The effect of assembling location on the performance of precast concrete beam under impact load

Abstract

With incorporation of assembling joints, precast concrete beams could behave very differently in resisting both static and dynamic loads in comparison to conventional reinforced concrete beams. With no research available on the dynamic behavior of precast concrete beams under impact load, a combined experimental and numerical study is conducted to investigate the dynamic response of precast concrete beams under impact load. The results were also compared with reinforced concrete beams. Four groups of concrete beams were tested with all beams designed with the same reinforcement, but different assembling locations were considered for precast concrete beams. The effects of the assembling location in resisting drop weight impact of precast concrete beams were analyzed. The influence of impact mass and impact velocity on the impact resistance of precast concrete beams were also investigated. The results revealed that the further the assembling location is away from the impact location, the closer the mechanical performance of the precast concrete beam is to that of the reinforced concrete beam. When the assembling location and the impact location coincided, the assembling region suffered from severe local damages. With increased impact velocity and impact energy, the damage mode of the precast concrete beams may change gradually from bending failure to bending–shear failure and eventually to local failure. In addition, the bonding around the assembling interface was found to be effective to resist drop weight impact load regardless of the magnitude of the impact velocity and energy.

Keywords

assembling beam finite element impact load precast reinforced concrete

Introduction

Precast concrete structure construction has become increasingly popular with less space required, less water, space, and construction materials consumed, and being environmentally friendly. Especially, nowadays, the state council of China has proposed to promote the application of precast structure constructions. As such, the research on the performance of precast concrete structures is becoming necessary and inevitable.

Recently, intensive research has been conducted on the performance of precast concrete structures under static and seismic loadings (Biondini et al., 2002; Cheok and Lew, 1991, 1993; Fan and Lv, 2011; Fan et al., 2007; Liu et al., 2011; Lv et al., 2008; Priestley, 1991; Priestley and Tao, 1993; Zhao et al., 2005). Some structures have to be designed to resist impact loads, due to the crashing of comparatively rigid heavy objects at low velocities, such as falling rocks in mountain areas and falling heavy loads dealt within factories, and even explosion loadings caused by gas explosion or terrorist attacks. However, to the best knowledge of the authors, no research on the performance of precast concrete (PC) beam under impact loading can be found from the literatures. However, many studies can be found on the mechanical performance of conventional reinforced concrete (RC) and composite structures under impact loadings (Fujikake et al., 2009; Kalamar et al., 2017; Kishi et al., 2001, 2002; Liao et al., 2014; Pham and Narayanswamy, 2014; Porcu, 2017; Saatci and Vecchio, 2009; Soleimani et al., 2007; Stochino, 2016; Wang et al., 2006; Yang et al., 2012).

For example, Kishi et al. (2001, 2002) conducted drop weight impact tests on 8 bent beams and 27 beams without web reinforcement and provided a calculation method of static load bearing capacity for both the bent beams and the beams without web reinforcement which is needed for impact resistance analysis. Saatci and Vecchio (2009) conducted drop weight impact tests on eight RC beams and analyzed its failure mechanism. Soleimani et al. (2007) conducted impact tests on RC beams, and their results indicate that inertia force cannot be ignored, and it has significant influence on the mechanical performance of the structure. Sha and Hao (2015) conducted pendulum impact tests on RC models strengthened by carbon fiber–reinforced polymer (CFRP) composites. The results show that compared with unstrengthened pier, the strengthened bridge pier has a higher impact resistance capacity and hence endures less structural damage under the same barge impact load. Xu and Zeng (2012, 2014) provided the moment and shear distribution curves under impact loading, in which the effect of inertia has been taken into account. They also performed secondary impact tests following the original tests.

In addition, a lot of research has adopted numerical methods to study the RC structures under extreme loadings. Thilakarathna et al. (2010) used LS-DYNA software to conduct finite element (FE) analysis of RC structural members under the coupling of axial load and impact, and also analyzed the vulnerability of RC.Meng (2012) and Dou (2015) conducted numerical analysis using different concrete material models and found that the simulation results varied with different concrete material model. Wang et al. (2006) also established a simplified model and derived the equation for RC beams under low-velocity impact action.

