Experimental Investigation on Shear Strength between Ultra-High-Performance Concrete and Normal Concrete Substrates

,


Introduction
China has the largest number of and most complex tunnels and underground works in the world, with the fastest speed of development in the future [1].Citing the 2021 Statistical Bulletin on the Development of the Transportation Industry, as of 2021, the number of highway tunnels in China reached 21,316, with a total length of 21,999.3km.Among these tunnels, the proportion of long, large tunnels is over 70%, and about 30% of them are in a subhealthy state [2].China's tunnel works have transitioned from the construction stage to a stage where construction and maintenance are both important [3].Tunnels are semienclosed underground structures with complex internal and external environments.It is difcult to eliminate their structural defects, and diseases such as cracking and water seepage are widespread.Under the adverse efects of water pressure, freezing and thawing, and fre, disasters such as chip of-falling, water gush, and collapse also occur from time to time [4,5].For urban rail transit, high-speed railways, highways, and other lines of communications, the traditional concrete jacketed arch reinforcement method is less applicable as it requires a long operation time and large structural thickness.In combination with the technical development of prefabricated treatment, mechanization, and new materials, the application of new treatment technologies that are fast and efcient with a light structure and no incursion has become a rigid demand and key R & D orientation for tunnel construction and maintenance [6,7].
Ultra-high-performance concrete (UHPC) is a new cement-based material with ultra-high strength, ultra-high toughness, and ultra-high durability that is widely used in the treatment and reinforcement of bridges and buildings [8] and has broad application prospects in the prefabricated reinforcement of tunnels [9].Te development of domestic prefabricated treatment technologies is lagging behind, and there is little research on prefabricated treatment technologies combined with new UHPC materials, so the related problems need to be solved.Statistics show that about 50% of reinforcement repair failures are caused by cracking of the bond interface [10].Te interfacial bond reliability of the damaged lining structure (NC) to grouted backfll (NC) to PCL segment (UHPC) is the key to determining the success of reinforcement.PCL is a prefabricated construction method in Japan.Te whole process is precast concrete linked.
Te interfacial bond strength consists of material bonding strength, mechanical bonding strength, and shear keyway shear strength, which depends on the surface roughness, adhesive, surface cleanliness, and water content.In practical engineering, interfacial cleanliness and wetness are the basic requirements, but the stability of interfacial adhesives is difcult to control and the efect varies [11,12].Te control of surface roughness is the most critical.Austin et al. [13][14][15][16] examined the efect of interface roughness on the bond behavior by using the experimental methods of slant shear, direct shear, pulling, and splitting for ordinary concrete materials.Tey came up with the interface treatment method of "sandblasting > interface grooving > wire brushing > profle chiseling >> original casting" and proposed the interface roughness calculation method.Tayeh et al. [17], Li and Rangaraju [18], and Feng et al. [19] used postcast UHPC repair as the object and obtained the same conclusions.Te investigation shows that for the damaged lining structure (NC) to the grouted backfll (NC) interface, a good efect of collaborative bearing can be achieved by NC sandblasting and chiseling, and the interface properties can be easily guaranteed.
For grouted backfll (NC) to the PCL segment (UHPC) interface, due to the high strength of precast UHPC, the efect of sandblasting and chiseling treatment is signifcantly reduced, and shear keyway or interface reinforcement is required to improve the interfacial bond strength [20].Diab et al. [21] and Jafarinejad et al. [22] used cylindrical slant shear tests to derive the mesh grooved treatment.Te average 28-day interface shear strengths of the precast and postcast ordinary concrete were 14.4 MPa and 19.5 MPa, respectively, which were 26% and 70% higher than the untreated results.Because of these shear strengths, postcast ordinary concrete mesh groove treatment can greatly improve its shear strengths.Tayeh et al. [17] used prismatic slant shear tests to derive the grid grooved treatment.Te 28day interface shear strength of precast NC-postcast UHPC was 13.63 MPa, 70% higher than the untreated result.Carbonell Munoz et al. [23], Long et al. [24], and Ganesh and Ramachandra Murthy [25] used the same test method with the same grooved form and derived the 28-day shear strengths of the UHPC-NC interface as 17.5 MPa, 20 MPa, and 13.3 MPa.Te above investigations show that grooving is benefcial to improving the interface shear capacity, but the improvement is limited and some of the experimental results cannot yet reach the requirement of 14 MPa (interface shear strength) specifed by the American Concrete Institute.Mansour and Fayed [26] concluded that the interface shear strengths of UHPC-NC increased by 9-16%, 22-29%, 32-42%, and 41-46% at 3%, 6%, 10%, and 13% of the grooved area, respectively.Increasing the density and scale of the groove and turning the grooves into keyways is another idea to improve the interface shear strength.Xiangguo and Zhang [27] analyzed the efects of shear key and shear angle on the interfacial bond strength and slip displacement of precast UHPC-post-bond NC with direct and slant shear tests.Te results show that the interface shear strength is increased by about 20% under single keyway treatment, and the larger the shear angle, the greater the interfacial bond strength and bond stifness.Yang et al. [28] and Zhang et al. [29] analyzed the interface shear behavior of precast NC-postcast UHPC with diferent keyway widths, depths, and gaps using direct shear tests.Te results show that the interface shear strength is about three times higher than that of the bond treatment after the multiple keyway treatment, and the regular trapezoidal-shaped keyway provides the greatest improvement in the interface shear capacity.In summary, the keyway treatment can signifcantly improve the interface shear strength and ultimate slip of UHPC-NC, and the degree of improvement is closely related to the number of keyways.Terefore, it can be used as an important orientation for the interface treatment of From this, the slant shear test method is used in this article to study the interface shear mechanical properties of precast UHPC-postcast NC under diferent numbers of interface keyways (0, 1, 2, and 3), while the infuence of interface agents (no interface agent, cement slurry, silica fume modifed cement slurry, and expansion agent modifed cement slurry) and postcast NC grades (C30, C35, C40, and C50) on them is considered.Te XTDIC system is an optical noncontact three-dimensional deformation measurement system for the measurement and analysis of object surface topology, displacement, and strain.Te full-feld strain monitoring system (XTDIC) and the universal testing machine are used to observe and analyze the interface failure form, slip process, and shear strength of UHPC-NC.Combining the test results with the design requirements for tunnel reinforcement treatment, the UHPC-NC interface bond shear strength calculation method and the slip model are refned, and the ftting results of relevant parameters are provided based on the experiment.

