Modelling and optimizing triple-glazed supply-air windows using an original CFD approach

• A simplified and accurate CFD model of supply-air triple-glazing windows has been developed. • Model validated using full CFD and experimental data from in-situ and lab studies. • Model accuracy within 6% of full CFD, with computation time reduced by a factor of 1000. • Speed and precision enable extensive sensitivity analyses and optimization studies. • A parametric study reveals optimal thickness and flow rate for maximum thermal performance Supply-air windows are simple, low-tech and low-cost systems that significantly improve the thermal performance of building envelopes by addressing both wall heat losses and ventilation. Most studies about the performance of these windows are either CFD or simplified models that are less accurate or require preliminary CFD calculations to characterize the specific convective exchanges. This paper presents an original modeling approach based on a simplified 2D CFD approach that allows for very short calculation times compatible with dynamic thermal simulations. The model makes it possible to adapt the performance of these windows in real time and in automatic way according to their dimensions and flow-rate and can be used as design tool. The approach is based on a known velocity field based on Poiseuille's theory between two plates, which has been validated based on numerical results from 2D full CFD simulations and experimental results from literature. This new model showed very similar results with deviations < 6% for thermal conductance and < 2% for solar factor, and demonstrated that a coarse mesh reduced to just a few hundred cells was sufficient to obtain accurate results with calculation times reduced to a few seconds. Finally, this model made it possible to quickly perform many simulations, which were used to feed into a broad sensitivity study. These simulations ultimately enabled an optimization study to be carried out, demonstrating the existence of optimum values of air gap thickness and flow rate to maximize the triple-glazing thermal performances.