Chapter 1 Introduction
Chapter 1
Introduction
1.1
Case Study: Pneumatically Actuated PDMS Fingers 1.2
Structural Mechanics: A Quick Review 1.3
Finite Element Methods: A Conceptual Introduction 1.4
Failure Criteria of Materials 1.5
Review
Chapter 1 Introduction
Section 1.1 Case Study: Pneumatically Actuated PDMS Fingers
Section 1.1
Case Study: Pneumatically Actuated PDMS Fingers
Problem Description
[2] A single nger is studied in this case. [1] The pneumatic ngers are part of a surgical parallel robot system remotely controlled by a surgeon through the Internet.
Chapter 1 Introduction
Section 1.1 Case Study: Pneumatically Actuated PDMS Fingers
5 [3] Geometric model. 4 Stress (MPa) 3 2 1 0 0 0.2 0.4 0.6 0.8 1.0 [4] The strain-stress curve of the PDMS elastomer used in this case.
Strain (Dimensionless)
[6] Undeformed shape. [5] As air pressure applies, the nger bends downward.
Chapter 1 Introduction
Section 1.1 Case Study: Pneumatically Actuated PDMS Fingers
Static Structural Simulations
[1] Prepare material properties.
[4] Set up loads and supports.
[2] Create geometric model.
[5] Solve the model.
[3] Generate nite element mesh.
[6] View the results.
Chapter 1 Introduction
Section 1.1 Case Study: Pneumatically Actuated PDMS Fingers
[8] Strains.
[7] Displacements.
Chapter 1 Introduction
Section 1.1 Case Study: Pneumatically Actuated PDMS Fingers
Buckling and Stress-Stiffening
Stress-stiffening: bending stiffness increases with increasing axial tensile stress, e.g., guitar string. The opposite also holds: bending stiffness decreases with increasing axial compressive stress. Buckling: phenomenon when bending stiffness reduces to zero, i.e., the structure is unstable. Usually occurs in slender columns, thin walls, etc. Purpose of a buckling analysis is to nd buckling loads and buckling modes.
[2] The upper surface would undergo compressive stress. It in turn reduces the bending stiffness.
[1] If we apply an upward force here...
Chapter 1 Introduction
Section 1.1 Case Study: Pneumatically Actuated PDMS Fingers
Dynamic Simulations
When the bodies move and deform very fast, inertia effect and damping effect must be considered.
When including these dynamic effects, it is called a dynamic simulation.
Chapter 1 Introduction
Section 1.1 Case Study: Pneumatically Actuated PDMS Fingers
Modal Analysis
A special case of dynamic simulations is the simulation of free vibrations, the vibrations of a structure without any loading.
It is called a modal analysis. Purpose of a modal analysis is to nd natural frequencies and mode shapes.
Chapter 1 Introduction
Section 1.1 Case Study: Pneumatically Actuated PDMS Fingers
Structural Nonlinearities
Linear simulations assume that the response is linearly proportional to the loading.
Deflection (mm) 0 -5 -10 -15 -20 -25 -30 0 40 80 120 160 200 [1] Solution of the nonlinear simulation of the PDMS nger. [2] Solution of the linear simulation pf the PDMS nger.
When the solution deviates from the reality, a nonlinear simulation is needed. Structural nonlinearities come from large deformation, topology changes, nonlinear stress-strain relationship, etc.
Pressure (kPa)
Chapter 1 Introduction
Section 1.2 Structural Mechanics: A Quick Review
10
Section 1.2
Structural Mechanics: A Quick Review
Engineering simulation: nding the responses of a problem domain subject to environmental conditions. Structural simulation: nding the responses of bodies subject to environmental conditions.
The bodies are described by geometries and materials. Environment conditions include support and loading conditions. Responses can be described by displacements, strains, and stresses.
Chapter 1 Introduction
Section 1.2 Structural Mechanics: A Quick Review
11
Displacements
{u} = {
uX
uY
uZ
[3] An arbitrary particle of position (X,Y, Z), before the deformation.
[5] The displacement vector {u} of the particle is formed by connecting the positions before and after the deformation.
[1] The body before deformation.
[4] After the deformation, the particle moves to a new position.
[2] The body after deformation.
Chapter 1 Introduction
Section 1.2 Structural Mechanics: A Quick Review
12
Stresses
{ }
X = YX ZX
XY Y ZY
XZ YZ Z
[3] This face is called negative X-face. [2] This face is called X-face, since the X-direction is normal to this face.
ZY
Z
Z
ZX
XY = YX , YZ = ZY , XZ = ZX
{ } = {
XY
YZ
ZX
[5] The Ycomponent of the stress on X-face.
XY
X
YX
XZ
YZ
[4] The X-component of the stress on X-face.
[1] The reference frame XYZ.
[6] The Z-component of the stress on X-face.
Chapter 1 Introduction
Section 1.2 Structural Mechanics: A Quick Review
13
Strains
[5] Translate AB C so that A coincides with A. The new conguration is AB C . Now CC is the amount of stretch of ABC in Y-face.
