The year 1946 was a good time to start a practice in Denver; the old firms had been closed during the war and the architectural firms had just been restarted with a few commissions. I had no trouble finding work, and was hired to design a large rigid frame storage facility for the American Crystal Sugar Company to be built in Moorhead, Minnesota, and this job lasted for several months. I rented an office, and within three months, I had an employee. The previous engineers must have not had much imagination or technical ability to design concrete continuous beams for office or school buildings, because the architects I worked with seemed delighted with the solutions for unusual framing problems that I suggested.

We designed a number of interesting concrete buildings in the first years, including a hospital with floors cantilevered from the central corridor. We were not well acquainted with the importance of shrinkage deflection in that climate. For example, we inspected a floor with a cantilever span of 10 feet that had sagged from its own weight a distance of five inches. The top had been covered with in impervious material which did not dry, but the bottom was uncovered and so dried out and shrank to create this huge deflection. For the hospital, fortunately, the construction contract was in two parts, with a considerable time between the construction of the frame and the cladding so we did not have this problem. Another hospital with the same design but by a different engineer, was not so fortunate, and had to be propped at the end of the cantilevers.

I became interested in perfecting the details used in drawings for the firm, and proceeded to write a book using personnel in the office to prepare the final drawings. The book was published by Prentice Hall and was called Handbook of Standard Structural Details for Buildings. For many years it was the standard reference for this subject and only last year was out of print. At first I tried to show a number of examples, but I found that this was very unsatisfactory, so I proceeded to give some of the architectural details for the building. Then I could develop the structural details, and write coherently on the reasons for their use. All drawings were made on paper (Clearprint), in pencil. I designed the format and the page size. There were no figure numbers, all drawings were to scale, and adjacent to the text. This was one of the first books set in cold type, with the galley proofs looking like blue prints.

Another great interest in the first years of my practice, was the design of steel rigid frames for buildings. A number of school buildings were being built, and frames worked well for gymnasiums. One of the design problems was the proper support of the inner flange of the knee which is in compression and tends to buckle. One solution was to use a channel with the flanges turned toward the inside of the knee as shown in Fig. 2.1. Then a simple pipe strut or a truss could be used between frames. The vertical tapered sections of the knee were made from rolled sections, split diagonally down the middle, turned end for end and rewelded. Later, I obtained the commission to design a series of production rigid frame standard buildings, for the Stran-Steel Corporation of Detroit. They were to be made from plate rather than rolled sections. In this case, the inner flange of the web was continued to the top of the rafter, and struts between frames were not necessary.

At one time, wide flange steel beams were difficult to obtain, so we designed a frame with two spans of 70 and 50 feet using only 15 inch I beams as shown in Fig 2.3. The crowning achievement was the design of several gymnasium frames with spans of 120 feet having clerestorys for which the frame followed the shape of the clerestory as shown in Fig 2.4, rather than having a structure of small members erected on top of the normal straight rafter. The moments at the knee were less that for a straight frame because the effective virtual hinge, (the highest point on the thrust line), was higher than the normal frame.

For the University of Wyoming at Laramie, we designed an arch with a 221 foot span, set on A frames so the springing is 22.5 feet above the floor, rather than from the floor level, which takes much more space for the arches to go outside the building. This is shown in Fig. 2.5. The analysis used a simple method for the determination of nonlinear stresses. The increase in stress due to the deflection is determined from the initial deflection by a simple equation. I wrote an article for Civil Engineering Magazine which was reprinted by the AISC as a reference and used for many years.

Single rigid frames were analyzed for both wind and vertical load by graphical methods. For two hinged frames you used a temporary closing point to determine the moments as a simple beam, you calculated the deflection at the springing due to these forces, and for a unit force at the springing, you determined the horizontal reaction. Then you could find the magnitude and location all of the forces and from the graphical presentation. The abandonment of these techniques, particularly the concept of the thrust line, is a great loss to the profession because they give a much better picture of structural action.

Multispan frames were analyzed, of course, by moment distribution methods; a cumbersome method as compared with matrix analysis with computers. Although the speed and precision on analysis is now greater, we have ceased to design may of the interesting structures that were then being designed. This is true of both rigid frames and concrete shells. Work with these frames led me to develop methods of analysis and to write a book. It was submitted to McGraw Hill but was rejected. I was very downcast and ceased all work on this project. I know now that I was too hasty and it is normal to have this experience. I taught a class at the University of Colorado Extension, on structural design and number of students from the Bureau of Reclamation. The Bureau became interested in its publication for its staff so the work was not totally lost.