You are on page 1of 6

Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 81 (2014) 498 503

11th International Conference on Technology of Plasticity, ICTP 2014, 19-24 October 2014,
Nagoya Congress Center, Nagoya, Japan

Model of curvature of crankshaft blank during the heat treatment after


forging
Andrzej Milenina, Tomasz Reca, Wojciech Walczykb, Maciej Pietrzyka,*
a

AGH University of Science and Technology, Al. Mickiewicza 30, Krakow, 30-059, Poland
b
,123,QVWLWXWHRI0HWDO)RUPLQJXO-DQD3DZD,,-3R]QD3RODQG

Abstract
The manufacturing process of large crankshafts is affected by undesirable bending of the shaft blank at the heating and heat
treatment stages, as well as at the forging step. In line elimination of the shaft blank bending is a time consuming and expensive
operation. Identification of reasons of the shaft bending and identification of possibilities of avoiding this bending is an
important theoretical and practical problem. In previous works, the Authors developed the finite element model and applied this
model to optimization of the forging process and to predict microstructure evolution. The model was extended to predict
bending of shaft and applied in the present work to simulate the influence of various parameters of forging on shaft bending.
The work is devoted to further extension of the model by including deformation of the crankshaft during heat treatment after
forging. The model estimates the contribution of the elastic-plastic deformation, including thermal expansion and dilatometric
effect due to transformations, on the deformation and bending of the shaft. This process generates also residual stresses. The
aim of the paper was to simulate phenomena occurring during heat treatment. A program based on the finite element method
was developed to solve the three-dimensional thermal and mechanical problems. Shaft material model was developed and the
elastic-plastic characteristics were implemented in the FE code. Heat exchange with the cooling medium and the dependence of
thermal properties on temperature and heat of phase transformations were accounted for in a solution of the thermal problem.
Dilatometric tests were performed to supply data for identification of the phase transformation model during cooling. The
results of calculations of bending of shaft during the process of heat treatment, as well as the distribution of residual stresses
and strains are presented in the paper. The calculations were performed for several modes of the heat treatment. Parameters
affecting bending of the shaft were identified.
2014 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
2014 The Authors. Published by Elsevier Ltd.
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and
and peer-review
peer-review under
under responsibility
responsibility of
of the
Nagoya
University
and Toyohashi
of Technology.
Selection
Department
of Materials
ScienceUniversity
and Engineering,
Nagoya University
Keywords: Residual stresses; Phase transformation; Numerical modeling; Crankshaft

* Corresponding author. Tel.: +0-000-000-0000 ; fax: +48-12-617-2921.


E-mail address: maciej.pietrzyk@agh.edu.pl

1877-7058 2014 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of the Department of Materials Science and Engineering, Nagoya University
doi:10.1016/j.proeng.2014.10.029

