Effect of vibration on the interface properties of welded steel joints and filled concrete in steel pipes | Scientific Reports

Oct 14, 2024

Scientific Reports volume 14, Article number: 20391 (2024) Cite this article

277 Accesses

Metrics details

Concrete-filled steel tubes (CFSTs) have been increasingly utilized in engineering due to their excellent mechanical properties. Ensuring a solid bond between a steel tube and concrete is essential for optimizing their synergistic effect. This study introduces an internally welded steel bar structure within the inner wall of a steel tube to enhance the bond properties at the connection interface. The influence of various configurations of steel bars welded to the inner surface of the tube on the bond strength is investigated considering the impact of vibration on the load-bearing capacity of the component. This study comprises two groups of specimens, one with vibration and one without vibration, for a total of ten specimens. Each group included CFST members with five distinct internal welded steel bar structures. The experimental results, including load–displacement curves and strain data of the steel tube, were used to assess the impact of the internal welded steel bar configurations on the steel–concrete interface. The sliding process is described by correlating test data with curves and observed phenomena. To comprehensively compare the effects of structural dimensions on the bonding and slipping properties of the welded bars, finite element simulations replicating the experimental conditions were carried out using ABAQUS software, and the simulation results agreed with the experimental observations. The study demonstrated that incorporating internal welded steel bars substantially enhances the bond strength of steel pipe–concrete interfaces. While vibration weakens the bond strength in CFST members, internal welded steel bars mitigate this effect. These findings improve the structural performance of CFST structures and their resilience to external vibrations.

Composite structures of concrete-filled steel tubes (CFSTs) can improve the overall mechanical properties of a component. A steel tube has a restraining effect on the core concrete, enhancing the compressive strength and ductility of the component. Moreover, the core concrete improves the overall yield strength of the component. Research has shown that the "composite effect" of combining steel tubes and concrete in CFST members results in significantly improved strength and ductility compared to those of individual components. Additionally, the overall ultimate load-bearing capacity of CFST members even surpasses the sum of the strengths of two materials1,2. Furthermore, in the construction process of CFST structures, formwork is unnecessary during concrete pouring, leading to significant savings in staffing, resources, and cost.

Notably, this composite structure also exhibits excellent bearing capacity. During the concrete pouring process, minimal staff are needed, which results in cost efficiency, while the tubular structure eliminates the need for formwork, and the finished structure has better collaborative work ability. Moreover, this composite structure has excellent mechanical properties. Due to these advantages, CFST bridges are widely used in civil engineering applications, particularly for large-span arch bridges, sea crossing bridges, and port terminals3,4,5.

To ensure the good mechanical properties of a CFST, coordinated interactions between the steel tube and concrete are necessary. Therefore, the bond strength between the two has played a critical role during the past few decades. Several researchers have studied the bond properties between steel tubes and infill concrete in CFSTs. For instance, Roeder et al.2. highlighted the application of CFSTs in construction and the importance of the bond strength and interface conditions on the overall performance and proposed an equation to estimate the bond stress. Chang et al.6. conducted experimental research on the expansion and contraction properties and bond strength of expanded cement CFSTs and traditional CFSTs. The results indicated that the use of expansive concrete is an effective method for enhancing the overall bonding strength. Aly et al.7. conducted a series of push-out tests on the strength and age of concrete and the loading scheme, and the results showed that the age of regular concrete causes a slight decrease in bond strength; however, conclusions regarding high-strength concrete were not drawn. Qu et al.8. conducted load reversal push-out tests on six rectangular CFST columns and found that macrobonding was the primary mechanism contributing to bond strength, followed by microbonding, and chemical bonding had little effect.

In recent years, it has become common practice to install structural devices on the inner wall of a tube to enhance the bond strength of a CFST9,10,11,12. Tao et al.13. Conducted a series of push-out tests on circular and rectangular CFSTs, and the results showed that the bond strength of stainless steel CFST columns was greater than that of carbon steel columns. The bond strength decreased with increasing age and section size. It has been suggested that internal welding of steel bars is an effective measure for improving bond strength, followed by welding shear nails and the use of an expansion tube. As the age of the concrete and the cross-sectional dimensions increase, the bonding strength decreases. Kai et al.14. conducted a series of push-out tests on the roughness of the inner wall of steel tubules with welded longitudinal ribs and setting pegs, and the results showed that the rougher the inner wall was, the stronger the bond performance was. The welded longitudinal ribs can significantly increase the bond strength, which is directly proportional to the length of the ribs; the pegs are also directly proportional to the length of the ribs; the bond strength of the pegs is directly proportional to the length of the ribs; and the number of pegs has a particular effect on the bond strength. Chen et al.15. Proposed that the built-in pattern in a steel tube can significantly improve the bond strength between the steel tube and concrete. Dong et al.16. Noted that the bond-slip curve of a CFST without additional devices was composed of an ascending section, a rapid descending section, and a residual section. Moreover, the most effective built-in measures were welded round ribs and vertical ribs.

