Statistical analysis of heat transfer with non-isothermal melting in 116 commercial PCMs using the apparent heat capacity method in ANSYS Fluent

Calorimetric measurements reveal that technical-grade phase change materials (PCMs) often exhibit complex phase change behavior, which is inadequately captured by the linear transition models commonly used in CFD software. This study statistically evaluates how variability in phase change behavior affects PCM melting. Numerical simulations are performed in ANSYS Fluent for 116 commercial materials, considering two-dimensional melting in a rectangular enclosure, using the apparent heat capacity (AHC) method with phase fraction–temperature curves derived from PCM heat capacity data. To isolate the effect of phase change behavior, other PCM thermophysical properties such as nominal melting temperature, latent heat and density were artificially homogenized, thereby eliminating additional sources of variability. The time to reach 80% state of charge (SoC) serves as a key metric for melting performance. Results show that variability in the phase fraction–temperature curves is substantial. Relative standard deviation values of up to 19.4% in melting time were observed under small temperature driving forces. Notably, statistical analysis shows that the mean melting time of the 116 commercial PCMs is shorter than that of a near-isothermal PCM, with a confidence level above 99.9%. Shape parameters of the characteristic peak in the AHC curve, including peak position and asymmetry, are identified as the primary contributors to the variability in melting performance. A positive correlation is found between the center temperature of the peak and melting times. Therefore, the center temperature is recommended over the PCM peak temperature for characterizing commercial PCMs, as it reflects the uneven distribution of latent heat around the nominal phase change temperature. • A statistical numerical analysis is conducted on the melting of 116 commercial PCMs. • Numerical models are implemented in ANSYS Fluent to compute PCM non-isothermal melting. • Calorimetric data of temperature-dependent PCM properties is used for the analysis. • The liquid fraction–temperature curves of the PCMs studied show significant variability. • The variability in liquid fraction–temperature curves affects relevant engineering KPIs.