projecting toolpaths onto 3D model

[paorawiremu]

Hi guys,
Success...I think? I did another engraving over the past few days, and I think the 3D model projection worked.
There are a few things that made the difference:
1.) Making sure my probe stylus ball was 1.5mm, not the default 2mm (this is not in Vectric software, but in CNCjs)
2.) Probing in smaller steps. I used 20mm steps and the probing time was 12 hours.
3.) Zeroing the probe and the cutting bit to the waste-board, and including the waste-board height in the point cloud map. I did this by manually amending the point cloud data to include an additional 20mm step on the right side, and then a drop-off down to zero. You can see this on the model. This means the Grid Import gadget generates a model the correct height from the machine bed.

The problem I did have was that when it started cutting the lines were too wide, ruining the details. I stopped it after a few hours and changed the flat depth from 2mm to 1.5mm, and restarted it, which didn't help much, so I lifted the spindle and additional 0.5mm. This was annoying because I wanted the lines to be 2mm deep, and ended up cutting them at 1mm. I suspect the 30degree bit I was using was not actually 30 degrees (something to test). Fortunately, even cutting at a depth of 1mm, I can see the machine followed the contour of the surface perfectly and didn't miss any sections.

[paorawiremu]

Just noting above, I called it an engraving, but it is a v-carve.
The cutting time was about 40 hours.

[LittleGreyMan]

Glad to see you sorted it out. Nice result.

I'm pretty sure Martin has the same question I still have: what is preventing you to vcarve before bending your composite, which is probably initially flat?

When you wrote:

it isn't perfectly flat

where you speaking of the ACM board before any operation?

[paorawiremu]

Ok, so, the bent ACM board was just for that test I was doing, so I had a taut, curved surface.

These actual works are on ACM board, screwed to a wooden frame, stretched with canvas (folded and stapled on the back), with about four-six litres of paint layered on it, which is the part I am engraving into. The paint doesn't like to be surfaced because it is too tacky.

It's a bit confusing to explain, so I was leaving those details out.

[paorawiremu]

Here is the side view. I built up the paint as evenly as I can, but on the larger works I can see from the point cloud some of them have a surface deviation of up to 2.5mm.

[LittleGreyMan]

Thanks for these explanations.

[adze_cnc]

One niggling question re: settings above.

If the model is 27.707mm as created via the probe and gadget why set the material thickness to be 30mm with a gap of 2.793mm above the mode? Why not tell the software the material thickness is the measured model thickness?

[paorawiremu]

You are absolutely right! I've changed this for the one in the machine at the moment.

[martin54]

How are you going to set your z zero position for the Z axis?
Just a thought thinking about adze_cnc's question about why set the material height at 30mm when the actual material height is less.
You are now probing outside of the material down to the spoilboard. Could you include a small piece of material of a known height (say 30mm) at one corner of the material, say the bottom right of your material piece. This would then give you an overall model height of 30mm plus a flat surface to set your z zero position for your tooling.
You are probing that area for your point cloud anyway so it won't add any time to the probe routine

[paorawiremu]

great ideas! Thanks!