6 Replies Latest reply on Jul 7, 2017 2:37 PM by Keith Frankie

    Edge Weld Connector

    Lesley Ott

      Are there any resources available that gives a detailed explanation for using the edge weld connector? Tutorials? Does a double sided weld actually signify a double weld, one on each side of the item being welded?

        • Re: Edge Weld Connector
          Casey Kluesener

          In order to effectively use the Edge Weld connector, you should review the paper by Weaver about the Throat Shear Method. It's easily googleable. There is also a solidworks Cosmos presentation, which predates the edge weld connector feature. This will give you a good background on how it works. It's good to know how it works, or you could easily get bad results. The edge weld connector basically automates the throat shear method. Unfortunately, you don't really have too much control over the output other than weld size. Other than the paper mentioned above, I haven't been able to find tutorials really. It's pretty straightforward once you figure out how to use it, but it should be used with caution, as if you don't set your model up correctly, you can get inncorrect results.

            • Re: Edge Weld Connector
              Casey Kluesener

              FYI, there is a bug in the Edge Weld Connector results for English units. It does not output the correct loading, but it is correct if you output it to metric units and convert back to english. Wish someone would have told me that...

              • Re: Edge Weld Connector
                Lesley Ott

                Thank you for the input. Here's what I have so far for SW2013.


                1. The global contact needs to be set to No Penetration.

                2. Each part to be welded (as a terminating part) has to be set up as a shell which includes all of the faces on that part. Failure to do this will give erroneous stress results for the individual components in the weldment. Shells can be set up using the Weld Connector. The mid-plane shell when used on complicated weldments gives erroneous results.

                3. Each weld has to be set up as single sided. Continuous welds are not possible. For example, a curved part with tangent legs will have to have an individual weld for the curve and each tangent.

                4. There are discontinuities at the weld start points. In my example, the transitions between the welds on a curve part with tangents are not the same at the end of the curve and the start of the tangent. Welds sizes may be different for each weld.

                5. To display the individual welds as green(pass) or fail(red), each weld will have to have its 'definition' individually opened and then closed without changes. I have not been able to determine a method to do them all at once. Showing the weld plot does not achieve this result.

                6. Edges can only be selected from the component selected as face 1. This face must be on the terminated (part to be welded) part. Although property manager indicates Face Sets only one face can be selected.

              • Re: Edge Weld Connector
                Aadi Patel


                for edge weld connector you can watch Weld check simulation on solidwork(weld pass or fail) - YouTube

                this video help not deep but.. you can further proceed.

                  • Re: Edge Weld Connector
                    Keith Frankie

                    Casey Kluesener which COSMOS paper are you referring to?  The COSMOS Companion?


                    Weaver's paper is available on his website.  It is a good read.


                    Casey do you have a SPR for that bug?  I have gotten some results in English units that are valid.


                    J. Mather recently recommended the "A Job Weld Done" video, and I found it to be very informative.


                    The attached part has 4 studies.  'Joint Normal' and 'Shear Weld Axis' verify expected SW behavior.  The input loading and weld sizing correctly results in a 1 inch weld throat.  'Shear and Normal' shows how both loads combine to produce the expected 1.4 inch weld requirement.  I didn't persevere to verify loading in bending.