In practical applications, several factors can decrease the bonding strength of a CFST, such as debonding and void formation. However, the influence of vibrations is also a critical factor that must be considered when applying concrete-filled steel arch bridge projects. The SEG Plaza in Shenzhen experienced significant internal detachment and deterioration of the damping ratio 20 years after its construction, which shocked domestic and international spectators. This contributed to the building becoming a vibration-sensitive structure, resulting in sudden vibrations. Researchers worldwide have conducted extensive experimental studies and finite element simulations to investigate the hysteresis characteristics and ductility of concrete-filled steel members. Moreover, moment–curvature hysteresis and axial force‒displacement restoring force models have been developed for compression–bending members, laying a theoretical foundation for dynamic calculations of concrete-filled steel structures17,18,19,20,21. Research on seismic performance based on CFSTs has primarily been performed by shaking table tests, which are mainly aimed at concrete-filled steel tubes in construction applications22,23,24,25,26. However, the influence of vibration on the bond-slip performance of CFST members has rarely been studied.

The main research objective of this study is to investigate the enhanced bonding strength between a steel tube and core concrete by using different forms of internally welded steel bars. The impact of vibrations on the bonding and slip behaviour of the components was also examined. Experiments were conducted using a push-out test method on two groups of ten CFST specimens. One group of specimens was subjected to vibrations, while the other group was not and, therefore, served as a comparison. The objective was to analyse the detailed distribution of bonding stresses during loading. A summary was made regarding the resistance of concrete-filled steel tube specimens with different forms of internal welded steel bars to the effects of vibrations by analysing and comparing the experimental results. This study provides valuable references for future practical engineering applications and scientific research in this field.

In the experimental program, there were two groups of specimens. Each group consisted of 5 CFSTs with five different shapes of inner welded steel bars, for a total of 10 CFSTs. One group was selected as the vibration control group, and push-out tests were carried out after the vibration. All tested steel tubes had an inner diameter of 370 mm, a wall thickness of 10 mm, and a height of 550 mm. The material of the steel tubes was Q235 steel. C60 self-compacting compensatory concrete was taken from the concrete used for casting the arch rib of an arch bridge in Guangxi, and its coordination is shown in Table 1. The diameter of the internally welded reinforcing bars was 10 mm. The material used for the internally welded reinforcing bars was HPB300. During pouring, an empty recess of 50 mm was reserved at the bottom of the tube so that the bond length of the connection interface in the tube was 500 mm. The parameters of the other specimens are shown in Table 2. Figures 1 and 2 include the design profile and actual photos of the components, respectively. Table 3 shows the mechanical properties of the materials used in accordance with the Chinese specification GB/T 228.1-201027.

Sectional view of the component design (all dimensions in mm): (a) R0, (b) R1, (c) R2, (d) R3, (e) RS, and (f) top view.

Photographs of the steel tube specimens (from left to right: R0, R1, R2, R3, and RS).

The vibration operation began after concrete curing was completed. Five CFSTs were used as the experimental group for vibration control after 28 days of concrete curing. The vibrator was firmly fixed to a steel plate prewelded to the front of each steel tube using bolts. A 1C101 acceleration sensor was installed at the centre of the rear side of the steel tube. The device was connected to the DH dynamic signal acquisition instrument and a laptop computer to detect the vibration frequency of the component. The power of the vibrator was 0.75 kW, and its frequency was 50 Hz. The aim is to ensure that the amplitude of each component is uniform and that each tube undergoes 30 min of vibration. After the vibration, a push-out test was conducted for comparison with the non-vibration group. Figures 3 and 4 show the arrangement of the vibrator and accelerometer, respectively.

Setup of the vibrator.

Setup of the acceleration sensor.

The push-out tests of the CFST were carried out on a kiloton press test bench at the College of Civil Engineering and Architecture of Guangxi University. The reserved end of the specimen was positioned on the bottom cushion block, secured by the wood block, and stripped at the bottom of the core concrete. The upper end of the wood block was firmly attached to the core concrete, while the lower end was fixed along the wood strip. This enabled the wood strip to move along with the bottom of the core concrete, facilitating the measurement of the slip. A circular steel block, slightly smaller than the diameter of the core concrete, was positioned on top of the specimen, and the loading was applied from above. The upper end of the test block served as the loading end, while the bottom end was the free end. The overall schematic diagram is shown in Fig. 5.

