You are on page 1of 8

A New Forming Limit Criterion and significantly increase the strength and workability of a

workpiece. However, fracture will occur on the workpiece


Ductile Fracture Prediction in the during forging for the improper blank profiles and incorrect
Forging of Sintered Powder Preforms die design.

Cheng-Chao Huang Ŧ and Jung-Ho Cheng † The plastic theories of porous materials have been
Department of Mechanical Engineering widely examined. A yield criterion for porous media was
National Taiwan University proposed by Gurson [1], and the scalar parameters were
No. 1, Sec. 4, Roosevelt Rd., Taipei, 106, Taiwan introduced by Tvergaard [ 2 ] into the criterion. The
deformation characteristics and fracture of iron compacts
were investigated by Spitzig et al. [3, 4, 5]. The finite
Abstract
element method (FEM) was used by Hartley et al. [6] to
The main purpose of this study is to investigate the analyze the effects of friction in upsetting of billets. The
forgability of sintered powder compacts, including the void influences of initial densities and friction in cold forging of
evolution and fracture prediction in the forming process. A powder preforms were researched by Jha and Kumar [7].
new ductile fracture criterion for porous media is developed Then, the finite element program ABAQUS was utilized by
based on the maximum principal strain energy per unit of Hom and McMeeking [8] to study the growth of voids. The
volume. It demonstrates that the strain energy density is processes of powder forging were also simulated by Huang
governed by the maximum principal stress in material. and Cheng [9].
Fracture will occur as the accumulation of strain energy
Ductile fracture criteria for a fully dense matrix were
density reaches the critical value at that point.
widely investigated based on the stress field, strain field, and
Experiments and Finite element analysis were the strain energy density due considerations. A critical value
conducted to establish the models. The material properties of in terms of plastic strain to study the instability in sheet
sintered porous preforms were obtained from both the metal forming was made by Keeler and Backofen [10].
uni-axial tension and compression tests. The finite element Upsetting experiments of solid cylinders and rings were
models were verified with experiments of upsetting under conducted by Kobayashi [11, 12] to obtain the fracture
different frictional conditions. Studies show that the results strain. A ductile fracture model based on the nucleation and
by the new criterion are more compatible with the growth of voids under tensile strain was studied by Roy et al.
experimental data than other previous forms. In addition, [13]. The prediction of instability and fracture of specimens
voids lead to the degradation of the strength and workability undergoing an upsetting test was addressed by Ettouney and
of materials. Friction contributes to the non-uniformity of Hardt [ 14 ]. Also, the fracture from cold upsetting was
deformation and density variation that causes the workpiece described by Darvas [15]. A criterion was developed by
to fracture. McClintock [16] for fracture by the growth and coalescence
Keywords: porous materials, powder forging, plasticity, of holes, and it was quoted by Sowerby and Chandrasekaran
FEM, fracture criteria, strain energy density [17] into the finite element model to predict fracture in
upsetting tests. The damage rate for spherical voids was also
considered by Rice and Tracey [18] in the fracture equation.
1. Introduction
Then, a forming limit criterion based on the hydrostatic
Powder forging combines powder metallurgy and component and effective stress was proposed by Vujovic and
forging technologies and thus possesses the advantages of Shabaik [19].
both processes that result in stronger yet versatile products
On the other hand, the forming criteria formulated by
with complicated geometry and arbitrary alloy compositions.
strain energy density were developed. The earliest
Preforms are prepared through various steps that involve
investigation based on the generalized plastic work was
powder mixing, compacting, and sintering, and the exact
made by Freudenthal [20]. Then, this criterion was modified
shape is obtained through a final forging procedure. Voids in
by Cockcroft and Latham [21] by the largest tensile stress.
sintered powder compacts exert a detrimental effect on the
Other workability criteria were also proposed by Brozzo et
mechanical properties, reducing the cavities that can
al. [22], Norris et al. [23], and Oyane et al. [24], including
the effects of the effective stress and hydrostatic component.
Ŧ
Graduate student, E-mail: r1852206@w3.me.ntu.edu.tw Recently, several previously published continuum criteria

