16 Replies Latest reply on Mar 10, 2014 9:09 PM by Andrew Emmert

    let's make a boat.

    Andrew Emmert

      Hey gang,

       

      I'm helping to design a boat from scratch for a school competition, and I'm tweaking my model for the hull and cap that we're going to be molding in fiberglass.

       

      The feature in question that I'm asking about here is the very first feature (Boundary-Surface1).  I'd like to make the surface so that the edge that forms the keel has a nice rounded fillet all the way down the edge of the keel.  (Found in Boundary-Surface1: sketch 5)  Unfortunately, right at the point in sketch 5 when the keel fillets up to the bow and stern of the hull (radius of 3in) I can't force the loft to maintain that nice rounded profile, I end up with a much sharper intersection.

       

      I've tried to set up my boundary surface so that sketch 5 is normal to profile, which should give the desired shape.

       

      Hopefully somebody here on the forum can take a look.

        • Re: let's make a boat.
          Tom dunn

          I made another surface by sweeping and arc shown in yellow is this sort of what your looking for? Tom

           

          ***I just looked back at my model and I see I moved the shape of the top and now it is messed up. I will try again.***

            • Re: let's make a boat.
              Andrew Emmert

              Tom,

               

              I like the way you think.  So we've got the swept feature first for the hull, and then I'm thinking either a lofted surface or a boundary surface.  I can get the loft to work, that's not a problem, although it does seem like the surface loses it's smooth curvature as it closes in on the bow/stern of the hull.  Perhaps it's an optical illusion, but it looks to me like the hull is getting squeezed lightly by a vise. I'm still pretty sure that has something to do with my guide curves or tangency control settings.

               

              the next hurdle will be the mirror and thicken features which need to come next.  I'm going to try to focus on making a solid body and then using the shell command to get an internal wall thickness  of 0.125 in

               

              thanks for the feedback, and I've included another model to play with (if you feel so inclined.)

              • Re: let's make a boat.
                Andrew Emmert

                Tom,

                 

                I don't think I can attach images directly to a personal message, so you should see the update to the discussion thread and my response here. 

                 

                As promised, here's the renderings of the models that I was able to complete.  So sorry that I can't show you pictures of the actual fiberglassed part.  (you're not missing much, the finished part is not exactly the sistene chapel of fiberglassing.)

                 

                Cheers,

              • Re: let's make a boat.
                Andrew Emmert

                So I've been working with the boundary surface for a few hours now and I'm as happy as I'm going to get with the resulting shape.  But my major struggle now is that the surface refuses to thicken.  the surfaces will thicken individually, changing the wall thickness is ineffective, as is finishing the solid and using the shell feature.

                 

                I'm at a loss.  I cannot force the surface to hold a .125 wall thickness.

                • Re: let's make a boat.
                  Tom dunn

                  Here is the hull shelled out to .125 thick. I mirrored the boundry surface over. knitted. added a surface over the top knitted again, made a solid, then shelled. Tom

                    • Re: let's make a boat.
                      Andrew Emmert

                      Tom,

                       

                      You got the shell, that's great.  But I'm so confused.  I spent quite literally 8 additional hours last night trying to make that shell and I had the exact method you posted.  Boundary surface, mirror boundary surface, planar surface, knit, and shell.

                       

                      On further inspection, your shell feature is showing the (wall thickness > minimum curvature) error when edited, but the model is still resolving.  My model wouldn't complete the shell feature regardless of the error, yours has.

                       

                      How come two guys with the same model, same settings, and same feature tree have different results?

                        • Re: let's make a boat.
                          Jerry Steiger

                          Andrew,

                           

                          Perhaps Tom made some minor modifications that he forgot to mention.

                           

                          Use the check tool and the curvature display to figure out where you have too tight a curvature. This will often show up at the edges of lofts or boundary surfaces. You may have to play with your tangency controls or even change your splines to clean them up. I seem to recall once having to remove tangency on an edge in order to get a smoother part, but that was a long time ago.

                           

                          Jerry S.

                            • Re: let's make a boat.
                              Andrew Emmert

                              Jerry,

                               

                              I thought about that also, and I checked out the minimum radius (same value.)  Tom doesn't have any curvature or tangent relations on the boundary surface, but I know I tried that as well while trying to troubleshoot the part last night.  Far as I can tell, Tom's part still works for an unspecified reason. 

                               

                              There is a problem, however.  When fabricating this hull, the team is going to be constructing the mold using cross-sectional cutouts from the model.  I'm planning on using intersecting surfaces with the interior of the hull, which is why the tangency and curvature consistency is so critical.  I think that because the Shell feature resolves in Tom's model, I'm unable to perform a 0in surface offset for trimming the cross-section surfaces to the interior of the hull, and I also am seeing some unusual radius geometry (I believe this is related to a guide curve from the boundary surface.)

