Given the recent, unexpected passing of John “Chip” Dean, I thought it only right to begin this new series on Truss Design Technology with reflections I collected from him. Three years ago, knowing Chip’s background as an iconic truss designer and pioneer in our industry, I asked him to recall his experiences, starting with his first job as a truss designer at Truss-Com in Sacramento, CA in 1978.
The recollections relayed in Chip’s notes reveal the manual steps he and his coworkers went through to produce a sealed drawing, and they paint a vivid picture of the beginnings of our business, and of an inspired learner who started from scratch.
Chip learned drafting working for a printing company, but after being laid off, found a new position for “a lumber company that also made wooden roof trusses, whatever the heck they were.” Like most 1960s truss plants, Truss-Com grew out of a lumber company, Steiner Lumber, when its owner, George Steiner, and his lumber salesman and World War II hero, Jim Dowd, recognized the growth potential of trusses. By the time Chip arrived, they had become one of Gang-Nail’s major customers, expanding to three locations. A key driver of their success was their recruitment of Mike Reeder five years prior, formerly Gang-Nail’s Chief Engineer in Miami, FL. Ironically, Mike moved to St. Louis, MO just prior to Chip’s arrival where he became my boss and VP of Engineering for Hydro-Air. Per Chip’s notes, “Mike’s meticulous nature resulted in beautiful, handwritten, skill sheets that made calculating loads like following a cooking recipe. Even though Mike was gone when I came to Truss-Com, everyone continued to use his worksheets every day.” Mike Reeder’s work, that Truss-Com’s competitors without engineers couldn’t match, led to their dominance in converting modernist-style, stick-framed roofs to trusses, as shown in the tail-bearing drawn by Chip. Per Chip’s notes: [For photo, See PDF or View in Full Issue.]
[My first job] was to create a truss layout. This was typically drawn to 1/8” scale and drawn on “D” size paper, (22” x 34”). All the trusses were drawn with a single line representing the centerline of the truss. Each truss was given a label and was shown on the drawing. Dimensions were added with special notes and other details as needed. Once the profiles were developed, then the truss design process began. Web configurations were based on maximum panel lengths and other physical conditions that may affect panel point locations. The next step was to fill out a computer input form so the truss could be analyzed by the computer.
Then there was the dreaded computer input process, which was easy for some, and painfully difficult for others. The input form was terse and required precision as all data was typed on lines with no spaces between the data fields. It was mind numbing. We had a loud Teletype machine in our room, and the truss designer would enter the input into the Gang-Nail timesharing system over a painfully slow 300 baud modem. After about 30 minutes, they would log back into the system and retrieve the output. If there was a single typo, or one missed space, the truss would not run and create an “exchange package” which resulted in pages and pages of random ASCII characters printed on expensive thermal printer paper. Once we got past that and had a successful computer run, the printout provided us lumber size, lumber grade and species, member forces, deflection, bearing reactions, and camber requirements, but nothing to do with connector plate sizes.
Once the output from the computer was downloaded and printed, the next step was drafting, creating a scale drawing of the truss with 2” scaled joint details for plating. The (plating) process involved writing the member forces from the computer output to the sketch noting whether the force is a compression joint or a tension joint. The designer would then take their 8 ½” x 11” mylar plating template which included each of the plates’ widths drawn to 2” scale with “x’s” denoting every tooth of the plate. The designer would determine the nails needed for each member of the joint. They would put the template on the 2” joint details and counted the nails until the joint was covered adequately. In most cases, we plated the joints symmetrically to make it easier for the shop.
At that point, the individual truss design was ready for checking and sealing. Truss-Com was ahead of its time because they had a professional engineer on staff, so our engineering turnaround could be as fast as required.
I absolutely loved my job. I found the truss industry truly fascinating. All the people I worked with were great and I loved the fact that we were designing and building structural solutions. The list of things I needed to learn seemed endless, but I felt confident that if I could master this job, it could become my life-long career. In fact, during my job interview the professional engineer, who became my boss, told me, “If you learn truss design, you will never have to worry about finding a job for the rest of your life.”
In 1979, Chip got to work with one of the first in-house computers, as Truss-Com purchased the HP 1000 E-Series minicomputer loaded with Gang Nail’s AutoTruss software, which ignited Chip’s love of computers. In 1980, when Chip moved to Seattle, WA to be close to his dad, he had the opportunity to work on a Tandy/Radio Shack TRS-80 computer, which opened the door for him to write a production scheduling program when he went to work for nearby Tilton Truss. In 1983, Tilton installed a Cromemco computer loaded with On-Line Data’s cutting program, giving Chip yet another chance to bolster his design skills, and he also developed a rapport with Gang-Nail’s engineers, as they reviewed Tilton’s drawings, which led to his next job as a Gang-Nail Tech Rep in the Northeast.
I first met Chip when he called on my plant, American Building Components, in Virginia in 1988. I was instantly impressed with his infectious enthusiasm and incredible knowledge, which I experienced throughout our association. He stands as an inspiring example for every truss designer and will be missed.