2016 Structural Engineering Traveling Fellowship
Integrating Secondary Goals into Structural Design

Nathan Brown traveled extensively to locations in Asia, Australia, Europe, and North America to carry out his research.

Nathan Brown
Massachusetts Institute of Technology
Department of Architecture

View Application Essay
View Final Report

Jury
William Bast
Martin Furrer
David Horos (Chair)
Catherine Wetzel
Todd Zima

If managed appropriately, a second significant design requirement can be turned into an opportunity. In the design of structures, the tenets of efficiency and economy are widely pursued goals. A number of talented designers have pushed beyond these basic requirements to achieve structures that are visually elegant as well. Many historically celebrated precedents are clear, sound responses to a particular condition—a column-cantilever following the moment diagram of a wind load, a funicular truss with appropriately sized sections, or a concrete shell shape determined by form-finding. When further design requirements are added to the process, however, the solutions cannot be as straightforward, since designers may need to manage tradeoffs between competing design goals while arriving at a geometry and form that satisfy a variety of conditions.

The design of modern buildings and other structures demands the ability to synthesize multiple design goals simultaneously. This is largely true because of an increased emphasis on overall performance and design sustainability, of which structural material efficiency is only one consideration. Modern computational methods have been developed to aid the effort of effectively managing different design goals, especially in the case where multiple design goals can be quantified, simulated, and measured. Yet, though computers can be helpful in rationalizing geometry and completing performance analyses, they cannot yet generate diverse typological possibilities, prioritize different design goals, or evaluate the aesthetic expression of a structure without input. There is still no substitute for human creativity, intuition, and experience in design.

Numerous historical examples exist of structural designs that clearly derive their form from the careful consideration of multiple, simultaneous goals. Early instances include churches that rose to soaring heights and enclosed volumes with heavy stone yet managed to create interiors washed with daylight. Other examples show how designers were able to solve uncommon structural problems while also achieving occupant comfort through the use of thermal mass and passive heating, cooling, and ventilation strategies. Recently, structural designs that pursue multiple goals include skyscrapers shaped to shed wind and self-shade, long-span stadiums on the leading edge of energy efficient design, and courtyard roofs that make compelling structural gestures while also performing well acoustically. There is significant value for an aspiring structural designer, and for the field of structural engineering as a whole, in studying these precedents for inspiration.

This report is a record of the 2016 SOM Foundation Structural Engineering Traveling Fellowship, which was planned to explore the design of such structures that are distinguished by their emphasis on fulfilling multiple objectives. While the report is an opportunity to organize and contemplate what I have learned during the fellowship, it is also meant to be a resource for other designers, researchers, or educators who are interested in the topic. The precedents in this report will hopefully inspire future creative structural designs, especially now that structural engineers are regularly expected to interface with many different types of design specialists, and making a building stand up is not the only noteworthy performance consideration.

Tokyo International Forum, Japan. © Nathan Brown.

Somf 2016 structural engineering nathan brown final report 02

Objectives in Design

At this point, it is important to discuss what exactly is meant by a “goal” or an “objective” in the context of design. In structural or other types of optimization, the term objective has an explicit definition—it is a measurable quantity that is being minimized or maximized by changing different variables in the problem, subject to a variety of constraints. The amount of structural material in a building could be an objective function worth minimizing as part of a design process. So could the simulated operational energy of that building, or the number of hours annually where natural daylighting is sufficient to meet its functional needs. Optimization is increasingly seen as a method for pursuing performance-driven design, and many of the objectives discussed in this report could be translated into a formal objective or cost function.

However, this report uses a much broader definition for what a design objective could be. Architectural and structural designs can be evaluated mathematically in some meaningful ways, but the success of a structure cannot be entirely reduced to numerical terms. The most important objective of a design might be something that is purely visual or aesthetic, even in the field of structural design, as the SOM Foundation Fellowship brief points out. Even among areas of design that are generally focused on quantitative goals, such as energy usage or artificial lighting, a desired design quality could sometimes be more difficult to capture, such as the way daylight bounces off a sculpture after it enters a skylight, or the sensation of warmth emanating from the thermal mass of a masonry wall. A substantial restriction on the definition of a secondary design objective to purely technical quantities risks ignoring important aspects of design and could undermine the goals of the fellowship.

An objective as discussed in this report can thus be a specific quantity such as structural material or energy usage, or it can be a nonquantifiable effect identified by the designer or clearly distinguishable from a visit to the structure. It could also include the response to an aspect of the site, such as a difficult geometrical problem that might be called a design constraint in other contexts. This loose definition of objectives allows for an extensive selection of interesting precedents, and it widens the applicability of the lessons learned throughout the fellowship. The terms “goal,” “aim,” and other similar words will be used synonymously as well, since each of these terms captures the broad situation in which a designer cannot simply consider one aspect of building performance or experience.

