Thin-Shell Concrete Buildings: Yesterday's Marvel, Today's Cast-Off

Thomas E. Boothby, The Pennsylvania State UniversityUniversity Park, Pennsylvania

Charlene K. Roise, Hess, Roise and Company Minneapolis, Minnesota

Figure 1. Kingdome under construction, 1975. King County Archive. The structure was imploded in 2000 when replaced by Safeco Field.

Figure 1. Kingdome under construction, 1975. King County Archive. The structure was imploded in 2000 when replaced by Safeco Field.

The period from about 1920 through 1970, the central time period of the Modern Movement in architecture, was marked by intense optimism about the future of construction and a desire to place architecture and engineering on a shared and rational basis. The thin- shell concrete structure was developed as an engineering solution to economically achieve large spans for industrial, commercial, and public structures, and was embraced by the architectural profession as a potent means of architectural expression. The optimism about this structural type was reflected in the large interest and popularity of the combined professions’ writing and symposia on the topic and in the proliferation of these structures throughout the world. Although originally a European invention, thin-shell concrete structures found a place in the building industry in the United States, and a significant population of such structures remain in the country. It has previously been shown that thin-shell concrete structures in America are an important part of the history of the architectural and engineering professions and worthy of preservation.(1)

It is further evident that these structures are enduring the same threats now that every type of historic structure has endured in the past. Regarded by the public and the professions as dated and ugly, influential works of architecture and engineering are allowed to decay, to become obsolete, to be removed from service, and finally to be demolished. An excellent example of the negative public attitudes towards thin-shell concrete structures, the poor public stewardship of thin-shell concrete structures, and the unthinking demolition of thin-shell concrete structures is the recent implosion of Seattle’s Kingdome (Figure 1). This is by no means the only example available of significant thin-shell concrete buildings of the mid-twentieth century being removed with public complicity, with public funding, and without public debate: I.M. Pei and Anton Tedesko’s Paraboloid in Denver (Figure 2), and Christiansen and others’ Rivergate Convention Center in New Orleans (Figure 3) also fit this paradigm.

Figure 3. Rivergate Convention Center under construction, April 1968. The Convention Center was demolished in 1995. Photo by Nick DeWolf.

Figure 3. Rivergate Convention Center under construction, April 1968. The Convention Center was demolished in 1995. Photo by Nick DeWolf.

It must be understood, however, that much of the public and the public agency owners’ attitudes towards the Kingdome arose from technical problems, particularly problems in maintaining the integrity of the roof. It is conceivable that if the technical problems had been better and more economically managed, the building itself would not have been so poorly regarded, and the public and the public agency owners of the building would not have so readily cast the building off in favor of a more modern 1930s retro retractable-roofed arena (Figure 4).(2) To underscore this point, examples follow of older buildings of this type that have all encountered and overcome similar problems, and remain, for the moment at least, part of our architectural heritage. The examples will include a much less well-known—but arguably more important—monument in the development of thin-shell concrete construction, which is at the present time preserved as a museum piece and used daily as a sports and entertainment venue: the Hershey Sports Arena (renamed the Hershey Park Arena). Other examples are a signature work by a major American architect, Saarinen’s Kresge Auditorium on the Massachusetts Institute of Technology (MIT) campus, and an enclosed athletic arena at the University of Illinois, Assembly Hall.

Thin-Shell Construction

The technology of thin-shell concrete structures and their technical evolution follow a long lineage through the domes of the Roman empire, the vaulting of medieval religious buildings, and ultimately the reapplication of domes, barrel vaults, and groin vaults with the technical innovation of reinforced concrete. Thin-shell concrete structures, essentially a European invention, were exported to the United States for commercial gain by a few patent-holding companies in the 1920s and 1930s, notably through the Zeiss-Dywidag system. The history of the development of thin-shell concrete structures in Europe is well- covered by Billington,(3) and the adaptation of this form to American practice is covered by Billington,(4) and by Boothby and Rosson,(5) and will not be described in detail here. The ideals of thin-shell concrete construction fit in particularly well with the ideals of the Modern Movement in architecture: the desire for primitive geometric forms, the quest to span large spaces with apparent ease, the widespread use of exposed structure and exposed concrete. The forms of thin-shell concrete structures seized the imaginations of the architectural and engineering professions, particularly in the 1960s and 1970s. During these two decades, thin-shell concrete construction became a canonical form for several types of buildings whose programs require large unobstructed spaces and whose formal programs require a plastic expression: industrial and warehouse structures, aircraft hangars and airport terminals, convention centers, and especially sports arenas, both indoor and outdoor.

