2008 Structural Engineering Traveling Fellowship
Aesthetics and Physical Models

Physical models have vast potential as visual, analytical, and design tools for the study and creation of aesthetically pleasing, technically challenging, efficient structures. Models are an ideal way to develop structural solutions through a balance of discipline and play. [1] Discipline as it pertains to structural design is the understanding of structural behavior and economic and material constraints, whereas play represents an exploration of form. For elegant forms to be efficient and economical to build, discipline should accompany play.

The Swiss engineer, Heinz Isler, engaged in a design process that exemplifies this balance; his process involved generating a series of models from which the form of the full-scale structure followed. In addition, Isler viewed completed works as opportunities to reflect on the process of design as well as a means of checking that the predicted results were reasonably achieved. Isler states that form-finding is only “the first link in a whole chain of investigations and the other links in the investigation are model tests, measuring of the first structure, 1:1 as we have it out [t]here; these are of primary importance.” [2] Isler’s process is tied to a physical understanding of behavior, and the Sicli Factory in Geneva, a free-form shell with seven supports, clearly demonstrates his methodology. To develop the complex form for the Sicli Factory, Isler created a series of hanging models, small-scale experimental models from which he could examine shell stresses and buckling capacity, and an architectural site model.

Concerns about long-term behavior led Isler to monitor the deflections of the Sicli Factory for almost twenty years after its completion. [3] Performance of the completed structure, not the complexity of the analysis, dictates a structure’s success. Isler, like other great structural artists relied on physical intuition gained from first-hand observation.

Today most physical modeling has been replaced by computer modeling. While the computer is an invaluable tool, it can compromise the development of one’s intuition about how structures behave. This report identifies how physical models have been used as visual, analytical, and design tools to create aesthetically pleasing, technically challenging, efficient structures both before and since the emergence of computer modeling. Physical models as complements to computer models still benefit practitioners today and promote more creative and rational form generation among college students as they prepare to become structural designers.

Conclusion

This report identified how physical models have been used as visual, analytical, and design tools in the past and suggested that despite the emergence of computer modeling, physical models can still benefit practitioners today. After the successful completion of a structure an engineer might commission a display model of the work. While display models may draw public attention to and allow engineers to reflect on their works, using visual models and prototypes during design can stimulate ideas and lead to unexpected changes. The physical model of the Severin Bridge is an exhibition model, but a similar model used early in the design process can give feedback about a structure’s proportion and relationship to its site. Computer graphics are also important in visualizing and refining a structure’s overall form and details and should complement physical models where modifications need to be realized quickly. Structural engineering like architecture is a visual art; generating study models and performing visual critiques should be as important in engineering as it is in architecture.

In contrast to visual models, which are more widely used by architects than engineers, analytical models have long been embraced by engineers for analyses too complicated or time consuming to perform by hand. Computer modeling has replaced physical modeling for well understood problems such as statically indeterminate network-tied arch and cable-stayed bridges, but not for predicting a structure’s wind response. For example, the Swiss Re Headquarters in London, completed in 2004, was tested by Rowan Williams Davies & Irwin (RWDI) in a wind tunnel. Computational fluid dynamics (CFD) continues to improve and give results closer to those observed physically. It is conceivable that CFD will eventually be reliable and quick enough to replace physical modeling. However, new forms and new methods of analysis will arise and require verification by physical means first.

Even the most well-conceived physical and computer analytical models can only approximate behavior and should complement simple hand calculations. If analysis is at best an approximation than some forms are easier to approximate than others. David Billington, when referring to the wide range of visually interesting shell forms that were generated by individuals who relied on and developed simple rather than complicated analysis techniques, notes that shells “carry forward the central scientific idea in structural art: the analyst of the form, being also the creator of the form, is free to change shapes so that complexity disappears.” [4] Here Billington does not mean that structures should only take on well-studied forms, but rather is emphasizing that engineers rely on experience to avoid unnecessary complications when creating new forms. Form-finding models like hanging chains and membranes and soap film take on rational forms and suggest how their full-scale equivalents will behave. These models may be of vastly different materials and scale than the final structure and therefore appear more like an abstraction than a facsimile. Engineers still need to exert judgment when assessing the validity of a form-finding technique, but the models are meant as a point of departure from which additional visual and analytical models are developed and used iteratively to inform the final design of a structure. Today, computational form-finding has replaced most physical form-finding and while physical form-finding is still informative as a means of developing intuition, physical modeling has a more direct benefit for the design of moveable and adaptable structures.

Physical models as complements to computer models have potential as visual, analytical, and design tools. At the small-scale one can generate and refine forms while approximating behavior. However, only at the full-scale can one make a proper aesthetic critique and confirm that the form safely carries the loads acting on it. For Isler, studying his completed works was an important part of his design process. Isler viewed each structure as an opportunity to reflect. This reflection is another form of discipline or striving to understand behavior and material constraints that informed his play or his future exploration of rational form. Today many of Isler’s shells like the BP Gas Station completed in 1968 and Bürgi Garden Center completed in 1971 are still used and are in excellent condition proving not only the success of these structures, but also the success of Isler’s process.

Isler, like other great structural artists relied on physical intuition gained from first-hand observation. Today most physical modeling has been replaced by computer modeling; however, computer modeling can compromise the development of one’s physical intuition. Because practitioners are more likely to use computer models than physical models, this physical intuition has to begin to be developed in college through the use of physical visual, analytical, and design models. Many institutes both abroad and in the United States already incorporate visual and analytical models into their curriculums, but more emphasis should be placed on creating design models as a means of generating more creative and rational forms and understanding the balance of discipline and play required to create efficient, economical, and elegant structures.

Student visits to built works and with practitioners should complement design exercises. In Europe, some practitioners are also full-time university faculty providing students with a mix of theory and practice in the classroom. However, in the United States while this is typical for architects it is not as common for engineers. As a result, universities in the United States usually emphasize theory and analysis rather than design. While this approach may teach students about fundamental structural behavior and how to analyze a particular set of members, it does not inspire creativity or elicit thinking about globally efficient forms. Visiting aesthetically pleasing built works and meeting with leading practitioners reassures aspiring engineers that the potential for new, visually interesting, and efficient forms exists. Physical models can help students and practitioners realize this potential.

Notes

[1] Concept presented in David P. Billington, The Tower and the Bridge (Princeton: Princeton University Press, 1983), 213–32.

[2] Heinz Isler, “New Shapes for Shells,” Bulletin of the International Association for Shell Structures 8 (1961): 123–30.

[3] John Chilton, The Engineer’s Contribution to Contemporary Architecture: Heinz Isler (London: Thomas Telford Publishing, 2000), 99.

[4] Billington, The Tower and the Bridge, 20.

Denmark, Norway, and UK

Germany

Switzerland

Spain