In this work, a simplified mathematical model, concerned with transient heat conduction as well as convective and radiative heat transfer, was developed to predict the variations of temperature and supercooling of the windshield during practical nocturnal cooling processes of a car. Final supercooling
was introduced as an indicator to evaluate the probability of occurrence of frosting. Following that, the Taguchi statistical method was used to conduct a parameter sensitivity analysis and then figure out the potential control strategies for frosting suppression. The results showed that relative humidity had the most significant influence on the distribution of supercooling during the nocturnal cooling period, whereas the initial temperature as well as the thickness and thermal conductivity of the windshield played a minor role in it. An increase in relative humidity resulted in a significant increase in
, which might be expected to trigger an earlier initiation of frosting. The emissivity of the windshield, concerned with the nocturnal radiation potential, showed a considerable effect on the response of
, whereas the influence of the total opaque cloud amount appeared to be largely limited. In addition, through a potential control of the thermal conductivity of the windshield,
just exhibited a very limited decline, thus contributing little to frosting mitigation. However, with a moderate potential control of the internal convective heat transfer coefficient, the frosting behavior might be effectively suppressed under a severe condition that favored the occurrence of icing. Besides, by introducing a combined control of the emissivity of the windshield and the internal convective heat transfer coefficient,
could be well reduced to a value below zero even as the relative humidity increased up to 90%, which was supposed to prevent the occurrence of frosting under a far severer condition.
