8 Replies Latest reply on Mar 25, 2015 3:17 PM by Jared Conway

    "Invalid contacts" In flow testing

    Christopher Harmon

      Hello.

       

      I am currently working on designing a boat hull for a high school engineering course.  More information on my overall project and process can be found here: Life Saving Technologies

       

      To put it simply, my overall process for modeling the hull has been to loft along guide curves.  The "journals" section of my website describes exactly how I did this.*  As natural with boat hulls, the individual lofted sections bend and twist as they should.  I am unable to use the "merge results" choice in the loft feature as it often results in a "Zero thickness" error.  So I have created over ninety lofts and none of them have merged. (I did use the "thin feature" and it also got rid of the problem, however I don't like the idea of using it for the whole model, and it is not allowing me to go back and deselect it in the feature.)  The model has small "cracks" in it where the lofts don't precisely match up.  Is there some type of surfacing tool I could use to fill this in?

       

      Getting to the root of the problem, I intend to flow test this model and see how it would move through the water as a part of my project.  When I run the flow simulation I get the error-

       

      "There is no fluid volume in the project. Please check geometry or boundary conditions."

       

      or

       

      "Fluid volume recognition has failed.  Please check the geometry."

       

      Furthermore, when I use the "check geometry" tool I receive the following errors:

       

      Status: FAILED. Problems detected

      Invalid contacts unresolved: 83

      Invalid contacts resolved: 12

      Invalid pairs: 1

      Analysis type: External

      Fluid volume: 0 m^3

       

      If posting screen shots would help to solve this problem, please let me know.

       

      Thank you.

       

      *Since the completion of my latest journal, I have finished all my loft sections.  The screenshots posted there are not of the completed model.

        • Re: "Invalid contacts" In flow testing
          David Hensel

          It might be helpful to post your geometry, or screenshots of the case you're working on in Flow Simulation.

           

          Is it a surface or solid geometry?  It sounds like you have some lofting issues, it may be best to resolve your geometry to something solid and robust, without any "cracks".

           

          From the look at your posts, you might consider using splines and use points from the boat plans for control.  In my experience, lofts work best along a continuous line, and it seems like you've got multiple segments between frames/profiles along the length of the boat.

          • Re: "Invalid contacts" In flow testing
            Jared Conway

            SolidWorks Quick Tip - 4 Things to Do Before Every Flow Simulation Analysis

             

            if your model is not "water tight" you're going to have an issue

            same things goes for if it doesn't survive a CTRL-Q with VOR enabled

             

            have you narrowed down what you want to test in flow simulation exactly? with the limits on free surface, it might be best to get a solid mode of the hull that is "close enough" but solid and run that vs try to get your model for visualization fixed up to work with flow.

              • Re: "Invalid contacts" In flow testing
                Christopher Harmon

                In the flow simulation, I want to test how the model would flow through water.  I did a CTRL-Q with Verification on Rebuild enabled.  Nothing happened.  The model is solid.

                  • Re: "Invalid contacts" In flow testing
                    David Hensel

                    As Jared said, there is no free surface functionality in Flow Simulation, meaning you cannot model the top of the water/waves on the boat.  For an approximate solution, you could model the boat below the waterline and neglect the region by the waterline.  You could get drag numbers from this, which would be more valid further away from the surface I'd guess. 

                     

                    Depending on the hull's dependence on waves, you might get something useful, or it might be a stretch.  If you had some known values from the boat plans, you could compare, but for further analysis (if you really want to get into it) I could recommend OpenFOAM for CFD modeling free surface/drag.  There are several examples, but the analysis could be a senior project/thesis in itself.  For example: OpenFOAM wigleyHull-Tutorial CFDI - YouTube  (not my work but seems comprehensive)

                      • Re: "Invalid contacts" In flow testing
                        Christopher Harmon

                        Thank you for the suggestion, I have looked into OpenFOAM before.  However, as you mentioned, it is a very challenging program to run and would use up to much of my allotted time.  The only data that I need is the drag data. So returning to the question of making this model work with a flow simulation, what would you suggest to fix all of these Zero-thickness errors?

                          • Re: "Invalid contacts" In flow testing
                            David Hensel

                            I would make another attempt at modeling the boat, starting from scratch with a different technique.  Or head over to the modeling forum and get some expert advice on surfacing.  Sometimes these tools, (surfacing in particular) can be picky about inputs, so it's good to learn alternatives or workarounds to get good geometry with no zero-thickness errors.

                             

                            I'd suggest modeling everything as a surface and then thickening it.  I'm guessing the lofts along the same lines are not lining up properly, or something else is going wrong.