
On an Autumn morning in my freshman year I walked into the Art Department at Duke University determined to combine Art and Computer Science into a program that would prepare me for a career in the emerging field of Computer Graphics. There I met a professor named Frank Smullin who had a great influence on me from then until his untimely death in 1983, just two short years later.
Duke at the time offered little in the way of a formal CG program. (I had heard anecdotally that the school had an arrangement with the University of North Carolina at Chapel Hill that Duke would specialize in AI and cede high level CG work to UNC.) Apparently I was at the wrong school, so I was particularly pleased to meet Frank that morning, huddled over what looked to me at the time to be amazing and exotic computer-generated images of a sculpture he was contemplating creating.
I knew I could learn a lot from this man, and I worked for Frank Smullin as much as I could in various capacities: sometimes in a work/study position for pay; sometimes as an independent study for college credit; sometimes just on my own time. I helped him write some of his software in BASIC on a Tektronix Vector Computer and I helped him assemble some of his small scale CAD/CAM sculpture prototypes using pink construction paper, cardboard tubes, Elmer's glue and an X-acto knife.

As an artist Frank had many styles. He was a strong draftsman capable of great realism in sculpture and drawing, but he often worked in a modern art idiom, using mitered cylinders. When he did this kind of work he called himself an Analytic Constructivist.
I once asked Frank why he liked cylinders so much. He explained to me that he liked the dual nature of the cylinder: when it was tall in comparison to its radius, the cylinder's resemblance to a one-dimensional line emerged; when it was short in comparison to its radius, its quality as a curved two-dimensional surface dominated.

Frank's sofware was, in the parlance of the day, a CAD/CAM system. It allowed him to engage in both computer-aided sculpture design (CAD) and computer-aided sculpture manufacture (CAM).
In its computer-aided design aspect, Frank's software worked as follows:
The software allowed Frank to maintain a list of vertices. It also contained a scheme for maintaining a vertex connectivity list, a means for representing those connections as mitered cylinders of a given radius, and the ability to draw the structure from any arbitrary angle.
Frank used a pen plotter to create his vector drawings, and (as raster graphics were beyond the scope of his hardware and software) he hand colored the drawings to give them an air of solidity.
When Frank was satisfied with his computer-aided design, he moved into the next phase: computer-aided manufacture.

When Frank was ready to prototype a design, he used a pen plotter to output a number of sine wave templates onto contruction paper.

The templates were usually scaled to be wrapped around cardboard paper towel tubes. Each sine wave, when wrapped, became a perfectly mitered cylinder. We cut these mitered cardboard cylinders with an X-acto knife and attached them together with Elmer's Glue.

The software could accomodate multiple cylinders meeting at a common point. Each line pair generated its own wave form. For a given cylinder all relevant wave forms would be superimposed, and the final template would be bounded by whichever portion of whichever wave was lowest.

The software could also handle off-center piercings of one cylinder through another through a routine I worked on with Frank which he called the Peanut generator. The programming of this routine took us some time back in the early 80's. I've simulated its functionality with relative ease today using Maya to planar-project an off-center circle onto an unwrappable file texture.

When Frank was ready to build a full-scale sculpture he would enlarge his templates, cut them into sheet metal, and have them rolled into tubes by a sheet metal fabricator. Frank did his own welding.

Frank had a wiry athleticism which made his sudden death (from a cerebral aneurism) in November of 1983 so unexpected. He was 40 years old -- two years younger than I am now.
Frank Smullin was the kind of person I greatly admire: part artist, part scientist, and full of curiosity. Sometimes when Novembers come around again I find myself thinking about him. The computer graphics industry has come so far in the last two decades. I'd like to see what he'd have been doing now.
Some contemporary artists of note:
Carlo Sequin (cites Frank Smullin as an influence)
UPDATE 2/2/2006
Here's a cardboard CAD/CAM project Paul Haeberli did in 1977

Hey, nice page, Joseph. Tetragranny is associated with some nice memories for me, as it stands on the grounds of my high school.
Yes, Frank Smullin was indeed an inspiration to me when I heard him give a talk at the Design Automation Conference in 1981.
It is wonderful, to have this review of Smullin's technique available on-line. I hope you keep this site active or allow me to copy it into my own domain.
I have recently given a talk on "Tangled Knots"
where I reference his intitial influence.
The paper is at:
http://www.cs.berkeley.edu/~sequin/PAPERS/ArtMath05_TangledKnots.pdf
The PowerPoint slides are at:
http://www.cs.berkeley.edu/~sequin/TALKS/AM05_TangledKnots.ppt
Carlo Sequin
Hi,
I went to computer camp at Duke (DUCK) in 1981 and 1982, and I'm pretty sure I visited the office/lab where you and/or Frank were working on this stuff!
At camp they took us to several offices/labs where interesting stuff was being done with computers.
It's been a long time, but I remember a small (dimly lit) office with what was probably that Tektronix vector display, and I remember seeing a design on the screen that looked like Tetragranny.
I later went to NCSSM, and was there in 1987 when Tetragranny was installed. I remember recognizing it as the thing I had seen at camp!
It's great to find your site and learn more about the origin of that sculpture!
Joe
Hi, I can�t understand how to add your site in my rss reader. Can you Help me, please :)
To be honest, I've never actually done that, but I assume the first step is to visit the small link labeled
* Subscribe to feed Subscribe to this blog's feed
Near the bottom of the right-hand column on this page. If you've already done that, and the process isn't working, then I'm not sure what to suggest next.