3 Replies Latest reply on Oct 15, 2015 4:24 PM by William Turk

    What is the best way to constrain flanges for FEA?

    William Turk

      I'm running FEA on an adapter with flanges, and I'm getting higher stresses than I think I should.  I've tried redefining the fixtures, and I get different results.  Here are the three scenarios I've tried:

       

      1) Constrain the vertical face of the pilot axially, constrain the cylindrical face pf the pilot radially, constrain the bolt holes radially to prevent rotation.  (This yields the biggest stress)

      2) Constrain the vertical face of the pilot axially, constrain the cylindrical face pf the pilot radially and rotationaly.

      3) Constrain the vertical face of the pilot axially, constrain the cylindrical face pf the pilot radially, constrain the outer diameter of the flange rotationaly.

       

      Which of these is most accurate?  I feel it should be 1), but I get "hot spots" that way.

        • Re: What is the best way to constrain flanges for FEA?
          Michael Schrider

          Assuming the "force" values represent bolt head or washer bearing pressure and that each flange of the adapter is mated to a similar rigid flange:  I would suggest:

          1. Restraining each of the flange faces normal to the faces.
          2. Radial restraint on each of the pilot cylindrical outside faces.
          3. Radial restraint on each of the bolt holes.
          • Re: What is the best way to constrain flanges for FEA?
            Albert Veldhuis

            Dear William,

             

            I'm not sure what you're trying to prove, but I´ll try to help.

             

            It looks like an FEA study for a 6" 2500# flange. You might want to check standard dimensions for such a flange first.

            According to ASME B16.5, blind standard ones should be 114.4mm thick, with 8pcs. 2" bolts. Yours is way too thin.

            Having remodelled that flange, you should get the results you expect when you apply an elastic support to the gasket surface of the larger flange. About 50kgf/cm/cm^2 should keep your model constrained. The other constraints can be deleted.

            I haven't recalculated your reaction forces, because you don't share the line stress or gasket data. According to standard you should apply gasket loads to the corresponding surfaces.
            There is a good example of how to determine the loads on these narrow face flanges in the ISO 13445/3 chapter 11,5.

             


            Hope this helps,
            Cheers, Albert.