Co-simulation exchange boundaries
Record how separately solved models exchange values and advance time.
Research note · November 14, 2026Research question
When do coupled simulators disagree because of their communication schedule?
Method
- Define two bounded components and list the variables exchanged between them. Record units, causality and the meaning of each communication point.
- Run the coupled case at several exchange intervals while retaining each component’s solver settings. Compare interface values and system response rather than assuming a finer internal step fixes coarse communication.
- Document how errors, rejected steps and unsupported rollback are handled. Preserve the orchestration policy with the run so the coupling method can be reproduced.
Records to retain
- Component versions and interfaces
- Communication step schedule
- Internal solver settings
- Interface traces and error handling
Source basis
FMI separates model interfaces from co-simulation execution semantics and communication points. A coupling policy is part of the experiment, not merely a transport detail between otherwise independent models.
Interpretation & limits
FMI compatibility does not by itself validate the coupled physics. This protocol does not announce FMI support in Imprint or guarantee that every imported model can roll back state.
Next experiments
- Build a two-component fixture with known coupling sensitivity.
- Compare two exchange schedules before adding more components.
References
Technical reference
Implementation details supporting this note, drawn directly from the maintained documentation. These are technical references, not additional papers or new experimental results. Expand a topic to inspect its method, fields and limits.