2 Replies Latest reply on Mar 31, 2018 5:48 AM by Maurits De Wolff

    meshing fails for truss with solids

    Maurits De Wolff

      Hello,

       

      I've found many messages on this forum from users not grasping the workings of simulation. I'am one of those.

       

      I applied 5mm Mesh Control to the hinge parts. But mesh fails for the construction as a whole time after time. Like other novices I have trouble to interpret the simulation diagnostic language Solidworks uses.

       

      In previous attempts I've once managed to get simulation to run. But structural members started flying off the jib part. The jib as a whole had degrees of freedom it should not have. So I fixed the connection between the jib itself and the (male part of the) hinge.

       

      Upon getting the Failure Diagnostics warning the male part(s?) of the structure can't be meshed I click OK. Then the meshing continues fairly rapidly and ultimately fails for all parts involved.

       

      I have consistently interference of structural members in the parts, where the 2.5" pipes intersect with the 4" pipes. I don't think that is right but I do not understand how to prevent this with dissimilar tubing. Miter feature doesn't work here as far as I can see.

       

      There is a gap of 2mm btween the face of the male and female hinge part. Ptototype is designed this way and even with no load in the crane in real life the hinges fail. I want to simulate what's happening inside and around the hinges. Therefore I excluded the pins from simulation. Torsion of the 2 truss parts is likely and of interest to me also. Both truss parts don't fail in real life. Which is the reason I made the overall mesh as coarse as possible.

       

      I'm running SW 2016 student edition. Could someone shine a light on this please?

       

      Thanks Maurits

        • Re: meshing fails for truss with solids
          Dan Golthing

          Using simulation without a strong understanding of what is actually happening isn't recommended.  You can get results that look good but are erroneous.  You should be supporting your results with classical calculations to verify that the software has been set up properly and has run properly.  There are issues that the software won't necessarily show, such as buckling in the linear static package.  Even in the buckling module, non-Euler buckling is not detected.

           

          You've got a mixed mesh situation happening.  beams and solids.  this rarely works. 

           

          You could change everything to solid mesh and give that a try.  Longer mesh time and longer run time, but much higher probability of success.  You will need small elements.

           

          Simulation is something that takes years to get good at.