Jun 222014

This post is part of a meta-series. Click here for a list of all posts in this series.

A friend of mine is moving out of state and needed to downsize the amount of stuff she had to move. Among the assets in question was a CNC router…which she sold to me!

Instead of manually cutting the cross-sections, I now can actually send the true cross-sections to the CNC machine and “print” out slices that I would have sanded into shape by hand previously.

Here’s the result of the first test run!

I realized that I had saved everything with non-linear colorspace, and when you’re doing stuff with heightmaps, non-linear colorspace screws up your curves and so on! Based on what I learned from this test, I’m now preparing to do a test with slightly adjusted positioning for the cross-sections…and proper linear color space data!

Nov 072013

This post is part of a meta-series. Click here for a list of all posts in this series.

I’ve been quiet on this front of late, but not idle. When we last left off, I had nearly finished gluing the cross-sections into place. Once finished, my concern about the main profile board proved well-founded, with the board making a gentle but noticeable arc from front to back. This meant the centerline of the entire helmet would be incorrect once finished.  However, I noticed that I could manhandle it into correct alignment. I hatched a scheme to create a platform for the helmet into which I would drill regular holes for dowels that would enforce the spacing between each profile. After doing just the center two and two toward the rear (around cross-section 8), I realized that the dowels just weren’t rigid enough for the idea to work. They bent too easily, meaning the heavy mass of cardboard was better at shifting their alignment than they were at keeping it aligned. I ultimately went with a simpler approach and tried to fix each of the cross-sections in place by anchoring them to other cross-sections with masking tape. It mostly worked.

Next came the insulation foam.

Continue reading »

Sep 192013

This post is part of a meta-series. Click here for a list of all posts in this series.

As of September 10, I finally had all of the cross-sections cut!


I’d been working in the basement, since the stairs were finally done, but all of the pieces felt a bit damp (not severely so, just…not as firm as one might expect).  I brought everything upstairs to dry out for the night before I started gluing the cross-sections to the profile. I’ve been doing that two-at-a-time, more than a little frustrated by how “slow” Elmer’s is to bond. Really, though, that’s just impatience talking.

As I’ve been going, I noticed that the thin cardboard of the main profile is actually bending a fair amount, which I’m going to have to figure out how to correct before I start filling in the foam. Would be a shame to go to all this trouble to achieve a fair amount of symmetry only to have my entire axis be wobbly!

I also had a minor crisis last night, when I realized that I had mislabeled cross-sections 25 and 27 (part of the “face”). I realized this when I went to put on cross-section 26 (i.e. after 25 had set) and it was larger than its predecessor. I checked my Blender file and all was well, but the profile I had in Blender did not at all match the profile I expected to see for 25, based on the printout template and the cross-section itself. Sure enough, I had mislabeled it (and 27) in Photoshop! Cue emergency surgery to slice that cross-section off of the main profile and replace it with 27.

I just attached the “true” 27 a few moments ago. Here’s how it’s shaping up:

(Bonus: In the background on the right, you can see the two toaster ovens that I’m going to use to form the plastic-warming oven for my vacuform table!)