(a) Layout of the test arrangement and (b) 3D schematic of the test arrangement.

Before commencing the test, the loading end was controlled to contact the specimen uniformly and to observe the output value of the test console. The loading and free ends of the system were fully in contact with the specimen to eliminate the effect of virtual displacement. Before commencing the test, the position of each displacement gauge was adjusted, and the readings of the strain and displacement gauges were reset. The loading was carried out in a displacement-controlled manner at a constant rate of 0.5 mm/min, and the process was stopped once the displacement reached 40 mm.

The main test measurements included the strain of the outer wall of the steel tube, the slip between the tube and the core concrete, and the displacement of the loading end and opposite end of the specimen. The loading displacement and the push-out force were read and controlled by the system software.

The test setup included four displacement transducers, two of which were linear variable differential transformers (LVDTs) placed at the loading end and two of which were wire-based rotary variable differential transformers (RVDTs) that convert angular deformations into linear displacements. The RVDTs were connected with the aid of pulleys, wires, and a wire encoder and attached to a wooden strip that moves identically to the core concrete. The real-time slip of the core concrete was obtained by measuring the downwards displacement of the wooden strip. During installation, perpendicular alignment between the wires and the wooden strip was ensured to minimize errors. A detailed diagram of the displacement meter arrangement metre is shown in Fig. 5.

The nine points used for measuring the strains on the outer wall of the steel tube were positioned in three rows, each including three points. The locations of the strain gauges, shown in Fig. 6, were carefully sanded to remove rust, then polished with sandpaper, and finally cleaned by wiping with alcohol. The lengths of the strain gauges were approximately 30 mm, and the gauges were pasted at the appropriate locations and protected with silicone gel. The wireless static strain measuring system DH3819 was used for data acquisition.

Layout of the measurement points: (a) front view and (b) top view.

Seven of the ten specimens were tested following the procedure explained in Section "Loading scheme", whereas the three remaining specimens were loaded further to full failure. During the initial loading phase, the specimens produced the piercing sound of friction; this indicated that the upper part of the concrete experiencing compression was rubbing against the inner wall of the steel tube. Slipping was primarily controlled by the mechanical connection behaviour, with additional contributions due to bonding and interlocking forces. The compression in the concrete increased with increasing loading displacement. A cracking sound was heard at the moment of failure in the weld that connected the steel bar to the steel tube, and from this point forwards, the bonding was primarily based on the mechanical interlocking between the intact welds and friction between the core concrete and inner wall of the tube. The gradual fracture of the joint welds ultimately decreased the mechanical connections out of work, and the slipping of the concrete core increased linearly with loading. After the cracking sounds ceased, the piercing sound of friction became prominently louder. This indicates that all of the welds had detached and that the bond was primarily based on the friction among the core concrete, internal steel bar, and internal wall of the steel tube. From this point forwards, the slipping increased rapidly until the loading process was terminated.

The lower end of the external steel tube in specimen S-R3 of the vibrated group and in specimens RS and S-RS of the two groups was destroyed when the loading displacement reached 10 mm, corresponding to a load greater than 3200 kN. During these tests, the sound of cracking was not heard. This is attributed to a greater number of welded joints preventing weld fractures and increasing the mechanical interlocking force. Due to the absence of infill concrete in the 50 mm bottom part of the steel tube, the material could not resist loading without failure, as shown in Fig. 7.

Failure mode of the steel tube.

After the end of the test, visible observations from the loading end revealed that the core concrete had been pushed out and that there were noticeable scratches on the inner wall of the steel tube. The specimens with an inner welded steel bar structure show significantly more debris than those without a steel bar structure. Figure 8 depicts the overall damage at the loading end of the non-vibration group, and the vibration group exhibits similar damage patterns.

Damage patterns at the loading ends of the specimens in the nonvibrated group for (a) R0, (b) R1, (c) R2, (d) R3, and (e) RS.

After the end of the test, both sets of test components were cut open, and their profiles were observed. Figure 9 depicts a cross-section of each component.

Specimen splits include: (a) R0, (b) R1, (c) R2, (d) R3, (e) RS, (f) S-R0, (g) S-R1, (h) S-R2, (i) S-R3, and (j) S-RS.

The measured push-out lengths for the nonvibrated and vibrated groups of specimens are presented in Table 4. The push-out length decreases as the number of internally welded steel rings increases. The push-out lengths in the specimens of the vibrated group tend to be greater than those in the nonvibrated group, but the trend is smaller when the number of rebar rings is greater. This suggests that vibration can influence the bond-slip properties of the specimen and that the use of internal welded rebar can improve the resistance against vibration.

