This paper develops and demonstrates a strategy for identifying the lack of dynamic interaction, or coupling, between potential design modifications to a vibrating base structure. Such decoupling may be exploited to efficiently compare the performance of competing engineering designs. In particular, it is shown how different design modifications may be represented as the addition or removal of substructures. When these substructures are uncoupled according to the metric developed here, the computational cost of determining the optimal system design can be greatly reduced. For example, if a designer considers seven possible modifications and wishes to examine all possible combinations of the modifications, 128 possible structures must be analyzed. However, if all modifications are dynamically uncoupled, significant computational effort need only be spent on eight of the possible structures in order to generate the responses for all remaining designs. Example problems demonstrate this cost reduction and illustrate cases where dynamic uncoupling occurs.
|Original language||English (US)|
|Number of pages||22|
|Journal||International Journal for Numerical Methods in Engineering|
|State||Published - Jan 27 2017|
All Science Journal Classification (ASJC) codes
- Numerical Analysis
- Applied Mathematics