[3] After deformation, ABC moves to AB C .
C
[4] To compare with original conguration, rotate AB C to a new conguration AB C .
C
C C
[1] The reference frame.
A
B
A
X
Strain on X -face =
BB AB
[7] And the vector DB describes the twist of ABC in X-face.
X =
BD DB , XY = AB AB
[2] Original conguration ABC. [6] The vector BD describes the stretch of ABC in X-face.
Chapter 1 Introduction
Section 1.2 Structural Mechanics: A Quick Review
14
Physical meaning of strains: The normal strain X is the percentage of stretch of a ber which lies along X-direction. The shear strain XY is the angle change (in radian) of two bers lying on XY-plane and originally forming a right angle. We can dene other strain components in a similar way.
{}
X = YX ZX
XY Y ZY
XZ YZ Z
XY = YX , YZ = ZY , XZ = ZX
{ } = {
XY
YZ
ZX
Chapter 1 Introduction
Section 1.2 Structural Mechanics: A Quick Review
15
Governing Equations
{u} = { { } = { { } = {
X
uX
uY
uZ
}
YZ YZ ZX ZX
Y Y
Z Z
XY XY
} }
Totally 15 quantities
Equilibrium Equations (3 Equations) Strain-Displacement Relations (6 Equations) Stress-Strain Relations (6 Equations)
Chapter 1 Introduction
Section 1.2 Structural Mechanics: A Quick Review
16
Stress-Strain Relations: Hooke's Law
X Y Z E E E Y Z X Y = E E E Z X Y Z = E E E = XY , YZ = YZ , ZX = ZX G G G X =
For isotropic, linearly elastic materials, Young's modulus (E) and Poisson's ratio ( ) can be used to fully describe the stressstrain relations.
XY
The Hooke's law is called a material model. The Young's modulus and the Poisson's ratio are called the material parameters of the material model.
G=
E 2(1 + )
Chapter 1 Introduction
Section 1.2 Structural Mechanics: A Quick Review
17
X Y Z + T E E E Y Z X Y = + T E E E Z = Z X Y + T E E E XY = XY , YZ = YZ , ZX = ZX G G G X =
If temperature changes (thermal loads) are involved, the coefcient of thermal expansion, (CTE, ) must be included.
If inertia forces (e.g., dynamic simulations) are involved, the mass density must be included.
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
18
Section 1.3
Finite Element Methods: A Conceptual Introduction
Basic Ideas
A basic idea of nite element methods is to divide the structural body into small and geometrically simple bodies, called elements, so that equilibrium equations of each element can be written, and all the equilibrium equations are solved simultaneously The elements are assumed to be connected by nodes located on the elements' edges and vertices.
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
19
Another idea is to solve unknown discrete values (displacements at the nodes) rather than to solve unknown functions (displacement elds).
In case of the pneumatic nger, the structural body is divided into 3122 elements. The elements are connected by 17142 nodes. There are 3x17142 unknown displacement values to be solved.
Since the displacement on each node is a vector and has three components (in 3D cases), the number of total unknown quantities to be solved is three times the number of nodes.
The nodal displacement components are called the degrees of freedom (DOF's) of the structure.
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
20
In static cases, the system of equilibrium equations has following form:
K D = F
{ } {}
The displacement vector {D} contains displacements of all degrees of freedom. The force vector {F} contains forces acting on all degrees of freedom. The matrix [K] is called the stiffness matrix of the structure. In a special case when the structure is a spring, {F} as external force, and {D} as the deformation of the spring, then [K] is the spring constant.
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
21
Basic Procedure of Finite Element Method
1. Given the bodies' geometries, material properties, support conditions, and loading conditions. 2. Divide the bodies into elements. 3. Establish the equilibrium equation: [K] {D} = {F} 3.1 Construct the [K] matrix, according to the elements' geometries and the material properties. 3.2 Most of components in {F} can be calculated, according to the loading conditions. 3.3 Most of components in {D} are unknown. Some component, however, are known, according to the support conditions. 3.4 The total number of unknowns in {D} and {F} should be equal to the total number of degrees of freedom of the structure.
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
22
4. Solve the equilibrium equation. Now, the nodal displacements {d} of each element are known. 5. For each element: 5.1 Calculate displacement elds {u}, using an interpolating method, {u} = [N] {d}. The interpolating functions in [N] are called the shape functions. 5.2 Calculate strain elds according to the strain-displacement relations. 5.3 Calculate stress elds according to the stress-strain relations (Hooke's law).
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
23
Shape Functions
Shape functions serve as interpolating functions, allowing the calculation of displacement elds (functions of X, Y, Z) from nodal displacements (discrete values).
[1] A 2D 4-node quadrilateral element
d6 d8 d7
Y
d5
{u} = N {d }
d4 d2
X
For elements with nodes at vertices, the interpolation must be linear and thus the shape functions are linear (of X,Y, Z).
d3 d1
[2] This element's nodes locate at vertices.
