Gradient Enhanced Vibration Control

1
Gradient Enhanced Vibration Control

• Alex Forrester Andy Keane
• Computational Engineering and Design Research
  Group
• University of Southampton

2
(No Transcript)
3

• Optimised using fast receptance code instead of
  FE analysis.
• Performance highly sensitive to small changes in
  node positions.

4
(No Transcript)
5
Robust design

• Partially converged GAs?
• Multi-objective optimisation of vibration energy
  and standard deviation of vibration energy?
• Multi-objective optimisation of vibration energy
  and gradients?
• Multi-objective optimisation of vibration energy
  and std. of gradient enhanced approximation?

6
Regular boom

• Energy 40.2048
• s 11.3525

7
GA

• Energy 14.6477
• s 6.3352

8
NSGA2 energy / gradients

• s estimated from AD gradients.
• GA based algorithm produces Pareto front of non
  dominated designs.
• Time saved compared to local DoE techniques

9
Best energy design

• Energy 8.6073
• Predicted s 1.9351
• 100pt DoE s 2.3423

10
Best std. design

• Energy 12.4644
• Predicted s 0.4767
• 100pt DoE s 2.2143

11
Damper

• Boom concertinaed to provide vibration control
  mounting
• Leads to problems with
• Acute joint angles
• Beam intersections

12
Acute joint angles
13
Beam intersection problems

• Add random perturbations?
• Gradient descent geometry repair?

14
(No Transcript)
15
Future work

• Gradient enhanced local approximations for robust
  design
• GA / gradient descent hybrid optimisation
• AD of active vibration control analysis
• Design, build, and test of passive and
  active/passive control structures