Control of Early-Age Cracking in Super-Long Mass Concrete Structures

Abstract

The early-age cracking problem in concrete has long been recognised by civil engineers and scientists because it can jeopardise the intended serviceability of concrete structures. However, the effects of various crack control methods are different. This paper carried out field monitoring on a super-long wall with different crack control measures and compared the temperature and strain development process of the wall. In the middle of the super-long wall, the pipe cooling method reduced the hydration heat of concrete by 13 °C via a vertical pipe arrangement, but the wall could reach the maximum tensile strain earlier than with the other methods. By embedding an I-shaped steel plate in the induced joint method, a structurally stiff mutation zone was generated, and the maximum strain was generated at the induced seam web. By calculating and setting a reasonable construction length, the alternative bay construction method reduced the internal tensile strain of the structure. The early-age cracking of super-long mass concrete structures is affected more by restrained shrinkage than by temperature, so it is difficult to control early-age cracking by addressing only one factor.

1. Introduction

Early-age cracking, which occurs during the first several days after concrete casting, remains an ongoing problem of considerable economic significance in the concrete construction industry. Early-age cracking can seriously compromise the durability, strength, and long-term performance of concrete structures [1]. This is of particular concern in structures with super-long shapes and large exposed surface areas, such as bridge decks, dams, highway pavements, industrial and residential floors, and massive foundations [2,3,4,5]. Early-age cracking can be caused by shrinkage (including drying shrinkage and autogenous shrinkage) or thermal expansion. Moreover, cracking or strain-softening damage may occur due to internal restrains, which are generated by the temperature and moisture gradients of volumetric deformations, or external restrains, which are generated by mechanical limitation of strain caused by stable concrete of previously cast layers [6,7,8,9,10].

The appropriate selection of constituent materials in concrete, the correct mix design of concrete, the application of coarse cement, and precooling of concrete components are some common measures to minimise the risk of early-age cracking in concrete structures. To further solve the problem of early-age cracking in concrete, several studies have been carried out to enhance the material properties. Shrinkage-reducing admixture (SRA), which was invented by Goto et al. [11] in 1982, has been effectively used to reduce the shrinkage of concrete. Weiss et al. [12] reported that cement paste containing 5% SRA showed swelling with an expansion rate of approximately 235 mm/m in the early stage of hydration, and autogenous shrinkage of cement paste containing SRA was reduced by approximately 370 mm/m within 24–168 h, compared with that of the control cement paste. Expansive agents have been used to compensate for concrete shrinkage. Liu et al. [13] reported that the restrained expansion ratio of concrete with expansive agents rapidly increases within the first 3 days. The restrained expansion ratio reaches its peak value after 14 days and then decreases with the elapsed time. Internal curing has also been considered the most effective method for reducing autogenous shrinkage of concrete [14]. Tankasala and Schindler [15] evaluated the effect of using lightweight aggregate (expanded slate) on the early-age cracking tendency of mass concrete mixtures. The results show that the addition of LWA to concrete delayed the time to cracking, with SLW concrete providing the best overall resistance to early-age cracking. Klemczak et al. [16] discussed the impact of the aggregate type on the concrete hardening temperatures, induced stresses, and the cracking risk in the massive foundation slab. Through analysis, it was revealed that gravel aggregate, consisting mainly of quartz, can be considered as the worst aggregate, while granite aggregate may be regarded as the best.

In contrast to using the diverse array of material control methods, engineers expect to control the early cracking of concrete structures through design and construction. There is a concern in water conservancy projects, as the main applications of mass concrete, about early-age cracking caused by temperature stress. Early design and construction measures to eliminate the temperature stress in concrete are mostly aimed at this kind of structure. Successful implementation of post-cooling was demonstrated at the Owyhee Dam in Oregon in 1931 [17]. Since then, the post-cooling system has gained popularity and is being used in many large concrete structures [18]. To prevent early-age cracking, the specific settings of induced joints in the dam can induce cracks at these positions, thus reducing the effect of temperature cracks, which occur commonly in roller compacted concrete (RCC) arch dams. An axial tensile test as part of the both-way interval induced joint test of RCC samples was carried out by Zhang et al. [19], in which quantitative analysis was performed on the variation rules of its equivalent strength under different ages and degrees of weakening, thus further optimising the fracture test of RCC-induced joints and providing a basis for the design of induced joints in actual engineering practice. Different from the mass concrete in water conservancy projects, for super-long concrete structures in civil buildings, the influence of external constraints must be considered. In most countries, permanent joints or post-pouring joints are used to release temperature stress. The space between expansion joints is 30–40 m, which is set based on experience. Based on many engineering practices, Wang [20] proposed the construction technology of ‘resistance’ and ‘release’ combined with alternate bay construction, which has achieved good results in practical engineering applications. Adding longitudinal unbonded prestressed reinforcement to the wall is currently the most common anti-crack measure. Jiao et al. [21] analysed the defects of the existing prestressing methods and proposed an improved prestressing technique, which has been verified in engineering applications.

