deform question

 From: wimverbe 22 Mar 2014  (1 of 4)
 I wanted to deform a gearwheel (just for advertising purposes, not CNC or 3D print), but ran into a problem. the deform target surface has an edge going to the middle, since it is a revolved curve. so the result is completely warped, but not with the desired centre. (quite a funky shape though...) can this be solved with deform, or should I take another approach? tnx wim Image Attachments:

 From: bemfarmer 22 Mar 2014  (2 of 4)
 6569.2 In reply to 6569.1 . EDITED: 19 Apr 2014 by BEMFARMER

 From: Michael Gibson 22 Mar 2014  (3 of 4)
 6569.3 In reply to 6569.1 Hi Wim, probably if you check your base surface by turning on its control points you'll see that the base flat disc is actually not a revolve, it's a 4-corner-point plane. That's built into revolve where if it sees a piece to be revolved that is a line perpendicular to the revolve axis, it replaces it with a plane primitive rather than using a revolved flattened disc. Normally for regular operations such as booleans it tends to be better to have things that are planar to be analytic plane objects rather than other kinds of surfaces that just happen to be squished down into the shape of a plane, but for your particular case here using Flow the particular layout of the surface is significant. You'll need to make the base surface in your case here by doing a revolve of something that has a little bit of curvature to it and then flattening the resulting surface, in order to get a flat revolved base shape like you want. Or also you could use "projective" mode of flow as well, that's described here with some example links: http://moi3d.com/forum/index.php?webtag=MOI&msg=5936.2 It's possible that you may run into problems using the regular non-projective mode using a revolved surface as the base surface, because the way regular surface-to-surface flow works is by dropping points onto the base surface and then evaluating the equivalent point on the target surface. But the pole area of a revolved surface kind of throws a monkey wrench into that since many points of the surface are stacked on top of each other right there. You may be able to get away with it if you don't have pieces of the model that drop down to the pole area, but it's not just the visible areas that get processed like that, if you have larger surfaces that are trimmed it's possible for the underlying surface to cross the pole area and be processed there as well. Hope this helps explain what you were seeing though, basically an optimization in revolve made your flat revolve structured different than what you wanted. - Michael

 From: wimverbe 23 Mar 2014  (4 of 4)
 Thanks again Michael for your more than clear explanation (as always), I tried both methods and it works splendidly! I am now trying to figure out how big the base and target surface should be so that the teeth of the gear do not deform but the inner part does. as you can see from the screendump, the placement of the 2 surfaces is crucial. 2nd from right is the projective method, the 2 left gears are with the non projective method, but with different placement of the 2 surfaces. the 2nd gear has the 2 surfaces aligned in the centre of the original gearwheel. to be honest, I hadn't noticed that little checkmark square with 'projective' till you mentioned it. the non projective method takes considerably more time to calculate. MoI rocks! wim Image Attachments: