How best to toolpath a feature like this?

[djstedman]

OK I have a part that is configured pretty much like what I am posting a pic of below, I can't share the actual part since its from a customer but suffice it to say it has a ton of scallops on it that look pretty much like what is shown below in green.. although the actual part is much longer so it wouldn't be possible to sidemill these scallops

 

If there were just one feature like this I might muddle through by creating geometry and doing a lot of manual coding, however since there are a lot of features I was hoping to maybe use a multi axis path powered by surfaces / solids..

 

I plan on doing these pockets as well as a bunch of other features with the part setup on a 4 Axis VMC, It seems to me this should be possible using multiaxis paths and restricting to 4 axis, but so far I have been sitting here cursing the part, my computer, and my inability to get it to make a decent path..

 

FYI I have MCX7 that I am using for this..

 

So .. any suggestions?

[htm01]

 

 

So .. any suggestions?

add the picture

[djstedman]

Ok Ok .. hitting attach helps .. it should be there now lol

[GoetzInd]

What rad. in the corners, material? Have you tried a pocket then using axis sub?

 

edit: Walls don't look like they're perp. to floor on ends so axis sub would only work for roughing. You would need to go multi axis locked to 4th at least for finishing.

 

Mike

[djstedman]

Well on the actual parts the rad is 10MM the walls are perpendicular though they do have a weird runout radius in the ends .. its gonna have to be a rough then finish later job no matter what..

 

Its being done in 300M steel (pretty much a fancy 4340 through hardening steel) Anyhow.. its going to be roughed then brought up to about 55HRC then it needs to be finished at hardness its the only way to maintain our tolerances

 

The pocket idea and axis sub might be a good way to go for roughing the majority of it though.. gonna have to attempt that..

 

Still looking for a nice smooth method for finishing since the plan is to profile the majority of these things with ballmills made for hard milling..

[Micromoldmaker]

Most Surface roughing tool paths would give tons of air cut time. Surface scallop finish with a boundary might work. You can do subsequent tool paths leaving stock. For example. Leave .100 on first tp . Leave .050 on second tp and son on.

[MotorCityMinion]

Opti Rough, raster, then 3d contour for the walls.

[Jim at Gentex]

I think Micro's approach is a good one.

Using a finish toolpath with 'stock to leave' is a timesaver for sure.

 

The great thing I've discovered about Mastercam is that there are a number of ways to accomplish a task.

That can be good or bad, depending on your point of view.

 

Sometimes you just need to try several different approaches and see which one looks best.

[djstedman]

I appreciate the input so thanks for that everyone, I was really looking for some more insight into the multiaxis paths though since I know I should be able to use the multiaxis paths to create a 4 axis solution for this. I know there are some 4 and 5 axis guys on this forum that could probably at least point me in the right direction.. I was messing around with parallel to surfaces and was soooo close .. but yet so far away..

 

Assuming I was dead set on a multi-axis path locked to 4-axis.. which would seem to be the best strategy for this type of a feature?

[crazy^millman]

Put a sample file and someone might be more inclined to do that.

[Jayson Kramer]

With waht I see on the picture in the past I do a mixture of Axis sub as a pocket for roughing then 5 axis swarf as a finish. as the crazy man stated share the file.

[djstedman]

Yeah I should have thought to attach the file itself and not just a picture, anyhow here is the file..

 

I removed the toolpaths I had in there that I was testing with because they weren't even close to working..

EXAMPLE_SCALLOP.MCX-7

[Jayson Kramer]

Ok the axis sub works great now as for the swarf , I think we need to use a different path to multiaxis surface cut due to the fingers sticking out.

[djstedman]

For the roughing after seeing people mention a pocketing path with axis sub, it made me think a bit and I have been messing around with making a dynamic mill path using axis sub and that looks like it might work pretty darned good

 

As for the finishing, when I tried any of the multiaxis paths my biggest problems came from it wanting to dive right through the part and it seemingly wanting to ignore check surfaces..

 

Then again I have never been able to get my mind wrapped around that collision strategies page having 4 options.. and to say the documentation for that stuff is weak is being kind of generous.. cant wait to see what you come up with.

