6 Replies Latest reply on May 1, 2013 4:01 PM by Jared Conway

    Difference between Sheet metal and Extrusion simulation

    April Dunham

      I have done a simulation using approximatley a 90 degree bend in it. I was having issues with the sheet metal feature so I decided to make it as an extrusion first. I then ran a simulation with incremented amounts of force and collected the results. Recently I re-ran the simulation after getting the sheet metal feature figured out. Using the same material and the same shape of a part, The resulting peak stress in the part was less than half of the peak of the extruded part. Can anyone help me in explaining why?

        • Re: Difference between Sheet metal and Extrusion simulation
          Bill McEachern

          Could be lots of things but first you should investigate the following:

          1. shell definition in the sheet metal part - thick or thin. I think it is thin by default and you may not be able to change it. YOu can offset a surface to the middle and delete the sold body for hte simulation and hten check the differences you get in thick and thin. A thick shell will produce correct results right at the bend in my expereince.
          2. Did you look at the shell results with the shell thickness rendered. If not then which side of hte shell were you looking at? Chekc htem both.
          3. The extruded part will need at least 2 elements through the thickness to produce a decent result and three is a better but two should do it in most cases.
          • Re: Difference between Sheet metal and Extrusion simulation
            Mike Pogue

            In addition to 2 elements through the thickness, you nee about 2 elements through a 90deg radius (8 around a circle). Both assuming you are using quadratic (high quality) elements. The two solutions should converge to the same answer if you use elements that are small enough.


            They won't. But they should.


            If you drive both solutions to convergence and they are still different, the solids are probably more reliable. There are fewer simplifying assumptions in solid elements. The advantage of shells is that they are less computation-intensive.

              • Re: Difference between Sheet metal and Extrusion simulation
                Bill McEachern

                I have tested the thick shells and for this type of case they are pretty much spot on with solids. If they are not close it might be better to look at finger trouble issues (ie your set up and interpretation).

                  • Re: Difference between Sheet metal and Extrusion simulation
                    Jared Conway

                    Shells vs solids should come out to pretty close numbers. Otherwise they wouldn't exist in the software. That being said, shells may be closer to the real answer if there aren't enough elements in the solid. (Another reason why shells are in the software) My experience is that they come pretty close. Thin vs thick, unless it is a real thick shell, I think the developers recommend going with thin shell formulation because it may act overly stiff. But I think before we go to that detail, we need to know if the displacements are the same and if the setups are equivalent. I could also imagine that the shell might be fixed on an edge vs a solid face/edge, the resulting displacements and stresses would be different.