5 Replies Latest reply on Sep 21, 2007 8:45 AM by genexxer genexxer

    Assembly Testing

    Matthw Wendlr
      Ok guys I have been gaining some valuable knowledge in the otherparts of this forum, now it is time to actually run some testing. As you can see below, I have a pro-modified drag race carrear end.  My company makes these and we would like to putsome numbers to it.  Basically what I need to do is run Cosmosas if a 4-link were hooked to it.  So I need to roughly breakdown 14,000 lb/ft after multiplication among the 4 differentbars, 2 push and 2 pull on the front and back of the 4-link holes. 

      It says that to run Cosmos on an assembly you have to run itthrough physical simulation, I am not really sure how I would dothat or if I really need to considering it isn't moving itself. I am looking for any insight you can give me to run anassembly in Cosmos selecting just the 4 holes in the 4-linkbrackets.
      Thank You,
      -Matt 
        • Assembly Testing

          I would set fixed constraints on holes youplan on attaching the bars to. This will essentially hold therear-end in place. Then you can apply the appropriate torque loads,likely somewhere on the axel shafts.

          What resultare you looking for? Stress? Delflection?
            • Assembly Testing
              Matthw Wendlr
              Mainly I am looking for deflection, we want to see how much andwhere it actually moves when the car is launched.  However Iwas hoping to do this with just the housing, I am trying to puttogether a somewhat replica center section to bolt to the fact ofthe housing but was not really planning on making shafts and whatnot.  

              The plan was to just split the torque multiplication among the 4holes 2 push and 2 pull and run it that way.  Is this notpossible or not the right option?
                • Assembly Testing

                  I think that may work. You will need toconvert the torques into the appropriate force, as seen by the4-link rods. For a restraint, I think you would want to hold it onbearing surface for the pinion gear. Which would be on the coveryou mentioned (assuming this is a for style rear).

                  This is actually apretty complicated scenario, all the force is created by the piniongear trying to walk up the ring gear. Which will create significantforce on its bearings and housing/cover.  I would want toinvestigate the cover as well as the housing.

                  You coulddefinitely get what you want out of CosmosWorks, but you shouldreally take some time (if you haven't already) to consider all theforces and reactions the instant the car launches (assumeeverything is static) and make a free body diagram. Once you havethis information you could them accurately define you loads andrestraints to get the answer you are looking for.

                  I hope this helpsand I didn't dig in to deep.

                  • Assembly Testing
                    Peter Gillespie
                    Matthew,

                    If you know the reaction forces at the various mounting holes and other locations, you can perform an analysis that way. If you don't have measured data, you're going to have to make assumptions, which can greatly affect the results, so be sure you consider all forces on the assembly.

                    If this is your intro to COSMOS, you've picked a fun one! From my experience, there are two things to be successful in COSMOS (FEA in general): knowing the program interface and the engineering mind to make correct assumptions. This particular model is going to be a challenge enough to get all of your assumptions right. I would recommend practicing on components that have fewer variables first so that you can get the COSMOS interface down. Otherwise, you're going to be wondering if the problem is in your assumptions or how you used COSMOS to apply those assumptions.

                    I've been in COSMOS over a year now, and I'm just getting comfortable with my understanding of what to expect from the program and how to translate assumptions to the COSMOS model.

                    Anyway, I've found a lot of help on these boards, and through COSMOS Companion. Hope you do as well!
                • Assembly Testing
                  genexxer genexxer
                  Hi folks...

                  I just admire the SYMMETRY. I risk being a fool...hey...reminds me of a song...

                  "Fools rush in...."

                  Cut the model (ie. node count) in half and set up a symmetry boundary condition on the cut face.

                  "where angels fear to tread...."

                  Then "run all studies" on twelve studies where each study has a unique push-pull pair of 2xholes selected as fixed restraint.

                  "but here I go again..."

                  The applied moment is one half the total torque in a half model.

                  "my heart above my head..."

                  "fools rush in..."

                  And uh...if the deflection is what is interesting to you...

                  "where wise men never go..."

                  nonlinear static analysis is recommended, for results that account for deformation in the stiffness matrix.

                  "but a wise man never fell in love..."

                  To get some quick lies use the linear solver since it also helps guide the mesh density setting.

                  "so how are they to know"