ZHAO Haitao, LI Jinghao, LIU Hui, XU Wen, LI Hua, WANG Penggang,HUANG Jie, ZHANG Yuming, PAN Li, JIANG Jianhua

(1. College of Civil and Transportation Engineering, Hohai University, Nanjing 210024, China; 2. College of Materials Science and Engineering, Southeast University, Nanjing 211189, China; 3. School of Civil Engineering, Chongqing University, Chongqing 400045, China;4. Agile Property Holdings Co., Ltd., Guangzhou 510000, China; 5. Jiangsu Sobute New Materials Co. Ltd., Nanjing 211103, China; 6. Department of Civil Engineering, Qingdao University of Technology, Qingdao 266033, China; 7. Digital Engineering Institute, Henan Water &Power Engineering Consulting Co., Ltd., Zhengzhou 450016, China)

Abstract: The pre-soaked shale employed as an internal curing agent and CaO employed as expansion agent were incorporated into concrete to investigate their effects on the mechanical properties and autogenous deformation of early-age concrete. We have conducted the relevant tests for setting time, mechanical properties, internal relative humidity and autogenous deformation of early-age concrete with shale or/and CaO incorporation. The results indicate that the set behavior is delayed by shale addition but is accelerated with CaO.The shale addition firstly enhances and subsequently decreases the strength, but CEA addition has a weakening effect. Additionally, shale or/and CaO incorporation deteriorates the elastic modulus. The shale and CaO incorporation significantly improve the internal relative humidity of concrete. The internal curing efficacy of shale could synergistically mitigate the autogenous shrinkage, that is, could enhance the expansion of CaO and then greatly reduce the contraction, which is significantly beneficial to impede the shrinkage-introduced cracks of early-age concrete.

Key words: early-age concrete; internal curing; pre-soaked shale; CaO-based expansion agent;mechanical properties; autogenous deformation

1 Introduction

High-performance concrete (HPC) has a popular application and good prospects in actual engineering owing to the outstanding workability, mechanical properties and durability[1-4]. However, the autogenous shrinkage (AS) of HPC is large because HPC is manufactured with a low water to binder ratio (w/b)and water content[5]. The large AS induces an extremely large tensile stress when the concrete structure is under restraint, which renders concrete vulnerable to generating the early-age cracks[6-8]and seriously weakening the concrete durability[9,10]. Herein, it is urged to mitigate the AS of HPC to prevent concrete from cracking at early age.

Incorporating expansion agent into concrete to manufacture the shrinkage-compensating concrete has been confirmed as an effective way of AS reduction.CaO-based expansion agent (CEA) is popularly applied among the expansion agents[11]. Yooet al[12]studied the AS of HPC incorporating CEA and stated that 68%AS of concrete was reduced in the presence of a 10%dosage of CEA. Similar conclusions were also obtained by other researchers[13-15]. However, some studies found that the CEA could not realize its expected expansion efficacy because thew/bof HPC was low,resulting in that CEA hydrated incompletely[16-18].Additionally, HPC incorporating CEA contracts sharply and substantially after expansion[11], which is greatly unbeneficial to early-age cracking control.

To solve the cracking problem of HPC, water curing is considered. However, the external curing water could not penetrate enough depth to cure the concrete because of the dense pore structure of HPC.Herein, internal curing (IC)[19,20]is significantly effective in mitigating the AS of HPC. Presently,pre-soaked shale lightweight aggregate has been investigated widely. Henkensiefkenet al[21]reported that the high content (14.3%-33.0%) of pre-saturated kilned expanded shale addition could maintain the concrete in an expansion state until 28 days, which minimized the risk of cracking induced by shrinkage of HPC. Joneset al[22]and Menget al[23]also stated similar conclusions. Nevertheless, incorporating shale only is still not enough for the concrete in the hostile serving environment which deserves a high crackresisting demand. Therefore, when CEA and shale are incorporated together in HPC, it is expected to reach a synergistic mechanism that shale and CEA cooperatively mitigate the AS,i e, shale desorbs moisture to promote CEA hydration and achieve more expansion; thereafter, shale continuously desorbs moisture that could impede the internal relative humidity (IRH) decline, which could inhibit concrete from contracting. However, there is not related study available at present. Besides, the mechanical properties of HPC are one of the most important performances[24],which would be weakened with expansion agent and IC agent incorporation. Therefore, we aim to understand the early-age mechanical properties and AS behavior of HPC incorporating shale or/and CEA experimentally.

