1 Introduction

Rock–shotcrete structure widely exists in practice, e.g., the underground openings after shotcreting and the rock-socked piles. The mechanical behavior of rock–shotcrete interface significantly affects the stability of those engineering structures. The rock–shotcrete interface is often suffered from dynamic loads, e.g., vibration, blasting and earthquake, propagating in the form of stress waves (**e et al. 2020). When S wave appears, the rock–shotcrete interface is subjected to dynamic shearing. As the rock–shotcrete interface is the fragile part of tunnel, its dynamic shear behaviors play an important role in structural stability (Hashash et al. 2001; Chen et al. 2014). In spite that many efforts have been devoted to the static shear behavior of rock–shotcrete interface, its dynamic shear response is still at its infancy and the corresponding failure mechanism remains unclear. Therefore, it is essential for the design and safety assessments of rock–shotcrete structures to investigate the dynamic shear behaviors of rock–shotcrete interfaces.

Static shear properties of the rock–shotcrete structure have been widely investigated (Saiang et al. 2005; Haque and Kodikara 2012; Andjelkovic et al. 2015; Jiang et al. 2021). For example, Moradian et al. (2012) studied the effect of roughness, normal load and displacement rate on shear behaviors of the rock–concrete specimen using acoustic emission technology, and found that these factors affect the adhesive bonding between rock and concrete. Tong et al. (2016) concluded that the bond characteristics of the rock–concrete specimen are dependent on the curing temperature and humidity, and the specimen’s residual shear strength depends on the applied normal stress. Tian et al. (2014) conducted experimental and numerical studies to analyze the shear behavior of rock–concrete specimens and found that the failure changes from sudden bond failure to gradual bond failure with the increase in normal stress. Zhao et al. (2018) carried out direct shear tests to investigate the shear behavior of rock–concrete specimens with different interfacial shear strengths, and observed three failure types under different normal loads. Previous findings show that interface roughness plays an important role in the shear properties of the rock–shotcrete structure. For example, Kodikara and Johnston (1994) carried out direct shear tests on rock–concrete specimens with regular or irregular triangle asperities under constant normal load and constant normal stiffness loading conditions, respectively, and observed that the specimen with regular triangle asperities shows different shear behaviors from that with irregular triangle asperities. Shen et al. (2019) found that the shear strength of rock–concrete specimens depends on the interface roughness, and proposed a two-element interfacial shear model to predict the shear strength. Mouzannar et al. (2017) concluded that there are two failure mechanisms, which were associated with interface roughness: one is controlled by shear bonding, and the other is by tensile bonding. Although many efforts have been made on static shear behaviors of the rock–shotcrete structure, the understanding of its dynamic shear behaviors is still at its infancy due to such factors as the limitation of relevant experimental equipment and methods. Rock–shotcrete structures experienced severe disasters such as rockburst and cavity collapse more frequently because of dynamic shearing (Sun et al. 2011; Fang et al. 2016). This necessitates studies on the dynamic shear properties of rock–shotcrete structures. Meanwhile, although interface roughness has obvious influences on mechanical properties of rock–shotcrete interface (Luo et al. 2017; Zhu et al. 2020), its influences under dynamic shearing are rarely studied.

By adopting a proper specimen shape, the dynamic stress equilibrium in the specimen could be achieved during laboratory tests using the split Hopkinson pressure bar (SHPB), and the dynamic shear test would be carried out successfully (Ma et al. 2018; Yuan et al. 2015; Peirs et al. 2012). In general, there are three types of specimens for dynamic shear tests, i.e., the “hatted” specimen, the double notch specimen and compact forced-simple shear specimens (S-shape specimen). Ran et al. (2017) and Zhou et al. (2017) adopted the “hatted” specimen to study the dynamic shear behavior of titanium alloy successfully. A double notch specimen was used by Guo and Li (2012) to investigate the dynamic shear response of titanium alloy, and the failure process was recorded by a high-speed camera. Arab et al. (2019) found that the stress equilibrium is much easier to be fulfilled by using an S-shape specimen. However, it is difficult for them to prepare specimen and directly observe the shear failure process. Therefore, it is of great importance to develop a proper method to obtain the dynamic shear properties and failure process of rock–shotcrete specimens.

