Peak Cladding Load Predictions
by Giacomo Lamberti; Zhu Huang; Mattia Ciarlatani
Motivation and Objective
Accurate estimates of mean and peak pressure distributions on buildings facades are essential to assess wind hazards. Our research aims to quantify, and improve, the accuracy of CFD results for cladding loads on high-rise buildings. The uncertainty in the CFD prediction of wind pressures on buildings arises primarily from two sources:
- uncertainty in the inflow boundary conditions representing the incoming atmospheric boundary layer, and
- uncertainty in the CFD model related to model choices such as the turbulence or subgrid model and wall model.
This study aims to establish methods to evaluate and quantify both types of uncertainties, and to validate the methods with available test data for a high-rise rectangular building.
Methods and Results
Initial analysis indicated that in Reynolds-averaged Navier-Stokes (RANS) simulations, the turbulence model uncertainties are dominant, and they are too large to meaningfully inform design pressures on cladding panels. In contrast, large-eddy simulations (LES) were found to have limited uncertainty due to the subgrid model and wall model (provided a sufficiently fine mesh is used), but the results are highly sensitive to the inflow conditions. Hence, the main focus of the research has been on validating LES, including an investigation of the inflow uncertainty in LES, and an exploration of multi-fidelity methods that could reduce the cost of an LES analysis.
We coupled a divergence-free digital filter inflow condition with an optimization method that enables tailoring the inflow conditions to obtain the desired turbulent wind flow at the location of the building in the simulation. This method has supported accurate validation and sensitivity analysis, demonstrating the predictive capability of LES, as well as establishing the importance of carefully documenting the incoming wind field in wind tunnel experiments used for validation. The figure below shows peak wind pressure coefficients near the downstream top corner of a high-rise building, comparing two wind tunnel tests and one of our LES results.
The video below shows iso-contours of Q-criteria around the time of a peak pressure events, indicating how the interaction between a large vortex advecting along the side facade and small vortices transported across the roof cause significant suction peaks.
With the predictive capability of LES demonstrated, the next objective is to speed up a design analysis using multi-fidelity methods. Initial analysis combining many RANS evaluations, for all wind directions with a 10 degree resolution, with a few LES at a smaller number of wind directions indicates this is a promising route for enabling the routine use of CFD for engineering design.
Related Publications
Lamberti, G., & Gorle, C. (2021). A multi-fidelity machine learning framework to predict wind loads on buildings. Journal of Wind Engineering and Industrial Aerodynamics, 214.
Lamberti, G., Amerio, L., Pomaranzi, G., Zasso, A., & Gorle, C. (2020). Comparison of high resolution pressure measurements on a high-rise building in a closed and open-section wind tunnel. Journal of Wind Engineering and Industrial Aerodynamics, 204.
Lamberti, G., & Gorle, C. (2020). Sensitivity of LES predictions of wind loading on a high-rise building to the inflow boundary condition. Journal of Wind Engineering and Industrial Aerodynamics, 206.
Conference Presentations
Acknowledgements & Resources
This material is based upon work supported by the National Science Foundation under Grant Number 1635137. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. The simulations were performed using the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation grant number CI-1548562.