Contents

1 Introduction 1
1.1 Governing eqs. of a vibrating flexible structure . . . . . . . . . . . . . 7
1.1.1 Example: axially vibrating slender beam . . . . . . . . . . . . . . . . . 9
1.1.2 Torsional vibration of a bar . . . . . . . . . . . . . . . . . . . . . . . 14
1.1.3 Transverse vibration of a beam . . . . . . . . . . . . . . . . . . . . . 14
1.2 Quasi-static response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.3 Coupled vibrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.3.1 Bending-torsion coupling . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.4 Plates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.5 Structures subject to initial stress . . . . . . . . . . . . . . . . . . . . . 27
1.5.1 Bending vibration of an axially-loaded beam . . . . . . . . . . . . . . 28
1.5.2 O.-O.-P. vibration of an axially-loaded plate . . . . . . . . . . . . . . . 30

2 Discretization Methods 31
2.1 Lumped-parameter models . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.1.1 S.-S. beam . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . 31
2.1.2 Other examples . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 33
2.2 Ritz-Galerkin method . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.2.1 R.M. for thin plates . . . . . . . .
. . . . . . . . . . . . . . . . . . . 38
2.2.2 Penalty method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.2.3 Dynamic vibration absorber . . . . . . . . . . . . . . . . . . . . . . . 41
2.2.4 Rayleigh’s principle . . . . . . . . .
. . . . . . . . . . . . . . . . . . . 42
2.2.5 Assumed modes method . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.3 Finite Element Method . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.3.1 Linear rod elements . . . . . . . .. . . . . . . . . . . . . . . . . . . 48
2.3.2 Beam FEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.3.3 Plate FEs . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . 51
2.3.4 Lumping approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.3.5 Singularities & static condensation. . . . . . . . . . . . . . . . . . . 52
2.4 Modal analysis for discrete systems . . . . . . . . . . . . . . . . . . . . 53
2.5 Damping . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 59
2.5.1 1-DOF damped system . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.5.2 N-DOFs damped system . . . . . . . . . . . . . . . . . . . . . . . . . 61
2.6 Modal analysis & peculiar situations . . . . . . . . . . . . . . . . . . . 66
2.6.1 Prescribed boundary motion . . . . . . . . . . . . . . . . . . . . . . . 66
2.6.2 Systems w/ rigid body modes . . . . . . . . . . . . . . . . . . . . . . 67

3 Numerical Methods 70
3.1 Solution of the eigenvalue problem . . . . . . . . . . . . . . . . . . . . 70
3.1.1Power method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.1.2Inverse power method . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.1.3Shifted-inverse power method . . . . . . . . . . . . . . . . . . . . . . 72
3.2 Direct time-integration methods . . . . . . . . . . . . . . . . . . . . . . 73
3.2.1 Central difference method . . . . . . . . . . . . . . . . . . . . . . . . 73
3.2.2 Newmark method . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

i
4 State-Space Formulation 77
4.1 Modal state-space representation . . . . . . . . . . . . . . . . . . . . . 80
4.2 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.3 Random loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.3.1 Ergodic process through LTI system . . . . . . . . . . . . . . . . . . . 90
4.3.2 Power Spectral Density . . . . . . . . . . . . . . . . . . . . . . . . . 95
4.3.3 Shape filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
4.3.4 Cross-PSD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

5 Control System Design 101

5.1 State-space C.S. design . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5.1.1 Control law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.1.2 Reconstruction of the state . . . . . . . . . . . . . . . . . . . . . . . 106
5.2 Pole placement technique . . . . . . . . . . . . . . . . . . . . . . . . . . 109
5.3 Optimal control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
5.3.1 Recovery of no-coupling case . . . . . . . . . . . . . . . . . . . . . . 113
5.3.2 Presence of random disturbances . . . . . . . . . . . . . . . . . . . . 114
5.4 Optimal observer: Kalman filter . . . . . . . . . . . . . . . . . . . . . . 115
5.5 Validity of separation principle . . . . . . . . . . . . . . . . . . . . . . . 117
5.5.1 Effect of uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . 117
5.5.2 Effect of neglected dynamics . . . . . . . . . . . . . . . . . . . . . . 117
5.6 LQR design guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
5.6.1 Maximum values weighting . . . . . . . . . . . . . . . . . . . . . . . 119
5.6.2 Eigenvalue shift (prescribed stability margin) . . . . . . . . . . . . . . 120
5.6.3 Implicit model following . . . . . . . . . . . . . . . . . . . . . . . . . 120
5.6.4 Sensitivity-weighted LQR . . . . . . . . . . . . . . . . . . . . . . . . 121
5.6.5 Frequency shaping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
5.7 Finite-horizon framework . . . . . . . . . . . . . . . . . . . . . . . . . . 123
5.7.1 Optimal tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.8 Gramians . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.8.1 Balanced realization . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.9 Direct-output optimal control . . . . . . . . . . . . . . . . . . . . . . . 127