Professor, the corresponding author, Fax: +886-2-2363-1755, were utilized by Clift et al. [25], Gouveia et al. [26, 27],
E-mail: jhcheng@w3.me.ntu.edu.tw Wifi et al. [28], Jain et al. [29], and Takuda et al. [30] in the
finite element models to predict fracture in the fully dense
1
matrix under various forming processes. is the deviatoric part of the macroscopic Cauchy stress
The workability of porous preforms is greatly concerned tensor σ ,
in forging applications. The forming limit in terms of the 1
strain ratio (major strain/minor strain) was applied by
σh = σ :I (4)
3
Downey and Kuhn [31]. A criterion for porous materials as a is the hydrostatic stress, and I is the identity matrix.
function of history of the hydrostatic stress was proposed by
The yield criterion of porous media as proposed by
Tabata and Masaki [ 32 ]. Then, a workability factor,
Gurson [1] is based on a rigid-plastic upper bound solution
proposed by Vujovic and Shabaik [19], describing the effects
for spherically symmetric deformations of a single spherical
of mean stress and effective stress was utilized by
void. The criterion combines the effects of deviatoric part
Abdel-Rahman and El-Sheikh [ 33 ] to investigate the
and hydrostatic stress, and is represented as follows:
forming limit of powder compacts in upsetting. Various
existing localization conditions and fracture criteria were 2
⎛σ ⎞ ⎛ ⎞
assessed by Lee and Zhang [34] in the finite element model Φ (σ , f ) = ⎜ ⎟ + 2 q1 f cosh ⎜ 3 q 2σ h (
⎟ − 1 + q3 f 2
)= 0 , (5)
⎜σ ⎟ ⎜ 2σ ⎟
to study the plastic flow and damage in compression of ⎝ y ⎠ ⎝ y ⎠
porous cylinders. The workability of porous materials was where f is the void volume fraction, σ y is the flow
also simulated by Zhang et al. [35] with a strain-based stress of a fully dense matrix, and q1, q2, and q3 are the
criterion determined by Lee and Kuhn [36]. material parameters.
Despite the close scrutiny of the ductile fracture criteria The presence of pressure in the yield conditions results
for a fully dense matrix or porous metals, the forming limit in nondeviatoric plastic strains. The plastic flow is assumed
of materials still cannot be described accurately. The main to be normal to the yield surface
purpose of this study, therefore, is aimed at investigating the
forgability of sintered preforms, including the void evolution ∂Φ ,
ε& p = λ& (6)
and fracture prediction in the formation stage (Fig. 1). First, ∂σ
the specimens were prepared from metal powder through the &
where λ is the nonnegative plastic flow multiplier.
compaction and sintering steps. The material properties of The hardening of a fully dense matrix is described as
sintered compacts were obtained from both the uni-axial σ = σ (ε mp ) , where ε mp is the effective plastic strain of
tension and compression tests. Then, the finite element the matrix. The evolution of effective plastic strain in the
models were established and verified with experiments of matrix is obtained from the following effective plastic work
upsetting under both frictionless and frictional conditions. In expression
order to govern the workability of preforms, a ductile
fracture criterion for porous materials is developed. (1 − f )σ y ε&mp = σ : ε& p . (7)
Sequentially, the cracking simulation in upsetting of porous
Observing the deformation mechanisms in the porous
billets with various initial relative densities was achieved by
metals, the total change of void volume fraction is partly due
using these models with the new criterion.
to the growth of existing voids and partly due to the
nucleation of new voids, stated as
2. Theoretical review
2.1 Yield criterion and void evolution f& = f&g + f&n , (8)
The plastic behaviors of metals have been generally where f&g is the growth rate of existing voids and f&n is
studied and described by constitutive equations. The von the change of nucleation of new voids. The growth of
Mises yield criterion, which is based on the shape distortion existing voids is based on the law of conservation of mass
energy for fully dense media, is in the following form: and is expressed as

Φ(σ ) = σ − σ y = 0 , (1)
f&g = (1 − f )ε& p : Ι . (9)
where σy is the flow stress of fully dense media,
On the other hand, the nucleation of new voids stems from
3 the micro-cracking or decohesion of the particle-matrix
σ = S :S (2)
2 interface only in the tensile stress state. Void nucleation at a
is the von Mises effective stress, material point, therefore, will not be considered in
compressive stress.
S = σ −σ h I (3)
2.2 Ductile fracture criteria
2
The forming limit in metalworking is a complex where σ 1 is the maximum principal stress and dε 1 is the
phenomenon that depends both on the materials and the corresponding principal strain increment; on the right hand
deformation process. Several ductile fracture criteria were of the equation, C cr is the critical constant for a fully
proposed by the formulation of strain energy density. The dense matrix toward fracture and m is the material
earliest attempt was made by Freudenthal [20], introducing a parameter, describing the sensitivity of the void volume
critical constant of the generalized plastic work per unit of fraction, f , on the critical value of strain energy density, to
volume at fracture, which is expressed as be determined experimentally. This criterion emphasizes the
dependence of the value of strain energy density and the
ε
∫ σ dε = C1 , (10) domination in fracture upon the level of the maximum
f

