Lecture 2: One Dimensional Problems: APL705 Finite Element Method

1/11/2017
Lecture 2: One‐dimensional
Problems
APL705 Finite Element Method
Steps in FEM Problems
1. System idealization – the given actual problem is broken
down into idealized elements
2. Element equilibrium – the equilibrium requirements are
established in terms of displacements or the state
variables (or primary unknowns)
3. Assembly of elements – the element interconnections
are used to develop a set of simultaneous equations in
the unknown state variables
4 Calculation of response –
4. Calculation of response – by solving the system of
by solving the system of
equations, the response of each element and thereby
that of the entire system under consideration is
determined
1
1/11/2017
Direct Stiffness Method
• The direct aprroach
Th
The process of obtaining the total
f bt i i th t t l or system stiffness matrix by
t tiff ti b
summation of individual element stiffness matrices is called
direct stiffness method.
n
K = ∑k( j)
i=1
Here k(j) are element stiffness matrices. Through the examples
discussed later, we will see that this approach is general and can
be applied to other non‐structural problems also. This method is
simple and gives a basic idea about obtaining the behaviour of a
finite element of a continuum.
Some Examples
• A single element model
Here we can consider a rod, spring, truss member, beam, pipe or
Here we can consider a rod spring truss member beam pipe or
any such simple structural element for analysis. Consider the
following single linear spring model as an example.
i k j
Ui Uj
fi fj
• Here
Here ii and j are the two nodes where the
and j are the two nodes where the
displacements and forces are present and k is
the stiffness (spring constant) of the linear
spring element under consideration
2
1/11/2017
Simple Examples
• Force – displacement relation
Let us consider the equilibrium of forces at nodes i and j:
fi = k(ui − u j ) = kui − ku j
f j = k(u j − ui ) = ku j − kui
Expressing these in a matrix form:
⎡ k −k ⎤⎧⎪ ui ⎫⎪ ⎧⎪ fi ⎫⎪
⎢ ⎥⎨ ⎬=⎨ ⎬ ⇒ ku = f
⎣ −kk k ⎦⎩⎪ u j ⎭⎪ ⎩⎪ f j ⎭⎪
• Here [k] – element stiffness matrix, {u}‐ nodal
displacement vector and {f}‐ element force vector
Direct Approach
• Two‐element model
Now let us consider the following linear elastic springs connected end
Now let us consider the following linear elastic springs connected end
to end as follows:
x
1 K1 K2 3
U1 U3
