3 Replies Latest reply on May 30, 2008 12:06 AM by Jerry Steiger

    failure to knit -sometimes

    John Kreutzberger
      I was designing a mold with a complex parting line last week. I make my parting surfaces manually because I seldom like the automatic parting surfaces that SW generates. That means that when I create my shut off surfaces, I allow SW to knit the results. Then I create the parting surfaces.

      I have been in the habit of bringing the molded part into a new assembly after completing the parting surface. I have gotten used to not knitting it manually to both the core and cavity surfaces in the context of the part. Rather, I copy the parting surface and the core (or cavity) surface into my new part file in the assembly and knit them there. On the last job, I did this and found that they would not knit. Then I went back to the part to see where the problem was and fix it. Just for the heck of it I tried to knit the surfaces in the part file and they knit with no difficulty. Therefore, I went back and made a copy of the parting surface and did my knitting for both sides in the context of the molded part. Then-back to my assembly file and copied the knitted surfaces. did my surface cut and completed the design.

      My question is: why would they knit in the part, but not in blank part being edited in the assembly?
        • failure to knit -sometimes
          Charles Culp
          The first thing I would check would be to do a rebuild (CTRL-Q) of both items with verification on rebuild turned on. See if that changes your results.
            • failure to knit -sometimes
              John Kreutzberger
              I am always working with verification on rebuild checked. I only un-check it as a very last resort. I know there are different opinions on this issue, but I have always worked this way. Any performance hit is worth it IMO. (my position on this issue is due to getting SW models from people who work the other way and not being able to scale them for shrinkage-or do anything else to them for that matter- due to all of the general faults in their model. I prefer to only create geometry that does not require me to turn off this option to get things to work.)

              In the case I cite, cntrl-q did not offer any clues. I did it after knitting in the part and all was well. Not sure what it would have done for me in the part I was editing in assembly as it was already bombing out. I did try it after copying the surfaces and before knitting with no positive results.

              by the way-SW 2008, sp3.1