Lecture3 PDF

Institute of Structural Engineering Page 1
Chapter 3
Variational Formulation &

the Galerkin Method
Method of Finite Elements I

Today’s Lecture Contents:
•  Introduction
•  Differential formulation
•  Principle of Virtual Work
•  Variational formulations
•  Approximative methods
•  The Galerkin Approach

The differential form of physical processes

Physical processes are governed by laws (equations) most probably
expressed in a differential form:

•  The axially loaded bar equation:
•  The isotropic slab equation:
•  The Laplace equation in two

dimensions:
(e.g. the heat conduction problem)


Focus: The axially loaded bar example.
Consider a bar loaded with constant end load R. Strength of Materials Approach

R
Equilibrium equation
R
f ( x) = R ⇒ σ ( x ) =
A
x
Constitutive equation (Hooke’s Law)
σ ( x)
R
ε ( x) = =
R E AE
f(x)
Kinematics equation
L-‐x
δ (x)
ε ( x) =
x
Given: Length L, Section Area A, Young'ʹs modulus E
Find: stresses and deformations.

Rx
Assumptions:
δ ( x) =
The cross-‐‑section of the bar does not change after loading. AE
The material is linear elastic, isotropic, and homogeneous.
The load is centric.
End-‐‑effects are not of interest to us.


Focus: The axially loaded bar example. The Differential Approach
Consider and infinitesimal element of the bar: Equilibrium equation
Δσ dσ
Aσ = A(σ + Δσ ) ⇒ A lim
! =0⇒ A =0
Δx→0 Δx dx
Constitutive equation (Hooke’s Law)
σ = Eε
Kinematics equation
du
ε=
dx
d 2u
AE 2 = 0 Strong Form
dx
Boundary Conditions (BC)
Given: Length L, Section Area A, Young'ʹs modulus E
Find: stresses and deformations. u(0) = 0 Essential BC
σ ( L) = 0 ⇒

Assumptions:
The cross-‐‑section of the bar does not change after loading.
du
The material is linear elastic, isotropic, and homogeneous. AE =R Natural BC
The load is centric. dx x= L
End-‐‑effects are not of interest to us.


d 2u
R
dx
u(0) = 0 Essential BC
x du
AE =R Natural BC
dx x= L
Definition
The strong form of a physical process is the well posed
set of the underlying differential equation with the
accompanying boundary conditions



d 2u
R
dx
x du
AE =R Natural BC
dx x= L
v This is a homogeneous 2nd order ODE with known solution:
Analytical Solution:
du(x)
u(x) = uh = C1x + C2 & ε (x) = = C1 = const!
dx



d 2u
R
dx
x du
AE =R Natural BC
dx x= L
C1 ,C2
To fully define the solution (i.e., to evaluate the values of parameters )
we have to use the given boundary conditions (BC):
C2 = 0
u(x) = uh = C1x + C2 ⎯du
⎯⎯⎯
u(0)=0
R
→ R
dx
=
EA C1 =
x= L
EA
Rx
⇒ u( x) = v  Same as in the mechanical approach!
AE

The Strong form – 2D case

A generic expression of the two-‐‑dimensional strong form is:

and a generic expression of the accompanying set of boundary conditions:

: Essential or Dirichlet BCs

: Natural or von Neumann BCs

Disadvantages
The analytical solution in such equations is
i.  In many cases difficult to be evaluated
ii.  In most cases CANNOT be evaluated at all. Why?
•  Complex geometries
•  Complex loading and boundary conditions

Deriving the Strong form – 3D case
su : supported area with prescribed displacements U su

sf
s f : surface with prescribed forces f
f B : body forces (per unit volume)
U : displacement vector
ε : strain tensor (vector)
σ : stress tensor (vector)


Kinematic Relations
ε = LU ⎡ ∂ ⎤
⎢ 0 0 ⎥
⎢ ∂x ⎥
⎢ ∂ ⎥
⎢ 0 0 ⎥
⎢ ∂y ⎥
⎢ ∂ ⎥
⎡ U ⎤ ⎢ 0 0 ⎥
∂z
U = ⎢ V ⎥, L=⎢ ⎥ , ε T = ⎡ ε xx ε yy ε zz 2ε xy 2ε yz 2ε xz ⎤
⎢⎣ ⎥⎦
⎢ ⎥ ⎢ ∂ ∂ ⎥
⎢⎣ W ⎥⎦ ⎢ 0 ⎥
∂y ∂x
⎢ ⎥
⎢ ∂ ∂ ⎥
⎢ 0 ⎥
∂z ∂y
⎢ ⎥
⎢ ∂ ∂ ⎥
⎢ 0 ⎥
⎣ ∂z ∂x ⎦