However, with no current research available on the performance of PC beams under impact loading, this study is to perform experimental and numerical studies on the impact response of PC beams under drop weight impact load. This study focused to understand the influence of the different assembling locations on the impact resistance and the characteristics of the performance of the PC beams under impact load. The effects of both impact mass and impact velocity on the impact resistance of the PC beams were also studied. The results were also compared with RC beams.

Experiment setup

Specimens and materials

In this study, four groups of concrete beams with the same reinforcement are tested. Group B1 are conventional RC beams, and all the rest are PC beams with different assembling locations, as shown in Figure 1. As presented in Figure 1, groups B2, B3, and B4 are PC beams with assembling locations at1/2, 1/3, and 1/4 of beam span, respectively. The dimension of the beams is 200 mm × 400 mm × 3300 mm (width × depth × length) with a clear span of 2900 mm and a shear span ratio of 4.14. In addition, different specimens are also prepared to be subjected to different impact conditions with varied impact mass and impact height as specified in Table 1.

Figure 1.

Design details of RC beam (B1) and PC beam (B2, B3, and B4) (unit: mm).

Table 1.

Details of impact parameters for each specimen.

Specimen	Impact speed (m/s)	Impact mass (kg)	Impact height (m)	Impact energy (J)
B1a	6.86	253	2.4	5951
B1b	13.28	253	9.0	20,493
B2a	6.86	253	2.4	5951
B2b	10.84	253	6.0	14,876
B2c	13.28	253	9.0	20,493
B3	6.86	253	2.4	5951
B4a	5.60	383	1.6	3927
B4b	6.86	253	2.4	5951
B4c	10.84	253	6.0	14,876
B4d	13.28	253	9.0	20,493
B4e	8.85	383	4.0	15,014

Impact mass $v = \sqrt{2 gh}$ ; impact energy Q = mgh.

The longitudinal rebars at the tensile part and compressive part are symmetrically placed, with 2Φ16 HRB400 rebars at each side. HRB400 is a hot-rolled China standard deformed steel rebar. The yield strength of the longitudinal rebars is 478.7 MPa, and the ultimate strength is 602.2 MPa. The stirrup is HRB400 rebars with Φ6@150, the reinforcement ratio is 0.19%. The yield strength of the stirrup is 414.5 MPa, and the ultimate strength is 478.7 MPa.

The rebars at the assembling location are connected by grout sleeves. During construction, the rebars are inserted into the grout sleeves at the beginning. Then, self-compacting micro-expansive cement-based material is poured into the grout sleeve and fills the voids between the grout sleeve and the rebars. The hardened grouting material binds closely with rebars and the inner wall of the groove of the grout sleeves and thus forming effective cohesion between the two. With such construction method, the resulted forces can be effectively transferred through the connected rebars, and the mechanical properties can meet the code requirements (G/T398-2012, 2012).

The structural member is designed according to GB50001-2010. The design value of the bending capacity is 57.2 kN, and the shear capacity is 91.3 kN. Figure 1 depicts the design of the beam specimens, and the unit of length is millimeter. The designed concrete grade for the RC beams and the precast part of the PC beams was C30, and the tested compressive strength was at 36.7 MPa. The concrete grade for the cast-in-place sections for PC beams was C40, and the tested compressive strength was 42.4 MPa. The length of the cast-in-place section of the PC beams is 500 mm.

Test setup

The impact tests on the concrete beams were carried out using a high-energy drop hammer test machine. The drop hammer test machine comprises machine frame, guide rail, drop weight, lifting device, and control device. The rail is installed on the machine frame, and the hammer is attached to the cable and the hoist which can move vertically along the rail. The hammer is clamped by the clamp when being lifted. When the impact test starts, the hammer is released from the clamp at certain height and drops along the track toward the beam in a movement close to free fall. Flat hammer is used as the drop weight, with a diameter of 200 mm and a mass of 188 kg. The impacting mass is varied by adding or removing steel plates, while the impacting velocity is varied by changing the releasing height of the hammer. The weight of each steel plate is 65 kg. The maximum lifting height of the hammer is 16 m. The beam is pin-supported at both ends, and uplift plates are placed at the two supports to avoid the beam breaking away from the support in the impact. The uplift plates are connected to the base pins by tie bars, and they do not affect the rotation of the beam. Sensors are placed at the hammer, and both supports to measure the impact force and support reaction. The test setup is presented in Figure 2.

