Simplified models for predicting thermal stratification in impinging jet ventilation rooms using multiple regression analysis.

Inappropriate thermal stratification characteristics can degrade the ventilation performance of impinging jet ventilation (IJV). In the present study, an analysis method using dimensionless parameters is proposed for predicting the thermal stratification, and the Box-Behnken design method is employed to determine the simulation cases of influencing factors for thermal stratification. The thermal performance in IJV cooling room is investigated by computational fluid dynamics (ANSYS Fluent), and the effects of supply air parameters, inlet area (S *) and height (h *) on thermal stratification are discussed, respectively. A further supply air parameter is represented by the thermal length scale (L m), which is applied to evaluate the combined effect of inertial force and thermal buoyancy on supply airflow. The results show that the driving force on supply airflow changes from buoyancy dominated to inertial dominated with increasing L m , leading to the thermal stratification profiles changing from thermal stratification to uniform distribution. For small L m , the thermal stratification profiles are similar for different inlet areas, while the dimensionless air temperature decreases with increasing inlet area. However, the larger inlet area promotes a more uniform temperature distribution for large L m. In addition, the thermal stratification profiles are similar regardless of the inlet height for both small and large L m conditions, indicating that the dimensionless air temperature is a weak function of inlet height. Finally, multiple regression models related to L m , S * and h * are developed for predicting the thermal stratification, which can provide a reference basis for the determination of temperature distribution in IJV cooling rooms. ● Thermal stratification (TS) of impinging jet ventilation (IJV) is studied. ● L m is used to evaluate the combined effect of inertial and buoyancy forces on TS. ● Prediction models of TS are developed for IJV cooling rooms (θ ank , Δ θ oz and Δ θ fc). ● Effect of L m on Δ θ fc is less than that on θ ank and Δ θ oz. [ABSTRACT FROM AUTHOR]