Efficient Methodology for Multibody Simulations With Discontinuous Changes in System Definition

A new algorithm is presented for accurately and efficiently simulating multi-scale multibody systems with discontinuous changes in system definitions as encountered in adaptive switching between models with different resolutions as well as models with different system topologies. An example of model resolution change is a transition of a system from a discrete particle model to a reduced order articulated multi-rigid body model. An example of model topology change is the transition of a system from a serial chain to closed kinematic loop topology or vice versa. An impulse-momentum formulation is implemented in a Divide and Conquer scheme which conserves the generalized momentum of the system and preserves the essential dynamics of the system. This process couples together the impulse-momentum equations of the individual bodies to formulate and solve the corresponding equations for the whole system using the boundary conditions. This results in an inherently substructured form of equations within the formulation making it highly conducive to adjusting / modifying model types within a substructure. The algorithm is applicable to general systems with either tree or closed loop topologies. This method maintains linear and logarithmic complexity in serial and parallel implementations respectively.