499

Andrzej Milenin et al. / Procedia Engineering 81 (2014) 498 503

1. Introduction
The technology of forging of crank shafts developed at the Metal Forming Institute in Poznan, Poland (TR
method following the name of the inventor) (Rut, Walczyk (2000), Rut, Walczyk (2002)) is the subject of this
paper. This technology gives very good results when shape of the crank webs is close to an ellipse. Application of
this technology to new constructions of crankshafts, which require large juts at the side of journals of main
bearings, was considered by Walczyk et al. (2011). In that technology an unsymmetrical preupsetting of the stock
material, which involves proper flow of the material needed to obtain required shape of the crank shoulders, was
introduced. Several possible variants of this process were investigated (Walczyk et al. (2011)) by numerical
simulations and loads, as well as flow of the material, were compared. An attempt of optimization of the tool
trajectories in this process was presented by Sztangret et al. (2011). The criterion of optimality in that work was
based on the comparison of the shape of material after forging with the assigned shape of the crank throw. A
change of the parameters of deformation in the TR process is connected also with a change in microstructure and
properties of material. Prediction of the changes during the numerical optimization requires the addition of the
microstructure evolution model to the finite element code (FEM) (Milenin et al. (2012)).
The manufacturing process of large crankshafts is affected by undesirable bending of the shaft blank at the
heating and heat treatment stages, as well as at the forging step. In line elimination of the shaft blank bending is a
time consuming and expensive operation. Identification of reasons of the shaft bending and identification of
possibilities of avoiding this bending is an important theoretical and practical problem. Therefore, the aim of the
paper was to develop the FEM model of curvature of crankshaft blank during the heat treatment and to simulate
phenomena occurring during heat treatment.
2. Description forging and heat treatment technology
The crankshafts investigated in this paper have nine cranks. Due to very large length of the forging billet (about
16 meters) the forging process of cranks is performed in nine cycles. One cycle contains:
1) local heating of forging of billet, 2) transport of forging billet from the furnace to TR forging device mounted
on the press, 3) forging process of one crank, 4) transport of forging billet to the furnace.
Last stage of manufacturing of crankshaft is heat treatment. In this stage a crankshaft is heated to the relevant
temperature and heat-treated. For heavy crankshafts the heat treatment is carried out in vertical position. The
numerical model of this stage is required. The model should predict a shape of the crankshaft and initial
temperature distribution after forging process. Therefore, the first step of investigation was numerical analysis of
forging process for cranks.
3. FEM modeling of crankshaft forging process
Numerical simulations of forging of cranks were done in Forge 2008 program based on the FEM. Boundary
conditions in the furnace, in air and between forging tools and crankshaft were defined using results published by
+DGDD (2013). Detailed description of the conditions and the results of simulation of the forging process was given
in (Milenin et al. (2012)). Temperature of the crankshaft after heating was about 1250C and it was uniform in the
whole volume of the material. Temperature of the crankshaft decreased during the transport to the press and
forging process. The measured temperature after forging process was 1170C on the surface of the crank.
4. Model of phase transformations
The objective of modelling of phase transformations was to supply data for calculation of dilatometric strains
and heat of transformation. This task can be efficiently completed by a simple model based on the JMAK equation:
X

1  exp kt n

where X is the volume fraction of a new phase and k and n are the model parameters.

(1)

500

Andrzej Milenin et al. / Procedia Engineering 81 (2014) 498 503

The coefficient k must vary with temperature in a way linked to the TTT diagram:
k

T  T a8 .
nose
kmax exp 

a7

(2)

In equations below, which complete the phase transformation model, parameters are introduced as a with
number in subscript. In equation (3) k max = a 5 /D J is the maximum value of k, T nose = A e3 + 400/D J a 6 is a
temperature position of the nose of the Gaussian function, where D J is austenite grain size. Coefficient a 7 is
proportional to the nose width at the mid height and a 8 is related to the sharpness of the curve. The following
equations describe well coefficient k respectively in pearlitic (k p ) and bainitic (k b ) transformations:
kp

a14
T ,

exp a13  a12

a16
100
DJ

(3)

kb

T .

a23 a22  a21

100

(4)

Incubation time was introduced for bainitic (W b ) and pearlitic (W p ) transformations:


WP

Wb

a9

Ae1  T

a11

a17 kb

Bs  T

a19

(5)

a u103
exp 10

RT

(6)

a u103
exp 18

RT

where T - temperature in C, T - temperature in K.


Start temperatures for the bainitic and martensitic transformations are functions of chemical composition:

Bs [ o C] a20  425[C]  42.5[Mn]  31.5[Ni]

(7)

M s [ o C] a26  a27CJ

(8)

where C J carbon concentration in the remaining austenite.


Fraction of austenite, which transforms into martensite, was calculated according to the model:
Fm

^1  exp 0.011 M

(9)

 T 1  Ff  Fp  Fb

where F f , F p , F b fraction of ferrite, pearlite and bainite with respect to the whole volume of material.
Coefficient n in equation (1) has numbers a 4 , a 15 and a 24 for ferritic, pearlitic and bainitic transformations,
respectively. The coefficients in the model were identified using inverse analysis for the data derived from the
dilatometric tests. Results of identification are given in Table 1. The results of calculation of phase transformations
in steel 34CrNiMo6 at different cooling rates are present in Fig. 1. Validation of the model of phase
transformations was performed on the basis of experimental data presented by Nrnberger et. al. (2010) (Fig. 2).
Table 1. Coefficients in the phase transfor4mation model obtained from the inverse solution for dilatometric tests.

