A cloud-based platform to predict wind pressure coefficients on buildings.

Area of Science:

• Building Science and Engineering
• Computational Fluid Dynamics (CFD)
• Sustainable Architecture

Background:

• Natural ventilation (NV) is crucial for energy-efficient buildings and indoor air quality.
• Accurate modeling of NV requires reliable wind pressure coefficients (Cp).
• A lack of accessible tools for predicting Cp on diverse building geometries limits NV application in design.

Purpose of the Study:

• To introduce a novel cloud-based platform for predicting wind pressure coefficients (Cp) on buildings.
• To automate the computational fluid dynamics (CFD) simulation workflow for Cp data generation.
• To provide a reliable and practical data source for natural ventilation (NV) modeling in architectural design.

Main Methods:

• Development of a cloud platform integrating automated CFD simulations.
• Automated generation of computational domains, meshing, solution, and post-processing.
• Validation against extensive wind tunnel experimental data for various building types and urban settings.

Main Results:

• The platform successfully predicts wind pressure coefficients (Cp) with high accuracy.
• Results show close agreement with wind tunnel experimental data across diverse case studies.
• Prediction error for openings was reduced by 60%-77% compared to EnergyPlus software.

Conclusions:

• The developed cloud platform is a reliable and practical tool for obtaining Cp data.
• It significantly overcomes limitations in current NV modeling for real-world building design.
• The automated CFD approach enhances the integration of NV strategies in sustainable architecture.