2 Experimental

2.1 Materials

The P.II 52.5 ordinary Portland cement satisfying GB 175-2009[25]was employed as the binder material.According to the GB 23439-2009[26], the CEA used expanded by 0.18% under restraint and cured in water at 7days (see Fig.1).

Fig.1 CEA used in experiment

The lightweight shale ranging continuously from 0.5 to 1 mm, weighing 2.2 grams per cubic centimeter,having a surface area of 600 m2per kilogram, and manufactured from porous waste brick was used (see Fig.2). The pure water absorption ratio of shale was tested according to ASTM C128-07[27]and was 0.16 g water/g shale at 3 days (see Fig.3). The cylindrical compressive strength of the shale was 3.3 MPa tested according to ASTMC1761/C1761 M-17[28]. The manufactured limestone sand, whose fineness modulus was 3.40 and maximum size was 4.75 mm, was used as fine aggregate. The crushed limestone distributing continuously from 5-25 mm and weighing 2 730 kg per cubic meter was used as coarse aggregate. To regulate the concrete workability, a liquid polycarboxylic-based superplasticizer was used. The water for mix and IC was tap water.

Fig.2 Shale used in experiment

Fig.3 Water absorption ratio of shale

2.2 Mix proportions

The dry-state shale mass sufficiently and internally curing 1 m3concrete was obtained by Eq.(1)[29]:

whereMshaleis the mass of dry-state shale, kg/m3;Cfis the cement mass in unit volume concrete, kg/m3;CSis the volumetric decrease induced by 100% cement hydration, and is 0.06 g water/g cement;αmaxis the maximum cement hydration degree expected, which is approximately [(w/b)]/0.36 with aw/bless than 0.36;Sis the saturation degree of the shale and considered as 1 after 3 day soaking in water;φshaleis water absorption capability at 3 days of shale.

TheMshalewas 180.6 kg/m3(absorbed-water content was 28.0 kg/m3), which was 20.4% of coarse aggregate in volume. A gradient shale contents in 1 m3concrete of 10%, 15% and 20% of coarse aggregate by volume with introducing IC water of 13.7, 20.5 and 27.4 kg/m3, respectively, were designed. Due to a compressive strength deterioration discussed in the next section, the maximum shale content used for the autogenous deformation experiment was 15%. The CEA content incorporated was 3.0% and 6.0% of cement by mass. Table 1 exhibits all the ten mixtures. Thew/bof all the mixtures was 0.35. WC035 represented the reference sample and SL15EA3 represented concrete with 15% shale and 3% CEA.

Table 1 Mix proportions of the concrete

2.3 Experimental programs

2.3.1 Autogenous deformation test

The autogenous deformation test was carried out according to the procedures in Ref. [30,31]. The deformation of the sample was measured by the deformation sensor (see Fig.4(a)), which simultaneously monitored the free deformation strain and temperature of the sample. The mold casting sample was polyvinyl chloride (PVC) tube in a cylinder shape with the bottom firmly stuck using epoxy resin on a wooden board (see Fig.4(b)). The inner wall of the PVC mold was covered by a layer of vaseline and then a layer of Teflon foil for eliminating the friction.

The shale was soaked for 3 days in water for mix and IC before testing, and the water left was employed to mix concrete. The CEA, if necessary, was premixed with cement thoroughly. The mixing procedure of all the mixtures was obedient to ASTM C192/C192 M-16[32]. The same batch of concrete was used for casting the samples of autogenous deformation, IRH,set behavior, and mechanical properties.

The sensor was placed vertically and centrally in the mold with a concrete height of 130 mm, then concrete was continuously poured until filling the mold(see Fig.4(c)). After vibrating, the mold top face was sealed immediately with paraffin wax. Three samples were prepared for one case, and the average value was employed as the final data. Such a data processing method was applied for the autogenous deformation,IRH, and mechanical properties tests. Finally, all the samples were placed in a constant temperature and humidity room (temperature of (20 ± 2) ℃ and relative humidity ≥ 95%).

Fig.4 Autogenous deformation test setup: (a) Deformation sensor; (b) Mold for autogenous deformation test; (c) Test setup for autogenous deformation and temperature

The starting point for AS was the initial setting time[33]. The measured strain included thermal deformation. Herein, autogenous deformation was calculated as Eq.(2)[34]:

whereεAS(t) is AS,με;εtotal(t) is free deformation measured, με;α(t) is the coefficient of thermal expansion (CTE),με/℃;T(t) is concrete internal temperature, ℃;T0is temperature when concrete initially set, ℃; andtis age since initial set, day.

