FEM Zabaras 2DFElementCalculations | Calculus Of Variations | Finite Element Method

# MAE4700/5700 Finite Element Analysis for Mechanical and Aerospace Design

Cornell University, Fall 2009 Nicholas Zabaras Materials Process Design and Control Laboratory Sibley School of Mechanical and Aerospace Engineering 101 Rhodes Hall Cornell University Ithaca, NY 14853-3801

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 1

Two-dimensional element calculations
• Similarly to 1D, we here also define the concept of a master element in natural coordinates (ξ ,η ) • Performing calculations on elements with curved sides as shown is difficult in the ( x, y ) coordinates.
– For example, the limits of integration will vary from element to element
Can we find a master element Ω that is mapped appropriately to each element Ωe ? Then with transformation Ω → Ωe we can perform all integrations for all elements at the master element.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 2

Two-dimensional element calculations
• We consider the master element to be a square with: −1 ≤ ξ ,η ≤ 1 • We define the map Ω → Ωe as follows

x = x(ξ ,η ) Te : y = y (ξ ,η )
• Note that the boundary segment ξ = 1 of Ω is mapped on the curved segment x = x(1,η ) y = y (1,η ) with η as a parameter. Similarly for the other 3 boundary segments.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 3

Two-dimensional element calculations
• The complete finite element mesh is seen as a sequence of transformations {T1 , T2 ,..., TE } with each of them mapping Te : Ω → Ωe , e = 1, 2,..., E • We need to figure out how we can transfer all needed calculations in each element Ωe to the master element Ω . • The plan is to make all these calculations defined in terms of the mappings {T1 , T2 ,..., TE }.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 4

Transformation equations
• We assume that the mappings

x = x(ξ ,η ), y = y (ξ ,η )

are differentiable. We can then write: ∂x ∂x dx = dξ + dη , ∂ξ ∂η ∂y ∂y dξ + dη dy = ∂ξ ∂η • We can write this as a system of ⎡ ∂x ∂x ⎤ Eqs: dx ⎢ ∂ξ ∂η ⎥ dξ
⎧ ⎫ ⎨ ⎬= ⎩dy ⎭ ⎧ ⎫ ⎥ ⎨ ⎬ ∂y ⎥ ⎩dη ⎭ ∂η ⎥ ⎦ Jacobian matrix [ J ] ⎢ ⎢ ∂y ⎢ ∂ξ ⎣
T

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 5

Transformation equations
• The Jacobian maps segments d ξ , dη from Ω to Ωe . We define this as the
⎧ dx ⎫ ⎨ ⎬= ⎩dy ⎭ ⎡ ∂x ⎢ ∂ξ ⎢ ⎢ ∂y ⎢ ∂ξ ⎣ ∂x ⎤ ∂η ⎥ ⎥ ∂y ⎥ ∂η ⎥ ⎦
T

⎧dξ ⎫ ⎨ ⎬ ⎩dη ⎭

transpose of the Jacobian matrix [ J ] for notational purposes

Jacobian matrix [ J ]

• Assume the inverse transformation exists:
Te−1 : Ωe → Ω, e = 1, 2,..., E

⎡ ∂y ⎧d ξ ⎫ 1 ⎢ ∂η ⎢ ⎨ ⎬= ⎩dη ⎭ | J | ⎢ − ∂y ⎢ ∂ξ ⎣

∂x ⎤ ∂η ⎥ ⎧ dx ⎫ ⎥⎨ ⎬ ∂x ⎥ ⎩ dy ⎭ ∂ξ ⎥ ⎦

Determinant of the Jacobian matrix ∂x ∂y ∂x ∂y T | J |= det [ J ] = det [ J ] = − ∂ξ ∂η ∂η ∂ξ
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 6

Transformation equations
ξ = ξ ( x, y )
Te−1 :
⎡ ∂y ⎧d ξ ⎫ 1 ⎢ ∂η ⎢ ⎨ ⎬= ⎩dη ⎭ | J | ⎢ − ∂y ⎢ ∂ξ ⎣ − ∂x ⎤ ∂η ⎥ ⎧ dx ⎫ ⎥⎨ ⎬ ∂x ⎥ ⎩dy ⎭ ∂ξ ⎥ ⎦

η = η ( x, y )

• From the equations above we can conclude that:
∂ξ 1 ∂y ∂ξ 1 ∂x = , =− , ∂x | J | ∂η ∂y | J | ∂η ∂η 1 ∂y ∂η 1 ∂x =− , = ∂x | J | ∂ξ ∂y | J | ∂ξ

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 7

Constructing the transformation Te
• We are interested for differentiable mappings {T1, T2 ,..., TE } of the form x = x(ξ ,η ), y = y (ξ ,η ) for each element that do not create gaps or overlapping between elements and which are easy to construct from the geometry of each element Ωe . • Based on an earlier lecture on construction of element basis functions, suppose that we have in place element basis functions N j (ξ ,η ) in Ω such that any function g (ξ ,η ) can be approximated as:
g (ξ ,η ) = ∑ g j N j (ξ ,η ), g j = g (
j =1 M

ξ j ,η j
coordinates of node j in t he master element

)

where M is the number of nodes in the master element.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 8

Constructing the transformation Te
• The finite element basis functions can be used to construct the mapping Te as follows:
x = ∑ x j N j (ξ ,η ),
j =1 M M

y = ∑ y j N j (ξ ,η ),
j =1

where ( x j , y j ) are the coordinates of node j in the element Ωe . • Can you convince yourselves that this mapping does not create gaps or element overlapping?