6 Thermoelasticity 129
6.1 Energy balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
6.1.1
Weak form of the thermal problem . . . . . . . . . . . . . . . . . . . 130
6.2 Mechanical problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
6.3 One-way coupled thermoelastic problem . . . . . . . . . . . . . . . . . 131
6.3.1 Quasi-static approximation . . . . . . . . . . . . . . . . . . . . . . . 132

ii
1 Introduction

1
1.1 Governing eqs. of a vibrating flexible structure

7
1.1.1 Example: axially vibrating slender beam

9
1.1.2 Torsional vibration of a bar
1.1.3 Transverse vibration of a beam

14
1.2 Quasi-static response

18
 1.3 Coupled vibrations
1.3.1 Bending-torsion coupling

21
1.4 Plates

22
1.5 Structures subject to initial stress

27
1.5.1 Bending vibration of an axially-loaded beam

28
1.5.2 O.-O.-P. vibration of an axially-loaded plate

30
 2 Discretization Methods
2.1 Lumped-parameter models
2.1.1 S.-S. beam

31
2.1.2 Other examples

33
2.2 Ritz-Galerkin method

34
2.2.1 R.M. for thin plates

38
2.2.2 Penalty method

40
2.2.3 Dynamic vibration absorber

41
2.2.4 Rayleigh’s principle

42
2.2.5 Assumed modes method

45
 2.3 Finite Element Method
2.3.1 Linear rod elements

48
2.3.2 Beam FEs

50
2.3.3 Plate FEs
2.3.4 Lumping approach

51
2.3.5 Singularities & static condensation

52
2.4 Modal analysis for discrete systems

53
2.5 Damping

59
2.5.1 1-DOF damped system

60
2.5.2 N-DOFs damped system

61
 2.6 Modal analysis & peculiar situations
2.6.1 Prescribed boundary motion

66
2.6.2 Systems w/ rigid body modes

67
 3 Numerical Methods
3.1 Solution of the eigenvalue problem
3.1.1 Power method

70
3.1.2 Inverse power method

71
3.1.3 Shifted-inverse power method

72
 3.2 Direct time-integration methods
3.2.1 Central difference method

73
3.2.2 Newmark method

75
4 State-Space Formulation

77
4.1 Modal state-space representation

80
4.2 Examples

82
4.3 Random loads

88
4.3.1 Ergodic process through LTI system

90
4.3.2 Power Spectral Density

95
4.3.3 Shape filter

97
4.3.4 Cross-PSD

98
5 Control System Design

101
5.1 State-space C.S. design

104
5.1.1 Control law

105
5.1.2 Reconstruction of the state

106
5.2 Pole placement technique

109
5.3 Optimal control

111
5.3.1 Recovery of no-coupling case

113
5.3.2 Presence of random disturbances

114
5.4 Optimal observer: Kalman filter

115
 5.5 Validity of separation principle
5.5.1 Effect of uncertainties
5.5.2 Effect of neglected dynamics

117
 5.6 LQR design guidelines
5.6.1 Maximum values weighting

119
5.6.2 Eigenvalue shift (prescribed stability margin)
5.6.3 Implicit model following

120
5.6.4 Sensitivity-weighted LQR

121
5.6.5 Frequency shaping

122
5.7 Finite-horizon framework

123
5.7.1 Optimal tracking
5.8 Gramians

125
5.8.1 Balanced realization

126
5.9 Direct-output optimal control

127
 6 Thermoelasticity
6.1 Energy balance

129
6.1.1 Weak form of the thermal problem

130
6.2 Mechanical problem
6.3 One-way coupled thermoelastic problem

131
6.3.1 Quasi-static approximation

132