Beam elements in LS-DYNA

Hello everyone,

I am checking the capabilities of beam elements in LS-DYNA v.971 Implicit solver. I created a benchmark problem of a small deformation of a cantilever beam under lateral point load at the free tip. Modelling the beam using only one element yields a ~ 15% overestimate of both free end deflection and first natural frequency compared to analytical solution. Well, that error quickly diminishes by increasing the number of elements to 6 elements. Therefore, I solved the same problem (again with one element) using both ANSYS and COSMOSM and the results were within 1% of the analytical value. I tried digging for the element formulations used in LS-DYNA, but could not identify the reason. I'd really like to discuss this problem with anyone who encountered it before.

 Regards,