Experimental Study of Interfacial Bond Properties between CGM and Existing Concrete

: To ensure that the cementitious grouting material (CGM) layer and existing concrete work together, the interfacial bond strength between CGM and normal concrete substrate was experimentally investigated by conducting direct and slant shear tests. The effects of connection interface form, concrete strength, interface shear key, and shear angle on the bond strength of CGM and normal concrete were discussed in detail. The results showed that for composite specimens, the interfacial bond strength was the highest for the triangular interface form, followed by the trapezoidal interface and smooth interface, which had the lowest. The interfacial bond strength increased as the existing concrete’s strength increased; the interfacial bond strength could be greatly improved under shear load in the slant section or when the shear key was set at the interface. Moreover, the interfacial bond strength increased with the increase in the shear angle. The interfacial bond–slip curves were analyzed, and a bond–slip model of the interface between CGM and concrete was proposed. A certain value of interfacial bond stiffness was also recommended.


Introduction
When an existing concrete structure's load-bearing capacity no longer meets the design requirements due to the increase in service time and the effects of natural environmental factors, repair and reinforcement are necessary to avoid costly reconstruction. The limitations of traditional approaches using fresh concrete as the repair material, such as long wet working time and difficulties associated with concrete pouring and vibration, have become increasingly prominent. A cementitious grouting material (CGM) is formulated with a high-strength material as the aggregate and cement as the binder, and it has high fluidity, micro-expansion, and anti-segregation capability. It also has good self-flow (vibration free), fast hardening, high early strength, stable performance, and convenient construction [1,2]. Using CGM instead of new concrete as a reinforcement material can overcome the limitations of traditional concrete methods [3]. Moreover, with the vigorous promotion and construction of prefabricated buildings, CGM is often used as a bonding material for connecting prefabricated concrete components. Whether for the retrofitting of concrete structures or as a bonding material for connecting assembly components, the reliability of the interfacial adhesive properties between CGM and concrete should be guaranteed to form a whole, and the two should work together. Statistics show that many concrete reinforcement structures develop cracks on the bonding surface during their service period, and the overall structure is damaged due to the cracking of the bonding surface, resulting in the failure of repair and reinforcement [4]. Therefore, the bond properties of the interface between CGM and existing concrete must be studied to maximize the excellent performance of CGM and ensure that the CGM layer and existing concrete work together.
Thus far, scholars have studied the interfacial bond of the repair layer and existing concrete and have achieved useful results [5][6][7][8][9][10][11][12]. Zambas [5] investigated the influence of shear angle on the interfacial bonding performance of new and old concrete and found that failure occurred along the interface when the shear angle was not more than 40 • , and the weaker concrete was crushed when the shear angle was more than 40 • . Ceia et al. [6] studied the influence of interface roughness on the interfacial bond strength between recycled aggregate concrete and natural aggregate concrete and found that the interfacial bond strength increased with an increase in roughness. Jafarinejad et al. [7] studied the bond strength between ultra-high-strength fiber reinforced cementitious mortar and conventional concrete. Their results showed that the bond strength depended greatly on the method of substrate surface treatment and pointed out that sandblasting was the best method for interface treatment and preparation. Mirmoghtadei et al. [8] examined the influence of interface treatment methods on the interfacial bond strength of normal concrete reinforced with metakaolin containing concrete and showed that the grooved acid etching treatment method provided the highest bond strength compared with other types of interface treatment techniques. Peng et al. [11] studied the interfacial bond strength between cementitious grouting material and ordinary concrete by using 30 • slant shear specimens and drilling holes at the interface and concluded that the bond strength increased with an increase in roughness. Yuan et al. [12] conducted a direct shear test to investigate the bond strength between cementitious grouting material and existing concrete and declaimed that the strength of cementitious grouting material had little influence on the interfacial bond strength.
As mentioned above, current studies on the interfacial bond properties of CGM and concrete is very limited, and the influence of different shear angles, different interface structures, and interface shear key settings remains unclear. Moreover, as far as we know, no literature on the interfacial bond-slip model (which is helpful for fine finite element simulation analysis) of normal concrete reinforced with CGM has been published. The bond performance at the interface is the key to ensuring the bearing capacity of reinforced or prefabricated structures. Thus, conducting in-depth and systematic research on the bond performance of CGM and existing concrete is necessary. Interfacial shear performance is an important index to evaluate interfacial bond performance, especially for areas with high shear stress where the possibility of shear failure is high. Interface shear resistance can be evaluated through direct [13] and slant [14] shear tests. The commercial CGM-I type grouting material suitable for repair and reinforcement engineering was used as the reinforcement material in this study. By considering the influencing factors, such as the existing concrete's strength, interface form, shear key setting, and shear angle, direct and slant shear specimens were designed and prepared from the perspective of engineering practicality. Loading tests were performed to systematically analyze the factors that affect the bond between CGM and concrete. A bond-slip model of the interface between CGM and existing concrete was proposed. The experimental data in this study can be beneficial for reasonably optimizing the structural design of the bond interface and improving the bearing capacity of reinforced structures. This study also provides a scientific basis for using CGM as a bonding material to connect assembly components.