Experimental Plan
2.1.Design of the Experiment.Te shear behavior of an interface between precast UHPC and postcast NC is investigated using a slant shear test.Te experimental procedures refer to the standard test method for bond strength of epoxy-resin systems used with concrete by slant shear (ASTM C882/C882M-13a).Based on the standard for test methods of concrete physical and mechanical properties (GB/T 50081-2019) and many scholars' fndings, the specimen size is taken as 100 mm × 100 mm × 300 mm, and the bond angle is taken as α � 30 °.
Tis experiment is designed into four working conditions (0, 1, 2, and 4) with the number of interface keyways as the main variable factor of the interface.As for selection of the keyway form, the inward concave keyway form is suitable for this experiment as actual prefabricated treatment of the tunnel is considered, and the outward concave keyway form will obstruct the grouting path and not be conducive to construction.As for selection of the keyway shape, a trapezoidal keyway is used to improve the shear behavior by considering the fndings of Zhou Jian-ting and Yang Jun.Dimensions of the specimen and trapezoidal keyway are shown in Figure 1.In addition, two variable factors of the keyway, i.e. the interface agent and strength of postcast concrete, are considered for certain cases.Four levels of each factor are designed as four interface agent treatments (no interface agent, cement slurry, silica fume modifed cement slurry, and expansion agent modifed cement slurry, respectively), and there are four grades of postcast concrete (C30, C35, C40, and C50).For the design of the experiment, K2-NA-40 is the reference group; three specimens are cast for each variable factor, and a total of 30 specimens are fabricated.Te main parameters of the specimen are shown in Table 1.

