Simulation of tennis net <-> ball collision?

[five1mon]

Hey All
I'm going to implement a tennis net for a video game. I already have the animation of the net in place (I think) through cloth simulation. What's missing is a physics simulation of the collision between the ball and the net?

I think the net could be modeled as an elastic rectangular surface, where the vertical edges are rigid and the horizontal edges are less elastic then the inner surface.

Can anybody direct me how to compute such a thing?

The thing has to run in real-time on quite low-end machines so the computations can't be too expensive(complicated).


Thanks

 

[DavidSnider]

The net swaying doesn't really have much effect on the game, so do you really think it's worth the effort to model it realistically? Maybe just do a collision detection, animate the net with your cloth simulation and have the ball bounce as if off a rectangular surface w/ added dampening.

 

[five1mon]

The net swaying doesn't really have much effect on the game, so do you really think it's worth the effort to model it realistically? Maybe just do a collision detection, animate the net with your cloth simulation and have the ball bounce as if off a rectangular surface w/ added dampening.

For the cloth simulation I'm using a mass-spring system where the movement is propagated through verlet integration (this is the standard form of cloth simulation in games). As input I need the ball to perturb the intersected vertices in order for the system work to work.

I totally agree that the system doesn't need to be realistic. But I would like to have the ball perturbing the vertices with a different magnitude dependent on if the ball hits in the middle, near the posts, ner the top etc. The nets will mostly be viewed up front and there will be cut scenes / re-plays where the net moves in slow motion.