To accurately analyse the causes of early-age cracking of mass concrete structures, several experimental methods have been carried out to test the early performance of concrete. The ring test is perhaps the most frequently employed experimental method for the determination of the likelihood of cracking in a cementitious system subjected to restrained shrinkage, mainly drying shrinkage. Fragkoulis Kanavaris provided a state-of-the-art review of the widely adopted ring test method to evaluate the risk of cracking of cementitious materials under restrained shrinkage [22]. Due to standardisation, the ring test is commonly used as a method to examine the cracking sensitivity of cement-based materials. Among the different methods of measuring the early-age autogenous deformation of cementitious materials, the corrugated tube method described in the ASTMC1698-09 standard is also one of the most used. Mateusz Wyrzykowski et al. [23] presented a review of this method, along with a statistical assessment of the results of an experimental study on three cement pastes. It was found that the method is sensitive enough to resolve different levels of shrinkage and its evolution in time, and that the effect of the operator is not significant. Many types of equipment have been developed for the assessment of restrained thermal, drying, and/or autogenous shrinkage, but there are several uncertainties and limitations with their use [24]. The thermal and shrinkage deformations of concrete are usually restrained by boundaries in practical engineering. Restrained cracking under different temperature and shrinkage histories could be simulated by a temperature stress testing machine (TSTM) [25,26,27]. TSTM analysed the mechanical properties, such as the influence of the elastic modulus, creep or stress relaxation, the thermal expansion coefficient, and the thermal conductivity coefficient, and how these properties influence the cracking sensitivity of concrete [28,29,30,31,32,33].

The existing testing equipment is mainly used to evaluate the early crack resistance of the material, so it is not effective for studying the strain development in super-long structures. Most of the early crack control measures applied to super-long concrete structures consider a single measure, and comparative analysis of various crack control measures is lacking. Therefore, in this paper, three different crack control measures were tested and analysed in a super-long concrete structure. Early-age cracking occurs in concrete structures due to temperature differences and stress development. Since no material control measures were adopted, the elastic modulus of materials was similarly affected by age in all cases. Therefore, the temperature and strain were studied as indicators for evaluating the early-age cracking of concrete structures.

2. Project Background

This study was based on a two-layer island platform underground station constructed in Xiamen, with dimensions of 209.40 m × 19.70 m in horizontal view. The underground structure has two layers, with a depth of 17 m. The underground diaphragm wall is composed of a primary lining and another lining, and the lining thickness is 0.8 m. The construction monitoring time with the alternate bay construction method was from November to December 2017, with an average temperature of 20 °C. The construction monitoring time of the pipe cooling method and the induced joint method was from July to August 2018, with an average temperature of 29 °C. Steel formwork was used for casting concrete and removed after 72 h of curing.

2.1. Materials

Grade 42.5 cement and fly ash were adopted as the binder in the project.

Table 1. Ingredient ratios of different materials (%).

Material	CaO	SiO2	Al2O3	Fe2O3	MgO	SO3
Cement	52.27	25.54	8.13	5.59	1.85	2.27
Fly ash	2.98	78.61	15.14	3.80	0.56	0.55

shows the chemical compositions of the cement and fly ash.

Table 2. Physical properties of cement.