Specimen Preparation. Te specimen production process is as follows:
(1) Experiment preparation: the place-holding wooden block of the same size as NC is made by laser cutting.(2) UHPC casting: the place-holding wooden block is fxed on one side of the mold as a built-in formwork.
Te mixed UHPC materials are then poured into the test mold and compacted using an insertion vibrator.
Te purpose of compaction is to remove air bubbles from the concrete, reduce pores, and make the concrete denser to ensure the quality of the structure.
And the fow value of the UHPC mixture used is about 1.8 mm.At the same time, a set of UHPC cubic test blocks is reserved.(3) UHPC curing: the UHPC specimens are covered with flms for curing at room temperature, and the formwork is removed after 24 hours to take out the place-holding wooden block.Wood chips and impurities should be removed from the UHPC interface.After 48-hour steam curing at 90 °C high temperature, the specimens are subject to curing at room temperature until the standard age of 7 °d.(4) NC casting: the UHPC is fxed on one side of the mold as a built-in formwork.Te mixed NC materials are then poured into the test mold and compacted using an insertion vibrator.At the same time, a set of NC cubic test blocks is reserved.(5) NC curing: after a 28-day curing period at room temperature, the combined specimens are taken out, and the surface shall be sanded, painted, and sprayed with scattered spots for treatment to complete the production of slant shear specimens.
Te specimen production process is shown in Figure 2.

Material Properties.
Te experiment is conducted using UHPC produced by a Zhejiang company focused on new types of building materials, whose main raw materials are cement, mineral powder, silica fume, quartz sand, steel fber, water reducer, and water, and the powder is premixed into bags.Te steel fber is about 10 mm long and 0.12 mm in diameter, with a weight ratio of about 7.1%.And the mix ratio of the UHPC matrix is shown in Table 2.

Advances in Materials Science and Engineering
Te raw materials of C30, C35, C40, and C50 concrete include cement, sand, stone, water, and water reducer, and the mix ratio is shown in Table 3.
Te reserved test blocks are tested for material properties in accordance with the fundamental characteristics and test methods of ultra-high performance concrete (T/CBMF 37-  Note.(1) Cement slurry is so prepared that the mass ratio of cement to water is 1 : 0.35, and the cement is the P II 52.5 cement.( 2) Silica fume-modifed cement slurry is so prepared that the mass ratio of cement to silica fume to water is 0.95 : 0.05 : 0.35, and the silica fume is the 970 silica fume.(3) Expansion agent-modifed cement slurry is so prepared that the mass ratio of cement to expansion agent to water is 0.9 : 0.1 : 0.35, and the expansion agent is the CSA type expansion agent made in Japan.3.During the experiment, the specimen is placed upright on the steel pad of the testing machine.Te specimen axis should be adjusted to align with the centers of the upper and lower platens.A force of 5 kN is preloaded for instrument and equipment commissioning.Te formal loading speed is controlled at 3 kN/s and loaded to failure.During the loading process, the load and displacement are recorded by the testing machine.Te XTDIC of XTOP 3D Technology (Shenzhen) Co., Ltd. is used to monitor the strain of the specimen.3.2.Failure Process.Te full-feld principal shear strain monitoring images (by XTDIC) of the specimens with typical failure modes under diferent ultimate load ratios are shown in Figure 5. Te analysis shows that, during the loading process, the red strain failure zone appears at the upper bond interface of the specimen frst.With the increase in load, the principal shear strain zone expanded and penetrated along the bond interface, and fnally, the specimen failed.Te analysis of the failure mode shows that the upper part of the bond interface failed frst during the loading process of the specimens.Te main reason is that the difference in material modulus between UHPC and NC causes the load bias, and the stress at the upper part of the bond interface is relatively large during the loading process, so the failure is early.During the loading process, the upper and lower tips are both the failure starting points, which are mainly caused by stress concentration.Te end-tangent treatment of the specimen interface has an optimization efect [27].

