7 Replies Latest reply on Dec 19, 2013 8:56 AM by Steven Kiley

    Nonlinear FEA of a Crimping Process

    Steven Kiley

      Hey guys,

       

      A colleague of mine sent me some parts to do a non-linear FEA study on. They are 6061-T6 components, where one has its rim crimped over a mating component. The object of the study is to find out how much force is required to crimp the rim. I am fairly new to SolidWorks simulation and I’m not really sure where to even begin. Could anyone walk me through the setup process, or point me to a tutorial that is similar to what I am trying to achieve?

      Thanks in advance for the help!

        • Re: Nonlinear FEA of a Crimping Process
          Bill McEachern

          It looks axisymmetric so use the 2D capbility. set up the probelm like you normally would with contacts etc. You need a elastoplastic material model - see if a bilinear model works (you just need to specify a tangent modulus - it is the slope of the second line of a  two line approximation to the stress strain curve) or get a stress strain curve for the material and use that. I would suggest a von mises plasticity model (pick it from the material in drop down).

           

          Maybe before you start you need to sort out what the simulation needs to achieve for you to say oh that's the force required? You need to come up with something like how much of the swagged bit is in contact with the part it is supposed to hold on when the force is released. Maybe just put a really large force on the end and see when it drives the swagged bit to full contact. That mightbe a good one and is a simpler simulation to run. Check out the 2D non-linear tutorial - it should get you most of the way. Good luck.

            • Re: Nonlinear FEA of a Crimping Process
              Steven Kiley

              Bill,

               

              I appreciate the response! I've been playing around with the 2D dynamic nonlinear analysis, but I can't seem to get it working the way I want. Please see the screen shot of what is happening. You'll notice I added a large steel ring to simulate the crimping operation, which I am trying to run into my model and cause deformation. Any advice?

                • Re: Nonlinear FEA of a Crimping Process
                  Bill McEachern

                  If it were me I would:

                  1. get rid of all the stuff that is not close to the action.
                  2. start with an NL static
                  3. Use a displacement condition and not a load to move the added ring (then use the reaction force to get the load, in the displacement condition make the displamcent go a bit further than you want - say time=1, disp=nominal gap, at t=1.1 add a squeeze factor - a small amount of say a few thou. If you get it to solve to t=1 you can restart and add some squeeze by changing the solve time to t=1.1)
                  4. try to ensure as uniform a mesh (constant element size, aspect ratios close to 1 everywhere). Over the bit to be crimped and added ring contact patch try to maintain a constant element size - it can be very helpful where large sliding occurs.
                  5. Put them in initial contact or as close as you can get it without interfereing - it seems like a rather large gap by the picture.
                  6. use large displacment formulation
                  7. Try it with elastic only materail properties - see if you can get that to work.
                  8. Do not use friction (maybe add it if you get it to work without it)
                  9. Use tria 3's to debug then tria 6's once you get it working with the tria 3's.
                  10. add plasticity - large strain option needs to be turned on if plasticity is enabled. us the stress strain curve if you have it. Try stiffer tangent modulus if solution fails due to large incremental strain
                  11. increase element density in plastic bit if solution terminates due to large incremental plastic strain..
                  12. If all that doesn't work maybe go to NL Dynamic with the same approach.
                  13. Metal forming is not SWX Sims best game - it maybe quite a challenge to get this to solve.

                  Good luck

                    • Re: Nonlinear FEA of a Crimping Process
                      Steven Kiley

                      Thanks for following up Bill. I will give these steps a try later tonight. Hopefully I will have some positive results to share.

                        • Re: Nonlinear FEA of a Crimping Process
                          Jerry Steiger

                          Steven,

                           

                          I agree with all of Bill's points and would emphasize his last point, that metal forming is not something SW Sim is particularly good at. To get a really good idea of what is happening, you will probably need to run software that will morph your mesh as you get large plastic deformations.

                           

                          Jerry S.

                          • Re: Nonlinear FEA of a Crimping Process
                            Jared Conway

                            I agree that this is going to be tough. Are you on subs and have a reseller you can work with on this? Once you get a reasonable setup and it isn't working if confirm with them that it is a software limitation. If you aren't sure about how to set it up, we could help you here at hawk ridge. Depending in your need, if it is a confirmed limitation, we also have abaqus if it makes sense for your design needs.

                             

                            In general, I'd say this will be tough but with the right setup, you should be able to get something. I'd recommend starting as simple as possible. If you can get the part to bend without the contact in 2d, don't go any further.

                    • Re: Nonlinear FEA of a Crimping Process
                      Steven Kiley

                      Thank you guys for your input. I ran it a couple times last night and I may have a result, but my number seems really high. After work tonight, I will post some screen shots of my results and we can discuss them.