Download as pdf or txt
Download as pdf or txt
You are on page 1of 40

Finite

element
method
Prof. Jinwoo Lee

© 2017 Cengage Learning


Isoparametric
Formulation 10
Chapter Objectives
✓ To formulate the isoparametric formulation of the bar element & Q4 element stiffness
matrices
✓ To illustrate by example how to evaluate the stresses at a given point in a plane
quadrilateral element using Gaussian quadrature
✓ To describe some higher-order shape functions for the three-noded linear strain bar, the
improved bilinear quadratic (Q6), the eight- and nine-noded quadratic quadrilateral (Q8
and Q9) elements, and the twelve-noded cubic quadrilateral (Q12) element
© 2017 Cengage Learning
10. The Isoparametric Formulation
• Isoparametric Formulation
➢ May appear somewhat tedious and confusing initially, but will
lead to a simplified computer program formulation

➢ Applicable for two- and three-dimensional stress analysis


and for nonstructural problems

➢ Allows for elements to be created that are nonrectangular


and have curved sides

➢ The term isoparametric is derived from the use of the same


shape functions (or interpolation functions) [N] to define the
element’s geometric shape.

© 2017 Cengage Learning


10. The Isoparametric Formulation
• Isoparametric Elements
➢ Isoparametric element equations are formulated using a
natural (or intrinsic) coordinate system, s, that is defined by
element geometry and not by the element orientation in the
global-coordinate system.

➢ There is a relationship (transformation mapping) between


the natural coordinate system, s, and the global coordinate
system, x, for each element of a specific structure.

➢ This transformation mapping must be used in the element


equation formulations.

© 2017 Cengage Learning


10.1 Isoparametric Formulation of the Bar Element Stiffness Matrix

• The natural coordinate, s, is attached to the element,


with the origin located at the center of the element
as shown below:
➢ The s-axis need not be parallel to the x-axis.
➢ We consider the bar element to have two degrees of
freedom — axial displacements u1 and u2.
➢ For when the s and x axis are parallel, they can be
related by:

where xc is the global coordinate of the element centroid:

© 2017 Cengage Learning


10.1 Isoparametric Formulation of the Bar Element Stiffness Matrix

• Step 1 : Select Element Type


➢ We express the natural coordinates, s, in terms of the
global as:

➢ To solve for the shape functions, we begin by relating


the natural coordinate to the global coordinate by:

➢ Keep in mind that from the figure:


➢ Solving we obtain:
➢ and in matrix form:

where the shape functions are:

© 2017 Cengage Learning


10.1 Isoparametric Formulation of the Bar Element Stiffness Matrix

• (Cont.)
➢ The linear shape
functions map the s
coordinate of any point
in the element to the x
coordinate.
➢ The displacement
function is defined by the
same shape functions:

➢ Since u and x are


defined by the same
shape functions, the
element is isoparametric.
© 2017 Cengage Learning
10.1 Isoparametric Formulation of the Bar Element Stiffness Matrix

• Step 3 : Define the Strain-Displacement Relation


➢ The strain-displacement relationship for the
isoparametric relationship is defined as:

➢ We seek:
➢ From before we obtain: and
➢ Thus:

= where

➢ The stress matrix is given by Hooke’s law as:

© 2017 Cengage Learning


10.1 Isoparametric Formulation of the Bar Element Stiffness Matrix

• Step 4 : Derive the Element Stiffness Matrix & Eqs.


➢ The stiffness matrix is:

➢ We must transform the coordinate x to s with the general


transformation:
where [J] is the Jacobian matrix.

➢ For the simple bar element:


➢ We obtain the stiffness matrix in natural coordinates:
➢ Substituting and integrating, we obtain:

➢ For higher-order elements, the integral in closed form becomes


very difficult. The use of numerical integration illustrates the distinct
advantages of the isoparametric formulation of the equations.

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• Step 1 : Select Element Type


➢ We consider the quadrilateral to have 8 degrees of
freedom.
➢ The element has straight sides but is otherwise an
arbitrary shape.

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• (Cont.)
➢ For the special case for a rectangular element

➢ Begin by assuming the global coordinates x and y are


related to the natural coordinates s and t as follows:

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• (Cont.)