In this study, a modified double notched rock–concrete specimen was proposed and adopted to conduct dynamic shear tests using the SHPB. Effects of strain rate and interface roughness on the dynamic shear behaviors of rock–concrete specimens were studied. Laser scanning was performed to obtain the failure surface morphology. The failure mechanism was analyzed. The findings will contribute to better understanding the mechanical behavior of rock–shotcrete structures, and facilitate the reinforcement of underground openings.

2 Experimental set-up

2.1 Specimen preparation

The sandstone from Neijiang, Sichuan, China was selected as the rock part of rock–concrete specimens, because of the relatively high adhesive force between the sandstone and concrete, which can avoid the debonding of rock–concrete interfaces in the process of specimen preparation (Zhu et al. 2020). According to the suggested methods from ISRM (1978) and Bieniawski and Bernede (1979), mechanical properties of the sandstone are obtained by the uniaxial compression test, Brazilian disc test, and direct shear test with a testing machine with a loading capacity of 1000 kN in the normal direction and 500 kN in the shear direction. The physical and mechanical properties of the sandstone are obtained and listed in Table 1.

Table 1 Physical and mechanical properties of sandstone, concrete
Full size table

C20 concrete was used as the concrete part of rock–concrete specimens in this study. The concrete consists of water, P.O. 42.5 cement, medium sand with a grain size range of 0.08–2 mm, and coarse limestone gravel ranging from 1 to 3 mm, with a weight ratio of 0.1, 0.13, 0.32 and 0.45.

Rock surfaces in nature, e.g., the tunnel surface, are often rugged with bulge parts and depression parts. For its easy preparation and effectiveness, the saw-tooth surface is often adopted in the laboratory to mimic the natural rock rough surfaces (Zhao et al. 2019; Cheng et al. 2021). In this study, rock–concrete interface with five sawtooth angles, i.e., 0°, 15°, 30°, 45° and 60°, were adopted, as shown in Fig. 1. The sawtooth length d is constant. By varying the sawtooth height h, different sawtooth angles can be obtained. In addition, the irregular sawtooth interface with combined sawtooth angles of 30°, 45° and 60° was also prepared to reflect the irregularity of natural rock surfaces, as shown in Fig. 1b. The sandstone sawtooth interface was fabricated with an engraving and milling machine to ensure the sawtooth accuracy.

Fig. 1
figure 1

Sandstone–concrete specimens with different interface roughnesses: a Regular sawtooth interface. From left to right, the sawtooth angles of the sandstone–concrete specimen are 0°, 15°, 30°, 45°, 60°, respectively; b irregular sawtooth interface. The mixed angle consists of 30°, 45° and 60° from left to right. l represents the length of the sandstone–concrete interface, d and h denote the length and height of the sawtooth respectively

Full size image

Before casting rock–concrete specimens, rock surface and coarse gravel were cleaned and dried to avoid the influence of contaminants. As shown in Fig. 2, the casting process follows four steps:

  1. (1)

    Apply the release agent to the designed shear specimen mould;

  2. (2)

    Put the sandstone part into the middle of the mould, and tighten the bolts on the side and bottom of the mould;

  3. (3)

    Pour concrete mixture into the mould, and vibrate the mould following the suggested method;

  4. (4)

    Demould after 24 h, and cure rock–concrete specimens in a standard curing box for another 28 days.

