Despite being such a common fixture in our built environment, the simple pre-engineered metal building (PEMB) system is often overlooked by architects. Why aren’t more architects using this system? Is it too limiting? Is not creative enough? These questions inspired us to team up on a research project, Parametric Designed Metal Buildings or PDMB Variations, to try and answer the question, “What other forms can metal buildings take?”
A parametric-based research project challenges the standard forms of PEMB
“Metal buildings are the dream that modern architects had at the beginning of the century come true, but they themselves don’t realize it. If they followed their own theories to the letter—form follows function, using mass production techniques to make cheap things with no frills—what you end up with is a metal building!” – David Byrne, “True Stories”
The team’s research methodology combined Roddy’s obsession with designing metal buildings and Haines’ passion for technology and computation to literally push the envelope. The research was structured on three key objectives:
- To advance the formal and aesthetic potential of the typical PEMB by using parametric modeling to generate formal variations from standard components.
- To provide variations on the standard PEMB sizes and profiles to create more dynamic forms, while maintaining the optimized structure, components and single-skin envelope.
- To create a data-rich, generative and qualitative design tool that can produce and evaluate an extensive catalog of design variations, calibrated to the specific needs of the consumer.
Initially, our curiosity was in unique older metal buildings on angular sites, typically located near freeways and railroad rights-of-way. These non-traditional trapezoidal building footprints caused us to wonder, “Why don’t we see more angular and irregular footprints in metal buildings?” The PEMB is a highly optimized system that uses standard components such as rigid frames, wall girts and roof purlins. These components are fabricated in a manufacturing facility then delivered and erected on-site. We felt that more could be explored in terms of this system’s form using the standard components and framing elements.
Utilizing only the rudimentary toolbox of PEMB standard components—frame, girt and purlin—we began to sketch and iterate formal ideas that were more dynamic than traditional PEMB boxes (Figure 1). Iterations resulted in similarities which enabled us to begin to categorize them based on emerging patterns and groupings. We continued the iterations to see just how many design variations could be generated. The sketching only took us so far in our exploration, so we continued by identifying parameters for three angular formal ideas we thought had the most potential: chamfer, cross-cut and compress.
Figure 1. Iterative sketches tested ideas of angular and dynamic geometric possibilities for PEMBs without introducing warped surfaces.
Using 3-D modeling and parametric software platforms Rhino and Grasshopper, we built parametric models for our exploration. Each parametric model comprises some combination of each of the following three primary properties (Figure 2):
- Formal operation—chamfer represents cut corner conditions, cross-cut represents a singular cut across the entire building or compress represents a perpendicular force pushing into the form.
- Building profile—symmetrical, asymmetrical or shed.
- Building size—small (30 feet by 60 feet), medium (80 feet by 100 feet) or large (100 feet by 200 feet).
Figure 2. The parametric modeling process used primary inputs of varying formal operations, building profiles and building sizes.
To standardize our modeling inputs, we established consistent rules for spacing of the rigid frames and secondary structural components. A plan grid with point coordinates of the structural components was established for each building size (Figure 3). Grasshopper was used to create the various 3-D forms, ultimately generating a total of 1,444 iterations based on the combinations of operation, profile and size. Each combination (i.e., chamfer/shed/30 feet by 60 feet) was subsequently catalogued and documented.
Figure 3. A plan point grid with standardized structural spacing shows the incremental design possibilities within the parametric model.
The resulting digital catalog allows for aesthetic evaluation and development of potential building designs (Figures 4 and 5).
Figure 4. The visual catalogue of the 1,444 variations provides inspiration for potential building ideas based on form and size.
Figure 5. A case study of a 100-foot by 200-foot PEMB dynamic transformation using the formal operations cross-cut and compress.
The data generated allowed us to test a performance criterion. The potential of this tool can integrate computational design allowing for metrics within the models to be analyzed and ideal prototypes selected (Figure 6). The digital catalog and data-rich environment we created demonstrates the potential of a more dynamic and performative metal building design process.
Figure 6. An example of a cloud-based data visualization platform, Thread, by Thornton Tomasetti’s Core Studio was used to explore data of the variations.
We believe that pre-engineered metal buildings to date have not lived up to the potential and promise of modernism. PDMB Variations offers a provocation for new possibilities of this banal typology by challenging the standard forms of PEMBs and exploiting the inherent intelligence of its highly optimized system to enhance design creativity through innovation.
We invite you to see our complete PDMB Variations research document through this link: www.metalarchitecture.com/PDMB_Variations.
Mark Roddy, FAIA, is a lecturer within the Department of Design at California State University, Sacramento. Formerly a partner and design director with SmithGroup, Roddy is recognized as a prolific designer and thought leader. Roddy was inducted into the American Institute of Architects’ prestigious College of Fellows in 2014. His projects have won more than 60 design awards, including a 2008 Metal Architecture Design Award for Metal Buildings, for an addition to a historic home in downtown Phoenix.
Ryan Haines is an architectural and computational designer at SmithGroup. He is a member of SmithGroup’s National Computation Committee, representing the Phoenix office and finds ways to apply computational thinking to projects at a variety of scales ranging from The San Francisco Ameneties (Public Toilets), The University of Arizona Applied Research Building and the UCSF Wayne and Gladys Center for Vision.