                               

                              I also found a youtube video addressing the subject, which leads me to believe that I'm using a boundary surface innappropriately in this scenario, since sketch 9 and open group 1 intersect with eachother.  This is supported by a check feature, which tells us that the minimum curvature is found near this intersection as well.  The video that I found at http://www.youtube.com/watch?v=uGZ7Ai0l3tA

                              suggests that a filled surface with a curvature relation is more suited to the type of surface I'm trying to construct, although I have yet to successfully replicate his method for my hull.

                               

                              thank you for thoughts,

                        • Re: let's make a boat.
                          Andrew Emmert

                          Ok,  so instead of focusing on using a boundary surface to construct the hull shape.  I'm trying to use a surface loft, which seems to make an acceptable shape and resolves correctly with a tangency control.  I've got the solid body completed, but the shell fails at a wall thickenss of .125, I know where the minimum curvature is while using the check feature, but I don't know what to do about it.

                           

                          any ideas?

                          • Re: let's make a boat.
                            Andrew Emmert

                            Update:

                             

                            I'm 90% sure I know what's causing the problem.  As you can see from the attached image, on the creation of the loft feature the tangency control isn't doing it's job, allowing the curvature to overshoot it's guide curve.  This happens in the area identified by the check tool as the source of the minimum curvature and also explains the unusual geometry when I attempt to use a surface offset to check for a degenerative surface.

                             

                            If we can get the loft to correctly hold a tangency relation to the swept feature, then I'm confident that the shell will work as expected and the rest of the part will be a cakewalk.  Unfortunately, I still can't get the loft feature to behave where I need a tangency relation in effect.  Perhaps someone watching this discussion can take a look at the attached model and tell me if I'm missing something obvious.

                              • Re: let's make a boat.
                                Jerry Steiger

                                Andrew,

                                 

                                I haven't opened your model (not enough time and still running SW2010), so I am just winging it.

                                 

                                Lofts aren't real good at following the Profiles or the Guide Curves or the Tangency Relations. It isn't very hard to make the Profiles, Guide Curves, and Tangency Relations incompatible by making the geometry too extreme for the Loft to keep them all consistent. Boundary Surfaces are better at all of these. Both Boundary Surfaces and Lofts have problems if you are building a surface with 3 sides instead of the 4 that they prefer. It is better to break a problematic surface up into a 4 sided part, built with a Boundary, and a 3 sided part, built with a Fill Surface.

                                 

                                Jerry S.

                              • Re: let's make a boat.
                                Tom dunn

                                It is my opion that your problem is at the very beginning. The arc you use to create the surface sweep needs to part of an ellipse, or conic in shape and not just a part of a radius. If you are trying to match the sides with a tangent relation to the hull something has to be different. I offset surfaces to the inside of the boat at a distance of .125" and there are small notchs on each end cut from the inside surfaces.

                                Tom

                                • Re: let's make a boat.
                                  Manuel Campos Costa

                                  Hi,

                                   

                                  Could anyone please try to explain to me the advantages of using loft (bodie) vs boundary surface vs boundary bodie?

                                   

                                  With the loft (bodie) I can't create the front end nose, but the shell looks much better than the thicken feature.

                                   

                                  The hull I'm trying to create is from a Hobie Cat 17:

                                   

                                  https://www.google.pt/search?q=hobie+cat+17&espv=210&es_sm=122&source=lnms&tbm=isch&sa=X&ei=guINU5P5Gs_Y0QXynYGIBg&ved=0CAkQ_AUoAQ&biw=1024&bih=640

                                   

                                  I also don't want to use free form since I wan't the hull to be updated with the modification of the dimensions that I will enter later.

                                   

                                  I have created three configurations so you can see the diferences I'm getting with the loft and with boundary surface and the boundary bodie.

                                   

                                  The one I like the most is the boundary Bodie with shell

                                   

                                  Thanks!

                                   

                                  Manuel Costa

                                    • Re: let's make a boat.
                                      Mark Biasotti

                                      Hi Manuel,

                                       

                                      Boundary is more or less a replacement generalized coons patch type of feature for Loft (with lots of special features and advantages over Loft)  that was introduced with SW 2007. It differs from Loft mainly because the 1st and 2nd direction curves (verses profile and guide)  have equal influence on the resultant surface. You should also take into account the Fill surface since Boundary or loft SHOULD NOT be used for 3 or 5> sided patches, but rather use Fill for 3 or 5> sided patches. If you'd like to understand it more, Charles Culp's site www.swtuts.com has a presentation I gave a number of years ago at SWW that explains it at length you can access it here

                                       

                                      Regards

                                       

                                      Mark

                                      • Re: let's make a boat.
                                        Jerry Steiger

                                        Manuel,

                                         

                                        Mark will know better than I, but I suspect that shelling a Boundary Body will usually give you the same results as a Thicken on a Boundary Surface. There are probably some cases where one or the other is preferable, but I suspect the differences are fairly subtle.

                                         

                                        Jerry S.

                                      • Re: let's make a boat.
                                        Tom dunn

                                        How does this look?