Fig. 2
figure 2

Specimen preparation procedure: a apply the release agent into the designed steel mould; b put the rock part into the mould; c pour the concrete into the mould; d the prepared rock–concrete specimen

Full size image

All specimens were polished following the ISRM suggested method (Muralha et al. 2014). The flatness and perpendicularity were less than 0.25 mm and 0.25°, respectively. 48 specimens were prepared, including sandstone, concrete and sandstone–concrete specimens.

2.2 Testing apparatuses

Figure 3 shows the testing apparatuses for the dynamic shear test on sandstone–concrete specimens. The SHPB with modified loading shear moulds was adopted. Both bars and the modified loading shear moulds are made of 60 Si2Mn steel. The elastic modulus and P-wave velocity are 211 GPa and 5102 m/s, respectively. The bars consist of striker bar, incident bar and transmitted bar with an identical diameter of 50 mm and lengths of 300 mm, 3000 mm and 2000 mm, respectively. A copper disc, which is 9 mm in diameter and 2 mm thick, was pasted on the head of the incident bar to eliminate the dispersion effect and achieve dynamic stress equilibrium in the tested specimen (Song and Chen 2004; Du et al. 2020). Two pairs of strain gages were pasted on the incident bar and transmitted bar, respectively, to collect wave signals. An 8-channel DL750 ScopeCorder digital oscilloscope with a 6-channel dynamic strain amplifier was adopted to record wave signals.

Fig. 3
figure 3

Diagram of the dynamic shear test system: a SHPB apparatus combined with the modified loading shear mould. The red rectangle solid line shows the whole SHPB apparatus, and the red rectangle dash line represents the modified loading shear mould with the double notch rock–concrete specimen; b dimensions of the modified loading shear mould. Dinner and Dexternal represent the inner and external diameter of the modified loading shear mould respectively

Full size image

The dimensions of the modified loading shear mould are shown in Fig. 3b. The mould’s inner diameter is only 0.05 mm larger than the bar diameter to fulfill the requirement of the tight junction between them. During the test, the two modified shear moulds were connected to the incident bar and the transmitted bar, respectively, and the specimen was installed between the two moulds, as shown in Fig. 3b. A 5 mm gap was reserved between the specimen bottom and the modified loading shear mould. A Photron SA-Z high speed digital camera combined with two lights was applied to record the shear failure process during the test. The frame rate and pixel were set as 100,000 fps and 640 × 280, respectively. In addition, a laser scanner Capture Mini with scanning precision of 0.03 mm was used to obtain fracture morphology after the test, and the obtained data were analyzed with the 3D scanning software Wrap (Edelsbrunner et al. 1998).

The dynamic shear strength of the rock–concrete specimen τd can be calculated by:

$$\tau_{{\text{d}}} = \frac{{\text{P}}}{2lw}$$
(1)

where P is the force applied on the specimen, l and w are the length and width of the sandstone–concrete interface, respectively. The dynamic force can be determined by the wave signals on the incident and transmitted bars (e.g., Chen et al. 2002; Han et al. 2022).

3 Results and analysis

3.1 Dynamic stress equilibrium

Figure 4 shows the dynamic stress equilibrium in a typical dynamic shear test on the sandstone–concrete specimen. It is seen that the sum of incident stress and reflected stress is approximately equal to the transmitted stress before the peak of transmitted stress, meaning that the sandstone–concrete specimen is in a state of dynamic stress equilibrium before it failed.

Fig. 4
figure 4

Diagram of the dynamic stress equilibrium of a typical dynamic shear test on the double notch sandstone–concrete specimen. i, r and t denote the incident, reflected and transmitted wave respectively

Full size image

3.2 Comparison among dynamic shear behaviors of sandstone–concrete, sandstone and concrete specimens

Figure 5 shows the dynamic shear behaviors of the sandstone–concrete specimen with a flat interface, sandstone and concrete specimen under a strain rate of about 180 s−1. For sandstone–concrete specimens, the shear failure occurred at the interface between sandstone and concrete, indicating that the sandstone–concrete interface is the fragile part. When subjected to dynamic shearing, all specimens behaved similarly, exhibiting a similar stress–strain curve. The strain herein denotes the axial strain of the tested specimen. However, the sandstone–concrete specimen has different dynamic shear strength and deformation characteristics from sandstone and concrete specimens. A tiny compaction stage is observed in the beginning, followed by an elastic stage until the peak stress is achieved. As the dynamic stress equilibrium in the specimen is difficult to fulfill after peak load, characteristics of the post-peak stage are not analyzed herein.

