4 Replies Latest reply on Apr 18, 2013 12:14 PM by Brandon Donnelly

    Simulation of thin beam based stage

    Werther Zhou

      I want to simulate the deflection of thin beam based stage like the picture shown below. The thickness of the thin beam is 1 mm.

      What kind of mesh should I draw if I want to get acurrate results? Any principles to follow?

      Thanks

       

       

        • Re: Simulation of thin beam based stage
          Werther Zhou

          How to know if my mesh is good enough?

          • Re: Simulation of thin beam based stage
            Werther Zhou

            I create the mesh like below, is it good?

              • Re: Simulation of thin beam based stage
                Brandon Donnelly

                A good way to check if a mesh is good is to try several different sizes of mesh. If your stresses are changing drastically, then you're results have not converged and you need a finer and finer mesh. If the changes are small in all but one area try using a mesh control. In addition, having an analytical solution (possible ballparked) to compare is always a good idea.

                 

                There are times where your mesh will never give you convergence. In these times, you should check your model for features that lead to stress concentrations (sharp corners, holes, etc). If you think about it there is no such thing as an infinitely sharp corner in reality. Think about zooming into what appears to be a sharp corner until you get to the atomic level, that atom would still be a "sphere" so it would have a radius to it. Even super computers would have trouble calculating a mesh that was fine enough to capture the effects of a fillet that tiny, but the point is all edges have some sort of "radius" to them.

                 

                Its all about your model and assumptions.

              • Re: Simulation of thin beam based stage
                Jared Conway

                standard rules are 2-3 elements across a thickness. if you can't do that without killing your computer, make sure you're using the appropriate idealizations like shells and beams.

                 

                make sure you're using high quality elements. take a look at some of the articles on mesh convergence.

                 

                in the end, you'll know your mesh is good when your solution is close to what you expect it to be.

                 

                (note, your pictures didn't make it for some reason)