[Chris Rizzo]

Axis sub with dynamic works great. I believe current posts support inverse time with dynamic toolpaths. X5 era they didn't out of the box and needed some trickery to work. I might be able to dig up what I did to make it work, if you can't get inverse time out.

[MrFish]

finishing option

EXAMPLE_SCALLOP-MARK.MCX-7

[crazy^millman]

Here is my take on this part. Hopefully it will give you some ideas how to think outside of the box.

 

 

Edited:

 

File removed email me and I will be glad to share it with you.

Edited November 12, 2013 by Crazy^Millman

[djstedman]

Mr.Fish .. that path looks great for finishing with a ball which i'm likely going to have to do

 

Crazy .. That Multisurface path is awesome.. would never have thought of doing it like that ..

 

Damn.. that's some great stuff from both of you .. definitely has got me going in the right direction.. Cant thank you enough for pointing me in the right direction!

[megaman]

crazy , i would have never thought of using multisurf on a 4 axiseven though i do exactly that tool path with 4 axis output on my 5 axis.

 

good stuff

[AlbertZini]

Here is my take on this part. Hopefully it will give you some ideas how to think outside of the box.

Crazy^Mill/Turn how you create vectors( lines control) ?

[crazy^millman]

I old schooled it. I take the curve from the geometry and then use the make points on the curve using create/point/segment then a number. I then use my auto cursor with right click to force it to only pick points to create the vectors from. Then take the time to pick them one by one and then go from there. Window for picking the vectors still needs a lot of work. You have the geometry to drive the path already and use the minus vector to put it at the depth and done.

 

Other thing is the Multisurface has depths of cut. I did not do it to keep the file size small so it was shareable.

[Aaron Eberhard]

Here's my take on the project... I didn't create any geometry at all, I morphed between the two red "surfaces" (I used solid selection), driving off of the green "surfaces" (again, solid selection). I used a Zig-Zag pattern set to From Center Away with my Linking (between slices) set to Blend Spline. This gives me a pseudo-high speed methodology, with all of my transitions happening in free space.

 

Note: In the picture, I reduced the step over so it was a bit clearer in a screen shot.

 

Roughing would be as easy as clicking the Depth Cuts options in Roughing if you wanted to do that. You could always do a swarf toolpath if you wanted to really clean the walls nicely as well..

EXAMPLE_SCALLOP-ACE.MCX-7

[crazy^millman]

Good Job Aaron glad to see again a different way to go about programming the part. Like I was telling a customer today programming is sometimes like art. All artist pick what defines the picture to them and though there are scientific principles that creativity is still important to consider when programming parts.

[Aaron Eberhard]

Good Job Aaron glad to see again a different way to go about programming the part. Like I was telling a customer today programming is sometimes like art. All artist pick what defines the picture to them and though there are scientific principles that creativity is still important to consider when programming parts.

 

Eh, mine was driven by laziness Errr... Did I say laziness? I meant Efficiency. Yeah, that's the word I couldn't be bothered to create geometry when there was a perfectly good solid sitting there Sorta like your Msurf!

 

You're absolutely right, it's as much art as science.

 

For those just getting started in multiaxis, we teach 4 things to think about the process, and the order to tackle them. Don't worry about the next step until you've completed the one you're on. I really recommend cranking down the surface tolerance and stepover, then working through an iterative design process one option at a time until you really understand the paths.

1. Cut Pattern - Get the toolpath pattern to look right on what you're trying to cut. Don't worry about how it'll tilt or what it will hit at this point.
    
   
2. Tool Axis Control - Now's the time to worry about how the tool is moving. In this case, I want it to point away from the surface, along the normal. There may be times when I want it to tilt through a curve, or to or from a point, etc.
    
   
3. Tool Tip control - Now we need to be concerned about what part of the tool tip is actually contacting the part. If this were a bull nose cutter, you'd need to have the flat always in contact, for example, or if you had a drive shape hidden under a wire mesh, you'd comp the tip to that mesh, etc.
    
   
4. Collision Control - This is the tricky one, as it gets the most people in trouble. Often, you don't even need it. But this is where you want to tell it, "tilt away from that wall if you're going to hit" or "Pull up along the tool axis if you're going to gouge."