Fig. 5
figure 5

Dynamic shear properties of the sandstone–concrete specimen with a flat interface, sandstone and concrete: a failure pattern; b shear stress–strain curve; c dynamic shear strength; d dynamic peak strain. The interface of the sandstone–concrete specimen is flat

Full size image

As shown in Fig. 5c, the dynamic shear strength of sandstone–concrete specimen (7.51 MPa) is far lower than that of concrete (10.77 MPa) and sandstone (26.88 MPa). The dynamic peak strain of sandstone–concrete specimen is 0.004, which is 17% and 8% of those of concrete (0.009) and sandstone (0.018), respectively. This indicates that the sandstone–concrete interface is more fragile than sandstone and concrete. In the post-peak stage, the friction of induced fracture surfaces contributes to the shear resistance. In addition, as the shear capacity of a flat interface depends mainly on the interfacial cohesion (Luo et al. 2017; Zhu et al. 2020), it can be deduced that the cohesion of the sandstone–concrete interface is smaller than that of sandstone and concrete, resulting in the lowest dynamic shear strength and peak strain of sandstone–concrete specimen compared with those of sandstone and concrete specimens.

3.3 Effect of strain rate

The rock–concrete specimen has an obvious strain rate effect when suffering from the compressive or tensile load, highlighting that the mechanical properties such as dynamic compressive strength and tensile strength increase with the strain rate (Luo et al. 2017; Gong et al. 2018). Figure 6 shows the effect of strain rate on the dynamic shear strength for the sandstone–concrete specimen with a sawtooth angle of 30° when subjected to dynamic shearing. The dynamic shear strength of rock–concrete specimens is highly strain rate dependent. The dynamic shear strength shows an increasing trend with increasing strain rate, and the strength increment is accelerated at a higher strain rate. The dynamic shear strength experienced a 92.6% increase from 6.11 to 11.77 MPa when the strain rate increased from 40 to 250 s−1. By observing the failure patterns shown in Fig. 6b, this accelerated increment can be attributed to the changed failure mode with increasing strain rate, which will be discussed in detail later.

Fig. 6
figure 6

Strain rate effect of sandstone–concrete specimens with a sawtooth angle of 30° under dynamic shear loading: a Dynamic shear strength; b failure pattern. The red rectangle solid line indicates sandstone shear failure, and the red oval solid line represents concrete shear failure. The loading end of the specimen is at the right of each high-speed image

Full size image

Figure 6b shows the shear failure pattern of the sandstone–concrete specimen under different strain rates. The shear induced fracture initiates from the interface, while the propagation path differs under different strain rates. When the strain rate is below 150 s−1, the induced fracture propagates along with the sandstone–concrete interface due to the weak interfacial strength, and the failure mode always manifests as an interfacial shear failure regardless of stain rate. In this range of strain rate, the increase in dynamic shear strength is mainly due to the strain rate effect on interfacial cohesion (Qian et al. 2009; Si et al. 2019). Under high strain rates, the strain energy released near the induced fracture tip is sufficiently high to cause the failure of sandstone and concrete, and consequently, sandstone and concrete shear-off fractures appear sequentially when the strain rate increases to about 180 s−1 and 230 s−1. As the sandstone and concrete failure consumes more energy than the interfacial failure, the dynamic shear strength of sandstone–concrete specimens is therefore increased with increasing strain rate. These findings indicate that attentions should be paid to the mechanical behaviors of sandstone section and concrete section of sandstone–concrete structure in addition to their interface when subjected to dynamic shearing of high strain rate.

