ACI MATERIALS JOURNAL

Title no. 109-M08

TECHNICAL PAPER

Effect of Phase-Change Materials on Properties of Concrete

by Pania Meshgin and Yunping Xi
This paper presents the results of an experimental investigation of using phase-change materials (PCMs) in portland cement concrete. The objective of this study is to improve the thermal properties of concrete as a structural material. A compression test, flexural test, drying shrinkage test and thermal conductivity test were conducted. PCMs are used in concrete mixtures as both sand replacement and additives. The results revealed that in the replacement method, the loss of compressive strength due to the addition of PCM is not as high as the additive method. For up to 20% sand replacement by PCM, the strength loss is not significant for structural applications. The specific heat of the concrete increased considerably. Thus, the thermal conductivity of the concrete is reduced and the insulation capacity of the concrete is improved. A flexural test, drying shrinkage test, and microstructure analysis provide a good understanding of the PCM-modified concrete. Overall, the results show that it is quite promising to use PCM in concrete to improve its insulation capacity with decreased thermal conductivity and, at the same time, that it is possible to keep the strength loss of the concrete in an acceptable range.
Keywords: compression; drying shrinkage; flexural; phase-change material; thermal conductivity.

INTRODUCTION Mechanical properties such as strength and stiffness are considered important material properties for structural materials including concrete. Recently, the thermal insulation and heat storage capacity of the structural materials have also been considered important material properties for the sustainable development of energy-efficient buildings. Specifically, how to improve the thermal insulation capacity of concrete materials becomes an important research topic. One of the most effective methods to improve the thermal resistance of concrete is to use phase-change materials (PCMs) as an additive. PCMs have high heat of fusion1 and, more importantly, the melting point of PCM is within the range of ambient temperature. The applications of PCM as an additive were originated from materials other than concrete. These applications provided inspiration for the concrete industry. Many researchers2,3 have studied different PCMs and their applications. Hawes et al.2 and Sharma et al.4 classified PCMs into three categories: organic, inorganic, and eutectic. Hawes et al.5 have shown that by adding a proper type of PCM in a material, the thermal storage capacity of the material can be enhanced. They have shown that organic PCMs, due to their compatibility with the building materials, would be the best choice for application in building envelopes. Chen et al.6 used a specific type of PCM made of small particles (called microencapsulated PCMs) embedded in energy-storing wallboard. Their results indicated that by including a proper amount of PCM in the inner surface of the wall, the thermal comfort of the building can be enhanced dramatically. Richardson and Woods7 analyzed the reduced unit weight of insulation materials when PCMs are used in the material. They concluded that the reduced unit weight by using PCM provided a substantial benefit. Li and ACI Materials Journal/January-February 2012

Li8 investigated the thermal performance of granular PCM composite, which was made of a granular porous materials and PCMs, and the result was quite promising. These are the applications of PCM in composite materials other than concrete. In recent years, PCMs have also been used in concrete. For instance, Cabeza et al.9 constructed a small house using concrete incorporating PCMs. Zhang and Zhang10 studied the effect of PCM on the thermal property of mortar. Most recently, Hunger et al.11 performed some experiments on self-consolidating concrete and showed how the porosity of the concrete was increased by using PCM, which implies that the strength of concrete with PCM may be reduced. Bentz and Turpin12 investigated the effect of PCM in concrete with lightweight aggregate (LWA) and showed that by using a lower-transition-temperature PCM in a bridge deck, a significant portion of the damage from freezing and thawing can be potentially avoided. The purpose of this research is to conduct a systematic study on the effect of PCMs on various properties of concrete containing PCMs, including mechanical and thermal properties and volumetric stability. Specifically, the compressive strength, flexural strength, drying shrinkage, and thermal behaviors of the PCM-modified concrete are investigated. These material properties are very important to ensure both short- and long-term performance of the concrete. Moreover, microstructure of PCM-modified concrete is studied by a scanning electron microscope (SEM). The results show how PCM microcapsules are incorporated in the microstructure of cement paste. RESEARCH SIGNIFICANCE To build energy-efficient buildings, it is important to apply every possible method to improve the thermal performance of all components used in the building envelope. Therefore, not only conventional insulation materials but also structural materials should be taken into account in terms of their contribution to the thermal performance of the building. While improving the thermal properties of structural materials, the importance of their mechanical properties should not be disregarded. The addition of PCMs in concrete can certainly improve the thermal properties of concrete but has some adverse impact on the mechanical properties of concrete. The effect of PCM on shrinkage of concrete has not been studied. Therefore, a comprehensive study on these material properties of PCM-modified concrete is necessary and important.

