My Third Project

I’m posting my experiences here in the hope that they’re useful for other people. If they are (or become) dull, useless, or annoying, tell me and I’ll stop.

So this project was to try to get up to speed with cutting real wood. I have a small bunch of recycled wood from wine casks - it’s supposed to be oak, but I don’t know for a fact. It’s stained red, and has tartrate crystals that I have to wash out.

I recently designed a butterfly pendant for my wife, and had it printed in gold-plated brass at Shapeways. I’m reasonably pleased with the results, and thought it might be an interesting experiment to try to cut the same model from the wood.

Metal results:

My first cutting attempt failed because the carpet tape didn’t hold and the wood slithered around. I also noticed that the outline was cutting outside of what I thought would be the work area. I was using the exact corner of the stock as my zero point, and hadn’t paid attention to the “Machining Margin” in the MeshCAM global parameters.

I contemplated changing the feed rates to try to prevent the wood from moving around, but instead followed the advice given by RoughVegabound in the thread here about vises. I sanded the wood down, and took extra care in placing the tape.

I also took special care to consider the machining margin when zeroing the Nomad this time.

MeshCAM estimated this design would cut in 28 minutes (noting that it doesn’t include time for rapid moves) with the 1/16th inch ball cutter. I started the run, and after 58 minutes, the job was complete:

A shot of the vacuum, and my design was revealed:

I hadn’t set the thickness of the stock to cut all the way through. This was intentional, as I suspected some of the details (such as the antennae) were too fine to hold together.

Here are the lessons learned from this project:

  • Gotta pay attention when securing the work.
  • The zero point in MeshCAM is the same as the zero point when jogging the Nomad, but the tool path may extend “behind” it. Pay attention to the margin.
  • Verify stock size! MeshCAM didn’t do a complete boundary cut on this design because the margin exceeded the stock size (see the extreme edge of the front wings) which it had computed based on the model size.
  • The time for rapid moves in a job may be substantial.
  • When there are details that are too small for the selected cutter, it may not be obvious that they won’t be cut from the view in MeshCAM. I looked at the toolpaths, and saw plunges at many of the smaller spots around the wings, but didn’t realize that those were just surface detail.
  • My recycled hardwood is too brittle for certain details. Some of the thinner areas broke away during machining. I should reserve this wood for larger designs.

1 Like

Don’t stop posting- we’re enjoying them.

Apollo tried a bracelet using some thin brass and some of our fixturing wax.

In addition to unaccounted rapids, the time estimate doesn’t include acceleration going around curves or up and down, which is calculated real-time on the Nomad.


Is your filtering was different than what you use with a hot glue gun? Where do you get the filtering wax?

Great post, thanks for sharing your experience and tips.

What other things have you made on the 883 and what are the key lessons learned?