a4
0.5756
a 16
0.53

a5
0.4668
a 17
69.48

a6
58.65
a 18
25.5

a7
7.843
a 19
1.0

a8
1.418
a 20
722

a9
4834
a 21
0.627

a 10
0.00016
a 22
0.0414

a 11
0.0003
a 23
0.0504

a 12
0.53
a 24
0.588

a 13
0.53
a 26
589.5

a 14
0.002
a 27
874.6

a 15
0.755

501

Andrzej Milenin et al. / Procedia Engineering 81 (2014) 498 503


800

steel 34CrNiMo6

temperature, oC

700

F start
P start
B start

600

A+F

B end
M start

500
400

50 15 6
80

0.15 0.04
0.3
0.02
0.6
1

A+B

A+F+B

300
200

cooling
rate, oC/s

B+M

F+B

100
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

log(time)

Fig 1. The results of calculation of phase volume fractions in steel


34CrNiMo6 at different cooling rates: (a) 0.01 C/s.

Fig 2. Validation of the model of phase transformations on the basis


of experimental data presented in the [7].

5. Heat transfer model during heat treatment


Heat transfer model was based on the solution of the heat conduction equation, taking into account the heat of
phase transformations:
c t U t

dt
dW

f dF f
dF
P dFp
m dFm

div>k t grad t @  U H trans


 H trans
 H btrans b  H trans
dW
dW
dW
dW

where U density, t temperature, W  time,

(10)

k t conductivity coefficient, c t  specific heat, V  effective

stress, [ effective strain ratio, H , H , H B , H M heat of phase transformation; F f , F p , F b F m volume


trans
trans
trans
trans
fractions of ferrite, pearlite, bainite and martensite with respect to the whole volume.
Thermophysical properties of steel are given by equations: c(t) = 0.2451t + 450.57 and k(t) = -0.0135t + 38.325.
The boundary conditions included convective heat transfer at the boundaries between the surface of forging and the
environment (adopted from the database of FORGE). The initial temperature distribution and FEM grid were
transferred from the results of forging process simulation. In the developed FEM program the type of elements was
the same, which was used in the program FORGE (tetrahedral simplex elements).
F

6. Mechanical model of thermal deformation


Mechanical model was based on the theory of small elastic-plastic deformations. Detailed description of the
FEM solution was given for example by Kopernik et al. (2011, 2014). Dilatation curves, which were used for
identification of the model of phase transformations, were also used to determine the thermal deformation. This
deformation depended on the cooling rate (CR), the temperature (t) and it was calculated from the equation:

Ht

(11)

H D t , CR  H D t0 , CR

The dependence of the characteristics of elastic plastic material on temperature was defined as:

E = 215307-104.91t

(12)

V p = 961.77 - 0.9138t

(13)

A simulation of unloading is a crucial problem in determining the residual stresses and strains, which cause
bending of the shaft. Modeling of unloading was realized according to the theorem of unloading. The following
algorithm was performed for each time step of the unloading:
x The distributions of stresses ^V ` and strains ^H ` before unloading are the solutions of the loading problem.
e
x Then, the specimen is loaded by an inverse force and a material is modeled as elastic. The fields of V and
e
H are obtained after this procedure.
x The actual state of a stress and a strain during the unloading stage is reached as a sum of the solutions:

^ `

^ `

502

Andrzej Milenin et al. / Procedia Engineering 81 (2014) 498 503

^V

unload

` ^V ` ^V `

(14)

^H

unload

` ^H ` ^H `

(15)

where ^H e ` - an elastic part of a strain, ^V e ` - an elastic part of a stress.


7. Example of simulation
As an example, the calculation of a cooling with the temperature of the shaft ranging from 860qC to 400qC was
considered. The simulation results are shown in Figs 3-6. When analyzing the stress state, two specific areas of the
crank pin indicated with the numbers 1 and 2 in Fig. 3a were considered. At the initial stage of cooling (Fig. 3) the
phase transformation has not yet begun. Analysis of the shape of the crank throw leads to the conclusion that the
bottom part of the crank pin (zone 2 in Fig. 3) was cooled intensively (Fig. 3a). Due to thermal shrinkage
significant tensile stresses appeared in this place. Since the entire surface of the crank throw is cooled faster,
compressive stresses appear in the inner part of the crank throw and tensile stresses appear in the outer part (Fig.
3b). Total change in the shape of the crank pin is characterized by a divergence of the upper parts. Further cooling
(Fig. 4) is associated with the beginning of phase transformations that corresponds to the nonlinear change of
thermal deformation in accordance with the dilatation curve. The direction of the bending of the crank pin is
reversed and the maximum tensile stress is moved to the zone 1 (Fig. 4b).

