6 Replies Latest reply on May 12, 2017 12:32 PM by Ryan Dark

    4 equal lifting lugs, one has higher stress than the others

    Brandon Cardoza

      Running a simulation on some lifting lugs for a project but cannot figure out why only 1 of the 4 lifting lugs has a higher stress than the rest in the proposed lug design. All components are equally spaced and the other 3 lugs do not have this single corner of high stress. On CustomerLift configuration I have one lift lug showing a lot more stress on the shell than the other 3. Once again evenly spaced so not sure what is going on.  Can anyone provide any sort of guidance as to why this is happening?

        • Re: 4 equal lifting lugs, one has higher stress than the others
          Kevin Pymm

          Brandon,

           

          I don't use simulation very much but when I opened your model I got a lot of mate errors, stating parts couldn't be moved to satisfy mates. Perhaps that is why you are getting a different result for one lug? Try a forced rebuild, Ctrl Q, to see if you get the same errors.

           

           

          Or try turning on Verfication on rebuild.....

           

            • Re: 4 equal lifting lugs, one has higher stress than the others
              Brandon Cardoza

              Kevin,

               

              The reason the mate error pops up is because the geometry is off by .004" (which if I could tell it to ignore I would). If you were to recreate the mate there would be no errors until you swap between configurations. The mate is placed on the main lug and an array was done off of that lug to position the other 4. The highest stress mark shows up on lug 3 of the array which I find the strangest out of everything.

            • Re: 4 equal lifting lugs, one has higher stress than the others
              Brandon Cardoza

              See attached photo showing the higher stress located on one lifting lug. LiftLugTest.jpg

                • Re: 4 equal lifting lugs, one has higher stress than the others
                  Ryan Dark

                  Hi Brandon,

                  Looking at the stress plot the results between each lug appear to be fairly close.  They are no orders of magnitude different from each other.  The sort of noise you see in your plot indicates the mesh as being too coarse for the model though.  I opened your model and it would appear that you have a single element across the thickness of your thin wall.  In addition to that the mesh quality plot shows that you have mesh elements with aspect ratios between 7-16 which will also impact the quality of the stress results (see below).  It is recommended that you have two high quality elements across the thickness of thin walls and that your aspect ratio be less than 10 (really less than 5 would be best).

                   

                   

                  Reducing your element size and applying mesh controls will help with all of these.  Ultimately, since the legs are not filleted to transition into the head model you will find that even with these improvements there will be stress singularities at the sharp corners where the lugs attach to the head.  The way to mitigate this is to model the lugs such that they contour perfect along the head and apply fillets between the lugs and the head (fillets the size of your welds would be good for this).

                    • Re: 4 equal lifting lugs, one has higher stress than the others
                      Brandon Cardoza

                      Ryan,

                       

                      Got any links or videos to show what you are describing to tweaking the mesh. It has been awhile since I used simulation and it was always for simple school projects. At my latest job I am not starting to get more and more into it so would like to be able to step my game up to compensate.

                       

                      Also in terms of the fillet weld to the head I am not sure I completely understand how you would go about getting the transition fillet. If you have any examples I could take a look at or once again some sort of video link that would be great. I know there is a way to add welds in when you treat items as weldments or sheet metal, could this be applied here (this is one area I am planning on trying to pickup more of but figured would ask as a quick question)?

                       

                      Thank you very much.

                        • Re: 4 equal lifting lugs, one has higher stress than the others
                          Ryan Dark

                          Brandon,

                          I don't see a particular video that goes over all the things you need.  This video shows techniques on using applying mesh controls starting at 9:13 (GoEngineer - Simulation: Troubleshooting Common Errors).

                           

                          For the fillet, since you have constructed this as an assembly, you will need to reconstruct the model as a part to facilitate the creation of fillets between the lugs and head.  I have done this below along with applying a split line at the base of the lug to allow me to localize a mesh control around the base of the lug.

                          The localized mesh control is for the sake of speed in this case as applying 0.25" elements to the entire model would make each solve take a much larger amount of time.

                          The result is a less noisy stress field at the high stress regions.