Fineness of the 80 μm Screen Margin (%)	Specific Surface Area (m2/kg)	Standard Consistency (%)	Loss on Ignition (%)	Time of Initial Setting (min)	Time of Final Setting (min)	Compressive Strength (MPa)
						3 d	28 d
1.726	345	25	1.75	161	218	27.4	50.4

shows the physical properties of cement. The fine aggregate and coarse aggregate used were qualified according to the Chinese standard JGJ 52-2006. The modulus of fineness of the fine aggregate was 2.62, with a soil content of less than 1%. The crushed limestone had a continuous grade with a size range from 5 to 31.5 mm.

Table 3. Concrete composition and compressive strength.

Code Grade	Concrete Composition (kg/m3)							Compressive Strength (MPa)
	Water	Cement	Fly Ash	Ground Slag	Sand	Gravel	Water Reducer	3 d	28 d	56 d
C45	151.2	294	94.5	31.5	713.2	1115.5	5.7%	47.2	60.0	63.7

shows the concrete composition and the compressive strength obtained in the mixing station.

2.2. Instrumentation

To evaluate the effect of different crack control methods, field tests of the concrete temperature and the strain of different sidewalls were performed for 7 days. The concrete strains of the walls were measured with vibrating wire strainmeters. The strainmeters were embedded at various locations of the wall in the middle of the concrete, with strain sensors to adjust thermal strains.

2.3. General Arrangement

2.3.1. Pipe Cooling Method

To avoid the early-age cracking of mass concrete structures due to temperature stresses, the early hydration heat of concrete must be effectively reduced. There are two methods of reducing the temperature of concrete [34,35]: precooling and post-cooling methods. The precooling method aims to reduce the temperature of concrete constituents, i.e., cement, water, and gravel, before they are mixed to form concrete. Post-cooling aims to remove the heat from the interior of the mass concrete through cooling pipes of circulating water embedded in the concrete.

A horizontal arrangement of the pipes in the pipe cooling method is typically used in mass concrete structures, which is suitable for super thick and large-area concrete structures but is not appropriate for slender structures such as super-long walls [36]. In this study, a vertical cooling pipe was used to increase the cooling effect of the wall, as shown in Figure 1. Post-cooling pipes were made of steel with a diameter of 30 mm and a thickness of 1 mm, and they were installed in the concrete wall at intervals of 1000 mm. During the construction of the structure, water at 22 °C was injected into the pipe at a rate of 3 L/min, and a pump was used to ensure the smooth flow of the cooling water. The supply of cooling water was discontinued gradually when the temperature of the concrete started to decrease.

The serpentine layout is shown in Figure 2. In the sidewall, additional steel bars should be erected to fix the cooling pipes. During the construction, attention should be paid to the protection of the pipeline to ensure a smooth flow of the cooling water. The position of the measuring points should be able to reflect the temperature change completely and accurately in the mass concrete and the change in the position under large strain. Five arrays of sensors were arranged equally at 3000 mm in the sidewall, as shown in Figure 2. Array 1 of sensors (11, 14, and 15) were placed near the inner surface of the segment. The interval between any two sensors was 2000 mm vertically. Other arrays were placed similar to array 1. Line set 1 (11, 21, 31, 41, and 51) was placed near the inside surface of the segment. Line set 2 (12, 22, 32, 42, and 52) was placed in the middle of the segment. Line set 3 (13, 23, 33, 43, and 53) was placed in the segment near the underground continuous wall.

2.3.2. Induced Joint Method

The design of the induced joint is based on the principle of limited stiffness to reduce structural shrinkage. When a small section with stiffness is artificially set along a certain direction of the structure, if the structure experiences adverse factors such as a temperature difference between the inside and outside, drying shrinkage, and foundation constraint, the induced joint section will open, and some structural energy will be released by cracks in the nearby areas to control the disorder and reduce the damage caused by cracks to the structure. The weakened sections created by the induced joints form an ‘outlet’ for the stress or deformation of the structure. Due to the reduction in the strength and rigidity of the structure, cracks are generated near these sections, and the energy is released. On the other hand, due to the separation of the induced joints, the length of the structure becomes shorter, and the stress of the structure is reduced.

