Student: Vivek Ojha
Sponsor: Air Force Research Laboratory (AFRL) at Dayton, Ohio
Numerical simulation of fluid structure interaction (FSI) is a complex and challenging problem that finds relevance across disciplines. The focus of our work is on the aeronautical applications of FSI where numerical simulation can be used to improve the understanding and prediction of coupled instabilities observed in aircraft. The interaction between the two systems is non-linear and involves multiple scales, thereby making the coupled system challenging to solve. Many approaches have been suggested for simulating fluid-structure interaction. This work implements a high-order partitioned approach which couples the two subsystems in a high-order manner without sub-iterations. The main motivation for developing a high-order FSI solver is to be able to accurately predict coupled outputs like lift, drag and structural deformation.
Increasing the accuracy of a simulation by refining the entire space-time mesh makes the simulation computationally expensive and inefficient. A better strategy is to evaluate errors in the output of interest and to adapt the mesh in the regions of space time which contribute most to the error. Such methods are called output-based adaptive methods. They offer a systematic approach for identifying regions of the domain that require more resolution for the prediction of scalar outputs of interest.
This project combines a high-order fluid structure interaction solver with an output-based mesh refinement strategy to demonstrate increased accuracy of coupled outputs at lower computational costs.