The Finite Element Method for the Analysis of Non-Linear ......Figure 1.1: Uniaxial Response of an...
Transcript of The Finite Element Method for the Analysis of Non-Linear ......Figure 1.1: Uniaxial Response of an...
The Finite Element Method for the Analysis ofNon-Linear and Dynamic Systems
Prof. Dr. Eleni Chatzi
Lecture ST1 - 19 November, 2015
Institute of Structural Engineering Method of Finite Elements II 1
Constitutive Relations
Overview so far
Material Nonlinearity
Large Displacements
Dynamic Analysis
Special Topics
Material Laws
The Contact problem
Fracture & Special Formulations (XFEM, SBFEM)
Institute of Structural Engineering Method of Finite Elements II 2
Constitutive Relations
Overview so far
Material Nonlinearity
Large Displacements
Dynamic Analysis
Special Topics
Material Laws
The Contact problem
Fracture & Special Formulations (XFEM, SBFEM)
Institute of Structural Engineering Method of Finite Elements II 2
Constitutive Relations
Previously we examined the kinematic equations formulation(displacement, strain displacement relations)
The next step is to determine appropriate constitutive relationshipsof the form:
Ļ = f(Īµ)
ex. linear analysis in 1D ā Ļ = EĪµ
When dealing with higher dimensions & incremental analysis, thisis written in tensor form for time t:
tĻ = tCijrstĪµrs
Institute of Structural Engineering Method of Finite Elements II 3
Constitutive Relations
It is necessary that kinematic and constitutive relations areappropriate. Previously we saw that in the large displacementformulation appropriate tensors need to be defined.e.g. TL Formulation ā Second Piola-Kirchhoff stresstensor, Green Lagrange strain tensor).
Therefore a problem involving large strains should also becombined with a material law that admits large strains.
However, we might be examining a problem of large displacements with
small strains. In that case we can still use the material laws defined for
classic engineering stress and strain measures (for small displacements) but
this time combined with the SP-K stress and G-L strain tensors.
Institute of Structural Engineering Method of Finite Elements II 4
Notation
Main Stress - Strain pairs:
Material Nonlinearity (small deformations)Engineering (or Nominal) Stress ĻEngineering Strain Īµ
TL formulation (large deformations)2nd Piola-Kirchhoff Stress SGreen-Lagrange Strain Īµ
UL formulation (large deformations)Cauchy (or True) Stress ĻAlmansi Strain ĪµA
Note that: Ļ =L
L0Ļ, Īµ =
Lā L0
L0
Institute of Structural Engineering Method of Finite Elements II 5
Solution Flowchart
General Solution process in incremental nonlinear FE
Known Solution at t:
Stresses šš” , strains šŗš” ,
Internal material parameters šæš”
Calculate at t+Īt:
Stresses ššāšš”+Īš”
Tangent stress strain matrix šŖšāš
Internal material parameters šæšāšš”+Īš”
ā¢ Elastic Analysis: directly obtain
ššāšš”+Īš” ,šŖšāš from šŗšāšš”+Īš”
ā¢ Inelastic Analysis: Integrate to get
ššāšš”+Īš” = š + ā« š šš”+Īš”šāšš
š”
š¼š = š¼šāš +š”+Īš”š”+Īš” Īš¼i
Calculate:
Incremental Displacement Vector Īš¼i:
š²šāšš”+Īš” Īš¼i = š¹š”+Īš” ā ššāšš”+Īš”
Then,
Known Quantities at iterations i- 1 :
Nodal Displacements at first Iteration: š¼šāšš”+Īš”
and hence
Element strains šŗšāšš”+Īš”
Repa
t til
l Con
verg
ence
Institute of Structural Engineering Method of Finite Elements II 6
Overview of Material Descriptions
We can discriminate amongst the following major classes ofmaterial behavior
Elastic, linear or nonlinear
Hyperelastic
Hypoelastic
Elastoplastic
Creep
Viscoplastic
Institute of Structural Engineering Method of Finite Elements II 7
Elastic Material
For an elastic material the stress is a function of strain only.The stress path is the same both in loading and unloading
Linear ElasticThe elasticity (constitutive)tensor components, Cijrs areconstant
Nonlinear ElasticThe elasticity (constitutive)tensor components, Cijrs area function of strain
š
šŗ
Example: Almost all materials under small stress
Institute of Structural Engineering Method of Finite Elements II 8
Elastic Material
(a) (b)
s s
e e
Figure 1.1: Uniaxial Response of an Elastic Material: (a) Linear, (b) Nonlinear
1.2.1 Cauchy Elastic Models
Some nonlinear elastic models are based on the general premise that for an elastic materialthe current state of Cauchy stress is a function of the current state of strain, and not of thehistory of strain. That is, Ļij = Fij (Īµkl), where Fij is some nonlinear function. Once thiscondition is satisfied, however, the inverse relation does not necessarily exist. Materials thatsatisfy this fundamental requirement of an elastic body are called Cauchy elastic materials.
