I was wondering what the opinions were on the subject of tool path techniques for 3D carving.
In terms of end result, time/effort to generate paths.
I have been using Fusion 3D Adaptive to try to carve a Eagle relief.
I am using Red Oak.
The mesh model was from a Lynx scanner and JMStudio application.
I short am still working though the process, but at this point I am generating the tool paths from a lightly cleaned ( reduce noise .002 ) mesh in Fusion.
On my second attempt I completed a part with 1/4 end mil and 1/4 ball mill. I had hurt the mesh model by heavy handed clean up so 1/4 ball mill was as good as needed. I got a temp license for the Fusion design extension so I could use the Convert command to create a Tspline surface model.
I created the Tool path from the surface model.
In Fusion , the tool path generation was a long experience, but worth it in education.
I had to segment the model due to multiple “unexpected results” reports. I basically divide and conquer.
The end result surface finish was beautiful, no touch required ( except for the effects of my own making).
While doing this, I got a temp license for CC pro which uses a parallel technique. The process to make the tool path in CC pro is very simple.
It generated the paths in a few minutes. What ???
I did not cut this path. I see that it is recommended to run a second finish path at the different angle to clean up.
I am on my third try in Fusion having much better tool path generation stability. The generation time will take about 16 hrs on my Dell G5 5500 laptop.
If you’re doing low-relief 3D work, which is much of the woodworking people do on a Shapeoko, you generally want a parallel finish. It will give a consistent, predictable finish.
Adaptive toolpaths try to keep the load on the tool constant, which is more important when removing large amounts of material. To maintain a constant load, the toolpath is going to frequently change direction, which is not generally what you want for a finish toolpath.
Regarding calculation time, adaptive toolpaths effectively have to run a simulation of the cut while calculating the cut, which makes it take a lot longer.
Comment on the red oak, I’d recommend trying white oak. White oak is a bit harder than red oak and not “stringy”. It will hold more detail than red oak and look sharper. In fact I just squared up some sides on a piece and cut my finger on the 90° edge! You can approximate the color of the red oak with whatever you use for a finish.
Early on in my woodworking career I used red oak because I liked the color plus it was cheaper. But the floor installer recommended white oak, I researched it and agreed. Since then I’ve built several large furniture pieces of white oak finished in the Stickley manner, or imitation thereof.
I have a stash of red oak , hence the reason for using it. I usually end up using whatever comes through my buddies place ( lot clearing and firewood). I get paid in wood.
I have noticed that the white oak must have long string, because splitting it is not fun.
That’s what the British navy cleared from eastern CT for their masts.
The Adaptive path allows for Conventional or Climb milling. Unless I select “Both ways” it cuts in the same direction ( and takes more time)
In Fusion 360 for the handful of 3D carves that I’ve done, I used an adaptive toolpath with a 1/4" endmill and then, if needed, a second adaptive toolpath with an 1/8" endmill to clear out any small areas that had excessive remaining stock finally followed up by a parallel toolpath with an 1/8" endmill.
If the 1/4" is small enough for your carves, I would just do the adaptive toolpath as you were but then use a parallel toolpath for the ball mill.