Specimen Design
In accordance with the principle that the specimen form should be as close as possible to the shear transfer mode of the joint surface between the repair material and the existing concrete in the reinforced structure, the Z-shaped specimens shown in Figure 1 were designed to conduct a direct shear test on the interface bond performance between CGM and existing concrete by referring to existing shear test data [15][16][17][18]. The uniformity of the shear stress distribution on the bonding interface and the simple operability of the loading test were also comprehensively considered. The interface forms were set as smooth, triangular, and trapezoidal. The smooth interface was used to simulate a flat interface without any roughening in the process of structure repair, and the triangle and trapezoidal Appl. Sci. 2022, 12, 12623 3 of 14 interfaces were employed to simulate an uneven interface during structure rehabilitation. Three groups of specimens with triangular interface shapes and concrete compressive strength grades of C35, C45, and C55 were designed to study the effect of existing concrete strength on the interface bond performance. The parameters of the specimens for the direct shear test are listed in Table 1. test were also comprehensively considered. The interface forms were set as smooth, triangular, and trapezoidal. The smooth interface was used to simulate a flat interface without any roughening in the process of structure repair, and the triangle and trapezoidal interfaces were employed to simulate an uneven interface during structure rehabilitation. Three groups of specimens with triangular interface shapes and concrete compressive strength grades of C35, C45, and C55 were designed to study the effect of existing concrete strength on the interface bond performance. The parameters of the specimens for the direct shear test are listed in Table 1.   Z-C35-I  C35  smooth  none  3  Z-C35-T  C35  trapezoidal  none  3  Z-C35-S  C35  triangle  none  3  Z-C45-S  C45  triangle  none  3  Z-C55-S  C55  triangle  none  3 To study the interfacial bond performance in the compressive-shear composite stress state and analyze the influence of different shear angles on the interfacial bond performance of CGM and existing concrete, four groups of specimens with and without shear keys were designed and prepared. The concrete strength grade was C35, and slant shear angle, α, was 30° and 45°, as illustrated in Figure 2. Two groups of specimens with shear keys using concrete strength grades of C45 and C55 and shear angle α = 45° were also designed and prepared to explore the influence of existing concrete strength on the interfacial bond performance. The dimensions of all specimens were 100 mm × 100 mm × 300 mm, and the dimensions of shear keys were 50 mm × 20 mm × 100 mm (same thickness as the specimens). The parameters of the specimens for the slant shear test are given in Table  2.
To study the interfacial bond performance in the compressive-shear composite stress state and analyze the influence of different shear angles on the interfacial bond performance of CGM and existing concrete, four groups of specimens with and without shear keys were designed and prepared. The concrete strength grade was C35, and slant shear angle, α, was 30 • and 45 • , as illustrated in Figure 2. Two groups of specimens with shear keys using concrete strength grades of C45 and C55 and shear angle α = 45 • were also designed and prepared to explore the influence of existing concrete strength on the interfacial bond performance. The dimensions of all specimens were 100 mm × 100 mm × 300 mm, and the dimensions of shear keys were 50 mm × 20 mm × 100 mm (same thickness as the specimens). The parameters of the specimens for the slant shear test are given in Table 2 with 3 Note: X denotes slant shear specimens, W denotes without shear key, and Y denotes with shear key.