Biquadratic shape functions Note that the common side between elements e and e+1 is uniquely defined in terms of the nodal coordinates that are shared by the 2 elements.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 9

Geometric interpretation of the Jacobian
• The value of |J| is the ratio of areas of elements at points x = x(ξ ,η ), y = y (ξ ,η ) and (ξ ,η ).

dA =| J | d A
• We need to select the element basis functions N i and ( xi , yi ) such that |J|>0 (area cannot be transformed to a line or a point or turned inside out!). • This guarantees that the mapping Te is invertable.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 10

Transformation equations
• Let us return to the definition of |J| and the earlier stated transformation equations:
∂x ∂y ∂x ∂y − ∂ξ ∂η ∂η ∂ξ ∂ξ 1 ∂y ∂ξ 1 ∂x ∂η 1 ∂y ∂η 1 ∂x = , =− , =− , = ∂x | J | ∂η ∂y | J | ∂η ∂x | J | ∂ξ ∂y | J | ∂ξ | J |= det J =

• Applying the mapping:
x = ∑ x j N j (ξ ,η ), y = ∑ y j N j (ξ ,η ),
j =1 j =1 M M

⎛M ∂ N j (ξ ,η ) ⎞⎛ M ∂ N j (ξ ,η ) ⎞ ⎛ M ∂ N j (ξ ,η ) ⎞⎛ M ∂ N j (ξ ,η ) ⎞ | J |= det J = ⎜ ∑ x j −⎜ ∑ xj ∑ yj ∑ yj ⎟⎜ ⎟ ⎟⎜ ⎟ ∂ξ ∂η ∂η ∂ξ ⎝ j =1 ⎠⎝ j =1 ⎠ ⎝ j =1 ⎠⎝ j =1 ⎠
∂ N j (ξ ,η ) ∂ξ ∂ N j (ξ ,η ) ∂η ∂ N j (ξ ,η ) ∂η ∂ N j (ξ ,η ) ∂ξ 1 M 1 M 1 M 1 M = , =− , =− , = ∑ yj ∑ xj ∑ yj ∑ xj ∂x | J | j =1 ∂η ∂y | J | j =1 ∂η ∂x | J | j =1 ∂ξ ∂y | J | j =1 ∂η
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 11

Example problem
• Consider the following master element.

• The basis functions for this element are:
1 N 1 (ξ ,η ) = (1 − ξ )(1 − η ) 4 1 N 2 (ξ ,η ) = (1 + ξ )(1 − η ) 4 1 N 3 (ξ ,η ) = (1 + ξ )(1 + η ) 4 1 N 4 (ξ ,η ) = (1 − ξ )(1 + η ) 4
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 12

Example problem
• Examine if the following mapping is invertable.

T1
T1−1 ?
M 3 x = ∑ x j N j (ξ ,η ) = 3 N 1 (ξ ,η ) + 3 N 2 (ξ ,η ) = (1 − η ) 2 j =1

T1 : 1 y = ∑ y j N j (ξ ,η ) = N 2 (ξ ,η ) + N 3 (ξ ,η ) = (1 + ξ ) 2 j =1
M

• Note that

3⎤ ⎡ 0 − ⎥ ⎢ 3 2 | J |= det ⎢ ⎥ = >0⇒ ⎢1 0 ⎥ 4 ⎢2 ⎥ ⎣ ⎦

the map T1 is invertable.
Ω

• |J| is the ratio of the area of Ω1 to area of

.

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 13

Example problem
• Examine if the following mapping is invertable. T2

T2−1 ?
M 3 x = ∑ x j N j (ξ ,η ) = 3 N 3 (ξ ,η ) + 3 N 4 (ξ ,η ) = (1 + η ) 2 j =1

T2 : 1 y = ∑ y j N j (ξ ,η ) = N 2 (ξ ,η ) + N 3 (ξ ,η ) = (1 + ξ ) 2 j =1
M

• Note that

⎡ ⎢0 | J |= det ⎢ ⎢1 ⎢2 ⎣

3⎤ 3 2⎥ ⎥ =− <0⇒ 4 0⎥ ⎥ ⎦

the map T2 is not invertable.

• The clockwise numbering of nodes should be avoided.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 14

Example problem
• Examine if the following mapping is invertable.

T3
T3−1 ?
M 1 x = ∑ x j N j (ξ ,η ) = N 2 (ξ ,η ) + 2 N 3 (ξ ,η ) = (3 + 3ξ + η + ξη ) 4 j =1

T3 :
M 1 y = ∑ y j N j (ξ ,η ) = 2 N 3 (ξ ,η ) + N 4 (ξ ,η ) = (3 + ξ + 3η + ξη ) 4 j =1

1 ⎡1 ⎤ (3 + η ) (1 + ξ ) ⎥ ⎢4 1 1 1 4 | J |= det ⎢ ⎥= + ξ+ η⇒ 8 ⎢ 1 (1 + η ) 1 (3 + ξ ) ⎥ 2 8 ⎢4 ⎥ ⎣ ⎦ 4

|J| cannot be zero inside the element so the map T3 is invertable.

• |J| is smaller near node 1, largest near node 3.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 15

Example problem
• Examine if the following mapping is invertable. T4

T4−1 ?
M 1 x = ∑ x j N j (ξ ,η ) =3 N 2 (ξ ,η ) + N 3 (ξ ,η ) = (4 + 4ξ − 2η − 2ξη ) 4 j =1

T4 :
M 1 y = ∑ y j N j (ξ ,η ) = N 3 (ξ ,η ) + 2 N 4 (ξ ,η ) = (3 − ξ + 3η − ξη ) 4 j =1

1 ⎡1 ⎤ (4 − 2η ) − (2 + 2ξ ) ⎥ ⎢4 1 4 | J |= det ⎢ ⎥ = (5 − 3ξ − 4η ) ⇒ 1 8 ⎢ − 1 (1 + η ) (3 − ξ ) ⎥ ⎢ 4 ⎥ ⎣ ⎦ 4

• All angles of quadrilateral elements need to be < π.

|J| is not >0 everywhere inside the element so the map T3 is not invertable. The marked region near node 3 -- above the line ξ = 5 3 − 4 3 η -- is mapped outside the element.

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 16

Finite element equations
• To solve our boundary value problem of interest examined earlier, we need to compute the following matrices and vectors ( 1 ≤ e ≤ E , 1 ≤ i, j ≤ N e ):
⎡ ⎛ ∂N ie ∂N e ∂N ie ∂N e ⎞ ⎤ j j e e e kij = ∫ ⎢ k ⎜ + ⎟ + bN i N j ⎥ dxdy ⎜ ∂x ∂x ⎟ ∂y ∂y ⎠ ⎥ Ωe ⎢ ⎝ ⎣ ⎦
P =
e ij

fi e =

Ωe

fN ie dxdy

∂Ωe h 2

pN N ds

e i

e j

γ ie =

∂Ωe h 2

γ N ie ds

• k and f are assumed to be functions of (x,y), whereas p and γ are assumed functions of s defining ∂Ω2 h ( x( s), y ( s)) • It remains to discuss how to compute these using the master element and the transformation Te.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 17

Master element calculations: Step 1
• We construct the mappings
N
e

Te : Ω → Ωe , e = 1, 2,..., E

as follows:

x = ∑ x j N j (ξ ,η )
j =1 N
e

y = ∑ y j N j (ξ ,η ),
j =1

• This mapping Te : Ω → Ωe , is completed defined by the nodal coordinates ( x j , y j ) of element e.

