CFD-CFD coupling: A novel method to develop a fast urban microclimate model

Abstract

Computational fluid dynamics (CFD) technique is well-known for its powerful capability of modelling the cross-ventilation. CFD is broadly coupled with energy simulation programmes, such as EnergyPlus, to expand its application in building energy calculation. During the coupling process, iterative calculation is required to improve the accuracy of the coupling method, which means the scale of the model should be designed with extra care so that the simulation can be practical in terms of computational cost. A coarse-resolution mesh (here denoted as CFDc) cannot properly represent the required local environment to study the underlying physics. However, a fine-resolution CFD microclimate model (here denoted as CFDf), which includes both indoor and outdoor spaces, always requires a large number of cells to capture the complexities of environment; this makes the coupling procedure a challenging problem. This study aims to propose an innovative, high-resolution and computationally cost-effective CFDf-CFDc model to overcome this gap. The full-scale CFDf works in an off-line manner at the preliminary simulation stage and generates detailed flow parameters at the interfaces for CFDc. At the dynamic stage of the simulation, CFDc participates in the coupling procedure with the other tools, for example, building energy simulation (BES) tool (e.g. EnergyPlus). The coupling of CFDf-CFDc-BES is about 200 times faster than an ordinary coupled CFDf-BES method with a significantly quicker and more stable achievement in the convergence.

Fast and dynamic urban neighbourhood energy simulation using CFDf-CFDc-BES coupling method

“…More details of boundary type analysis and the comparison of the flow characteristics' modelling by CFDf-BES and CFDf-CFDc-BES methods have been reported in Ref. [40]. Figure 6 shows RMSE of the normalized local velocity of these lines over the meteorological wind speed.…”

Comparison between CFD simulation and affinity laws and analysis of the degree of reaction.

“…Computational fluid dynamics (CFD) is the most current technology in the study of fluid flow used in many applications (Sathish & Mohanavel, 2018;Wehinger, 2019;Zhang & Mirzael, 2020;Chavan, Vitankar, Shinde, & Thorat, 2021). Experimental methods of turbomachinery performance prediction involve greater consumption of time and financial resources compared to the CFD approach, which has been widely used in fluid flow simulation in turbomachinery (Abdelmadjid, Mohamed, & Boussad, 2013;Danish, Khan, Umer, Qureshi, & Ma, 2014;Liu, Wang, Wang, Huang, & Jiang, 2014;Nataraj & Singh, 2014).…”