Time-domain coupling about mphys HOT 2 CLOSED

There is a demonstrator of the OpenMDAO black box approach in the examples directory using a modal structure time integratory. It proves the concept is doable, but the time step OM problem needs to be generalized to work with more than just the modal solver

from mphys.

Modal solver example

• prototype ways to handle the various modes people would want to use: async file IO, checkpointing algorithm, all in memory, and don't save time states (primal analysis only)
• establish a standard for time domain builders. Follow precoupling, coupling, postcoupling concept. May need to add preintegrator and post integrator options too.
• Determine how to handle cost functions that cannot be computed step by step (contributions accumulated during time loop
• submit POEM to request mode where multiple cost functions are handled in one solve linear/ apply linear call.

from mphys.