Figure 5. Structural diagrams for Arch, Vault, and Shell Structures. (a) beam bending: compressive and tensile stresses develop in top and bottom; (b) compressive stresses in arch; (c) plate bending (bending in one direction only shown); (d) meridional (along latitude lines) compressive stresses in domes; (e) beam supporting plate (after Billington (18)); (f) barrel-vaulted shell, showing development of bending stresses through the overall depth of the vault, compressive stresses perpendicular to vault axis, and edge-stiffening beam (after Billington (19))

Figure 5. Structural diagrams for Arch, Vault, and Shell Structures.

(a) beam bending: compressive and tensile stresses develop in top and bottom; (b) compressive stresses in arch; (c) plate bending (bending in one direction only shown); (d) meridional (along latitude lines) compressive stresses in domes; (e) beam supporting plate (after Billington (18)); (f) barrel-vaulted shell, showing development of bending stresses through the overall depth of the vault, compressive stresses perpendicular to vault axis, and edge-stiffening beam (after Billington (19))

A beam, a horizontal structure supported at two ends, can only resist transverse gravity loads by bending, developing large compressive stresses in the top and large tensile stresses at the bottom (Figure 5(a)). An arch, however is restrained horizontally at its two points of support, and resists applied transverse loading (such as vertical loads due to gravity) not by bending but by developing compressive stresses throughout the cross section of the arch (Figure 5(b)). Similarly, a horizontal plate resists transverse load by bending about two axes (Figure 5(c)), while if properly shaped into a vault or dome, the structure can be made to resist gravity loads by compressive “membrane forces,” that is compressive forces in the direction of longitude or latitude (Figure 5(d)). This principle was exploited by ancient builders for the construction of massive long-span structures in masonry that resisted their own weight by compressive stresses throughout the structure. However, the shape of any structure cannot be exactly matched to the loading, and variable, unanticipated, and unbalanced loads (such as wind, snow drift, maintenance, or construction loads) will also arise. These conditions result in bending, in addition to the compressive membrane forces. Ancient masonry vaults and domes resist these bending stresses by increasing the thickness of the shell: the additional weight produces an initial compressive force, which offsets tension due to bending, and the added thickness increases the lever arm of the internal forces resisting bending. Because the steel reinforcing is capable of resisting tension, vaults or domes built with reinforced concrete can resist moderate bending stresses in addition to the compressive membrane forces. With this ability to resist bending stresses, the vault or dome can be made thinner. The resulting reduction in the weight of the structure allows the further attenuation of the shell. A process of experimentation through the 1920s and 1930s first allowed the elimination of stiffening beams and finally resulted in the understanding that the roof structure could be made very thin, ending in achievements such as the Kresge Auditorium’s (113-foot span) 3-inch shell thickness,(6) or the Kingdome (660-foot span) with its 5.5 inch to 15-inch variable shell thickness.(7) (The Kingdome shell was also provided with stiffening ribs.) Where the carefully shaped curved form of a thin shell is interrupted, some kind of edge stiffening, using a beam or diaphragm, is required (Figure 5(e)). The proper design of edge stiffening was often overlooked in thin-shell concrete construction with unfavorable short-term and long-term consequences.

However, the most significant technical problems associated with thin-shell concrete buildings relate to the maintenance of the integrity of the roof. Because the reinforced concrete in a thin-shell concrete structure is meant to resist bending as well as compression, it necessarily develops cracks. Even the most moisture-impervious concrete will, on cracking, allow the passage of water.