More specifically, the general functional form relationship for Cauchy elastic models iswritten in indicial form as
Ļij = Ī±0Ī“ij + Ī±1Īµij + Ī±2ĪµimĪµmj + Ī±3ĪµimĪµmnĪµnj + Ā· Ā· Ā· (1.1)
or, in direct formĻ = Ī±0I + Ī±1Īµ+ Ī±2Īµ
2 + Ī±3Īµ3 + Ā· Ā· Ā· (1.2)
where I is the second-order identity tensor. The Ī±i (i = 1, 2, 3, Ā· Ā· Ā· ) represent model parame-ters whose values are determined from the results of laboratory experiments. Further detailspertaining to such parameter determination are given in the following example.
.Example 1.1: First-Order Cauchy Elastic Model
A first order Cauchy model corresponds to the case of isotropic linear elasticity. As such,only the first-order terms in equation (1.1) are retained, giving
Ļij = Ī±0Ī“ij + Ī±1Īµij (1.3)
In order to determine the values of the parameters Ī±0 and Ī±1, consider two simple lab-oratory experiments. First consider simple shear deformation. Since there is no volumetricstrain, and since only one nonzero component of strain exists (say Īµ12 = Īµ21), it follows that
Ļ12 = Ī±1Īµ12 = Ī±1Ī³122
(1.4)
Solving for Ī±1 gives
2
Institute of Structural Engineering Method of Finite Elements II 9
Elastic Material
For the case of an elastic material we already saw that the TLFormulation (used for large deformation analysis) yields:
t0Sij = t
0Cijrst0Īµrs
The elasticity tensor for 3D stress conditions is defined as:
tCijrs = Ī»Ī“ijĪ“rs+ Āµ(Ī“irĪ“js+ Ī“isĪ“jr)
where Ī» and Āµ are the Lame constants and Ī“ij is the Kronecker delta,
Ī» =EĪ½
(1 + Ī½)(1ā 2Ī½), Āµ =
E
2(1 + Ī½)(homogeneous isotropic material)
Ī“ij =
{0 i 6= j1 i = j
Institute of Structural Engineering Method of Finite Elements II 10
Elastic Material
Important NoteāThe 2nd Piola-Kirchhoff (PK2) stress and Green-Lagrange straintensor components are invariant to rigid body motions.ā
For problems with small strains we can take advantage of thisobservation and use any constitutive relationship that has beendeveloped for engineering stress and strain measures by justsubstituting with the PK2 stress and Green-Lagrange strain
This observation can be extended to all problems with largedeformations but small strain conditions such as the elastic orelastoplastic buckling problem and the collapse analysis of slenderstructures.
Institute of Structural Engineering Method of Finite Elements II 11
Hyperelastic Material
Hyperelastic (rubberlike) materialsexhibit an incompressible response(volume preserving), pathindependence and no energydissipation.
The stress is now calculated throughthe strain energy functional W
t0Sij =
āW
āt0Īµij
Figure: Stress-strain curves for varioushyperelastic material models.