(a)

(b)

Fig 3. Modeling of crank deformation during cooling after 4 hours (before the start of the phase transformation): (a) the temperature distribution,
(b) mean stress distribution. Scale of displacements 200:1.

(a)

(b)

(c)

Fig 4. Modeling of crank deformation during cooling after 6 hours (after the start of the phase transformation, which causes a change of the
bending direction): (a) - the temperature distribution, ( b) mean stress distribution, ( c) the distribution of the ferrite and the pearlite contents.
The scale of displacements 200:1.

After completion of the intensive phase of transformation thermal deformation becomes linear again and
deformation due to nonuniform temperature distribution becomes dominant (Fig. 5a). In consequence, the lower
colder crank pin part (zone 2) is bent as shown in Fig. 5. Stress in zone 1 again changes sign and becomes
compressive (Fig. 5b). At the final stage of cooling process the temperature begins to equalize (Fig. 6a), resulting
in the unloading volume of metal that deforms plastically. This leads to a last qualitative change in the direction of
the bending of the crank pin and the appearance of residual stresses and strains, as shown in Fig. 6b.

503

Andrzej Milenin et al. / Procedia Engineering 81 (2014) 498 503

(a)

(b)

(c)

Fig 5. Modeling of crank deformation after 11 hours of cooling (after the intensive phase transformation, which again changes the direction of
the crank deflection): (a) the temperature distribution, (b) the distribution of mean stress, ( c) the distribution of ferrite and pearlite contents.
The scale of displacements 200:1.

(a)

(b)

(c)

Fig 6. Modeling of crank deformation after 17 hours of cooling (after the phase transformation was completed and relaxation started, which
again changed the direction of the crank deflection): ( a) the temperature distribution, (b) the distribution of the mean stress, (c) the distribution
of X - displacement. The scale of displacements 200:1.

8. Conclusions
1. Mathematical model of crankshaft deformation in the conditions of heat treatment, taking into account the
elastic-plastic deformation and phase transformations, was developed.
2. It was shown in the paper that in the presence of phase transformation, cooling process is accompanied by a
three-time changing of stress sign and the direction of bending of the crankshaft. It is due to the nonlinearity of
thermal deformation during the phase transformation and unloading of metal after the temperature becomes
uniform.
Acknowledgement: Financial assistance of the NCBiR, project number PBS1/B6/3/2012, is acknowledged.
References
Rut, T., Walczyk, W., 2000. Obrbka plastyczna Metali. 11 5-8 (in Polish).
Rut, T., Walczyk, W.: Part I., 2002. Archiwum Technologii i Automatyzacji, 22, 177-186; Part II. Archiwum Technologii i Automatyzacji 22,
187-196 (in Polish).
Walczyk, W., Milenin, A., Pietrzyk, M., 2011. Computer aided design of new forging technology for crank shafts. Steel Research International
82, 187194.
Sztangret, Milenin, A., Sztangret, M., Walczyk, W., Pietrzyk, M., Kusiak, J., 2011. Computer aided design of the TR forging technology for
crank shafts sensitivity to model and process parameters. Computer Methods in Materials Science 11, 237-242.
Milenin, A., Walczyk, W., Pietrzyk, M., 2012. Numerical modeling of microstructure evolution during forging of crank shafts. Steel Research
International 83, 808816.
+DGDD%,PSOHPHQWDWLRQRIWKHKHDWEDODQFHLQWKHILQLWHHOHPHQWVROXWLRQWRWKHWHPSHUDWXUHILHOGRIWKHSODVWLFDOOy deformed material.
International Journal of Thermal Sciences 71, 172181.
Nrnberger, F., Grydin, O., Schaper, M., Bach, F. -W., Koczurkiewicz, B., Milenin, A., 2010. Microstructure transformations in tempering
steels during continuous cooling from hot forging temperatures. Steel Research International 81, 224233.
Kopernik, M., Milenin, A., Major, R., Lackner, J.M., 2011. Identification of material model of TiN using numerical simulation of
nanoindentation test. Materials Science and Technology 27, 604-616.
Kopernik, M., Milenin, A., 2014. Numerical modeling of substrate effect on determination of elastic and plastic properties of TiN nanocoating
in nanoindentation test. Archives of Civil and Mechanical Engineering 14, 269277.