Moreover, several other factors have made these buildings especially susceptible to moisture penetration. The placement of concrete in vertically curved configurations and in very thin panels is difficult at best, and problems with improper consolidation of concrete in thin-shell concrete buildings are widespread. A bare concrete roof plane is a troublesome substrate for the application of roofing material, and the experience with applying roof membranes directly to concrete was minimal at the time of the construction of thin-shell concrete structures. The irregular forms of thin-shell concrete roofs, including near vertical surfaces, sharp corners, re-entrant corners, and non-developable surfaces create extraordinary challenges for the application of standard membrane roofs. In many cases, the structural technology of thin-shell concrete roofs was unhappily wedded to some less successful experiments in the development of new roofing technologies, such as spray-on roofs or alternative insulation schemes.

In the remainder of the paper, the history of repairs to four representative buildings will be sketched, beginning with the lamentable history of the Kingdome, continuing in reverse chronological order, to Assembly Hall at the University of Illinois, MIT’s Kresge Auditorium, and the Hershey Sports Arena.

The Seattle Kingdome

The design of the King County Stadium (completed in 1975 and later renamed the Kingdome) was a joint venture including Skilling, Helle, Christiansen, and Robertson, although most of the credit for the form of the roof is due to Jack Christiansen. In the words of David Billington, “The Kingdome is by far the most economical large-scale fixed covered stadium ever built . . . . The Kingdome is also the most visually dramatic roof structure of any large covered stadium anywhere, both from above and from the inside It is an extraordinarily striking thin-shell design . . . .”(8)

Figure 6. Kingdome, Seattle, Washington, 1975. Concrete placement in Shell Roof. This photograph shows the real difficulties of concrete placement on a  steeply and variably sloping roof. A very stiff concrete is being used, difficult to consolidate properly in the best of conditions. Photograph courtesy of Seattle Times.

Figure 6. Kingdome, Seattle, Washington, 1975. Concrete placement in Shell Roof. This photograph shows the real difficulties of concrete placement on a  steeply and variably sloping roof. A very stiff concrete is being used, difficult to consolidate properly in the best of conditions. Photograph courtesy of Seattle Times.

The form of the roof is a scalloped shell, consisting of radial stiffening ribs across the 660-foot diameter, roughly in the shape of an umbrella, with inverted barrel vaults draped between the ribs. Concrete placement and consolidation in the shell portions of the roof structure was quite difficult, as evidenced by the photograph of this operation (Figure 6). As a result, according to a 1994 article in the Engineering News-Record, low density and honeycombing of the concrete were widespread.(9) The roof is constructed with a patented cemented composite wood-fiber board known as Tectum, that has a layer of asphalt felt at the top to receive a roof membrane— however, contrary to the usual application of this material, it was placed below the roof deck, and used as a stay-in-place form liner. As a result of this, the insulation was very poorly bonded to the roof. By 1994, 19 years after the opening of the stadium, the roof was in very poor condition due to lack of cleaning and attack by seagulls. The original roof was a spray-applied urethane foam material, a very poor roofing material. Water from Seattle’s frequent rains had penetrated the roof membrane and found easy paths through the poorly consolidated concrete. To make matters worse, in a roof replacement scheme under construction, the roof membrane was stripped, and the roof deck was being cleaned with very high-pressure water spray. According to the Engineering News-Record:

After frequent leaks and repeated repair attempts, the county decided to strip off the moisture-prone urethane foam in 1993. Without engaging an engineer for the $4 million scheme, the county substituted a modified cement grout and silicone elastomeric coating that, it turned out, failed to bond well. Furthermore, the new system lacked insulation, in violation of the state energy code . . . . In March 1994, they switched to pressure washing at 25,000 to 40,000 psi. Management’s reaction highlighted a rift; the Kingdome’s facilities manager looked after the stadium’s interior and the county’s facilities manager oversaw the exterior. To no avail, the interior’s manager complained about the hydroblasting. On July 19, just before a Seattle Mariners baseball game, four waterlogged ceiling tiles [the Tectum stay-in-place formwork/insulation] fell to the floor of the vacant stadium.(10)
Figure 7. The Kingdome, Seattle, Washington, 1986, after completion of the roofing repairs. Exterior insulation has been used around the perimeter to eliminate ponding. Photograph courtesy of Seattle Times.

Figure 7. The Kingdome, Seattle, Washington, 1986, after completion of the roofing repairs. Exterior insulation has been used around the perimeter to eliminate ponding. Photograph courtesy of Seattle Times.