Yoyogi National Gymnasium, Tokyo. © Nathan Brown.

Somf 2016 structural engineering nathan brown final report 03

Travel Fellowship Goals

The purpose of the proposed research is thus to gain a better understanding for how these multiple design goals can be effectively pursued at once, especially when both quantitative performance objectives and fuzzier, generally subjective, qualitative goals are important to the designer. This question, or a version of it, has been present at every step of my education and professional career. From my initial exposure to both architecture and structural engineering up through my current role as a researcher, I have gradually learned more about how buildings, other types of structures, and the built environment in general come into existence. The opportunities associated with the SOM Foundation fellowship have considerably accelerated this training in both design process and precedents, as the best way to absorb the intent and execution of a structure is quite often to experience it firsthand. In one sense, the goals of the fellowship are fulfilled simply by this exposure to the myriad ways in which designers wrestled with and ultimately synthesized different design requirements, as evidenced by their final products.

Yet the intent of a research fellowship suggests that the outcome should include something more than a simple record of this exposure—a way to evaluate, or at least organize what has been learned. When explaining this fellowship to design professionals during my trip, a few asked a specific question: how exactly do you intend to assess the level of success achieved by each design on your itinerary? These buildings have been collectively experienced by millions of people, ranging from experts in the fields of architecture and engineering to more casual patrons or observers. Many of the most famous buildings have already been extensively documented. For my fellowship research to build on existing scholarship and further the goals of the SOM Foundation, this report must offer a distinctive framework or lens for understanding these building precedents and the influences that drove their design processes. Thus, although the building descriptions in this report are intended to provide at least some basic context, history, and notable achievements for each one, the entries also explain how each example fits within a broader research lens.

Kimbell Art Museum, Fort Worth, TX. © Nathan Brown.

Somf 2016 structural engineering nathan brown final report 04

Research Lens and Report Organization

This lens has a number of components. In keeping with the fellowship’s aim of fostering an appreciation of the aesthetic potential in the structural design of buildings and bridges, the travel itinerary highlights precedents that make noticeable structural gestures and solve decisively structural challenges. In this way, the research is grounded in the field of structural engineering. Assessments of the studied buildings are primarily made in terms of structural efficiency, clarity of form and expression, level of technical proficiency and innovation, durability, and other facets within the field.

Although the structures documented in this report were chosen because they are exemplary in one or more of these structural aspects, the examples are diverse in terms of age, geography, scale, ambition, and fame. As such, a contemporary observer can often appreciate especially creative or technically sound aspects of a design, while also criticizing other decisions, often with the benefit of hindsight. In order to compare the different structures and most effectively extract concrete and valuable design lessons, it is tempting to place these building assessments on some type of scorecard. A system of rankings, or an evaluation of performance across set categories, would communicate the outcomes of this research in a way that is comfortable, or even potentially useful, for many engineers.

At the same time, it is important to acknowledge the limitations of a ranking methodology for addressing the initial question about design for multiple objectives. First, despite everything that can be learned through site visits, conversations with designers, and even drawings or secondary sources about a building, it is difficult to fully know every constraint acting on a particular designer. This is especially true of the older or more obscure structures, which are not as well documented. If this research fellowship was addressing only one quantifiable building aspect such as its carbon footprint, and reliable information could be found about material quantities for each one, a ranked list would be more useful.

However, as the research was proposed and conducted, it would be unfair to evaluate the buildings and their vastly different intents and expectations according to a rigid scale. In addition, each of the buildings on the list was conceived of by a designer who has more experience than I do. In most cases their knowledge and understanding of design and construction totally dwarf mine, since I am an early career researcher and aspiring designer. Also, my primary training is in structures, and so I am not an expert in all the other aspects of performance-based building design—energy, daylighting, acoustics—that are discussed in terms of how they interact with structural considerations. Although the fellowship has allowed me to visit many high-quality design examples and I consequently have some context for critique and comparison, it would be dangerous to presume that I know more than the experienced designers who built these remarkable structures.

Furthermore, when the notion of aesthetics is involved, both in the limited sense often used by engineering historians to describe visual quality, as well the more general definition of being concerned with beauty, the question of evaluation becomes even more murky. A method for evaluating aesthetic quality would have to leave the field of structural engineering and probably enter the field of architecture. Architects are of course comfortable sharing their opinions on the expressiveness and experience of buildings, and this report similarly contains discussions of many nonquantifiable aspects of structures. However, trying to fit aesthetic assessments into an overall rubric would require value judgements that may yield less-valuable results.

