A numerical study on the performance of a three-dimensional solar chimney with a combined CFD – ANFIS method

In the studies of solutions for green buildings, the Computational Fluid Dynamics (CFD) method has been widely used, particularly for solar chimneys. Performance of a solar chimney for natural ventilation or cooling can be assessed numerically in terms of the induced flow rate and temperature rise through the chimney. As the CFD technique has been employed successfully by many researchers for small-scale solar chimneys, applications of this method to large-scale systems, particularly three-dimensional (3D) ones are extremely difficult due to computational cost. As an attempt to apply the CFD method to 3D solar chimneys, in this work, we built an Artificial Intelligence – based ANFIS (model) to predict performance of a simple 3D vertical solar chimney. The ANFIS model was trained with the data generated by an CFD model for various applied heat fluxes. The trained ANFIS model was then employed to predicted performance of the chimney at other heat fluxes. The results show that the ANFIS model is able to predict well the induced flow rate and temperature rise through the chimney with the maximum difference with the CFD results of less than 5.0%. Therefore, this study shows a possibility of combining CFD and ANFIS methods for large-scale solar chimneys.