Community
Fusion Manufacture
Talk shop with the Fusion (formerly Fusion 360) Manufacture Community. Share tool strategies, tips, get advice and solve problems together with the best minds in the industry.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Request for guidance to generate 4-axis toolpath that follows surface isolines (flow lines) about simultaneous rotary axis

12 REPLIES 12
SOLVED
Reply
Message 1 of 13
cgoody
602 Views, 12 Replies

Request for guidance to generate 4-axis toolpath that follows surface isolines (flow lines) about simultaneous rotary axis

Aloha all - does anyone out there know how to or can give advice on how to generate a toolpath for a cut that follows the isolines or flow lines of a given surface or body, that is rotated about a longitudinal axis (XYZ & A) for a long and organically curved object such as a surfboard or a kayak?  X, Y, and Z axes are linear like a CNC router bed, but with a long rotary axis (A) along the x-axis.  X, Y and Z are all on a form of gantry that moves over the entire A-axis.

Currently, I'm using open source BoardCAD, with a custom python based script used to generate the toolpath and g-code.   Its not perfect and looking to improve tooplath quality as well as obtain a little more control over it.  I've dipped my toes into the water with Fusion 360 to check it out, but having a hard time getting anything close to what I need.  I've attached a video of the custom machine I built so you can see how it cuts the board while rotating it, largely keeping the tool tip perpendicular to the surface at all times.

I'm trying to replicate this type of cut program in 360 but no luck so far.  The best I could get is using the 3 and 5 axis parallel rotary method, but it makes some really non-intuitive moves on the surface (see attached screen grab) such as circles and loops.   I would really like to get a toolpath that generally follows the isolines of the surface, or at least stays roughly parallel to the long axis of the board for smooth cuts.

Any constructive advice greatly appreciated...

Mahalo

Chris

12 REPLIES 12
Message 2 of 13
a.laasW8M6T
in reply to: cgoody

you can do this with the Flow toolpath(it follows the isocurves) but there is a bit of a trick to make it work 4 axis which is a little hard to explain.

 

If you could share your file I can show how its done.

 

the reason Rotary parallel doesn't work great is that the tool it always pointing towards rotary centerline where as with Flow we can point towards the Surface normal(and give a lead angle if necessary)

Message 3 of 13
cgoody
in reply to: cgoody

Interesting, thanks for that.  I've attached the 360 file as a *.f3z archive -- just LMK if that format doesn't work, I'll try send it another way.

Mahalo

Message 4 of 13
a.laasW8M6T
in reply to: cgoody

Unfortunately when I try applying Flow to your part it keeps crashing Fusion on my end.

 

Here's a video and file on how to get flow 4 axis

Message 5 of 13
cgoody
in reply to: cgoody

That was really an excellent instructive video--I never would have figured that out.  mahalo nui for taking the time to make that, its much appreciated.  Think I almost have it figure out, but of course there's always something.  I could get it to work with the isolines in the wrong orientation, but not in the one I need (see screenshot).  NC kernel throws an error about no passes to link and empty toolpath (2nd screenshot).  Guessing it didn't like the v-direction.  Interesting how u-direction worked fine tho.  

Attaching the Flow setup dialog for geometry, passes, and multi-axis tabs as well--mind having a peak at those for anything glaringly wrong?  So close to success...

Message 6 of 13
a.laasW8M6T
in reply to: cgoody

Thinking about it a bit more,

Probably the reason it fails is due to the isocurves collapsing to a point in that direction

alaasW8M6T_0-1700273473542.png

 

The toolpath algorithm is probably having trouble with that.

 

All your settings look fine

 

Message 7 of 13
cgoody
in reply to: a.laasW8M6T

You were absolutely right -- after trimming both ends to remove the flow line  'singularity', it works perfectly.  Thanks again for the help.  

Separately, I noticed that when I use a bullnose tool, it creates a 'zig-zag' feature due to the curvature of the board and the shape of the tool (see first two screenshots).  Any idea for a way to suppress those types of moves (i.e., sudden reversals in direction)?  I'm cutting thru foam at a pretty good clip, and hitting one of those momentary reversals in direction along x-axis creates some unwanted shock as the momentum of the x-carriage wants to keep plowing forward.   The ballnose endmill has none of the issues noted above (third screenshot), but doesn't cut as nice as the bullnose with its nice flat bottom....

Mahalo

Message 8 of 13
a.laasW8M6T
in reply to: cgoody

The zig zag is caused by the contact point changing from heel to toe of the tool(in the cutting direction), it can be a problem with using Bull nose tools with multi axis.  because you are limited to 4 axis there really isn't anything you can do about this other than change to a Ballnose tool.

Message 9 of 13
a.laasW8M6T
in reply to: cgoody

You could try giving some lean angle so the contact point stays on the side of the tool, I'm not sure how well that will work out though.

alaasW8M6T_0-1700513912457.png

 

Message 10 of 13
cgoody
in reply to: a.laasW8M6T

Expert advice.  Played around with the lean angle values a little bit, and noticed that the 'heal-toe' zigzag goes away with a modest amount of lean.  Pictured here at 5 deg.

Just for my own clarification--the lean angle would be relative to the rotation axis yeah?  Is the value +/- or is it absolute to one side or the other depending on the sign (i.e., only -y or +y)?

And lead/lag would be about a perpendicular axis like Y yeah, which in my case is fixed b/c its only 4-axis..

mahalo nui!

 

Message 11 of 13
a.laasW8M6T
in reply to: cgoody

Hi

 

See here:https://help.autodesk.com/view/fusion360/ENU/?guid=Fusion_CAM_how_to_generate_multi_axis_machining_5... 

 

Basically Lead tilts the tool in the direction of travel, and lean tilts sideways relative to the direction of travel.

Because you are moving along your rotary axis lead will have no effect as the tool is locked at 90°

but lean is still able to be applied.

If you change the U/V direction to the other way so the tool is going around the rotary then you would use a lead angle rather than lean

 

You can give negative lead/lean values too depending on which way you want the tool to tilt

 

Message 12 of 13
cgoody
in reply to: cgoody

I noticed that with the retract heights for this setup (while using the tool-z-axis along the machine's rotary axis which in turn lies on the machine x-axis), it wants to put the retract heights along the tool-z and not the machine-z, as shown in the attached screenshots.  Do you happen to know any tricks to get it to put the safe-z and retract heights back along the machine-z?  mahalo nui

Message 13 of 13
a.laasW8M6T
in reply to: cgoody

No , unfortunately because this is kind of a work around(to get 4 axis motion) it does retract like that.

You can change the direction of the retracts along the X axis by flipping the tool orientation but that's it

 

Hopefully in the future There will be a "Proper" 4 axis mode which will include a cylindrical Clearance height mode, also probably a dropdown box to select 4 axis mode and which axis you want to rotate around, like what there is currently for the Deburr toolpath

alaasW8M6T_0-1701054322051.png

 

 

or geodesic

alaasW8M6T_1-1701054500218.png

 

You could play with Geodesic, It wont really follow the flowlines AND you have to use a Ball-nose tool but you could get a close approximation

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report