Testing Phenomenon and Failure Process
Te analysis of the failure process shows that for specimens without a keyway (Figure 5(a)), the strain failure zone appears early, starting at about 40% of the ultimate load.,However, the strain diagrams at 60% and 80% limit loads are almost identical to those at 40% limit loads.Before the overall slip of the specimen interface, the strain change is small with a feature of brittle failure.For specimens with keyways (Figures 5(b)-5(f )), as the keyway increases, the contact area of the interface increases, and the shear strain failure zone appears relatively late, starting at about 70% to 80% of the ultimate load.With the increase in load, the   Comparing the failure process of the specimens with and without the keyway, it can be seen that without the keyway, with the increase in load, the specimens failed directly through the interface; with the keyway, the principal strain value in the keyway zone is small, and the specimens are frst subject to slip failure along the bond interface without a keyway, and then the strain in the keyway zone expands and penetrates until the specimens failed.It can be seen that the presence of the keyway has the ability to limit the interface slip.

Range of Infuence.
Te strain is measured by a strainometer, and the ultimate load is measured and displayed simultaneously by a universal testing machine.When extracting the ultimate load, strain is also ultimate strain.For the UHPC and NC principal strains along the midperpendicular direction of the bond interface, the standard division of ε u � 0.33 is used to obtain the main range of infuence at diferent numbers of keyways, as shown in Figure 6.From Figure 6(a), it can be seen that the principal strain at the UHPC-NC bond interface of the specimen is the largest under the load and decreases along the normal sides of the interface; the strain at the UHPC side attenuates more rapidly with distance.From Figure 6(b), it can be seen that the range of the infuence zone increases under the action of shear keys.When the number of shear keys is 2, the range of the infuence zone tends to be stable.Eventually, the major range of infuence of UHPC will be about 16.2 mm and about 17.5 mm for the NC side.the shear force acting on the specimen to the area of the bond interface, τ � Pcos 2 α/A, where, P is the axial ultimate load and A is the cross-sectional area of the slant shear specimen.Te calculated interface shear bond strength τ, failure mode (M), coefcient of variation (COV), and average shear strength (τ) of diferent specimens are shown in Table 5.

General Status of Interface Shear Strength.
Overall, the maximum interface shear strength is the largest (13.9 MPa) when the interface is treated with three keyways and postcast C40 concrete.Te maximum is smaller than the acceptable range of bond shear strength 14 MPa-21 MPa for the 28 d slant shear test as specifed in the guide to materials selection for concrete repair (ACI 546.3R-14); it is also much smaller than the shear strengths of the UHPC-NC interface obtained from the tests conducted by other scholars for the working conditions of precast NC and postcast UHPC, using interface treatments such as grooving and sandblasting (e.g., Table 6).And "the strengths measured in Table 6 has the variation only in the interface treatment."Tis also shows that the interface shear capacity of the treatment forms of precast UHPC and postcast NC is more unfavorable, and the guarantee of interface properties is more critical to the success of the reinforcement as a whole.In addition to the keyway treatment measures, interface reinforcement can be introduced to ensure good interface mechanical properties.

