5 Replies Latest reply on Dec 19, 2013 7:58 PM by Shaun Densberger

    Finite element result validate through calculation

    Sur S

      Hello and a good day. I need a help regaring the theoretical calculation of my design which is not same as von misses stress result. Tq

        • Re: Finite element result validate through calculation
          Jerry Steiger

          Sur,

           

          You will have to give us considerably more information for anyone to be able to help you.

           

          Jerry S.

          • Re: Finite element result validate through calculation
            Jared Conway

            Recommend starting with something simple that you can calculate and post it here with both the calcs and analysis setup along with a description of the issue.

             

            Also, look at the validation examples In the software to see that it had been checked. Most times the issue is setup or interpretation. For example using stress at a singularity as a measuring point.

              • Re: Finite element result validate through calculation
                Sur S

                The plate dimension

                 

                Length      = 200mm

                Width       =100mm

                Thickness = 5mm

                The force given on the plate surface area is 49N

                With A extruded cut of hole with 1mm depth and 25mm of diameter which I had defined as fixed geometry.

                In this case I'd try to calculate by hand calculation but the result are not same as in the von misses result in the solidwork. Calc.jpg

                  • Re: Finite element result validate through calculation
                    Jared Conway

                    1. You are comparing a 2d problem to a 3d problem. Is that a fair comparison? You may want to consider making the problems more consistent.

                    2. You are comparing results at what looks like a singularity to your hand calc. You may want to look at our article on blog.hawkridgesys.com about mesh convergence.

                    3. From a setup perspective, use a split line instead of a cut for your restraint.

                    4. I don't see any documentation of your assumptions for your calcs or your simulation. This may help you understand the differences,

                    • Re: Finite element result validate through calculation
                      Shaun Densberger

                      I'm having a hard time understanding what you're trying to model and what you're calculating by hand, but from what I understand:

                       

                      You have a plate you a uniform surface load on top and the plate is supported by a circular rod at the center. If so, then here are my thoughts:

                       

                      For your FEM:

                      • You cut into the material to get your surface region. Removing material (as well as creating sharp reentrant corners) will greatly amplify stress.
                      • It doesn't seem as if you performed a convergence study; your results don't have any error tied to them.
                      • You've modeled the support with a fixed constraint. Does the real system behave as a fixed constraint?

                       

                      For your hand calculations:

                      • You're calculating your shear stress using F/A. That's "average" shear, whereas you need to calcualte maximum shear. In other words, you're assuming a constant shear stress distribution through the thickness, when this is not the case.
                      • You don't consider bending stresses at all.
                      • If you're going to do this with a 2D assumption, you need to apply axi-symmetric assumptions.