Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword
Like this
2Activity
0 of .
Results for:
No results containing your search query
P. 1
Micromechanical modelling of polycrystalline materials under non-proportional loading paths

Micromechanical modelling of polycrystalline materials under non-proportional loading paths

Ratings: (0)|Views: 83|Likes:

More info:

Published by: Ryszard B. Pecherski on May 12, 2009
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

05/11/2014

pdf

text

original

 
Volume 52 2007 Issue 2
K. KOWALCZYK-GAJEWSKA
, Z. MRÓZ
, R. B. PĘCHERSKI
,
∗∗
MICROMECHANICAL MODELLING OF POLYCRYSTALLINE MATERIALS UNDER NON-PROPORTIONALDEFORMATION PATHSMODELOWANIE MIKROMECHANICZNE MATERIAŁÓW POLIKRYSTALICZNYCHDLA NIEPROPORCJONALNYCH ŚCIEŻEK DEFORMACJI
The rigid-plastic crystal plasticity model with single yield surface of 2n-degree is applied to simulate the polycrystallinebehaviour and the crystallographic texture development under non-proportional deformation paths. The role of two controllingparameters: the amplitude and frequency for the processes of tension or compression assisted by cyclic torsion of thin-walledtubes made of copper is analysed. The effect of micro-shear bands on the reduction of global hardening rate is described bymeans of the contribution function of shear banding in the rate of plastic deformation. The conclusions drawn from the studycan find also application in the extension of the analysis for high strength and hard deformable materials.Przeprowadzono symulacje zachowania się polikryształu oraz rozwoju tekstury krystalograficznej dla nieproporcjonalnychścieżek deformacji wykorzystując sztywno-plastyczny model plastyczności kryształu z powierzchnią plastyczności stopnia 2
n
.Przeanalizowano wpływ dwóch parametrów kontrolnych: amplitudy i częstości dla procesów rozciągania lub ściskania przyudziale cyklicznego skręcania cienkościennych rurek wykonanych z miedzi. Efekt działania mikropasm ścinania w postaci re-dukcji globalnego modułu wzmocnienia odkształceniowego został opisany poprzez funkcję udziału pasm ścinania w przyrościedeformacji plastycznej. Wnioski wysnute z powyższego studium mogą znaleźć docelowo zastosowanie w rozszerzeniu analizydla materiałów o podwyższonej wytrzymałości oraz materiałow trudno odkształcalnych.
1. Motivation
Metal forming processes such as extrusion assistedby cyclic torsion have demonstrated essential advantageswith respect to classical forming processes. The signif-icant reduction of required load for forming, growth of ductility, possible reduction of dissipated energy of form-ing and finer grain structure are the main beneficial fac-tors, cf. e.g. [5, 10, 11]. One of possible explanationsof these advantages is the appearance of a multi-scalehierarchy of shear localization modes, so-called shearbanding, that replaces the crystallographic multiple slipor twinning and reduces strain hardening. In the paperthe constitutive model developed in [13] will be appliedto study macroscopic behaviour of polycrystalline mate-rials that are subjected to non-proportional deformationpaths as so called KOBO-type processes related withcyclic torsion [10]. Plastic anisotropy evolution connect-ed with crystallographic texture development will be alsostudied.In the next section the concise description of theconstitutive model is given. More information can befound in [12] and [13]. In section 3 the results of nu-merical simulation are presented and discussed. Finally,in the last section the conclusions are outlined.
2. The constitutive modelling2.1. The single grain model with the yield surfaceof 2n-degree
The rigid-plastic model is considered. In order tostudy the texture development as well as the polycrys-talline behaviour for the advanced plastic deformationprocess the large strain formulation of the crystal plastic-ity theory has to be applied [3]. We use the multiplicativedecomposition of the total deformation gradient into theelastic part and the plastic part
F
 p
. The assumption ismade that the elastic stretches are negligible comparing
INSTITUTE OF FUNDAMENTAL TECHNOLOGICAL RESEARCH PAS, 00-049 WARSZAWA, 21 ŚWIĘTOKRZYSKA STR., POLAND
∗∗
INSTITUTE OF STRUCTURAL MECHANICS, FACULTY OF CIVIL ENGINEERING, CRACOW UNIVERSITY OF TECHNOLOGY, 31-155 KRAKÓW, 24 WARSZAWSKA STR., POLAND
 
