By Aaron Mazeika, PE, SE
In 1968, as construction workers hoisted steel beams more than 1,100 feet above Chicago’s lakeshore, the John Hancock Center reached its full height—an achievement that would forever change not only Chicago’s skyline, but the entire field of tall building design and engineering. Hancock’s “braced tube” structural system, devised by SOM engineer Fazlur Khan together with architect Bruce Graham, introduced a new era of efficient skyscraper design. However, the foundations of that design innovation were laid even earlier, and today, more than 50 years later, designers at SOM are continuing to improve upon the system.
The evolution of the braced tube structural system began with the initial conception of the Tube Frame structural system implemented by Khan at the Chestnut-Dewitt apartment building in Chicago (completed in 1966). Up until that point, rigid frame structures had typically been conceived as an array of two-dimensional frames passing through the building in two orthogonal plan directions, and along each column line. The stiffness of the frames was limited by practical considerations—how closely columns could be spaced without compromising the functionality of the interior space. By using closely-spaced columns around the building perimeter, Khan discovered that the structure could perform like a thin-walled tube cantilevering from the ground; densifying the columns at the perimeter could be accommodated without impacting the building function, while enhancing the stiffness of the perimeter tube. This could allow the interior columns to resist only gravity loads, and therefore be spaced further apart, resulting in more functional and flexible interior spaces. In part, this structural innovation was achieved by conceiving of the structure as functioning three-dimensionally, with the tube acting as a 3D box section rather than an array of 2D frames. Designed at the advent of computer-based structural engineering analysis, this behavior was confirmed by performing more complex analyses than previously possible through hand calculations.
With the development of the braced tube system for the John Hancock Center in Chicago (now known as 875 North Michigan Avenue), Khan rationalized that an even more efficient structural system for a tall building could be achieved by replacing the perimeter frame tube (as used at Chestnut-Dewitt) with a diagonalized steel braced truss on the perimeter faces of the building. Whereas the frame-tube resisted lateral deformations through flexural and shear strains in the spandrels and perimeter columns, the braced tube resisted lateral building deformations through axial strains in the truss members. In the same way that a diagonalized truss is more efficient than a Vierendeel truss in spanning and bridging applications, the use of a diagonalized truss on the perimeter of the John Hancock Center established a new benchmark for the minimum structural material usage for a tower of this height. While the innovation of employing a braced tube for the lateral system of the John Hancock Center resulted in a step change in design efficiency, the precise geometry of the bracing employed was not the most efficient possible, as subsequent research into optimal truss geometries has demonstrated.
Starting around 2008—approximately 40 years after the design and construction of the John Hancock Center—SOM partner and structural engineer William F. Baker led a research team within SOM to explore optimal truss geometries. Building on studies on continuum optimum frames by Michell and discrete optimum trusses by Prager, the team applied this knowledge to increase the efficiency of SOM’s tall building designs. One finding was that the X-brace configuration of the John Hancock Center facade geometry could be improved by moving the central intersection node upwards, at the three-quarters height of the braced bay. This would be the optimal point when stiffness controlled the design, and buckling of the individual compression members was not a factor. This optimal point could be lower if compression buckling of the individual members became a controlling factor. This simple adjustment to the geometry of the braced bay offers an approximate 10 percent increase in the material efficiency of the brace. It is interesting to think that the iconic X-brace geometry of the John Hancock Center would likely have been different had the design team had knowledge of discrete optimal truss geometries at that time.
Following the research team’s findings, SOM’s design and engineering team quickly found opportunities to implement new knowledge about optimal truss geometries in the firm’s active design work. The first constructed example was at 100 Mount Street, a 35-story office tower in North Sydney, Australia, completed in 2019. This reinforced concrete linear bar building is designed to maximize daylight and open space for tenants, with a high-performance closed cavity facade and the core offset to one end of the floorplate. To prevent a torsional response to lateral wind and seismic loading, a stiff, lateral force-resisting element was required at the opposite end of the building, where a primarily transparent facade was required. The optimal discrete truss geometries developed by SOM’s research team provided the most effective and material efficient solution.
