pei’s proto-tube structures

Inspired by Lee Bey’s Southern Exposure, which celebrates the often-ignored architecture of Chicago’s South Side, my weekend mornings have involved skyscraper scavenger hunts to find some of the city’s tall buildings that have never made the canonical history books. Most of these are residential blocks–some public housing, but also numerous developer towers that extend and complicate the tidy narratives of downtown’s commercial architecture.

Several of those to come, but this morning I turned inland through Hyde Park to zip past I.M. Pei’s University Apartments from 1961. These are well known for the way they sit in the middle of 55th Street and for their role in the infamous plan to “renew” the neighborhood around the University of Chicago. Those are both well-traveled tales, but I have a parallel interest in them that’s actually tied directly to those more celebrated towers downtown…

One of the best known stories of postwar Chicago is the development of the tube structure by Fazlur Khan and a team of engineers and architects at Skidmore, Owings, and Merrill. Khan recognized that beyond 35-40 stories, a skyscraper is most efficient if it’s thought of as a giant column, subject to all kinds of random bending and buckling loads from wind. The theoretically ideal shape for a tall column is a hollow cylinder–a tube–and Khan realized that by concentrating a skyscraper’s structure around the perimeter, he could take advantage of this principle on a large scale. The results were towers like the Brunswick, which distilled its structure into dozens of small columns around the exterior, all fixed to deep spandrel girders and thus working together as four giant (but heavily perforated) shear walls.

SOM | Brunswick Building
Brunswick Building, Dearborn and Washington Sts. SOM, 1965 (Skidmore, Owings, and Merrill)

But, like any other structural innovation, the Brunswick didn’t come from nowhere, and Hal Iyengar, one of Khan’s trusted engineers, recalled later that the idea for stiff perimeter connections and oversized columns that would approximate a hollow tube structure had played a role in engineering the Equitable Building, at Michigan Avenue and the River, in 1963-65.

Brunswick Building, Dearborn and Washington Sts. SOM, 1965 Typical Floor Plan showing Perimeter Structure. (Skidmore, Owings, and Merrill)

But even that idea seems to have come from somewhere else–New York, to be exact. Concentrating structure around the perimeter means, necessarily, that the 19th century fight between windows and structure was back on, and the compression of window wall and bearing structure into a single plane was employed by I.M. Pei on two projects in 1961–Kips Bay Plaza Apartments in New York, and Hyde Park’s University Apartments. These were both mid-rise buildings, but they were significant for the way their structures suggested a return to exterior bearing walls, albeit ones with wildly different proportions since they were made of reinforced concrete instead of simple masonry. Such a circling back to bearing wall construction was lost on many, including the New York Times’ Glenn Fowler, who reported on the two projects as:

Kips Bay Plaza, I.M. Pei & Associates, 1963. Typical floor plan showing perimeter structural walls. (Pei, Cobb, Freed).

“…a new method of concrete construction, in which the facade of a building serves as its supporting framework…”

Glenn Fowler, “Facade of Building Forms Structural Support for High-Rise Apartments:  New York Times, Apr 23, 1961, pp. 449.

Bruce Graham, interviewed for the Art Institute of Chicago’s Oral History Project, recalled that the Brunswick wasn’t as innovative as it was made out to be–“the idea of the tube was not ours. It was taken from a building by Pei in New York,” he told her, and even Chicago partisan Carl Condit acknowledged, in 1964, that the concrete tube had originated with Pei:

“…the Brunswick is the largest building to date with external bearing walls of rigid-frame concrete trusses (also known as load-bearing screen walls and window truss walls).  A wall of this kind is the most recent structural innovation in multi-story American building but it springs from a long and complex history….It was first adapted to the construction of entire building walls by the architect I. M. Pei in the design of two apartment groups built in 1959-61 in New York and Chicago: the Kips Bay Plaza Apartments in New York and the University Apartments on 55th Street in Hyde Park.  Loewenberg and Loewenberg were the associated architects on the latter buildings.”

Carl W. Condit, “The New Architecture of Chicago.”  Chicago Review, 17:2/3, special issue on New Chicago Writing and Art.  1964.  11

University Apartments, 55th St., Chicago. I.M. Pei & Associates. Partial typical (top) and ground (bottom) floors showing tightly spaced columns around perimeter. (Architectural Record, April, 1962).

Much more on those associated architects in a future post–but what’s most interesting to me about Condit’s reporting is that Kips Bay was actually finished in 1963, while the University Apartments–on SOM’s doorstep–were finished in 1961, just as the Brunswick was being announced. While University apartments are also mid-rise–at 470 feet long and just 94 feet tall, they’re the ultimate groundscrapers–it’s easy to imagine the construction site being of particular interest to Khan, Graham, and the other designers contemplating a high-rise concrete structure and primed by their experience on the Equitable to be thinking of exterior structures.

University Apartments, 55th St., Chicago. I.M. Pei & Associates. Exterior view with detail of concrete bearing wall. (Architectural Record, April, 1962).