Analysis of Interface Shear Strength and Infuencing
Factors.Te test results of interface shear strength with diferent numbers of keyway treatments are shown in Figure 7.It can be seen that the specimen has the weakest shear capacity when no interface keyway is treated, and the maximum shear strength is 2.1 MPa; with the increase in the number of keyways, the interface shear capacity increases linearly.When the number of interface keyways is 3, the maximum shear strength is 13.9 MPa.
Combining the testing failure phenomenon with the analysis of the failure process, it can be seen that the keyway limits the interface slip deformation, improves the interface roughness and friction coefcient, and thus achieves the shear capacity enhancement.Combined with the relevant fndings of Pedro on the roughness characteristic index, the average keyway height R z at the interface is taken as the roughness characteristic index, R z � A J cosα/A, where A J is the width of the specimen multiplied by the waist length of the keyway trapezoid and multiplied by the number of keyways.According to this formula, the roughness indexes of 0, 1, 2, and 3 shear keys are 0, 1.32 mm, 2.645 mm, and 3.97 mm, respectively.Te ftted relationship between interface shear strength and roughness is obtained as follows: Te shear strengths of the specimens with diferent postcast concrete strength grades are shown in Figure 8.Since the strength of the cube specimens cured under the same conditions as the C35 concrete is approximately equal to the strength of C30 and lower than that of C35, they can be neglected in the analysis or considered as postcast C30 concrete.Te analysis shows that there is no trend of increase or decrease in the interface shear capacity with the increase in the grade of the postcast concrete.Tis is different from Júlio's fndings on the interfacial shear behavior for precast NC and postcast UHPC and is also related to the dense properties of UHPC.It also shows that for precast UHPC reinforcement, increasing the interface roughness and providing shear keyways or reinforcement during 8 Advances in Materials Science and Engineering precast production are more efective treatments than improving the performance of postcast materials.Te shear strength of the specimens with diferent interface agent treatments is shown in Figure 9.When treating specimens, the shear strength with no interface agent is slightly higher than that of cement slurry, silica fume modifed cement slurry, or expansion agent modifed cement slurry, but the overall diference is not signifcant.Tis shows that for precast UHPC and postcast NC treatment conditions, the interface agents cannot penetrate each other to improve the shear capacity of the interface.

Analysis of Interface Shear Stifness and Infuencing Factors
4.2.1.Load-Slip Curve.During the experiment, the vertical load and displacement are obtained by the universal testing machine.Te interface shear stress and slip are obtained after decomposition (a complete specimen is divided into two parts along the bonding interface) along the bond interface.Slip is measured by a universal pressure testing machine's vertical displacement and then calculated by the Pythagorean theorem along the bond interface.And the shear stress-slip curve relationship is drawn as shown in Figure 10.During the experiment, the data of individual specimens are abnormal and are not indicated in the fgure.

Interface Slip Model.
Combined with the shear stressslip curve of the slant shear specimen and the testing phenomenon and failure process, the specimen slip can be divided into four stages: (1) Interface slip stage: the shear capacity at this stage is dominated by the interfacial bonding and mechanical bonding, and the stifness is relatively small; (2) Keyway strengthening stage: the UHPC-NC interface shear capacity reaches the limit, the keyway concrete plays a role, and the shear stifness increases and tends to stabilize; (3) Shear yielding stage: the UHPC-NC interface slips, keyway concrete is subject to shear yielding, shear stifness gradually decreases, and shear strength grows in a nonlinear mode until it reaches its load-carrying limit; (4) Specimen failure stage: the penetrated damage to the keyway concrete causes a rapid decrease in the load-carrying capacity of the specimen and loss of load-carrying capacity.In summary, the shear stages are simplifed to a linear process, and the interfacial bond-slip model of precast UHPC and postcast NC is obtained as shown in Figure 11.Te expression is as follows: 10 Advances in Materials Science and Engineering where τ is the interfacial bond shear strength (MPa); τ 0 is the bond strength before the interface slip; S h is the maximum slip at the interface slip stage; τ h is the maximum shear stress at the interface slip stage; K h is the average shear stifness at the interface slip stage; τ u is the ultimate bond strength of the specimen; S q is the slip when τ u is reached at the keyway strengthening stage; K j is the average shear stifness at the keyway strengthening stage; and S u is the slip when τ u is reached in the experiment.Te ftting results of the interfacial bond strength, stifness, and slip with diferent numbers of keyway treatments are shown in Table 7. Te correlation systems of the regression equations are all greater than 0.90, and the regression values have great goodness of ft with the experimental values.Terefore, the model can accurately refect the interfacial bond-slip mechanical properties of precast UHPC and postcast NC.Te second-stage stifness results obtained from the experiment are similar to the fndings of Advances in Materials Science and Engineering Yang et al. [28] on the interfacial bond behavior of UHPC-NC and lower than those of Xiang-guo and Zhang [27].Te main reason concluded from the preliminary analysis is that a larger interfacial shear key was used by Xiang-guo et al.

