Decoding Details: Integrating Physics of Assembly in Discrete Element Structures

SMArchS Thesis 2016
Advisor: Caitlin Mueller; Readers: Brandon Clifford, Larry Sass

Architecture is intrinsically the coordination of parts to form a whole, and the detail is the critical point where this coordination is resolved. Between technical and perceptual constraints, details are geometrical solutions and organizational devices that negotiate physics, construction, assembly, materials, fabrication, economy, and aesthetics, all at once.

Over centuries, detail formulas have been created, tested and revised by builders, architects, engineers and fabricators; collected in catalogs and magazines, they have been usually documented in two-dimensional sections that silence all intervening forces. While masters with knowledge in construction and materials are able to iterate through different possibilities creating novel details, usually less experienced designers can only reproduce standard solutions. In the era of digital design and fabrication, in which material and building information can be parametrically linked and massively computed, can we challenge what we can build with a new way of looking at details?

This thesis introduces the concept of synced detailing, where conflicting constraints are resolved in the details. As a case study, stability and assemblability are studied on a structurally challenging discretized funicular funnel shell. The goal is to eliminate scaffolding during assembly using only joint details. Finite element (FE) analysis is performed at every step of the assembly sequence to show global and local instability. Local translation freedom (LTF) analysis shows the range of feasible assembly directions. Detailing knowledge is studied and encoded in shape rules to create a detail grammar. Real time visual feedback of the constraints informs the designer to apply these rules to create joints that satisfy across a range of priorities. This method is generalizable for other constraints, allowing architects to create novel solutions informed by quantifiable analysis and encoded knowledge.