Flexural Failure Research Articles

Experimental and numerical study on the impact resistance of ultra-high performance concrete strengthened RC beams

To enhance the impact resistance of built reinforced concrete (RC) members, the effectiveness of using ultra-high performance concrete (UHPC) for strengthening RC structures was investigated in this study. The mechanical properties of UHPC were evaluated by the uniaxial compression, tension and flexural bending tests. Drop weights with hemispherical and wedge-shaped indenter were adopted to impact the beam specimens with and without UHPC strengthening. A total of six beams, including three control RC beams and three UHPC jacketed RC (RC-UHPC) beams, were tested. The beams were prestressed in the axial direction with a 200 kN force. The test results revealed that the UHPC jackets improved the structural impact resistance. With an impact mass of 411 kg and an impact velocity of 4.95 m/s, the maximum and residual deflection of the RC-UHPC specimen decreased by 15.3 % and 21.1 % as compared to the RC control specimen, and the failure mode shifted from diagonal shear failure to flexural failure. To further investigate the dynamic responses of the beams, a detailed finite element model was established and validated with the test results in terms of the impact force, structural deflection and damage profile. The dynamic shear force and bending moment distribution diagrams were numerically derived to examine the failure mechanism of the test specimens. Finally, a parametric study was conducted to evaluate the effect of different impact locations and UHPC jacket length on the impact resistance of strengthened RC beams.

Nonlinear finite element analysis of non-symmetrical punching shear of rectangular flat slabs supported on square columns

Different tests and calculation methods for determining the punching shear of flat slabs are based on symmetrical punching shear. However, in practical engineering, most flat slabs are not subjected to symmetrical punching shear due to the differences of two orthogonal directions in load, slabs reinforcement, column section size and slabs plane dimensions. The calibrated nonlinear finite element model (FEM) effectively expands the test database, performs variable parameter analysis, and verifies the calculation method. In this paper, the FEM of isolated slab-column connections is developed using ABAQUS and calibrated based on 14 test results from the literature. The simulation of punching behavior of 49 slab-column connections is carried out, the corresponding effects of the shear span ratio and the slab aspect ratio are evaluated. The results indicate that the shear span ratio and the slab aspect ratio significantly affect the crack development, reinforcement stress, stiffness, and punching shear strength. Furthermore, the slab aspect ratio influences the symmetry of the response. With the increase in the slab aspect ratio, the failure mode changes from symmetrical punching shear failure to non-symmetrical punching shear failure and flexural punching failure. The finite element analysis (FEA) prediction results of the ultimate load are compared with the calculation results of four equations of ACI, EC2, CSCT-simplified, and CSCT (ψX − ψY). The results confirm that the prediction using equations of CSCT (ψX − ψY) is reasonable, ACI method gives conservative results, whereas the safety factor of EC2 and CSCT-simplified is less than 1.0.

Flexural behavior of pre-damaged and repaired reinforced concrete beams with carbon fiber reinforced polymer grid and engineered cementitious composite

In this study, the flexural behavior of pre-damaged reinforced concrete (RC) beams repaired with carbon fiber reinforced polymer (CFRP) grids and engineered cementitious composite (ECC) was investigated. Fifteen beams were designed and subjected to four-point bending tests. Three types of specimens were tested: undamaged, pre-damaged with removed load (i.e., loaded to the target level and then unloaded), and pre-damaged under sustained load. The main testing parameters were strengthening material (ECC alone, or CFRP grid plus ECC), CFRP grid thickness, main reinforcement ratio, preloaded level, and repairing process. The experimental results showed that most of the repaired beams exhibited flexural failure caused by concrete crushing, and a few others failed by debonding. The proposed repairing technique was effective in enhancing the flexural stiffness and bearing capacity of pre-damaged RC beams. For the repaired beams, the flexural bearing capacities did not significantly decrease as the preloaded level increased from 0 to 0.45 and slightly decreased when the preloaded level exceeded 0.65. A mathematical model based on the plane section assumption was proposed to predict the flexural capacities of the repaired beams, considering whether the load causing the damage was removed or sustained. The predicted capacities were in good agreement with the experimental results.