ACI Materials Journal, V. 109, No. 1, January-February 2012. MS No. M-2010-165.R2 received April 25, 2011, and reviewed under Institute publication policies. Copyright 2012, American Concrete Institute. All rights reserved, including the making of copies unless permission is obtained from the copyright proprietors. Pertinent discussion including authors closure, if any, will be published in the November-December 2012 ACI Materials Journal if the discussion is received by August 1, 2012.

71

Pania Meshgin is a Postdoctoral Appointee at Sandia National Laboratories in the Department of Civil, Environmental and Architectural Engineering at the University of Colorado at Boulder, Boulder, CO. He received his BS in civil engineering from the Bahai Institute for Higher Education, Tehran, Iran, in 2001; his MSc in civil engineering from the University of New Mexico, Albuquerque, NM, in 2006; and his MSc and PhD in civil engineering from the University of Colorado at Boulder in 2009 and 2011, respectively. Yunping Xi is a Professor in the Department of Civil, Environmental and Architectural Engineering at the University of Colorado at Boulder. He received his BS in civil engineering from the Beijing Institute of Civil Engineering and Architecture, Beijing, Peoples Republic of China, in 1982; his MS in structural engineering from the Central Research Institute of Building and Construction, Beijing, in 1985; and his PhD in structural engineering from Northwestern University, Evanston, IL, in 1991. His research interests include the experimental and theoretical study of the durability of construction materials and the sustainable development of construction materials.

MATERIALS AND MIXING METHODS USED IN EXPERIMENTAL PROGRAM A PCM, MPCM-28 wet cake, was used in this study. The melting point of the MPCM-28 wet cake is 28C (82F). This specific product was chosen based mainly on its melting point, which is the closest one to room temperature among all commercially available PCM products. The PCM was microencapsulated, as shown in Fig. 1, which is an image of MPCM-28 wet cake taken by a low-vacuum scanning electron microscope (SEM). The white microspheres shown in Fig. 1 are made of melamine formaldehyde (MF

Fig. 1Microstructure of MPCM-28 wet cake. Table 1General properties of MPCM-28 wet cake
Typical properties Appearance Form Capsule composition Core material Particle size (mean) Melting point Heat of fusion Specific gravity Temperature stability Thermal cycling White to slightly off-white color Wet cake (70% solid, 30% water) 85 to 90% weight % PCM 10 to 15 weight % polymer shell Paraffin 17 to 20 micron 28C (82F) 180 to 195 kJ/kg (77.43 to 83.89 BtuIT/lb) 0.9 Extremely stableless than 1% leakage when heated to 250C (482F) Multiple

Melamine) polymers and the core material is paraffin (n-Octadecane). The encapsulated PCM was selected for this study because the polymeric microcapsules can survive the wet and high-pH environment during the casting process of the concrete. Table 1 shows MPCM-28s general properties based on the manufacturers data sheet.13 When PCM is added to the mixture of concrete materials as an additional component, there are two ways to adjust the mixture design of the concrete. PCM can be used to replace a certain percentage of fine aggregate (sand) in the concrete mixture, which can be called the PCM replacement method. PCM can also be used as an additive in the concrete mixture, which can be called the PCM additive method. With a fixed amount of PCM, the effect of PCM on the thermal properties of concrete is fixed. As shown in later sections, however, the effect of PCM on the mechanical properties of the concrete is different depending on the mixture design method used. Therefore, there are two phases in this study. In the first phase, the focus was to compare the difference of the two mixture design methods. So, both the PCM replacement method and PCM additive method were used for preparing concrete samples. The effect of the two different mixture design methods on the compressive strength of PCM concrete was studied. In the second phase, the focus was to study properties of the PCM-modified concrete with various volume percentages of PCM. Therefore, only one mixture design methodthe replacement methodwas used. The effect of PCM contents on the various properties of PCM concrete was studied. With the PCM replacement method, different volume percentages of PCM were used to replace the same volume percentages of fine aggregate. For instance, when 5% of the fine aggregate needs to be replaced by the PCM, the weight of the PCM can be readily calculated by knowing the density of PCM being equal to 0.9 kg/m3 (0.0562 lb/ft3). In the PCM additive method, the amounts of gravel and sand remain the same, and various amounts of PCM were added to the mixture. The concrete mixtures consisted of 630 kg/m3 (39.3 lb/ft3) fine aggregate (sand), 1160 kg/m3 (72.4 lb/ft3) coarse aggregate, 350 kg/m3 (21.8 lb/ft3) Type I portland cement, 200 kg/m3 (12.5 lb/ft3) water, and four different amounts of microencapsulate PCM. The sieve analysis based on ASTM C136-0614 was conducted on the fine aggregate and the coarse aggregate being used in the concrete mixtures. Figure 2 shows the particle size distribution of the aggregates. Tables 2 and 3 show different mixture designs using PCM replacement and PCM additives in Phase 1 of the study. As one can see in the two tables, 5, 10, 15, and 20% replacements (by volume) and additions were used as experimental parameters to examine the effect of PCM on the compressive strength of the PCM-modified concrete. The same amounts of PCM were used in both the PCM additive and replacement methods to examine the effects of the two different design methods on the compressive strength of PCM-modified concrete (refer to Tables 2 and 3). There are three steps in the mixing process. First, all the dry materials such as cement, sand, and aggregate are mixed for 1 minute. Second, the water is gradually added to the mixture and the mixing lasts for another minute, and then PCM is added to the concrete mixture; this step takes 1 minute as well. It should be noted that it is important to add the PCM as the very last component to prevent the damage of microencapsulate during the mixing process. ACI Materials Journal/January-February 2012