3.4 Effect of roughness

3.4.1 Shear stress state of sandstone–concrete specimen

To reveal the mechanism behind failure mode change of sandstone–concrete specimens, force analysis was carried out. Figure 7 shows the force diagram of the sliding fracture (interfacial shear failure) and shear-off fracture of the sandstone–concrete specimen (concrete shear failure or sandstone shear failure). When shear failure appears at the sandstone–concrete interface, as shown in Fig. 7a, the equilibrium equation along the interface can be obtained as follows:

$${\text{S}}_{F} - R_{F} - R_{C} = 0$$
(2)

where SF represents the component of the external shear force along with the sandstone–concrete interface. RF and RC are the friction force and cohesive force of the sandstone–concrete interface, respectively.

Fig. 7
figure 7

Force analysis of the sandstone–concrete specimen with different failure modes: a Sliding fracture; b shear-off fracture. L and d here denote length and height of the sawtooth respectively, l means the length of broken sawtooth, α shows the sawtooth angle, F is the external shear load. SF represents the component of the external shear force along with the sandstone–concrete interface. RF and RC are the friction force and cohesive force of the sandstone–concrete interface, respectively. FI, FC and FS are the required external shear forces for interfacial shear failure, concrete shear failure and sandstone shear failure respectively. The red area is the concrete shear failure

Full size image

SF, RF and RC can be calculated as:

$${\text{S}}_{F} = F \times \cos \alpha$$
(3)
$$R_{C} = \frac{d}{\sin \alpha } \times {\text{c}}$$
(4)
$$R_{F} = N_{F} \times \tan \varphi = F \times \sin \alpha \times \tan \varphi$$
(5)

where F denotes the external shear load. NF is the normal tension loaded on the sandstone–concrete interface. α is the sawtooth angle. φ and c are the internal friction angle and cohesive force of the sandstone–concrete interface, respectively. d represents the sawtooth height.

In addition, the relationship between d and the sawtooth length L can be described as:

$$d = \frac{L}{2} \times \tan \alpha$$
(6)

Substitute the Eqs. (3)–(6) into Eq. (2), F can be derived as:

$$F = \frac{L \times c}{{2\cos \alpha (cos\alpha - \sin \alpha \tan \varphi )}}$$
(7)

Equation (7) implies that F increases with α. In conjunction with Eq. (1), the shear strength of the sandstone–concrete specimen increases with the increase of the sawtooth angle α.

From Eq. (7), it can be deduced that F is only related to L, internal friction angle φ and cohesive force of the interface c. If L keeps invariable. c and φ would dominant the interfacial shear failure.

When the shear-off fracture appears, i.e., shear failure at concrete section or sandstone section of the sandstone–concrete specimen, as shown in Fig. 7b, the required external shear force for the occurrence of sliding fracture, FI, can be calculated as:

$$F_{I} = F \times \frac{l}{L} = \frac{l \times c}{{2\cos \alpha (\cos \alpha - \sin \alpha \tan \varphi )}}$$
(8)

where l represents the length of the broken sawtooth.

The required external shear forces for concrete shear failure (FC) and sandstone shear failure (FS) can be expressed as:

$$F_{C} = l \times c_{C}$$
(9)
$$F_{S} = l \times c_{S}$$
(10)

where cC and cS denote the cohesion of concrete and sandstone, respectively.

From Eqs. (8)–(10), it is seen that FI increases with α, while FC and FS is independent on α. When α increases to a critical angle, FI is larger than FC or FS, leading to the shear-off fracture of the sandstone–concrete specimen. Therefore, with increasing α, the failure mode changes from interfacial shear failure into mixed failure (i.e., interfacial shear failure combined with concrete shear failure or sandstone shear failure), leading to an increase in dynamic shear strength.