Specimen Preparation
Each test specimen was made of two materials; normal concrete was used as the substrate, and CGM was used as the reinforcement material. The cement, sand, and stones used to prepare the concrete were produced by Beijing Guoxingsheng Hardware and Building Materials Company, and their cement strengths were 32.5, 42.5, and 52.5 MPa, respectively. Medium sand with a fineness modulus greater than 2.8 and crushed stones with aggregate sizes ranging from 5 mm to 10 mm were used as fine aggregate and coarse aggregate, respectively. Potable tap water was utilized for mixing. According to the Chinese code Specification for mix proportion design of ordinary concrete [19], the materials and concrete mix procedure was designed. After many trials, the mixture proportions of each concrete strength were determined and are presented in Table 3. The CGM used as overlay materials in this study was CGM-Ⅰ type commercial cementitious grouting material from Zhongde Xinya Company. In accordance with the instructions for use, the strength and fluidity of CGM were measured after trial preparation. The result showed that the water added to the dry material accounted for 11.9%.
Note: X denotes slant shear specimens, W denotes without shear key, and Y denotes with shear key.

Specimen Preparation
Each test specimen was made of two materials; normal concrete was used as the substrate, and CGM was used as the reinforcement material. The cement, sand, and stones used to prepare the concrete were produced by Beijing Guoxingsheng Hardware and Building Materials Company, and their cement strengths were 32.5, 42.5, and 52.5 MPa, respectively. Medium sand with a fineness modulus greater than 2.8 and crushed stones with aggregate sizes ranging from 5 mm to 10 mm were used as fine aggregate and coarse aggregate, respectively. Potable tap water was utilized for mixing. According to the Chinese code Specification for mix proportion design of ordinary concrete [19], the materials and concrete mix procedure was designed. After many trials, the mixture proportions of each concrete strength were determined and are presented in Table 3. The CGM used as overlay materials in this study was CGM-I type commercial cementitious grouting material from Zhongde Xinya Company. In accordance with the instructions for use, the strength and fluidity of CGM were measured after trial preparation. The result showed that the water added to the dry material accounted for 11.9%. The preparation process of all test specimens was as follows. First, the normal concrete substrate was poured in accordance with the mix proportions listed in Table 3. Second, the mold was removed after 24 h of film curing and the concrete specimens were sent to the standard curing room for curing. Third, after 28 days of standard curing (at a temperature of 20 ± 3 • C and relative humidity greater than 95%), the surface of the substrate was cleaned to remove dust and oil and then dried with a cloth. Fourth, the concrete substrate of the specimen was placed in the original mold with the contact surface perpendicular to the bottom surface. The remaining half of the mold was filled with CGM, and the mold was removed after 24 h and the concrete-CGM test specimens were sent to the standard curing room (at a temperature of 20 ± 1 • C and relative humidity greater than 95%). Lastly, a loading test was conducted after the CGM standard curing for 28 days. While pouring the concrete substate part of the test specimens, three concrete blocks with dimensions of 150 mm × 150 mm × 150 mm were cast at the same time, and when pouring the CGM part of the test specimen, three CGM blocks with dimensions of 100 mm × 100 mm × 100 mm were made at the same time. The measured average test-day compressive strengths of the concrete and CGM are listed in Table 4.