72

Adding the PCM toward the end of the process prevents damage in the outer shell due to the high shear associated with the mixing of concrete. TESTING PROCEDURES To have a better understanding of the effects of PCM on the properties of the PCM-modified concrete, the following tests were conducted: the compressive strength test, flexural strength test, drying shrinkage test, and thermal conductivity test. These four tests were selected for this study because the PCM and the microcapsules are not high-strength, solid

materials like the regular components in concrete and thus their effect on the strength of concrete was a major concern. Their effects on the drying shrinkage of concrete are also important when the phase transformations take place at different temperatures. The thermal conductivity test is especially important for PCM-modified concrete because the modifications in the thermal properties of the concrete due to added PCM will make the concrete suitable for many potential applications. Compressive strength This test was done according to ASTM C873/ C873M-04e1.15 Six cylindrical specimens of 50.8 x 101.6 mm (2 x 4 in.) were cast for each mixture design and the compression test was conducted on three of the cylinders at 7 days of age and the remaining three at 28 days of age. The compression test was conducted using a MTS machine with displacement control. The loading rate was kept constant at 0.0381 mm/s (0.0015 in./s) for all specimens at all ages. Three-point bending flexural test This test was done according to ASTM C78-09.16 Two 50.8 x 50.8 x 203.2 mm (2 x 2 x 8 in.) prism specimens were prepared for this test. The three-point bending flexural test provides values for the modulus of elasticity in bending Ef, flexural stress sf, flexural strain ef, and the flexural stressstrain response of the material. The flexural stress (also called modulus of rupture) can be calculated based on the following equation
f = 3 pL 2bd 2

(1)

Fig. 2Particle size distributions of fine and coarse aggregate.

where sf is the flexural stress; p is the load at the given point on the load-deflection curve; L is the support span; b is the width of the beam; and d is the depth of the beam. In this study, the flexural test was performed using a universal testing machine with displacement control. The loading rate was 0.01 mm/s (0.000394 in./s) and the sampling rate was 0.1 second.

Table 2Mixture designs using PCM replacement method

Group PCM0 PCM5-R PCM10-R PCM15-R PCM20-R Cement, kg/m3 (lb/ft3) 350 (21.8) 350 (21.8) 350 (21.8) 350 (21.8) 350 (21.8) Coarse aggregate, kg/m3 (lb/ft3) 1160 (72.4) 1160 (72.4) 1160 (72.4) 1160 (72.4) 1160 (72.4) Fine aggregate, kg/m3 (lb/ft3) 630.0 (39.3) 598.5 (37.4) 567.0 (35.4) 535.5 (33.4) 504.0 (33.7) PCM, kg/m3 (lb/ft3) 0.00 10.66 (0.67) 21.32 (1.33) 31.97 (2.00) 42.63 (2.66) Water, kg/m3 (lb/ft3) 200 (12.5) 200 (12.5) 200 (12.5) 200 (12.5) 200 (12.5)

Table 3Mixture designs using PCM additive method

Group PCM0 PCM5-R PCM10-R PCM15-R PCM20-R Cement, kg/m3 (lb/ft3) 350 (21.8) 350 (21.8) 350 (21.8) 350 (21.8) 350 (21.8) Coarse aggregate, kg/m3 (lb/ft3) 1160 (72.4) 1160 (72.4) 1160 (72.4) 1160 (72.4) 1160 (72.4) Fine aggregate, kg/m3 (lb/ft3) 630.0 (39.3) 630.0 (39.3) 630.0 (39.3) 630.0 (39.3) 630.0 (39.3) PCM, kg/m3 (lb/ft3) 0.00 10.66 (0.67) 21.32 (1.33) 31.97 (2.00) 42.63 (2.66) Water, kg/m3 (lb/ft3) 200 (12.5) 200 (12.5) 200 (12.5) 200 (12.5) 200 (12.5)

