9 Replies Latest reply on Jul 6, 2015 1:47 PM by James Riddell

    Prediction of behavior: is it going to collapse?

    Stefano Tiburzi

      I run a simulation of an applied load on an aluminum profile insisting on a small recatangula area (pretending to be the tire imprint). The yield resistance is 170 Mpa, strentgh is 200 MPa. Apparently only in a small area this value has been reached (it's not event throurg the thickness), whereas the remaining volume stay below. This body is going to collapse or have to be strenghten? maybe it's a stupid question.

       

      profile.jpg

        • Re: Prediction of behavior: is it going to collapse?
          Andrei Popov

          Of course your stress in the red area is 326 MPa which is way over 170 Mpa so it will buckle in the plastic deformation and then it will crack.

          Even if it stays with a fracture it won't take long until collapse, especially after several cycles of loading.

          • Re: Prediction of behavior: is it going to collapse?
            Nandish Datti

            Since the max stress has reached a value of 326 MPa the structure will fail for sure but if you want to know whether the structure is going to buckle first or yield then you should carry out a buckling analysis.

            From the buckling analysis you can calculate the lowest buckling load factor (BLF) and you can compare that with the stress factor of safety (FOS) received from the static analysis. If BLF is lesser than FOS then the structure is going to buckle first else the structure will fail by yielding.

              • Re: Prediction of behavior: is it going to collapse?
                Stefano Tiburzi

                Unfortunately I cannot run a buckling analysis, I have a Solidworks Premium license with Simulation Standard.

                Buckling analysis shoudn't meant to be applied for beam elements with high slenderness? Or It's just fit with aluminum extruded profile like this one?

                Last but not least important thing I'm worried about this kind of simulations puzzle me when 99% of the structure pass the test while 1% are far above tensile strenght in a tiny volume or even on the surface of an element. What are your conclusions on that case? Do you suggest to your customers to redesign the assembly or simply consider the better element that is 100% below tensile strength even if the ratio between cost increment and general safety  are very high?

                Than you for your answers!

                  • Re: Prediction of behavior: is it going to collapse?
                    James Riddell

                    This does not appear to be a buckling case if I judge your applied load(s) correctly - small rectangle on top surface with tire pressure pressing down?  It does appear as if the compression load due to bending has been exceeded.  Have you tried running this in 2D?  Do you get similar results?  Are you restraining the model correctly?  Have you done the hand calculations to see what bending stress should be present?

                      • Re: Prediction of behavior: is it going to collapse?
                        Stefano Tiburzi

                        I think I applied the correct restraint as far as I know this part would work. No, I haven't run in 2D because the load aren't suited (they aren't symmetrcial nad cannot be projected correctly on a 2D simulation) and I avoided hand calculations because internal stresses cannot be got with the standard equation being an aluminum profile with a variable section in the z-direction.

                        I run another part with p-adaptive method and i got this:

                        Appunti01.jpg

                        On the left I plotted iso-section above yield strenght (part run with symmetrical fixture); the covergence stopped and there's a tiny volume above 410 MPa which is tensile strenght for S275JR steel. How can this be fixed? Have to change the entire project or can I solve in other way, i.e.welding a plate. Could this be a numerical error? Could the hole edge increase locally the stress so high?