Measuring Point Arrangement
Given that the relative bond slip between CGM and the concrete interface needed to be obtained in the test, two rebound-type self-resetting linear variable differential transformer (LVDT) displacement sensors with an accuracy of 0.01 mm and a range of 10 mm were symmetrically arranged on both sides of the adhesive interface to eliminate the possible eccentricity effect in the test, as shown in Figure 3. However, the displacement observed in this way includes the deformation of the concrete and CGM itself. According to the conservation law of work, the relative slip of the interface should not include the influence of the deformation of the concrete and CGM itself. Four strain gauges numbered C1, C2, G1, and G2 were symmetrically pasted to the concrete and CGM of the specimen to measure the deformation of the two materials, as shown in Figure 3. Hence, the relative slip, S, of the interface should be where S LVDT is the average value of the displacement measured by the displacement meter, S C is the average value of concrete displacement, and S G is the average value of CGM displacement. Figure 3 shows that the LVDT displacement meter measured the relative slip within the range of 250 mm in the middle of the specimen, and the strain gauge measured the material strain in the corresponding area, that is, the CGM and concrete should have a strain in the range of 200 − 25 = 175 mm. The concrete strain and grout strain in this range Appl. Sci. 2022, 12, 12623 6 of 14 were assumed to be the same, so S C = ε C l = 175ε C and S G = ε G l = 175ε G , where ε C and ε G are the measured average strains of the concrete and CGM, respectively.
Appl. Sci. 2022, 12, x FOR PEER REVIEW 6 of 15 strain in the range of 200 − 25 = 175 mm. The concrete strain and grout strain in this range were assumed to be the same, so 175  are the measured average strains of the concrete and CGM, respectively.

Test Scheme
A spoke-type LTH-LX pressure sensor with a range of 500 kN was placed at the bottom of the specimen (on the bottom steel plate of the loading machine) to measure the pressure transmitted to the bottom of the specimen, which was the load borne by the specimen. The bearing capacity of the specimen, the displacement value of the displacement meter, and the strain value of the strain gauge were determined synchronously through the DHDAS dynamic signal test and analysis system. For the test specimen to be in a pure shear stress state, attention should be paid to the placement and alignment of the specimen. A preload of 1 kN was imposed on the specimen before each test to ensure that the displacement meter and force sensor were firmly fixed and that the test device was connected reliably. After the preloading was completed, the connection of the instrument and the correctness of the data display were checked. When no error was found, the instrument was reset to zero, and the formal loading and data collection were initialized. The test was terminated when the interface or the test specimen failed. A photo of the specimen loading is given in Figure 4. The displacement control method was adopted for loading, and the loading rate was 0.1 mm/min.

Test Scheme
A spoke-type LTH-LX pressure sensor with a range of 500 kN was placed at the bottom of the specimen (on the bottom steel plate of the loading machine) to measure the pressure transmitted to the bottom of the specimen, which was the load borne by the specimen. The bearing capacity of the specimen, the displacement value of the displacement meter, and the strain value of the strain gauge were determined synchronously through the DHDAS dynamic signal test and analysis system. For the test specimen to be in a pure shear stress state, attention should be paid to the placement and alignment of the specimen. A preload of 1 kN was imposed on the specimen before each test to ensure that the displacement meter and force sensor were firmly fixed and that the test device was connected reliably. After the preloading was completed, the connection of the instrument and the correctness of the data display were checked. When no error was found, the instrument was reset to zero, and the formal loading and data collection were initialized. The test was terminated when the interface or the test specimen failed. A photo of the specimen loading is given in Figure 4. The displacement control method was adopted for loading, and the loading rate was 0.1 mm/min. A group of specimens were poured for trail loading before the formal test to ensure the success of the test, and the result showed that cracks appeared at the L-shaped corner of the specimen earlier than they did in the interface area. Thus, by referring to the method A group of specimens were poured for trail loading before the formal test to ensure the success of the test, and the result showed that cracks appeared at the L-shaped corner of the specimen earlier than they did in the interface area. Thus, by referring to the method used in Reference [16], a fiber-reinforced polymer cloth was pasted on both sides of the specimen for reinforcement after curing was completed to avoid this phenomenon. This method can also prevent the specimen from being damaged due to eccentric compression during the test. For the direct shear composite specimens with a smooth interface, the failure mode observed in the test was always adhesion failure at the interface, as presented in Figure 5a; the direct shear strength mainly depended on the chemical adhesion action and friction between CGM and the concrete substrate. For the composite specimens with a trapezoid or triangle interface form, interfacial failure with partial concrete substrate failure and CGM failure was observed, as shown in Figure 5b; part of the trapezoidal or triangular serrations on the connection interface were cut off, and the direct shear bond strength gradually depended on the properties of the concrete substrate and CGM. Hence, the increased concrete strength of the composite specimens resulted in high shear bond strength of the composite specimens with trapezoid or triangle interface forms. A group of specimens were poured for trail loading before the formal test to ensure the success of the test, and the result showed that cracks appeared at the L-shaped corner of the specimen earlier than they did in the interface area. Thus, by referring to the method used in Reference [16], a fiber-reinforced polymer cloth was pasted on both sides of the specimen for reinforcement after curing was completed to avoid this phenomenon. This method can also prevent the specimen from being damaged due to eccentric compression during the test.

