When two simulation systems are connected via HLA how are the physics engines linked?
For example, if two armoured vehicles (entities) tank A from federate A and tank B from federate B collide how is the collision physics resolved?
As I understand it, there are collision and detonation interactions that convey who hit who and the velocity but is this enough to fully resolve the collision effects?
Presumably, each federate is responsible publishing the collision interaction message and applying those message to its own entity physics engine. I can't help but think, surely there is more to it that this.
Any details or resources would be greatly appreciated.
In HLA, the RTI doesn't calculate anything it merely facilitates the communication between the federates and keeps track of time. If you have a physics engine in your federated simulation, you either have to have it implemented as its own federate that other federates need to query to resolve the physics, or each federate keeps track of their own physics.
In your example, two federates each publish a single object of type "Tank" which has attributes "Geometry" and "WorldCoordinates". Lets say they keep track of their own physics. Each federate should subscribe to every other "WorldCoordinates" of every physical object in the region, and calculate when there is a collision between their owned object and the colliding object. If you have a separate federate calculating physics (i.e. it's a submarine simulation and red team positions are masked), then the physics federate needs to subscribe to all the "WorldCoordinates" and "Geometry" attributes and calculate collisions based on that then send an interaction to the federates that own those objects.