16 Replies Latest reply on Oct 17, 2013 5:51 PM by Unspecified Unspecified

    Fin modelling problem

    Tom Arthur

      Hi,

      I'm doing a job which requires me to model a surfboard fin... based on a pre-defined profile, which I am trying to match exactly.

       

      The problem I have is that the surface modelling method i'm using, requires the edges to be rounded as the last step... which means the fin shrinks considerably away from my outline. Please see attached image (the sketch line is my target outline... used to create the shapewith lofted profiles). I've also attached the part.

       

      Can anyone suggest a better method so that I end up with the exact side profile, whilst maintaining good control over the cross section. I don't necessarily need to use surface modelling over solid, but it seems to me that I would have the same problem.

       

      I'm trying to end up with a clean model which can be exported as an stl and rapid prototyped.

       

      Thanks in advance.

        • Re: Fin modelling problem
          Mark Kaiser

          As I'm sure you know, if you have two surfaces meeting along an edge, using a fillet on that edge will always remove material.  I'm guessing you'd have to add a surface between your two fin sides, that would get removed during the fillet, or make your fin sides oversized.  I don't really like those methods, hoping someone else has a better suggestion for you.

           

          That's one small fin!

           

          Untitled.png

          • Re: Fin modelling problem
            Robert Stupplebeen

            Try making an extruded surface of the fin profile and then use boundary surface or loft adding a tangent constraint to the newly created extruded surface. This will make it so that a fillet is not needed.

             

            I hope this helps.

            Rob Stupplebeen

            • Re: Fin modelling problem
              Tom Arthur

              Thaks for the response.

              Unfortunately i need a variable fillet.. which isn't possible with these methods if i understand correctly. More of a round on the front edge (1-2mm), reducing along the length to a very small (0.5mm) rad on the back edge.

              • Re: Fin modelling problem
                John Burrill

                Tom this the best I could do.

                The problem is that you've laid out your profiles under one set of assumptions, filleted the part under another and are trying to satisfy both paradigms simulteneously.

                Well, nothing says, 'make it fit' like the deform feature so that's what I did.  There was some buckling of the surface near the point of the skeg, so I had to delete those faces and replace them with surface fills.  I wasn't able to get good continuity ouf of tangent and curvature fills, so I used  a straight contact.  As a result, the spine isn't tangent to the side faces.  To make matters worse, I didn't realize you were on an older version, so the file I uploaded is in 2012.

                REWORKING_FIN.PNG

                The question you have to be able to answer when you're preparing the model to apply your fillet is what how should the other surfaces grow in order to change it.

                What I would recommend (which is not what I did) is creating the spine as a boundary patch. 

                To set that up, first, chop the thing in half.  There's no point in doing something twice when you can mirror half of it and get the same result.  You'll need to create planes perpendicular to your profile sketch at the locations where you change radii in your variable fillet; sketch the cross sections at those  points using arcs  with curvature-constrained tangent splines between the them and intersection curves from your lofted surface (this might be easier to do in a 3D sketch btw).  On the lofted surface, use a spline-on-surface and draw a curve passing through the endpoints of the curvature splines.  split the lofted surface side face with that spline (using an intersection method).  extrude your profile sketch in the opposite direction of the side you're working  on.  This ribbon will give you tangency control on the finished product  Create a boundary with the edge of the ribbon thats normal to the front plane and the split curve on the lofted sketch as the direction 1 curves and the cross sections with your curve transitioning to the lofted surface as your direction 2 curves.  Set the tangency control in direction 1 to curve continuous.  and set everything in direction 2 to be normal to the profile.

                Build that and you'll have a strip surface that does what your fillet was supposed to, but it matches your spine and merges with the lofted surface.

                Use replace face or the new intersect tool in 2013 to trim and knit your new surface into yor lofted surface.

                If that sounds too convoluted, then you need to rethink your loft sections.

                • Re: Fin modelling problem
                  Robert Stupplebeen

                  Attached is a model with a boundary surface. I also modified the splines to hopefully reflect your variable fillet's intent. I hope this helps.

                  Rob Stupplebeen

                  Fin-RStupplebeen.jpg

                  • Re: Fin modelling problem
                    Tom Arthur

                    Thanks guys.

                    Robert, your boundary surface looks good, but the surface at the tip of the fin is distorted... i've had a play with the various tangency options but can't see to smooth it out.

                     

                    John, thanks for the explanation... like you say, unfortunately i'm still on 2011 so can't open the part, but will try to work through the steps myself tonight. Will let you know how i go.

                     

                    Tom

                    • Re: Fin modelling problem

                      Hi Tom (and others here)

                       

                      I don't have time to try it but I'l lbet that a sweep with guide curves woudl make a pretty nice solution. I think the key would be to creat a simple ellipse or 4 point spline as the profile and also make a dedicated path that follows the curve of the fin, located around the center portion of the fin. Also using constant option migh help.

                       

                      Regards

                       

                      Mark

                       

                      Edit - I found a model similar and using the method I discribed and have attached it here.

                        • Re: Fin modelling problem
                          John Burrill

                          Mark, this reminids me of the hard-hat lesson in the old SW2000 esssentials class.  It seems counter-intuitive that SW would know how to handle a profile being shrunk to a point in a sweep.  Out of curiosity, what was the role of the projected curve in shaping the surface and how di you size it?  (btw, putting the horizontal relation righ ton the spline handle was a clever way to control tangency.  I'll have to remember that)

                          • Re: Fin modelling problem
                            Unspecified Unspecified

                            Hi Mark,

                             

                            New to Solidworks, if im being honest im struggling with it.

                            Im working on a project and your "fin using sweep" model is a great start.

                            In modelling thruster fins on a surfboard they tend to be asymmetric.

                            I tried cutting your model with the right plane but it would not import into FLUENT.

                            Are you able to post a document with that same model just cut in half?

                             

                            Any help is appreciated

                             

                            Brett

                          • Re: Fin modelling problem
                            Michael Maz

                            Did you ever get this sorted Tom?