Failure Modes of Direct Shear Specimens
For the direct shear composite specimens with a smooth interface, the failure mode observed in the test was always adhesion failure at the interface, as presented in Figure  5a; the direct shear strength mainly depended on the chemical adhesion action and friction between CGM and the concrete substrate. For the composite specimens with a trapezoid or triangle interface form, interfacial failure with partial concrete substrate failure and CGM failure was observed, as shown in Figure 5b; part of the trapezoidal or triangular serrations on the connection interface were cut off, and the direct shear bond strength gradually depended on the properties of the concrete substrate and CGM. Hence, the increased concrete strength of the composite specimens resulted in high shear bond strength of the composite specimens with trapezoid or triangle interface forms.

Failure Modes of Slant Shear Specimens
For the slant composite specimens without shear keys, the failure mode observed in the test was always adhesion failure at the interface, as shown in Figure 6a; the slant shear strength mainly depended on the chemical adhesion action between CGM and the concrete substrate. For the slant composite specimens with shear key at the interface, shear key being cut off was observed as the failure mode when the shear angle was 30 • , as presented in Figure 6b; the slant shear strength mainly depended on the shear strength of the shear key's own material. As the shear angle increased to 45 • , interfacial failure with concrete substrate and partial CGM failure was observed, as illustrated in Figure 6c. The slant shear bond strength gradually depended on the properties of the concrete substrate and CGM. Hence, the increased concrete strength of the composite specimens resulted in high shear bond strength of the composite specimens with shear keys. key being cut off was observed as the failure mode when the shear angle was 30°, as presented in Figure 6b; the slant shear strength mainly depended on the shear strength of the shear key's own material. As the shear angle increased to 45°, interfacial failure with concrete substrate and partial CGM failure was observed, as illustrated in Figure 6c. The slant shear bond strength gradually depended on the properties of the concrete substrate and CGM. Hence, the increased concrete strength of the composite specimens resulted in high shear bond strength of the composite specimens with shear keys.

Interfacial Bond Strength Analysis
The average bond strength of the interface [16] was used as the shear bond strength of the interface between CGM and existing concrete in this study, and it can be calculated as where max  is the interface bond strength (MPa); F is the failure load (N); α denotes the shear angle (α = 0° for the direct shear test and α = 30° and 45° for the slant shear test); and A is the interface bond area, which was considered to have a nominal value of 100 × 100 = 10,000 mm 2 for the direct shear specimen and 100 × 100/sinα mm 2 for the slant shear specimen. Table 5 presents the shear bond strength of each specimen.