In some ways, the question of how to effectively pursue multiple design goals simultaneously, even if one goal is explicitly structural or especially dominant, cuts to the heart of design itself. In my experience presenting research on the topic of multiobjective optimization, which is a computational approach that can help navigate performance-based design spaces, architects and engineers have strong opinions about its usefulness and role in the design process. These sentiments largely stem from underlying views about the more complicated question of how buildings should be designed in general. Some designers are noticeably defensive against computational methods for managing different design objectives, due to skepticism about the ability of computers and their programmers to fully understand the complexity of design, fear that optimization will be too deterministic, or because they see architecture as an expressive human endeavor and do not want to cede control. Others are more optimistic about computation but are frustrated that although computers have become ubiquitous for documentation and simulation, they are too limited in their ability to affect the iterative design process owing to a combination of technical issues and lack of imagination. Despite these disagreements, it is at least clear that advances in computation have challenged architecture and engineering schools and practices to wrestle with deep questions about the design process, and how to both teach and execute design synthesis.

Masdar City, Abu Dhabi. © Nathan Brown.

Somf 2016 structural engineering nathan brown final report 05

Since the fellowship topic of design for multiple objectives touches on such a fundamental question about design, one of the most directly useful outcomes is a collection of the different approaches to multiobjective design I have discovered during my research. This collection amounts to a somewhat loose categorization, and is understood to be both overlapping and incomplete, rather than exhaustive. A discussion about different design methods for multiple objectives will be included in the final chapter. These approaches have been collected through the building visits, design meetings, and research symposia enabled by the SOM Foundation fellowship, which I will also discuss briefly to bring up comments expressed by designers without reference to any building in particular.

First, however, I will present the structures and let them speak for themselves. Engaging with precedents is important for both students and practicing designers, and I hope to share what I have learned through this fellowship with both. I have already begun to use knowledge of these design examples in class lectures and design reviews to contextualize performance-based design, demonstrate the creative possibilities that multiobjective structural thinking can facilitate, and hopefully inspire new directions for architectural studios and engineering design projects. In this vein, giving clear descriptions of the structures of my SOM Foundation fellowship and discussing them in terms of design for multiple objectives is the most direct way to influence the practice and teaching of structural engineering in the future.

The selected design examples are broken into categories based on their main secondary objective aside from structure: daylighting, acoustics, energy, low carbon, and site. These categories sometimes overlap, and certain structures could easily fit into multiple categories. This categorization nevertheless highlights important and often dominant goals that interact with structure during the design process. Each structure is presented with basic background information, along with a discussion of its relevance to the topic of the fellowship.

Los Manantiales, Mexico City. © Nathan Brown.

Somf 2016 structural engineering nathan brown final report 06

Mexico and the United States

Boston
Washington, D.C.
Eureka Springs
Bentonville
Fort Worth
Seattle
La Jolla
Mexico City

Austria, France, Germany, Portugal, Spain, Switzerland, and the United Kingdom

London
Carmarthen
Bath
Cambridge
Norwich
Paris
Lausanne
Fribourg
Zurich
Linz
Munich
Stuttgart
Manheim
Cologne
Madrid
Guimarães

United Arab Emirates

Dubai
Abu Dhabi

Japan

Tokyo
Kyoto
Naoshima
Teshima
Osaka

Australia and New Zealand

Sydney
Melbourne
Auckland
Somf structural engineering nathan brown 2016 01

Nathan Brown
Massachusetts Institute of Technology
Department of Architecture

Nathan Brown

is currently a student in the Building Technology program within the Massachusetts Institute of Technology (MIT) Department of Architecture. He earned his undergraduate degree in Civil and Environmental Engineering at Princeton University in 2012, along with certificates in Architecture and Urban Studies. At Princeton, Brown received the W. Mack Angus Prize for an outstanding academic record and high achievement in engineering, the highest award in his department. His senior thesis on the geometry and structure of hypar shells was advised by Sigrid Adriaenssens. While at MIT, Brown has been advised by Caitlin Mueller as part of the Digital Structures research group, where he studies interactions between structural considerations and other architectural performance criteria in conceptual design. He has also worked with John Ochsendorf and the Structural Design Lab, an interdisciplinary research group in Civil Engineering and Architecture. Brown’s research in the area of computational multiobjective optimization for structural and architectural design has been presented at both design and technical venues, including conferences for the International Association of Shell and Spatial Structures, the Engineering Mechanics Institute, and the Boston Society of Architects. Brown plans to graduate in June 2016 with a Master of Science in Building Technology, specializing in structures. Upon completing the fellowship, he hopes to apply what he has learned about design and computation to a variety of structural engineering projects.