15 Replies Latest reply on Mar 24, 2015 12:10 PM by Dave Branco

    Need to make an equation driven surface...

      All:

       

      I need to define a part with that has a solid surface which is a 4th order polynomial of form z = f(r, theta); cylindrical coordinates.

       

      can soldworks do this? I can see how to do equation driven curves, but not surfaces. Do i need to convert to cartesian z = f(x,y) first?

       

      help!

       

      Thanks in advance all

       

      Jason

        • Re: Need to make an equation driven surface...
          Jerry Steiger

          Jason,

           

          I don't believe SolidWorks has equation driven surfaces. I think you would have to use equation driven curves and loft through them. You may or may not be able to get a shape that is good enough for what you want to do. If you are looking for optical quality results for a mirror or lens you are probably going to be disappointed. Your choices on how many profiles to use and where to place them are probably going to be crucial to getting the "best" surface.

           

          I doubt that converting to cartesian coordinates would help, but it could depending on the shape of your surface. If your shape lends itself to profiles on the theta planes, then cylindrical is most likely the way to go. If you do go with profiles on the theta planes you may want to another type of surface, probably a fill with constraint curves, in the area around r=0 to avoid a singularity in your loft.

           

          You should probably read Matt Lombard's surfacing book, SolidWorks Surfacing and Complex Shape Modeling, before you start.

           

          Jerry Steiger

          • Re: Need to make an equation driven surface...
            Mahir Abrahim

            This macro I wrote a while back might be helpful.  It automates equation based surfaces.  Circular surfaces made using cylindrical or spherical equations don't always work property, but you can still use the macro to create a set of curves for ceating a lofted or boundary surface.  I'm actually working on the macro to make it more robust, so stay tuned. Below is an example of a boundarysurface made with curves created with my macro.

             

            http://www.3dcontentcentral.com/macros/download-Macro.aspx?id=6290&catalogid=171

             

            Sample SurfPlot Surface.png

             

            Edit: Added picture

            • Re: Need to make an equation driven surface...
              Robert Stupplebeen

              Jason,

               

              I also share your desire.  For optics any spline based surface wiggles too much to be useful.  I have submitted an enhancement request and talked to every one at SW during SWW that I could.  With equation driven curves in 2009 and being able to drive them in design tables in 2010 maybe by 2011 we will have equation driven surfaces.  In the mean time this code has worked for me http://www.sharewareconnection.com/surfx3d.htm.  I hope this helps.

               

              Rob Stupplebeen

                • Re: Need to make an equation driven surface...
                  Harold Brunt

                  Rob,

                   

                  Thanks for posting the link. I just downloaded the software and I'm hoping to find the time to get familiar with it soon. I am curious what optical aplications you are using the surfaces for? I have had issues with the accuracy of the surfaces created using the equation driven curve sketch tool in SW09 as well. I would like to be able to edit the surfaces (which I can do in SW) but I need higher accuracy to be able to use the surfaces in optical calculations or simulations. An accurate conic (section) surface controlled using a RHO value would be great.

                   

                  Harold

                    • Re: Need to make an equation driven surface...
                      Robert Stupplebeen

                      Harold,

                       

                      I am in the opthalmic industry.  So we are trying to fix the optics of natural eyes.  What is your industry?  Most of our real optical work is done in optics codes.  For SW we are more looking at mechanical aspects of the design so a micron off is OK for our application in SW.  We model conics as revolved partial ellipses dimensioning the major and minor axes using a design table converting from R&K.  I hope this helps.

                       

                      Rob

                        • Re: Need to make an equation driven surface...
                          Harold Brunt

                          Rob,

                           

                          We are a Photonics Engineering consulting and small manufacturing company. The majority of our projects require high precision in the modeled surface for the simulations to be accurate. Any time I need to use a conic section I am limited to the Optis interface. SW is a dandy tool for most everything I need but the lack of true accurate conic surfaces is the one thing that is missing (but I seem to be a small minority).

                           

                          Also, thanks Mahir. Down loaded the macro. Very cool.

                           

                          Harold

                            • Re: Need to make an equation driven surface...
                              Robert Stupplebeen

                              Harold,

                               

                              What do you do with the model that you need the precision: CNC, ray tracing?  Don't get me wrong I agree that we need better optical surfaces but there might be a good workaround if there is such a thing.  Thanks.

                               

                              Rob

                                • Re: Need to make an equation driven surface...
                                  Harold Brunt

                                  Ray,

                                   

                                  CNC from files, raytracing withing the 3D environment, etc. We don't always have the same path to a final design. ZEMAX is the primary tool for sequential raytrace and lens design. Our OE supplies the data to me and I model the parts from scratch (instead of importing the iges). If/when the design prescription gets changed I can update the files like any other top-down SW assembly file. Unless the design includes aspherics then it gets a little messy.

                                   

                                  For non-sequential applications (we are doing a lot with LEDs lately) we often start in SW. One of the benefits of having an add-in lighting package is that any SW generated surface can be defined as an optical surface. That doesn't mean it will be an accurate surface. My work-around has been to generate all my conics in Optis or to use a design table to generate a spline with a ton of points. I can use the conics in the tool bar (no hyperbolics though) and the formula driven curves to get close to what I want when designing a non-imaging optical system but ultimately I have to hand it back to the OE to confirm in ZEMAX.

                                   

                                  What are you suggesting?

                                   

                                  Applogies Jason if this is off topic.

                                   

                                  Harold

                                    • Re: Need to make an equation driven surface...
                                      Robert Stupplebeen

                                      Harold,

                                       

                                      That is very similar to our work flows.  We have a CAD model that we CNC the non-optical parts from then merge that with the mathematical optical surfaces.  I have separate FE and optical models connected by SigFit.  Juggle all 3, hope I don't drop one then perform all the different analyses, optimize the design and spit out a final design.  It is not ideal however it works.

                                       

                                      Looks like for the time being we will both be in the same circle but hopefully not for eternity.

                                       

                                      Rob

                            • Re: Need to make an equation driven surface...

                              Mahir, Rob:

                               

                              thanks for the macros. I'll give them a try and see if this works.

                               

                              This application is for metal mirrors with arbitrary shape functions on the backside, so that they can be "bent" to vary the radius of curvature without introducing too much surface error from a pure sphere.

                               

                              thanks!

                               

                              Jason

                              • Re: Need to make an equation driven surface...

                                Mahir, Rob:

                                 

                                thanks for the macros. I'll give them a try and see if this works.

                                 

                                This application is for metal mirrors with arbitrary shape functions on the backside, so that they can be "bent" to vary the radius of curvature without introducing too much surface error from a pure sphere.

                                 

                                thanks!

                                 

                                Jason