Multi-scale analysis on thermal properties of cement-based materials containing micro-encapsulated phase change materials

Highlights

• FEM model was developed to predict thermal properties of mortar/concrete with PCM.

• The model was validated with experiments; it also agrees with analytical model.

• Multi-scale model can be applied for design optimisation.

• Thermal storage wall with optimum amount of PCM improves energy performance.

Abstract

The incorporation of phase change materials (PCMs) in building envelopes for passive thermal storage can enhance the thermal mass effect and thereby reduce energy consumption. In this investigation, multi-scale analysis of cementitious mortar and concrete containing microencapsulated PCM (MPCM) was performed experimentally and using numerical simulations. A three-dimensional two phase random composite model, which can be integrated with finite element method, was developed to predict the effective thermal properties of cementitious mortar and concrete with MPCM. MPCM was considered as inclusions in a continuous mortar matrix and the latent heat of PCM was incorporated into the simulations. The results showed that the effective thermal conductivity is strongly correlated with the volume fraction of PCM and is independent of the spatial distribution of the inclusions. These predictions were within the upper and lower bounds of parallel and series analytical models and agreed well with the experimental data (correlation coefficient 0.96 for concrete and 0.98 for mortar). Numerical simulations of the macro-scale behaviour of mortar and concrete with PCM for passive thermal storage showed a reduction in the maximum heat flux and time lag effect subjected to diurnal temperature variations. However, an optimum amount of PCM should be selected to fully exploit these passive systems. The developed models can be applied for optimising the design of composites to achieve the best thermal performance.

Introduction

Phase change materials (PCMs), which have high energy storage density per unit mass, can be used to improve the thermal mass effect of building envelopes [1]. Cementitious materials integrated with PCM has shown significant benefits in terms of improving thermal performance and reducing thermal stresses during hydration [2]. Enhanced thermal performance with improved specific heat capacity and reduced thermal conductivity was reported in different studies [3], [4]. Experimental and numerical studies have shown that PCM-integration improves the thermal comfort in buildings [5], [6]. Incorporation of PCM in external walls reduces the magnitude of diurnal indoor temperature fluctuation thereby improves the thermal comfort of the occupants [7]. Experimental and numerical study by Sayyar et al. [8] on PCM embedded gypsum wall board has shown reduced energy consumption while maintaining required thermal comfort levels. Combination of active and passive systems has suggested in some studies [9], [10] to enhance the thermal performance of PCM embedded systems. While cement-based materials containing PCMs have been shown to improve the thermal properties of concrete, adverse effects on mechanical properties have been reported in literature [11]. Meshgin and Xi [12] reported loss of compressive strength and suggested to use appropriate replacement levels in mix designs. Furthermore, method of containment of PCM can affect the hydration process and strength development in concrete [13]. Thus, it is vital to identify the optimum amount of PCM in cementitious mortar and concrete to achieve acceptable strength properties for building applications.

Recent developments of microencapsulated PCMs (MPCM) provide better heat transfer by creating finely dispersed PCMs with high surface area and direct integration with cementitious materials [14], [15]. The effective thermal properties of mortar or concrete with MPCM depend on several variables, including the mortar/concrete microstructure, PCM volume fraction, inclusion geometry, particle orientation, particle distribution, and the thermal conductivity of each component. Although the effective thermal properties are typically characterised experimentally, the evaluation of each variable is difficult and time-consuming due to complexity. When analytical and numerical approaches are both considered, they provide significant advantages in the evaluation of the effective thermal properties of a composite. In general, construction materials are multi-phase composites composed of two or more constituent phases with different physical, thermal and mechanical properties. The thermal properties of the constituent materials of mortar and concrete can also be analysed at a multi-scale level, ranging from the micro to macro scale. However, it is imperative to numerically simulate cement-based materials consist of PCMs at different scales to accurately predict and optimise thermal, mechanical properties. Due to material heterogeneity, predicting the effective thermal properties of multi-phase composites is a challenging task. Series and parallel models are the most widely used analytical models to predict the thermal conductivity of material [16], [17], [18]. A study of PCM concrete composite by Shi et al. [19] using series and parallel models assumed a well-aligned particle distribution, thereby not accounting for the heterogeneous nature of construction materials. A generalised self-consistent (GSC) model was used by Meshgin and Xi [20] to predict the effective thermal conductivity of PCM concrete. The presence of constituents with high contrast in properties can impact the accuracy of analytical model results [21]. Analytical methods based on homogenisation have limitations in terms of approximating the particle geometry, and the spatial distribution and orientation of the constituents [22], [23]. Thus, analytical methods are less likely to capture these complex characteristics of composites. To this effect, parametric studies should be conducted using numerical and experimental methods.

Different numerical methods including finite element methods (FEMs) have been used to study effective homogenised properties of composites with high accuracy [24]. Capability of handling irregular geometries and simulation of wide engineering applications are some of the main advantages in FEMs [25], [26]. Multi-physics analysis related to developed FE models on heterogenous materials namely on thermal, mechanical, thermo- mechanical properties of composites is reported in some studies [27], [28]. The computational method for homogenisation often uses the concept of a representative volume element (RVE) due to modelling complexities [29]. RVE (or unit cell) is regarded as a volume of a heterogeneous material that is sufficiently large to capture all the statistical variations in the composite material [30], [31]. FE analysis on heterogenous materials was used to analyse the effect of matrix properties on effective thermal properties of the composite [32], [33]. Floury et al. [34] studied particle geometry, orientation, volume fraction and particle properties on effective thermal conductivity of theoretical composite material. Transient heat transfer during phase change of PCM is often represented using enthalpy method or heat capacity method [35]. Numerical simulation of natural convection during phase change was studied by Amin et al. [36], [37] and is suitable for spherical inclusions. However, numerical studies on the effective thermal properties of mortar and concrete containing MPCM are scarce and model validation is uncommon [38]. Thus, it is vital to develop numerical tools in predicting the effective thermal properties of cementitious materials with MPCM composite to identify the optimum parameters that govern their performance in applications. This method will provide more flexibility where the relative effects of each variable affecting the critical properties can be evaluated individually.

This investigation aims to develop a numerical model to predict the effective thermal properties of cementitious mortar and concrete containing MPCM. A three-dimensional two phase random composite model was developed to analyse the effect of the spatial distribution, geometry and volume fractions of inclusions in a continuous matrix. The model focuses on different hierarchical stages; mortar matrix with MPCM inclusions, Mortar and MPCM matrix with coarse aggregates inclusions. Results from the extensive experimental program conducted in this investigation, as well as analytical models found in the literature, were used to validate the developed numerical model. Furthermore, a macro-scale passive thermal storage wall was simulated to evaluate the thermal performance of the system in building applications. The developed multi-scale model links the constituent properties and the effective thermal properties of the composite. Therefore, it can be used to optimise the design of cementitious materials with PCM to achieve better thermal performance with acceptable strength levels for building applications.