Interface Slip Infuencing Factors. Te ftting results
of the shear stifness at the interface slip stage and the keyway strengthening stage with diferent numbers of keyways are shown in Figure 12.Te analysis shows that at the interface slip stage, there is no obvious tendency pattern between the interface shear stifness and the number of keyways.Te main reason is that this stage is the initial stage of loading, and the initial gap and displacement have a large infuence on the results.In addition, the critical slip value itself is small at this stage, which is more infuenced by other inherent errors.At the keyway strengthening stage, the shear stifness frst increases linearly with the number of keyways and then tends to stabilize with the maximum shear stifness of about 20 MPa/mm.Te main reason is that when the interface roughness reaches a certain level, the interface slip stifness is mainly infuenced by the NC modulus and then tends to stabilize.
Te ftting results of the critical slip at the interface slip stage and the keyway strengthening stage with different numbers of keyways are shown in Figure 13.At the interface slip stage, when there is no shear keyway, the slip is large and the specimen is subject to slip failure directly; when the shear keyway exists, the specimen slip is reduced by the shear keyway constraint, and the critical slip grows with the increase in the number of keyways.At the keyway strengthening stage, the maximum slip increases linearly with the increase in the number of keyways, and the continuous improvement of the interfacial ultimate shear capacity is achieved after the stifness stabilization.Te main reason for the formation of this pattern is similar to that for the stabilization of stifness: because the number of keyways grows, more interface UHPC and NC are involved in the shear process, and the relative slip strain at the interface decreases, realizing the improvement of ultimate strain and shear capacity on the side.

Conclusion
Tis article focuses on the key scientifc issues on interfacial bond shear behavior between precast UHPC and postcast NC used in prefabricated treatment of the tunnel, carries out experimental investigation on slant shear strength under working conditions of diferent numbers of keyways, interface agent treatments, and postcast NC grades, analyzes the failure process and failure mode of UHPC-NC specimens under compression shear, shear strength of bond interface and infuencing factors, shear stifness, and slip model, and draws the following conclusions: (2) Te full-feld principal strain analysis (by XTDIC) shows that the strain in the keyway zone is much smaller than that in the nonkeyway zone, and the presence of the keyway has the ability to disperse the Advances in Materials Science and Engineering load and limit the interface slip.Te principal strain decreases along the normal sides of the interface, and the range of infuence of UHPC is about 16.2 mm and about 17.5 mm for the NC side.In practice, the thinnest position of the treatment structure should not be lower than 2 cm.(3) Comparing the fndings of this experiment with those of other scholars and specifcations, the overall interfacial bond shear strength of precast UHPCpostcast NC is low, and the maximum interface shear strength of 13.9 MPa obtained from the experiment is still smaller than the recommended value of 14-21 MPa specifed in ACI 546.3R-14.For the design of prefabricated treatment of the tunnel, interface shear reinforcement can be introduced to ensure collaborative loading.
(4) Te interfacial bond shear behavior of precast UHPC-postcast NC is little afected by the interface agent and the grade of postcast concrete and increases linearly with the increase in the number of keyways.Equating the keyway to the interface roughness index R z , the relationship between interface shear strength and roughness is f s � 1.664 + 3.030R z .In engineering applications, the interface shear capacity can be improved by increasing the equivalent roughness.
(5) Te interface slip of UHPC-NC under the action of the keyway can be simplifed into four stages: the interface slip stage, the keyway strengthening stage, the shear yielding stage, and the specimen failure stage.Accordingly, the interface slip model of precast UHPC-postcast NC is proposed, and the fndings of key parameters such as interfacial bond strength, stifness, and slip at diferent stages are ftted.
(6) At the interface slip stage, there is no tendency pattern between the fndings of shear stifness and the number of keyways due to the initial gap.At the keyway strengthening stage, the shear stifness frst increases linearly with the number of keyways and then tends to stabilize with the maximum shear stifness of about 20 MPa/mm.Te maximum interface slip grows with the number of keyways, resulting in an increase in overall shear capacity.
In this experimental investigation, the fndings of interfacial bond mechanical properties of partial precast UHPC-postcast NC are obtained in combination with the slant shear test method, which can be used in the design of the prefabricated tunnel segment.However, some problems were also found in the experiment.For example, the interfacial bond behavior is generally low, so interface shear reinforcement should be provided.Te dispersion of stifness fndings at the interface slip stage is large, so further experiments should be conducted.Te compression-shear state of UHPC-NC for tunnel reinforcement treatment is complex, so it cannot be fully refected by this experiment alone.Next, further in-depth investigations will be conducted on these specifc problems to realize the implementation of prefabricated tunnel reinforcement treatment technology in China at an early date.