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 18

Master element calculations: Step 2
• Let g(x,y) be any function of (x,y) in Ωe. We can convert g to a function of (ξ ,η ) defined in Ω as follows:
g ( x, y ) = g ( x (ξ ,η ), y (ξ ,η )) ≡ g (ξ ,η )

• The transformation x(ξ ,η ), y (ξ ,η ) is defined in Step 1. N e ( x, y ) are simply • The element shape functions j obtained from N j (ξ ,η ) as follows:
N e ( x, y ) = N j ( x(ξ ,η ), y (ξ ,η )), j = 1, 2,..., N e j

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 19

Master element calculations: Step 2
• The derivatives of
∂N e ( x, y ) j ∂x ∂N e ( x, y ) j ∂y = ∂ N j ∂ξ ∂ N j ∂η = + ∂ξ ∂y ∂η ∂y

N e ( x, y ) j

with respect to x and y are:
∂x N ∂ N k (ξ ,η ) ∂x N ∂ N k (ξ ,η ) = ∑ xk , = ∑ xk ∂ξ k =1 ∂ξ ∂η k =1 ∂η
e e

∂ N j ∂ξ ∂ N j ∂η + , ∂ξ ∂x ∂η ∂x

where:

∂y N ∂ N k (ξ ,η ) ∂y N ∂ N k (ξ ,η ) = ∑ yk , = ∑ yk ∂η k =1 ∂η ∂η k =1 ∂η
e e

• Finally, with |J| the Jacobian of

Te , using

∂ξ 1 M ∂ N k (ξ ,η ) ∂ξ 1 M ∂ N k (ξ ,η ) ∂η 1 M ∂ N k (ξ ,η ) ∂η 1 M ∂ N k (ξ ,η ) = , =− , =− , = ∑ yk ∑ xk ∑ yk ∑ xk ∂x | J | k =1 ∂η ∂y | J | k =1 ∂η ∂x | J | k =1 ∂ξ ∂y | J | k =1 ∂η

we conclude:

∂N e ( x, y ) j ∂x ∂N e ( x, y ) j ∂y

1 ⎧∂ N j M ∂ N k (ξ ,η ) ∂ N j M ∂ N k (ξ ,η ) ⎫ ⎪ ⎪ = − ∑ yk ∑ yk ⎨ ⎬, | J | ⎪ ∂ξ j =1 ∂η ∂η k =1 ∂ξ ⎪ ⎩ ⎭ 1 ⎧ ∂N j M ∂ N k (ξ ,η ) ∂ N j M ∂ N k (ξ ,η ) ⎫ ⎪ ⎪ = + ∑ xk ∑ xk ⎨− ⎬ | J | ⎪ ∂ξ k =1 ∂η ∂η jk =1 ∂η ⎪ ⎩ ⎭

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 20

Note regarding convergence rates
• We had presented earlier that the asymptotic rates of e convergece are only valid if wh ( x, y) = N w j N ej ( x, y) ∑
e

j =1

contains complete polynomials of degree k for smooth interpolated functions. • However, with the transformation
N e ( x, y ) = N j ( x(ξ ,η ), y (ξ ,η )), j = 1, 2,..., N e j

the shape functions N j ( x, y ) may not even be polynomials! • It can be shown that if the master basis functions N j (ξ ,η ) contain complete polynomials of degree k and |J|>0, the error estimates still hold.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 21

e

Master element calculations: Step 3
• The domain integrals of interest now take the form:
Ω

∫ g ( x, y )dxdy = ∫ g ( x (ξ ,η ), y (ξ ,η )) | J | d ξ dη ≡ ∫ g (ξ ,η ) | J | d ξ dη
e

Ω

Ω

G (ξ ,η )

• Our integrals are now in the domain of Ω
Ω

∫ g ( x, y )dxdy = ∫ g ( x (ξ ,η ), y (ξ ,η )) | J | d ξ dη ≡ ∫ G (ξ ,η )d ξ dη
e

Ω

Ω

• As in 1D, we use Gauss integration (to be discussed):
Ω

∫ G (ξ ,η )d ξ dη ≈ ∑ G ( ξ i ,ηi )
i

N

i =1

wi

Gaus s Gauss int egration int egration po int s weights

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 22

Master element calculations: Step 4
• In numerical calculation of e.g. the element stiffness,
⎡ ⎛ ∂N ie ∂N e ∂N ie ∂N e ⎞ ⎤ j j e e e kij = ∫ ⎢ k ⎜ + ⎟ + bN i N j ⎥ dxdy ⎜ ∂x ∂x ∂y ∂y ⎟ ⎥ Ωe ⎢ ⎝ ⎠ ⎣ ⎦

we often need to consider functions k(x,y). It is common to use the finite element interpolant to approximate this function, i.e.
k h ( x, y ) = ∑
e

N

j =1

kj
k(x ,y )
j j

N e ( x, y ) j

= ∑ k j N j (ξ ,η )
e

N

j =1

This requires only the nodal values of k. Similar calculation can be done for integrals involving b(x,y) and f(x,y).
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 23

Master element calculations: Step 5
• We need to consider a number of integrals that involve boundary integration:
Pije =

• Lets introduce the restriction of the basis functions N j (ξ ,η ) on the side ξ = 1 . We denote them as:θ (η ) ≡ N j (1,η ), j = 1, 2,.., Ne . These functions are only non-zero for nodes on the boundary ξ = 1 . The 2 integrals above can be | j (η ) | is the Jacobian of the computed as: transformation of η to s
P =
e ij ∂Ωe h 2