to the analysed plastic strains so that the elastic part isrestricted to the rigid rotation
R
. It is easy to showthat such form of the deformation gradient results in theadditive decomposition of the velocity gradient
L
intothe elastic spin
, the plastic spin
 p
and the rate of plastic deformation tensor
D
 p
(the symmetric part of theplastic part of the velocity gradient)
L
=
+
 p
+
D
 p
.
(2.1)It is generally accepted that during the elastic regimecrystallographic lattice and the material undergo thesame deformation while plastic deformation does not in-duce lattice motion which is described only by the elasticspin. Plastic deformation occurs by slip in the specifiedcrystallographic direction
m
on the specified crystallo-graphic plane with the unit normal
n
. The number
andtype of slip systems
{
n
,
m
}
depend on the lattice type. Inthis paper the numerical simulations are performed forthe f.c.c. crystals with 12 slip systems
{
111
}
<
110
>
.The plastic part of the velocity gradient is then describedas follows
L
 p
=
 p
+
D
 p
=
 M 
=
1
˙
γ 
m
n
(2.2)In the classical S c h m i d multi-surface plasticitythe slip on the considered slip system is initiated if theresolved shear stress
τ
=
m
·
σ
·
n
reaches the criticalvalue
τ
c
being the material parameter. Instead of thisformulation the regularized S c h m i d law (the crys-tal plasticity with single yield surface) is used [12]. Inview of this model the crystallographic slip mechanismis activated when the following yield condition for theC a u c h y stress
σ
is fulfilled
 f 
(
σ
)
=
12
n
 M 
=
1
τ
τ
c
2
n
m
=
0
,
(2.3)where
n
is the positive exponent and
m
is the materialparameter independent of the lattice orientation [7]. Therate of plastic deformation tensor is associated with thisyield condition
D
 p
=
˙
λ
 M 
=
1
1
τ
c
τ
τ
c
2
n
1
P
(2.4)and for the plastic spin
 p
the following specification isused
 p
=
˙
λ
 M 
=
1
1
τ
c
τ
τ
c
2
n
1
W
,
(2.5)where
P
and
W
are symmetric and skewsymmetricparts of 
m
n
diad, respectively. Comparing the aboveflow rules with the kinematic description (2.2) one mayeasily identify that˙
γ 
=
˙
λ
1
τ
c
τ
τ
c
2
n
1
.
(2.6)In its mathematical structure this model is similar to therate-dependent formulation [4], however, the referencevelocity, which is the material parameter in the latterapproach, is here replaced by the plastic multiplier˙
λ
obtained from the consistency condition for the yieldsurface (2.3). The explicit formula for˙
λ
may be foundin [7]. In the case of deformation controlled process itmay be calculated from the formula˙
λ
=
D
 p
 M 
=
11
τ
c
τ
τ
c
2
n
1
P
,
where
A
=
√ 
A
·
A
=
 
 A
i
 A
i
.Slip occurs by motion of dislocations that have toovercome both short-range and long-range obstacles. Ac-cumulation of dislocations makes its further movementmore and more difficult. This qualitative description of physical nature of slip is captured in the framework of crystal plasticity by the hardening rule for the critical re-solved shear stress
τ
c
. In this study we use the formulaproposed in [3]˙
τ
c
=
 M 
q
=
1
h
rq
|
˙
γ 
q
|
,
(2.7)where hardening matrix
h
rq
enables to account for thelatent and self-hardening by setting the parameter
q
1in its specification
h
rq
=
h
q
(
q
+
(1
q
)
δ
rq
)
.
For the function
h
q
responsible for the self-hardening weassume the following equation [2]
h
q
=
h
o
1
τ
qc
τ
sat 
 β
(2.8)with the material parameters
{
h
o
,τ
qc
0
,τ
sat 
,β
}
that have tobe identified in the experiments. Here
τ
c
0
are the initialvalues of 
τ
qc
.Most of the results discussed in this paper is ob-tained applying the presented model. The material pa-rameters have been identified with use of the experimen-tal results for polycrystalline copper reported in [9], [1].They have been obtained by fitting the stress-strain curvefor the simple compression. More details concerning theidentification procedure may be found in [14].
 