3.4.2 Regular artificial interface

Figure 8 shows the shear stress–strain curves of sandstone–concrete specimens with different sawtooth angles. All the sandstone–concrete specimens present similar shear stress–strain curves, which can be divided into four stages, as shown in Fig. 8b:

Fig. 8
figure 8

Dynamic shear stress–strain curves of sandstone–concrete specimens with different sawtooth angles at a strain rate of about 50 s−1. a The sawtooth angle ranges from 0° to 60°; b typical stage division of the sandstone–concrete specimen with sawtooth angle

Full size image

(I) Compaction stage. In this stage, the pre-existing cracks and voids in the interface between sandstone and concrete are compacted. The curve manifests as the upper concave, indicating the shear stress increases slowly with strain. For specimens with different sawtooth angles, there exist some differences in terms of the stress-strain curves. With the increase in sawtooth angle, the duration of this stage increases, probably attributed to the enhanced mechanical interaction effect between sandstone and concrete at relatively great sawtooth angle. When the sandstone–concrete interface is flat, the mechanical interaction between sandstone and concrete is slight, resulting in a tiny short compaction stage, as shown in Fig. 5b. The higher the sawtooth angle is, the more significant the mechanical interaction is (Zhu et al. 2020). With increasing sawtooth angle, the contact area between sandstone and concrete also increases, which is likely to cause increased voids and cracks adjacent to the interface. The increased voids and cracks require relatively higher shear force and more time to compact. It is the combined effect of the mechanical interaction and the contact area that leads to different compaction stages of sandstone–concrete specimens with different sawtooth angles.

(II) Elastic stage. At this stage, the shear stress of the sandstone–concrete specimen increases almost linearly with strain. Additionally, the curve’s slope, i.e., shear modulus, increases with the increase in sawtooth angle due to the enhanced mechanical interaction between sandstone and concrete.

(III) Adhesive failure stage. In this stage, the shear stress-strain curves of sandstone–concrete specimen present as upper convex shape, showing a slower increase in stress with increasing strain. With increasing sawtooth angle, this stage becomes longer and more obvious. This is related to the length of cracking. On the one hand, the increased contact area between sandstone and concrete causes a longer cracking. On the other hand, the cracking path varies with the failure mode that changes from interfacial failure to concrete or sandstone sawtooth failure. In addition, the stage duration of the sandstone–concrete specimen is shorter than that of concrete or sandstone specimens, as shown in Fig. 5b.

(IV) Softening stage. The sandstone–concrete specimen appears macroscopic cracking and finally breaks in this stage after the peak. As the dynamic stress equilibrium in the specimen was not well achieved at this stage, no more analysis was carried out in the present study.

Figure 9 shows the shear fracture evolution of the sandstone–concrete specimen with a sawtooth angle of 30°. Shear fractures initiated at the sandstone–concrete interface closer to the transmitted bar, when the shear stress was about 90% of the peak shear strength. The shear fractures propagated along the opposite direction of the wave propagation and coalesced into a main shear fracture. All sandstone–concrete specimens show a similar failure process.

Fig. 9
figure 9

Failure process of a sandstone–concrete specimen with the sawtooth angle of 30°: a Dynamic shear stress as a function of time, where three dotted lines denote 280 us, 360 us and 460 us, respectively, from left to right; b the evolution of shear fractures (0 us, 280 us, 360 us and 460 us) where the incident bar is located on the right of the figure. The red rectangle solid line shows the shear failure area

Full size image

Figure 10 presents the dynamic shear properties of sandstone–concrete specimens with different sawtooth angles. The dynamic shear strength of the sandstone–concrete specimen shows a roughly linear growth with increasing sawtooth angle, experiencing a 257% increase from 3.13 to 11.18 MPa when the sawtooth angle increases from 0° to 60°, whereas the peak shear strain increases from 0.001 to 0.009 correspondingly. The peak strain increment shows a rapid growth at small sawtooth angles yet slows down at larger sawtooth angles.