Interfacial Bond Strength Analysis
The average bond strength of the interface [16] was used as the shear bond strength of the interface between CGM and existing concrete in this study, and it can be calculated as where τ max is the interface bond strength (MPa); F is the failure load (N); α denotes the shear angle (α = 0 • for the direct shear test and α = 30 • and 45 • for the slant shear test); and A is the interface bond area, which was considered to have a nominal value of 100 × 100 = 10,000 mm 2 for the direct shear specimen and 100 × 100/sinα mm 2 for the slant shear specimen. Table 5 presents the shear bond strength of each specimen. Table 5 indicates that, for the direct shear specimens with a concrete strength grade of C35, the interfacial shear bond strength from high to low was in the order of triangular interface, trapezoidal interface, and smooth interface. The bond strengths of the specimens with triangular and trapezoidal interfaces were much larger than that of the specimen with a smooth interface (i.e., 178.76% and 156.64% larger than that of the smooth interface, respectively). This result was obtained because the bond strength of the specimen with a smooth interface was provided only by friction force and the chemical adhesive force generated with the existing concrete when CGM was poured. Meanwhile, the bond strengths of the specimens with trapezoidal and triangular interfaces were mainly provided by the shear strength of serrated concrete and CGM itself at the interface, and the friction force and chemical adhesive force between them were relatively small. The bond strength of the specimen with a triangular interface was 8.62% higher than that of the specimen with a trapezoidal interface because the actual bonding area between CGM and concrete of the triangular-interface specimen was larger than that of the trapezoidal-interface specimen. Table 5 indicates that the bond strength of the interface increased with the increase in the existing concrete's strength, but the increase rate decreased. For example, for the direct shear specimen with a triangular interface without shear key, when the existing concrete strength increased from C35 to C45 and C55, the interface shear bond strength increased from 3.15 MPa to 3.50 and 3.73 MPa, and the increase rate decreased from 11.11% to 6.57%. For the slant shear specimen with shear key at a shear angle of 45 • , when the existing concrete strength increased from C35 to C45 and C55, the interface shear bond strength increased from 11.52 MPa to 14.47 and 15.54 MPa, and the increase rate decreased from 25.61% to 7.39%. 15.52

Influence of Existing Concrete's Strength
Note: The column "discreteness" in the table shows the absolute value of the ratio of the difference between the maximum or minimum value and the median value of the bond strength of the three specimens in the same group to the median value. When one of them is greater than 15%, the adopted interfacial bond strength value, τ max , is taken as the minimum value of the three values; when both of them are less than 15%, the adopted interfacial bond strength value, τ max , is the average of the three measured values. Figure 7 presents a comparison of the interfacial bond strengths of specimens with and without shear keys at different shear angles. The bond strength of the interface was greatly improved when the shear key was set for the same shear angle. Specifically, the interfacial bond strengths of the specimens with shear angles of 30 • and 45 • increased by 60.42% and 37.31%, respectively. Hence, setting shear keys can effectively improve the interface bond strength.

Influence of Shear Angle
The comparison of the interfacial bond strengths of the specimens in the Z-C35-I, X-C35-30-W, and X-C35-45-W groups showed that when the shear key was not set, the interfacial bond strength of the slant shear specimen was greatly improved compared with that of the direct shear specimens; the larger the slant shear angle, the more the interface bond strength was improved. Compared with the direct shear specimen at a shear angle of 0 • , the interfacial bond strength increased by 409.73% when the shear angle was 30 • and increased by 642.48% when the shear angle was 45 • . The presence of and the increase in pressure significantly improved the interfacial bond strength.
The comparison of the interfacial bond strengths of the specimens in the X-C35-30-W, X-C35-45-W, X-C35-30-Y, and X-C35-45-Y groups revealed that when the shear angle increased from 30 • to 45 • , the bond strength of the interface without shear keys increased from 5.76 MPa to 8.39 MPa (an increase of 45.66%), and the bond strength of the interface with shear keys increased from 9.24 MPa to 11.52 MPa (an increase of 24.68%). With the increase in the slant shear angle, the increase rate of the interfacial bond strength of the specimens with shear keys exhibited a decreasing trend, that is, the effect of shear key tended to decrease as the compressive stress on the specimen increased.
According to the ACI standard [20], the interface bond strength of a direct shear specimen at 28 days should reach 3-4 MPa, and the interface bond strength of a slant shear specimen should reach 14-21 MPa. The direct shear specimens with smooth and trapezoidal interfaces in this study did not meet these requirements, but the direct shear specimens with triangular interfaces all met the above-mentioned ACI requirements. With regard to the slant shear specimens, except for the X-C45 and X-C55 specimens that met the ACI requirements, the rest did not meet the ACI requirements.
Appl. Sci. 2022, 12, x FOR PEER REVIEW 11.11% to 6.57%. For the slant shear specimen with shear key at a shear angle of 45 the existing concrete strength increased from C35 to C45 and C55, the interface she strength increased from 11.52 MPa to 14.47 and 15.54 MPa, and the increase rate de from 25.61% to 7.39%. Figure 7 presents a comparison of the interfacial bond strengths of specime and without shear keys at different shear angles. The bond strength of the interf greatly improved when the shear key was set for the same shear angle. Specific interfacial bond strengths of the specimens with shear angles of 30° and 45° incre 60.42% and 37.31%, respectively. Hence, setting shear keys can effectively imp interface bond strength.