ACI Materials Journal/January-February 2012

73

Drying shrinkage The standard test method for the drying shrinkage of concrete based on ASTM C157/157M-0817 was used in this study. The drying shrinkage tests were conducted only on specimens prepared by the PCM replacement method. Two mixture designs for PCM replacement at 10 and 20% volume of sand were used for the specimens in the drying shrinkage test. These two mixture designs were chosen to cover a lower and higher range of PCM content in the mixture design. Two prism specimens 25.4 x 25.4 x 255 mm (1 x 1 x 10 in.) were prepared for each mixture design. Measurements were started after 7 days of moist curing in a standard curing room. The relative humidity of the lab remained at approximately 40% during the testing period. Thermal properties In general, thermal diffusivity a, density r, and specific heat (or heat capacity) cp of a material can be measured directly, and then thermal conductivity k of the material can be calculated by using the following equation

specific heat. By keeping the crucible of the sample and the reference the same size, under the same temperature and made of the same material, the effects of parasitic convective and radiative losses are automatically canceled. Based on the difference in the heat (DSC signal) required for heating the samples, the ratio of specific heats of the sample and the standard can be easily determined. Equation (3) shows how this information is used to calculate the specific heat capacity of each sample.
CP , Sample = CP , Standard mStandard DSCSample DSC Baseline (3) mSample DSCStandard DSC Baseline

k = C p

(2)

Direct measurement of thermal diffusivity a is popular because it only requires measurement of temperature history due to a thermal perturbation on the sample. This is much easier than measuring heat flux, as it is required in many steady-state methods for obtaining thermal conductivity k.18 An elemental analyzer device was used for thermal diffusivity measurements. On the other hand, differential scanning calorimetry (DSC) for measuring the specific heat was conducted using a differential scanning calorimeter machine19 at a scanning rate of 10C/min (50F/min) in the temperature range of 25 to 55C (13 to 131F) in a nitrogen atmosphere. In general DSC analysis, two crucibles with the same size are heated through the same temperature-versus-time program. The reference crucible is always empty, while the other crucible, in turn, is empty, carries the standard (sapphire) sample, and finally, the sample with unknown

It should be noted that the specific heat can only be calculated under an exothermic process. Moreover, for the DSC analysis, the system should be programmed for the desired temperature range, which was 25 to 55C (13 to 131F) in this case. Thus, in the heating cycle (endothermic), the temperature was first brought down to 25C (13F) and was held for 4 minutes; then, the heating cycle started up to 55C (131F). When the temperature reached this highest desired temperature (55C [131F]), it was held for an additional 3 minutes. Then, the cooling cycle (exothermic) can be started by cooling the system to the lowest desired temperature (25C [13F]). Similar to the specimens for the drying shrinkage test, two mixture designs for PCM replacement of 10 and 20% of sand were used for the specific heat test. Specimens with dimensions of 10 x 10 x 1.5 mm (0.39 x 0.39 x 0.06 in.) were cut from concrete specimens and their thermal diffusivity was measured at 7 and 28 days of age. Because the microcapsules of PCM are very small, the sample size of 10 x 10 x 1.5 mm (0.39 x 0.39 x 0.06 in.) can be used to study the effect of PCM on the thermal properties of the cement paste incorporating PCM microcapsules. EXPERIMENTAL RESULTS AND DISCUSSIONS Compressive and flexural strengths The average compressive strengths of the two groups of samples at 7 and 28 days are shown in Tables 4 and 5.

Table 4Compressive strengths and strength reductions using PCM to replace sand
Group PCM0 PCM5-R PCM10-R PCM15-R PCM20-R Total volume fraction PCM and sand 0.209 0.209 0.209 0.209 0.209 7 days Average stress, MPa (ksi) 18.27 (2.65) 15.24 (2.21) 15.17 (2.20) 16.96 (2.46) 14.75 (2.14) Percentage change 0.00 16.60 16.98 7.17 19.25 21.44 (3.11) 19.44 (2.82) 18.13 (2.63) 17.79 (2.58) 15.79 (2.29) 28 days Average stress, MPa (ksi) Percentage change 0.00 9.30 15.40 17.19 26.31

Table 5Compressive strengths and strength reductions using PCM as additive

Group PCM0 PCM5-A PCM10-A PCM15-A PCM20-A Total volume fraction PCM and sand 0.209 0.217 0.225 0.233 0.240 7 days Average stress, MPa (ksi) 18.27 (2.65) 13.24 (1.92) 10.69 (1.55) 11.51 (1.67) 10.34 (1.50) Percentage change 0.00 27.41 41.67 36.96 43.40 21.44 (3.11) 16.20 (2.35) 12.89 (1.87) 13.24 (1.92) 12.55 (1.82) 28 days Average stress, MPa (ksi) Percentage change 0.00 24.42 39.93 38.38 41.67

