2011/3/13 dave <[email protected]>: > It it takes 5 axis then I'm missing something because it looks like a > truncated cone which intersects with a cube. If so, then along the axis > of rotation it is a pair of straight lines. So you calculate and machine > one pass then rotate and do another.....
No, actually, it is bit more complicated than intersection of cube and cone. I used the previously mentioned "lofting" function. The circle transitions to square. 2011/3/13 andy pugh <[email protected]>: > On 12 March 2011 22:30, Viesturs Lācis <[email protected]> wrote: > >> So the problem is finding out the necessary tilt angle, which >> corresponds to the slope of the edge. > > I think that it is non-trivial. The problem is deciding which point on > the bottom curve matches which point on the top curve (and > vice-versa). Once you have pairs of points then the mathematics is > relatively simple. I think that it should take particular point of top curve and then find the point on bottom curve. The horizontal projection of the line, connecting this pair of points, has to be perpendicular to top curve. So there is special case for sharp corner on top curve. > The solution probably involves finding the set of shortest lines from > top to bottom, and then bottom-to-top. You will often find a situation > where a sharp corner at the top matches up with an entire arc below, > and vice-versa, so need to solve in both directions. I do not think that I would need both directions. And yes, sharp corners on top contour is special case, so I think that can be avoided - put them on bottom. > I think I would do it in Matlab. I am familiar only with MathCad, but I do not know, if they are equal. 2011/3/13 Dave Caroline <[email protected]>: > Dave its water jet/plasma so the head has all the rotation and 5 axis needed, > Viesturs, I cannot see the item as heekscad bombed on the stl file but > gcode can handle the maths so you can can just do the maths in the > gcode program. There is also IGES file, if that helps. Let me know, if I can save it in another file format. > Also then is another way which you can develop from sheet metal practice > http://search.sheetmetalworld.com/news/articles/197.cfm > and develop the paths in gcode break into as many triangles as needed > for accuracy I think that this article tells, how to create particular part from sheet metal - how to design correct layout so that it can be bended/welded together to get the right part. 2011/3/13 dave <[email protected]>: > Ah, got it. rotate and tilt is almost all it needs. > Is this for mental exercise or does this part have a purpose? Honestly - I just would like to produce some range of samples, that would demonstrate, what a 5 axis waterjet machine is capable of. It is not intended to be a part with particular use, just demonstration purposes. > > (5 axis cut - arc at top, straight line at bottom) > (commented code is for max5kins - note, no A axis) > (tested on 5axis sim) > (assuming table at Z=0) > > #10 = 1 (material thickness) > #11 = 2 (radius, and the square's diagonal) > > G10 L2 P2 Z[-1*#10] > > G55 G00 Z[#10] > (X[#11] Y0 A0 B0) > X[#11] Y0 B0 C0 > M3 > > #1 = 0 (#1 is used as the parameter for the equations) > O101 while [#1 lt 1] > G1 Z[#10] F10 > #21=[cos[#1*90]*#11] (X) > #22=[sin[#1*90]*#11] (Y) > #24=[atan[[[1-#1]*#11]-#21]/[#10]] (A, for AB) > #25=[atan[[#1*#11]-#22]/[#10]] (B, for AB) > #26=[#24+#25] (B, for BC **pretty sure this is not > right - how do we really calculate this?**) > (X[#21] Y[#22] A[#24] B[#25]) > X[#21] Y[#22] B[#25] C[#1*90] > > #1 = [#1+.01] > O101 endwhile > > M2 > Wow, thank You a lot! One thing I have yet to learn are this kind of code - with variables and calculations in the code. I think that now I have very good starting point to start figuring out, how can fit all this together. Thanks all of You, guys, for the ideas! Viesturs ------------------------------------------------------------------------------ Colocation vs. Managed Hosting A question and answer guide to determining the best fit for your organization - today and in the future. http://p.sf.net/sfu/internap-sfd2d _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