Figure 2.

Experimental setup.

Testing and measurement

The impact parameters set for each specimen beams are presented in Table 1. The impact energy Q can be expressed as Q = mgh, in which m is the mass, g is the gravity, and h is the impact height. The impact velocity v can then be described as √2gh. In this study, two impact masses 253 and 383 kg are used. And the lifting height of the hammer ranges from 1.6 to 9 m. The impact velocity of the hammer ranges from 5.6 to 13.28 m/s.

During or after the testing, the impact force of the hammer was obtained through the load cell in the hammer. The support reaction was measured by the load cells at the two end supports. Also the mid-span deflection was measured by the displacement meter located in the mid-span at the bottom of the beam. In addition, the failure mode was captured after each test.

Test results and discussion

The experimental results from the drop weight impact tests on both conventional RC beams and PC beams are presented in this section.

Failure mode

Figure 3 presents the failure mode of each specimen under the drop weight impact load. To clearly observe the cracking pattern and the failure mode of the beam, all cracks are marked with red color markers. As the cracking pattern for each specimen is detailed in Figure 3, they would not be detailed below.

Figure 3.

Failure mode of RC beams (B1) and PC beams (B2–B4) under impact.

The impact velocities of specimens B1a, B2a, B3, B4b are the same at 6.86 m/s. The failure mode of B1a, B3, and B4b are similar, with radial propagation of cracks as shown in Figure 3. The general failure mode of the specimen was bending–shear failure. The results indicate that under the same impact load, when the assembling location was close to the impact point, the failure mode of PC beams was similar to that of RC beam. When the impact point coincided with the assembling location, the PC beam could experience the most severe damages.

Comparing the specimens in Batch B2, the impact energy of B2b with an impact velocity of 10.84 m/s is higher than that of B2a with an impact velocity of 6.86 m/s. There were bending cracks with small width at mid-span in B2b. Furthermore, the shear cracks from impact point to supports propagated further and passed through the cast-in-place part. The width of bond cracks at the interface between cast-in-place part and precast part was relatively large. Spalling occurred at the tension side of the interface. The failure mode can be seemed as a shear failure. Specimen B2c had the highest impact energy with an impact velocity of 13.28m/s. However, B2c had small width of cracks at the mid-span and shear span. For the cast-in-place section, more severe damage was observed with the concrete failed at both compression and tension sides. With grout sleeves and stirrups exposed, the failure mode of the specimen was classified as local failure.

Comparing the specimens in Batch B4, Specimen B4a has the lowest impact energy with an impact velocity of 5.6 m/s. Bending cracks were formed at mid-span, and one diagonal crack was formed at each side of the shear span. This specimen experienced bending failure. The impact velocity of specimen B4b was 6.86 m/s. The compression side of the mid-span suffered severe damage with concrete collapsing at the edges, and hence, this failure mode of the beam was deemed as bending–shear failure.

The impact velocity of B4d was 13.28 m/s, which was the highest in Batch B4. It was found that the damage of B4d was the most severe one among this group. The concrete at the mid-span completely collapsed. The impact mass of B4e was 383 kg with an impact velocity at 8.85 m/s. The impact energy was close to that of B4c, but the damage was more severe than that of B4c. The concrete at the compression side was crushed. Small cracks were formed at the interface between the PC and the cast-in-place concrete, under both low and high impact energy. With the increase in the impact velocity and impact mass, the cracks in the bonding area was not further developed. This indicates that the bonding strength of the assembling location at two ends were effective and adequate to resist impact loads.

Time history curves of impact force

Figure 4 provides the time history curves of impact force for each beam. The time history curve of impact force for B2c is not available due to breakdown of the load cell to measure the impact force of the hammer during the impact. From Figure 4, each time history curve of impact force has a primary peak. The value of the peak had the largest value but with very short duration. In addition, there is also a secondary peak in the curves for B1a, B2a, B3, B4a, and B4b. The curves for B1b, B2b, B4c, B4d, and B4e dropped to 0 after the primary peak. This may be due to the fact that B1b, B2b, B4c, B4d, and B4e were severely damaged locally and the stiffness was reduced significantly. There was no rebound of the drop hammer observed, and hence, there was no second impact happened.

Figure 4.

Time histories of impact force.