∂Ωe h 2

pN ie N e ds j

γ ie =

∂Ωe h 2

γ N ie ds

pN N ds = ∫ p (η )θ i (η )θ j (η ) | j (η ) | dη
e i e j −1

1

ds =| j (η ) | dη dy

ds 1/2 ds = ⎡ dx 2 + dy 2 ⎤ ⎣ ⎦ dx
1/2

γ ie =

Notes: (a) Use FE interpolants of p(η ), γ (η ) (b) Use 1D Gauss integration.
MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 24

∂Ωe h 2

γ N ie ds = ∫ γ (η )θ i (η ) | j (η ) | dη
−1

1

⎡⎛ ∂x (1,η ) ⎞2 ⎛ ∂y (1,η ) ⎞ 2 ⎤ | j (η ) |= ⎢⎜ + ⎥ ∂η ⎟ ⎜ ∂η ⎟ ⎥ ⎠ ⎝ ⎠ ⎦ ⎢⎝ ⎣

Example problem using 3-node triangular elements
• Consider 2D heat conduction using the general problem discussed in an earlier lecture. • We are computing T(x,y) such that:
−∇ • ( D∇T ) = f ( x, y ) T = T on ΓT q = − D∇T • n = q on Γ q

• Recall that the final weak form is: ⎛

⎡ ∂N1 ⎢ ∂x e ⎡B ⎤ = ⎢ e ⎣ ⎦ ⎢ ∂N 1 ⎢ ∂y ⎣

⎞ ⎜ ⎟ nel nel T T T ⎜ T T ⎟ T T ⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤ {w} ∑ ⎣ Le ⎦ ∫ ⎣ B e ⎦ [ De ] ⎣ B e ⎦ d Ω ⎣ Le ⎦ {d } = {w} ∑ ⎣ Le ⎦ ⎜ ∫ ⎣ N e ⎦ f d Ω − ∫ ⎣ N e ⎦ qd Γ ⎟, ∀{wF } e =1 e =1 Ωe Γe ⎜ Ωe ⎟ q ⎜ ⎟ Ke { fΩe } { fΓe} ⎝ ⎠ e e e
∂N 2 ∂x e ∂N 2 ∂y ∂N nen ⎤ ∂x ⎥ ⎥ ⎡N e ⎤ = ⎡N e e ∂N nen ⎥ ⎣ ⎦ ⎣ 1 ... ∂y ⎥ ⎦ ...
e e N 2 ... N nen ⎤ ⎦

⎡ KE ⎢K ⎣ FE

K EF ⎤ ⎡ d E ⎤ ⎡ f E + rE ⎤ ⎥⎢ ⎥ = ⎢ f ⎥ KF ⎦ ⎣dF ⎦ ⎣ F ⎦

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 25

Example problem using 3-node triangular elements
• Consider the heat conduction problem in the figure. We assume isotropic conductivity k=5 WoC-1 and a source term f=6 Wm-2. The BCs are as shown. • We want to compute the temperature field using 2 linear triangular elements.

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 26

Example problem : Stiffness, element 1
• Using expressions derived before for triangular elements, we can write: 3
⎡ ∂N1e ⎢ ∂x e ⎡B ⎤ = ⎢ e ⎣ ⎦ ⎢ ∂N 1 cons tan t ⎢ ∂y matrix ⎣
e ∂N 2 ∂x e ∂N 2 ∂y

∂N 3e ⎤ ∂x ⎥ ⎥ = 1e ∂N 3e ⎥ 2 A ∂y ⎥ ⎦

e e ⎡ y2 − y3 ⎢ e e ⎣ x3 − x2

1 e y3 − y1 e x1e − x3

e y1e − y2 ⎤ ⎥ e x2 − x1e ⎦

2

e e e e e e e e 2 Ae = ( x2 y3 − x3 y2 ) − ( x1e y3 − x3 y1e ) + ( x1e y2 − x2 y1e )

1

⎡K e ⎤ = ⎣ ⎦

Ω

e e e e e e ∫e ⎡ B ⎤ [ D ] ⎡ B ⎤ d Ω = ⎡ B ⎤ k ⎡ B ⎤ A ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ T T

• For e=1:

1 ⎡ −0.5 1 −0.5⎤ ⎡ B1 ⎤ = ⎢ ⎣ ⎦ 2 −2 0 2 ⎥ ⎣ ⎦

⎡ 5.3125 −0.625 −4.6875⎤ 1 K 1 = B1 B1kA1 = ⎢ −0.625 −0.625 ⎥ 2 1.25 ⎢ ⎥ ⎢ −4.6875 −0.625 5.3125 ⎥ 3 ⎣ ⎦
T

1

2

3

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 27

Example problem : Stiffness, element 2
• Using expressions derived before for triangular elements, we can write: 3
⎡ ∂N1e ⎢ e ⎢ ∂xe B = ⎢ ∂N1 cons tan t matrix ⎢ ∂y ⎣
e ∂N 2 ∂x e ∂N 2 ∂y

2

∂N 3e ⎤ ⎥ 1 ∂x ⎥ = e ∂N 3e ⎥ 2 A ∂y ⎥ ⎦

e e ⎡ y2 − y3 ⎢ e e ⎣ x3 − x2

1 e y3 − y1 e x1e − x3

e y1e − y2 ⎤ ⎥ e x2 − x1e ⎦

1

e e e e e e e e 2 Ae = ( x2 y3 − x3 y2 ) − ( x1e y3 − x3 y1e ) + ( x1e y2 − x2 y1e )

Ke =

Ωe

B e [ D e ]B e d Ω = B e BkAe
T T

• For e=2:
2 T

⎡ 0 0.5 −0.5⎤ B2 = ⎢ 0 ⎥ ⎣ −2 2 ⎦

−10 0 ⎤ ⎡ 10 ⎡ K 2 ⎤ = ⎡ B ⎤ ⎡ B 2 ⎤ kA2 = ⎢ −10 10.625 −0.625⎥ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎢ ⎥ ⎢ 0 −0.625 0.625 ⎥ ⎣ ⎦

2

4

3

2 4 3

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 28

Example problem : Assembled stiffness
• Assembling the 2 stiffness matrices gives:

0 ⎤ ⎡ 5.3125 −0.625 −4.6875 ⎢ −0.625 11.25 −0.625 −10 ⎥ 1 T 1 1 2 T 2 2 ⎥ [K ] = ⎡L ⎤ ⎡K ⎤ ⎡L ⎤ + ⎡L ⎤ ⎡K ⎤ ⎡L ⎤ = ⎢ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎢ −4.6875 −0.625 5.9375 −0.625⎥ ⎢ ⎥ −10 −0.625 10.625 ⎦ ⎣ 0

