Worst clever idea I've had all week

Things You've Made
1

I was noticing that the wood grain in some 2x4’s has a pretty strong contrast between light and dark areas. It got me to thinking, could I use the natural rings as a colorant for a sign. This would mean the foreground would be cut to just the right depth to expose a dark ring, and the background would be cut slightly deeper to expose a light ring. The fun part is that the rings are irregular and buried within the wood so you can’t see where they are. You have to infer.

This is a terrible idea.

So I first surfaced the board with a 1" surfacing bit and then took pictures of the end and the surface. In inkscape I tried to fit circles to the rings and recorded their coordinates. Then on the surface I measured the width and position of the exposed part of one of the rings. A crude approximation would be if the same ring pattern passed through the wood in the X direction without changing shape but perhaps translating somewhat in the Y and Z directions.

end_grain_fit
end_grain_fit.png1195×539 749 KB
surface_grain_fit
surface_grain_fit.png987×597 693 KB

Then in OpenSCAD I created a 3D model of the internal ring which I can’t see, and projected my 2D pattern onto it.

grain-following-model
grain-following-model.png765×620 18.7 KB

After some CAM it was ready to go. It’s a bit scary because I have only one shot at this. No second chances without measuring a new piece of wood. And the end result was…

FAIL.

I missed the Z depth by enough that the foreground and background do not really stand out, and at the ends you can see the error where the pocket does not line up with the rings.

20200105_035442
20200105_035442.jpg4032×2268 2.28 MB

20200105_035245
20200105_035245.jpg4032×2268 2.14 MB

I think if you really wanted to use this technique it would be possible with a more elaborate process:

  1. Approximately model the 3D grain pattern
  2. Cut a ‘waffle’ pattern, perhaps sine waves or triangle waves, just above the ring you are targeting
  3. Use the grain appearance in the waffle pattern to refine the 3D model of the ring (and do not allow steppers to disengage)
  4. Model and cut your desired pattern using the refined, hopefully very precise, model of the grain pattern

It seems the darkest part of the ring is right at the boundary where it transitions from light to dark, and then it gradually gets lighter from there. It might be good to carve just above that dark transition and then sand down until the dark area is exposed, for the maximum contrast. For the light-colored recessed areas, you would have to get the depth right but it’s less sensitive to hitting the precise depth exactly.

Ultimately a fail, at least so far, but I thought you guys might enjoy. This is just too amusing not to share.

2 Likes
2

Hmm, The O side is pretty close, trying to catch the exact spot on the rings is a pretty tough target to hit.

Me…hardware guy, would just shim up each side as needed to try and catch a ring with a finishing pass. I see how that might not be as fun as software for you though.

If you could pull it off though that would be insane!

3

Well, technically Columbus failed to get to China, but he managed a pretty good haul never the less.

Same here as far as I am concerned. I haven’t done any curved surface milling like this but seeing something like this keeps me interested.

4

The problem is, wood is too irregular. The grains don’t take a straight path through the board.

Pine is also a very fast growing wood, so it tends to bend and rotate as the tree grows.

I wonder if zebra wood or one of the varigated hard woods would be more consistent.

5

I think you almost nailed it. Just a tad deeper. You might be abke to sand right down to it. The right side would be getting darker, but you were just a tad shallow on the left. Also, put some laquer or poly on it, the dark will get darker.

Really interesting. Thanks for sharing.