as a junior engineer before he had completed his B.S. degree from UC Berkeley. He later earned his master’s degree while working full time as a resident engineer at Caltrans. For several years he logged a full day at a job site in the Central Valley and then made the six-hour round trip to USC for evening classes. Even then he was always the first one to the office in the morning.
After accumulating a decade of field experience, Roberts was assigned to Caltrans’ Design Section 2 in 1962. His mentors in the bridge design section were engineers who would go on to be public works directors and chief executive officers of engineering firms. The early 1960s were a time of tremendous change in bridge engineering. Computational tools had recently become available. Roberts learned quickly, worked fast, and asked questions constantly. He soon not only mastered the conventional bridge design techniques of the time but also absorbed the importance of the new computer-aided wave of bridge design. His bridge projects grew in importance as his experience grew.
In the 1960s, most steel bridges were simple. Curved steel bridges used short spans with straight girders flared to accommodate the curves. The slide rule was used for most design calculations. The first large bridge design that Roberts was responsible for as a senior engineer was the Tuolumne River Bridge in California’s Tuolumne County. That bridge was substantially different from other steel bridges of the time. It was a six-span, 1,400-foot-long bridge on a 1,200-foot radius with a 350-foot main span and an 8 percent super-elevation. Time was of the essence, as the bridge would be constructed over a river that was soon to turn into Lake Don Pedro. Roberts’s efforts on this project exemplified the leadership and engineering skills that he demonstrated throughout his career.
The design standards of the time did not fully address many of the structural issues presented by this particular bridge. Research was under way nationally on curved steel girders with spans as long as 150 feet, but these lengths were substantially shorter than the planned spans on this bridge. Preliminary