Flexural performance of prefabricated composite beams with grouped bolt shear connectors under positive bending moments

Four prefabricated steel–concrete composite beams (PSCCBs) with grouped bolt shear connectors were designed and tested to investigate their flexural performance under positive bending moments. The effect of shear connection degrees and segmented precast concrete (PC) slabs on the flexural performance of PSCCBs was investigated. Experimental results show that all the specimens failed due to the yielding of steel girders and concrete crushing in the mid-span regions, presenting typical characteristics of flexural failure. Compared to the PSCCB with a monolithic PC slab, the PSCCBs with segmented PC slabs experienced a short fast increasing stage of mid-span deflection after the steel girders yielded, of which the initial flexural stiffness and ductility coefficient decreased by 47.8% and 44.1% at most, respectively. Then, finite element (FE) models of PSCCBs with different design parameters were developed to assess their flexural behaviors under positive bending moments. Numerical results demonstrate that the initial flexural stiffness was significantly affected by the girder web height and the length of the middle PC slab. Enlarging the gap width, the length of the middle PC slab, and the girder web height are beneficial for the ductility of PSCCBs. The research findings might provide references for the design of PSCCBs with segmented PC slabs and grouped bolt shear connectors.

A comparative study on the seismic performance of RC beam-column joints retrofitted by ECC, FRP, and concrete jacketing methods

Reinforced concrete (RC) beam-column joints in many old buildings do not have appropriate seismic detailing that allows them to provide sufficient strength and stiffness in strong seismic events. To improve the seismic performance of these deficient beam-column joints, this study proposes a new retrofitting method based on high-strength steel bar mesh and Engineering Cementitious Composite (ECC), a fiber-reinforced cementitious material that shows high tensile ductility and damage tolerance. The proposed method reduces the required size of the retrofitting jacket through the combined use of ECC jackets and high-strength steel bar mesh. The performance of the proposed ECC jacketing method was compared with that of concrete jackets and fiber-reinforced-polymer (FRP) jackets. Experimental program included cyclic testing of four exterior beam-column joints, i.e., one control specimen and three others retrofitted using ECC, concrete, and FRP jacketing. Results showed that (1) reinforced ECC jackets confined the joint core and prevented localized damage in the joint zone via formation of dense and fine cracks, (2) the beam-column joint retrofitted using ECC jackets with high-strength steel mesh exhibited the favorable beam flexural failure, while the other three joints showed joint shear failure, (3) compared to other two retrofitting methods, the reinforced ECC jacket more effectively increased the strength, stiffness retention, energy dissipation capacity, deformation capacity, and collapse drift. In addition to experimental study, suitable strength models were suggested to evaluate the strength and failure patterns of the retrofitted exterior beam-column joints.

Static behaviours of steel-concrete-steel sandwich beams with novel interlocked angle connectors: Test and analysis

New steel-concrete-steel sandwich beams with interlocked angle connectors (SCSSB-IACs) were first proposed, and their static behaviours were studied via conducting two-point loading tests. The failure mode, load–displacement response, resistance, and stiffness of the SCSSB-IACs were obtained. Moreover, the influences of thickness of steel faceplate, spacing of IACs, thickness of concrete core, shear-span ratio and interlocking bolts on static performances of the SCSSB-IACs were discussed. Two failure modes of the SCSSB-IACs were observed from the static tests, including flexural and shear failure. Moreover, test results indicated that the resistance and stiffness of the SCSSB-IACs were significantly affected by the thickness of steel faceplate, spacing of IACs and thickness of concrete core. In addition, the load–strain curves and deformation shapes of the SCSSB-IACs were obtained and discussed. Finally, analytical models were developed for predicting elastic stiffness and resistance of the SCSSB-IACs. The analytical models were also validated against the test data.

Failure Mechanisms of Fibre Reinforced Shotcrete: Numerical Simulations Considering Local Variations in Thickness and Bond Strength

Abstract Fibre-reinforced shotcrete is the most common support method for hard rock tunnels in the Nordic countries. The design of shotcrete is often based on empirical methods or simplified analytical equations, which neglect variations in mechanical properties and shotcrete thickness. Data collected from the field shows that significant variations in shotcrete thickness and bond strength should be expected during tunnel construction. However, how this affects the structural behaviour and capacity of the shotcrete lining is unknown. Moreover, the design philosophy for shotcrete assumes that the primary failure modes of shotcrete, i.e. bond and flexural failure, can be treated separately. This was derived based on observations of experiments in a laboratory environment. Therefore, the focus of a finalized doctoral project was to develop a numerical framework to simulate the structural behaviour of fibre-reinforced shotcrete in interaction with hard rock and rock bolts. The effect of variations in shotcrete thickness and bond strength was studied through numerical simulations to increase the understanding of its effect on the failure load of the lining. The results indicate that the most important parameter is the mean value of the shotcrete thickness and bond strength around a narrow perimeter of the block.’

