LS-DYNA Air Blast Techniques: Comparisons with Experiments
Numerical simulations used to predict events are always challenging. Among the challenges is establishing some basis for confidence in the results when no experimental results exist, i.e. a prediction. While there is no assurance that all the necessary physics have been included in the model, e.g. strain rate effects, until the experimental results are available for comparison, there are procedures the user can adopt in model development that will build confidence in the modeling. The first of these confidence building procedures is mesh refinement. In refining the mesh, or key parts of the mesh, the key results should converge. The results may not converge to the experimental result, due to possible missing physics or data, but a convergent model is an indication of a well posed model. Conversely, if the mesh refinement does not produce a converged result, this is an indication of an ill posed model. While some (all too few) users understand the value of mesh refinement, even fewer users appreciate the confidence provided by solving the problem at hand using different solution strategies. Too many users apply the method they know, even if alternative, or possibly better, solution techniques exist. LS-DYNA offers a menu of solution strategies and the knowledgeable user takes advantage of several of the solution strategies when predictions are required.
https://www.dynamore.de/de/download/papers/2015-ls-dyna-europ/documents/sessions-g-1-4/ls-dyna-air-blast-techniques-comparisons-with-experiments/view
https://www.dynamore.de/@@site-logo/DYNAmore_Logo_Ansys.svg
LS-DYNA Air Blast Techniques: Comparisons with Experiments
Numerical simulations used to predict events are always challenging. Among the challenges is establishing some basis for confidence in the results when no experimental results exist, i.e. a prediction. While there is no assurance that all the necessary physics have been included in the model, e.g. strain rate effects, until the experimental results are available for comparison, there are procedures the user can adopt in model development that will build confidence in the modeling. The first of these confidence building procedures is mesh refinement. In refining the mesh, or key parts of the mesh, the key results should converge. The results may not converge to the experimental result, due to possible missing physics or data, but a convergent model is an indication of a well posed model. Conversely, if the mesh refinement does not produce a converged result, this is an indication of an ill posed model. While some (all too few) users understand the value of mesh refinement, even fewer users appreciate the confidence provided by solving the problem at hand using different solution strategies. Too many users apply the method they know, even if alternative, or possibly better, solution techniques exist. LS-DYNA offers a menu of solution strategies and the knowledgeable user takes advantage of several of the solution strategies when predictions are required.