7 Replies Latest reply on May 4, 2012 12:30 PM by Jerry Steiger

    Testing labresults in simulation

    Nikolai Hansen


      I've done some pressuretests of IPE80 (S355) beams in a hydraulic press to test their capacity. I'm trying now to see if I can recreate the tests in Solidworks. The setup is like this (without the stiffeners):

      IPE80 press.jpgIPE80 without stiffeners.jpg

      The beam lies free between the foundation and the press, only held in place by friction. The press doesen't move or rotate in any direction.


      For the simplicity of it I first tried with a shell model. I've added rotation fixtures on the top flange to simulate the press. Here are the results for the buckling and the static test:

      IPE80 buckling.JPGIPE80 Static.JPG


      The buckling test was applied with a force of 1 N so the buckling load is 180,47 kN if I understand this right.


      In the static test I lowered the force until I could get a maximum tension of 355 N/mm2 which is the yield strength of S355. The force here was 113 kN.


      My questions are:

      - Is buckling and static simulation the right tests to replicate this experiment?

      - I've just applied force and runned the tests. Is is this simple or do I need to do something more with the results?

      - The material (Solidworks DIN Materials - DIN Steel (Structural)) doesn't seem to apply to the results. When I change the material the results are the same in both buckling and static. Am I doing something wrong?

      - All the materials have a yield strength of 275 N/mm2. This can't be right?


      This is my first go with Solidworks so I hope you can bear with me

        • Re: Testing labresults in simulation
          Anthony Botting

          Hi Nikolai: I'm fairly certain the Euler Buckling calculation (in SW Simulation) will not apply for this case because the slenderness ratio of the web is too low.  I found a reference on Engineer's Edge, link attached. It might explain some of what's happening. - Tony



          • Re: Testing labresults in simulation
            Anthony Botting

            Hi Nikolai: Just for kicks I ran it, but examine mode 2 of the buckling study and found the same buckled shape as in your lab test. Mode 1 allows horizontal slip of the top flange (unlike the lab test)- so at least the shape at mode 2 is like the lab test. But, again these calculations are for linear behavior and only apply for a much "taller" web, so I believe there is some localized yielding going on in the lab test, then it buckles. That means you'd have to conduct a nonlinear simulation with yielding to try and replicate the lab test.

            Regarding switching materials - you would get similar answers if the material modulus is the same (because the linear buckling and static solutions use material modulus and not the strength - other that posting a factor of safety in the static simulation).

            I did find too, many of the DIN materials have the same strength in SW material database - so you might want to make your own material (just copy and paste one of the existing ones, change the values, rename it, and save it). Hope that helps a little.  - Tony

              • Re: Testing labresults in simulation
                Nikolai Hansen

                Thank you for the feedback!


                Yes, I figured out from my calculations that the slenderness was too low for buckling to be a problem. But it gave a pretty good graphic display of the buckling shape (I've tested with stiffeners too) from the experiments, as you said.


                It's been a while since I've had FEM-theory but if it's the E-module that is being used to calculate the stiffness that explains a lot


                Looks like my laptop isn't good enough to run a nonlinear static simulation. I'll try back at school tomorrow.

                But is it not possible to use regular static simulation? I thought if I knew the average maximum tensions in the beam based on the buckling/yield load from the experiments and the cross-sectional area, I could run a static simulation with a given load and then use the ISO-clipping tool to see if there was any tensions within the beam that exceeded the maximum allowed tension?