This study employs Eq.(3) to calculate the CTE of concrete[35]:

where α28is CTE at day 28 andmis an aggregate typedependent parameter. In the present study, αkis 8.0 με/°C approximately andmis 2.0.

2.3.2 Internal relative humidity measurement

The IRH at the center of the 100 mm-wide cubic sample was tested by the method in Ref.[35]. The humidity sensor used was HC2A-S (Rotronic) type with an accuracy of ±0.8% (see Fig.5(a)). The humidity data was continuously and automatically acquired every 10 minutes for 28 days. Before and after testing,NaCl, KBr, KCl, and K2SO4with RH values at 20 °C of 75.5%, 81.7%, 85.1%, and 97.6% were utilized to adjust the humidity sensors.

Fig.5 Humidity test setup: (a) Humidity sensor; (b) PVC tube and spacer rod; (c) Samples for humidity test

The PVC tube was designed to contain and locate a humidity sensor, as displayed in Fig.5(b). Two rectangular openings and holes were opened on the PVC tube to ensure that the IRH tested by the humidity sensor was the IRH of concrete to the possible greatest extent. An enclosed space of PVC tube was created by installing an O-ring above the humidity sensor and set to monitor the concrete IRH accurately.

Before casting concrete, the PVC tube was vertically placed with the bottom touching the mold bottom. A spacer rod was inserted into the PVC tube.It was noted to keep the PVC tube vertical in the mold during vibrating the sample and then the sample was sealed immediately with plastic film and aluminum foil paper (see Fig.5(c)). After the sample initially setting, the spacer rod was taken away carefully, then a humidity sensor was placed and the PVC tube top was sealed with a rubber stopper. The IRH samples were placed in the same room along with the autogenous deformation samples.

2.3.3 Set behavior

The set behavior was monitored by the capillary pressure device in Ref.[36]. The top of the sample was sealed well by the plastic film for preventing water evaporation.

2.3.4 Mechanical properties

The compressive and splitting tensile strength were measured according to methods in GB/T 50081-2002[37]. The dynamic elastic modulus (DEM) samples were prismoids with a size of 150 mm× 150 mm× 300 mm, and DEM was monitored by the ultrasonic device and calculated by Eq.(4)[38]：

whereEd(t) denotes DEM, GPa;ρdenotes the concrete density, kg/m3;v(t) denotes ultrasonic velocity, m/s;μcudenotes Poisson’s ratio and is 0.2.

The samples were demolded after day 1 and tested at day 3, 7, 14 and 28.

3 Results and discussion

3.1 Setting time

The setting times results are presented in Fig.6.The test results indicate that setting time are positively related to the shale content. The initial (final) setting time of SL10, SL15 and SL20 are 113.90% (113.57%),118.40% (117.76%) and 121.88% (123.45%) of the reference sample, respectively. The portlandite saturation behavior could embody the cement hydration and then determine the concrete set behavior[39]. The shale desorbs moisture, which decreases the portlandite concentration, leading to retardation in the concrete set. Meanwhile, as the CEA content increases, the setting times decrease slowly, which means the CEA addition could accelerate the concrete set. Comparing with WC035, the 6% CEA addition renders the concrete initial (final) setting time earlier by 5.4%(6.0%). The CEA hydration accelerates the portlandite saturation, giving rise to advance in setting times[40].Such development trends are suitable for the concrete incorporating shale and CEA,i e, the set of concrete incorporating shale becomes faster with CEA content while that of concrete incorporating CEA becomes slower with shale content. The initial (final) setting time of SL10 decreases by 1.5% (1.5%) and 2.7%(2.3%) compared with SL10EA3 and SL10EA6,respectively. The initial (final) set of EA3 is slowed by 15.8% (15.2%) and 18.4% (19.2%) compared with SL10EA3 and SL15EA3, respectively.