74

ACI Materials Journal/January-February 2012

The strength reductions are shown in the same tables. Figure 3 shows the comparison of the test data obtained from the 7-day-old samples, and Fig. 4 shows the comparison of the test data obtained from the 28-day-old samples. One can see that with an increasing amount of PCM, the compressive strength of PCM-modified concrete decreases. This result confirms the finding from Hunger et al.s work.11 This result can be easily explained based on the properties of PCM, which is not a solid material with high strength and high stiffness. Therefore, its addition in the hardened concrete will definitely result in a strength reduction. The authors goal is to find the extent and trend of the strength reduction and minimize the strength reduction. As discussed previously, with a given amount of PCM in concrete, two different methods were used to add PCM in the concretethat is, the replacement method and the additive method. From Fig. 3 and 4, one can clearly see that the strength reductions associated with the PCM replacement method are less than those of the PCM additive method for all four amounts of PCM used in this study. This is an important result because the improvement of the thermal properties of a concrete material depends on the amount of PCM in the concrete. With a fixed amount of PCM in concrete, one must minimize the strength reduction of concrete due to the addition of PCM. The present test data indicate that the PCM replacement method results in less strength reduction and thus should be adopted in practice. Then, the question is why the PCM replacement method resulted in less strength reduction than the PCM additive method. This may be explained by the volume fractions of the constituents in PCM-modified concrete. The compressive strength of concrete depends on several mixture design parameters, such as the water-cement ratio (w/c), cement content, and gravel-sand ratio. In this study, the w/c was kept constant for both the PCM replacement and PCM additive methods. Therefore, the effect of other design parameters became important. Considering the PCM microcapsules as soft sand particles, in the PCM replacement method, the total volume fraction of fine aggregate (natural sand plus PCM) remained as a constant with an increasing amount of PCM in the mixture. So, the strength reduction of the concrete is mainly due to the increasing amount of soft sand (that is, PCM microcapsules) in the concrete, whereas in the PCM additive method, the total volume fraction of fine aggregate (natural sand plus PCM) increased and, thus, the volume fraction of cement decreased with an increasing amount of PCM in the mixture. Therefore, the strength reduction of the concrete is due to two concurrent mechanisms: 1) the increasing amount of soft sand (PCM microcapsules) in the concrete, which is the same as the PCM replacement method; and 2) the decreasing amount of cement (lowered cement content). It is worthwhile to point out that from the test data shown in Fig. 3 and 4, the strength reduction due to the addition of PCM is not substantial up to 20% of volume replacement of fine aggregate, and there are many applications in which the reduced compressive strength will still be acceptable. Based on the test data shown in Fig. 3 and 4 and the aforementioned analysis, it was decided that the rest of the experimental program (Phase 2) would be conducted on the specimens prepared by the PCM replacement method. To this end, two groups of specimens were prepared: PCM10-R and PCM20-R. PCM10-R had 10% replacement of natural sand, which is a representation for a medium ACI Materials Journal/January-February 2012

Fig. 3Comparison of compressive strengths at 7 days.

Fig. 4Comparison of compressive strengths at 28 days. range of PCM replacement. PCM20-R had 20% replacement of natural sand, which is a representation for a high range of PCM replacement. A three-point bending flexural test was conducted on Specimens PCM10-R and PCM20-R at 28 days. Figure 5 shows the test data of the flexural strength tests. By increasing the amount of PCM in the concrete mixture, the flexural strength of the concrete decreased. The extent of strength reduction is approximately 6%, which is not significant for the 10 to 20% PCM replacement. Hawes et al.2 also showed that the flexural strength of the PCM wallboards is comparable with the flexural strength of the conventional wallboards, which is a valuable confirmation of the test results. 75

Fig. 8Comparison of drying shrinkage between PCM10-R and PCM20-R. replacement method), the drying shrinkage increases by approximately 25%. The increase in drying shrinkage can be explained by the low stiffness of PCM particles and the high moisture content of the specific PCM used in this study, which is MPCM-28 wet cake. The main component in the concrete responsible for the drying shrinkage is cement paste. The amount of cement paste in concrete depends on the cement content and w/c. Higher cement content and higher w/c in concrete leads to a higher potential of drying shrinkage. On the other hand, the potential of drying shrinkage is restrained by the solid framework formed by aggregates (gravel and sand) inside the concrete. With the same shrinkage potential generated by the cement paste, the actual shrinkage depends on the stiffness of the aggregates. The higher the stiffness of the aggregate framework, the lower the actual drying shrinkage of the concrete. In Phase 2 of this study, the cement content and w/c were fixed, and thus the difference in the measured drying shrinkage came from the difference in the stiffness of the solid framework of aggregate including PCM particles. With 20% PCM replacement of fine aggregate, the stiffness of the aggregate framework is lower than that of the concrete with 10% replacement. Therefore, the shrinkage of PCM-modified concrete with 20% PCM replacement is higher. Using the present test data, an empirical formula was developed for characterizing the drying shrinkage of PCM-modified concrete
PCM w ) sh = 0.02 ln t + ( 100 c
( 20 + 2) PCM

Fig. 5Flexural strengths of PCM-modified concrete.