➢ In matrix form expressed as:

where the shape functions are:

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• (Cont.)
➢ The physical shapes of the shape functions as they
vary with natural coordinates are shown in Figure 10-4

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• Step 2 : Select Displacement Functions

➢ The displacement functions within an element are now


similarly defined by the same shape functions as are
used to define the element geometric shape; that is:

➢ u and v are displacements parallel to the global x and y


coordinates.
© 2017 Cengage Learning
10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• Step 3 : Define the Strain-Displacement Relation


➢ Recall from Chapter 6 that for the Q4 rectangle the strains
are given in global coordinates by:

➢ Normal strain variations are given in the figure.

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• (Cont.)
➢ To construct the element stiffness matrix, we must
determine the strains which are now functions of s and
t. Apply the chain rule to express s and t as functions of
x and y, yielding:
𝜕𝑓 𝜕𝑥 𝜕𝑦 𝜕𝑓
𝜕𝑠 = 𝜕𝑠 𝜕𝑠 𝜕𝑥
𝜕𝑓 𝜕𝑥 𝜕𝑦 𝜕𝑓
𝜕𝑡 𝜕𝑡 𝜕𝑡 𝜕𝑦

➢ are all known


from previous equations. We seek .

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• (Cont.)
➢ Using Cramer’s rule (Appendix B) to solve we obtain:

where the determinant in the denominator is the


determinant of the Jacobian matrix [ J ]:

➢ We want to express element strains as:

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• (Cont.)
➢ Start with the usual relationship between strains and
displacement:
➢ represent the partial of any variable
we put inside the parenthesis.

➢ Substituting and evaluating the determinates:

where is the determinant of [ J ] given earlier.


© 2017 Cengage Learning
10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• (Cont.)
➢ Substituting above equations again we obtain the
strains expressed in terms of the natural coordinates
as:

where

and and

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• (Cont.)
➢ We can express strains in terms of the shape functions and
global coordinates in compact matrix form as
where

➢ So then,

➢ We have [B] expressed as a function of s and t and thus


have the strains in terms of s and t.
© 2017 Cengage Learning
10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• (Cont.)
➢ The explicit form of [B] can be obtained

where for i = 1, 2, 3, and 4

and

© 2017 Cengage Learning


10.2 Isoparametric Formulation of the Q4 Element Stiffness Matrix

• Body & Surface Forces


➢ The element body-force
matrix will be determined
numerically from:

➢ The surface-force matrix


for Figure 10-6 is:

© 2017 Cengage Learning


10.3 Gaussian Quadrature
• Gaussian Quadrature
➢ The Gaussian Quadrature is a method for numerical
evaluation of definite integrals commonly used in finite
element work.
➢ The amount of sampling points is determined to achieve a
certain level of accuracy and are located symmetrically with
respect to the center of the interval.
➢ In general, Gaussian quadrature using n points (Gauss
points) is exact if the integrand is a polynomial of degree 2n
– 1 or less.
➢ If the function is not a polynomial, Gaussian quadrature is
inexact, but it becomes more accurate as more Gauss points
are used.
➢ We will give a brief overview of a general case of the
Gaussian quadrature using one sampling point.
© 2017 Cengage Learning
10.3 Gaussian Quadrature
• Gaussian Quadrature One Sampling Point
➢ We start by considering the following integral for evaluation:
where y is a function of x.
➢ We might choose (sample or evaluate) y at the midpoint y(0) = y1
and multiply by the length of the interval as shown in the figure.
➢ We arrive at I = 2y1 a result that is exact if the curve happens to be
a straight line.
➢ This formula can be generalized as:

where n is the amount of sampling points, yi is the value of each


point, Wi is the weight factor

© 2017 Cengage Learning


10.3 Gaussian Quadrature
• Gaussian Quadrature in 2-Dimensions
➢ In two dimensions, we obtain the quadrature formula by
integrating first with respect to one coordinate and then
with respect to the other as:

➢ We need not use the same


number of Gauss points in each
direction
➢ For example, a four-point Gauss
rule is shown in the figure and
yields:
© 2017 Cengage Learning
10.4 Evaluation of the Stiffness Matrix by the Gaussian Quadrature

• Evaluation of the Stiffness Matrix

➢ For a quadrilateral element,


the stiffness matrix can be
evaluated as:

➢ A flowchart to evaluate [k] for


an element using four-point
Gaussian quadrature is given.

