Introduction

One of the undergraduate engineering courses that left a deep and lasting impression on me was a course on innovation and aesthetics in engineering, taught by David Billington at Princeton University. One example I remember vividly is the comparison between the Eiffel Tower and the Saint Louis Arch, both iconic monuments of steel that have come to symbolise their cities. According to Billington, the Eiffel Tower is an example of good engineering, as it sits on four legs and is inherently stable, whereas the Saint Louis Arch rests on only two points, and requires feats of foundation engineering to keep it standing. So, when I read the story of a skyscraper in New York City that had to undergo secret repairs thanks to a query from a New Jersey engineering student, I had a hunch that the student must have been one of Billington’s. And I knew then that I wanted to come back and investigate this issue in greater detail some day. Both my hunch and my wish have since been realised.

The Citicorp Tower story

[caption id="attachment_51062" align="alignright" width="666"] Fig. 2 Near-wake instantaneous particle traces behind a square cylinder for two orientations (from Dutta10)[/caption] The award-winning, wedge-topped, 59-storey Citicorp Building in Manhattan features striking columns at the middle of its four sides rather than its corners (Fig. 1b). This remarkable configuration was due to the existence, at one corner, of a church (now demolished), that refused to be bought out, but did grant the use of the space above it. Forty-one years ago, in 1978, a student working on an undergraduate thesis telephoned the chief engineer on the building project, Le Messurier, to ask about the stability of the building under the effect of winds impacting its corners, also known as quartering winds. The chief engineer told the student that the 1970 New York City building code only required consideration of head-on or face winds, that is, winds that hit the flat sides of the square building. The student, originally reported to be a male engineering student, was identified decades later as a female architecture student at Princeton, Diane Hartley (1). How form and function influence each other, a central theme in Billington’s legacy, was the basis of Hartley’s question, which, unbeknown to her, started an extraordinary chain of events. According to the commonly accepted story, after Hartley’s call, Le Messurier recalculated some numbers. He was alarmed by the results and flew to Canada to consult with the engineers who had performed wind tunnel tests on a model of the building. Then, he initiated emergency repairs that were performed in secret, at night, while the building was in full use in the daytime. On September 1, 1978, repairs were still unfinished as Hurricane Ella and its high winds bore down on New York City. If the building toppled, it could knock down neighbouring skyscrapers like a series of dominos. Officials in the know were ready to call for a massive evacuation of downtown Manhattan when, much to their relief, the storm veered away. This story was only revealed in 1995 by a journalist from 'The New Yorker' (2). In the following decades, the chief engineer of the Citicorp Building gave many lectures on engineering ethics, using it as an example (3, 4, 5, 6, 7).

Modern analysis

However, some of Le Messurier’s close collaborators disagreed with 'The New Yorker' account and stated that the building suffered from more defects than the article revealed. Curiously, Le Messurier, while he was widely lauded for his ethical behaviour, did not set out at the time to change the New York City building code, which, by implication, was dangerously negligent in not requiring that quartering winds be considered in the design of tall buildings. Remarkably also, in distilling lessons learned for future generations of engineers in his many lectures, Le Messurier released neither the internal engineering report that detailed the building’s structural deficiencies, nor the wind tunnel tests that would have allowed independent scrutiny. In a 2002 article, architecture professor Eugene Kremer (5) stated that he was “perplexed by the absence of a re-evaluation of the conventional wisdom on this celebrated case”. Undoubtedly, the absence of data has something to do with it. In recent years, Emil Simiu, other researchers at the National Institute of Standards and Technology in Gaithersburg, Maryland, and myself have developed a method called Database-Assisted Design (DAD) (8). This rigorous procedure is enabled by the development of hardware capable of simultaneously measuring and recording time histories of wind pressures at hundreds of taps, and the availability of computer resources needed for processing large amounts of data economically and rapidly. Let me explain. The state of the art in 1970 is represented by the following table: [caption id="attachment_50891" align="alignright" width="295"] Table RS9-5-1 Design wind pressure on vertical surfaces (psf of projected solid surfaces) vs. height zone (ft above curb level), excerpted from the1970 Building Code of the City of New York.[/caption] In comparison, DAD uses simultaneous measurements on a wind tunnel model sampled (in this case) by 500 pressure taps 1,000 times a second for 30 s for each of a dozen wind directions. Not only are the wind pressures measured in a much more fine-grained fashion in space, over the entire building surface, but the fine resolution in time also captures the dynamic fluctuations of the chaotic phenomenon that is wind. These measurements enable the calculation of the dynamic responses of a building along and across the wind direction. It should be noted that the designers of the Citicorp building did not just rely on the four pressure values specified by the building code, but performed their own wind tunnel tests. At the time, though, they could not collect many measurements simultaneously, and only recorded the reaction forces and overturning moments at the base of the building for various wind directions. [caption id="attachment_51065" align="alignright" width="473"] Figure 1. a) TPU pressure tap areas on square building of aspect ratio 5, and b) schematic of Citicorp Building (measurements from 100 out of 125 taps per each façade, totaling 400 taps, were used, shaded areas).[/caption] We did not have access to the proprietary wind tunnel tests of the Citicorp building, but we found in the Tokyo Polytechnic University (TPU) aerodynamic database (9) wind pressure time histories for an isolated building in open terrain with square cross section and depth to height ratio 1:5, modelled at a scale of 1:478.   By mapping 400 of the 500 pressure taps from the model onto the actual building (shaded parts of Fig. 1), we were able to capture the fluctuations of the wind pressure on all faces of the building simultaneously and for a dozen wind directions. (Due to building symmetry, only one-eighth of all possible wind directions need to be tested, every five degrees.) With the wind pressures thus determined with high spatial and temporal resolution, we calculated the building’s response with the same temporal resolution, turning the spotlight on overall behaviour, such as overturning moments, and on individual members, such as the eight-storey chevron braces, which were the focus of the repairs. In the past, a great deal of engineering judgment had to be exercised to select the combinations of pressures most detrimental to the building. With today’s computers, this is no longer necessary. In total, we performed 30,000 structural analyses that covered the entire duration (30 s) of the wind tunnel tests. These analyses accounted for the dynamics of the building in response to the time-varying wind pressures, which reflected, among other things, the vorticity shed in the structure’s wake (Fig. 210). Vortex-shedding causes across-wind motions.