The load vs. displacement curves for each specimen in the vibrated group and in the nonvibrated group are presented separately as P–S diagrams in Fig. 10, and the curve of each specimen can be divided into ascending, descending, and residual parts remaining after the decline.

Load‒displacement or P‒S diagrams for specimens (a) R0, (b) R1, (c) R2, (d) R3, and (e) RS.

Three characteristic points appear in each P–S diagram. The initial point (Ss, Ps) marks the starting point for the rapid development of slipping, and the peak point (Su, Pu) indicates the appearance of the maximum load in the slipping process. The residual point (Sr, Pr) signifies the endpoint for the descending part. In Table 5, α1 = Ps/Pu, and α2 = Pr/Pu. In the three specimens with a damaged steel tube, the tube yielded after reaching the peak point, and the descending and residual parts did not follow the principles of the P–S diagram in other specimens; moreover, the residual point cannot be set for comparison.

According to all the P–S diagrams in Fig. 9, the bond load increases steadily within the ascending part of the curve, and once the peak load has been reached, some of the bond components are gradually deactivated, and the curve enters the descending part. After the end of the descending part, the curve shows a flat and slightly increasing trend, except in the specimens with the yielded steel tube.

The average bond stress τ is the load P divided by the contact area A between the steel tube and infill concrete. The bond strengths τs, τu and τr correspond to points (Ss, Ps), (Su, Pu), and (Sr, Pr), respectively, in the diagrams. The energy dissipation of various components in the specimen also serves as an essential indicator of efficiency. The dissipation value W is the energy absorbed per unit volume of the specimen to produce a particular value for the slip displacement. It can be calculated as the area under the load–slip diagram of the specimen as follows28:

where V is the volume of the specimen, S is the slip displacement and the load P is a function of S.

In the residual phase of the behaviour, because most of the primary bonding forces no longer contribute to resisting the load, it is important to analyse the energy consumption values in both the ascending and descending phases. Since all the specimens reach the residual phase after passing a displacement of 15 mm, the energy consumption value for the displacement of the first 15 mm is used as the representative value when comparing the energy consumption during the slipping process. The average bond strength and energy consumption capacity of each specimen are presented in Table 6.

According to Fig. 9, the P–S diagrams of the specimens exhibit a consistent linear behaviour and manner of change, except for the three specimens with a damaged steel tube. The distinct peak points indicate the transformation from one phase to the next. For specimens R0 and S-R0 without any mechanical bonding devices, the initial loading activates a chemical bond force, a mechanical interlocking force, and frictional resistance as components of the bonding force. With increasing loading, a slight displacement appears in the core concrete to reduce the bond strength, and the concrete also experiences longitudinal compression at the same time as the transverse expansive extrusion of the wall of the steel tube, which increases the friction and mechanical interlocking with the tube. When the free end of the concrete infill experiences displacement, chemical bonding is no longer present, mechanical interlocking and frictional resistance are insufficient to counteract the shear force of the connection interface, and a sudden decrease in the bearing capacity occurs. A reduction in the load and continuous damage to the concrete at the connection interface reduce the frictional and mechanical gripping forces. After a certain amount of slipping, the lateral expansion of the concrete gradually equilibrates the friction force, mechanical gripping, and connection shear force. In the specimens with welded steel bars, the mechanical gripping forces are considerably greater than those in the specimens without welded steel bars.

The importance of the descending part of the curve for the members with internal welded reinforcement is more pronounced. After this part of the curve, a gradual increase in the load occurs. The peak point indicates that there is a mechanical gripping force between the tube and welds on the ring steel bar. Additionally, when the connecting welds are damaged, the mechanical part of the total connection force disappears, and only the frictional share of the reinforcement remains.

The internal welded reinforcement significantly improved the bond strength, the maximum bearing capacity of the R1 specimens was more than three times that of the R0 specimens, and the energy consumption doubled. As the number of reinforcement rings increases, the bond strength and energy consumption capacity improve gradually. The bond behaviour of the R3 specimens is better than that of the R2 members, which is better than that of the R1 members. Although the steel tubes in specimens RS and S-RS, which are the two specimens with spiral reinforcement, were damaged, an increase in the maximum load-bearing capacity occurred, and no apparent slipping in the concrete could be found at the end of the test. Therefore, spiral connection reinforcement is the most superior form of reinforcement.