Innovative bolted-steel box for seismic-resistant dry-precast concrete beam-column connection

Currently, many buildings are constructed with precast concrete methods. However, they still use the cast in-situ method, especially for the beam and column connection system where there are some obstacles in the casting process and the possibility of installing the reinforcement bars on site not following the design. Hence, there is potential for structural failure in that part of the connection. To reduce the connection constraints, it is proposed that the method of connecting precast concrete with an innovative connection system that is simple, lightweight, fast in the manufacturing process, can be constructed quickly and effectively on-site, and easily meets the specifications of connection quality. This proposed type of connection is a dry joint type that only uses bolts for the connection system with steel box components. This study will prove that the dry joint system can receive the moment and shear load due to earthquakes still in the elastic phase, not experiencing plasticity. The concept of connection planning does not change the concept of the Special Moment Frame System which applies the concept of Strong Column Weak Beam and Plastic Hinges. The connection is ensured in strong conditions, there is no yielding and damage at the time of a large earthquake, so the failure is not in the connection but occurs in the precast concrete block element that is at the end of the beam outside the planned connection in flexural failure with ductile behavior, not shear failure.

The Effect of Binder Types of Strengthening Panel for Shear Strengthening Reinforced Concrete Beams

This paper presents the shear behavior, crack patterns, and failure modes of reinforced concrete beams strengthened with geopolymer and Portland cement mortar panels. The specimens were prepared in the form of geopolymer and Portland cement mortar panels with a size of 300 mm x 700 mm, each consisting of 4 pieces. The reinforced concrete beams specimen were produced with dimension of 150 mm x 300 mm x 2300 mm and static monotonic loading test was carried out. To ensure the strength of the geopolymer and concrete used is in based on the target, a compressive strength test is carried out at the age of 28 days, with 3 samples each. According to the results of the loading test of reinforced concrete beams, the maximum load on the beam strengthed with geopolymer and Portland cement mortar panels increased by 34.98% and 29.54%, respectively, compared to the control beam. The addition of panels to reinforced concrete beams causes it’s behavior to change, becoming more ductile. This also has an impact on the crack patterns and failure modes. CB beams experience shear failure, which is indicated by the appearance of diagonal cracks along the shear area to the point of loading, whereas mortar panel reinforced beams experience flexural failure.

Seismic behaviors of T-shaped steel-concrete-steel sandwich composite walls using enhanced C-channels

This paper developed the T-shaped steel-concrete-steel sandwich composite (TSCSSC) walls using enhanced C-channels, and investigated their seismic behaviors through three cyclic loading tests. The parameters studied in the cyclic loading test programme included wall thickness of the steel tube in boundary CFST columns and aspect ratio of TSCSSC wall specimens. The test results showed that TSCSSC walls exhibited flexural failure modes characterized by infilled-concrete crushing, tensile fracture of steel tubes, weld seams, and local buckling of steel faceplates and tubes. The skeleton hysteretic P-Δ curves of TSCSSC walls exhibited a typical three-stage working mechanism, i.e., elastic, nonlinear development, and recession stage. Increasing the thickness of steel tubes in the web wall improved the strength and energy dissipation capacities. Reducing the aspect ratio of TSCSSC walls improved strength and stiffness, but exhibited lower peak bending resistance owing to the effect of acting shear force on wall-base cross section. Finally, prediction equations were developed to estimate the peak bending resistance of TSCSSC walls. Validations against reported test results confirmed their limited prediction errors (≤10%). • T-shaped steel-concrete-steel sandwich composite (TSCSSC) walls with enhanced C-channels is proposed. • TSCSSC walls exhibited flexural failure and plump hysteretic load-drift behavior. • Increasing the wall-thickness of boundary steel tube improves seismic behaviors of TSCSSC walls. • TSCSSC walls with low aspect ratio failed in combined flexure and shear mode. • Developed prediction equations predict well the bending resistance of TSCSSC walls.