From the beginning, the Kingdome does not appear to have ever been particularly well-liked by the public in Seattle. A 1993 readers’ poll in the Seattle Times (taken well before the ceiling repair fiasco) identified the Kingdome as Seattle’s worst building.(11) After the Seattle Mariners were unable to resume their home schedule for the remainder of 1994, roof repair crews suffered two fatalities and a subsequent delay of an additional two months in the completion of the repairs, and roof repairs initially estimated at $4 million were completed at a cost of over $50 million in public funds,(12) the public had no love left for the Kingdome (Figure 7). Regina Hackett, in a commemorative piece entitled “That Massive Concrete Shell Had Few Friends Left in the End,” which ran on 27 March 2000, the date of the implosion, says “Willam Bain of NBBJ, the architectural firm in charge of the Kingdome’s design, refused to acknowledge it even had a chief designer, even though George Loschky is usually given the dubious honor.”(13) The implosion of the Kingdome seems to have been met with the kind of public glee that often characterizes such unhappy events.

Assembly Hall, University of Illinois

Figure 8. Assembly Hall (now State Farm Center), University of Illinois at Urbana-Champaign

Figure 8. Assembly Hall (now State Farm Center), University of Illinois at Urbana-Champaign

The initial history of the roof at the 1962 Assembly Hall at the University of Illinois is similar to that of the Kingdome (Figure 8). The original roof coating, a spray-on urethane coating, failed prematurely. Due to the irregular geometry of the structure, the 1979 re-roofing was a field-applied waterproof coating with a white finish coat. The re-roofing cost over $1 million in 1979 (over $6.00 per square foot). A 1995 set of repairs to the roof cost $250,000, and the roof is scheduled for replacement in less than ten years at a cost of over $2 million in 1995 dollars.(14) In this case, however, the roofing problems were detected and corrected in a timely manner, and a single authority, the Operations and Maintenance Department of the University of Illinois, is managing the facility and making appropriate decisions about its continued maintenance.

Kresge Auditorium

Kresge Auditorium has presented a number of technical challenges in construction and in subsequent rehabilitations. Although the shell itself, a triangular segment of a sphere, is adequate, the stiffening system provided at the edges of the shell proved totally inadequate when, on removal of the formwork in March 1954, the shell suffered excessive creep deflections. A month later, the edges of the shell had to be jacked back to a maximum midspan deflection of five inches and provided with permanent supports to resist further creep. This incident undoubtedly opened cracks in the shell that were later observed in the major repair program in 1979 and which allowed a path for moisture ingress. The roofing system was also inadequate and had to be replaced repeatedly up to 1979, when a major roof replacement program, coupled with structural repair, had to be undertaken. These repairs were necessitated by the effect of twenty-four years of moisture penetrating the roofing membrane, passing along the top of the concrete shell, and entering the shell through cracks and poorly consolidated areas of concrete. Particularly severe deterioration occurred along the edge beams and abutments, where the water beneath the roof membrane accumulated without any relief path. The repair program undertaken in 1979, however, was very thorough, being programmed from the start to include the important step of completely removing the roofing membrane for observation of the shell, edge beams, and abutments, identification and repair of all deteriorated areas of concrete, and finally, installation of a carefully designed standing-seam copper roof.(15)

Hershey Sports Arena

Built in 1936, the Hershey Sports Arena helped bring a new and creative method of dome construction to the United States. The significance of the facility lies not only in its unique construction but also in its association with Anton Tedesko, one of the foremost concrete-shell designers of the period. The property, reflecting an important step in the evolution of concrete-shell design in this country, is one of the first American buildings to incorporate the innovative construction method known as the Zeiss-Dywidag system. Moreover, the distinctive structure was built at the behest of an extraordinary man, Milton S. Hershey, founder of both Hershey Chocolate Corporation and Hershey, Pennsylvania, a community affectionately known as “Chocolate Town, U.S.A.”

