22 Replies Latest reply on Jan 28, 2010 3:32 PM by Bill McEachern

    Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force

    Ryan Laplante

      I get the exact same stress distribution and deformed shape of my model if I spec it as 3k or 6k aluminum or alloy steel, and range from 2-200 lbs of force.

       

       

      I tried closing down model and re-opening, re-starting cosmos express, as well as using the start -over command available through cosmos expresss.

       

      I also went back through and did the cosmos express tutorial and guess what??  Same results for the load bearign hook at 1 Lb as at 1500lbs.  This is following the tutorial  and already having it done a few times.

       

      Q1.

      Is this a bug in SW 2008 SP3.1?

       

      Q2.  We need to change material properties and thickness and run cosmosexpress 10-20 times with different parameters of input to judge deflection in metal tables.

       

      Q3.  I get the correct FOS - example for the hook at .2 Lb the FOS 92714 is at 2,000 lbs is  FOS is 5.6228   the stress distribution and the deformed shape of the model do not change.

       

      I am trying to get this information to china on about 40 tables by the end of the week?

       

      I have 2010 sitting here but will probably lose a weeks worth of work changing over, and would rather not till SP2 lol.

        • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
          Bill McEachern
          Stress is independent of material for small displacements & linear behavior. Deflections on the other hand are highly dependent on material stiffness.
            • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
              Dale Dunn

              I'm an ME that hasn't formally studied FEA, and this has piqued my curiosity.

               

              Aluminum has about 1/3 the modulus of elasticity of steel. I would expect this to give different stress/strain gradients on a loaded part. Are you saying the small displacement assumption of linear statics effectively removes this from play? If I do a simple test with SimulationXpress, I do get slightly different results for maximum stress (33418 for steel, 33043 for aluminum, using identical geometry and element size). If this was two parts with slightly different geometry, I'd call that insignificant. With everything but material the same, I'd expect the solver to converge to identical numbers for maximum stress if you are correct that stress is identical. So what is going on?

               

              Now then, I would expect a linear static analysis to result in a deformed shape that looks identical for both cases. The differences would be too subtle to notice, apart from the numbers.

            • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
              Josh Brady
              I believe the deformed shape is always a gross exaggeration of the actual deformation intended to show you how the thing will deform - not by how much.  For most parts, if there was enough deformation for you to see it on the actual part, the part would be destroyed.  However, in reality there is always some deformation, no matter how small the load.
              • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
                Derek Bishop
                Are you an engineer? It doesn't sound like you understand the basic concepts relating to undertaking stress analysis. If the analysis is critical you should have it undertaken by a suitably qualified engineer.
                • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
                  Anthony Botting
                  Stress is a made-up concept (force per unit area). Plastic or steel result in same stress. The strength is material-dependent and you'd have to obtain test data to assign strength values to each material. The shape should not change with material (assuming isotropy), and all statements assuming linear-elastic, small displacement theory is valid.
                  • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
                    Corey Wallace

                    Ryan,

                     

                    Use the Von Misses data and URES data given after the analysis to make the case to your customer.  You should see a difference here.  As others have mentioned the visual displacement that Solidworks generates is probably mostly for show.

                      • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
                        Ryan Laplante

                        What does it matter if I am an engineer or not , I am trying to apply the fundamentals of the program.  I have been working with SW since 2004.

                         

                        The COSMOS tutorial could be done by a 6th grader.


                        I expected the visual data to represent the mathematical equations used to derive the FOS.  Is this too much to ask?

                         

                        Correlation between results and visual representations of the data should be accurate but this forum is telling me -

                         

                        "the visual displacement that Solidworks generates is probably mostly for show."

                         

                        What I don't get is my table deforming like it has 2 tonnes on it whether I put 2 pounds or 2oolbs on it.

                         

                        I thought that that is what this is for - analysis - what is it showing?

                         

                        From a design standpoint its critical to find out the deflection is with a given load, visually.

                          • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
                            Bill McEachern

                            Settle down boys, these forums are not a place for where this is heading.

                            The deflection plot in CosmosExpress is set to automatic which means it shows the displacements scaled so that the visual presence on the screen shows the character of the displacements so one can make the call on whether that character lines up with the intent/expectations of the analysis. I never use the express product, but if you want to visually show the differences in the paid for product you would adjust the scale to show the effect you desire. If the scale was set to 1:1 - also known as true scale - you would not see any noticeable deflections hence the need to have them amplified. the actual deflections should be available in on the legend and if 2 lbf gets you 0.002 inches then 2000 lbf will get you 1000 x .002". so if you are not getting this relationship then something is wrong with the analysis full stop.

                              • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
                                Ryan Laplante

                                I just wanted confirmation on the analysis being visually incorrect.

                                 

                                COSMOS is supposed to be used for thsese things right?

                                 

                                 

                                9. View and  evaluate the results.  This is where the job of a design engineer comes in, because it's not just about viewing but also evaluating the  results.
                                - Animate the  results to see if it moves as one would expect it to move in real  life.
                                - Is von Mises  equivalent stress appropriate or do you also need to create other stress plots, like principal or shear stresses?  Cartesian or cylindrical coordinates? Is  strain more appropriate?
                                - Novice users can  use the Analysis Advisor for help in interpreting results.

                                See the second bullet point?

                                Now follow the COSMOS tutorial and vary the amount of weight on the hook, the visual deformation is the same, no matter the weight, on my computer.

                                All I was looking for was confirmation that other users saw the same effect in their animations - the disconnect between the numbers input and the visual representation of those numbers.  This has ben confirmed be several users on this thread and so THIS THREAD IS OVER>

                                I will update you on the status of COSMOS express in 2010 after i upgrade this weekend.
                              • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force

                                Ryan,

                                 

                                As near as I can tell, your question still has not been answered.  If I understand you correctly, you're wondering why the results look visually the same regardless of the magnitude of the load.  Is this correct?

                                 

                                The answer is, the stress distribution will be more or less the same, regardless of the load, providing it doesn't exceed the yield strength of the material.  SW scales the graphical representations of deformation so you can see them.  In most circumstances, your part is going to deform thousandths of an inch, so it would be very hard to see, so they scale it up to a visible deformation and SW notes the scale on the deformation page.  If you are concerned about whether the actual stress is changing with the load, look at the table of Von Mises stresses noted on the stress plot.  It should scale according to the load.

                                 

                                I hope this is helpful.

                            • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
                              Brian Zias

                              Another thing to remember is the linear nature of response in Xpress.  If you apply a 2 lb force, and get a maximum displacement and stress, then that maximum displacement and stress will always increase linearly as you increase the magnitude of the force.  e.g. 200 lbf will produce exactly 100 times what you see for the 2 lbf.  It's up to you to determine when those displacements and stresses violate small deformation and linear elastic material assumptions.  But you shouldn't have to run 10-20 times each model if all you change is magnitude of force.

                               

                              Regards,

                              Brian

                              • Re: Cosmos express giving same results - steel vs aluminum - 2-200 lbs of force
                                Bill McEachern

                                Ryan,

                                the limitations are imposed by the marketing department not the "software" per se. It is the same software as the stuff you pay bigger bucks for - it is just packaged up to get people interested in it. Then they hope you actually pay for the more advanced products where the limitation you are concerned with don't exist.