Fig.6 Setting time of concrete: (a) Incorporating shale; (b)Incorporating CEA; (c) Incorporating shale and CEA

3.2 Mechanical properties

The experimental results of compressive strength are displayed in Fig.7. The compressive strength enhances with age by a gradually decreasing enhancement rate. Additionally, the compressive strength increases with shale dosage when the shale content ≤ 15%, and then begins to decline. The 3-day compressive strengths of SL10, SL15 and SL20 are 103.2%, 105.0% and 92.1% of WC035, respectively.Meanwhile, the strength-enhancing effect by 10% and 15% shale addition increases to 4.3% and 10.0% at 28 day, respectively. The moisture desorbed from the shale contributes to the promotion of cement hydration[41],and makes the pore structure denser[42,43]; besides, the cement paste can penetrate the surface of shale[44],which mechanically interlocks them with each other and renders the cohesive force increase[45]. These two reasons contribute to enhance compressive strength.Nevertheless, the strength of the shale is inferior to the coarse aggregate replaced[46], which weakens concrete compressive strength. In the case of relatively small shale content addition, the strength-enhancement effect is superior to the strength-deterioration effect, and the compressive strength increases[47]; andvice versa,which is validated by test results.

Fig.7 Compressive strength of concrete with different shale and CEA contents

The compressive strength decreases with the increase of CEA content. The compressive strength of EA3 and EA6 is 94.9% and 92.0% of WC035 at 3 day, and increases to 99.5% and 99.1% at 28 day,respectively. This is because that cement is partially replaced by the equivalent-mass CEA in this study,leading to a fewer hydration products contributing strength and further a weaker skeleton. Meanwhile,it may be related to the microcracking resulted from CEA-induced excessive expansion pressure[48].

The effect of shale on the compressive strength in the presence of CEA is similar to that of concrete with shale alone, so is the CEA. The CEA-induced strength loss of shale concrete narrowed with age, which is because the CEA plays its role mainly during the early age. Besides, the concrete compressive strength incorporating shale and CEA is 2.0%-9.0% higher than that of WC035 at 28 day.

Fig.8 illustrates the splitting tensile strength results. The evolution trend of the splitting tensile strength is similar to the compressive strength.The splitting tensile strength at day 28 of concrete incorporating shale and CEA is 3.4%-9.1% higher than that of WC035.

Fig.8 Splitting tensile strength of concrete with different shale and CEA content

Fig.9 depicts the DEM results. Regardless of the presence of shale (CEA), a higher CEA (shale) content incorporation obtained a lower DEM. The 28-day DEM of WC035 decreases by 12.3% and 5.8% when 20%shale and 6% CEA is added, respectively. Besides, the DEM loss percentage of concrete incorporating shale and CEA is lower than the sum of those with shale and CEA addition alone. For instance, the 28-days DEM loss of SL15EA6 is 11.4% and is lower than the sum of SL15 (7.9%) and EA6 (5.8%). The DEM represents the stiffness of cementitious materials. Therefore, the lower the DEM, the less the restraint on the autogenous deformation, hence the stronger the deformation capacity of concrete[49].

Fig.9 Dynamic elastic modulus of concrete with different shale and CEA contents

3.3 Internal relative humidity

The evolution of IRH is displayed in Fig.10-Fig.12. The IRH of all samples saturated firstly and then unsaturated[50]. The test results show that the shale and CEA incorporation significantly improve the IRH.

Fig.10 IRH of concrete with shale addition

Fig.12 IRH of concrete incorporating shale and CEA: (a) 3% CEA; (b) 6% CEA

The time of duration of the saturation stage is greatly extended by the shale addition and is intensified with shale content, as Fig.10 shows. The duration time of the saturation stage of SL10 and SL15 occur 2.0 and 5.5 days later than that of WC035, respectively.When the cement hydration consumes the water in the capillary pores, the meniscus occurs and the capillary pore pressure exerts, which compels the shale to release the pre-absorbed water to alleviate the water consumption and maintain a longer saturation period.When IRH starts to decline, the IRH decline is impeded with shale incorporation and increases with shale dosage. WC035, SL10 and SL15 have a 28-day IRH of 87.9%, 92.3% and 95.4%, respectively, which indicates that a higher shale content incorporation in concrete can better inhibit self-desiccation development.

Compared with shale addition, the CEA incorporation has a relatively weaker effect on IRH,as Fig.11 shows. The self-desiccation begins slightly later with the CEA addition. EA3 and EA6 begin selfdesiccation 12.5 and 28.4 hours earlier than WC035,respectively. This is because the water in the sample with aw/bof 0.35 for CEA and cement hydration is insufficient, and CEA reacts with water at a faster rate than cement. Thus the water consumption in CEA concrete is faster than the reference sample at early age, and subsequently the IRH drops earlier.Meanwhile, during the unsaturation stage, the CEA incorporation could inhibit the IRH decline. The IRH of WC035, EA3 and EA6 at day 28 are 86.8%, 88.9%and 90.5%, respectively. This is attributed to that the CEA hydration consumes less water compared with the equal-mass cement, which leads to that the more water is left in capillary pores of CEA concrete than the plain concrete.