The impact energy of specimens B1a, B2a, B3, and B4b was very close to each other, and their time history curves of impact force were similar. Such results indicated that the PC beams and the RC beams could have similar contact stiffness when the design parameters were all the same. When comparing the time history curves of impact force within Batch B1, B2, and B4, it is found that the peak of the impact force increased with the increase in impact velocity. The impact energy of B4d and B4e was very close, but when comparing the time history curves of impact force of B4d and B4e, it can be seen that the primary peak of B4d was significantly higher than that of B4e. The secondary peak was not substantial after the primary peak, with the same oscillating pattern. This indicates that the peak of the impact force could be determined by impact velocity. With the same impact energy, the higher the impact velocity, the larger the impact force could be exerted on the PC beam. The RC beam also has the similar performance under impact load as indicated by the previous literatures.

Time history curves of support reaction

Figure 5 provides the time history curves of the support reaction for RC beam B1a and PC beams batch B4. As can be seen from the time history curves of the reaction force of B1a and B4b, the peak values of reaction force of both curves are very close. When comparing it with the corresponding time history curves of the impact force, the impact force fluctuated around 0 when the peak values of reaction force occurred, such that most of the reaction force at support was balanced/canceled out by inertia force. This shows that the dynamic behavior of RC beams and PC beams were similar. As can been seen from the time history curves of the reaction force of B4a, B4b, B4c, and B4d, the value of the main peak increased with the impact velocity. However, the fluctuation of reaction force for each specimen was gradually dissipated to 0 after the peak, respectively. This shows that higher impact velocity resulted in higher reaction force at support, and the fluctuation of reaction force after the peaks was weaker and stiffness reduction was greater, thus the damage was more severe. This is also consistent with the observed failure mode of the beams. When comparing B4a with B4e, impact mass did not significantly affect the reaction force at supports of PC beams with the same impact energy for both beams.

Figure 5.

Time histories of reaction force.

Time history curves of mid-span deflection

Figure 6 presents the time history curves of mid-span deflection of each beam. The mid-span deflection of B3 is not available because the displacement meter for B3 fell off during the test. As can be seen from the time history curves of mid-span deflection, B1a, B2a, and B4b had the same trend of displacement variation. B2a presented with the largest mid-span deflection, with a maximum value at 34.6 mm and a residual maximum deflection at 22.6 mm. It indicates that the stiffness of B2a was dropped more than that of other beams, and its stiffness against impact was small. The time history curves of mid-span deflection of B1a and B4b almost overlapped with each other, which indicates that the impact stiffness as well as the deterioration trend for both beams were similar, and their impact resistance were also comparable to each other. Comparing the time history curves of mid-span deflection of Batches B2 and B4, it is found that with the increase of impact velocity and impact energy, the maximum mid-span deflection and the residual displacement increased correspondingly. When B4d was compared with B4e, it is found that when the impact energy was the same, the higher the impulse, the larger the mid-span deflection was.

Figure 6.

Time histories of mid-span deflection.

Previous studies (Kishi et al., 2001; Tachibana et al., 2010) proposed the relationship between the maximum mid-span deflection of RC beams under impact loading and their static bending capacity. Such relationship can be expressed as

δ_{\max} = α E / P_{us}

(1)

where δ_max is the maximum mid-span deflection of the beams, E is impact energy, P_us is static bending capacity, and α is the fitting coefficient.

Based on the available research, it is found that the maximum mid-span deflection of the PC beams was related to the impact energy as well as its static bending capacity. Such relationship of the maximum displacement and the input impact energy can be found in Figure 7, where the x-axis is impact energy E, and the y-axis is the maximum mid-span deflection of the beam; and the slope is α/P_us.

Figure 7.

Relationship of the mid-span deflection and impact energy E.

From Figure 7, the relationship between the mid-span deflection and the impact energy of the PC beams of B4 batch almost overlaps with that of the RC beams B1. It indicates that when the assembling location of the PC beam was located at the two ends of the beam, the mid-span deflections of the PC beam under impact load could be similar to that of the RC beam with the same design parameters, and their impact resistance would be comparable.

