01 / Computational Design

The Dual Reality of Topological Performance.

Topology, morphology and fabrication in discretised free-form shells.

Date
January – March 2026
Context
UCL / Digital Ecologies
Group project
Methods
Python, force-density form-finding, Karamba3D, K-means, regression
Catalogue of twelve boundary geometries across five shell topology patterns
Boundary geometries and topology patterns / comparative shell catalogue

How topology shapes performance

This study compares beam-based gridshells and plate-based continuous shells across twelve boundary geometries, five topology patterns, and three levels of planarity deviation. A custom Python workflow connects topology generation, force-density form-finding, geometric evaluation, and Karamba3D structural analysis.

K-means clustering and multiple linear regression are used to investigate the relationship between geometric organisation and structural performance. The findings inform an application to the British Museum Great Court roof, connecting quantitative evaluation with topology selection and fabrication decisions.

Project documentation

04 PLATES
Topology generation and force-density form-finding
01 / Topology generation and force-density form-findingView full size ↗
Planarity, tortuosity and finite-element evaluation
02 / Planarity, tortuosity and finite-element evaluationView full size ↗
Dataset analysis, clustering and regression
03 / Dataset analysis, clustering and regressionView full size ↗
British Museum Great Court test and design implications
04 / British Museum Great Court test and design implicationsView full size ↗