Fig. 10
figure 10

Dynamic shear strength and peak strain of sandstone–concrete specimen with different sawtooth angles. The strain rate is about 50 s−1

Full size image

Figure 11 shows the failure morphologies of sandstone–concrete specimens with different sawtooth angles. When the sawtooth angle increases from 0° to 30°, shear failure appears at the sandstone–concrete interface. Some remained adhered concrete can be found at these interfaces. The volume of concrete adhered on the fracture surface of the sandstone section to some extent can reflect the magnitude of interfacial cohesion. The fracture morphologies of sandstone–concrete specimens were acquired by conducting laser scanning. By comparing the volumes of sandstone before and after dynamic shearing using the software Wrap, as shown in Fig. 12a and b, the adhered concrete volume can also be acquired. When the sawtooth angle increases from 0° to 15° and 30°, the adhered concrete volume increases significantly from 5 to 635 mm3 and 1745 mm3, respectively, meaning that the interfacial cohesion is enhanced. As a result, the shear resistance of sandstone–concrete specimens improves. It is worthy to note that except for the effect of increased contact area between sandstone and concrete, the mechanical interaction between sandstone and concrete also plays an important role in dynamic shear behavior when the sawtooth angle increases from 0° to 15°. The coupling effect of contact area and mechanical interaction leads to a significant increase in the dynamic peak shear strain of sandstone–concrete specimens.

Fig. 11
figure 11

Fracture surface of sandstone–concrete specimens with different sawtooth angles under dynamic shear loading, the sawtooth angle ranges from 0° to 60°. Red curves denote to the adhered concrete on the rock surface, red rectangle solid lines show sandstone shear failure, and red oval solid lines represent sandstone shear failure

Full size image
Fig. 12
figure 12

The calculation process of the adhered concrete volume: a Surface morphology of sandstone with the sawtooth angle of 30° before dynamic shearing; b surface morphology of sandstone with the sawtooth angle of 30° after dynamic shearing

Full size image

Both concrete and sandstone shear failures can also be observed when the sawtooth angle equals 45°, attributed to the enhanced mechanical interaction between sandstone and concrete. With increasing sawtooth angle, the mechanical interactions of sandstone sawteeth and concrete sawteeth would be enhanced. And the enhanced mechanical interaction is the consequence of the decreased friction force on the sandstone–concrete interface in essence, leading to the increased interfacial shear resistance. As a result, the local shear force applied on the sawteeth may exceed concrete shear strength or sandstone shear strength in some cases, leading to the sandstone shear failure or concrete shear failure (Cao et al. 2019; Zhang et al.

5 Conclusions

The main conclusions of this paper are listed as follows:

  1. (1)

    Dynamic shear behaviors of sandstone–concrete structures greatly differ from those of sandstone or concrete. Both the dynamic shear strength and peak strain of the sandstone–concrete interface are less than those of sandstone or concrete.

  2. (2)

    Dynamic shear strength of sandstone–concrete specimen has an obvious strain rate effect, experiencing a growing trend with increasing strain rate, because of the increased interfacial cohesion and the change of failure mode.

  3. (3)

    Interface roughness has great influences on the dynamic shear behaviors of sandstone–concrete specimens, attributing to the change in contact area and mechanical interaction between sandstone and concrete. Both the dynamic shear strength and the dynamic peak shear strain of sandstone–concrete specimen increase with interface roughness. For specimens with rough interface comprising of different sawtooth angles, the smallest sawtooth angle dominates the dynamic shear strength.

  4. (4)

    The failure mode of sandstone–concrete specimen depends on the interface’s sawtooth angle, internal friction angle as well as the cohesion of the interface, sandstone and concrete. In addition, a coupling effect of strain rate and interface roughness on the failure mode is observed.