© 2017 Cengage Learning


10.4 Evaluation of the Stiffness Matrix by the Gaussian Quadrature

• Evaluation of the Stiffness Matrix


➢ In explicit form for four-point Gaussian quadrature (now
using the single summation notation with i = 1, 2, 3, 4),
we have:

➢ Go through Example 10.4 in the text to practice


evaluating a stiffness matrix for a quadrilateral element.
© 2017 Cengage Learning
10.4 Evaluation of the Stiffness Matrix by the Gaussian Quadrature
10.4 Evaluation of the Stiffness Matrix by the Gaussian Quadrature

• Answer
10.4 Evaluation of the Stiffness Matrix by the Gaussian Quadrature

• (Cont.)
➢ We first evaluate [J] at each Gauss point
10.4 Evaluation of the Stiffness Matrix by the Gaussian Quadrature

• (Cont.)
➢ Using the following equations, we can evaluate [B]
10.4 Evaluation of the Stiffness Matrix by the Gaussian Quadrature

• (Cont.)
➢ For instance, we obtain the final form of [B(-0.5773, -
0.5773)] as

➢ For plane stress condition, [D] is


10.4 Evaluation of the Stiffness Matrix by the Gaussian Quadrature

• (Cont.)
➢ Using Eq., the matrix [k] becomes
10.4 Evaluation of the Stiffness Matrix by the Gaussian Quadrature

• Evaluation of the Element Stress


➢ The stresses are not constant within the
quadrilateral element.
➢ In practice, the stresses are evaluated at the same
Gauss points used to evaluate the stiffness matrix [k].
➢ For a quadrilateral, we get four sets of stress data.
➢ To reduce the data, it often practical to evaluate the
stress matrix at the s = 0, t = 0 locations of the element.
➢ The common method used in computer programs is to
evaluate the stresses in all elements at a shared node
and then use an average of these element nodal
stresses to represent the stress at the node.
➢ Stress plots obtained in these programs are based on
this average nodal method.
© 2017 Cengage Learning
10.5 Higher-Order Shape Functions
• Higher-Order Shape Functions
➢ In general, higher-order element shape functions can
be developed by adding additional nodes to the sides of
the linear element.
➢ This results in higher-order strain variations and
convergence occurs at a faster rate using fewer
elements.
➢ The trade-off is that there is a substantial increase in
required computational power.
➢ Another advantage of higher-order elements is that
curved boundaries of irregularly-shaped bodies can be
approximated more closely than simple straight-sided
linear elements

© 2017 Cengage Learning


10.5 Higher-Order Shape Functions
• Improved Bilinear Quadratic (Q6)
➢ An improved element to remove the shear locking
inherent in the Q4 element is to add two internal
degrees of freedom per displacement function, creating
the Q6 element.

➢ The displacement functions are shown:

© 2017 Cengage Learning


10.5 Higher-Order Shape Functions
• (Cont.)
➢ The displacement field is enhanced by modes that describe
the curvature as shown in Figure 10-16.
➢ There is a possibility that the edges of two adjacent elements
may be incompatible as shown in Figure 10-17:

© 2017 Cengage Learning


10.5 Higher-Order Shape Functions
• Quadratic Rectangle (Q8 & Q9)
➢ The eight-noded element also called “Q8” element has
four corner nodes and four mid-side nodes.
➢ The shape functions of the Q8 element are based on
the incomplete polynomial given:

➢ Two forms of shape functions are required to describe


the geometry, one for corner nodes and one for mid-
side notes.

© 2017 Cengage Learning


10.5 Higher-Order Shape Functions
• (Cont.)
➢ To evaluate the stiffness
matrix for the eight-noded
quadratic isoparametric
element, we use the nine-
point Gauss rule (often
described as a 3 X 3
rule).
➢ By adding a ninth node at
s = 0, t = 0, we create an
element called a “Q9”.
➢ This is an internal node
that is not connected to
any other nodes.
© 2017 Cengage Learning
10.5 Higher-Order Shape Functions
• Comparing the Higher-Order Elements

© 2017 Cengage Learning

You might also like