Determining feasible design parameter domains of impinging jet ventilation and passive chilled beam combined systems using an integrated CFD-data-driven framework

Ren, Tao, Sun, Meiyu, Mo, Shentong, Gu, Yuqian, Cheng, Jin, Hu, Jun, Kang, Yanming and Zhong, Ke 2026. Determining feasible design parameter domains of impinging jet ventilation and passive chilled beam combined systems using an integrated CFD-data-driven framework. Building and Environment 305 (PartB) , 115277. 10.1016/j.buildenv.2026.115277

Full text not available from this repository.

Abstract

Impinging jet ventilation (IJV) and passive chilled beam (PCB) combined (IPC) systems have been recognized as an efficient air distribution strategy for public buildings. However, the feasible design parameter domains of IPC systems under varying cooling load conditions remain unclear because the system performance is jointly affected by IJV supply parameters (vs, Ts) and the PCB cooling ratio (η). An integrated framework combining CFD simulations, Sobol sensitivity analysis, a backpropagation neural network (BPNN), and the non-dominated sorting genetic algorithm III (NSGA-III) is proposed to determine the feasible design parameter domains. The results indicate that, with increasing cooling loads from 100 to 250 W/m2, the indoor airflow pattern changes from a downward cold jet dominated flow to a buoyancy-driven turbulent flow, resulting in heat and contaminant accumulation in the occupied zone. Accordingly, the feasible design parameter domains initially expand and subsequently shrink, accompanied by a shift toward higher supply velocities and larger η. Sobol sensitivity analysis reveals that η exhibits significant nonlinear interactions with energy efficiency (ET) and ventilation efficiency (ε), indicating its critical role in improving system performance. A BPNN surrogate model of IPC systems is established, and prediction accuracy with R2 > 0.98 is achieved for eight key performance indicators. Finally, the optimal operating solutions are obtained using NSGA-III combined with entropy-weight TOPSIS. Validation results confirm that the optimized solutions effectively alleviate overcooling and overheating under different cooling load conditions, maintain draft dissatisfaction below 20%, and achieve a balanced improvement in thermal comfort, indoor air quality, and energy efficiency.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Schools > Architecture
Publisher:	Elsevier BV
ISSN:	0360-1323
Date of Acceptance:	14 September 2026
Last Modified:	28 Sep 2026 12:45
URI:	https://orca.cardiff.ac.uk/id/eprint/189841

Actions (repository staff only)

Edit Item