In engineering, the induced joint method is usually used to reduce the amount of longitudinal reinforcement passing through the induced joint section and the concrete area of the section to induce orderly cracking of the structure. This approach is different from the conventional induced joint method used to reduce the structural stiffness and release the early-age temperature stress of the structure. In this study, an I-beam was inserted into the structure to create a sudden change in stiffness, which caused cracks in the web of the I-beam to release stress. At the same time, an I-beam flange plate was used to prevent the cracks from expanding to the surface of the structure and to prevent cracks from extending through the structure, as shown in Figure 3. The real-time data were collected, including the temperature and strain of concrete near the induced joint, and the date of the concrete in the middle of the segment. The distance between the two induced joints was 8000 mm. Both induced joints were 5000 mm away from the construction joints on both sides. The layouts of sensors are shown in Figure 4.

In addition, due to the influence of the constraint of the ground beam at the bottom of the sidewall, compared with the layout plan of the wall without an induced joint, the spacing of each group of measuring points should be appropriately increased to meet the layout requirements, and the layout of measuring points should be appropriately reduced in the upper area on both sides of the induced joint. Five arrays of sensors were arranged along the segment. Arrays 1, 2, 4, and 5 were located on both sides of the induced joint. Array 1 of sensors (11 and 14) were placed in the inner surface of the segment near the induced joint. Measuring point 11 was 1000 mm away from the construction joints at the bottom of the sidewall. The interval between any two sensors was 2000 mm from bottom to top. Arrays 2, 3, and 5 were placed similar to array 1. Array 3 was located in the middle of the sidewall, with three strain gauges that were arranged at heights of 1 m, 3 m, and 5 m. Along the cross-section, the strain scores were divided into three groups—areas located near the outer surface layer of the wall, the inner surface layer of the wall, and the middle layer of the wall.

2.3.3. Alternate Bay Method

The alternative bay construction method was used to set the construction joint in the vertical direction of the concrete structure and divide the concrete structure into several blocks according to a certain size by using the construction joint. The adjacent blocks were poured at intervals. After the concrete block was poured and had a large shrinkage deformation, it was connected to the other concrete blocks and poured as a whole. According to the nonlinear relationship between the temperature shrinkage stress and the structure length, the temperature difference and shrinkage deformation in the early stage (7–10 days) of concrete were relatively large, and the method of short-distance stress release was adopted to address the large early-stage shrinkage. After the early-stage large temperature difference and shrinkage of concrete, other blocks were poured to connect the concrete into a whole unit to address the subsequent small shrinkage [20].

The alternate bay construction method is a comprehensive measure to control the early-age cracking of concrete structures. The construction scheme includes a series of procedures, such as concrete mix proportion design, a reduction in the boundary constraints, setting of the construction length, and development of a maintenance system. The construction measures of the alternate bay construction method in this project were as follows:

• The construction length and construction interval time of the structure were calculated according to the site working conditions. The construction interval time was 7 days, and the construction length at one time was 18 m;
• The temperature of the concrete entering the formwork was controlled within 30 °C;
• A 15 mm thick piece of Styrofoam was attached between the primary lining and the other lining to allow volume expansion of the concrete;
• To eliminate defects from fast heat conduction in the steel formwork, the formwork was removed 3 days after the concrete was poured, maintenance was performed in time, and the maintenance system was adjusted according to the measured data.

The measuring points in the sidewall could comprehensively and accurately reflect the temperature change in the mass concrete and the change in the large strain location. The location and installation of sensors are also shown in Figure 5. Five arrays of sensors were arranged equally at 3000 mm in the sidewall. Array 2 of sensors (21, 24, and 25) were placed near the inner surface of the segment. The interval between any two sensors was 2000 mm vertically. Other arrays were placed similar to array 2. Line set 1 (11, 21, 31, 41, and 51) was placed near the inside surface of the segment. Line set 2 (12, 22, 32, 42, and 52) was placed at the middle of the segment. Line set 3 (13, 23, 33, 43, and 53) was placed in the segment near the underground continuous wall.

2.3.4. Summary

Approaches to decrease early-age cracking in concrete structures are covered in the list shown in

Table 4. Approaches to decrease early-age cracking in concrete structures.

