I am soooo disappointed!

[Burkhardt]

Modest progress on the box body. But at least I got the new lid inlay cut. I used a home-made fractal design to avoid the pesky copyright uncertainty and to find out how detailed such an inlay can be made.

The new piece of cherry was less brittle but I believe what made a major difference was to soak the inlay blank surface with ample thin CA glue (and a whiff of accelerator spray after that). It reduces the tendency to fuzz and stabilized the tiny protruding slivers somewhat. I also used the even sharper 1/8" version of the Kyocera 30-degree bit. Total machine time for both pieces was about 55 minutes.

[shilala]

That's gonna look great. Can't wait to see it!!!

[scottp55]

G. , While playing with buttons, I wound up with some lacquer that had been thinned about 80%(Yeah, what do I know), and to use it up- I coated the button blanks after surfacing and had way less fuzz and tearout.
So when I started Vinlays and Paul Z thought a thin CA coat before second pass(he recommends Always doing a second pass)(I just duplicate toolpath and do full depth and double or triple X,Y,Z MS) would stiffen wood. I TRIED and didn't like results, BUT next time tried the thinned lacquer-- Worked Great! Penetrates DEEP/Flashes in 10-15 minutes and doesn't appear to affect CA bond when gluing and stops CA penetration(darkening). A good swipe with solvent alcohol(been working oily woods anyway) and a 320G "scuff" on the flat works. Nice thing is it's so thin it doesn't "pool" in the bottom of the cuts noticably. Seems to stiffen fibers a LOT, and then a bamboo "Shish-Kebab" stick knocks everything loose without scratching anything like a dental pic sometimes does.
Just saying

[scottp55]

Oh. Just used unstirred Deft Brushing Lacquer poured off the top(no retardants) first time.
Last batch was Watco Clear Lacquer.
Heck of a lot cheaper than CA

[Burkhardt]

That's gonna look great. Can't wait to see it!!!

Here it is. Looks quite nice, and very meaningful to learn the limits of the technique. But these fractal patterns are actually not that good for inlays because by definition they lead to unlimited level of detail that can not be machined and gets washed out. It also looks kind of busy to me. But hey, I am not going to sell it.

[myxpykalix]

Actually G, i accomplishes it's purpose, at least in this photo. When i look at it starting at the top it looks to me as if the top is rounded and further away from the camera, as you scroll down the center is closest and as you go to the bottom the wood curves away from the camera and you are looking at the underside of a rounded object.

So the illusion works.....

[scottp55]

G., Came out way better than I thought it might!! BUT VERY useful for finding limits and improving technique! Any way to set on your Fractal program a limit for size, or would you have to group and "sweep" away the too small ones like I do for "noise" in imported bitmaps like the tree I just did(interesting that some "noise" I didn't zoom in enough to see still inlaid and looks like fly specks .
Think maybe denser wood or different speeds and you would have "Gotten" it!
Reminds me of some Escher's I tried way back in the beginning with the same problem with the "Infinity Fade"
Did you try the Vac Bag with "pressure plates?" (think I might know what you're talking about now". Epoxy? (would Vacuum Bagging need to be epoxy?).
Great Work!

[Burkhardt]

I did one first try with epoxy and that does work quite well but I found that I used a 30 degree bit while the path was calculated for a 60 degree bit. So that was not a good fit and I machined it off and started over. But for this (correctly machined) result I ran out of epoxy and tried the Titebond again. This time I diluted it with a good sip of water so that it becomes really liquid and can be brushed on thinly. Also that allows wiping excess glue from the surface and bigger pockets with a shop towel.

It would be nice to have an easy way of eliminating too small detail automatically. But I am only aware of the bitmap vectorization detail settings and the bezier curve conversion. Both don't do enough for that purpose (or may I am not using it right). Maybe I need to "despeckle" the image first in a photo editor before vectorization.