Fig.11 IRH of concrete with CEA addition

However, the combined incorporation of shale and CEA into concrete has a bit different from the above discussions. The CEA addition could prolong the saturation period of shale concrete, embodying as SL10EA6 (6.7 days) ＞ SL10EA3 (5.8 days) ＞ SL10(4.6 days), and SL15EA6 (10.2 days) ＞ SL15EA3 (8.7 days) ＞ SL15 (8.0 days). The IC water released from pre-wetted shale helps concrete going through the CEA-induced fast water consumption period at early age, thereafter the hydration of CEA requires less water and then contributes to a longer moisture saturation period. During the unsaturation stage, the IRH is improved by increasing shale (CEA) content in the presence of CEA (shale). The above-mentioned action mechanism and the synergistic mechanism of shale and CEA on IRH contribute to a better IRH performance than that with shale or CEA addition only.

3.4 Autogenous deformation

3.4.1 Autogenous deformation of concrete incorporating shale

The autogenous deformation development of WC035, SL10 and SL15 are exhibited in Fig.13. The positive value stands by the expansion deformation,otherwise the shrinkage deformation. All mixtures share a similar autogenous deformation trend, featuring by the expansion stage (0-t0days,t0is the age of the maximum expansion deformation (MED)) and the contraction stage (t0-28 days).

Fig.13 Autogenous deformation of concrete with shale addition: (a) the first 3 day; (b) the first 28 day

WC035 occurs a slight expansion during the first few hours, which is usually due to the re-absorption of the bleed water[51]. There is still some bleed water left although the sponge is adopted to remove the possible bleed water during the test in this study. Moreover,for SL10 and SL15, the time-dependent tendency of expansion deformation relies on the IC water amount of shale,i e, a larger shale content prompts the concrete to expand for a longer time and a bigger magnitude. The motivation may be that the water desorbed from shale could render the C-S-H layers which is unsaturated in the loww/benvironment of HPC, and open access for IC water continually entering into the interlayer pores which renders C-S-H gel swell and causes macroscopic expansion of cementitious materials; besides, the moisture increase decreases the surface tension of C-S-H gel and leads to a further expansion[52].Additionally, as discussed earlier, the DEM decreases with shale content (see Fig.9); therefore, the concrete with a higher shale content is restrained by a lowerstiffness skeleton and expands at a larger magnitude and for a longer period, which is proved by the test results.

After the early-age rapid expansion, the shale concrete exhibits a significant contraction. However,shale incorporation can significantly improve the AS behavior of concrete. A higher shale content incorporation has a better IC efficacy and could obtain a smaller AS. The 28-day AS of concrete declines by 49.3% and 82.4% with 10% and 15% shale incorporation, respectively. According to the above mention, the IC water released from shale keeps a higher IRH in concrete compared with the reference sample (see Fig.10); therefore, it delays the selfdesiccation, inhibits the increase of capillary pore pressure, and mitigates the AS[53]. Moreover, the 15%shale mixture even begins to shrink at day 15 and exhibits a very small shrinkage.

Meanwhile, the shrinkage rate is important for early-age crack control because a low rate is beneficial to sufficiently develop the tensile strength as well as the relaxation effect reducing tensile stress via the creep.The shrinkage rate can be calculated as Eq. (5)[54]:

whereR(t) ist-day shrinkage rate,με/day; andεsh(t) ist-day shrinkage strain,με.

The contraction rates of WC035, SL10 and SL15 are shown in Fig.14, which are consisted of a sharp decrease within the first 7 days, a slow decrease until 20 days, and subsequently a stable stage. The lower contraction rate is obtained by a higher shale content incorporation. Hence, the shale incorporation could effectively decrease the contraction rate, which is beneficial to help concrete resisting crack generation.