Approaches	Measures
Pipe cooling method	A vertical cooling pipe is used to remove the heat from the interior of the mass concrete through cooling pipes of circulating water embedded in the concrete.
Induced joint method	I-beam is inserted into the structure to create a sudden change in stiffness, which causes cracks around the web of the I-beam to release stresses.
Alternate bay method	This systematic method includes a series of procedures, such as concrete mix proportion design, a reduction in the boundary constraints, setting of the construction length, and development of a maintenance system.

. The pipe cooling method focuses on reducing the hydration temperature; the induced joint method focuses on releasing the strain produced by the constraint; the alternate bay method comprehensively considers the effect of hydration temperature and constraint on the early-age cracking in concrete structures.

3. Results and Discussion

3.1. Pipe Cooling Method

The hydration heat and strain history measured in the middle of the sidewall with the pipe cooling method from 0 to 168 h are shown in Figure 6. The plots depict the temperature and strain change with time at the bottom of the sidewall. As shown in the figure, the concrete temperature development trends at various locations were similar. Figure 6a shows the temperature changes along the wall length. The maximum temperature of the sidewall appeared at point 32 in the middle of the sidewall after casting for 24 h, and the temperature peak was 66.9 °C. After reaching the maximum temperature, the concrete temperature decreased slowly until the measuring temperatures were close to the atmospheric temperature, and the temperature–time curve became flat. Figure 6c shows temperature changes through the wall thickness. The temperature at the centre was close to that near the underground continuous wall, and the temperature of the internal face was the lowest. After 80 h, the temperature of each point approached the same.

Different from the temperature curves at different positions, the strain curves at different positions changed greatly. Figure 6b shows the strain changes along the wall length. The maximum strain of the sidewall was located at point 32 in the middle of the wall. The strain was compressive strain within 0–8 h after casting, which gradually changed to tensile strain after 8 h and slowly decreased after 26 h. The cooling effect of the sidewall became weak far away from the water inlet of the cooling pipe, and a large constraint appeared between the sidewall and the underground continuous wall. Figure 6d shows strain changes through the wall thickness. The maximum strain of the sidewall was also located at point 32. Points 33 and 31 were restrained by underground continuous wall and steel formwork, respectively, and the strain attenuation was gentle.

3.2. Induced Joint Method

Figure 7 shows the hydration heat and strain history of the sidewall with the induced joint method from 0 to 168 h and describes the changing trend of temperature and strain with time at the measuring points at the bottom of the sidewall. The temperature changes along the wall length are presented in Figure 7a. The maximum temperature of the sidewall concrete appeared after casting for 20 h, which was measured at the induced seam web point 42, and the maximum temperature was 79.9 °C. After 20 h, the concrete temperature decreased slowly until the internal temperature was close to the atmospheric temperature, and the temperature–time change curve reached a plateau. The temperature changes in the wall thickness are presented in Figure 7c. The maximum temperature of the sidewall cross-sections appeared in the induced seam web.

The strain changes along the wall length are presented in Figure 7b. The maximum strain of the sidewall was measured at point 22 of the induced seam web. The strain was compressive strain within 0–6 h after casting and gradually changed to tensile strain after 6 h. The strain reached the peak after 48 h and slowly decreased. The temperature changes in the wall thickness are presented in Figure 7d. The maximum strain value of the sidewall cross-sections appeared in the induced seam web. The construction of I-steel-induced joints changed the stiffness of the corresponding wall, which caused massive strains.

3.3. Alternate Bay Construction Method

Figure 8 shows the hydration heat and strain history of the sidewall with the alternate bay construction method from 0 to 168 h, and it depicts the changing trend of temperature and strain with time at the bottom of the sidewall. The temperature changes along the wall length are presented in Figure 8a. The maximum temperature increased after casting for 24 h. The maximum temperature point was measured at the middle part of the wall, and the temperature peak was 55.7 °C. After 24 h, the concrete temperature decreased slowly until the measuring temperatures were close to the atmospheric temperature, and the temperature–time curve became flat. The temperature changes in the wall thickness are presented in Figure 8c. The maximum temperature of the wall cross-section appeared in the middle of the wall.