principal stress, σ 1 . Ductile fracture will occur as the


0

where σ is the effective stress, dε is the effective strain accumulation of the maximum principal strain energy
increment, ε f is the effective strain at fracture, and C1 is density in material meets the critical value at that point.
the critical constant for a fully dense matrix. However, it Emphatically, the existence of voids leads to the exponential
cannot be used in the state of compressive stress during degradation of the critical value for porous metals at fracture.
deformation for all the positive value of strain energy This new criterion is applied to predict fracture in the
density. powder forging simulation.
Then, a criterion based on the critical constant of the 2.3 Material parameters
largest tensile strain energy density was modified by
The Gurson-Tvergaard yield criterion is adopted to
Cockcroft and Latham [21]. It recognizes the importance of
describe the plastic behavior of porous metals in this study.
tensile stress component in fracture,
Emphatically, the scalar parameters, q1, q2, and q3, are
εf ⎛σ * ⎞ determined empirically. On the other hand, the critical
∫ 0
σ ⎜⎜
⎝σ ⎠
⎟⎟d ε = C 2 , (11) constant for the fully dense matrix Ccr and the material
parameter m in the new fracture criterion, Eq. (14), are
where σ is the largest tensile stress and C 2 is the
*
also obtained by fitting the data from experiments.
critical constant. Nevertheless this form is devoid of
presenting the physical meaning of the strain energy density
3. Experimental work
in material. On the other hand, the forming criteria
combining with the level of both the maximum principal 3.1 Specimens preparation
stress, σ 1 , and the hydrostatic stress, σ h , was proposed The porous specimens used in the experiments are
by Brozzo et al. [22] by means of an empirical modification prepared from water-atomized iron powder. Their
of the above mentioned criterion, Eq. (11), constituents and particle size distributions are shown in
Table 1. The irregular particle shape, which increases the
ε 2σ 1
∫ dε = C3 . (12)
f
mechanical locking between powders, can enhance the green
0 3 (σ 1 − σ h )
strength of compacts and reduce the volume contraction
The fracture criterion for porous metals, including the during the sintering step. Various compacting pressures and
effect of porosity, was investigated by Tabata and Masaki the floating die method are utilized to obtain cylindrical
[32], and is presented as follows: billets with different green densities yet more homogeneous
distribution of the particles. The compacted billets are 12.5

ε σ ⎞
∫ ⎜ A + h ⎟d ε = B ⋅ ρ r 0 , (13) mm in diameter, and the height is controlled by weight to
f C
0
⎝ σ ⎠ make the aspect ratio (height/diameter) less than 1.5 to
where A, B and C are material constants to be determined by prevent the specimens from buckling during compression
experiments and ρ r 0 is the initial relative density of the tests.
materials. This equation also shows the importance of the
All the compacts were sintered in dry hydrogen at 1200
effect of hydrostatic stress on the fracture of porous
℃ for 1 hour and then cooled down to room temperature. In
materials during loading.
order to determine the mechanical properties of the matrix,
In order to actually govern the workability of porous some porous specimens were fully densified by rolling.
metals, a new ductile fracture criterion for porous media are Since the microstructure plays a significant role in plastic
developed based on the maximum principal strain energy per behavior, the process of recrystallization in dry hydrogen at
unit of volume in this research, which is expressed as 600℃ for 10 minutes was needed to dispel the hardening
εf induced in the rolling process.

0
σ 1 d ε 1 = C cr ⋅ e − mf , (14)