1 2 3 4

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 29

Example problem : Load calculations – e=1,2
• The shape functions for 3-node triangular elements are
3
⎡ N e ⎤ = ⎡ N1e ⎣ ⎦ ⎣ N1e =
e N2

2

N 3e ⎤ ⎦

3 1 2

1 e e e e e e e e ( x2 y3 − x3 y2 + ( y2 − y3 ) x + ( x3 − x2 ) y ) e 2A 1 e e e e e N 2 = e ( x3 y1e − x1e y3 + ( y3 − y1e ) x + ( x1e − x3 ) y ) 2A 1 e e e e N 3e = e ( x1e y2 − x2 y1e + ( y1e − y2 ) x + ( x2 − x1e ) y ) 2A
e e e e e e e e 2 Ae = ( x2 y3 − x3 y2 ) − ( x1e y3 − x3 y1e ) + ( x1e y2 − x2 y1e )
e Ω

1

{ f } = f ∫ {N }
Ωe

e T

• For e=1: • For e=2:

⎡1⎤ This expression fAe ⎢ ⎥ 1 can be proved for dΩ = 3 ⎢⎥ constant f! ⎢1⎥ ⎣⎦

⎡1⎤ ⎡1⎤ ⎡ 2 ⎤ fA1 ⎢ ⎥ 6 × 1 ⎢ ⎥ ⎢ ⎥ { f 1} = 3 ⎢1⎥ = 3 ⎢1⎥ = ⎢2⎥ ⎢1⎥ ⎢1⎥ ⎢ 2 ⎥ ⎣⎦ ⎣⎦ ⎣ ⎦ ⎡1⎤ ⎡1⎤ ⎡1⎤ fA2 ⎢ ⎥ 6 × 0.5 ⎢ ⎥ ⎢ ⎥ { f 2 } = 3 ⎢1⎥ = 3 ⎢1⎥ = ⎢1⎥ ⎢1⎥ ⎢1⎥ ⎢1⎥ ⎣⎦ ⎣⎦ ⎣⎦

1 2 3 2 4 3

⎡ 2 ⎤ ⎡2⎤ ⎢ 2 + 1⎥ ⎢ 3 ⎥ 1 T 1 2 T 2 { f } = ⎡L ⎤ { f } + ⎡L ⎤ { f } = ⎢ ⎥ = ⎢ ⎥ ⎣ ⎦ ⎣ ⎦ ⎢ 2 + 1⎥ ⎢ 3 ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ 1 ⎦ ⎣1 ⎦

1 2 3 4

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 30

Example problem: Applying natural BCs
• For elements with segments on Γq, we need to compute
{ f } = − ∫ ⎡N ⎣
e Γ Γe q e T

3

2 1

⎤ qd Γ ⎦

• We only need to worry about e=2:
⎡ 1 ⎤ 2 2 2 2 2 2 2 2 ⎢ 2 A2 ( x2 y3 − x3 y2 + ( y2 − y3 ) x + ( x3 − x2 ) y ) ⎥ 0 ⎡ ⎤ ⎢ ⎥ T 1 2 2 2 2 ⎡ N 2 ⎤ | y =1 = ⎢ 2 ( x3 y12 − x12 y3 + ( y3 − y12 ) x + ( x12 − x3 ) y ) ⎥ = ⎢ 0.5 x ⎥ ⎣ ⎦ ⎢ ⎥ ⎢2A ⎥ ⎢ −0.5 x + 1⎥ ⎢ 1 ⎥ ⎣ ⎦ 2 2 2 2 2 2 2 2 ⎢ 2 ( x1 y2 − x2 y1 + ( y1 − y2 ) x + ( x2 − x1 ) y ) ⎥ ⎣2A ⎦ y =1

• Thus:

0 ⎡ ⎤ ⎡ 0 ⎤ { fΓe } = −∫0 ⎢ 0.5 x ⎥ 20dx = ⎢−20⎥ ⎢ ⎥ ⎢ ⎥ ⎢ −0.5 x + 1⎥ ⎢ −20 ⎥ ⎣ ⎦ ⎣ ⎦
2

2 4 3

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 31

Example problem: Final system of equations
• Assembling the load vector gives the following:
⎡ 2 ⎤ ⎡ r1 ⎤ ⎡ r1 + 2 ⎤ ⎢ 3 ⎥ ⎢r ⎥ ⎢ r + 3 ⎥ { f } = { f Γ } + { f Ω } + {r} = ⎢ ⎥ + ⎢ 2 ⎥ = ⎢ 2 ⎥ ⎢ −17 ⎥ ⎢ r3 ⎥ ⎢ r3 − 17 ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎣ −19 ⎦ ⎣ 0 ⎦ ⎣ −19 ⎦

r1 , r2 , r3 ,

are unknown heat fluxes at the nodes with prescribed temperature (short of reaction forces!)

• The final system of equations is:
0 ⎤ ⎡ 0 ⎤ ⎡ r1 + 2 ⎤ ⎡ 5.3125 −0.625 −4.6875 ⎢ −0.625 11.25 −0.625 −10 ⎥ ⎢ 0 ⎥ ⎢ r2 + 3 ⎥ ⎢ ⎥⎢ ⎥ = ⎢ ⎥ ⎢ −4.6875 −0.625 5.9375 −0.625⎥ ⎢ 0 ⎥ ⎢ r3 − 17 ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ −10 −0.625 10.625 ⎦ ⎣T4 ⎦ ⎣ −19 ⎦ ⎣ 0

• Partitioning this system gives: T • `Reaction fluxes’:⎡⎢ 5.3125 −0.625 −4.6875
−0.625 11.25 ⎢ ⎢ ⎣ −4.6875 −0.625 −0.625 5.9375

4

=−

⎡ 0 ⎤ 0 ⎤⎢ ⎡ r1 + 2 ⎤ 0 ⎥ ⎢ ⎥ = r +3 ⎥ ⇒ −10 ⎥ ⎢ ⎥ ⎥⎢ 0 ⎥ ⎢ 2 ⎥ ⎢ r3 − 17 ⎥ −0.625⎦ ⎢ ⎥ ⎣ ⎦ ⎣ −1.788⎦

