On Thursday, June 16, 2011 12:56:34 AM Kent A. Reed did opine: > Gentle persons. > > Recall that, on 6/3/2011 3:41 PM, Viesturs Lācis wrote: > > Hello, gentlemen! > > > > I would like to ask, if anyone can suggest me a way to generate a > > g-code for a single-time situation: > > The part consists of 3 pieces of tubes - 2 of them need to be welded > > to main piece. > > What I need is to get a g-code that describes the intersection lines > > of each 2 tube pieces. > > I have SolidEdge model, IGES, STEP and STL models for the whole part. > > > > Does anyone have any idea, how do they usually generate toolpaths for > > such kind of machines? > > > > Viesturs > > I replied that it seemed pretty straightforward to work up and Gene > Heskett then suggested I put my workup on the Wiki. > > I realize this may be moot since Viesturs received g-code to play with > from Stuart Stevenson but I still like the intellectual challenge. > > I assumed that any two intersecting cylinders (aka tubes or pipes) have > equal diameters and that their centerlines lie in a plane. If that isn't > so, then the following must be replaced by something much more > complicated. > > With my assumptions, working out the analytic geometry for the > intersection lines and for the normals to those lines was > straightforward, but I've discovered that I don't understand the g-codes > available to emc2 as well as I thought I did. (Do you think the fact > that I've never touched any 5-axis machine, let alone a PUMA-like > robotic arm, has anything to do with this?) > > With my assumptions, the intersection lines Viesturs seeks are two > half-ellipses, each lying in its own plane, with the two planes > perpendicular to one another. I had in mind an elegant g-code solution > which would, for each half-ellipse, define the plane containing the > curve to be a working plane, and then send 2D-points and 2D-vectors > (actually, these would be mathematical functions defined in subroutines) > that define the half-ellipse and its normal in the working plane. In > this solution, the welding head would trace around the curve while > rotating about an axis perpendicular to the plane so as to remain > pointing inward along the normal. > > Now that I've read and reread the relevant material in the Users Guide, > I'm not sure I can do that. Am I just misunderstanding or do I have to > send 3D-points and 3D-vectors that define each half-ellipse and its > normal in space? > > Also, is there any purpose served if I post my analytic geometry to the > Wiki as a "first half" score to the problem, so to speak? My wife is > going under the surgeon's knife next week and it may be a while before I > get to the "second half". > I don't see why not, maybe someone else will finish it while you are tending to the missus.
You folks are concentrating on the welding, but before you can weld it, the ends of the pipes must be prepared. If they butt meet at an angle, and are the same size, then its a simple miter saw problem. If however a space frame where the pieces are triangulated, even with dis-similar sized pipes, then again it seems like a simple jigging setup problem, one where you want to drag the end of the tube being welded to the side of the other, slowly past a ballnose that is tracing a half circle whose diameter matches the other pipe. Once one pipe is properly seated on the other, and all such pieces are assembled, then rather than welding, I would be tempted to 'furnace braze' the whole assembly in one pass through the oven. Obviously not as strong as properly welded, but done in one swell foop, and with less stored stress than welding would leave. For the weld by emc, it strikes me that this same profile of motion could be done, with the 'wrist' holding the migs wire feeder being auto turned so that it always splits the angles. I can visualize it, but I have no clue how the math that controls the 'wrist' angle can be done in real time as it moves around the joint. Don't forget too, that generally speaking, one must move the fire in a 'j' stroke, always pushing the heat ahead of the puddle. Maybe it can be done, but I've had relatively poor luck just trying to push the puddle in a straight line, so I usually revert to a j stroke of about 3/16" out & back, with the dwell at the 'bottom' of the j on the heavier gauge metal if there is a difference. OTOH, I've never tested for a cert for arc welding of any kind, so take that with a suitable measure of salt. My work with a smith wrench has stood the test of time quite a few times though. One has so much more control over the composition of the weld puddle with gas that you can do almost anything with it. Even magnesium, but you'd need to be better than me. A lot better. ;-) > Regards, > Kent > > > ------------------------------------------------------------------------ > ------ EditLive Enterprise is the world's most technically advanced > content authoring tool. Experience the power of Track Changes, Inline > Image Editing and ensure content is compliant with Accessibility > Checking. http://p.sf.net/sfu/ephox-dev2dev > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users Cheers, gene -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Q: Do you know what the death rate around here is? A: One per person. ------------------------------------------------------------------------------ EditLive Enterprise is the world's most technically advanced content authoring tool. Experience the power of Track Changes, Inline Image Editing and ensure content is compliant with Accessibility Checking. http://p.sf.net/sfu/ephox-dev2dev _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