Strain Compatibility – Saint Venant principle
⎡ ∂2 ∂2 ∂2 ⎤
⎢ ∂y 2 0 −2 0 0 ⎥
⎢ ∂x 2 ∂x∂y ⎥
⎢ ∂2 ∂2 ∂2 ⎥
⎢ 0 0 −2 0 ⎥
⎢ ∂z 2 ∂y 2 ∂y∂z ⎥
⎢ ∂2 ∂2 ∂2 ⎥
⎢ 2 0 0 0 −2 ⎥
∂z ∂x 2 ∂x∂z ⎥
L1ε = 0 ⇒ L1 = ⎢ 2
⎢ ∂ ∂2 ∂2 ∂2 ⎥
⎢ 0 0 − − ⎥
⎢ ∂y∂z ∂x∂z ∂x 2 ∂x∂y ⎥
⎢ ∂2 ∂2 ∂2 ∂2 ⎥
⎢ 0 0 − − ⎥
⎢ ∂x∂z ∂z∂y ∂x∂y ∂y ⎥
2
⎢ ⎥
⎢ 0 ∂2 ∂2 ∂2 ∂2 ⎥
0 − −
⎢⎣ ∂x∂y ∂z 2 ∂x∂z ∂y∂z ⎥⎦


Equilibrium Equations
Body loads: L2 τ + f B = 0
τT = ⎡⎣τ xx τ yy τ zz τ xy τ yz τ zx ⎤⎦ , L 2 =LT
Surface loads:
⎡ l 0 0 m 0 n⎤
where N = ⎢⎢0 m 0 l n 0 ⎥⎥
sf
on s f we have Ντ − f =0
⎢⎣0 0 n 0 m l ⎥⎦
l, m and n are cosines of the angles between the normal on the
surface and X, Y and Z


Constitutive Law
τ = Cε
C is elasticity matrix and depends on material properties E and ν

(modulus of elasticity and Poisson’s ratio)
−1
ε=C τ

Differential Formulation
Summary -‐‑ General Form 3D case:

•  Boundary problem of the linear theory of elasticity: differential
equations and boundary conditions
•  15 unknowns: 6 stress components, 6 strain components and 3
displacement components
•  3 equilibrium equations, 6 relationships between displacements and
strains, material law (6 equations)
•  Together with boundary conditions, the state of stress and
deformation is completely defined

Principle of Virtual Work

•  The principle of virtual displacements: the virtual work of a system of
equilibrium forces vanishes on compatible virtual displacements; the
virtual displacements are taken in the form of variations of the real
displacements
•  Equilibrium is a consequence of vanishing of a virtual work
Internal Virtual Work External Virtual Work
∫ ∫ f dV + ∫ U f f dS + ∑ U iT R iC
SfT S
ε T
τ dV = U T B
V V Sf i
Stresses in equilibrium with applied loads
Virtual strains corresponding to virtual displacements

Principle of Complementary Virtual Work
•  The principle of virtual forces: virtual work of

equilibrium variations of the stresses and the forces on
the strains and displacements vanishes; the stress field
considered is a statically admissible field of variation
•  Equilibrium is assumed to hold a priori and the
compatibility of deformations is a consequence of
vanishing of a virtual work
•  Both principles do not depend on a constitutive law

Variational Formulation
•  Based on the principle of stationarity of a functional,

which is usually potential or complementary energy
•  Two classes of boundary conditions: essential
(geometric) and natural (force) boundary conditions
•  Scalar quantities (energies, potentials) are considered
rather than vector quantities

Principle of Minimum

Total Potential Energy
•  Conservation of energy: Work = change in potential,
kinetic and thermal energy
•  For elastic problems (linear and non-‐‑linear) a special case
of the Principle of Virtual Work – Principle of minimum
total potential energy can be applied
•  U is the stress potential: τ ij = ∂U / ∂ε ij
1 T
U = ∫ ε Cε dv
v
2
W = ∫ f Udv + ∫ f Uds
BT Sf
v s