Influence of Shear Angle
The comparison of the interfacial bond strengths of the specimens in the Z-C C35-30-W, and X-C35-45-W groups showed that when the shear key was not se terfacial bond strength of the slant shear specimen was greatly improved compar that of the direct shear specimens; the larger the slant shear angle, the more the i bond strength was improved. Compared with the direct shear specimen at a she of 0°, the interfacial bond strength increased by 409.73% when the shear angle was increased by 642.48% when the shear angle was 45°. The presence of and the inc pressure significantly improved the interfacial bond strength.
The comparison of the interfacial bond strengths of the specimens in the X W, X-C35-45-W, X-C35-30-Y, and X-C35-45-Y groups revealed that when the she increased from 30° to 45°, the bond strength of the interface without shear keys in from 5.76 MPa to 8.39 MPa (an increase of 45.66%), and the bond strength of the i with shear keys increased from 9.24 MPa to 11.52 MPa (an increase of 24.68%). W increase in the slant shear angle, the increase rate of the interfacial bond strengt specimens with shear keys exhibited a decreasing trend, that is, the effect of sh tended to decrease as the compressive stress on the specimen increased.
According to the ACI standard [20], the interface bond strength of a direct she imen at 28 days should reach 3-4 MPa, and the interface bond strength of a sla specimen should reach 14-21 MPa. The direct shear specimens with smooth and t dal interfaces in this study did not meet these requirements, but the direct shear sp with triangular interfaces all met the above-mentioned ACI requirements. With r the slant shear specimens, except for the X-C45 and X-C55 specimens that met