Fig. 6Drying shrinkage test data of PCM10-R.

(4)

Fig. 7Drying shrinkage test data of PCM20-R. Drying shrinkage For the drying shrinkage test, two specimens were prepared for each group and were kept in the curing room for 7 days; then, the drying shrinkage test and measurement of length change were started in room temperature. Figures 6 and 7 show the test data for the drying shrinkage tests. The solid lines in the two figures represent the curve fitting of the averaged test data. Then, the two solid lines were plotted in Fig. 8 as a comparison of the drying shrinkage of PCM-modified concrete with 10 and 20% replacement. The comparison of the two curves shows that by increasing the amount of the PCM in the concrete from 10 to 20% (the 76

where w/c is the water-cement ratio; t is the time, in days; and PCM is the volume percentage of PCM replacement of fine aggregate in the concrete, which in this study is either 10 or 20. In the equation, the first term is the time-dependent drying shrinkage, which represents the drying shrinkage of regular concrete, and the second term is not time-dependent, which represents the additional shrinkage due to the addition of PCM. It is important to note that Eq. (4) is valid only for the mixture designs used in this study with PCM replacement between 10 and 20%. For example, the shrinkage of the concrete with 15% replacement at 30 days can be calculated by the equation. For other mixture designs, the equation for the drying shrinkage of concrete will be different, mainly in the first term of Eq. (4). In addition to the effect of PCM on the stiffness of the solid framework of aggregate, there is another important factor that may contribute to the higher shrinkage of PCM-modified ACI Materials Journal/January-February 2012

Fig. 9Specific heat of MPCM-28 wet cake. (Note: 1 J/(g.K) = 1 J/(g.C).) concrete. As shown in Table 1, MPCM-28 wet cake contains 30% moisture and thus more PCM in the concrete mixture will bring more moisture in the mixture. As a result, more shrinkage strain will be observed in specimens with a higher volume percentage of PCM. Thermal properties The thermal tests of the PCM-modified concrete revealed interesting results. To understand these results better, the DSC analysis was performed on MPCM-28 wet cake alone, which had 30% water content on the exterior of the capsules. The water content should be eliminated from the particles prior to the DSC analysis, otherwise the DSC graph would not provide the accurate result and the joules per gram at the PCMs phase-change temperature will be a combined contribution of PCM and water. Figure 9 shows the specific heat of the MPCM-28 wet cake under different temperatures. Because the specific heat is calculated based on the ratio method as described previously, the peaks observed on the DSC curve and cp curve are happening at the same temperature, which is the melting temperature. Therefore, from Fig. 9, it can easily be seen that the melting point is 29.2C (84.6F). This value is slightly different from the melting point of MPCM-28 wet cake and that is because residual water remained on the exterior of the PCMs after the predrying treatment. Additionally, one can also observe the amount of energy needed for the melting process, which ACI Materials Journal/January-February 2012 is approximately 373.6 J/g (160.73 BtuIT/lb), representing the area beneath the cp curve in the endothermic cycle. Moreover, Fig. 10 shows that the peak in the endothermic cycle appears at approximately 29.2C (84.6F), which is the same as the peak observed in Fig. 9. The onset of the melting process, however, is at 23.5C (74.3F). The other peak observed in the exothermic cycle is 14.9C (58.8F), corresponding to the solidification process. The onset in this cycle is 22.4C (72.3F), which is the temperature for the solidification to take place. Additionally, Fig. 11 shows the specific heat capacity of different PCM-modified concrete specimens at different ages. The peaks in specific heat curves for PCM-modified concretes are all in the range of the melting point of the MPCM-28. Again, the water content in the wet cake causes the difference between the peak point and the actual melting point of the MPCM-28. Because the PCM particles were used in the concrete mixture without the predrying process, these peaks for the concrete are not exactly at 28C (82.4F) (the melting point of PCM). One important observation from Fig. 10 is that increasing the amount of PCM will increase the specific heat capacity of the concrete. This increase is approximately 19% when PCM replacement changed from 10 to 20%. Another observation is that the older the specimens, the higher the specific heat capacity. In other words, aging is a significant factor in increasing the specific heat capacity of the concrete. For instance, in the case of 77

Fig. 10Differential scanning calorimetry of phase-change material. (Note: 1 mW = 0.0007376 ft.lbf/s = 0.0034121 Btu/h.) Table 6Thermal diffusivity of PCM-modified concrete
Sample PCM10 PCM10 PCM20 Age, days 7 28 28 Thermal diffusivity, m2/s (ft2/s) 0.89 (2.92) 1.14 (3.74) 0.96 (3.15)

Fig. 11Specific heats of PCM-modified concretes with different amounts of PCM at different ages. (Note: C = (F 32)/1.8; 1 J/(g.K) = 1 J/(g.C).)