However, the slope of the curve in Figure 7 for PC beams B2, assembled at the mid-span, is relatively small. With the impact energy increased, the maximum mid-span deflection of B2 was smaller than that of B4 with different assembling location. This may be explained by the following reason. As the impact energy increased, the concrete at mid-span of both B2 and B4 beams collapsed, and the rebar at mid-span was exposed. The grout sleeve in the assembling location of B2, at the mid-span, could help B2 beams retain certain stiffness. While B4 may have lost the stiffness straight away with collapsed concrete at its mid-span. In addition, the results also suggest the effectiveness of the connecting rebars within the grout sleeve under impact load.

The impact force versus mid-span deflection relationship curves

Figure 8 gives the impact force–displacement curves of each RC and PC beam. As can be seen from Figure 8, displacement is almost zero when the impact force reaches its peak value, and the displacement starts increasing as the impact force drops after its peak value. In Figure 8, it can also be seen that the response of displacement lags behind the impact force, and therefore, the peak of the impact force forms a rough triangular area with the displacement. During this stage, the impact force was to generate the acceleration of the beam, and most impact force was canceled out by the inertia force. The response of the PC beams was consistent with that of the RC beams in this regard.

Figure 8.

Impact force versus mid-span deflection relationship curves.

Dynamic response of PC beams

Under impact loading, only part of the impact force acting on the beam can be balanced by the reaction force at the support, while the other part of the impact force will be used to generate the acceleration of the beam, which will generally cause larger displacement at the mid-span compared with the beam under static loading condition. According to d’Alembert principle, a dynamic balance equation is derived below

\begin{matrix} S (t) + R (t) = P (t), then \\ S (t) = P (t) - R (t) \end{matrix}

(2)

where S(t) is the inertia force, R(t) is the reaction force at support, and P(t) is the impact force. Take beam B4b as an example, the different force versus time curve is plotted in Figure 9, respectively.

Figure 9.

Histories of the sum of support reaction and inertia force and impact force for beam B4b.

As can been seen from Figure 9, at the very beginning of the impact process, inertia force reaches its peak as quick as the impact force and the inertia force are used to balance the impact force, which in turn verifies that the impact force acts on the acceleration of the beam at this stage. After the peak value, the inertia force was used to balance the reaction force at the support, and the displacement of the beam continued to increase.

Numerical modeling

FE model

A FE model was established using LS-DYNA software package based on the experimental study. FE analyses were conducted based on this model for B1a, B2a, B2b, B3, B4a, B4b, B4c. For simplification, the hammer used in the experimental study was simplified as a cylinder with the same diameter in the FE model. The mass of the cylinder was also the same as that of the hammer to ensure that the impact energy was the same. Both reinforcement and concrete were separately modeled. The concrete, the hammer, and the supports were modeled by solid164 elements, and the reinforcement was modeled by link160 elements. There are many material models available to simulate the dynamic response of concrete under impact loading, such as the Riedel–Hiermaier–Thoma (RHT) model (Riedel and Hiermaier, 1999), Karagozian and Case concrete (KCC) model (Malvar et al., 1997), Holmquist–Johnson–Cook (HJC) concrete model (Polanco-Loria et al., 2008), continuous surface cap model (CSCM) (Murray et al., 2007), and concrete damaged plasticity (CDP) model (Amadio et al., 2017). In this study, CSCM model was used to simulate the dynamic behavior of concrete as it can capture the varied material behaviors under different loading conditions. It also considers the strain hardening, post-peak softening and strain rate effect (Meng, 2012; Murray et al., 2007). The plastic kinematic model, which was suitable to model isotropic and kinematic hardening plasticity, was employed to model the reinforcement. The strain rate effect was also considered in this model (Jones, 1989). The rigid-body constitutive model was used to simulate the supports. In the experimental study, the connection behavior of the grout sleeves was good enough when the grout sleeves were fully grouted. As such, the grout sleeve connection was not particularly modeled in this FE model. To differentiate the cross section with grout sleeves from the cross section without grout sleeves, the cross-sectional area of rebar was increased in the model. Due to the discontinuous construction between the precast and the on-site-casting section, there were weak connections (interface) in between. Therefore, the concrete strength was weakened at the interface in the FE model. However, to what extent the concrete strength was reduced in the interface still needs to be further investigated. In this FE model, the concrete strength at assembling location was set to be 10% of the original strength, conservatively. In this FE model, the contact between the hammer and concrete was modeled as “CONTACT_AUTOMATIC_SURFACE_TO_SURFACE,” while the interaction between the hammer and the rebars, the hammer and the concrete, the hammer and the supports were all defined as “CONTACT_AUTOMATIC_NODES_TO_SURFACE.” After convergent study, the element size was set as 10 mm. In the model, it was found when the maximum plastic strain exceeded 0.1, the FE results agreed well with the experimental results. Therefore, when the maximum plastic strain in the element exceeded the number 0.1, such element would be deleted. The established FE model of the PC beam is presented in Figure 10.