3
Table 1 Powder constituent and particle size distribution where σ first yield is the first yield stress of the porous
Properties of water-atomized iron powders (KIP 301A) materials and f is the porosity in the first yielding state.
Chemical analysis Screen analysis The fitting curve meets the fully dense axis at -126.31 MPa,
Constituent % mesh Size(μ m) % which is similar in magnitude to the first yield stress of the
Fe Bal. +60 +250 - fully dense matrix in the uni-axial tension test (Fig. 2).
M. Fe - +80 +180 ≦ 2
In consideration of the Gurson-Tvergaard yield criterion,
C ≦ 0.01 +100 +150 ≦ 15 each of the material parameters contributes different effects
Si ≦ 0.05 +150 +106 10~30 to the plastic behavior of the porous materials. In this study,
Mn 0.10~0.25 +200 +75 10~40 these scalar parameters are determined from experiments by
P ≦ 0.025 +250 +63 ≦ 25 fitting the non-linear polynomial function, Eq. (15), into Eq.
S ≦ 0.025 +325 +45 5~30 (5). The values of Young’s modulus, E, Poisson ratio, ν ,
O ≦ 0.25 -325 -45 10~30 the first yield stress of the fully dense matrix, σ first yield ,
and the scalar parameters of the porous metals are all shown
3.2 Simple tension and frictionless compression tests in Table 2.
for material parameters
The material properties of sintered porous preforms Table 2 Material constants
were obtained from both uni-axial tension and compression Matrix property Porous parameter
E [GPa] ν σ first yield [MPa] q1 q2 q3
tests. In the simple tension test of the matrix, the specimens
180 0.27 126 1.81 1.00 2.80
were made of the fully densified pure iron. They measure 50
mm in gauge length, 12.5 mm in width, and 2 to 3 mm in 3.3 Fracture upsetting test of porous billets
thickness. The material constants, Young's modulus, Poisson Upsetting experiments without Teflon sheets were made
ratio, first yield stress, and plastic behavior of the matrix to perform the frictional compression and to determine the
were obtained in tensile loading. A strain gauge and an fracture stroke of porous billets with various initial relative
extensometer were also utilized to monitor the deformation densities, and the load-displacement curves toward fracture
in the axial and transverse directions. The tensile true were obtained. The cylindrical billets crack first on the
stress-true strain curve of a fully dense matrix is shown in equatorial surface because the frictional forces inhibit the
Fig. 2, indicating that the first yield stress is about 126 MPa. deformation of materials at contact interfaces of the billets
The plastic behavior of the matrix reveals a power law and flat die. The tension and compression tests are
relation σ = k (ε m ) , where the coefficient k is about
p n
conducted using the MTS 810 material testing system, and
543.63 and the strain hardening exponent n is about 0.3. the experiments are compared with the finite element
Sequentially, the porous cylindrical billets with various analysis.
initial relative densities were tested in uni-axial compression
loading, and Teflon sheets were used as a lubricant to 4. Finite element verification
maintain the uniform deformation of these specimens in the
4.1 Finite element model construction
frictionless upsetting. However, the porous materials are not
incompressible in the test because the void volumes change ABAQUS/Standard [ 37 ] was used to perform the
under the volumetric plastic strain of the materials. The modeling and the forging of porous compacts. The material
current area is imperatively needed to calculate true stress in properties and deformation characteristics are very
the billets. Thus, a diameter gauge, as shown in Fig. 3, is complicated in real cases, and some reasonable assumptions
constructed to measure the current diameter, which gives the are needed to simplify the analysis:
current area under the circular cross-section assumption. 1. The void distribution is homogeneous in the specimen.
2. The matrix is perfectly bonded and incompressible.
For each cylindrical billet, the first yielding point is
3. The cylindrical billets are uniformly deformed during the
found via the intersection of the line that shifts 0.2% off the
frictionless upsetting.
elastic region and the stress-strain curve. The first yield
stress-porosity data, with various initial relative densities, The finite element mesh and boundary conditions of the
are given in Fig. 4. The increase in void volumes leads to the cylindrical billet in the open flat die upsetting analysis are
degradation of the first yield stress. A third order polynomial shown in Fig. 5, with a quarter of each cylindrical billet and
function is used to fit these data, and the empirical curve is die being modeled for the axisymmetric profile, and the
obtained as follows: boundary conditions are specified on the line of symmetry.
The discretization is biased toward the outward region to
σ first yield = −126.31+ 482.66f − 692.70f 2 + 389.14f 3 , (15) account for the larger amounts of deformation and the
4
contact situations. The 8-node reduced-integration quadratic density
axisymmetric solid element is used for the billet, and the Bulk forming processes effect complicated distribution
rigid surface is used for the open flat die. of stress, which dominates the cavity evolution in the porous
4.2 Upsetting simulation under frictionless and materials. At the current stage of this study, we assume that
frictional conditions the change in void volume fraction is all brought about by
In order to verify finite element results with experiments, the variation of existing voids and that the nucleation of new
the upsetting processes with two different frictional voids in the tensile stress is negligible.
conditions were considered. The coefficient of friction is In order to demonstrate the fracture prediction in
given as zero in the frictionless analysis, which preserves powder forging, the user's subroutines USDFLD and
uniform deformation of the billets. The results are used to URDFIL in ABAQUS are implemented into the models.
examine whether the inputted experimentally determined What the subroutine can do is completely up to the user. In
material data truly represent the actual bulk material this study, the procedure of simulation with user's subroutine
behavior. Then, the frictional upsetting analysis is carried is shown in Fig. 9. The maximum principal strain energy per
out using an equivalent coefficient of friction, 0.3, which unit of volume in material during deformation is
gives a similar bulged profile to the specimens of frictional accumulated at each integration point by the user's
experiments. subroutine USDFLD. Consecutively, the simulation will be
The compressive true stress-true strain curves and the stopped by the user's subroutine URDFIL as the
true stress-porosity curves with initial relative densities of accumulation of maximum principal strain energy density at
0.6897, 0.8071, and 0.9058 under frictionless conditions are that point reaches the critical value in any time increment,
and this indicates the occurrence of fracture in the
shown in Figs. 6 and 7. The results indicate that finite
element simulations closely match the measured data with workpiece.
slight discrepancies. Fig. 6 depicts that the yield strength and The new forming limit criterion developed in this study
loading capacity of the workpiece deteriorate with the emphasizes the dependence of the initiation of ductile
presence of voids. In Fig. 7, the curves reveal the evolution fracture upon the level of the maximum principal stress. The
of voids during the frictionless upsetting. Void reduction is critical values of the density of maximum principal strain
governed by the volumetric plastic strain of the materials, energy for porous metals with various initial relative
caused by the hydrostatic pressure. Errors between the finite densities at fracture are obtained through the finite element
element results and experiments are less than 5%, which calculation with the new criterion. The critical values of