Fig. 12Thermal conductivity of PCM-modified concrete with different amount of PCM at different ages. (Note: C = (F 32)/1.8; 1 W/(m.K) = 1 W/(m.C).) 78

Specimen PCM10-R, the specific heat increased by 24% from 7 to 28 days of age. The results of thermal diffusivity in room temperature are provided in Table 6. These data show that the thermal diffusivity decreases with an increasing amount of PCM, whereas the aging has an opposite effect on the thermal diffusivity, meaning the older the specimens, the higher the thermal diffusivity. The specific heat, thermal diffusivity, and Eq. (2) were used to determine the thermal conductivity of each PCM sample. Figure 12 shows the result of thermal conductivity of different PCM-modified concretes. This figure illustrates that increasing the amount of PCM in the concrete results in a lower thermal conductivity, but the older the concrete, the higher the thermal conductivity; this is similar to the result of the heat capacity. Adding more PCM in the concrete results in higher specific heat and thus lower thermal conductivity, which are beneficial effects for improving the thermal insulation property of building envelopes. Hunger et al.11 also concluded similar results in his work. In general, materials with high thermal diffusivity rapidly adjust their temperature to that of their surroundings, and ACI Materials Journal/January-February 2012

this is due to the fact that they can conduct heat quickly in comparison to their volumetric heat capacity.20 Therefore, materials with low thermal diffusivity, such as PCMmodified concretes, are more desirable in the application of building envelopes. Microstructure of PCM-modified concrete Figures 13 and 14 show the micrographs of the PCM microcapsules within the concrete taken by an SEM. Several PCM particles are distributed throughout the mixture. These SEM images provide a good understanding of the bonding between cement paste and the PCMs outer layer, which is quite strong to ensure a high compressive strength of the PCM-modified concrete. Some research12 has shown that the PCM particles were destroyed or partially destroyed in the mixing process. To avoid the damage of PCM particles in the mixing process, the PCM particles are added at the very last step during the mixing process. CONCLUSIONS The effect of PCM on some properties of concrete was investigated by a systematic experimental study. Based on the results of this study, the following conclusions are drawn: The results have shown that any addition of PCM particles in concrete will decrease the compressive and flexural strength of the concrete in comparison with conventional concrete. With up to 20% PCM replacement of fine aggregate (sand), however, the deduction in compressive and flexural strength is not too large and there are many structural applications for the resulting PCM-modified concrete. There are two mixture design methods for adding PCM particles into concrete. PCM can be used to replace a certain percentage of fine aggregate (sand) in the concrete mixture, which is the PCM replacement method. PCM can also be used as an additive in the concrete mixture, which is the PCM additive method. Both methods can be used for the mixture design of PCM-modified concrete. The present test data show that the strength reductions associated with the PCM replacement method are less than those of the PCM additive method. The drying shrinkage test results show that with an increasing amount of PCM, the drying shrinkage of the PCM-modified concrete increases, which is due to the low stiffness of the PCM particles that reduces the overall stiffness of the aggregate framework in the concrete. An empirical equation was developed to characterize the shrinkage of PCM-modified concrete with mixture designs in this study and with different PCM contents from 10 to 20% replacement of fine aggregate. The thermal analysis results of the PCM-modified concrete show that the PCM can alter the thermal conductivity of the PCM-modified concrete. Specifically, any increase in the amount of the PCM in the concrete mixture increases the heat capacity of the PCM-modified concrete, decreases the thermal diffusivity of the concrete, and thus decreases the thermal conductivity. Because the change in thermal diffusivity at different temperatures is not as significant as the change in the heat capacity, it can be concluded that using more PCM in the concrete mixture can ACI Materials Journal/January-February 2012

Fig. 13SEM micrograph of PCM-modified concrete.