Table 1 :
Parameters of specimens.

3. 1 .
Failure Mode.During the experiment, there are four main failure modes for the specimen: (1) Complete interface failure (Class A): the specimen is subject to slip failure along the bond interface, and both material interfaces are smooth after slip.No damage and spalling of UHPC and NC were observed on both sides of the interface, and no other forms of cracking of UHPC and NC were observed outside the interface, as shown in Figure 4(a).(2) Interface failure + NC shear failure (Class B): UHPC-NC combined specimens are subject to Class A failure at the nonkeyway treated position.Tere are dense cracks on the NC side at the nonkeyway treated position, with failure eventually penetrating along the direction of the bond interface.Except for a small amount of breakage outside the interface and at the tip of UHPC and NC, the failure is caused by stress concentration based on a preliminary analysis, as shown in Figures 4(b)-4(d).

( 3 )
NC compression failure (Class C): during the loading process, the nonkeyway treated position of the interface is subject to Class A failure frst.As the load increases, the cracks expand vertically along the NC side, and the keyway concrete cracks simultaneously.After the vertical cracks penetrated the interface cracks, the specimen failed.After the failure, a large amount of NC will adhere to the UHPC keyway position.(4) Interface failure + NC compression failure + NC slip failure in the keyway (Class D): a special failure mode emerged during the experiment of K3-NA-40.During the loading process, the condition of the specimen is frst consistent with the Class C failure.When the upper NC crack penetrated, the NC and UHPC in the upper side shear keyway were detached by tension.As the load increases, it is subject to interface slip failure, as shown in Figure 4(f ).

Figure 6 :
Figure 6: Te range of the strain damage zone.(a) Principal shear strain distribution.(b) Te range of the infuence zone.

( 1 )
Tere are four typical failure modes of UHPC-NC under compression shear, namely complete interface failure (Class A), interface failure + NC shear failure (Class B), NC compression failure (Class C), and interface failure + NC compression failure + NC slip failure in the keyway (Class D).Class A failure is sudden and should be avoided in the works.

Table 4 .
Te 28 d cubic (the specimen size is taken as 100 mm × 100 mm × 100 mm) compressive strength of UHPC obtained from the test is 139.0MPa.Te 28 d cubic (the specimen size is taken as 150 mm × 150 mm × 150 mm) compressive strengths of C30, C35, C40, and C50 concrete specimens are 30.6MPa, 28.6 MPa, 40.9 MPa, and 52.1 MPa, respectively.Due to the insufcient strength of the C35 concrete test results, the K2-NA-35 working conditions in the subsequent analysis are not representative, and their relevant results are for reference only.
2.4.Loading and Monitoring Schemes.Te electro-hydraulic servo pressure testing machine of Shanghai Hualong Test Instruments Corporation is used for loading.Loading and monitoring devices are shown in Figure

Table 4 :
Cube compressive strength test results of UHPC and NC (unit: MPa).