Principle of Minimum

Total Potential Energy
•  Total potential energy is a sum of strain energy and

potential of loads P = U − W
•  This equation, which gives P as a function of deformation
components, together with compatibility relations within
the solid and geometric boundary conditions, defines the
so called Lagrange functional
•  Applying the variation we invoke the stationary condition
of the functional
P

dP = dU − dW = 0

Variational Formulation
•  By utilizing the variational formulation, it is possible to
obtain a formulation of the problem, which is of lower
complexity than the original differential form (strong
form).
•  This is also known as the weak form, which however can
also be a_ained by following an alternate path (see
Galerkin formulation).
•  For approximate solutions, a larger class of trial functions
than in the differential formulation can be employed; for
example, the trial functions need not satisfy the natural
boundary conditions because these boundary conditions
are implicitly contained in the functional – this is
extensively used in MFE.

Approximative Methods
Instead of trying to find the exact solution of the continuous
system, i.e., of the strong form, try to derive an estimate of what
the solution should be at specific points within the system.

The procedure of reducing the physical process to its discrete
counterpart is the discretisation process.


Approximative Methods
Variational Methods Weighted Residual Methods
approximation is based on the
start with an estimate of the the solution and
minimization of a functional, as those
demand that its weighted average error is
defined in the earlier slides.
minimized
•  Rayleigh-Ritz Method
•  The Galerkin Method
•  The Least Square Method
•  The Collocation Method
•  The Subdomain Method
•  Pseudo-spectral Methods


Focus: The axially loaded bar with distributed load example.

ax d 2u
AE 2 = −ax Strong Form
dx
x du
AE =0 Natural BC
dx x= L
C1 ,C2
To fully define the solution (i.e., to evaluate the values of parameters )
we have to use the given boundary conditions (BC):
3
C2 = 0
ax
u(x) = uh + u p = C1x + C2 − ⎯du
⎯⎯⎯
u(0)=0
R
→ aL2
6EA dx = EA C1 =
x= L
2EA
aL2 ax 3
⇒ u( x) = x−
2EA 6EA



ax
aL2 ax 3
u( x) = x−
2EA 6EA
x
3 2.5
2.5 2
2
1.5
1.5
1
1
0.5 0.5
0 0
0 0.5 1 1.5 2 0 0.5 1 1.5 2
Length (m) Length (m)

Weighted Residual Methods


ax
R and its corresponding strong form:
x d 2u
AE 2 = −ax Strong Form
dx
Boundary Conditions (BC)
Given an arbitrary weighting function w
that satisfies the essential conditions and u(0) = 0 Essential BC
additionally: σ ( L) = 0 ⇒
du
AE =0 Natural BC
If then, dx x= L



ax
Multiplying the strong form by w and integrating over L:

Integrating equation I by parts the following relation is derived:



ax
Elaborating a li_le bit more on the relation:



ax
Elaborating a li_le bit more on the relation:
why? why?
x
Therefore, the weak form of the problem is defined as
Find such that:
Observe that the weak form involves derivatives of a lesser order than the
original strong form.

Weighted Residual Methods start with
an estimate of the solution and
demand that its weighted average
error is minimized:
•  The Galerkin Method
•  The Least Square Method
•  The Collocation Method
•  The Subdomain Method
Boris Grigoryevich Galerkin – (1871-1945)

•  Pseudo-spectral Methods
mathematician/ engineer

The Galerkin Method

Theory – Consider the general case of a Example – The axially loaded bar:
differential equation:
Try an approximate solution to the
equation of the following form
Choose the following approximation
Demand that the approximation

where are test functions (input) and satisfies the essential conditions:
are unknown quantities that we need
to evaluate. The solution must satisfy
the boundary conditions.
The approximation error in this case is:
Since is an approximation, substituting
it in the initial equation will result in
an error:

The Galerkin Method

Assumption 1: The weighted average error of the approximation should
be zero

The Galerkin Method

be zero
But that’s the Weak

Form!!!!!
Therefore once again integration by parts leads to

The Galerkin Method

be zero
But that’s the Weak

Form!!!!!
Therefore once again integration by parts leads to
Assumption 2: The weight function is approximated using the same
scheme as for the solution
Remember that the
weight function must
also satisfy the BCs
Substituting the approximations for both and in the weak form,

The Galerkin Method

The following relation is retrieved:
where:

The Galerkin Method

Performing the integration, the following relations are established:

Or in matrix form:

The Galerkin Method
that’s a linear system of equations:
and that’s of course the exact solution. Why?

