[TheMadHungarian]

Hi all,


Here's an issue my colleagues and I are stuggling for the last few weeks: we are simulating the progress of the blast shockwave in air at certain altitude.


The mesh is set up as a 1/8th of a complete sphere and each flat "side" is a symmetryPlane. The high p/T/U originates from the inside smaller spherical surface (called "blast") while the outer larger surface of the sphere is atmosphere (a waveTransmissive patch).
The boundary conditions are prescribed per the Friedlander equation (although negative T and p is meaningless as both T and p are absolute, so they donot dip below ambient values). In essence it means the T and p on the "blast" surface are described as time-dependent conditions: type uniformFixedValue; uniformValue table (.....


What we see during simulation (ParaFoam - Filters - Data Analysis - Plot over Line) is that the pressure wave is moving outward and decreasing...this is what you'd expect. But we also see a standing wave of low pressure located close to the inner sphere. That low pressure does not seem right since it just "sits" there, not moving, not decreasing even with 0.5 seconds simulation time. I dont think it makes sense at all since the pressure would equalize after a while.

This is a quick moving event, the blast wave comes out at about Mach 3, temps are 5000K at t=0 and decreasing as well as the blast pressure is 450 psi (3.1e6 Pascal) then decreasing.

We use rhoCentralFoam with laminar flow. The medium is air.



One problem is that since the velocity U at t=0 is defined as surfaceNormalFixedValue - this does not allow to describe the U over time although we could do that with T and p.


I have the case tar.gz attached...any input or insight would be useful. What's missing? What's not set up properly?
Are we not using the right kind of BC for T, p and/or U ?



Thanks!

[opedrofunk]

Hi Andy,

Came across this post. Interesting idea. There has been substantial progress recently extending OpenFOAM (e.g. the blastFoam solver) for high-explosive and air blast simulation (https://github.com/synthetik-technologies/blastfoam). If you and your group are working in this area, it might be of interest.

Best regards,
Peter