19 = −1.788 10.625

⎡ r1 ⎤ ⎡ −2 ⎤ ⎢ r ⎥ = ⎢ 14.88 ⎥ ⎥ ⎢ 2⎥ ⎢ ⎢ r3 ⎥ ⎢18.1175⎥ ⎦ ⎣ ⎦ ⎣

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 32

Example problem: Postprocessing
• The DOF in each element (nodal temperatures) are now given as:
⎡0 ⎤ ⎡ 0 ⎤ {d 1} = ⎢0⎥ , {d 2 } = ⎢−1.788⎥ ⎢ ⎥ ⎢ ⎥ ⎢0 ⎥ ⎢ 0 ⎥ ⎣ ⎦ ⎣ ⎦

• The flux vectors can then be computed for each element as:
⎡0⎤ ⎡ q1 ⎤ 1 ⎡ −0.5 1 −0.5⎤ ⎢ ⎥ ⎡0 ⎤ x {q1} ≡ ⎢q1 ⎥ = −k ⎡ B1 ⎤ {d 1} = −5 2 ⎢ −2 0 2 ⎥ ⎢0⎥ = ⎢0⎥ ⎣ ⎦ ⎣ ⎦⎢ ⎥ ⎣ ⎦ ⎣ y⎦ ⎣0⎦
⎡ 0 ⎤ 2 ⎡ qx ⎤ 0 0.5 −0.5⎤ ⎢ 4.47 ⎥ ⎡ ⎤ {q 2 } ≡ ⎢ q 2 ⎥ = −k ⎣ B 2 ⎦ {d 2 } = −5 ⎡−2 2 2 ⎥ ⎢ −1.788⎥ = ⎡17.88⎤ ⎢ ⎢ ⎥ ⎣ ⎦⎢ ⎣ ⎦ ⎣ y⎦ ⎣ 0 ⎥ ⎦

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 33

Example problem using 4-node quadrilateral elements
• We repeat the calculations of the earlier example using one 4-node quadrilateral element. We use again `matrix notation’.

• The element coordinates are:
⎡ xe ⎣

⎡ x1e ⎢ e x e y ⎤=⎢ 2 ⎦ ⎢ xe 3 ⎢ e ⎣ x4

y1e ⎤ ⎡ 0 1 ⎤ e⎥ y2 ⎥ ⎢ 0 0 ⎥ ⎥ =⎢ e ⎥ ⎢ 2 0.5⎥ y3 ⎢ ⎥ e⎥ y4 ⎦ ⎣ 2 1 ⎦

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 34

Example problem using 4-node quadrilateral elements
• The basis functions in natural coordinates are:
1 N 1 (ξ ,η ) = (1 − ξ )(1 − η ) 4 1 N 2 (ξ ,η ) = (1 + ξ )(1 − η ) 4 1 N 3 (ξ ,η ) = (1 + ξ )(1 + η ) 4 1 N 4 (ξ ,η ) = (1 − ξ )(1 + η ) 4

• We need to compute the Be matrix relating derivatives wrt x and y and nodal values.
⎡ ∂T e ⎤ ⎡ ∂ N 1 ⎢ ⎥ ⎢ ∂x ⎥ ⎢ ∂x ⎢ = ⎢ ∂T e ⎥ ⎢ ∂ N e ⎢ ⎥ ⎢ 1 ∂y ⎦ ⎢ ∂y ⎣ ⎣
e

∂N 2 ∂x ∂N 2 ∂y
e

e

∂N3 ∂x ∂N3 ∂y
e

e

e

⎡ ∂T e ⎤ ⎢ ⎥ ⎢ ∂ξ ⎢ ∂x ⎥ = ⎢ ⎢ ∂T e ⎥ ⎢ e ⎢ ⎥ ⎢ ∂ N1 ⎣ ∂y ⎦ ⎢ ∂ξ ⎣

⎡∂N e 1

∂ξ ∂η + ∂x ∂η ∂x ∂ξ ∂ N 1 ∂η + ∂y ∂η ∂y
e

e ∂ N1

e ∂N 2

∂ξ ∂η + ∂ξ ∂x ∂η ∂x
e e

e ∂N 2

[B ]

e ⎡T e ⎤ ∂N 4 ⎤ ⎢ 1 ⎥ ⎥ e ∂x ⎥ ⎢T2 ⎥ e ⎥⎢ e⎥ ∂ N 4 ⎥ ⎢T3 ⎥ ∂y ⎥ ⎢T4e ⎥ ⎦⎣ ⎦

∂ N 3 ∂ξ ∂ N 3 ∂η + ∂ξ ∂x ∂η ∂x ∂ N 3 ∂ξ ∂ N 3 ∂η + ∂ξ ∂y ∂η ∂y
e e

e

e

∂ N 2 ∂ξ ∂ N 2 ∂η + ∂ξ ∂y ∂η ∂y

e e ⎡ e⎤ ∂ N 4 ∂ξ ∂ N 4 ∂η ⎤ ⎢T1 ⎥ ⎥ e + ∂ξ ∂x ∂η ∂x ⎥ ⎢T2 ⎥ ⎥⎢ e⎥ e e ∂ N 4 ∂ξ ∂ N 4 ∂η ⎥ ⎢T3 ⎥ + ∂ξ ∂y ∂η ∂y ⎥ ⎢T4e ⎥ ⎦⎣ ⎦

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 35

Example problem using 4-node quadrilateral elements
⎡∂N e ⎢ 1 ∂ξ ⎡ Be ⎤ = ⎢ ⎣ ⎦ ⎢ e ⎢ ∂ N1 ⎢ ∂ξ ⎣ ∂ξ ∂η + ∂x ∂η ∂x ∂ξ ∂ N 1 ∂η + ∂y ∂η ∂y
e e ∂ N1 e ∂N 2