When comparing the results between the vibrated group and the nonvibrated group, the most apparent change is observed in specimen S-R1. Compared with that in R1, the peak in the curve significantly decreases by 311 kN. In S-R2, S-R3, and S-RS, the peak points are also slightly reduced after vibration. The peak point in the curve of the S-R0 specimen is not affected by vibration and is 10 kN greater than that in R0; however, the residual part of the curve after vibration is significantly lower than that of the nonvibrated specimen, and the peak point reaches that of the nonvibrated specimen. The residual part of the curve of the vibrated group is lower than that of the nonvibrated group when reaching the residual point, and on average, the bond strength in the vibrated group is slightly lower than that in the nonvibrated group. Similarly, in the S-R0 and S-R1 specimens, there is a reduction in energy dissipation. This indicates that vibration can decrease the bond properties of the members somewhat, especially in the residual part of the curve.

To investigate the development law of longitudinal interfacial interactions in members with different internal structures, longitudinal strain gauges were arranged on the outer wall of the steel tube, and the measuring points are shown in Fig. 6. The diagrams of the longitudinal strains in the tests of the vibrated group and the nonvibrated group are presented in Fig. 11.

Diagrams of the longitudinal strains measured on the outer wall of the steel tube of the specimens: (a) vibrated group and (b) nonvibrated group.

The diagrams of the longitudinal strains of the outer wall of the steel tube in both groups follow similar patterns. The strain in each specimen increases with increasing distance from the loaded end. Furthermore, the changes in the strain at each measuring point are relatively consistent, forming a triangular distribution with smaller values at the upper end and larger values at the lower end. This suggests that the interfacial force experienced by the combination-shaped steel tube is generally distributed uniformly from the loading end to the push-out end. With an increasing number of inner welded steel rings, the level of compression in the concrete layer by layer becomes superimposed, and the ring bars work cooperatively. Among these cases, the longitudinal strains in specimens RS, S-R3, and S-RS increase more significantly than those in the other members at a distance of more than 250 mm from the loaded end, which causes destructive deformations in the unfilled part of the steel tube. This indicates that the behaviour of the connection interface in each specimen is relatively uniform and that all the reinforcing rings work well together.

The engineering structures to be designed can only be analysed correctly by using proper constitutive material models in the FEM software ABAQUS. Additionally, the CFSTs discussed in this paper contain two materials that should be given different constitutive models and further contact relationships when simulating their connections.

The ABAQUS framework mainly provides concrete damaged plasticity and smeared crack concrete models. The plastic damage model can reflect the stiffness degradation and unilateral effect of concrete under loading by setting parameters that simulate the crushing mechanism and irreparable damage of core concrete. This paper uses the plastic damage model for concrete in the finite element simulation. The elastic modulus of the concrete, Ec, is calculated according to ACI31829:

where \(f_{{\text{c}}}{\prime}\) is the compressive cylinder strength in MPa, and Poisson's ratio is equal to 0.2.

The stress‒strain relationship for the compressed core concrete, accounting for the restraining effect of the steel tube on the core concrete as a coefficient \(\xi\), is calculated according to Han Linhai30,31 as follows:

where

where \(\varepsilon_{c}\) is the compressive strain corresponding to the stress \(\sigma_{c}\) of the concrete, \(\varepsilon_{c0}\) is the strain corresponding to the peak stress \(\sigma_{c0}\) of the concrete, and \(\xi\) is the coefficient for the constraint effect given as:

where \(f_{y}\) is the yield strength of the tube material, \(A_{s}\) is the cross-sectional area of the tube, \(f_{cu}\) is the axial compressive strength of the core concrete, and \(A_{c}\) is the cross-sectional area of the core concrete.

The secondary plastic flow model describing the behaviour of the steel tubes and bars in this paper and the primary constitutive stress‒strain relationships are taken as follows:

where \(f_{y}\) is the yield limit of steel, \(E_{s}\) is the elastic modulus of steel = 206,000 MPa, Poisson's ratio is 0.3, \(\varepsilon_{e} = 0.8{{f_{y} } \mathord{\left/ {\vphantom {{f_{y} } {E_{s} }}} \right. \kern-0pt} {E_{s} }}\), \(\varepsilon_{e1} = 1.5\varepsilon_{e}\), \(\varepsilon_{e2} = 10\varepsilon_{e}\), \(\varepsilon_{e3} = 100\varepsilon_{e1}\), \(A = 0.2{{f_{y} } \mathord{\left/ {\vphantom {{f_{y} } {\left( {\varepsilon_{e1} - \varepsilon_{e} } \right)}}} \right. \kern-0pt} {\left( {\varepsilon_{e1} - \varepsilon_{e} } \right)}}^{2}\), \(B = 2A\varepsilon_{e1}\), and \(C = 0.8f_{y} + A\varepsilon_{e}^{2} - B\varepsilon_{e}\).