Things get even more interesting, though, when you dive into University Apartments’ engineering, because the structural engineer for the project was August Komendant, better known for his work with Louis Kahn and, at the time, New York’s premiere concrete engineer. Komendant and Pei developed a styrofoam forming system for University Apartments that produced super-smooth finishes–a technique not dissimilar to that adopted last year by SOM for their Stereoform Slab project. And two years after University Apartments, when Kips Bay was being completed and Kahn was struggling with mixtures and formwork on the Salk Institute in the summer of 1963, he had Komendant approach Pei for advice, and Pei generously shared his office’s concrete specs. While its local legacy is a decidedly mixed one, University Apartments’ more subtle influences were widespread…

6 thoughts on “pei’s proto-tube structures

  1. Two other interesting early tubes (or proto-tubes) are the CBS Building in NYC by Eero Saarinen, (1960 design, construction completed 1965) with Weidlinger Associates, as structural engineers; and the IBM Building in Pittsburgh (1961-1963) with Curtis and Davis, architects; and Worthington, Skilling, Helle and Jackson, structural engineers.

    Liked by 1 person

    • Indeed—I’d forgotten about IBM, but CBS was paired with the Brunswick in an Architectural Forum article—the (New York) critics thought Saarinen’s building was better architecturally, but praised the structure of Brunswick…


  2. Fazlur Khan’s biography by his daughter Yasmin Sabina Khan (also a structural engineer) agrees that Kips Bay and University Apartments may “have influenced Khan’s thought processes.” “Pondering the grid elevations of the University Apartments could not have been more convenient for him; he and Liselotte had moved to a top-floor apartment in one of the complex’s 10-story buildings in 1962.”
    But she also points out, in the caption to a photo of Kips Bay, that “although the columns are closely spaced, the grid lacks substantive spandrel beams, an essential component of the framed tube structural system.”


    • Good catch–I’d forgotten about this passage…Its spandrel beams are fairly thin, but their connections to the columns look to me like they’re intended as moment connections–which would explain the curving corners. Further research needed, clearly, thanks for pointing this out!


  3. See also Society Hill Towers in Philadephia (I.M. Pei) and Dewitt-Chestnut Apartments in Chicago (SOM). They are officially dated 1964 and 1963 respectively, all though it would be of interest to untangle the actual dates of design versus completion and see what there is to learn about precedence and influences.

    University Apartments is clearly exploring the exterior bearing wall and core methods, though ultimately 10 stories in a linear block is a different structural problem than a 30-40 story (or higher) tower. Society Hill and Dewitt-Chestnut are closer relatives to Brunswick in that way.

    Notable also are the differing transitions from the upper stories to the ground floor. Pei draws little attention to the transfer of the forces from the grid of the upper elevation to the ground floor columns in his examples. Brunswick has many fewer columns on the ground floor, in order to accommodate a street level commercial use and probably to create less of barrier to the sidewalk. Brunswick employed a deep exterior wall structural mezzanine to span these larger distance and distribute the much greater bearing loads of a tower to fewer columns. Dewitt-Chestnut in contrast makes almost a plinth at the mezzanine level, though like Pei’s buildings, since the ground floor columns are only spaced at two times the upper verticals, they don’t need Brunswick’s more substantial mezzanine structure.

    It might be said that the Dewitt-Chestnut almost has some Classicism emerging in the ground floor treatment, at least in the sense it evokes a trabeated colonnade. Considering it stands across from Mies’s LSD buildings, SOM designers, already well-versed in his work, are said to have been paying attention to those famous towers. It would be hard not to, placing their own grid right next door to his.

    Though the materials differ, Brunswick in plan is rather Miesian, too, perhaps more so than some of his own buildings were. It’s also hard not to see a little Seagram Building in the proportions and the mechanical floor treatment, among other features.

    Liked by 1 person

    • Agreed–there’s a whole range of approaches to ground floors of tube buildings in SOM’s work. I think the best ones are the Graham/Khan design for the two Shell Plaza buildings in Houston, where you can see the forces collecting as columns branch out around the lower floors.

      Dewitt-Chestnut is fascinating to me because it shows how the principle has to be fine-tuned to buildings of much greater height–in addition to the very classical solution to the ground floor. I had an undergrad research assistant model it last year, and one of the things he found is that its columns and spandrel beams both get gradually thinner as the building rises. This makes sense in terms of bearing loads, obviously, but I suspect it also has to do with wanting greater ductility in the upper floors–something that Nervi developed in the Vierendeel columns in the Pirelli Tower, Milan. It also makes for a very subtle but notable visual effect–a slight forced perspective that emphasizes the very vertical proportions of the tower. That, I think ties in with your comment about both the Seagram’s and Brunswick buildings’ ‘attics,’ both of which seem remarkably close to the proportions of a good Corinthian capital.

      All good stuff–thanks for posting your thoughts. There’s more coming on this, including some work on the ‘heirs’ to the University Apartments’ structural system in other apartment buildings in the city through the 1960s…


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s