Skip Navigation

Our Article Library

Back to Library

Truss Design Technology, Part II: Designing Before Computers

Issue #18327 - October 2026 | Page #10

By Joe Kannapell

This month’s article returns to a topic begun last month in “Inside Early Truss Design”: the movement of design technology from the drafting board to the computer. During this period, the only tools available were engineering theory and painstaking manual work. Specialized methods were developed for premade wood trusses in the early 1950s, by collaboration between Cal Jureit, a professional engineer at Wingerter Laboratories, and Miami building officials. When, in 1960, Jureit was elected President of the newly formed Truss Plate Institute (TPI), his methodology was incorporated into the first industry design criteria. The resulting standard, Design Specification for Light Metal Plate Connected Timber Trusses, TPI-60, included the requirement that truss designs be supported with both drawings and computations.

To apply this technology in the pre-computer days, a draftsman did the drawing and an engineer did the calculations, with both tasks consuming several hours of labor to produce a single drawing. The only physical tools used in the preparation were cumbersome mechanical calculators and imprecise blueprint machines. [For images, See PDF or View in Full Issue.]

Initially, only the truss plants with PEs on staff could provide such drawings. Those without engineers had to communicate truss details verbally with plate suppliers or outside engineers, a process fraught with errors. To lessen engineering requests, Sanford Truss and Timber Engineering Company (TECO) pioneered the practice of illustrating a series of designs on a single drawing, sometimes called series engineering. For example, the typical Florida design shown here covered a series of residential spans from 16’-0” to 32’-0” for a TECO nail-on connector that satisfied all TPI-60 requirements. An efficient design was rendered by varying plate and lumber designs as a function of span. For example, a 32’ truss required five nails on each top chord and six nails on each web, for a total of 22 nails on each face of the truss and a total of 44 nails at the peak.

This drawing was part of a 12-page manual from TECO that included multiple roof pitches. TECO wisely distributed its manual to thousands of lumberyards free of charge to promote its nail-on plates. Likewise, Carol Sanford amassed an even larger number of engineering manuals that covered three different plate types, East and West Coast loadings, and a wide variety of truss configurations.

By 1960, the production process was greatly simplified by the automated Xerox 1824 printer technology capable of coping 18” x 24” drawings. This innovation sped up the conversion from cumbersome blueprints to plain paper, enabling text to be typed instead of hand lettered. By 1970, clerical staffs typed individual sections of drawings including single lines of text, such as “#1 Southern Pine or Fir” that draftsmen would painstakingly affix to manually drafted truss drawings in a cut-and-paste process.

Later generations of Xerox machines had the capability to obscure the borders of individual pieces of paper, called cut-lines, producing a clean, homogeneous appearance. The advantage of using typed text was that a large quantity of critical engineering details and notations could be appended to appropriate sections of hand-drawn graphics. For example, the critical notation, “1 x 4 lateral bracing,” could be inserted adjacent to compression webs. Moreover, this information would retain its legibility even when copied multiple times for submittal to building departments or for field use.

In 1969, the large, but highly productive, Xerox 7000 could reduce the size of 18” x 24” drawings to a standard letter-size format that could be subsequently hole-punched and included in standard binders. In 1971, I noticed a large clerical staff filling binders at Hydro-Air Engineering, especially for customers like Material Fabrication (later Shelter Systems) who faced a wide array of code jurisdictions.

Also at this time, I programmed an IBM mainframe computer to print much of the text for series drawings, along with a CalComp pen plotter to print graphic details, thereby eliminating much of the drafting and clerical work, but still requiring paste-up of series drawings.

By the early 1970s, the truss business benefitted from a wave of innovations that would only continue at an even faster pace, as fabricators got their hands on a computer for the first time!

You're reading an article from the October 2026 issue.

Search By Keyword

Issues

Book icon The Monthly Publication for the SBC and Off-Site Construction Industry