f1 (1) 2 (2) f3
U2
f2
• Here
Here again the nomenclature is similar at nodes
again the nomenclature is similar at nodes
and elements. Now we have 3 nodes and two
elements with their stiffness as k1 and k2
respectively.
3
1/11/2017
Direct Approach
• Simple Two‐element Example
Let us extend the idea of to a two element system now:
For Element 1
⎡ k −k ⎤⎧⎪ u ⎫⎪ ⎧⎪ f (1) ⎫⎪
⎢ 1 1
⎥⎨ 1 ⎬ = ⎨ 1 ⎬
⎢⎣ −k1 k1 ⎥⎦⎪⎩ u2 ⎪⎭ ⎪ f2(1) ⎪
For Element 2 ⎩ ⎭
⎡ k ⎧ ⎫
−k2 ⎤⎧⎪ u2 ⎫⎪ ⎪ f1(2) ⎪
⎢ 2 ⎥⎨ ⎬=⎨ ⎬
⎢⎣ −kk2 k2 ⎥⎦⎪⎩ u3 ⎪⎭ ⎪ f2(2) ⎪
⎩ ⎭
• For assembling the system or total stiffness matrix,
there are two approaches: (i) Equilibrium of forces
(ii) Superposition of element matrices
Assembling Stiffness Matrix
• (i) Considering Equilibrium of forces
N d 1 f1(1) = R1 ⇒ k1u1 − k1u2
Node
Node 2 f2(1) + f1(2) = R2 ⇒ k1u1 + (k1 + k2 )u2 − k2 u3
Node 3 f2(2) = R3 ⇒ k2 u2 + k2 u3
⎡ k −k1 0 ⎤⎧ u1 ⎫ ⎧ R1 ⎫
⎢ 1 ⎥⎪ ⎪ ⎪ ⎪
⎢ −k1 k1 + k2 −k2 ⎥⎨ u2 ⎬ = ⎨ R2 ⎬
⎢ ⎥⎪ ⎪ ⎪ ⎪
⎢⎣ 0 −k2 k2 ⎦⎥⎪⎩ u3 ⎪⎭ ⎪⎩ R3 ⎭⎪
• (ii) Superposition of element matrices
4
1/11/2017
Assembling Stiffness Matrix
• (ii) Superposition of element matrices
Here we take each element matrix of 3x3 size and then just add them
t th t bt i th
together to obtain the system stiffness matrix
t tiff ti
⎡ k −k 0 ⎤⎧ u ⎫ ⎧ 1 ⎫
f1( ) ⎪
⎢ 1 1
⎥⎪ 1 ⎪ ⎪⎪ ⎪
⎢ −k1 k1 0 ⎥⎨ u2 ⎬ = ⎨
f2(1) ⎬ element 1
⎢ ⎥⎪ ⎪ ⎪ ⎪
⎢⎣ 0 0 0 ⎥⎦⎪⎩ u3 ⎪⎭ ⎪
0 ⎪
⎩ ⎭
⎡ ⎤⎧ u ⎫ ⎧ ⎫
⎢ 0 0 0 ⎥⎪ 1 ⎪ ⎪ 0 ⎪
⎪ ⎪
⎢ 0 k2 −k2 ⎥⎨ u2 ⎬ = ⎨
f1(2) ⎬ element 2
⎢ ⎥⎪ ⎪ ⎪ 2 ⎪
⎣ 0 −k2 k2 ⎦⎪⎩ u3 ⎪⎭ ⎪⎩
f2( ) ⎪⎭
⎡ k ⎧ ⎫
−k1 0 ⎤⎧ u1 ⎫ ⎪ f1(1) ⎪⎪
• Superposition ⎢ 1 ⎥⎪ ⎪ ⎪
⎢ −k2 (k1 + k2 ) −k2 ⎥⎨ u2 ⎬ = ⎨ f2(1) + f1(2) ⎬ E1+ E2
⎢ ⎥⎪ ⎪ ⎪ ⎪
⎢⎣ 0 −k2 k2 ⎥⎦⎪⎩ u3 ⎪⎭ ⎪ f2( )
2
⎪⎭
⎩
Boundary Conditions and Loading
• Treatment of B.Cs and Loading
Let the displacement u1=0 and loads R2=R3=P
⎡ k −k1 0 ⎤⎧ 0 ⎫ ⎧ R1 ⎫
⎢ 1 ⎥⎪ ⎪ ⎪ ⎪
⎢ −k2 (k1 + k2 ) −k2 ⎥⎨ u2 ⎬ = ⎨ P ⎬
⎢ ⎥⎪ ⎪ ⎪ ⎪
⎢⎣ 0 −k2 k2 ⎥⎦⎩ u3 ⎭ ⎪⎩ P ⎪⎭
• Now we can obtain the reaction and displacements as
R1 = −k1u2 ⎧⎪ u ⎫⎪ ⎧⎪ 2P / k1 ⎫⎪
⎨ ⎬=⎨ ⎬
2
⎩⎪ u3 ⎭⎪ ⎪⎩ 2P / k1 + P / k2 ⎭⎪
• Therefore the reaction is
R1 = −2P

Lecture 2: One Dimensional Problems: APL705 Finite Element Method

Uploaded by

Copyright:

Available Formats

Lecture 2: One Dimensional Problems: APL705 Finite Element Method

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lecture 2: One Dimensional Problems: APL705 Finite Element Method

Uploaded by

Copyright:

Available Formats

1/11/2017

You might also like