Skip to main content
Image
GSA - Notre Dame de Paris

Description : The Geometry–Structure–Architecture (GSA) Laboratory at ENSA Paris-Malaquais focuses on the intersection of engineering sciences and architecture. Since 1990, it has explored innovative structural, geometric, and material approaches, combining theoretical, historical, and experimental research to advance architectural design.

Une image contenant texte, Police, capture d’écran, blanc

Le contenu généré par l’IA peut être incorrect. 

Technical/scientific Challenge:

Standard calculation methods prove inadequate for a non‑standard construction composed of stone blocks and mortar joints, dating back to the 12th century for the oldest sections. Scientific insight was required to support the project management overseeing the restoration work.

Solution:

After developing their models and collecting measurements directly on site, the scientists used high‑performance computing to accurately simulate the deformations and stresses within the building. This intensive calculation required 25,000 CPU hours on the CRIANN supercomputer in 2021.

Business impact:

The block‑by‑block finite element modeling with interfaces, implemented in the CEA’s Cast3m software, allows for a highly detailed representation of the cathedral’s structure. By meshing even a quarter of a vault with over a million elements, the simulations provide precise insights into the building’s behavior, despite requiring six days of computation. The model realistically captures openings and observable cracks in the flying buttresses and the choir, offering valuable information for restoration and structural analysis.

Benefits:

  • High‑fidelity structural modeling of a 12th‑century cathedral, including stone blocks, mortar joints, and vaults
  • Ability to simulate complex deformations and stress distributions with over 1 million finite elements
  • Realistic representation of openings, cracks, and flying buttresses to guide restoration decisions
  • Intensive computation validated by 25,000 CPU hours on the CRIANN supercomputer, demonstrating the capacity for large-scale heritage simulations
  • Integration of site measurements into predictive models, bridging on-site data and numerical simulations