Re: file size-processing time

[camguy]

I just completed some tests on this. i created a coons surface approx. 10.0" x 17.5" using 4 large arcs. the resulting surface (s1)had control points of 199 x 16, degree 3x 3, file size 81kb. i created three surface finish parallel toolpaths

op-1 cut tol. .001" time 4 sec.

op-2 cut tol. .0001 time 44 sec.

op-3 cut tol. .00001 time 8 min. 53 sec.

I then converted the arcs to splines .001" tolerance, and created a new surface (s2)which had control points of 13 x 16, degree 3 x 3, file size 11kb. i then repeated the toolpaths.

op-1 cut tol. .001 time 2 sec.

op-2 cut tol. .0001 time 17 sec.

op-3 cut tol. .00001 time 4 min. 18 sec.

As you can see, there is a substantial difference in processing times. File size does have an effect depending on the number of control points on the surface and cut tolerance also makes a big difference. What i usually do when creating toolpaths is start a multisurf finish, zero stock, cut tolerance of .010". This should process very quickly. If i notice the processing slow to a crawl on a surface, i stop and analyze the surface. if there is a large number of control points i will try to reduce the bounding geometry (remnodes0 and recreate the surface. This can be a little time consuming, but it can really speed up cutter path creation. Other benefits are smaller file size, faster shading and dynamic rotation.

[BernieT]

Very good point, the more complicated the surface - longer calc times. And surfaces created with splines are much more complicated than ones created with lines and arcs. Also, Mcam tends to create surfaces with lots of control points in order to match the accuracy set in the config file.

I like to translate complicated surfaces into Rhino3D and rebuild them to fewer control points but within an acceptable tolerance, then translate back to Mcam. I have had very good results doing this. Rhino lets you adjust the number of control points and tells you how close the new surface is to the original; you can keep changing the number of control points until you get a new surface as accurate as you need.(I usually try for within .001 of the original)

BerTau

PS. I just noticed that Camguy's spline surfs were simpler than the arc/line surfs, this is opposite of what I have experienced.

Hmmmm

[ 08-30-2001: Message edited by: BerTau ]