2 Replies Latest reply on Mar 21, 2011 9:05 AM by Bill McEachern

    Any advice for composite, nonisotropic beam models?

    John Fitch

      I will be creating a tapered beam (a spar for a wing) made with graphite layers.  This will generally be a cylindrically or elliptically shaped beam that tapers down in diameter.  The 'taper' will probably be stepped piecewise to accommodate another manufacturing need.  Because the wing has lift and twist, the fiber direction may spiral around the beam, out toward the end. 


      I realize that using Simulation for this is probably more instructive than definitive.  But it could help with some optimization studies, and we thought it might be useful to give it a try.


      Before I jump in, I wondered if some people out there had advice or tips.


      I have read a related post:





        • Re: Any advice for composite, nonisotropic beam models?
          Jerry Steiger



          Notice the stunning silence? I think this is an indication that SW Simulation may not be the best tool for the job. I've been reading back through the posts on this forum to familiarize myself a bit with Simulation before giving it a try. (We have one seat and have a seat of ANSYS that I've been using. Because ANSYS is not very designer-friendly, we are hoping to expand our usage of Simulation.) I noticed another post where a person analyzing stresses in a layered composite structure found that the Factor of Safety tool gave him nonsensical answers. I see a number of question involving composites going unanswered. I think this is an indication that SW hasn't put the effort into composite models that is needed to use it for serious design. It seems that this is a highly specialized, complex, emerging technology that may not be ready for a more general purpose tool like Simulation to make an entrance. I would love to be proved wrong.


          Jerry Steiger

          • Re: Any advice for composite, nonisotropic beam models?
            Bill McEachern

            Your options are:

            1) Modle as beams: Get a laminate calculator and compute the material stiffness of each section of the beam, then add orthotropic - maybe isotropic might be sufficient by the sounds of it, material definition for each section, and apply to each section of beam. Divide up the beam into separate solids for each thickness change.

            2)Model as shells: build a surface model with separate surfaces for each ply change, add materials to materails library, add laminate shell definitions.


            The one caveat is the handling of directions - simulation doesn't handle non-orthoganl directions all that well. You can approximate directins by breaking hte surfaces up in to smaller patches and adding co-ordinate systems to each one to et a better approximation. This is not simulations sweet spot but you can probably get a ball park estiamte.


            I would advise that you do some simple tests that you can compare against known data or other calculation methods and figure out the subteleties.