7 Replies Latest reply on Mar 30, 2015 10:08 PM by Stephen Callegari

    local failure (triaxial stress) at every node

    Stephen Callegari

      Asme section viii div 2 and api 6a require a local failure (triaxial stress) at every node is checked. This means the sum of the principal stresses must be less than 4Sm (For linear fea).


      Is there a more efficient way to export these stresses or perhaps create a custom stress plot?


      Exporting tabular stresses through simulation on large models can take upwards of 8 hours for each principal stress. This is not very practical.



        • Re: local failure (triaxial stress) at every node
          Attilio Colangelo

          I have had similar requirements for Section VIII Div 2 FEA,  It's one of the things that drove me to learn the Simulation API which would be the most efficient (and maybe only) way to accomplish this.  It does require familiarity with Visual Basic (VBA) and object oriented programming methods


          I have some sample code ir you want to go this route.

          • Re: local failure (triaxial stress) at every node
            Alessandro Tornincasa

            New in SOLIDWORKS Simulation 2015 is the ability to plot triaxial stress.



            Hope this helps.



            • Re: local failure (triaxial stress) at every node
              Stephen Callegari


              Thanks. I am definitely interested. I could also use this for the strain based elasto plastic  local failure criteria checks. We're you able to create your own stress plots?



              Awesome! Thanks for the heads up. This is great news. I have not installed 2015 , but I will now. Much appreciated.


              Thanks for the responses.



                • Re: local failure (triaxial stress) at every node
                  Attilio Colangelo

                  It's great they added that tri-ax plot.  It seems every time I hit a wall with SW Sim they expand features so I don't have  to go to a different package .


                  The API gives you access to all the plot settings programmatically but does not create plot types outside of the built-in types.  What I have done is write results to an Excel sheet and charted results.    In particular if you're doing pressure vessel work the API gives access to all the linearized stress components according to ASME VIII Div 2.  Here's an excerpt of code:


                  Set CWObject = swApp.GetAddInObject("SldWorks.Simulation")

                  Set COSMOSWORKS = CWObject.COSMOSWORKS

                  Set ActDoc = COSMOSWORKS.ActiveDoc()

                  Set StudyMngr = ActDoc.StudyManager()

                  Set cwResults = Study.results

                  linearizedStress = cwResults.GetLinearizedStressValues(swsStressComponentVON, 10.423, 6.111, -22.409, 147, 6.4537, -22.409, 10, 0, errCode)


                  It's not very well  documented in SIm API help so it takes some trial and error to get it right.  Then Excel API methods would be used to write to a spreadsheet.

                    • Re: local failure (triaxial stress) at every node
                      Stephen Callegari

                      In the asme code for strain based local damage failure. The code instructs to compare the total equivalent plastic strain (from non linear analysis) to the limiting strain for each point in the structure. 


                      1. In solidworks we can list tabular results for total equivalent strain or plastic strain. I am not clear which strain should be compared to the calculated limiting strain for each point.


                      2. I can only list tabular results for element values of strain and not nodal values. Is there a means to list nodal strain results. Is it necessary  per the code?

                        • Re: local failure (triaxial stress) at every node
                          Attilio Colangelo

                          Hi Steve,


                          This area of the code is not clear.  I have rarely gone to the plastic approach but of course each case is different.  Stress linearization seems to be a good middle ground if you're dealing with local stress risers.


                          Hope that helps.

                            • Re: local failure (triaxial stress) at every node
                              Stephen Callegari

                              Atillo. Im with you. Linear fea takes less time and computing resourses to run but more post processing. Non linear fea is a more pass/fail criteria, but takes longer to solve and is more technical. I am learning and think I am ready to take the next step using the nonlinear approach. I have studied the asme section viii div 2 code,  the asme example  problem manual, and have worked with my reseller on questions regarding settings and stress strain curve inputs versus the code requirements.


                              I have an answer on the strain requirements. It appears the code requires the plastic strain (Not total elastic + plastic strain) to be compared to the limiting strain. By total strain, I belive the code implies plastic strain + forming strain. This confused me.


                              I have not found any formal clarification as to whether the strain should be nodal or elemental. However, if the mesh is sufficiently  converged (5% or less is our criteria) then the elemental and nodal values should be close in magnitude. Simulation non linear fea only outputs elemental strain values (based on my experience - unless I am missing something ).