Stochastic analysis of time series temperature in battery cooling with ejector bubble generator

Catrawedarma, IGNB
Ton, Sefri
Fiveriati, Anggra
Astyanto, Achilleus Hermawan

Mechanical Engineering for Society and Industry
Vol. 6 No. 1 (2026): Issue in Progress

This study identifies flow patterns in the cooling channels and analyzes temperature during LiFePO4 battery cooling using an ejector bubble generator. The cooling fluids included quiet air, circulating air, and bubbles, with airflow rates ranging from 0.1 to 1.5 lpm. The temperature patterns were analyzed using probability density functions (PDFs), and Sample entropy—PDFs were quantified using mean, variance, skewness, and kurtosis. The results showed the formation of a slug film, an elongated slug film, and clustered bubbles on the bottom wall of the battery pack. There was a 3.75% and 5.98% decrease in the maximum temperature and thermal resistance, respectively, at an airflow rate of Qa = 1.5 lpm. The farther the thermocouple is from the bubble-generator nozzle and the greater the supplied airflow, the lower the PDF's kurtosis. The greater the airflow, the lower the entropy.