An improved 2DOF model for dynamic behaviors of RC members under lateral low-velocity impact

Under lateral low-velocity impact, the RC slender members, i.e., beams and columns, commonly exhibit the local punching-shear as well as the global shear, flexural, flexure-shear and instability failure modes. However, the existing theoretical models are incapable of simultaneously predicting the above failure modes and the corresponding dynamic responses of RC members. This study aims to develop a two-degree-of freedom model (2DOF) model to predict the dynamic behaviors of RC members exhibiting the above-mentioned failure modes under the lateral low-velocity impact. Firstly, the combined dynamic flexural and shear resistance function of RC members is established and verified by rapidly concentrated loading tests on RC beams. Secondly, an improved 2DOF model is proposed by incorporating the combined dynamic resistance function, P-δ effect and the shear wave propagation effect for the local and global response stages of members. Then, the improved 2DOF model is comprehensively validated by eight series of existing drop hammer tests (165 scenarios), in terms of the failure modes, impact forces and mid-span deflections of RC members. It is further found that, under impact loadings, the failure mode of the static flexure-critical RC beams would transit to the punching-/flexure-shear failure, which is more sensitive to the impact velocity than the impactor mass, and can be effectively inhibited by increasing sectional shear capacity. Finally, aiming to assist the impact-resistant design and evaluation of RC members, the corresponding flowchart is given.

Durability of BFRP bars and BFRP reinforced seawater sea-sand concrete beams immersed in water and simulated seawater

Seawater sea-sand concrete (SSC) members reinforced with fiber-reinforced polymer (FRP) have considerable development potential. However, an adverse effect of alkalinity in seawater sea-sand concrete on the long-term performance of FRP was found, and the durability of FRP reinforced SSC structures in seawater is always a significant concern. Thus, this study examined the durability of basalt-fiber-reinforced polymer (BFRP) bars reinforced SSC beams in simulated seawater and tap water. Tensile and four-point bending tests analyzed the tensile properties of the BFRP bars embedded in SSC and the flexural performance of the BFRP-reinforced SSC beams respectively. The tensile strength of the BFRP bars decreased with the immersion time and the environmental temperature. Further, the tensile strength of the BFRP bars declined more severely in tap water than in seawater. The failure of the BFRP-reinforced SSC beams changed from shear to flexural failure owing to the reduction of the balanced ratio of the BFRP-reinforced beams in seawater and tap water. Moreover, the ultimate load-carrying capacity of the BFRP-reinforced SSC beams decreased, and they showed brittle failure characteristics. Finally, the ultimate load-carrying capacity of the BFRP-reinforced beams was predicted as a function of the degradation of the tensile strength of the BFRP bars. The developed equation conservatively estimated the flexural-load-carrying capacity of the BFRP-reinforced beams in actual oceanic environments.

Flexural behaviors of prefabricated composite beams with segmented precast concrete slabs

The flexural behaviors of prefabricated steel-concrete composite beams (PSCCBs) with segmented precast concrete (PC) slabs were investigated through four-point bending tests. Experimental results showed that all the tested PSCCBs failed in flexural failure. The PSCCBs with segmented PC slabs exhibited good deformability and ductility, and the displacement ductility coefficient was not less than 13.3; the ultimate load-carrying capacity of the beams decreased by 8.7% when the shear connection degree reduced from 1.0 to 0.65. Three-dimensional finite element (FE) models were developed to evaluate the influence of parameters, such as the gap width (GW) between adjacent PC slabs and the length of the middle PC slab (LM), on the flexural behaviors of PSCCBs. Numerical results showed that the LM and GW are tightly correlated with the flexural behaviors of PSCCBs with segmented PC slabs, but their effect on the ultimate load-carrying capacity is negligible. A prediction method for the mid-span deflection of PSCCBs with segmented PC slabs was proposed based on the numerical analyses. The comparisons show that the theoretical and numerical results agree well with each other. The proposed method could be used for the preliminary prediction of the mid-span deflection of PSCCBs with segmented PC slabs when the design load is no more than half of the load-carrying capacity.

Behaviour of mass timber members subjected to consecutive blast loads

This paper presents the results of an experimental and analytical investigation into the behaviour of full-scale glued-laminated timber (glulam) beams and cross-laminated timber (CLT) panels subjected to consecutive shock-tube-simulated blast loads. Superficial crack development was observed in the glulam specimens even when loaded within their elastic range. However, these cracks did not appear to affect the overall stiffness and strength of the specimens, as proven through subsequent loading and verification of the experimental resistance curves. Extensive propagation of rolling shear damage was prevalent within the transverse layers of the CLT specimens once flexural failure initiation occurred. Following significant rolling shear damage, the composite nature of the CLT panels were observed to degrade, corresponding to a loss of stiffness and strength. An equivalent single-degree-of-freedom (eSDOF) modelling methodology, capable of taking into consideration the dynamic behaviour of both specimen types and damage accumulation due to subsequent loading, is presented and validated using the experimental test data.

