Post By 5 axis Fidia guy
Post By 5 axis Fidia guy
I have this part I am looking to program and I need some advise on the steps YOU would take to rough, semi finish and finish this part. What tool paths would you use and why? Any feedback would be great.
Z Level Rough - Clear bulk of the material
Rest Rough - Rough out stock where the first tool couldn't fit
Rest Rough - Step down again in tool size clearing more stock
Equi Distant offset - Semi Finish
Equi Distant offset - Finish pass
Pencil Trace - Clean up intersections
Am I missing anything?
( Equi on parting surface )
I would probably break up the 3d offsets to work on the deeper lugs individually, at least on the finish pass. Doing 3d offset on the whole part for me generally doesn't yield the best surface finish.
Originally Posted by jelrod1
On a part like that simple Z-Level finish would be best.
Just make sure stepdown on shallow radiuses is smaller than on steep regions.
The exact steps I would probably take on this part using surfcam would be:
--SRM Roughing stepping down tool radii to below the smallest part radius
--3d offset semi finish on parting surface
--3d offset semi finish on cavity not going into deep pockets (would draw capped surfaces over the pockets including the radii)
--3d offset semi finish on pockets
--3 separate 3d offset finishing passes on the 3 areas
All of this is to get the best surface finish possible and is usually necessary in tough to machine materials. This is also assuming there are no other tight corners or areas that the print deems critical that would need more individual attention.
Last edited by jelrod1; 08-03-2012 at 05:21 AM.
Reason: Changed word roughing to semi finish for clarification
Al why not post an IGES of the file...
My method, and my opinion only. You will always get a better finish, doing a Z level slice toolpath on steep walls, no exceptions. So in the "sole" area I would break up the toolpath and z level the steep areas, and raster the shallow areas, some cams have the ability to do this automatically for you. Otherwise you must be a little creative. On the parting line, I would use a raster toolpath, you will get a much better result, on your picture with your 3d offset you can see how the toolpath slows down and collapses on its self, that will give you a poor result on the finished part, I ALWAYS avoid that at all costs, ANY sharp direction change is a no-no. Also please break up your molding area and your shutoff parting line areas, ALWAYS and no exceptions, you will never and I don't care what machine or control you have, keep a sharp shutoff edge just throwing a toolpath over the entire block.
Last edited by 5 axis Fidia guy; 08-03-2012 at 10:23 AM.
5 axis Fidia, You make good points on the parting surface and the sharp directions causing problems. I will still stick to using 3d offset on the pockets since they have radii at the top and are pretty consistent shapes. If you isolate them and use a top down approach it machines the side walls with a constant z level slice with the contour, and also includes the radius. I cant get that kind of control out of z level finishing when the radii are involved. I will take back what I said about the parting surface finishing. I think the approach you listed will definitely yield a better surface finish. I enjoy seeing the different methods people use to machine the same parts. On the pocket below, the bottom was the shutoff so it wasn't priority. I could not get this kind of consistent toolpath on the side walls without 3d offset. z level finishing using either scallop or constant z left the radii raggety if that's a word because it wouldn't follow the contour with the step down.
So to get better finish on the parting surface you would cut with as little change in direction as possible?
free image hosting
( Before off set surface )
Here is another question for you. Since I have isolated the parting service to "raster" or what we call planar, basically a back and forth tool path, what do you do for the edges. It's always a good idea to start and finish of the part right. So in your CAM package does it give you an option to start off and finish off the surface, or is it all surface driven like BobCAD.
For me I would need to extend the surface out on the inside and outside of the part. This way I force it to start off the part and finish off the part.
(Using the un trim surface, then extend surface option I can get a little more surface to cut and restrict with a boundary)
Yes, I do believe using as little change as possible on direction will be best as 5 axis fidia pointed out. As to what angle you make your cut, that will be geometry driven to determine what's best, or you can use flow surface to make a planer cut follow the surface directions. That is where using 3d offset for that is useful if you can live with the surface finish. Most of my parts are pattern equipment that is going against sand, so I usually can unless it's an iron corebox on it's parting surface.
On the other question, I usually extend surfaces to make it start and end off the part if necessary. Machining right to the edge never leaves it crisp. You just have to be careful that you gouge check to make sure the tool doesn't run into something on the inside. It doesn't look like it would on your part.
Extending the edges is always good practice, however depending on your cam you might not need to do it with a surface. I use PowerMill and on the raster "planer" toolpath I would just add a lead in and lead out with a surface normal arc, which is basically what you did without actually creating a surface. And as mentioned ealier the direction really depends on the block shape sometimes I go 0 degrees, some 90, some 45 etc.. Looking closer at the sole area without actually twidling it around on my screen, I would agree, 3d offset might be the best answer for getting the lower fillets nice.
surface blend between the cavity and outside of block would work great on this.
It is hard to tell by the picture I posted but the floor is actually a shutoff to a window through the part. That surface is something like a 50 inch radius. Nothing on the part can be done with flat tooling except the main plateline of the pattern. The part had somewhere around 60 of these windows in it and 3d offset was sufficient for surface finish of the shutoff since it had nothing to do with the part. The sidewalls of the windows however needed to be really smoothe. It is also hard to tell but the whole feature is on a radius. I have better luck getting a smooth tool path on contoured pockets using 3d offset over z finish. I have better luck with z finish on steep walls with other shapes in them also that seem to drive 3d offset crazy. For me it comes down to trying different methods to figure out which one is best for that geometry. I am anxious to see how some of the new tool paths do if this new version would ever come out.
Call me old-fashioned but I still prefer to finish my floors with flat end mills QUOTE]
That's not called old-fashioned, that's called efficient.
On this part everything is curved so not many "flat" areas to cut. What I am looking at now is roughing out the cavity. My thought is I can rough and rest rough to get down into the smaller areas and then come back with the equi distant to finish. I'll post some screen shots of the tool paths as I go.
This is what I've come up with for the 1st roughing tool.
Ok well to start I have a few assumptions
1. You have proper rights to the file that you are using for testing or as a demo.
2. Material = Aluminum
Area Clarence, 1" or .75" EM, .1" Depth of cut, 75% step over
Semi Rough Parting Line
3D Offset, .375 BM, .05" Step over
Semi Rough Cavity
3D Offset, .375 BM, .05" Step over
Finish Parting Line
Raster, .1875 BM, .008" Step over "Rest milling in the toe area may be needed"
1. Cap all of the cleats
2. 3D Offset, .1875 BM, .008" Step over
3. Uncap cleats and create boundaries for the unfinished areas
4. Constant Z Finish Cleats, .125 BM
5. Raster Finish Cleat Shallows, .125 BM
6. Multi-Pencil Rest, .093 BM, .002 Step over, "An additional .046 Rest May be needed"
Also if you are using this for a demo, I would suggest adding Bob-Cad as a subtracted geometry in the middle of the plate.
Thank you Loner,
Those steps seem to work well. Creating the boundaries for all the cleats is turning out to take some time. So instead of capping each I just extruded a surface to stop the tool path from going down an deeper.
free image hosting