Student: Christopher Sullivan, Graduate Student in Mechanical Engineering, University of Iowa
Faculty Advisor: Hiroyuki Sugiyama
Adaptive Model Order Reduction Techniques for High-Fidelity Multibody Vehicle Models
High-fidelity computational models play a critical role in vehicle design and performance evaluation. These models, however, take significant amounts of computational time, even with modern high-performance computing. To address the limitations of existing high-fidelity computational models in multibody dynamics simulations, my research focuses on developing an adaptive model order reduction technique for high-fidelity multibody vehicle models in order to decrease computational time while limiting decreased accuracy. One of the areas which I am pursuing is off-road mobility prediction for ground and space vehicles, which is highly computationally intensive due to the complex interactions between vehicles and deformable terrain. If real-time computations were possible, control decisions based on high-fidelity models could be made at every point in time, enhancing vehicle autonomy capabilities on sandy terrain in challenging missions. Inaccurate prediction, on the other hand, could result in a vehicle/rover being immobilized in the sand, jeopardizing the completion of important missions.