SWARF TOOL VECTOR DIFFERENT FOR EACH PASS

[lowcountrycamo]

Hi Guys,

We build mostly aerostructure and I use mostly swarf to finish part walls.  I am always fighting the tool vector.  It will change for different z passes and offset passes.  On this part I am using sync with upper and lower chains.  I have a floor but no wall surfaces.   But depending on the part I often use normal to guide and sometimes tilt line.  Does anyone know why this is happening? See pic

[crazy^millman]

Without seeing some sample geometry no. I strongly recommend you either make it do what you want with Curve 5 axis and vectors or make surfaces to help control smoothness. The ModuleWorks Swarf will work with only chains and maybe it will provide some smoothing you're not getting if using the old school swarf. That said chains can be an odd duck and again without something to review just guessing. Are they lines and arcs? Are they symmetrical and have the same number of entities? If splines have you rebuilt them to have equal number of control points for upper and lower chains? Verisurf free tools is the method for doing that.

If it something you cannot share here you have my email and I will be glad to take a look.

[lowcountrycamo]

I would love to get your feedback.   If this does not come through, I could try another way.  Thanks

TEST.MCAM-CONTENT

[crazy^millman]

I created edge curves at every surface intersection and changed the process to use them in the strategy. It looks better to me. You might try kicking down the angle change from 1.5 degrees to .1 if on the machine still not giving smooth motion. The other option is flood it with code and run no more than .01 in the maximum distance.

Hopefully 5th Axis Vices doesn't send me another cease and desist for using the word 5th Axis in that picture.

[Aaron Eberhard]

I believe you're having that issue because the centerline of the two arcs isn't the same when are offset by .010 (your stock to leave for the first pass), and with Damping turned on, it can't really contour to the corners well anyway.    When you offset by .010, the radius is exactly the same size as your cutter, which is going to always cause a bit of issues with the collision algorithms.    This is exagerated since you're using a bull nose cutter and causing gouges around the floor that will be cut out.

Also, agreed with Ron on using tilt lines here, especially because you're so close to your cutter size.

Just changing to lines and the "Pattern Layers" Distance to .008 will give you this:

Oh, one more thing with this toolpath... If you're near the limit of where the tool is at the same radius as your corners, don't be afraid to really crank up (down?) the tolerance.   Right now you're at .0002 which is going to be way better than stock, but don't be afraid to try .00005 or .00002 (note the extra 0), that will give it way better data to calculate fit with.

 

[lowcountrycamo]

Gentlemen, much better!  Thanks for the advice.

[gcode]

I tried a different approach, parallel with tilt lines.

The motion looks pretty good, but I can't figure out why it's making 2 passes.

I also lied about the tool's length of cut. That helped alot. I think it would help with the swarf toolpath as well

 

TEST.zip

[crazy^millman]

Just now, gcode said:

I tried a different approach, parallel with tilt lines.

The motion looks pretty good, but I can't figure out why it's making 2 passes.

I also lied about the tool's length of cut. That helped alot. I think it would help with the swarf toolpath as well

 

TEST.zip 1.33 MB · 0 downloads

He wanted 2 passes so the fact you got them is a good thing.

[gcode]

Yeah, but why.. I can't figure it our?????

[lowcountrycamo]

Gcode, 

I love this!  I have tried to make Parallel work but could never get the tool to stay true to the wall.  Many thanks for the help.

[gcode]

18 minutes ago, lowcountrycamo said:

Gcode, 

I love this!  I have tried to make Parallel work but could never get the tool to stay true to the wall.  Many thanks for the help.

I think lying about the tool's length of cut has a lot to do with this working. I'm completely baffled as to why it's making two passes though.

It looks like it should run pretty smoothly

You might get better results from the swarf toolpath if you lie about the tool's LOC.

I guess you could lie about the tool's LOC in Mastercam, then use the correct tool definition in Vericut.

You can also define the corner surfaces as feed zones and slow it down in the corners then speed it back up on the straightaways .

[lowcountrycamo]

Part came out beautiful.  This Mag1 is a really nice machine.  And the apps guys are the best I have worked with.    The day we purchased this machine I was given a 200-page pdf outlining what they have found has worked and what has not.  Vibrations analysis, RAH model numbers, RobbJack cutter catalogue for high HP/RPM mills, efficient toolpath use of accell/decell, etc. We purchased and I learned the Cut Pro software (tap test).   We are having much success, and the owner is happy.    Thanks again for the advice.  

[crazy^millman]

38 minutes ago, lowcountrycamo said:

Part came out beautiful.  This Mag1 is a really nice machine.  And the apps guys are the best I have worked with.    The day we purchased this machine I was given a 200-page pdf outlining what they have found has worked and what has not.  Vibrations analysis, RAH model numbers, RobbJack cutter catalogue for high HP/RPM mills, efficient toolpath use of accell/decell, etc. We purchased and I learned the Cut Pro software (tap test).   We are having much success, and the owner is happy.    Thanks again for the advice.  

Pictures or it didn't happen.

[Aaron Eberhard]

On 9/7/2023 at 12:54 PM, gcode said:

 I'm completely baffled as to why it's making two passes though.

Huh, me too...  That's really odd!