[ander]

I want to model boundary conditions from a blast wave, where the origin of the blast wave is located outside the solution domain, it might be far off. I want the time-history of the blast wave to be specified by some semi-empirical method. I think these methods typically let you specify the equivalent amount of TNT, and how far away you're located, and it gives you a pressure-time history and perhaps some more.

I wonder if this information (provided by some semi empirical relation) could be sufficient or not to implement boundary conditions for a 2D Euler solver. I recall that you should specify as many field variables as there is incoming waves relative to the boundary, so if the incoming wave is supersonic you should set all 4 field variables, and if it's subsonic you should specify 3 (and calculate the last one from a value inside the domain(?)).

I guess the blast prediction methods that only provide pressure-time history are typically used for structural mechanics where you only care about the forces on the structure. Could it still be possible to estimate the rest of the needed flow variables somehow?

Thanks!