Fig.14 Contraction rate of concrete with shale addition

3.4.2 Autogenous deformation of concrete incorporating CEA

The evolution of autogenous deformation of CEA concrete is depicted in Fig.15. The CEA concrete exhibits a steep expansion, a sharp decrease and subsequently a stable stage[11]. As expected, the CEA incorporation evidently increases the expansion deformation. The CaO reacts with water and produces portlandite, providing crystalization pressure and swelling pressure. However, the hydration of CEA is rapid and almost complete within 24.0 hours, which is unbeneficial to AS-induced cracking control because of the small elastic modulus of concrete and the small chemical pre-stress within such period.

Fig.15 Autogenous deformation of concrete with CEA addition: (a) the first 3 day; (b) the first 28 day

During the contraction stage, the contraction trends of WC035, EA3 and EA6 are similar. But the contraction rate of concrete with CEA addition is slightly lower than that of the reference sample.Besides, it is noted that contraction value increases with CEA content. The contraction value increases approximately from 140 to 160 and 180 με as the CEA content increases from 0 to 3% and 6%. The reason is that a part of the water is absorbed by portlandite to induce swelling pressure[55]. Therefore, the higher the CEA dosage, the larger the absorbed water, the intenser the self-desiccation and the larger the contraction.Although the contraction is quite large, the shrinkage compensation efficacy of CEA is still obvious. A 25.4%and 44.0% reduction in 28-day AS are achieved by 3%and 6% CEA addition, respectively.

3.4.2 Autogenous deformation of concrete incorporating shale and CEA

The autogenous deformation results of concrete incorporating shale and CEA are presented in Fig.16,which also includes the expansion and contraction stage and are significantly affected by the combined addition of shale and CEA.

Fig.16 Autogenous deformation of concrete with the combined use of shale and CEA

During the expansion stage, the main interesting feature is that the sample incorporating shale and CEA shows a larger MED than the sample with shale or CEA only, and even larger than their sum. When the same CEA content is used, the higher shale addition results in a remarkably larger expansion deformation.The peak expansion value of SL15EA3 (SL15EA6)is 62.67% (76.09%) higher than that of SL10EA3(SL10EA6) and occurs 8.3 (19.5) days later. The combined incorporation of shale and CEA in concrete can not only enhance the expansion magnitude but change the process of expansion. As above-mentioned,the cement and CEA hydrate incompletely in the plain concrete withw/bequaling to 0.35. As expected, the IC water desorbed from pre-wetted shale can effectively improve the hydration of CEA, resulting in a high expansion efficiency, especially for the samples with 15% shale. Additionally, as Fig.9 shows, the DEM of concrete is reduced by the shale and CEA combined incorporation, leading to a smaller resistance against the expansion pressure generated by CEA and resulting in a higher expansion.

During the contraction stage, the contraction is characterized by two features according to shale content. One is that the contraction of 10% shale concrete shows a decreasing trend, although the curves drop slowly and slightly, which indicates the relatively lower shale incorporation could not adequately inhibit the shrinkage. Meanwhile, in the case of 15% shale,the autogenous deformation almost does not decrease and even shows a continuously slight expansion, which is mainly attributed to the IC efficacy of shale (see Fig.10). This deformation behavior could achieve an adequate relaxation effect which reduces shrinkageinduced tensile stress via creep. Besides, all samples keep expanding throughout the whole test period,indicating that the combined use of shale and CEA can completely eliminate the AS of concrete. Herein, it can be concluded that the shale enables to embody its effect at the contraction stage as well as the expansion stage.

Consequently, an expected synergistic mechanism is achieved, that is, the shale provides moisture to improve CEA expansion efficacy at early age; afterward, the water from pre-wetted shale continuously inhibits the contraction development.Such a mechanism is excellently beneficial to control shrinkage-generated cracking.

4 Conclusions

a) The shale or/and CEA addition strongly influence the setting times and mechanical properties.The shale addition postpones the setting times while the CEA has opposite results. The shale incorporation enhances the strength when the content ≤ 15% but weakens the strength when the content ＞ 15%, and continuously decreases the DEM. Meanwhile, the CEA incorporation deteriorates the mechanical properties.

b) Shale addition greatly inhibits the IRH decline and impedes the self-desiccation. The larger the shale content incorporation is, the longer the IRH maintains saturated, and the higher the IRH is during the unsaturated period. CEA addition has a slight improving effect on IRH. The combined incorporation of shale and CEA could further improve the IRH.

c) Incorporating shale into concrete could significantly mitigate the AS. Shale and CEA used together could obtain a synergistic effect on the AS,i e,the shale desorbs moisture to improve CEA expansion efficacy early and then inhibits the contraction later,which is effective for controlling AS-induced cracking.