8 Replies Latest reply on May 9, 2007 12:49 PM by Brian Hoerner

    Impact in cosmosworks

    Designer Designer
      Dear cosmos users....

      Any body know is it possible to create an impact study in cosmosworks ?
      What I mean for example is impact on a car with car or maybe car with a wall.

      See image below, I get those picture from web.

      Thank You...!!!

      -DD
      SW 07 Sp3.0
        • Impact in cosmosworks
          Lin Shaodun
          A drop test? Use hard floor, and you need to calculate the equibirum G force due to impact.

          It will take a few hundrund hours to mesh and calculate as it's a non-linear process, please make sure you are using 64bit and have enough RAM.

          BTW, COSMOSworks result is just for rough (how rough? I don't know) reference only.
          • Impact in cosmosworks
            Designer Designer
            Thank`s for your info Lin...
            What do you mean about rough ? It`s mean the result is not accurate ?

            -DD
              • Impact in cosmosworks
                Lin Shaodun
                Just guessing, i don't think COSMOSworks can handle complex geometry like a car body. ( Anyone please correct me if I wrong)
                • Impact in cosmosworks
                  Peter Gillespie
                  I couldn't say how accurate, but you should be able to get results that are comparable to real life IF you are able to model it accurately. If you're not an expert using COSMOS, I would think there would be assumptions made that would limit the accuracy of your results on such a complex setup. But I have read of companies that make vehicles using COSMOS (though I don't know how they use it).

                  And I haven't worked with drop testing in COSMOS, so I don't know how that compares.
                • Impact in cosmosworks
                  genexxer genexxer
                  In cwx 2008 beta there is a new module in called "dynamic". It looks like the same solver as the astar module in geostar. this allows the direct integration of mass normalized eigenvectors obtained from natural frequency analysis into amplitude. So the effects of material and modal damping can be simulated. Linear approximations of a highly nonlinear event are usually only valuable as trend analyses.
                    • Impact in cosmosworks
                      Vince Adams
                      To build on what Genexxer said, impact problems can be solved in 2008 but crashing a car is a totally different story. If that is really what you are hoping to do (or something of similar complexity), I'd suggest enlisting the aid of someone who has done this before. This is a very complicated problem. The problems that are best suited for typical FEA nonlinear transient dynamics solutions (which use an implicit (vs. an explicit) solver are relatively slow and simple contact situations. Implicit solvers don't handle high speed, short duration events that well. How slow is slow? Tough call... another good reason to hook up with someone who's got some solid experience with these problems.

                      Good luck!
                    • Impact in cosmosworks
                      Brian Hoerner
                      One program we are looking at for crash testing simulation is LS-Dyna ( http://www.lstc.com/), this direction was given to me by a professor in North Dakota, we have found the same as what Lin, Vince and Genexxer have stated with CosmosWorks, and especially that drop testing won't do the job,