may have been caused by the slightly non-uniform billets at fracture upsetting are shown in Table 3, where the
deformation of billets in experiments and Luders band right crack initiates on the equatorial surface. The increase of
after the first yield that is not modeled in the finite element voids in the specimens reduces the deformation capability
analysis. and the critical value of maximum principal strain energy
As another example of validation, the same upsetting density at fracture.
process with frictional conditions was computed. The Table 3 Critical value of maximum principal strain energy
load-displacement curves toward the initiation of crack on density at fracture upsetting with various initial relative
the porous billets with initial relative densities of 0.7190, densities
0.8236, and 0.8962 are shown in Fig. 8. Voids decrease the
Initial Height×Diameter Fracture MPSED at fracture Porosity
loading capacity and workability of the workpiece. Friction relative stroke by by FEM at
density [mm×mm] exp.[mm] [MJ/m3] fracture
at contact interfaces of the billet and flat die contributes to
0.7190 15.36×ψ 12.50 -7.85 9.61 (Equat. surf.) 0.1765
the non-uniformity of deformation and density variation that
0.8236 13.19×ψ 12.51 -6.36 13.54 (Equat. surf.) 0.1108
cause the porous billets to bulge out and fracture on the
0.8962 12.22×ψ 12.52 -6.12 21.45 (Equat. surf.) 0.0635
equatorial surface, where larger amounts of voids exist. The
fracture strokes of porous billets with various initial relative 5.2 Fracture upsetting of sintered porous billets
densities, determined from the frictional upsetting tests,
were then utilized to formulate the forming limit criterion in According to the data in Table 3, the critical value of
the finite element simulation. maximum principal strain energy density is exponentially
degraded by the existence of porosity. The exponential decay
function is utilized to fit these data, and they are shown
5. Ductile fracture prediction in upsetting of together in Fig. 10. The critical constant for the fully dense
porous preforms matrix C cr is equal to 34.28 and the material parameter
5.1 Calculation of maximum principal strain energy m is equal to 7.69 in the forming limit curve. Then, this
5
curve is applied to construct the forming limit criterion in (3) Voids lead to the degradation of the strength and
the models. forming limit of a workpiece.
Ductile fracture prediction in upsetting of sintered Through the computer simulation of the forming process
preforms with various initial relative densities was fulfilled with the new ductile fracture criterion, we can then study the
by using the new fracture criterion. For the instance of deformation characteristics and fracture even in forging of
simulation, the fracture upsetting of porous billet with the complex applications, such as gear blanks, connecting rods,
initial relative density of 0.82 under frictional conditions is etc. It would be extremely helpful in improving the
shown in Fig. 11. The crack initiates on the equatorial preparation of sintered preforms, the design of the forging
surface, and then propagates toward the contact sides, which die, and the quality of products. Additionally, in order to
phenomenon is consistent with the evidence in the govern the total volume change of voids accurately, it also
experiments. In addition, the results of fracture upsetting of requires an effective model to consider the mechanism of
porous billets with various initial relative densities that void nucleation in material during forging, to which we will
computed by different fracture criteria are exhibited in Table devote ourselves in the future.
4. However, the position of crack initiation is misled by
other criteria, which are constituted with either the effective Acknowledgements
stress or the hydrostatic component. The analysis by the new
workability criterion is more compatible with the The authors gratefully acknowledge Professor K. S.
experimental data than other forms that proposed previously. Hwang and P/M Laboratory in the Institute of Material
Hence, we can true govern the forming limit of sintered Science and Engineering of NTU for the experimental
porous compacts by the validation of upsetting process. support of the P/M procedure. We also wish to thank Fatigue
and Fracture Laboratory in the NTU Department of
Table 4 Critical value (C.V.) of strain energy density in Mechanical Engineering, which provided us with a material
fracture upsetting of porous billets with various initial testing system that enabled us to ascertain the mechanical
relative densities by different fracture criteria properties of porous metals.
Initial εf εf σ
⎛ *
⎞ εf ⎛ σh ⎞ εf
∫ σ dε = C1 ∫0 σ ⎜⎜ ⎟⎟dε = C2 ∫ ⎜ A+ ⎟dε = B⋅ ρr0 ∫ σ1dε1 = Ccr ⋅e−mf
C
relative
density
0 ⎝σ ⎠ 0
⎝ σ⎠ 0 References
C.V. = 456.60 C.V. = 13.02 C.V. = 0.4790 C.V. = 9.61
0.7190 (Upper corner) (Equat. surf.) (Upper corner) (Equat. surf.) [1] Gurson, A. L., 1977, “Continuum Theory of Ductile Rupture by Void
C.V. = 483.20 C.V. = 19.54 C.V. = 0.4886 C.V. = 13.54 Nucleation and Growth: Part I – Yield Criteria and Flow Rules for
0.8236 (Upper corner) (Equat. surf.) (Upper corner) (Equat. surf.)
Porous Ductile Media,” J. of Eng. Mat. and Tech., Trans. ASME, 99,
C.V. = 620.00 C.V. = 30.62 C.V. = 0.5372 C.V. = 21.54
0.8962 (Upper corner) pp. 2-15.
(Equat. surf.) (Upper corner) (Equat. surf.)
[2] Tvergaard, V., 1981, “Influence of Voids on Shear Band Instabilities
Under Plane Strain Conditions,” I. J. of Frac., 17, pp. 389-407.
[3] Spitzig, W. A., Smelser, R. E., and Richmond, O., 1988, “The
6. Conclusions Evolution of Damage and Fracture0 in Iron Compacts with Various
Initial Porosity,” Acta Metall., 36, pp. 1201-1211.
Powder forging processes enhance the shape precision [4] Biner, S. B., and Spitzig, W. A., 1990, “Densification of Iron
and the strength of products. Forming simulation by the Compacts with Various Initial Porosity Under Hydrostatic Pressure,”
ductile fracture criteria for porous media helps to understand Acta Metall. Mater., 38, pp. 603-610.
[5] Spitzig, W. A., 1990, “Effect of Hydrostatic Pressure on Deformation,
the workability and forging limit of sintered preforms. The Damage Evolution, and Fracture of Iron with Various Initial
deformation characteristics and the prediction of fracture in Porosity,” Acta Metall. Mater., 38, pp. 1445-1453.
[6] Hartley, P., Sturgess, C. E. N., and Rowe, G. W., 1980, “Influence of
forging of sintered compacts were presented in this paper. Friction on the Prediction of Forces, Pressure Distributions and
Experiments and finite element analysis were conducted to Properties in Upset Forging,” I. J. Mech. Sci., 22, pp. 743-753.
establish the models. From the results, we draw the [7] Jha, A. K., and Kumar, S., 1986, “Deformation Characteristics and
Fracture Mechanisms during the Cold Forging of Metal Powder
following conclusions: Preforms,” I. J. Mach. Tool Des. Res., 26, pp. 369-384.
(1) Finite element simulation of fracture upsetting of [8] Hom, C. L., and McMeeking, R. M., 1989, “Void Growth in
Elastic-Plastic Materials,” J. of Appl. Mech., Trans. ASME, 56, pp.
sintered compacts, verified with experiments under 309-317.
both frictionless and frictional conditions, are achieved [9] Huang, C. C., and Cheng, J. H., 2001, “Forging Simulation of
by using the Gurson-Tvergaard yield function and the Sintered Powder Compacts under Various Frictional Conditions,”
Submitted to I. J. Mech. Sci., March.
new ductile fracture criterion for porous media. [10] Keeler, S. P., and Backofen, W. A., 1963, “Plastic Instability and
(2) The new fracture criterion is developed based on the Fracture in Sheets Stretched over Rigid Punches,” Trans. ASM, 56,
pp. 25-48.
maximum principal strain energy density, which [11] Kobayashi, S., 1970, “Deformation Characteristic and Ductile
recognizes the dependence of the domination in Fracture of 1040 Steel in Simple Upsetting of Solid Cylinders and
fracture upon the maximum principal stress.