While our knowledge of optimal truss geometries has advanced significantly, changes in the construction industry have presented new barriers to the implementation of trussed tube buildings. Most notable is the transition from steel to reinforced concrete as the preferred construction material for tall building structures. When the John Hancock Center was built, all-steel construction was the method of choice for very tall building structures, but in the 1980s, as high-strength concrete technology improved, the economics of tall building construction shifted to using reinforced concrete where possible. The shift from all-steel is significant, because even though steel is ideal for the brace elements due to its isotropic characteristic (meaning they can be subjected to both tension and compression forces in equal magnitude), steel braces are difficult to incorporate into a reinforced concrete building due to an incompatibility between the long- term behavior of steel and concrete. Concrete columns tend to shorten over time when subjected to sustained gravity load, due to a phenomenon called creep, whereas the creep in the steel braces is insignificant in comparison. Typically, the braces’ configuration allows them to act as an alternate gravity load path to the columns. As the concrete columns creep, gravity loads gradually transfer from the columns to the braces, increasing the forces in the braces over time and requiring them to be larger, and therefore much less materially efficient.
For 100 Mount Street, SOM’s design and engineering team solved this problem by incorporating a special sliding detail at the central node of the brace. The design goal was to isolate the lateral and gravity load paths. If all the gravity load could be kept within the reinforced concrete columns, then the braces would experience only equal and opposite forces due to lateral wind and seismic loads. The bracing system would then be a symmetrical system subject to perfectly antisymmetric loading. This condition simplifies the load transfer at the central node such that only vertical loads need to be transferred from one side of the central node to the other. A sliding detail consisting of a series of interlocking horizontal plates enables this load transfer, but also allows a “release” that isolates the bracing system from gravity loads. When the columns shorten due to creep, the two halves of the central node simply slide slightly towards each other, allowing the columns to shorten without load being transferred to the braces. In a similar way, thermal strains in the braces can be released without inducing axial loads in the bracing.
While the sliding detail at 100 Mount Street was successful and continues to perform as designed, it was complex to fabricate and labor-intensive to install. The design team soon found an opportunity to improve this concept in the design of another high-rise building, 800 Fulton Market, completed in Chicago in 2021. This project is at the edge of the Fulton Market Historic District with protected views to the south and southwest due to the low scale of those historic buildings. With a large- scale office building immediately to the north of the site, the preferred architectural massing was a rectangular long-span bar building with a narrow, offset core on the north side of the site. Early contractor price feedback favored post-tensioned concrete beams over a steel gravity frame, and preliminary analysis suggested supplemental lateral stiffness was required over that provided by the slender core. This series of conditions lead to a design solution similar to 100 Mount Street, with reinforced concrete gravity frame and a steel braced lateral force resisting system, and therefore an opportunity to improve on the design of the brace nodes.
At 800 Fulton Market, the team discovered that all the performance objectives of the central node could be achieved by replacing the complex sliding mechanism with a much simpler geometric mechanism: simply displacing the central node of the braced bay 24 inches out of the plane defined by the four corners of the braced bay. The four diagonal brace members in each bay form the edges of a shallow, horizontal pyramid. When the reinforced concrete columns eventually shorten due to creep, the central node simply moves further out of plane, forming a slightly ‘taller’ pyramid. Similarly, when subjected to temperature variations, the braces lengthen and shorten and the central node automatically moves inwards or outwards to reconfigure the geometry without inducing any forces in the bracing. In this way, the forces acting on the braces are limited to wind, which are equal and opposite axial forces, thereby stabilizing the geometry of the central node. When one diagonal is in compression and tends to buckle outwards, the other diagonal is automatically in tension, holding the central node in position.
The detailing of all the connections in the brace system allows this geometric reconfiguration with only elastic deformations of the steel plates. Instead of a pin-jointed hinging mechanism, a high-ductility “hinging plate” is used at areas of maximum deformation demand. The hinge is formed by a simple rectangular steel plate with the stiffness of the hinge designed by optimizing the height, length, and thickness of the plate. The design challenge of the hinge is to minimize its flexural stiffness to reduce stresses induced as the brace geometry reconfigures, while maintaining adequate strength to transfer the required shear between the two halves of the node. The plate also is designed to withstand the expected deformations elastically and at stress levels and cycle counts that mitigate any concerns about fatigue in the plate itself. To alleviate concerns about fatigue in the welds connecting the hinging plate to the remaining parts of the steel node, the hinging plates are machined down to a dog-bone profile from a thicker plate, significantly reducing the stress levels at the welding points. Beyond the hinging plate, the geometry of the nodes is all determined by analytical optimization. A deliberately oversized design space is analyzed and the material in areas of low stress is removed, until only the useful material remains. The sculptural nature of the nodes is an honest expression of the force flow through the nodes.
SOM has a long history of designing buildings through an integrated approach, producing engineering solutions that are expressed in the building architecture. These buildings’ designs enable an inherent understanding of how they work, and they tend to be well received by the general public. 800 Fulton Market is a prime example, with its state-of-the-art integrated design and research-driven structural systems that are carefully detailed to optimize performance and artfully expressed in the architecture.■