1 Reply Latest reply on May 28, 2008 7:00 PM by Bill McEachern

    Beam mesh - bad results

    Adam Sandler
      I'm analyzing a large lattice structure composed mainly of steel angle. Some members are bolted with a single bolt at each end (hinged in one direction), while others are bolted with 2 bolts at each end or welded (rigid end condition).

      I've had all kinds of issues getting this model set up correctly in Cosmos Works because there is no way to determine direction 1 and direction 2 for an angle cross section. I've found a very long work-around to determine which is direction 1 and 2, but it turns out not to make any difference in the results - the results can't be correct. Then I tried just using the plain hinged end condition (direction 1 and 2) for members that are bolted with a single bolt at each end, but again the results don't change at all. And when I look at the bending moment diagrams for those beam, they are carrying a moment. Shouldn't a hinged end condition prevent the beam from carrying a moment?

      Then I tried making all the members with single bolted ends into truss members (although I don't see what the difference is between a truss and a beam with hinged end condition). The results changed only slightly and surprisingly (suspiciously) resultant displacement decreased. But again when I look at the bending moment diagrams for the truss members, they are carrying a moment.

      Am I missing something here? How can a truss (or beam with hinged ends) carry a moment? Shouldn't they be 2 force members (tension or compression only)? Has anyone else come across this?

      Thanks,
      Adam
        • Beam mesh - bad results
          Bill McEachern
          Hi Adam,

          Shouldn't a hinged end condition prevent the beam from carrying a moment? If the beam is carying a distributed load then it will have a bending moment distribution along the span though zero at the ends if it is hinged. Otherwise it should not have a moment

          I didn't think trusses worked at all. Do they?

          You should fire up Geostar - you can do this whole thing in about an hour. Essentially you put in all the key points (every connection essentially) where the , then put in lines where the beams go. Mesh the beams and you can contol the end fixity. When you are meshing the beams there is a third point required to orient the beams cross section and there is a plotting option to show the directions. The stuff works in Geostar and it won't take that long to do it.

          It would be nice if the stuff worked with something else other than rigid connections though. They should grey them out when the stuff doesn't work.
          Bill