The basic stress‒strain modeling results for the steel materials are shown in Fig. 12.

Basic stress‒strain modelling for the steel materials.

In the calculation, the tie constraint is used for the welded joint between the steel tube and steel bar to simulate its interface contact. There is friction and cohesion on the contact surface of the steel tube and the concrete, and its friction contact model includes tangential Coulomb friction and normal hard contact. In this paper, the cohesive force model in ABAQUS was used, and a surface-based cohesive behaviour model that ignores the interface thickness was adopted. This is more in line with the definition of the actual situation and requires the setting of three-phase parameters, including the slope of the linear elasticity phase, i.e., the stiffness, the damage initiation criterion, and the damage evolution process. The initial linear elastic behaviour is described by the elastic constitutive matrix shown below:

where \(\sigma_{n}\) is the normal stress, \(\sigma_{s}\) and \(\sigma_{t}\) are the tangential stresses, \(\delta_{n}\) is the normal relative displacement, \(\delta_{s}\) and \(\delta_{t}\) are the tangential relative displacements, and K is the elastic stiffness.

In the tests described in this paper, the core concrete remained uncracked, and the maximum elastic stress criterion was used when assessing the damage criterion for the strength of the concrete. When the maximum value in Eq. (12) becomes unity, this criterion is fulfilled.

where \(\sigma_{n}\) represents the normal stress; \(\sigma_{s}\) and \(\sigma_{t}\) represent the tangential stresses; and \(\sigma_{n}^{0}\), \(\sigma_{s}^{0}\) and \(\sigma_{t}^{0}\) represent the maximum elastic stress at the point of contact during the elastic behaviour. This state requires the input of three constant covariants \(\sigma_{n}^{0}\), \(\sigma_{s}^{0}\) and \(\sigma_{t}^{0}\), which correspond to the three variables \(\sigma_{n}\), \(\sigma_{s}\) and \(\sigma_{t}\).

A set of five components was established via ABAQUS software; these components had internally welded steel bars of various forms. The specific dimensions of each component used in the simulation process were identical in each case considered. Four-node shell elements S4R were used for the steel tube, the core concrete was established by eight-node 3D solid elements C3D8, two-node linear 3D truss elements T3D2 were used for the steel reinforcement, and the end plates were coupled to the reference points at the ends of the specimen by setting rigid body constraints in the interaction module. The models are shown in Fig. 13, and the meshing models are shown in Fig. 14. Displacement loading was also adopted up to 40 mm, similar to the tests.

FEM of the R0, R1, R2, R3 and RS tests.

Meshing of the specimen (a) the specimen as a whole and (b) the wall of the steel tube.

The diagrams of the bond stress in the inner wall of the tube for the group of five specimens are presented in Fig. 15 at the moment of displacement loading, which is equal to 10 mm for each specimen. The results reveal that the bond stress value of R0 is the smallest among all the specimens and is much smaller than that of the R1 member. This indicates that the internally welded reinforcements significantly improved the bond strength of the members during the loading process, which is consistent with the test results. As the number of internally welded reinforcement rings increases, the cohesive stress in the inner wall of the steel tube decreases slightly due to the increased pressure on the internal reinforcement rings. The cohesive stress of the RS member is the smallest and is slightly lower than that of R3. This indicates that the construction of spiral reinforcement in the inner wall can increase the interfacial load-bearing capacity the most. Additionally, the change in reinforcement construction leads to a slightly different distribution of interfacial stress compared to that of the other components. Specifically, the RS member exhibits maximum interfacial cohesive stress at the loading end, reducing the interfacial stress at the free end. Consequently, this stress distribution enhancement helps improve the overall structural performance.

Bond stress distributions in the steel tube: (a) R0, (b) R1, (c) R2, (d) R3, and (e) RS.

The analysis of the contact interface bond stress reveals that the implementation of internal welded reinforcement greatly enhances the interfacial load-carrying capacity of a CFST. Furthermore, as the number of reinforcement rings increases, the interfacial load-carrying capacity becomes even more vital. Among the designs, the section with spiral reinforcement exhibits the highest load-carrying capacity and effectively optimizes the stress distribution at the interface.

According to the simulation results processed with ABAQUS software, the P‒S curves of each member were plotted for comparison with the test results (Fig. 16). The diagrams show that the curves derived from the simulation results are consistent with those from the tests and that the overall behaviour indicates a similar trend, including the ascending, descending, and residual phases. In the simulation of the RS member, after the displacement of the loading end reaches 10 mm, the steel tube at the lower end starts to yield, which agrees with the test results.

Comparison of the simulated and experimental diagrams.