6
Rings,” J. of Eng. for Indus., Trans. ASME, 92, pp. 391-403. [35] Zhang, X. Q., Peng, Y. H., Li, M. Q., Wu, S. C., and Ruan, X. Y.,
[12] Kobayashi, S., 1976, “Workability of Aluminium Alloy 7075-T6 in 2000, “Study of Workability Limits of Porous Materials under
Upsetting and Rolling,” J. of Eng. for Indus., Trans. ASME, 98, pp. Different Upsetting Conditions by Compressible Rigid Plastic Finite
800-806. Element Method,” J. of Mat. Eng. and Perf., 9, pp. 164-169.
[13] Roy, G. L., Embury, J. D., Edward, G., and Ashby, M. F., 1981, “A [36] Lee, P. W., and Kuhn, H. A., 1973, Metall. Trans., 4, pp. 969-972.
Model of Ductile Fracture Based on the Nucleation and Growth of [37] ABAQUS/Standard User’s Manual, Version 5.8, 2000, Hibbit,
Voids,” Acta Metall., 29, pp. 1509-1522. Karlsson & Sorensen Inc.
[14] Ettouney, O., and Hardt, D. E., 1983, “A Method for In-Process
Failure Prediction in Cold Upset Forging,” J. of Eng. for Indus.,
Trans. ASME, 105, pp. 161-167.
[15] Darvas, Z., 1985, “The Forming Limit and the Fracture Mode in Cold
Upsetting,” Mat. Sci. and Eng., A70, pp. 101-110.
[16] McClintock, F. A., 1968, “A Criterion for Ductile Fracture by the
Growth of Holes,” J. of Appl. Mech., Trans. ASME, 35, pp. 363-371.
[17] Sowerby, R., and Chandrasekaran, N., 1986, “The Prediction of
Damage Accumulation when Upsetting AISI 1045 Steel Specimens,
Based on McClintock’s Model,” Mat. Sci. and Eng., 79, pp. 27-35. Porous specimens preparation
● Powder compaction
[18] Rice, J. R., and Tracey, D. M., 1969, “On the Ductile Enlargement of ● Sintering