Figure 10.

FE model.

FE results

Characteristics of failure mode

Figure 11 presents the damage contour of the beams obtained from the FE modeling. For RC beam B1a, PC beams B3 and B4b, the failure mode obtained in FE analysis were all bending–shear failure. The cracks propagated radially from the impact point. The elements in the bending region at mid-span were deleted due to element strain exceeding a given threshold value. In addition, the crack development in the region from the impact point to the shear span was significant as several shear microcracks and bending microcracks were formed in the shear span. The cracks in B3 were more concentrated than those in B1a and B4b. For B3 and B4, there were clear tiny cracks at the interface between precast and cast-in-place segments. The results indicate that the further the assembling location was from the impact point, the closer the failure zone to that of the RC beams.

Figure 11.

Comparison of the numerical results on failure modes of selected beams.

The failure mode of B2 is different from the common bending failure and shear failure as shown in Figure 11. The failure mode was significantly influenced by the assembling region. The damage happened around the assembling region and more specifically at both sides of the interface.

When comparing the PC beams in group B4, the impact velocity of B4a was relatively low, and consequently, its failure mode was less severe. And the failure mode was close to bending failure (Figure 11). In comparison to B4a, the impact velocity of B4b was higher. The cracks in B4b were further developed compared with the cracks in B4a, and its failure mode was bending–shear failure. With a much higher impact velocity of B4c, its damage was much more severe than the former. The cracks were very concentrated toward to the impact point, and the bending cracks at mid-span were not obvious. Furthermore, the development of the shear cracks from the impact point to the shear span was relatively severe. Concrete tends to be crushed in both compression and tension sides. And the failure mode of this PC beam was categorized as shear failure. Above all, the FE results discussed above agreed well with the experimental results discussed in previous section.

Characteristics of the impact force and mid-span deflection

Table 2 presents the peak impact force of selected beams for comparison of the FE modeling and experimental results. B1a, B2a, B3, and B4b were presented as they had the same impact energy and the impact velocity of the impact test. The obtained primary peak impact force for these specimens from FE modeling was very close to 1400 kN. When compared to the experimental results, the ratio is ranging from 0.93 to 1.12 as shown in Table 2. The difference was within 15% which indicates the good agreement between the FE results and experimental results.

Table 2.

Comparison of test results and FE results.

Specimen	Impact energy (J)	Peak impact force (kN)			Max mid-span deflection (mm)
Specimen	Impact energy (J)	Test	FE	FE/Test	Test	FE	FE/Test
B1a	5951	1245	1393	1.12	30.00	29.7	0.99
B2a	5951	1341	1468	1.09	34.58	29.5	0.85
B2b	14,876	2324	2218	0.95	77.10	72.7	0.94
B3^a	5951	1354	1394	1.03	–	28.6	–
B4a	3921	1086	1236	1.14	21.10	19.6	0.93
B4b	5951	1470	1364	0.93	29.60	29.9	1.01
B4c	14,876	2046	2100	1.03	76.1	86.6	1.14

FE: finite element.

The mid-span deflection of B3 was not captured because the displacement meter for B3 fell off during the test.

Few other examples are also presented in Table 2 to show the effects of impact energy on the impact force, including B2b in Group B2, B4a, and B4c in Group B4. Comparing the results for group B2 and B4, it is found that with the increased impact velocity and impact energy, the peak impact force was greatly increased as indicated by the experimental results. And the ratio of FE results versus the test results was also within 15% difference, which again confirmed the capability of the FE model in simulating both RC and PC beams under impact force. Also, with the increased impact velocity and impact energy, the difference between the FE result and experimental result became smaller. This could be due to the fact that when impact velocity and the impact energy were increased, the consequent proportion of the energy loss from the input energy became smaller. In sum, the FE modeling results agreed well with the experimental results, and the FE modeling can accurately simulate the characteristics of the impact force for both RC and PC beams under impact load.

