2D Muzzles. We have here an adaptive 2D computation of a cannon blast. That was our first attempt using isotropic mesh refinement. We already have some anisotropic axisymmetric results (thanks olowee) with a small aspect ratio of 5. Those results, which show not so badly resolved shocks (we think ;-), were computed on one pentium III machine in less that 2 hours, including 100 mesh refinements. The same computation was taking 8 hours with the isotropic technique (and the results were not as good).
Pressure contours (left) and adapted meshes in the axi case
3D Muzzles. We are starting some non adaptive 3D cases. Here is a closup of the cannon next to the muzzle break. The image shows velocity vectors at an early stage of the computation.
Pressure contours (left) and velocity vectors (right) near the muzzle
An animation of a 3D blast coming out of a cannon, taking ito account the ground. This is a non-adaptive computation, I'm going to provide adaptive aniso images A.S.A.P.
