Since my arrival at Algor 7 years ago, my team & I have been striving to improve the accuracy of the Algor Simulation software. These efforts have increased since the acquisition of Algor by Autodesk and have steadily been ramping up. I am going to post our progress in this thread.
Those of you familiar with Algor Sim know that Algor has a set of verification models called the Algor Verification Examples (AVE). There are 110 of these models and these models are run every release. These model have know results from well-known FEA references. Of course, Algor Sim passes this quality check every release.
A less biased test is to use models from a third party such as NAFEMS. We have run seven linear structural/modal and 21 nonlinear structural models based on NAFEMS models. In each of these cases, Algor Sim results match within reasonable tolerance to the published NAFEMS results. This benchmark activity was done with both Algor Sim 2011 and 2012 (in-development).
Software Development Manager
We have also been running various verification models as supplied by our competitors.
We ran the 6 LS-DYNA models supplied with earlier versions of ANSYS. Algor MES successfully ran all 6 models and the results matched within tolerance to LS-DYNA.
You mentioned comparison of Algor against NAFEMSbenchmarks, (National Association Finite Element Modeling Systems NAFEMS dot org)
1) Where can we see those results for perspective on what to expect on Algor simulation uncertainty ("accuracy")?
2) I am especially interested in performance of nonlinear vs linear modeling.
We are still debating on the best way to publish these results. Hence I don't have a document to send to you at at the moment. However, I recommend reviewing the results for the Algor Verification Manual (AVE) models. This set of models and manuals were part of earlier versions of Algor's installation. The manual is (or at least was) in the Subscription Center for download as well.