23 Replies Latest reply on Jan 24, 2014 5:28 PM by Jared Conway

    sensor issue

    Andres Gomez

      Hi guys. I am having some "trouble" with the given values of some sensors.

      The model is this:

      Sin título.jpg

       

       

       

      I ran the analisys for r/d=0,005 and the program determined que maximum normal stress at some location (x,y,z) . Then I put a sensor right at that location and I ran a design study for determining the stresses with the variation of r/d

       

      r: fillet radius

      d: shaf diameter

       

      The isssue is that the value of the sensor doesn´t match the maximun value given by the program for r/d=0,005.

       

      Can the sensor somehow modify the mesh?

        • Re: sensor issue
          Jared Conway

          please post some screenshots of your results and some numbers. how much exactly was it off for example?

           

          if you just care about the max, why use a sensor at that specific location? why not use model max?

           

          my suggestion here would be similar to what we discussed in another thread on this problem, evaluate more than just single point locations. i think you might be missing the general behavior of the problem.

            • Re: sensor issue
              Andres Gomez

              1.jpg2.jpg3.jpg

              the stress 1 is given by a surface sensor over the fillet radius

              the stress 2 is given by a puntual sensor at 15° of the particion line, this location is where the photolelasticity predicts the maximum normal stress

              the stress 3 is given by a puntual sensor (point 7 in the fig), this location was determined by simulation.

               

              Comparing, the location of the maximun values of FEA and photoelasticity are very close each other, but it seems that based on the stress1 values, the location of the maximum is not constant since stress 1 doesn´t match with stress 3

               

              Am I right?

                • Re: sensor issue
                  Jared Conway

                  without seeing hte stress on the part, it is hard to tell if this is a sensor problem or not

                   

                  if you probe/list select those locations do you get what you want?

                   

                  i'm also not sure i follow what you're trying to match. in your original post you mention your results for r/d of X doesn't match the sensor but then you gave us the results from a design scenario. which one do we look at? or are all of them showing the behavior?

                    • Re: sensor issue
                      Andres Gomez

                      would I post a picture of the stress?

                       

                      yes, all of them show the behavior....I ran an analysis for r = 0,5mm. Based on the maximum result of a surface sensor over the fillet, I put a sensor at that location.

                       

                      I ran a design study for diferent values of r/d, and for r=0,5mm I realized that stress1(surface max sensor) doesn´t match stress 3(puntal sensor based on the max value of the surface sensor for r = 0,5mm)

                       

                      for example, in the design study for r= 0,5mm, stress1= 73,416 and stress 3=72,721

                        • Re: sensor issue
                          Jared Conway

                          i'm having trouble synthesizing what the problem defintion is

                           

                          i think you're saying that you run an analysis

                          and then you made some sensors based on that analysis

                          then you ran the analysis with the design study and the results were different

                           

                          if that is the case, what i think would help better understand the problem is to look at the stresses and try to reverse the problem. ie, go back to your original analysis and use the sensors, do they report the same way as the design study or not? or go to the design study analysis and look at the full results, do they match your previous analysis.

                  • Re: sensor issue
                    John Pruyn

                    I have a design study that I have run on a simple formed flat spring.  I've allowed a couple of the dimensions and the material thickness to vary, set an upper limit on the max vonmises stress and and a lower limit on deflection.  The study attempts to minimize the mass of the part.  It cranked for a while and found several candidate designs not violating any bounds and selects the optimum based on mass.

                     

                    Slide1.JPG

                     

                    The problem came in when I want to look closer at this optimal solution.  For that I go back to the parent study, update the geometry with the values the design study found and run the simulation.  The trouble is that where the design study found displacement of 5.1 mm and stress of 703 MPa the parent study finds displacement of 22.1 mm and stress of 3028 MPa.

                     

                    Slide2.JPG

                     

                    Slide3.JPG

                     

                    So am I missing something here?

                     

                    Thanks

                    John Pruyn