The design of the 1950 Tacoma Narrows Bridge began shortly after the 1940 collapse of its predecessor "Tacoma Narrows Bridge (1940)". In July 1941, the Washington Toll Bridge Authority (WTBA) appointed Charles E. Andrew[3] (who had been involved in the design and construction of Gertie, the previous bridge, as a consultant) as chief engineer and chairman of the consulting board in charge of the design of the new bridge. Members of the design board included Theodore von Kármán, Glenn Woodruff, and the firm of Sverdrup and Parcel of Chicago, Illinois. To lead the design team, Andrew selected Dexter R. Smith as lead designer and architect. As early as October 1941, less than a year after the Gertie collapse, the WTBA had devised the initial configuration of the structure, very similar to the original 1940 design drawn up by Clark Eldridge. The construction cost was estimated at 7 million dollars (about 152 million in today's terms).
As the original bridge became a major asset during the short period it was in service, the Navy pushed hard for the adoption of a combined highway/rail bridge to replace Gertie, and proposed a steel cantilever bridge in place of the suspension bridge. However, the additional steel required to build such a structure would have increased the construction cost by an additional $8.5 million (about $185 million today), ruling out any possibility of such a structure ever being built.
Furthermore, the proposed design required new laboratory tests. A purely mathematical solution to designing the suspension bridge was not possible, because very little was known about the forces that brought down the first structure. In light of this fact, the engineers chose to build scale models of the design and test them in a wind tunnel specially built for this purpose at the University of Washington. According to Charles Andrew, "the only way to attack the problem was to design a bridge, then build a model of that design and subject it to the action of the wind." The test was carried out by Professor F. B. Farquharson,[4] who had investigated the movements of Gertie before her collapse on November 7, 1940.
From late 1941 onwards, Professor Farquharson (and also von Kármán, who did his work at the California Institute of Technology in Pasadena "Pasadena (California)") continued to advance the design of the new bridge. In 1943, work was underway in a laboratory with a wind tunnel specially conceived for this project, built on the campus of the University of Washington in Seattle. The facility was large enough to house a scale model of the entire bridge up to 30.5 m, as well as section models for various tests. After Farquharson confirmed that Gertie collapsed due to excessive board flexibility and aerodynamic forces, tests were performed on designs projected by Smith. All the new designs had wind-permeable decks composed of a highly rigid and deep "Armor (structure)" lattice, instead of the box girder of metal plates used in the 1940 bridge that was brought down by the wind.
Tests on the design of the new bridge began in November 1943 and continued until 1945. The studies included 200 different configurations, so that the wind forces affected the bridge models at more or less 45 degrees with respect to the direction perpendicular to the deck. Subsequently, testing was performed on a design with wind-open louvers arranged in the roadway, which added even greater stability against torsional movement. A design with a lower lateral reinforcement in the stiffening truss was also tested to test resistance against lateral movement. Additionally, a design was tested with motion damping devices placed on the platform in three locations: one on each tower (at each end of the main span and each side span of the tower) and a set of mid-span damping devices on each main cable. Each of these steps in the design and testing phase was performed to reduce as much lateral and torsional motion as possible.
After dedicating $80,000 ($1,353,937 in today's terms) to the design and testing of the new bridge, the project was completed on December 5, 1945. The WTBA finalized and approved the revised designs of Dexter's plans (submitted in December) in April 1946, and minor revisions continued through September. The new structure had a construction cost of $8.5 million (US$144 million in today's terms).
The final Tacoma Narrows Bridge designs, once completed, presented a drastic contrast to Leon Moisseiff's project. Instead of a deck formed by a thin box girder, a reinforced lattice beam permeable to the passage of air, with a depth of 10 meters, was installed. The new towers would be 7.7m taller and 6.4m wider than Gertie's original towers. The bridge's main cables went from 440 mm in diameter to 510 mm, and the anchor blocks would accept a load 1.6 times greater than that planned for the original bridge. However, some elements of the Galloping Gertie were incorporated into the 1950 bridge. The pedestals of the towers were enlarged and increased by 5.2 m. At the west end, a 140 m section of the approach viaduct was maintained with the same 2.4 m deep box girders as the Gertie main deck. This approach viaduct reused three support towers, two of them based on the original light framework and the third with the structural complexity and design of one of the main towers of Gertie, with spans of 46 m. This span of viaduct, after structural examination, was retained and used as part of the 1950 bridge design, with additional stiffening reinforcement added to the tower closest to the shoreline (officially known as Tower #3 on the design drawings), and the widening of the top of the pier to accommodate the widened deck of the new bridge.
The deck deck itself was an important innovation in suspension bridge design. Traffic lanes on typical suspension bridge roads are separated by broken paint lines, a solid stripe, or a set of two paint strips. In the final design adopted in 1950, the 14 m wide roadway was divided into four traffic lanes, each 2.9 m wide. Each lane was separated from the next by a grid 840 mm wide and 76 mm deep to allow the passage of the wind. Bordering the outside of each road, two other 480 mm wide grids were arranged, forming sidewalks 130 mm high, and a third central sidewalk 0.91 m wide, also formed by a grid, served to separate the two roads. The section was completed by separate 1.2 m high railings.