Institute of Structural Engineering Method of Finite Elements II 12
Hyperelastic Material
Hyperelastic Material Models
Saint Venant-Kirchhoff model
W (Īµ) =Ī»
2[tr(Īµ)]2 + Āµtr(Īµ2)
and the second Piola-Kirchhoff stress can be derived as
S = Ī»[tr(Īµ)]I + 2ĀµĪµ
Ī», Āµ are the Lame constants
Mooney-Rivlin model
W (Īµ) = C1(I1 ā 3) + C2(I2 ā 3)
where C1 and C2 are empirically determined material constants and
I1 = tr(C) = C11 + C22 + C33
where C is the Cauchy-Green deformation tensor (see Lecture 4) and
I2 =1
2[(I1)2 ā tr(C)2]
Institute of Structural Engineering Method of Finite Elements II 13
Inelasticity
Elastoplasticity, Creep and Viscoplasticity are types of Inelasticbehavior
Elastic behavior ā stresses can be directly calculated from the strain
Inelastic behavior ā the stress at time t depends on the stress strainhistory
In the incremental analysis of inelastic response we had three main scenarios
Small displacements-rotations / small strains ā use linear elasticsolution, engineering stress and strain measures
Large displacements-rotations / small strains ā use TL formulationby substituting the appropriate stress - strain measures (PK2,Green-Lagrange) in the place of the engineering stress and strainmeasures
Large displacements-rotations / large strains ā use either TL or ULformulation, more complex constitutive laws
Institute of Structural Engineering Method of Finite Elements II 14
Elastoplasticity
In this formulation we encounter a linearly elastic behavior until yieldand usually a hardening post yield behavior
Examples: Metals, soild and Rocks when subjected to high stresses.Institute of Structural Engineering Method of Finite Elements II 15
Elastoplasticity
The strain and stress increments are given by:
dĪµrs = dĪµE
rs + dĪµP
rs dĻij = CE
ijrs (dĪµrs ā dĪµP
rs)
where CE
ijrs are the components of the elastic constitutive tensor and dĪµrs, dĪµE
rs,
dĪµP
rs are the components of the total strain increment.To calculate the plastic strains we use the following three properties:
Yield Function fy(Ļ, ĪµP)
fy < 0 ā Elastic behaviorfy >= 0 ā Plastic or elastic behavior depending on the loading condition
Flow ruleThe yield function is used in the flow rule in order to obtain the plasticstrain increments
dĪµP
ij = Ī»āfyāĻij
Ī» is a scalar to be determined
Hardening ruleThis specifies how the yield function is modified during the progression ofloading.
Institute of Structural Engineering Method of Finite Elements II 16
Elastoplasticity
Example: Von Mises yield criterion (in 3D):
fy = 0ā (Ļ11 ā Ļ22)2 + (Ļ22 ā Ļ33)
2 + (Ļ11 ā Ļ33)2 + 6(Ļ2
12 + Ļ223 + Ļ2
31)ā 2Ļ2y = 0
Institute of Structural Engineering Method of Finite Elements II 17
Elastoplasticity
Isotropic & Kinematic hardening Rules
In the case of isotropic hardening, the yield surface expandsuniformly.
In the case of kinematic hardening, the size of the yield surfaceremains unchanged and the center location of the yield surfaceis shifted. (Bauschinger effect)
Institute of Structural Engineering Method of Finite Elements II 18
Elastoplasticity
Cyclic Loading
PLASTICITY. Flow rule for kinematic hardening
Obviously, for a reversed loading process like the one in the cyclic loading diagram of Fig. 1, the
isotropic hardening will lead to a cyclic test behaviour according to the solid line OABCDE of Fig. 2
(in which the length of line segment BC is the same as that of line segment AB). It is, however, a well-
established fact that in most materials there is a Bauschinger effect, by which a reversed loading will
E
t
D
C
B
A
s
s
O
Fig. 1 Cyclic loading
Dā
Cā
A,E
B,D
C
O
Fig. 2 diagram of uniaxial cyclic test.
Isotropic hardening
Kinematic hardening
Hardening
PLASTICITY. Flow rule for kinematic hardening
Obviously, for a reversed loading process like the one in the cyclic loading diagram of Fig. 1, the
isotropic hardening will lead to a cyclic test behaviour according to the solid line OABCDE of Fig. 2
(in which the length of line segment BC is the same as that of line segment AB). It is, however, a well-
established fact that in most materials there is a Bauschinger effect, by which a reversed loading will
E
t
D
C
B
A
s
s
O
Fig. 1 Cyclic loading
Dā
Cā
A,E
B,D
C
O
Fig. 2 diagram of uniaxial cyclic test.
Isotropic hardening
Kinematic hardening
Institute of Structural Engineering Method of Finite Elements II 19
Thermoelastoplasticity and Creep
This behavior exhibits time effect of increasing strains under constant loadsor decreasing stress under constant deformations (relaxation)Typical examples of such behavior are metals at high temperatures
The thermal strain(Īµ = Ī±āT ) and the creep strain now enter the
formulation of the stress strain relationships.Institute of Structural Engineering Method of Finite Elements II 20
Viscoplasticity
Viscoplasticity describes the rate-dependent inelastic behavior ofsolids. Rate-dependence in this context means that the deformationof the material depends on the rate at which loads are applied. Theinelastic behavior that is the subject of viscoplasticity is plasticdeformation which means that the material undergoes unrecoverabledeformations when a load level is reached. Rate-dependent plasticityis important for transient plasticity calculations.
The main difference between rate-independent plastic andviscoplastic material models is that the latter exhibit not onlypermanent deformations after the application of loads but continueto undergo a creep flow as a function of time under the influence ofthe applied load.
Typical examples of such behavior are Polymers and Metals
Institute of Structural Engineering Method of Finite Elements II 21
Viscoplasticity
Institute of Structural Engineering Method of Finite Elements II 22