I am using SPH to simulate Flyer-Plate impact at impact velocities in the range of 1000m/s to 2000 m/s. For an impact velocity of 500m/s, the code predicts the free surface velocity of the target as observed through VISAR in experiments. But at 1000m/s, particles start flying even before the shock wave reaches the free surface of the target. I found that the corner particle in the target which first comes in contact with the flyer gets some spurious acceleration in both the logitudinal (direction of impact) direction and transverse direction.I did not use any artificial damping. So, I thought if damping the oscillations would help.
I tried doing conservative smoothing as proposed by Hicks, Liebrock, Applied Mathematics and Computation, 150 (2004), pages 213-234 (http://www.sciencedirect.com/science/article/pii/S0096300303002224). This prevents the flying of particles even at 5000m/s impact velocity. However, this smoothing is not controlled and hence a lot of information is lost in the process. Dicontinuities become weak or vanish completely. I could not get the predicted pressures and velocities in the target material.
Kindly help me in this regard.
Siva Prasad AVS.