Furthermore, the maximum mid-span displacement of those beams discussed above obtained from FE modeling is also compared with the test results in Table 2. From Table 2, the maximum mid-span deflection of B1a, B2a, B3, and B4b obtained from FE, with the same impact energy, is very close to 30mm, ranging from 28.6 to 29.9 mm. The stiffness of each beam model was very close to each other. Compare group B2 and B4 specimens, it is found that as impact energy and impact velocity increased, the maximum displacement of beam was increased accordingly. When comparing the FE results with the test results, the ratio were ranging from 0.85 to 1.14, providing the differences within 15%, which indicates the good agreement between FE modeling and experimental results in simulating the deflection of both RC and PC beams under impact loads.

Conclusion

In this study, drop weight impact tests were conducted on four groups of beams including nine PC beams in comparison with two conventional RC beams. The influence of assembling location on the impact resistance of PC beams under impact load was investigated. The effects of impact mass and impact velocity were also studied. FE models were established using LS-DYNA. It is found that the FE model can accurately simulate the characteristics of the dynamic response of both RC and PC beams under impact load, and the FE results agreed well with experimental results.

Based on the results, it is found that the assembling location could significantly influence the impact resistance and the response characteristics of the PC beams under drop weight impact load. When the impact location being further away from the assembling location, the PC beams could behave similarly to the RC beams. Such similarity includes the failure mode, contact stiffness, and integral stiffness against impact. However, the PC beam could experience very severe local damage even under a low-intensity impact when the impact happened right on the precast assembling location. With impact velocity and impact energy increasing, the damage to the PC beam could become more and more severe with the failure mode gradually changed from bending failure to bending–shear failure, and eventually to local failure. However, it needs to be mentioned that the interface bonding in the PC beams was found to be effective in resisting impact load regardless of the magnitude of the impact velocity and energy applied in this study.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Key Basic Research and Development Program of China (no. 2015CB058000), the Natural Science Foundation of China (nos 51678018 and 51508294), and the Beijing Natural Science Foundation (no. 8172010).

References

Amadio

Bedon

Fasan

et al . (2017) Refined numerical modelling for the structural assessment of steel-concrete composite beam-to-column joints under seismic loads. Engineering Structures 138: 394–409.

Biondini

Ferrara

Negro

et al . (2002) Results of pseudo dynamic test on a prototype of precast R.C frame. In: Proceedings of the 14th national CTE congress, Mantova, 25 November.

Cheok

Lew

(1991) Performance of precast concrete beam-to column connections subject to cyclic loading. PCI Journal 36(3): 56–67.

Cheok

Lew

(1993) Model precast concrete beam-to-column connections subject to cyclic loading. PCI Journal 38(4): 80–92.

Dou

(2015) Experimental study on the behavior of high strength reinforced concrete beams under impact loading. PhD Thesis, Beijing University of Technology, Beijing, China.

Fan

(2011) Seismic response analysis of a jointed. China University of Mining and Technology Journal 40(4): 544–548.

Fan

Zhao

et al . (2007) Pseudo dynamic test on jointed precast concrete frame structures. Journal of Earthquake Engineering and Engineering Vibration 27(6): 97–105.

Fujikake

Soeun

(2009) Impact response of reinforced concrete beam and its analytical evaluation. Journal of Structural Engineering 135(8): 938–950.

G/T398-2012 (2012) The grouting coupler for rebars splicing (in Chinese).

10.

Jones

(1989) Structural Impact. Cambridge: Cambridge University Press.

11.

Kalamar

Bedon

Eliášová

(2017) Assessing the structural behaviour of square hollow glass columns subjected to combined compressive and impact loads via full-scale experiments. Engineering Structures 143: 127–140.

12.

Kishi

Ikeda

Mikami

et al . (2001) Dynamic behavior of RC beams under steel weight impact loading-effects of nose-shape of steel weight. In: Proceeding of third international conference on concrete under severe conditions, Vancouver, BC, Canada, 18–20 June, pp. 660–667.Vancouver, BC, Canada: University of British Columbia.

13.