Furthermore, in the actual test, the steel pipe wall at the reserved end of the RS member failed, preventing us from obtaining the actual bond-slip curve. To address this issue in the simulation, the thickness of the steel pipe was increased to improve its yield strength. Specifically, the wall thickness of the steel pipe was increased from 10 to 14 mm. The P–S curve of the unyielding steel pipe wall is shown in Fig. 16. According to a comparison between Figs. 16 and 17, the member constructed with internal spiral reinforcement has the best interface load-carrying capacity. The simulation results of the whole specimens agree with the test results, demonstrating the accuracy of the simulation method.

RS components with increased wall thickness.

This study focused on enhancing the bond strength between the core concrete and steel tube by introducing various forms of internally welded reinforcements on the inner wall of steel tubes and investigating the effect of vibration on the bond strength of individual members. Two groups of ten members with different internally welded reinforcement structures were set up, and push-out tests were conducted. By setting the parameters, the interfacial load‒carrying capacities of the different member forms were analysed and compared, and the results for the nonvibration group were compared with those for the vibration group. Moreover, finite element simulations were also carried out to supplement and verify the results, and the following conclusions were obtained:

Internally welded steel reinforcement can significantly improve the interface load-bearing capacity and energy consumption capacity of concrete-filled steel tubes (CFSTs).

The combination of simulation and experimental results shows that the bond strength of the members gradually increases with an increasing number of reinforcement rings, and the member RS with the established spiral reinforcement structure demonstrates the best cross-sectional load-carrying capacity. Compared with member R0 without an internal welded reinforcement structure, the maximum load-carrying capacity is increased by six times, and the average bond strength and energy consumption capacity are also increased by 4–6 times.

The test results show that vibration reduces the bond strength of all types of members, which is more evident in the residual part of the P–S curve. When analysed in conjunction with the final push-out lengths of the different members, the design measures of the internal welded reinforcement also reduce the impact of vibration on the components.

A finite element simulation is carried out to observe the cohesive stress at the interface, indicating that the use of internally welded reinforcement bars enhances the bond strength of the members and that the construction of RS specimens is optimal and can improve the stress distribution at the interface. The agreement between the simulation results and the test value curves verifies the reliability of the simulation method.

The research in this paper provides valuable reference suggestions for the practical engineering application of steel pipe concrete, especially in concrete-filled steel tube arch bridges and buildings that require seismic capacity. Measuring internally welded steel reinforcement substantially improves the interfacial bearing capacity of a CFST, reduces the occurrence of void defects, attenuates the influence of vibration, and improves the mechanical and seismic performance of the members in actual engineering. In the future, the formula and optimization of the interface properties of welded steel pipes and filled concrete will be studied.

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Han, L. H., Li, W. & Bjorhovde, R. Developments and advanced applications of concrete-filled steel tubular (CFST) structures: Members. J. Constr. Steel Res. 100, 211–228. https://doi.org/10.1016/j.jcsr.2014.04.016 (2014).

Article Google Scholar

Roeder, C. W., Cameron, B. & Brown, C. B. Composite action in concrete filled tubes. Struct. Eng. 125, 477–484 (1999).

Article Google Scholar

Song G. Study on Restraint Effect and Axial Compression Behavior of Large Diameter Concrete Filled Steel Tube, Xihua Univ., Chengdu, (2020).

Deng L. Study on Ultimate Bearing Capacity of Large Diameter Concrete-Filled Steel Tube Members, Chongqing Jiaotong University, Chongqing, (2021).

Deng, N. C. et al. Bond-slip performances of ultra-high performance concrete steel tube columns made of a large-diameter steel tube with internally welded steel bars. Materials https://doi.org/10.3390/ma16103836 (2023).

Article PubMed PubMed Central Google Scholar

Xu, C. et al. Push-out test of pre-stressing concrete filled circular steel tube columns by means of expansive cement. Construc. Build. Mater. 23(1), 491–497. https://doi.org/10.1016/j.conbuildmat.2007.10.021 (2009).

Article Google Scholar

Aly, T. et al. Incremental collapse threshold for pushout resistance of circular concrete filled steel tubular columns. J. Construct. Steel Res. 66(1), 11–18. https://doi.org/10.1016/j.jcsr.2009.08.002 (2010).

Article Google Scholar

Qu, X., Chen, Z., Nethercot, D. A., Gardner, L. & Theofanous, M. Load-reversed push-out tests on rectangular CFST columns. J. Construct. Steel Res. 81, 35–43. https://doi.org/10.1016/j.jcsr.2012.11.003 (2013).