Voids in Triaxial Stress Fields,” J. Mech. Phys. Solids, 17, pp.


Frictionless compression test of porous billets
201-217. Uniaxial tension test of fully densifized sheets

[19] Vujovic, V., and Shabaik, A. H., 1986, “A New Workability Criterion
for Ductile Metals,” J. of Eng. Mat. and Tech., Trans. ASME, 108, pp. Frictional upsetting to
bulge out and crack
Acquire material parameters
for finite element analysis
245-249.
[20] Freudenthal, A. M., 1950, The Inelastic Behaviour of Solids, Wiley,
Verify FEM with
New York. compression test

[21] Cockcroft, M. G., and Latham, D. J., 1968, “Ductility and the
Workability of Metals,” J. of Inst. of Met., 96, pp. 33-39. Frictional upsetting simulation
Capture deform mechanism
[22] Brozzo, P., Deluca, B., and Rendina, R., 1972, “A New Method for
the Prediction of Formability Limits in Metal Sheets. Sheet Metal Develop fracture criteria
Forming and Formability,” Proceedings of Seventh Biannual Conf. of Predict forging limit

I. Deep Drawing Res. Group.


[23] Norris, D., Reaugh, J., Moran, B., and Quinnonesa, D., 1978, “A Fig. 1 Flow chart of this research
Plastic Strain Mean-Stress Criterion for Ductile Fracture,” J. of Eng.
Mat. and Tech., Trans. ASME, 100, pp. 279-286.
[24] Oyane, M., Sato, T., Okimoto, K., and Shima, S., 1980, “Criteria for 425
400
Ductile Fracture and their Applications,” J. of Mech. Work. Tech., 4, 375
350
pp. 65-81.
True stress [MPa]

325
[25] Clift, S. E., Hartley, P., Sturgess, C. E. N., and Rowe, G. W., 1990, 300
275
“Fracture Prediction in Plastic Deformation Processes,” I. J. of Mech. 250
225
Sci., 32, pp. 1-17. 200
[26] Gouveia, B. P. P. A., Rodrigues, J. M. C., and Martins, P. A. F., 1996, 175
150
“Fracture Predicting in Bulk Metal Forming,” I. J. of Mech. Sci., 38, 125