The strain changes along the wall length are presented in Figure 8b. The maximum strain of the sidewall was located at point 32 in the middle of the wall. The strain was a compressive strain in the period from 0 to 28 h after casting and gradually changed to tensile strain after 28 h. The strain reached its peak after 94 h and slowly decreased. The temperature changes in the wall thickness are presented in Figure 8d. The maximum strain of the sidewall cross-section appeared in the middle of the wall. Under the combined action of temperature and constraint, the structure generated massive strains in the middle of the wall.

3.4. Relationship between Temperature and Strain

Both temperature and constraints are the main reasons for the early cracking of super-long structures. Figure 9 shows the variation trends of temperature and strain in construction applications with different crack control methods and the correlation between temperature and the early deformation of structures. In the sidewall with the pipe cooling method, the temperature peak value and the strain peak value of the wall were reached at the same time. The hydration heat can be effectively controlled by the post-cooling method. In the sidewall with the induced joint method, the time at which the wall reached the peak strain was slightly delayed by the peak temperature increase. The induced joint effectively released the constraint strain in the structure, but the temperature still affected the development of early strain in the structure. In the sidewall with the alternate bay construction method, the time for the wall to reach the strain peak obviously lagged the peak temperature rise. A reasonable construction process and a reduction in constraints effectively delay the development of structural strain.

3.5. Analysis of Different Control Methods

Figure 10 presents the concrete hydration heat and strain histories evaluated under different measures from 0 to 168 h. The change in temperature and strain in the middle of the sidewall with time is described. The temperature in the sidewall reached the peak nearly at the same time for all tested methods, and it did not change with a change in the control method; however, the shape of the time variation changed with the temperature curve. The gradient of the temperature drop was the steepest with the induced joint method and the gentlest with the alternate bay construction method. After removing the influence of ambient temperature, the pipe cooling method reduced the peak temperature rise by 13 °C, compared with the induced joint method.

The strain vs. time variation curve of the wall showed a substantial influence on different control measures. It is noted that the strain histories obtained from all projects showed the typical early-age behaviour trend, which is the initial compressive strain that was gradually released and transformed into tensile strain. In the compression strain section of the strain history under different measures, the pipe cooling method had the smallest strain, the induced joint construction had the largest, and the alternate bay construction showed the most drastic change. The maximum tensile strain was obtained by the pipe cooling method at 26 h, and the maximum tensile strain was obtained by the alternate bay construction method at 94 h. The induced joint method continued to increase slowly after 40 h. The tensile strain of the super-long concrete structure is caused by both temperature and constraint, with constraint being the main cause. Controlling only the temperature cannot relieve the early tensile strain of super-long concrete structures. Constraint release can be induced to lower the early structural strain to a certain extent. Simultaneous temperature control and constraint release can effectively reduce the early tensile strain of super-long concrete structures.

4. Conclusions

In this study, three kinds of crack control methods applied to super-long mass concrete structures were evaluated in an underground continuous wall of a subway station platform. The following conclusions may be drawn based on the results of this study:

• The vertical pipe cooling method with water cooling in the pouring period was applied, which reduced the heat of hydration in the concrete structure. The temperature of the sidewall structure reached its peak value in 24 h and reached the ambient temperature in 120 h. The pipe cooling method reduced the hydration heat by approximately 13 °C and effectively reduced the internal temperature strain of the structure. The time variation law of wall strain and the temperature was synchronous. The vertical pipe cooling method reduces the hydration heat of materials and is more suitable for early-age cracking structures that are only sensitive to temperature changes.
• An I-beam was inserted into the structure to create a sudden change in stiffness, and this induced joint method reduced the internal restraint strain of the structure. The maximum temperature and strain peak in the structure appeared in the web, the strain in the middle of the structure grew slowly, and the changes in strain over time lagged behind the changes in temperature. The induced joint method releases the restraint strain and is more suitable for early-age cracking structures that are only sensitive to constraints.
• The alternate bay construction method is a comprehensive measure to control the early-age cracking of concrete structures. By reasonably setting the construction length and the construction interval time, the alternate bay construction method reduces structural constraints, optimises the maintenance system, delays the development process of structural temperature and strain, and reduces the internal tensile strain of the structure. The peak values of the temperature and strain of the structure appeared in the middle of the wall, and the change in strain over time lagged behind the change in temperature. The alternate bay construction method, which comprehensively considers the influence of both temperature and restraint, is the most effective measure.