Interfacial Bond-Slip Curve
The bond-slip relationship and bond stiffness of the interface are important indicators that reflect the bond performance of the interface. The test results of this study showed that the interfacial bond-slip curves of all the specimens were basically similar. They all exhibited a slow upward trend at the initial stage of specimen loading. After reaching the peak, the load suddenly and sharply decreased, and the displacement increased rapidly, resulting in a relatively insufficient data acquisition in the descending phase. Therefore, considering the reliability of the regression analysis data, this study preferred to safely ignore the bearing capacity of the descending phase, and only the test data of the ascending stage (before failure) of the interface bond-slip curve were used for regression analysis. Figures 8 and 9 show the ascent test curves of interfacial shear bond stress, τ, versus the interfacial slip, s, of each group of specimens. Most of the rising sections of the test specimens were close to a straight line. Therefore, the linear method was used to conduct a regression analysis of the test curves. The regression equation was assumed to be where τ is the interface bond stress, τ 0 is the critical bond stress before the interface slipped, K is the interface bond-slip curve stiffness, s is the relative slip of the interface, and s u is the slip when linear segment bond strength, τ u , is reached. The specific values and correlation coefficients of each group of specimens are listed in Table 6. For the specimens of the Z-C35-T and Z-C35-S groups, to conduct a thorough regression analysis, the ascending section was divided into two stages (elastic and damage stages), and the regression analysis was carried out. Hence, each parameter of the groups in Table 6 had two values. 0 slipped, K is the interface bond-slip curve stiffness, s is the relative slip of the interface, and u s is the slip when linear segment bond strength, u  , is reached. The specific values and correlation coefficients of each group of specimens are listed in Table 6. For the specimens of the Z-C35-T and Z-C35-S groups, to conduct a thorough regression analysis, the ascending section was divided into two stages (elastic and damage stages), and the regression analysis was carried out. Hence, each parameter of the groups in Table 6 had two values.  Table 6 shows that most of the correlation coefficients of the linear regression equations of each group of specimens were close to 1, indicating a good correlation with the experimental results. The ratio of the bond stress at the end of the linear growth (the linear segment bond strength) obtained by linear regression to the ultimate bond strength was generally above 90%. The critical bond stress of each group of specimens before slippage occurred at the interface did not exceed 11.02% of the linear segment bond strength. In addition, with the increase in the existing concrete's strength or shear angle, the bond strength and slip of the specimen increased, and the interfacial bond strength and slip of the specimen with shear keys further increased.     Table 5 for specific values). (4) The column of "Linear growth proportion" in the table indicates the ratio of the linear segment bond strength to the ultimate bond strength, namely, Linear growth proportion(%) = τu τmax × 100%, which is used to measure the degree of linear growth of the interfacial adhesion-slip rising segment. Table 6 shows that most of the correlation coefficients of the linear regression equations of each group of specimens were close to 1, indicating a good correlation with the experimental results. The ratio of the bond stress at the end of the linear growth (the linear segment bond strength) obtained by linear regression to the ultimate bond strength was generally above 90%. The critical bond stress of each group of specimens before slippage occurred at the interface did not exceed 11.02% of the linear segment bond strength. In addition, with the increase in the existing concrete's strength or shear angle, the bond strength and slip of the specimen increased, and the interfacial bond strength and slip of the specimen with shear keys further increased.

Interface Bond-Slip Model and Recommended Value for Stiffness
As revealed by the experiments, the interface slip was very small before the interface bond stress reached the peak value; after reaching the peak value, the load suddenly and sharply decreased as the slip increased, and the failure mode of the specimen was sudden brittle failure. From the point of view of safety, the bearing capacity at the descending stage could be ignored. Except for the two groups of specimens, Z-C35-T and Z-C35-S, the initial stage (most areas) of the rising phase of the interface bond-slip curves could be represented by a linear regression line, that is, the ascending segment was linear growth. For the late stage of the ascending segment (close to the ultimate bond strength), although it was a curve growth, the proportion of linear growth was mostly above 90%. Thus, the ascending segment could be simplified by approximately assuming that the bond stress increased linearly to the ultimate bond strength.
Given these reasons, by referring to References [21,22] and considering that the critical bond stress of all specimens before interface slippage did not exceed 11.02% of the linear segment bond strength, this study posits that the interfacial bond-slip model of CGM-reinforced existing concrete structures adopts the form shown in Figure 10, and the equation is where K is the average bonding stiffness of the interface, i.e., K = τ u /s u ; τ and τ u are the bond stress corresponding to any slip value, s, of the adhesive interface and linear segment bond strength, respectively; and s u is the slip value corresponding to linear segment bond strength, τ u . The values of K, τ u , and s u are listed in Table 6.

Conclusions
In this study, the interfacial bond properties of CGM and a concrete substrate were studied through direct and slant shear tests. The following main conclusions were derived: (1) For the direct shear specimen without shear key, the bond strength between CGM and the concrete interface was the highest when the adhesive interface form was triangular, followed by the trapezoidal and smooth interface (the lowest). Hence, the interface of assembly components should be set as a triangle, and the concrete surface should be roughened when conditions allow to effectively bond with CGM. Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this article are available from the corresponding author upon request.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.