Congratulations to our Postdoctoral Fellow, Dr. DING Zhe for receiving the Best Oral Presentation at Hong Kong Scholars Annual Symposium 2021.
Project: Topology Optimization of Viscoelastically Damped Structures Under Time-Dependent Loadings: Sensitivity Analyses and Computational Considerations
Topology optimization has grown into a pervasive and versatile tool for developing novel conceptual designs in a wide variety of applications. Despite this growth, the bulk of topology optimization research remains focused on undamped or viscously damped systems, with only a handful of studies investigating viscoelastic damping systems. Usually, the transient response topology optimization problems for viscoelastically damped systems are computationally expensive. Therefore, there is a great desire to reduce the computational cost for such problems.
The sensitivity analyses are the most time-consuming part in topology optimization. In our research, some efficient sensitivity analysis methods, such as model order reduction techniques and discretize-then-differentiation method, are developed for capturing the derivatives of the response functions for viscoelastically damped systems. Both adjoint variable method and direct differentiation method are used. The computational considerations, including accuracy, consistency, computational cost and implementation effort, are discussed and compared for the proposed methods. The proposed sensitivity analysis methods are incorporated into a density-based macrostructural topology optimization framework and verified by some examples.
The results indicate that computational costs are significantly reduced by the proposed topology optimization framework. Also, the proposed sensitivity analysis methods are feasible to be extended to microstructural and concurrent topology optimization.
To learn more about the project: Click here!