The Galerkin Method

Now lets try the following approximation:
Which again needs to satisfy the natural BCs, therefore:

The weight function assumes the same form:
Substituting now into the weak form:
( )
EAu2 − x ( ax ) dx = 0 ⇒ u2 = α
L2
L
w2 ∫
0 3EA
Wasn’t that much easier? But….is it correct?

The Galerkin Method

Strong Form Galerkin-Cubic order Galerkin-Linear
3
2.5
1.5
0.5
0
0 0.5 1 1.5 2
Length (m)

The Galerkin Method

We saw in the previous example that the Galerkin method is based on the approximation
of the strong form solution using a set of basis functions. These are by definition
absolutely accurate at the boundaries of the problem. So, why not increase the
boundaries?
element: 1 element: 2
1 2 3

Instead of seeking the solution of a single bar we chose to divide it into three
interconnected and not overlapping elements

The Galerkin Method

element: 1 element: 2
1 2 3

The Galerkin Method
1 2 3

The Galerkin Method

The weak form of a continuous problem was derived in a
systematic way:
This part involves only the
solution approximation
This part only involves the

essential boundary
conditions a.k.a. loading

The Galerkin Method

And then an approximation was defined for the displacement field, for example
Vector of
degrees of
freedom
Displacement
field Shape Function
Matrix
The weak form also involves the first derivative of the approximation
Strain field Strain Displacement

Matrix
The Galerkin Method

Therefore if we return to the weak form :
and set:
The following FUNDAMENTAL FEM expression is derived
Why??
or even be]er

The Galerkin Method
EA has to do only with material and cross-‐‑sectional

properties
We call The Finite Element stiffness Matrix
Ιf E is a function of {d} Material Nonlinearity
Ιf [B] is a function of {d} Geometrical Nonlinearity

Lecture3 PDF

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Lecture3 PDF

Diunggah oleh

Hak Cipta:

Format Tersedia

Institute of Structural Engineering Page 1

Variational Formulation &

Method of Finite Elements I

Today’s Lecture Contents:

Method of Finite Elements I

The diﬀerential form of physical processes

• The isotropic slab equation:

• The Laplace equation in two

Method of Finite Elements I

The diﬀerential form of physical processes

Method of Finite Elements I

The diﬀerential form of physical processes

The diﬀerential form of physical processes

Method of Finite Elements I

The diﬀerential form of physical processes

v This is a homogeneous 2nd order ODE with known solution:

Method of Finite Elements I

The diﬀerential form of physical processes

Method of Finite Elements I

The Strong form – 2D case

: Natural or von Neumann BCs

Method of Finite Elements I

Deriving the Strong form – 3D case

su : supported area with prescribed displacements U su

Method of Finite Elements I

Deriving the Strong form – 3D case

Method of Finite Elements I

Deriving the Strong form – 3D case

Method of Finite Elements I

Deriving the Strong form – 3D case

Method of Finite Elements I

Deriving the Strong form – 3D case

C is elasticity matrix and depends on material properties E and ν

Method of Finite Elements I

Summary -­‐‑ General Form 3D case:

Method of Finite Elements I

Principle of Virtual Work

Internal Virtual Work External Virtual Work

Stresses in equilibrium with applied loads

Virtual strains corresponding to virtual displacements

Principle of Complementary Virtual Work

• The principle of virtual forces: virtual work of

Method of Finite Elements I

• Based on the principle of stationarity of a functional,

Method of Finite Elements I

Principle of Minimum

Method of Finite Elements I

Principle of Minimum

• Total potential energy is a sum of strain energy and

Method of Finite Elements I

Method of Finite Elements I

Method of Finite Elements I

Method of Finite Elements I

The diﬀerential form of physical processes

Method of Finite Elements I

The diﬀerential form of physical processes

Method of Finite Elements I

Weighted Residual Methods

R and its corresponding strong form:

Method of Finite Elements I

Weighted Residual Methods

Multiplying the strong form by w and integrating over L:

Integrating equation I by parts the following relation is derived:

•  The isotropic slab equation:

•  The Laplace equation in two

v This is a homogeneous 2nd order ODE with known solution:

Summary -‐‑ General Form 3D case:

•  The principle of virtual forces: virtual work of

•  Based on the principle of stationarity of a functional,

•  Total potential energy is a sum of strain energy and

EA has to do only with material and cross-‐‑sectional