∂ξ ∂η + ∂ξ ∂x ∂η ∂x
e e

e ∂N 2

e ∂N3

∂ξ ∂η + ∂ξ ∂x ∂η ∂x
e e

e ∂N3

e ∂N 4

∂ N 2 ∂ξ ∂ N 2 ∂η + ∂ξ ∂y ∂η ∂y

∂ N 3 ∂ξ ∂ N 3 ∂η + ∂ξ ∂y ∂η ∂y

∂ξ ∂η ⎤ ⎥ + ∂ξ ∂x ∂η ∂x ⎥ ⎥ e e ∂ N 4 ∂ξ ∂ N 4 ∂η ⎥ + ∂ξ ∂y ∂η ∂y ⎥ ⎦
e ∂N 4

• We can write this matrix as:
⎡ ∂ξ ⎢ ⎡ B e ⎤ = ⎢ ∂x ⎣ ⎦ ⎢ ∂ξ ⎢ ∂y ⎣ ⎡∂N e ∂η ⎤ ⎢ 1 ∂x ⎥ ⎢ ∂ξ ⎥ ∂η ⎥ ⎢ e ⎢ ∂ N1 ⎥⎢ ∂y ⎦ ⎣ ∂η
e ∂N 2 e ∂N3

∂ξ ∂N 2 ∂η
e

∂ξ ∂N3 ∂η
e

⎡ ∂y e ⎥ ⎢ ∂ξ ⎥ 1 ⎢ ∂η = e ⎥ | Je |⎢ ∂x e ∂N 4 ⎥ ⎢− ⎥ ⎣ ∂η ∂η ⎦

e⎤ ∂N 4

⎥⎢ ⎥ ⎢ ∂ξ ⎢ e ∂x e ⎥ ⎢ ∂ N 1 ⎥ ∂ξ ⎦ ⎢ ∂η ⎣

∂y ∂ξ

e⎤⎡

e ∂ N1

e ∂N 2

e ∂N3

e⎤ ∂N 4

∂ξ ∂N 2 ∂η
e

∂ξ ∂N3 ∂η
e

⎥ ∂ξ ⎥ e⎥ ∂N 4 ⎥ ∂η ⎥ ⎦

This comes from an earlier derivation:

⎡ ∂y e ⎢ ⎧dξ ⎫ 1 ⎢ ∂η ⎨ ⎬= e ⎩dη ⎭ | J | ⎢ ∂y e ⎢− ⎣ ∂ξ

∂x e ⎤ − ⎥ ∂η ⎥ ⎧ dx ⎫ ⎨ ⎬ e ⎥ dy ∂x ⎩ ⎭ ⎥ ∂ξ ⎦

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 36

Example problem using 4-node quadrilateral elements
⎡ ∂y e ⎢ 1 ⎢ ∂η e⎤ ⎡B = ⎣ ⎦ | Je |⎢ ∂x e ⎢− ⎣ ∂η
e ∂y e ⎤ ⎡ ∂ N 1 − ⎥⎢ ∂ξ ⎥ ⎢ ∂ξ ⎢ e ∂x e ⎥ ⎢ ∂ N 1 ⎥ ∂ξ ⎦ ⎢ ∂η ⎣
e

∂N 2 ∂ξ ∂N 2 ∂η
e

e

∂N3 ∂ξ ∂N3 ∂η
e

e

e ∂N 4 ⎤ ⎥ ∂ξ ⎥ e⎥ ∂N 4 ⎥ ∂η ⎥ ⎦

Inverse of the Jacobian [ J

]

x e = ∑ x e N j (ξ ,η ), j
j =1 4

4

• The Jacobian matrix can be computed from:y e =
⎡ ∂x ⎢ ∂ξ ⎡J e ⎤ = ⎢ ⎣ ⎦ ⎢ ∂x e Jacobian ⎢ matrix of ⎣ ∂η
e

j =1

∑ y j N j (ξ ,η ),
e

element e

⎡∂N e ∂y ⎢ 1 ⎥ ∂ξ ⎥ ⎢ ∂ξ = e⎥ ⎢ e ∂y ⎥ ⎢ ∂ N1 ∂η ⎦ ⎢ ∂η ⎣
e⎤

e ∂N 2

e ∂N3

∂ξ
e ∂N 2

∂ξ
e ∂N3

∂η

∂η

e ⎤ ⎡ xe ∂N 4 ⎢ 1 ⎥ e ∂ξ ⎥ ⎢ x2 e ⎥⎢ e ∂ N 4 ⎥ ⎢ x3 e ∂η ⎥ ⎢ x4 ⎦⎣

e y1 ⎤ ⎥ e y2 ⎥ e⎥ y3 ⎥ e⎥ y4 ⎦

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 37

Example problem using 4-node quadrilateral elements
• For element 1:
η −1 1−η 1+η ⎡J1⎤ = 1 ⎡ ⎣ ⎦ 4 ⎢ξ − 1 −ξ − 1 1 + ξ ⎣
⎡0 1 ⎤ −η − 1⎤ ⎢ 0 0 ⎥ ⎡0 0.125η − 0.375 ⎤ ⎢ ⎥ ⎥ ⎢ 2 0.5⎥ = ⎢1 0.125ξ + 0.125⎥ 1− ξ ⎦ ⎣ ⎦ ⎢ ⎥ ⎣2 1 ⎦

⎡1 + ξ ⎤ ⎢ 3 − η 1⎥ −1 ⎡J1⎤ = ⎢ ⎥ , det J 1 = −0.125η + 0.375 ⎣ ⎦ ⎢ 8 0⎥ ⎢η − 3 ⎥ ⎣ ⎦

⎡1 + ξ ⎤ 1⎥ ⎢3 −η 1 ⎡η − 1 1 − η 1 + η 1⎤ ⎡B = ⎢ ⎥ ⎢ ⎣ ⎦ ⎢ 8 4 ξ − 1 −ξ − 1 1 + ξ 0⎥ ⎣ ⎢η − 3 ⎥ ⎣ ⎦

−η − 1⎤ 1−ξ ⎥ ⎦
T 1 1 T

• The element stiffness is:
⎡ K1 ⎤ = ∑ ⎣ ⎦ i =1
int

⎡ 1⎤ ∫ ⎡ 1⎤ ⎡ 1⎤ ⎡ 1⎤ ⎡ 1⎤ ⎡ 1⎤ ⎡ 1⎤ 1 ⎣ K ⎦ = Ω ⎣ B ⎦ ⎣ D ⎦ ⎣ B ⎦ d Ω = −∫1 −∫1 ⎣ B ⎦ ⎣ D ⎦ ⎣ B ⎦ | J | d ξ dη ⇒
1