Kishi

Mikami

Matsuoka

et al . (2002) Impact behavior of shear-failure-type RC beams without shear rebar. International Journal of Impact Engineering 27(9): 955–968.

14.

Liao

Zhang

Tian

(2014) Experimental research on impact resistance properties of RC beams strengthened with high strength steel wire mesh and high performance mortar. Journal of Vibration Shock 33(12): 200–206.

15.

Liu

Song

Jiang

(2011) Experimental study on seismic performance of post tensioned precast prestressed concrete frame. Journal of Building Structures 32(2): 24–32.

16.

Fan

Zhao

(2008) Pseudodynamic test on a reduced scale jointed precast concrete frame structure. Journal of Building Structures 29(4): 58–65.

17.

Malvar

Crawford

Wesevich

et al . (1997) A plasticity concrete material model for DYNA3D. International Journal of Impact Engineering 19: 847–873.

18.

Meng

(2012) Experimental and Numerical Simulation Study on Reinforced Concrete Beam under Impact Loading. Changsha, China: Hunan University.

19.

Murray

Abu-Odeh

Bligh

(2007) Evaluation of LS-DYNA concrete material model 159. Report no. FHWA-HRT-05-063. Washington, DC: Federal Highway Administration.

20.

Pham

Narayanaswamy

(2014) An optimization tool for impact analysis of composite structures. Procedia Engineering 75: 3–8.

21.

Polanco-Loria

Hopperstad

Borvik

et al . (2008) Numerical predictions of ballistic limit for concrete slabs using a modified version of the HJC concrete model. International Journal of Impact Engineering 35: 290–303.

22.

Porcu

(2017) Ductile behavior of timber structures under strong dynamic loads. In: Concu

(ed.) Wood in Civil Engineering. Rijeka: InTechOpen, pp. 173–196.

23.

Priestley

MJN

(1991) Overview of PRESSS Research Program. PCI Journal 36: 50–57.

24.

Priestley

MJN

Tao

(1993) Seismic response of precast prestressed concrete frames with partially debonded tendons. PCI Journal 38(l): 58–66.

25.

Riedel

Hiermaier

(1999) Penetration of reinforced concrete by BETA-B-500-numerical analysis using a new macroscopic concrete model for hydrocodes. In: Proceedings of the 9th international symposium on interaction of the effect of munitions with structures, Berlin, 3–7 May.

26.

Saatci

Vecchio

(2009) Effects of shear mechanisms on impact behavior of reinforced concrete beams. ACI Structural Journal 106(1): 78–86.

27.

Sha

Hao

(2015) Laboratory tests and numerical simulations of CFRP strengthened RC pier subjected to barge impact load. International Journal of Structural Stability and Dynamics 15(2): 1450037.

28.

Soleimani

Banthia

Mindess

(2007) Behavior of RC beams under impact loading: some new findings. Fracture Mechanics of Concrete and Concrete Structures 2: 867–874.

29.

Stochino

(2016) RC beams under blast load: reliability and sensitivity analysis. Engineering Failure Analysis 66: 544–565.

30.

Tachibana

Masuya

Nakamura

(2010) Performance based design of reinforce concrete beams under impact. Natural Hazards and Earth SystemSciences 10(6): 1069–1078.

31.

Thilakarathna

HMI

Thambiratnam

Dhanasekar

et al . (2010) Numerical simulation of axially loaded concrete columns under transverse impact and vulnerability assessment. International Journal of Impact Engineering 37(11): 110–112.

32.

Wang

Song

(2006) A calculation method of reinforced concrete beam under low velocity impact. Acta Armamentarii 27(3): 399–405.

33.

Zeng

(2012) Experimental study on the impact-resistant behavior of RC beams without shear-resistant rebar. Journal of Civil Engineering 45(9): 63–73.

34.

Zeng

(2014) Experimental study on the behaviors of reinforced concrete beams under impact loadings. Journal of Civil Engineering 41(2): 51–61.

35.

Yang

Lam

NTK

Zhang

(2012) Evaluation of simplified methods of estimating beam responses to impact. Journal of Structural Stability and Dynamics 12(3): 1250016.

36.

Zhao

Liu

(2005) Experimental study on seismic behavior of precast concrete beam-column sub-assemblage with fully assembled joint. Journal of Earthquake Engineering and Engineering Vibration 25(1): 81–87.