Learning Forum

[rji229]

Hello,

I'm using the SPH method in LS-DYNA to simulate the high-speed impact problem (Using a metal ballistic projectile to impact a metal plate). I find that the internal energy is not consistent with different numbers of particles (different meshes) for the plate and the projectile. The results are listed below, which are the coarse mesh (70*70*19=93100 nodes), medium mesh (107*107*28=320572 nodes), and fine mesh (151*151*40=912040 nodes).

All the other setups are kept the same for the three cases. As can be seen, the internal energy is not convergent for these cases. For example, the internal energy at 1e-5 s is 2.8e9 (coarse mesh), 3e9 (medium mesh), and 3.5e9 (fine mesh) respectively. These results are calculated by the R 13.0 solver. I also tried to use R 14.0 solver but the results still show a different trend. Even for some cases, the job will stuck at 33% or 46% and I don't know why.

I would like to know how LS-DYNA calculates the internal energy and how I can verify it. We know that the internal energy can be divided into two parts. One is the strain energy, which can be calculated by particle strain and stress. The other is heat, which can be calculated by the particle temperature change. For strain energy, I plan to output all the stress and strain of all the particles in each timestep, then I can calculate the increment of strain δε, then time σ, and the total volume and sum, which is the internal energy. For the heat, I plan to output all the temperature change, then times the total mass and specific heat capacity, which is the total heat generated. However, in some cases, I find that the sum of total strain energy calculated and total heat calculated does not equal to the total internal energy directly plotted in *MATSUM.

 

Best,

Rong

[Ashish Khemka]

Hi,

Unfortunately this question is beyond the scope for ANSYS employees to answer on a public forum, so hopefully other users can chime in.

Regards,

Ashish Khemka