Experimental study of a modular cooling system for ultra-fast e-mobility charging cables using dielectric fluid

• This paper studies a subcooled flow boiling using the dielectric fluid. • A modular cooling system was suggested for ultra-fast EV charging cables. • This system accounts for cable inclination and asymmetric boiling behavior. • The maximum allowable charging current reaching 3257.99 A. • A new correlation was developed considering inclination effects. This study proposes a modular cooling system based on subcooled flow boiling using the dielectric fluid HFE-7100 to effectively dissipate high thermal loads generated in ultra-fast electric mobility charging cables. The system’s performance was experimentally evaluated under horizontal and downward inclination angles. To mitigate performance degradation-such as dryout from extended channel lengths, flow maldistribution, and localized overheating-common in single-channel designs, the cooling cable was divided into multiple modules with independently controlled cooling. Experiments were conducted on a single module at mass fluxes of 500, 700, and 900 kg/m²·s and inclination angles of 0°, 30°, and 60°, to analyze heat transfer characteristics under orientation changes encountered during charging. The influence of surface position (upper vs. lower) on heat transfer was also investigated. Results showed that increasing mass flux enabled higher heat fluxes, with the maximum allowable charging current reaching 3257.99 A-a 365.43% increase over conventional systems. Even at 500 kg/m²·s, the current reached 3071.76 A, a 338.82% improvement. The upper surface temperature was consistently higher than the lower. While inclination effects were negligible at high mass fluxes, they were significant at 500 kg/m²·s; at 30°, the maximum allowable current decreased by 22.36% from the horizontal case. Experimental data were compared with existing correlations; Moles & Shaw performed best (MAPE 17.89%). A new correlation was developed to address overestimation and inclination effects, reducing error to 4.75%. Under real-world conditions, the system enabled 1,263 kW of power, charging an 84 kWh battery in 3 min 59 s-72.3% faster than a 350 kW commercial charger.