Results

The results surprised us, and we spent considerable time making sure they were correct. The analysis (11) shows that corner winds are less demanding than face winds. The along- and across-wind overturning moments in the corner wind case are about 20 per cent and 50 per cent lower, respectively, than their counterparts in the face wind case. [caption id="attachment_51067" align="alignright" width="524"] Figure 3. Along- and across-wind overturning moments for two orientations (red circles mark the peak resultant overturning moments).[/caption] Figure 3 shows the results of all 30,000 analyses: each cloud contains 30,000 points, with the co-ordinates of each point representing the overturning moments along and across the wind, respectively. The peak resultant overturning moments are marked by red circles. It turns out that the across-wind response dominates in the case of face winds (Figs. 2 and 3). In Fig. 3a, the results cloud is wider along the horizontal axis than along the vertical, meaning that, for face winds, the peak overturning moment is greater across wind than along wind. In comparison, Fig. 3b shows a circular results cloud, implying that, for corner winds, the peak overturning moments across wind and along wind are equally strong.[caption id="attachment_51069" align="alignright" width="524"] Figure 3. Along- and across-wind overturning moments for two orientations (red circles mark the peak resultant overturning moments).[/caption] Measurements of base moments in the 1970s wind tunnel tests would have reached the same conclusion. The peak axial forces in the mid-side columns and the peak demand-to-capacity indexes of the chevron braces induced by corner winds are lower by 20 per cent to 30 per cent than their counterparts due to face winds. Surprisingly, the original answer to Hartley’s question was right: our investigation confirms that the building code of the city of New York in effect in the early 1970s can be interpreted as meaning that the design for wind of structures with a square shape in plan may be performed by assuming the wind loads to act perpendicularly to a face of the building; that is, if the building were safe for face winds, then it would be safe for corner winds, whether the diagonal braces were welded or bolted together.

Discussion

In this study, wind loads on the wedge-shaped crown of the structure (Fig. 1b) is neglected, but this should not affect the conclusion on the relative importance of the effects of corner winds and face winds. More significantly, wind loading in this study is for open terrain, whereas in actuality, there is significant interference from adjacent buildings. How this affects the conclusion is impossible to ascertain in the absence of the actual wind tunnel test measurements. However, to our knowledge, the interference of neighbouring buildings on the wind loads has never been mentioned in discussions of the Citicorp building’s strengthening, including the interpretation of the building code of the time, Hartley’s initial question, and the justification for the repairs. We further note that, when the Citicorp building was designed, wind tunnel testing was at an early stage of its development. Even modestly reliable statistics of hurricane wind speeds and directions on the basis of which mean recurrence intervals of wind-induced effects could be estimated were unavailable. Structural analysis was largely done by hand, as was evidenced by the interruption of the corner columns every eight storeys, for the purpose of making each eight-storey stack statically determinate and easier to analyse.

Conclusion

We started our analysis with the goal of quantifying the need for the repairs and how closely catastrophe was averted by the wise question of a student and the brave and urgent actions of the chief engineer, leading to dramatic, clandestine repairs done under the cover of darkness in a bustling city. Instead, we found that Hartley’s question should not have caused any alarm, and Le Messurier’s original response was correct. The modern, powerful experimental and computational methods we used produced results that suggest, when interpreted in light of the similarities and differences between our computer model and reality, that the basis for the decision to strengthen the structure after its completion deserves to be revisited.