N =2 N =2
int

1 1 ⎛ ⎡ B1 ⎤T ⎡ B1 ⎤ | J 1 | ⎞ WiW j , (ξ i ,η j ) = (± , ± ),(Wi ,W j ) = (1,1) ∑ k⎜ ⎟ ⎣ ⎦ ⎣ ⎦ ⎠ξ =ξ ,η =η 3 3 j =1 ⎝
i j

k[ I ]

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 38

Example problem using 4-node quadrilateral elements
• The element stiffness is finally given as:
⎡ 4.76 −3.51 −2.98 1.73 ⎤ ⎢ 4.13 1.73 −2.36 ⎥ 1 ⎤ ⎢ −3.51 ⎥ ⎡K = ⎣ ⎦ ⎢ −2.98 1.73 6.54 −5.29 ⎥ ⎢ 1.73 −2.36 −5.29 5.91 ⎥ ⎣ ⎦

• The nodal force is (Gauss integration not shown):

{ fΩe }

⎡Ne ⎤ ⎢ 1⎥ ⎡ 2.5⎤ e⎥ ⎢ ⎢ 2.5⎥ 1 1 1 1 N2⎥ e ⎤T e ⎤T 1 (−0.125η + 0.375)dξ dη = ⎢ ⎥ = ∫ f ⎡ N ⎦ d Ω = ∫ ∫ f ⎡ N ⎦ | J | dξ dη = ∫ ∫ 6 ⎢ ⎣ ⎣ ⎢ e⎥ ⎢ 2 ⎥ Ω −1 −1 −1 −1 N3⎥ ⎢ ⎢ 2 ⎥ ⎣ ⎦ ⎢ e⎥ ⎣N 4 ⎦
e

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 39

Example problem using 4-node quadrilateral elements
• We now need to compute the contribution to element load from the natural boundary conditions on the element e=1 at ξ=−1.

{ }

⎡1 ⎤ (1 − η ) ⎥ ⎢2 ⎡ −20 ⎤ ⎢ ⎥ ⎢ 0 ⎥ 1 1 T T 0 ⎥2−0 ⎛ ⎡ N e ⎤ | J1 |⎞ e e⎤ ⎢ ⎥ fq = − ∫ q ⎡ N ⎦ d Γ = − ∫ q ⎜ ⎣ ⎦ dη = − ∫ 20 dη = ⎢ ⎟ ⎣ ⎢ 0 ⎥ 2 ⎢ 0 ⎥ ⎠ξ =−1 Ω −1 ⎝ −1 ⎢ ⎥ |J | ⎢ −20 ⎥ 1 ⎣ ⎦ ⎢ (1 + η ) ⎥ ⎢2 ⎥ ⎣ ⎦ ⎡ r1 − 17.5⎤ ⎢ r + 2.5 ⎥ • Assembly of all the nodal ⎥ { f } = { f Ω } + { f Γ } + {r } = ⎢ 2 ⎢ r3 + 2 ⎥ contributions gives: ⎢ −18 ⎥ ⎣ ⎦
e 1

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 40

Example problem using 4-node quadrilateral elements
• We finally obtain:
⎡ 4.76 −3.51 −2.98 1.73 ⎤ ⎡ 0 ⎤ ⎡ r1 − 17.5⎤ ⎢ −3.51 4.13 1.73 −2.36 ⎥ ⎢ 0 ⎥ ⎢ r + 2.5 ⎥ ⎢ ⎥⎢ ⎥ = ⎢ 2 ⎥ 6.54 −5.29 ⎥ ⎢ 0 ⎥ ⎢ r3 + 2 ⎥ ⎢ −2.98 1.73 ⎢ 1.73 −2.36 −5.29 5.91 ⎥ ⎢T ⎥ ⎢ −18 ⎥ ⎣ ⎦⎣ 4⎦ ⎣ ⎦

which gives

T4 = −3.04.

• The nodal DOFs are:

{d1}

⎡ 0 ⎤ ⎢ 0 ⎥ ⎥ =⎢ ⎢ 0 ⎥ ⎢ −3.04 ⎥ ⎣ ⎦

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 41

Example problem using 4-node quadrilateral elements
• The flux components at each Gauss point are:

{

⎡T1e ⎤ ⎡ ∂T e ⎤ ⎡1 + ξ ⎤ ⎢ e⎥ 1⎥ ⎢ −k ⎥ ⎢3 −η ⎢T ⎥ 1 ⎡η − 1 1 − η 1 + η ∂x ⎥ 1 ⎢ ⎡ B e ⎤ ⎢ 2 ⎥ = −k ⎢ ⎥ ⎢ q (ξ ,η ) = = −k ⎣ ⎦ e 4 ξ − 1 −ξ − 1 1 + ξ ⎢ ∂T e ⎥ ⎢ 8 ⎢T3 ⎥ 0⎥ ⎣ −k ⎢ ⎥ ⎢η − 3 ⎥ ⎢ e⎥ ∂y ⎦ ⎣ ⎦ ⎣ T4 ⎦ ⎣

}

⎡ 0 ⎤ −η − 1⎤ ⎢ 0 ⎥ ⎢ ⎥⇒ ⎥ 1− ξ ⎦ ⎢ 0 ⎥ ⎢ −3.04 ⎥ ⎣ ⎦

1 ⎫ ⎡0.8979 ⎤ ⎧ 1 1 q ( , − )⎬ = ⎢ ⎨ 3 3 ⎭ ⎣3.5916 ⎥ ⎩ ⎦ ⎧ 1 1 1 ⎫ ⎡ −2.2965⎤ ⎨q ( , ) ⎬ = ⎢ 3 3 ⎭ ⎣ 19.793 ⎥ ⎩ ⎦ ⎧ 1 1 1 ⎫ ⎡ 4.9482 ⎤ ⎨q (− , ) ⎬ = ⎢ ⎥ 3 3 ⎭ ⎣19.793 ⎦ ⎩ 1 ⎥ ⎡5.8042 ⎤ ⎢ 1 1 q (− , − ) ⎥ = ⎢ ⎢ 3 3 ⎦ ⎣3.5916 ⎥ ⎣ ⎦

MAE 4700 – FE Analysis for Mechanical & Aerospace Design N. Zabaras (10/06/2009) 42