Numerical model for determining the effective heat capacity of macroencapsulated PCM for building applications

Abstract

This paper presents a finite difference model of macroencapsulated PCM panels coupled with the genetic algorithm for the determination of effective heat capacity of whole panels via inverse method. This provides an accurate characterization of the thermal properties of macroencapsulated PCMs for building envelope applications. A novel definition of the effective heat capacity is proposed based on the superimposition of two Gaussian curves, applicable to any PCM whose phase transition is characterized by a single peak. Three PCMs were tested, subjected to temperature variation rates typically experienced in building envelopes: 0.5 °C/h and 1 °C/h. Surface temperature and heat flux were measured and used in the inverse method procedure. The developed model is accurate, as numerical results greatly agree with the experiments: the root mean square difference between the experimental and numerical heat fluxes ranged between 0.543 and 1.246 W/m2. Significant differences in the effective heat capacity were found between the whole macrocapsule and small quantities of PCM (specified in the datasheets). The effective heat capacity specified in the datasheets is sensibly greater than that of the whole macrocapsules determined through the inverse method: the specific heat in the solid phase was up to 107.39 % higher in the datasheet values, the specific heat in the liquid phase up to 184.04 %, and the peak effective heat capacity, between 18.28 % and 164.11 %. The same happened to the enthalpy: datasheet values were 61.24 % – 175.55 % greater than inverse method results. This proves that latent heat is overestimated if small quantities of PCM are analyzed, and not the whole panels. The scale effect was assessed by comparing two capsules with the same material, but with different quantities of PCM: 0.5 kg and 1 kg. A greater mass of PCM over the total mass of the capsule implies a different relationship between the effective heat capacity and temperature, with higher peak heat capacity. The capsule with 1 kg of PCM showed a peak effective heat capacity up to 30.65 % greater than that of the panel with 0.5 kg of PCM. Thus, adequate modeling in building applications requires characterization of whole macroencapsulated PCMs. The determination of the relationship between temperature and effective heat capacity of macroencapsulated PCMs presented in this work could easily be incorporated into other simulation software, facilitating the assessment of adaptive envelopes with PCM macrocapsules.