References

1.) Hartley, DL (1978). Implications of a Major Urban Office Complex: the Scientific, Social, and Symbolic Meanings of Citicorp Center, New York City, BSE thesis, Princeton University, Princeton, NJ. 2.) Morgenstern, J (1995). 'The Fifty-Nine-Story Crisis'. The New Yorker, New York City, N.Y., May 29, 45-53 https://www.newyorker.com/magazine/1995/05/29/the-fifty-nine-story-crisis [cited 4.27. 2018] 3.) LeMessurier, W (1995). 'The Fifty-Nine-Story Crisis: A Lesson in Professional Behavior', MIT Mech. Engg. Colloq. 17 Nov. 1995, https://medium.com/@seagertp/the-disaster-that-wasnt-nyc-c-1977-eea621d28eff, https://www.youtube.com/watch?v=um-7IlAdAtg [cited 09.13.2018] 4.) National Academy of Engineering Online Ethics Center (2014) 'William Le Messurier The Fifty-nine-story crisis: a lesson in professional behavior' http://www.onlineethics.org/Topics/ProfPractice/Exemplars/BehavingWell/lemesindex.aspx, with addendum by Caroline Whitbeck “The Diane Hartley Case,” http://www.onlineethics.org/DianeHartley.aspx updated 02.11.2014 [cited 09.13.2018] 5.) Kremer, E (2002). "(Re)examining the Citicorp Case: ethical paragon or chimera?" arq, Cambridge Journals, 6(3), Sept. 269-276 https://doi.org/10.1017/S1359135503001763, http://www.crosscurrents.org/kremer2002.htm [cited 4.27.2018] 6.) 99% Invisible '04.15.14 Podcast on Structural integrity' https://99percentinvisible.org/episode/structural-integrity/ 7.) Brady, S (2015). 'Citicorp Center Tower: how failure was averted', Engineer’s Journal, Dublin, Ireland 8.) ASCE (2016). Minimum design loads for buildings and other structures, §C31.4.2, ASCE 7-16, American Society of Civil Engineers, Reston, VA https://www.asce.org/structural-engineering/asce-7-and-sei-standards/ [cited 4.27.2018] 9.) Tamura, Y (2012). TPU Aerodynamic Database http://wind.arch.t-kougei.ac.jp/system/eng/contents/code/tpu [cited 4.27.2018] 10.) Dutta, S (2006). Sensitivity of a square cylinder wake to orientation and oscillation in the intermediate Reynolds number regime, Ph.D. thesis, Indian Institute of Technology, Kanpur, India http://nptel.ac.in/courses/112104039/thesis4/dutta_thesis.pdf [cited 4.27.2018] 11.) Park, S, Duthinh, D, Simiu, E and Yeo, D (2019). 'Wind effects on a tall building with square cross section and mid-side base columns: a database-assisted design approach.' ASCE J. of Structural Engineering, Vol. 145, No. 5, May, online Mar. 6, https://ascelibrary.org/doi/abs/10.1061/%28ASCE%29ST.1943-541X.0002328 Acknowledgments: We are grateful to Diane Hartley for enlightening discussions of her undergraduate thesis, and to Christa Cleeton of the Seeley G. Mudd Manuscript Library at Princeton University for making available high-resolution copies of the Citicorp Building blueprints. This study is dedicated to the memory of David Billington, an inspiring teacher to the author and thesis supervisor to Hartley.

About the author

Dat Duthinh, pictured, is a research structural engineer with the US National Institute of Standards and Technology in Gaithersburg, Maryland. For the last several years, he has developed methodologies and computational tools to 1) improve the design of buildings to resist multiple hazards, such as hurricanes and earthquakes, for greater efficiency and risk consistency; 2) advance the state of the art of fire-structural analysis of buildings, in particular the use of adiabatic surface temperatures to model the heat flux to structural surfaces, and the transfer of temperature results from thermal analysis to structural analysis, which typically uses different types of finite elements; 3) advance the state of the art of design of buildings to resist high winds with direct use of wind tunnel measurements (database-assisted design) and explicit accounting of wind directionality and probability of occurrence. He also contributed to the development of the technical basis for standards for high-strength concrete (HSC) structures, especially in the area of shear strength, and for the use of fiber-reinforced polymers (FRP) in infrastructure (repair of concrete structures and connections of FRP members). Prior to US government service, he worked on the design and construction of the Hibernia Gravity Based Structure, a massive concrete platform offshore Newfoundland, Canada, designed to resist the impact of icebergs and the onslaught of waves. One of the most exciting projects during his decade spent studying ice-structure interaction was an Antarctic expedition he led to measure the pressure and force of impact of icebergs against vertical structures. He took part in the investigation of the New York City World Trade Center collapse, served on several panels of the National Academies, was a visiting professor at the University of Rome La Sapienza in 2010, and is associate editor of the 'American Society of Civil Engineers Journal of Structural Engineering'. He is the author or co-author of 120 journal papers, book chapters, conference articles and reports, and the recipient in 2007 of the United States Department of Commerce Silver Medal for Scientific/Engineering Achievement. Dat Duthinh holds a bachelor degree from Princeton University, a master degree from the University of Delaware and a PhD from Cornell University, all in civil/structural engineering.