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
24
For elements with nodes at vertices as well as at middles of edges, the interpolation must be quadratic and thus the shape functions are quadratic (of X,Y, Z). Elements with linear shape functions are called linear elements, rst-order elements, or lower-order elements. Elements with quadratic shape functions are called quadratic elements, second-order elements, or higher-order elements. ANSYS Workbench supports only rst-order and second-order elements.
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
25
Workbench Elements 3D Solid Bodies
[4] Tetrahedron.
[1] 3D 20-node structural solid. Each node has 3 translational degrees of freedom: DX, DY, and DZ.
[3] Quadrilateralbased pyramid.
[2] Triangle-based prism.
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
26
2D Solid Bodies
[5] 2D 8-node structural solid. Each node has 2 translational degrees of freedom: DX and DY.
[6] Degenerated Triangle.
Chapter 1 Introduction
Section 1.3 Finite Element Methods: A Conceptual Introduction
27
3D Surface Bodies
[8] Degenerated Triangle
[7] 3D 4-node structural shell. Each node has 3 translational and 3 rotational degrees of freedom: DX, DY, DZ, RX, RY, and RZ.
3D Line Bodies
[9] 3D 2-Node beam. Each node has 3 translational and 3 rotational degrees of freedom: DX, DY, DZ, RX, RY, RZ.
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
28
Section 1.4
Failure Criteria of Materials
Ductile versus Brittle Materials
A Ductile material exhibits a large amount of strain before it fractures. The fracture strain of a brittle material is relatively small. Fracture strain is a measure of ductility.
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
29
Failure Points for Ductile Materials
Mild steel is a typical ductile material. For ductile materials, there often exists an obvious yield point, beyond which the deformation would be too large so that the material is no longer reliable or functional; the failure is accompanied by excess deformation.
y
Stress [1] Stress-strain curve for a ductile material. Strain [3] Yield point. [2] Fracture point.
Therefore, for these materials, we are most concerned about whether the material reaches the yield point y .
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
30
Failure Points for Brittle Materials
Cast iron and ceramics are two examples of brittle materials. For brittle materials, there usually doesn't exist obvious yield point, and we are concerned about their fracture point f .
f
Stress
[2] Fracture point.
[1] Stress-strain curve for a brittle material. Strain
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
31
Failure Modes
The fracture of brittle materials is mostly due to tensile failure. The yielding of ductile materials is mostly due to shear failure
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
32
Principal Stresses
A direction in which the shear stress vanishes is called a principal direction. The corresponding normal stress is called a principle stress.
Y
Y
[5] Mohr's circle. [7] Point of minimum normal stress.
[8] Point of maximum shear stress.
[4] Other stress pairs could be drawn.
[2] Stress in the base direction.
( X , XY )
( Y , XY )
XY
XY
[1] Stress state.
X
XY XY
Y
[3] Stress in the direction that forms 90o with the base direction.
[9] Another Point of maximum shear stress.
[6] Point of maximum normal stress.
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
33
At any point of a 3D solid, there are three principal directions and three principal stresses. The maximum normal stress is called the maximum principal stress and denoted by 1 . The minimum normal stress is called the minimum principal stress and denoted by 3 . The medium principal stress is denoted by 2 . The maximum principal stress is usually a positive value, a tension; the minimum principal stress is often a negative value, a compression.
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
34
Failure Criterion for Brittle Materials
The failure of brittle materials is a tensile failure. In other words, a brittle material fractures because its tensile stress reaches the fracture strength f . We may state a failure criterion for brittle materials as follows: At a certain point of a body, if the maximum principal stress reaches the fracture strength of the material, it will fail. In short, a point of material fails if
1 f
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
35
Tresca Criterion for Ductile Materials
The failure of ductile materials is a shear failure. In other words, a ductile material yields because its shear stress reaches the shear strength y of the material.
It is easy to show (using Mohr's circle) that
max =
1 3 2 y 2
We may state a failure criterion for ductile materials as follows: At a certain point of a body, if the maximum shear stress reaches the shear strength of the material, it will fail.
y =
Thus, the material yields if
1 3 y
(1 3 ) is called the stress
intensity.
In short, a point of material fails if
max y
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
36
Von Mises Criterion for Ductile Materials
In 1913, Richard von Mises proposed a theory for predicting the yielding of ductile materials. The theory states that the yielding occurs when the deviatoric strain energy density reaches a critical value, i.e.,
w d w yd
It can be shown that the yielding deviatoric energy in uniaxial test is
w yd =
2 (1+ ) y 3E
And the deviatoric energy in general 3D cases is
wd =
2 2 2 1+ 1 2 + 2 3 + 3 1 6E
) (
) (
Chapter 1 Introduction
Section 1.4 Failure Criteria of Materials
37
After substitution and simplication, the criterion reduces to that the yielding occurs when
2 2 2 1 1 2 + 2 3 + 3 1 y 2
) (
) (
The quantity on the left-hand-side is termed von Mises stress or effective stress, and denoted by e ; in ANSYS, it is also referred to as equivalent stress,
e =
2 2 2 1 1 2 + 2 3 + 3 1 2
) (
) (
