inc., Philadelphia, USA, 1996, pp. 2–10. 2nd ed., Springer, New York, 2009, p. 21. 1086
1055 34. R Aagaard, J Hattel, W Schafer, IL Svensson, and PN Hansen: 53. A. S. Pandit: Ph.D. Thesis, Department of Materials Science and 1087
1056 AFS Trans., 1996, vol. 104, pp. 659–67. Metallurgy, University of Cambridge, 2011. 1088
D
1057 35. C. Charbon and M. Rappaz: Adv. Mater. Res. (Durnten-Zurich, 54. JJ Kramer, GM Pound, and RF Mehl: Acta Metall., 1958, vol. 6 1089
1058 Switz.), 1997, vol. 4–5, pp. 453–60. (12), pp. 763–71. 1090
1091
TE
EC
RR
CO
UN