Summary
Manually take each simulation step in co-simulation app(s) rather than using the step callback.
Rationale
In the initial implementation of a co-simulation example (#537), the client side uses the step_callback in the interface to Ida::runSimulation to trigger network exchanges. This is already a problem since we want to remove the step callback parameter altogether. Also, we want to manually drive the solver steps so that we can synchronize time step size between the two instances in the co-simulation.
Description
Don't call runSimulation. Instead, drive the simulation by manually taking each step on both client and server side between data exchanges.
Additional information
No response
Summary
Manually take each simulation step in co-simulation app(s) rather than using the step callback.
Rationale
In the initial implementation of a co-simulation example (#537), the client side uses the
step_callbackin the interface toIda::runSimulationto trigger network exchanges. This is already a problem since we want to remove the step callback parameter altogether. Also, we want to manually drive the solver steps so that we can synchronize time step size between the two instances in the co-simulation.Description
Don't call
runSimulation. Instead, drive the simulation by manually taking each step on both client and server side between data exchanges.Additional information
No response