HSM Raster gouges

[MotorCityMinion]

I'm getting small, erratic moves in the raster tool path that look like the're gouging the part in verify. No arc filter turned on. Tried it at 0 and 90 deg. 0 deg. being parallel to the surfaces looked a bit better, but not much. Is this the norm for this path? Maybe I'm applying it to the wrong type of surfaces? The Finish parallel path seemed to be much better at 0 or 90 deg. The first image uploaded is the solid, not verify. Sorry about the huge pics.

 

Any help would be appreciated., Thanks.

 

 

[Colin Gilchrist]

Try using the Surface Finish Blend toolpath.

 

Always use the filter, even if you turn off the "output arc" settings.

 

Do you intend for the toolpath to "roll" over the edges? If not you can create some fence surfaces, or extend the trimmed surface edges and use a containment boundary to keep the toolpath from rolling over the part edges.

 

The best toolpath for what you are trying to accomplish (in my opinion) would be Blend.

 

You would need to create two lines that define the "drive" direction (similar to cut direction in Flowline).

 

Also, with Blend the tip of the tool will drive to the line, so you might have to offset the lines to get the tool to cut all the way down to the front edge...

 

HTH,

[MotorCityMinion]

Thanks for the input. Yes, I've tried the blend, it's awesome. Tried it on the nose of this part in fact. Boundaries were used with the path in the pic, leads were set at default .078 linear, arc entry and exit.

 

"Always use the filter, even if you turn off the "output arc" settings." Are you referring to the Blend tool path, Raster toolpath, or all surfacing tool paths in general?

 

That option, (Raster path) with the filter set to 1:1, .0002 TT, arcs disabled, gave me an even nuttier path.

 

I'll try the blend tool path on all those surfaces and see what happens.

 

Back to the main topic. What is Raster actually good/used for? The problem I'm having seems to be localized in an area of the solid where one rad isn't tangent to the next rad. The valley area in general. Everything else looks good.

[MotorCityMinion]

bump

[Colin Gilchrist]

Raster is a fairly good toolpath for what are termed "shallow" surfaces. Where all the surfaces are close to flat. This is due to the way all the "Parallel" style toolpaths work. Raster (High Speed version of parallel) works by taking a plane that is oriented like the Front plane and making multiple slices of the geometry. It then offsets these slices and compensates for the tangency of the tool along that offset slice. It then links these compensated slices together forming the tool's path.

 

This works great on a set of surfaces, but the problem is that the toolpath does not adjust the stepover to maintain an even scallop height (like flowline does). Take a look at the picture of the toolpath you first posted.

 

In the middle where the part surfaces are flatter, your scallop will be the smallest (not where it goes up over the bumps, but you get the idea). Then there is the angled ramp in the back. The scallop will be larger on this wall because of the angle. The closer the surface gets to vertical, the larger the actual distance between the slices the toolpath creates.

 

You can see this in the nose area. The distance between the slices appears to increase as the surface drops off...

 

Back to your problem (which is causing you to think Raster/Parallel toolpaths are crap right?).

 

First off, if you are using a tool that is close to the radii on either edge of that valley, that could be adding to your problems.

 

One thing I noticed is that you are using .0002 for your tolerance. This doesn't give the toolpath algorithm very much "room" to smooth out the geometry. I'm guessing that if you bumped your tolerance up it would really improve the toolpath quality. I'd say try .0005, .0008, .001, .0015, .002, and .003 and look at the results.

 

I'm guessing you won't have to bump it up very much to see much better results.

 

The other thing that could affect a parallel toolpath is the dreaded "Gap Settings". (Gap settings don't exist in the HSM toolpaths)

 

Gap settings tell the toolpath how to connect each of the slices that the algorithm makes on your set of surfaces.

 

The Gap settings tell Mastercam when to retract, and when to keep the tool down on the surfaces (continuing the toolpath in most cases).

 

Try a parallel toolpath with the same settings as your Raster toolpath.

 

In the Gap settings, set your gap size to inches and enter 1.0.

 

Set your Motion < Gap size, drop down to "Smooth".

 

Turn on "optimize cut order".

 

Under "Advanced settings", set the radio button to "Only between surfaces solid faces".

 

This should give you a fairly nice toolpath with the filter bumped up a bit. That and/or a smaller tool will probably improve the toolpath quality a lot.

 

HTH,

[MotorCityMinion]

Colin, thanks again. I'm relatively new to the surfacing features in MC and the tool path algorithm you speak of just went over my head. I use the parallel path frequently. (It's the raster that I knew nothing about). Parallel was my first choice but in this case with the sides tapering from 0 to 5, then 10 deg, in x,y and z, I was getting some funny moves, not as uniform as I would like. I'm probably being abit greedy in trying to do all these features in 1 op. Transitions between sloping surfaces also caused some erratic moves with the parallel path.

 

Gap defaults are set at .50", and the first thing I do when it looks like the tool is rolling over the edge is shut off the edge roll over entirely in advanced settings, then look to see what changes. That part is less than .60 wide. The rolling over appearance in the cutter path pictured is actually a arc feed in from the vertical, and it pulls up on the exit. It just looks funky in that view.

 

"First off, if you are using a tool that is close to the radii on either edge of that valley, that could be adding to your problems."

 

The tool I'm using is less than 25% of the actual rad that's getting cut in the problem area

 

The profile of that part has a tolerance of p/m .0004 on the profile and I can't risk getting any facets on it, which always seems to happen when I mess with the arc filters. (Lack of experience). They also call out a a$$nine tolerance of p/m 0.00.01 deg on the side angles.

 

BTW, I did end up applying the blend tool path to the entire profile, 37 features in 1 SHOT! The tool stayed down and looked smooth all over. The solid was created in Solidworks by mwa, and as usual, all SW solids machine great in MC. The raster path, as you can see, also did a fine job with those 37 features, of staying down without any really funky entry or exit moves and looks like it has much potential in the right application. I haven't given up on it.

 

I tried the blend path on a much simpler part that was created with MC solids, 4 simple features, all tangent, and it ended up looking rather $hitty transitioning from one feature to the next, which perplexes me considering the geometry pictured in the part above worked fine with that same path.

 

Just for kicks, I'll put the raster path back in, then bump up the tolerance and see what pops. I may even run it back thru SW and make those problem rads Tangent just to see if it helps with the raster path.

 

Once again, thanks for your time. I'm sure alot of peeps here are going to get some good pointers out of this post.