Fig. 14SEM image showing bond between PCMs outer layer and cement paste. provide a better material for the building in terms of thermal insulation. SEM analysis showed that there is almost no damage to the PCM outer layer, and there is a decent bonding between the cement paste and the PCM particles. Therefore, the PCM microcapsules and the mixture procedure used in this study are quite effective to avoid premature damage of PCM particles during the concrete mixing process. ACKNOWLEDGMENTS

The authors would like to thank R. Yang and S. J. Thiagarajan from the Department of Mechanical Engineering at the University of Colorado at Boulder for providing equipment for the thermal analysis throughout this research. Their sincere help is greatly appreciated.

b Cp Cp,Sample Cp,Standard D d DSCBaseline DSCSample DSCStandard Ef k L mSample mStandard P PCM t

= = = = = = = = = = = = = = = = =

width of beam specific heat specific heat of sample specific heat of standard (sapphire) sample maximum deflection at center depth of beam difference in heat for baseline difference in heat for sample difference in heat standard modulus of elasticity in bending thermal conductivity support span mass of sample mass of standard load at given point on load-deflection curve amount of phase-change material time in days

NOTATION

79

w/c a ef esh r sf

= = = = = =

water-cement ratio thermal diffusivity flexural strain shrinkage strain density flexural stress

1. Wikipedia, Phase-Change Material, http://en.wikipedia.org/wiki/ Phase_Change_Material, 2009. 2. Hawes, D. W.; Feldman, D.; and Banu, D., Latent Heat Storage in Building Materials, Energy and Buildings, V. 20, 1993, pp. 77-86. 3. Lencer, D.; Salinga, M.; Grabowski, B.; and Hickel, T., A Map for Phase-Change Materials, Nature Materials, V. 7, 2008, pp. 972-977. 4. Sharma, A.; Tyagi, V. V.; Chen, C. R.; and Buddhi, D., Review on Thermal Energy Storage with Phase Change Materials and Applications, Renewable and Sustainable Energy, V. 13, 2009, pp. 318-345. 5. Hawes, D. W.; Banu, D.; and Feldman, D., Latent Heat Storage in Concrete, Solar Energy Materials, V. 19, 1989, pp. 335-348. 6. Chen, C.; Guo, H.; Liu, Y.; Yue, H.; and Wang, C., A New Kind of Phase Change Material (PCM) for Energy-Storing Wallboard, Energy and Buildings, V. 40, 2008, pp. 882-890. 7. Richardson, M. J., and Woods, A. W., An Analysis of Phase Change Material as Thermal Mass, Proceedings of the Royal Society: A Mathematical, Physical and Engineering Science, V. 464, 2008, pp. 1029-1056. 8. Li, Z., and Li, X., Development of Thermal Insulation Materials with Granular Phase Change Composite, Advances in Construction Materials, 2007, pp. 741-748. 9. Cabeza, L. F.; Castellon, C.; Nogues, M.; Medrano, M.; Leppers, R.; and Zubillaga, O., Use of Microencapsulated PCM in Concrete Walls for Energy Savings, Energy and Buildings, V. 39, 2007, pp. 113-119.

REFERENCES

10. Zhang, Y., and Zhang, X., Phase Change Material Effect on Heat Property of Mortar, JournalTongji University, V. 35, No. 7, 2007, pp. 965-969. 11. Hunger, M.; Entrop, A. G.; Mandilaras, I.; Brouwers, H. J. H.; and Founti, M., The Behavior of Self-Compacting Concrete Containing MicroEncapsulated Phase Change Materials, Cement and Concrete Composites, V. 31, 2009, pp. 731-743. 12. Bentz, D. P., and Turpin, R., Potential Applications of Phase Change Materials in Concrete Technology, Cement and Concrete Composites, V. 29, 2007, pp. 527-532. 13. Microtek, MicroPCM Products and Corresponding Product Data Sheets, http://www.microteklabs.com/micropcm_products.html, 2009. 14. ASTM C136-06, Standard Test Method for Sieve Analysis of Fine and Coarse Aggregates, ASTM International, West Conshohocken, PA, 2006, 5 pp. 15. ASTM C873/C873M-04el, Standard Test Method for Compressive Strength of Concrete Cylinders Cast in Place in Cylindrical Molds, ASTM International, West Conshohocken, PA, 2004, 4 pp. 16. ASTM C78-09, Standard Test Method for Flexural Strength of Concrete (Using Simple Beam with Third-Point Loading), ASTM International, West Conshohocken, PA, 2009, 4 pp. 17. ASTM C157/C157M-08, Standard Test Method for Length Change of Hardened Hydraulic-Cement Mortar and Concrete, ASTM International, West Conshohocken, PA, 2008, 7 pp. 18. Bejan, A., and Kraus, A. D., Heat Transfer Handbook, John Wiley & Sons, Inc., New York, 2003, 1496 pp. 19. Holman, J. P., Heat Transfer, McGraw-Hill Science, New York, 2001, 752 pp. 20. Richardson, M. J., The Application of Differential Scanning Calorimetry to the Measurement of Specific Heat, Compendium of Thermophysical Properties Measurement Methods, Springer, Plenum Press, New York, 1992, pp. 519-545.

80

ACI Materials Journal/January-February 2012