Comparison of fiber reinforcing methods of composite resin: A flexural strength and stereo microscopy study.

This study aimed to compare the effect of fiber reinforcing methods on the flexural strength and failure modes of indirect composite resins. Based on the reinforcement methods, the bar specimens (3 × 3 × 25 mm) were divided into five groups (n=20). Glass or polyethylene fibers were used for reinforcement of indirect composite resins. Fibers were either light polymerized and mixed with indirect composite resin or mixed with indirect composite resin after resin application and polymerized together. Indirect composite resin without fiber reinforcement was used as control. All five types of specimens were stored in distilled water at 37°C for 24 h. Half of the specimens were additionally thermocycled. Then the specimens were tested in a three-point bending test. Failure types were examined and categorized by using stereo microscope. The data were analyzed using two-way ANOVA and Tukey HSD test. Flexural strength was found to be significantly higher for fiber-reinforced indirect resin composites than control. However, the fiber-reinforced groups did not present any significant difference. Analysis revealed aging does not affect the flexure strength of fiber reinforcement of indirect composite resin. The study concluded that the flexure strength of indirect composite resins was improved with fiber reinforcement. Different fiber reinforcement methods demonstrated similar effects on the flexure strength of indirect composite resin. Reinforcement with glass or polyethylene fibers presented the potential to improve the mechanical properties of indirect composite resins. RESEARCH HIGHLIGHTS: Flexural strength of indirect composite resins are affected by the reinforcement of composites with glass or polyethylene fibers. Aging with thermocycling has no effect on the flexural strength of the indirect composite resins, however can cause catastrophic failures in material.

Theoretical and numerical analysis of flexural toppling failure in soft-hard interbedded anti-dip rock slopes

Soft-hard interbedded structures are commonly found in anti-dip rock slopes and endow the slopes with special properties. In this work, a simple two-beam model was first used to discuss the possible failure modes of rock layers with different lithology. The exact failure sequence of rock layers depends on their tensile strengths, elastic moduli, and thicknesses. Then, based on the deformation compatibility of rock layers, a novel theoretical method called the DCM for assessing the stability of soft-hard interbedded anti-dip rock slopes was proposed. A centrifuge test and a numerical model were used to verify the feasibility of the DCM. The calculated critical g-level of 47 g for the centrifuge model and a safety factor of 0.96 for the numerical model agree well with the results of the centrifuge test (45 g) and numerical simulation (0.98). Moreover, the traditional limit equilibrium method (LEM) was also adopted to explore these two models. The calculated results of 68 g and 1.79 using the traditional LEM were found to be much larger than the validated values, with deviations of more than 50%. This finding implies that the traditional LEM is not suitable for assessing the stability of soft-hard interbedded anti-dip rock slopes. In addition, the failure evolution of toppling failure in soft-hard interbedded anti-dip rock slopes was investigated using numerical modelling. The results show that the soft and hard rock layers do not fail at the same time; the rock layers of one type of lithology are destabilized first, then the other rock layers begin to fail until the entire slope undergoes flexural toppling.

Experimental investigation on seismic performance of plain engineered cementitious composite beams

The authors proposed using plain engineered cementitious composite (ECC) in construction without aligned reinforcement. In this study, ECC with a tensile strain capacity of up to 10% was produced, which has a comparable deformability level to steel reinforcement. To demonstrate the feasibility of plain ECC construction, five plain ECC beams and two steel reinforced concrete reference beams were prepared, and four-point bending reversed cyclic loading tests were conducted on the beam specimens. The experimental tests indicated that the plain ECC beams showed more stable hysteresis loops with improved energy dissipation capacity, deformation restoring capacity, and lower stiffness degradations than normal RC beams. Two categories of failure modes were observed, i.e., flexural failure mode and shear failure mode. With the increase in the shear span ratio, the failure mode of ECC beams changed from shear failure to flexural failure. Additionally, the increase in sectional dimension had a positive influence on the stiffness and energy dissipation capacity of ECC beams. It is experimentally verified that the plain ECC beams have sufficient load-bearing capacity under static and seismic loading.

