VM149

VM149
Residual Vector in Mode-Superposition Harmonic Analysis

Overview

Reference:

J.M.Dickens, J.M. Nakagawa, M.J. Wittbrodt, “ A Critique of Mode Acceleration and Modal Truncation Augmentation Methods for Modal Response Analysis”. Computers & Structures, Vol.62, pp.985-998, 1997.

Analysis Type(s):

Modal Analysis (ANTYPE =2)

Harmonic Analysis (ANTYPE =3)

Element Type(s):

Spring-Damper (COMBIN14)

Structural Mass (MASS21)

Input Listing:vm149.dat

Test Case

A mode-superposition harmonic analysis is performed on a spring-mass model for two cases:

Case 1: Extracting all available modes
Case 2: Extracting one mode and residual vector (RESVEC)

The following values are computed for both cases and compared against the target values shown in the reference:

  • Peak displacement amplitude along X direction at node 4

  • First and second peak force amplitude for the spring element

  • Frequency corresponding to first peak displacement and first peak force amplitudes

Because the target values are taken from figures, they are reported to an appropriate accuracy in the results comparison section. The modal truncation (MT) method shown in the reference corresponds to the residual vector method (Case 2).

Figure 205: Spring-Mass Model

Spring-Mass Model

Material PropertiesGeometric PropertiesLoading
Spring Stiffness (K) = 1000 N/m
Mass1 = 1.0 Kg
Mass 2 = 0.5 Kg
Damping ratio = 2%
Total Length = 5m
Force along X direction at node 4 = 1N
Excitation frequency: 3 – 70 Hz

Analysis Assumptions and Modeling Notes

The spring-mass model is represented using 2D COMBIN14 and 2D MASS21 elements without rotary inertia. The X and Y axis of the spring model are inverted from the reference to model it along global X axis. The model is fixed at both the ends, and displacement along the Y direction is constrained on all nodes. In order to obtain four distinct modes, the mass at node 5 is set to half the value of the other three masses defined at nodes 2, 3, and 4. A modal analysis is performed first using the Block Lanczos eigensolver. A mode-superposition harmonic analysis is then performed with an excitation frequency range of 3-70 Hz and a force load along the X direction at node 4. A damping ratio of 0.02 (2%) is defined using the DMPRAT command. Calculations are done in /POST26 to determine the displacement and force amplitudes. Results are retrieved using the *GET command.

Results Comparison

TargetMechanical APDLRatio
Extracting all available modes
UX_MAX (m)0.002500.002261.106
F_MAX (N)10.0000010.040350.996
Frequency @ UX_MAX (Hz)10.1000010.102001.000
Frequency @ F_MAX (Hz)10.1000010.102001.000
F_MAX2 (N)4.500004.288671.049
Extracting one mode and Residual Vector
Residual Mode21.865021.865231.000
UX_MAX (m)0.002500.002261.106
F_MAX (N)10.0000010.050020.995
Frequency @ UX_MAX (Hz)10.1000010.102001.000
Frequency @ F_MAX (Hz)10.1000010.102001.000
F_MAX2(N)4.500004.729240.952

Figure 206: Displacement Amplitude and Spring Force Amplitude Versus Frequency for Case 1

Displacement Amplitude and Spring Force Amplitude Versus Frequency for Case 1

Figure 207: Displacement Amplitude and Spring Force Amplitude Versus Frequency for Case 2

Displacement Amplitude and Spring Force Amplitude Versus Frequency for Case 2