2.2. Simplified description of the reductionof the hardening rate due to micro shear bands
General conclusion that follows from experiments[5, 10, 11] is that by persistent changing of the defor-mation path it is possible to reduce significantly theglobal strain hardening rate. The inventors of KOBOmethod explained their pivotal idea as follows. Expe-rience gained from the conducted tests allow them to ar-gue that the change of mode of plastic flow destabilizesthe material substructure that has been developed andleads to the initiation of the second mode of deforma-tion. The homogeneous crystallographic slip is graduallyreplaced by the localized heterogeneous deformation bymicro-shear bands.In order to account for the reduction of the hardeningrate due to appearance of micro-shear bands we followthe concept introduced in [18]. The plastic velocity gra-dient
L
 p
is decomposed into two parts: one connectedwith the crystallographic slip
L
 pslip
and one connectedwith micro-shear bands
L
 p MS
. For the crystallographicslip the rate-independent single grain model presentedin the previous subsection is used. The constitutive ap-proach for the part connected with the micro-shear bandswas thoroughly described in the series of papers [13,18–21].In the present analysis the simplified model describ-ing the effect of micro-shear bands is applied. The yieldcondition for initiation of micro-shear bands is the sameas the yield condition for slip mechanism (2.3). In [13]two flow rules for micro-shear bands were considered.In this paper we focus on the model in which the flowrule for the part of the rate of plastic deformation
D
 p MS
connected with micro-shear bands is associated with theyield condition, so both tensors
D
 p MS
and
D
 pslip
have thesame directions. Using the following notation:
D
 p
=
,
D
 pslip
=
slip
,
D
 p MS
=
 MS
 MS
=
 MS
the total rate of plastic deformation reads
D
 p
=
˙
λ
1
 MS M 
=
1
1
τ
c
τ
τ
c
2
n
1
P
,
 MS
<
0
,
1) (2.9)and the constitutive function
 MS
describes the contribu-tion of shear banding in the rate of plastic deformation.Since the hardening rate depends on the incrementof crystallographic slip (Eqs (2.6) and (2.7)), the highercontribution of micro-shear bands the higher is the re-duction of the global strain hardening rate. According tothe invoked experimental results the reduction of strainhardening is more pronounced for plane strain com-pression than for the simple compression [9]. Followingthis observation the
 MS
function of logistic dependenceon accumulated plastic strain proposed and identified in[19], [16], was extended in [13] to incorporate the effectof the type of strain path so that the following formulawas proposed
 f 
 MS
(
 ξ 
)
=
 f 
 MS
1
+
exp(
a
b
 ξ 
)
,
where
 ξ 
=
32
ε
 peq
(1
α
|
cos3
θ 
|
)
,
0
α
1 (2.10)and˙
ε
 peq
=
23
D
 p
·
D
 p
,
|
cos3
θ 
|
=
3
√ 
6
|
det(
 µ
D
 p
)
|
,
 µ
D
 p
=
D
 p
D
 p
.
The coefficient
α
is an additional material parameter thatdescribes the following ratio
α
=
1
ξ 
cos3
θ 
=
1
 ξ 
cos3
θ 
=
0
for the same value of the accumulated plastic strain fortwo extreme types of strain paths. In the next section thenon-proportional strain paths will be analysed, therefore,contrary to the analysis in [13] the distinction has to bemade between magnitude of plastic strain
E
 p
and theaccumulated plastic strain
ε
 peq
. For the purpose of thisanalysis the second one must be used.For the plastic spin connected with micro-shearbands we use the rule expressed in [6] that relates theplastic spin to the non-coaxiality factor between corre-sponding the plastic strain rate tensor and the C a u c h ystress
 p MS
=
ω
¯
 f 
 MS
D
 p MS
σ
σ
D
 p MS
D
 p MS
σ
σ
D
 p MS
,
(2.11)where
 p
=
ω,
 p MS
=
ω
 MS
¯
 f 
 MS
=
ω
 MS
ω.
For the second constitutive function¯
 f 
 MS
connectedwith micro-shear bands two possibilities are considered,namely¯
 f 
 MS
=
0
,
and¯
 f 
 MS
=
 MS
.
(2.12)In the first case there is no plastic spin due to micro-shearbanding and in the second case the contribution of 

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->