Research on the Bearing Performance of Prefabricated New Type Subway Track Slab with Basalt Fiber

Currently, most ballastless tracks in Chinese subways are traditional cast-in-situ concrete structures, which require long construction progress. In contrast, prefabricated ballastless tracks can greatly reduce the construction period. However, there are few theoretical and experimental analyses on the prefabricated new type of subway slab track. Hence, to study the mechanical properties of the new type of subway track slab, static bending crack tests of one standard slab track and nine prefabricated full-scale track slabs with basalt fiber were carried out in this paper. The influence of the reinforcement arrangement method and the basalt fiber content on the performance of the track slabs was studied. Results showed that with the increase in the basalt fiber content in the range of 0.1–0.3%, the load–strain curve growth rate increased, the stiffness decreased, and the flexural resistance reduced. All the specimens exhibited a linear portion (elastic) and then a nonlinear portion (plastic) followed by flexural failure. During the elastic stage, the basalt fiber helped to improve the rigidity of the track slab. During the plastic stage, the influence of the basalt fiber content on the failure load varied with different reinforcement arrangements. With the reinforcement arrangement F2 and 0.1% basalt fiber content, the specimen had the smallest load–strain curve growth rate and the best overall flexural performance and crack resistance, and its strain was 23.35 με at an elastic limit of 95 kN. The results can provide a reference for the design of the prefabricated slab ballastless track.

Performance Flexural of RC Beams Without Concrete at Tension Cross-section

This study aims to analyze the flexural capacity of RC without concrete in a tension cross-section using an experimental method. The number of specimens is three pieces, namely a spiral reinforced concrete beam (SBC) and a vertical reinforced concrete beam (CBN); both of these blocks are without concrete in the cross-section of the reinforcement and 60D tensile steel reinforcement in the support area, where D is the primary diameter, and a conventional concrete beam as the control beam (CB). The beam size is 3100×150×200 mm. The beams are supported by simple supports with a span of 3000 mm. The concrete in the structural beam elements, which work optimally to withstand the load, is the outermost fibre part of the side, while the concrete on the tension side does not have a direct role in determining the magnitude of the resisting moment. Therefore, the quality of the concrete in the concrete beam section must be optimized, while the concrete in the tension section must be minimized. Eliminating concrete in tension areas reduces the construction's self-weight and use of concrete-making materials. The main variables in this research are bending behaviour and crack pattern. The beam specimens were tested with two-point loading monotonically. By observing the crack pattern and failure mode, the results showed an increase in the capacity load of SBC by 21.58% CBN but a decrease of 27.57% compared to the CB control beam. Flexural cracks and beam failures resembled under-reinforcing. The flexural capacity was analyzed based on static analysis and then validated by calculating the ratio between the theoretical nominal moment and the experimental moment. This finding shows that changing the conventional shear reinforcement model to spiral can increase the flexural of the beam without concrete in the tension cross-section. Doi: 10.28991/CEJ-2022-08-11-014 Full Text: PDF

Experimental study on flexural behavior of hollow steel-UHPC composite bridge deck

The application of UHPC might reduce the weight of the bridge deck and improve its crack resistance and durability. This study proposes a hollow steel-UHPC composite deck composed of the hollow steel tube and UHPC without shear connectors. The flexural test was performed on the hollow steel-UHPC composite deck to evaluate the effects of the embedded steel tubes, steel bars, UHPC material, and flange thickness. In addition, the mechanical properties of hollow steel-UHPC composite decks were compared with waffle decks and solid concrete decks. Results indicated that the hollow steel-UHPC composite deck had excellent bending capacity and material utilization. The load-deflection curve of the composite decks could be divided into four stages: elastic stage, working stage with cracks, nonlinear stage, and fully plastic stage. Without any additional shear connection, the steel-UHPC interface slip did not occur until the load reached 0.89Pu. In the ultimate limit state, the strains at the mid-span cross-section of steel and UHPC were almost the same, suggesting that there was a good composite effect between the embedded steel tube and the outer UHPC. The embedded steel tube had a significant influence on improving the original stiffness (K0), the stiffness at the working stage with cracks (K1), and the flexural capacity (Pu) of composite decks. The flexural failure of the hollow steel-UHPC composite deck was controlled by the tensile zone. Finally, a theoretical formula was proposed to predict the flexural capacity of the hollow steel-UHPC composite deck.