Stabilized Co-Simulation of Coupled Problems Including Fields and Signals

Abstract

Co-simulation is becoming an increasingly integral and indispensable technique for solving today’s challenging engineering problems. By means of this code coupling technique, the engineering problem is partitioned as an assembly of different subsystems exchanging solution information at run time. The inherent advantage of co-simulation in contrast to the monolithic approach is that it allows the (re)use of well-established and specialized simulation software to be combined, with minor alterations. Furthermore, co-simulation allows different fidelity models to be combined at different stages of the design process. Unfortunately, this partitioned treatment of the individual system poses stability and accuracy challenges.

A novel co-simulation algorithm is introduced, referred to as the Interface Jacobian-based Co-Simulation Algorithm (IJC\-SA), which overcomes present stability issues. The algorithm can solve co-simulation scenarios involving an arbitrary number of fields and signals. Due to the fact that the IJCSA is based on the residual form it handles algebraic loops in a natural manner. Furthermore, the individual simulators can run in parallel without flow dependency reducing the wall-clock time of the simulation, since the subsystems do not have to be executed using the classical Gauss-Seidel pattern. A thorough stability analysis of the IJCSA is presented.

In order to demonstrate the applicability, several industrially relevant examples are solved by using the IJCSA. The shown examples range from a fully coupled fluid-structure-signal interaction with closed-loop control to fully coupled emergency brake maneuver of a wind turbine. Here the interaction of the generator/gearbox, flexible composite blades, control unit and the three-dimensional flow field is taken into account. Furthermore, the simulation results are validated against measurement data from the National Renewable Energy Laboratory (NREL) Unsteady Aerodynamics Experiment Phase VI, performed in the NASA AMES wind tunnel.