By post-1950 standards, the structure is quite timid, consisting of a barrel vault shell, with a shell thickness of about 2.5 inches, spanning 232 feet, but provided with heavy stiffening ribs. The completed roof covers an area 232 feet wide and 340 feet long, and its crown is situated 100 feet above the arena’s playing surface. The roof’s two end units each measured 52.5 feet in width. The outermost roof arch supports were positioned 32 feet from either Figure 7. The Kingdome, Seattle, Washington, 1996, after completion of the roofing repairs. Exterior insulation has been used around the perimeter to eliminate ponding. Photograph courtesy of Seattle Times.end wall, thus both end units were cantilevered approximately 19.5 feet over each support. All three intermediate roof units measured just over 78 feet in length. Two arches spaced roughly 39 feet apart supported each one. The roof units were cantilevered about 19.5 feet on either side of the arches. A stiffening rib was provided at the cantilevered ends of all the roof units.

The shape of the shell, being a developable surface, could easily be roofed with conventional built-up roofing materials, while the conservatism of the structural design has made the integrity of the roofing system easier to maintain than the examples given above. The management of the arena and the remainder of the Hershey Park facilities has paid attention to the condition of the building, with regular outside consultant inspections and a conservative maintenance schedule. As a result, the building and its fixtures have been kept in a remarkable state of integrity, even as the building continues to receive heavy use as a sports and entertainment venue.

Conclusions

An important legacy of the Modern Movement in architecture, as well as a daring technical innovation, thin-shell concrete buildings are increasingly threatened by public disdain and functional obsolescence. In recent years, some particularly important thin-shell concrete buildings have been removed, often with the use of public funds. The poor public perception is exacerbated by some of the vexing technical problems frequently encountered in these buildings. The long-term preservation of thin-shell concrete buildings will require not only the development of improved awareness of the importance and worth of these buildings, but also the resolution of important technical problems of concrete repair and rehabilitation and roofing. The political management of technical issues is arguably more important than the application of appropriate technology. A quotation by Richard K. Sandaas, Kingdome interim director, from a review of the roofing repairs of the Kingdome is particularly revealing. “The lesson learned here must be to avoid this situation in the future by planning and programming, and by funding maintenance, operations, and capital improvements with a long- term vision.”(16)

Acknowledgements

Professor M. Kevin Parfitt assisted in the assembly of material on the case studies, and Carmen Gerdes assisted in clarifying the explanations of the structural behavior of thin-shell buildings. Dennis Gardner conducted historical research.

Notes

1. Thomas E. Boothby and Barry T. Rosson, “Preservation of Historic Thin-Shell Concrete Structures,” Journal of Architectural Engineering 4, no. 1 (March 1998): 4–11.
2. The Kingdome will be replaced by two sports venues, an as- yet-unnamed football and soccer stadium on the site of the Kingdome, and Safeco Field, the retractable-roofed baseball arena, which is described by ballparks.com as “. . . built to resemble the great ballparks of yesteryear.” (http://www.ballparks.com/baseball/american/seabpk.htm). Descriptions of the new football and soccer stadium, as well as clips of the Kingdome implosion can be obtained at http://www.seahawks.com.
3. David P. Billington, The Tower and the Bridge (Princeton: Princeton University Press, 1983).
4. David P. Billington, “Anton Tedesko: Thin Shells and Esthetics,” Journal of Structural Engineering 108, no. 11 (November 1982): 2539–2554.
5. Boothby and Rosson, “Preservation of Historic Thin-Shell Concrete Structures.”
6. Edward Cohen, Norval Dobbs, and William Combs, “Inspection, Analysis, and Restoration of MIT Kresge Auditorium,” in Rehabilitation, Renovation, and Preservation of Concrete and Masonry Structures, American Concrete Institute Special Publication, SP-5 (Detroit: American Concrete Institute, 1985), 95–125.
7. “Fancy Formwork Shapes Largest Concrete Dome,” Engineering News-Record 194, no. 16 (1975): 18.
8. David Billington, letter to the editor, Seattle Times, 15 June 1997.
9. “Falling Ceiling Tiles Uncover Problems,” Engineering News- Record, 1 August 1994.
10. “One Steep Roof Repair,” Engineering News-Record 234, no. 8 (27 February 1995): 30–32.
11. “City’s Worst Building Described as a ‘Blight’,” Seattle Times, 28 March 1993, B2.
12. “One Steep Roof Repair.”

This article originally appeared in Preserving the Recent Past II, published by HPEF, National Council for Preservation Education, and the National Park Service. The conference proceedings are available as a free digital download