Transition to Tech Tools

This is the story of my difficult transition from sculpting with traditional tools and media (clay, pencils, carving wax, etc.) to using CAD software and other tech tools. Using computers still feels a little like "cheating", but the benefits of modern methods were too great for me to pass up!

woodpecker-keychain-05.jpg
barnOwl-pendant-06.jpg

Figure 1. Some hand-carved birds

If I were to draw a pie-chart of the waking hours of my life, one of the largest wedges would still be labeled "wax carving". Without an "undo" button, or the ability to create backups, each minute of these endless hours held the possibility of irretrievable loss. It is possible to repair a broken wax with a heat-pen, but there are limits to this ability... and there is no way to retrieve a model lost to an error in the casting process.

I was aware of computer design and 3d fabrication technologies, and it seemed obvious that these would soon allow people to make products like mine with less effort and greater flexibility. I was given a Vic-20 in 1980, and I was also exposed to a first gen. Mac, and I enjoyed using even the crude design tools available then, so I was primed to embrace new technologies.

Aside from making design easier, computers also allow one to do things that impossible or prohibitively difficult by traditional means. One can scale or mirror an object using a pantograph, but it is usually easier to start from scratch. I had been creating derived products (figure 2) manually, adding a company logo to a keychain or affixing shank to turn it into a drawer-pull, and I was excited by the possibilities of working in virtual space.

Freeform computer sculpting was available in the 90s, but it was beyond my financial reach, however when I began work on a submarine model (Figure 3, left), I thought I had the perfect candidate for a computer-generated object. Unfortunately, I had a tough time modeling the submarine using CAD, and ended up carving it by hand.

 

Two years after creating the sub, I was asked to make its mother ship, and this time, it was hand-carving that turned out to be too challenging. After a unit conversion issue that was proportionately as costly to me as the loss of Mars Climate Orbiter was to NASA, I got my first successful Machine generated piece (Figure 3, right).

owl-keychain-10.jpg
graduation-owl-keychain-01.jpg
graduation-owl-keychain-02.jpg
penguin-keychain-01.jpg
holiday-penguin-keychain-01.jpg
holiday-penguin-keychain-02.jpg

Figure 2. Originals from the hand-carved era (left), along with some "derived products" (center and right)

alvin-and-atlantis-02.jpg
atlantis-models-02.jpg

Figure 3. hand carved submarine (a), failed hand-carving of mother ship (b), faulty 3d-print (c), and lost-wax casting of final model (d)

After completing the ship, I continued to make most of my pieces by hand, while searching for software with which I could easily model animals and other organic shapes. About a decade later, in 2009,  I finally found software that worked for me, and made my first CAD modeled piece, an ammonite. Incidentally, this was a replacement for a hand-carved ammonite that I had lost to a molding accident two years earlier.

My earliest computer generated pieces were printed on a machine capable of high detail, but which used a fragile, wax-like material. Because of this, it was still necessary to go through the process of lost-wax casting before I had an original model from which to make a vulcanized rubber mold for pewter casting.

The next advance came when Steriolithography (SLA) printers became capable of suitable detail, and I discovered that their output was durable enough to create molds from. This advance cut several weeks out of the process, and eliminated inaccuracies introduced by the extra step (shrinkage, porosity, etc.), however another problem emerged.

Over the course of about six years, I would periodically get 3d prints which would develop cracks during the molding process (Figure 4). This costly problem occurred in items created with several resins, and on several makes of 3d printer, and there seemed to be no way to isolate the cause.

002-photosilver-casting.jpg
casting01_photosilver-02.jpg

Figure 4. castings made from 3d-print (showing pattern of cracks introduced in the vulcanized-rubber molding process).

sla-tesc-discs-10-02.jpg
sla-tesc-discs-02-01.jpg

Figure 5. 2 sets of 3d printed test discs made to diagnose the cause of persistent and unpredictable cracking. These were exposed to varying degrees of UV postcuring, IPA soaking, and water.

In spite of the crackle issue, CAD and 3d printing made things a lot easier, and opened up new design possibilities, and in 2014, when high resolution 3d printers became cheap enough for me to afford, things got even better! With my own printer, I could have a new design in hand​ at a tenth of the cost, and within hours rather than days. I was truly able to "prototype", and to handle physical objects before all design decisions had been made. It was another big step forward, however there was still the crackle...

Throughout my life, I've been steeped in the notion that the way to solve matters like the crackle mystery was to isolate the variables, and to test them carefully and without bias. With my own 3d printer, I could now do this, however the lessons of scientific objectivity are easier to preach than to practice.

At some point in the years of dealing with SLA crackle, I'd had it put into my mind that the cause was excessive soaking in isopropyl alcohol (IPA). IPA was used in all prints, and hence all crackly prints, and so it was easy to see it as the culprit... and aside from that, the cracks looked like just the sort of thing that might be caused by a volatile liquid seeping into the resin. I printed test cylinders (figure 5, left), exposed the subjects the post-processing variables (IPA, UV postcure, and water), I carefully noted the outcomes... and then ignored my own results!

In the first round of testing, I had numbered the cylinders after printing, and so I assumed that my results (which pointed to UV-postcure as the cause), were the result of experimental error. After another couple of months of dealing with crackle, and with gooey, poorly cleaned prints, I re-ran my experiment with pre-printed cylinders (Figure 5, right), and finally learned the truth. I gave away my UV lamp, and have not seen the crackle since.

And so, the process of creating new products has been radically simplified, and I have not even mentioned how the internet enables research (no more waiting for weeks for books on interlibrary loan, or traveling to museums to make sketches). We've arrived at a time in which a semi-skilled person with a CAD workstation, can create what once took teams of skilled artisans, and this trend seems bound to continue. I hope that technical tools won't result in creative atrophy for society and I hope it hasn't done so for me. Some of the products I am able to make now, involve features that I could never have realized in the past, such as precisely interlocking and articulating parts. I hope I've succeeded in making technology be the tool of progress and not just laziness.