pp. 361-372. 100


75
[27] Gouveia, B. P. P. A., Rodrigues, J. M. C., and Martins, P. A. F., 2000, 50 Uni-axial tension test of matrix
25
“Ductile Fracture in Metalworking: Experimental and Theoretical 0
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22
Research,” J. of Mat. Proc. Tech., 101, pp. 52-63.
[28] Wifi, A. S., Abdel-Hamid, A., and El-Abbasi, N., 1998, True strain
“Computer-Aided Evaluation of Workability in Bulk Forming Fig. 2 Tensile true stress-strain curve of the fully dense matrix by
Processes,” J. of Mat. Proc. Tech., 77, pp. 285-293. experiment
[29] Jain, M., Allin, J., and Lloyd, D. J., 1999, “Fracture Limit Prediction
Using Ductile Fracture Criteria for Forming of an Automotive
Aluminum Sheet,” I. J. of Mech. Sci., 41, pp. 1273-1288.
[30] Takuda, H., Mori, K., and Hatta, N., 1999, “The Application of Some Cylindrical billet
Criteria for Ductile Fracture to the Prediction of the Forming Limit of
Sheet Metals,” J. of Mat. Proc. Tech., 95, pp. 116-121.
[31] Downey, C. L., and Kuhn, H. A., 1975, “Application of a Forming
Limit Concept to the Design of Powder Preforms for Forging,” J. of
Eng. Mat. and Tech.,Trans. ASME, 97, pp. 121-125.
[32] Tabata, T., and Masaki, S., 1977, “A Fracture Criterion for Porous Diameter gauge
Upper flat die
Materials and its Application to the Shape of Sintered Preforms in
Forging,” J. of Eng. Mat. and Tech., Trans. ASME, 99, pp. 16-22.
[33] Abdel-Rahman, M., and El-Sheikh, M. N., 1995, “Workability in
Forging of Powder Metallurgy Compacts,” J. of Mat. Proc. Tech., 54,
pp. 97-102.
[34] Lee, J. H., and Zhang, Y., 1996, “A Finite-Element Work-Hardening
Lower flat die
Plasticity Model of the Uniaxial Compression and Subsequent Failure
of Porous Cylinders Including Effects of Void Nucleation and Fig. 3 The construction of diameter gauge in compression test
Growth – Part II: Localization and Fracture Criteria,” J. of Eng. Mat.
and Tech., Trans. ASME, 118, pp. 169-178.

7
First yield porosity Displacement [mm]
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
0 0

-10
-20
First yield stress [MPa]

-20
-40
-30
-60 -40

Load [KN]
-80 -50

-100 -60

Exp. (ργ=0.7190)
-70
-120
FEM -80
-140 Exp. (ργ=0.8236)
Experiment data FEM -90
-160 Exp. (ργ=0.8962)
3rd order polynomial fitting curve -100
FEM
2 3
-180 σ = -126.31 + 482.66 f - 692.70 f + 389.14 f Fracture point -110

-200 -120

Fig. 4 Compression of first yield stress-porosity data by Fig. 8 Compression of load-displacement curves toward fracture
experiment and the non-linear fitting curve under frictional conditions by experiment and FEM
Develop fracture criterion of porous materials
Upsetting flat die ε
∫ σ 1 d ε 1 = C cr ⋅ e − mf
f

Accumulation of maximum principal strain


energy density at each integration point by
user's subroutine *USDFLD
z
Complete every time increment
during forging simulation
Biased mesh * URDFIL
of 1/4 of billet
Check with critical value No
at integration point
r
Yes
Fracture occurs in material
Boundary conditions on finite element program stop running by
the line of symmetry user's subroutine *URDFIL

Output the position at fracture


in the workpiece

Fig. 5 Finite element mesh and boundary conditions of Fig. 9 Flow chart of ductile fracture prediction during forging
cylindrical billet in the upsetting analysis with the user's subroutine USDFLD and URDFIL

True strain 40
-0.5 -0.4 -0.3 -0.2 -0.1 0.0
0 35
strain energy density

Critical value by FEM


Maximum principal

Exponential decay fitting curve


Exp. (ργ=0.6897)
30 MPSED = 34.28 exp( -7.69 * f )
FEM -50
True stress [MPa]

Exp. (ργ=0.8071)
25
FEM -100
Exp. (ργ=0.9058)
FEM 20
-150
15

-200
10

-250 5

-300 0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

-350 Porosity at fracture

Fig. 6 Compression of true stress-strain curves under frictionless Fig. 10 Data of the critical value of maximum principal strain
conditions by experiment and FEM energy density-porosity at fracture and the fitting curve

Porosity
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Initial height of billet H0= 13.19 [mm]
0 Open flat die upsetting U= - 6.36 [mm]

-50
True stress [MPa]

-100

-150

-200
Exp. (ργ=0.6897)
-250 FEM
Exp. (ργ=0.8071)
FEM
-300 Exp. (ργ=0.9058) Crack on the
FEM
equatorial surface
-350

Fig. 7 Compression of true stress-porosity curves under Fig. 11 Crack initiation in fracture upsetting of porous billet
frictionless conditions by experiment and FEM under frictional conditions by experiment and FEM

You might also like