Article Google Scholar

Jiang, S. F., Han, L. H. & Qiao, J. C. Initial discussion on the bond problems of steel and concrete in concrete filled steel tubles (CFST). J. Harbin Univ. C E. Architecture 02, 24–28 (2000).

Google Scholar

Yang, Y. F. & Han, L. H. Research on bond behavior steel and concrete of selfcompacting concrete filled steel tubes with rectangular sections. Indust. Constru. 11, 32–36 (2006).

Google Scholar

Liu, Y. J. et al. Interfacial bond-slip properties of PBL stiffened rectangular steel pipe concrete. J. Build. Sci. Eng. 32(05), 1–7 (2015).

ADS Google Scholar

de Fátima, M. et al. Study of bolts used as shear connectors in concrete-filled steel tubes. Eng. Struct. 231, 111697. https://doi.org/10.1016/j.engstruct.2020.111697 (2021).

Article Google Scholar

Tao, Z. et al. Bond behavior in concrete-filled steel tubes. J. Construct. Steel Res. https://doi.org/10.1016/j.jcsr.2015.12.030 (2016).

Article Google Scholar

Xu, K. C., Huang, C. L. & Chen, M. C. Research on impact of steel tube inner surface roughness on interfacial bonding behavior of CFST. Build. Struct 34, 420–424 (2013).

Google Scholar

Chen, L. H., Dai, J. X. & Jin, Q. L. Push-out test study on the bond strength between square checkered steel tube and concrete. Build. Struct 45, 17–20+26 (2015).

Google Scholar

Dong, H. et al. Bond-slip behavior of large high-strength concrete-filled circular steel tubes with different constructions. Constr. Steel Res. 167, 105951 (2020).

Article Google Scholar

Guo YH. Seismic performance and damping vibration table test of a submerged steel-tube concrete arch bridge, Xi'an University of Architecture and Technology, Xian, (2010).

Xiong F, Research on seismic performance of steel-tube concrete arch bridge. Sichuan University, Sichuan, (2002).

Yang, Y. F., Han, L. H. & Fan, X. Z. Current status of research on dynamic properties of steel pipe concrete. J. Harbin Archit. Univ. Haerbin 05, 40–46 (2005).

Google Scholar

Lv, X. L. & Lu, W. D. Experimental study on seismic performance of steel-tube concrete columns under repeated loads. J. Build. Struct. https://doi.org/10.1406/j.jzjgxb.2000.02.001 (2000).

Article Google Scholar

Yang, S. C., Wang, F. M. & Qu, P. Effect of core concrete devoiding on mechanical properties of steel pipe concrete members. J. Chongqing Jiaotong Univ. (Natural Science Edition) 03, 360–365+486 (2000).

Google Scholar

Zhang, D. et al. Seismic performance of bridge with unbonded posttensioned self-centering segmented concrete-filled steel-tube columns: An underwater shaking table test. Soil Dyn. Earthq. Eng. 138, 106350. https://doi.org/10.1016/j.soildyn.2020.106350 (2020).

Article Google Scholar

Yue, C. Z. & Zheng, Y. L. Shaking table test study on seismic behavior of underground structure with intermediate columns enhanced by concrete-filled steel tube (CFT). Soil Dyn. Earthq. Eng. 127, 105838. https://doi.org/10.1016/j.soildyn.2019.105838 (2019).

Article Google Scholar

Dong, H. Y. et al. Shaking table tests on concrete-filled steel tubular-framed building assembled with microcrystalline foam boards. Structures 34, 2098–2114 (2021).

Article Google Scholar

Sun, Z. Y. et al. Shaking table test of concrete columns hybrid reinforced by steel/FRP bars. J. Build. Eng. 48, 103938 (2022).

Article Google Scholar

Hu YN, Research on vibration characteristics and damage localization of steel pipe concrete debonding. Chongqing University, Chongqing, (2020).

GB/T228.1–2010; Metallic Materials-Tensile Testing—Part 1: Test Method at Room Temperature. China Standard Press: Beijing, China, (2010).

Chen, Z. P., Li, J. C. & Zhou, J. Axial compressive behavior of built-in carbon-steel marine concrete-filled stainless steel tubular columns. Constr. Steel Res 202, 107796 (2023).

Article Google Scholar

ACI 318. Building code requirements for structural concrete. Detroit (USA): American Concrete Institute, (2019).

Han, L. H. Steel pipe concrete structures-theory and practice (Third Edition) 26–35 (Science Press, 2016).

Google Scholar

Han, L. H., Ghyb, C. & Zhong, T. B. Performance of concrete-filled thin-walled steel tubes under pure torsion. Thin-Walled Struct. 45(1), 24–36 (2007).

Article Google Scholar

Download references

The authors thank the National Natural Science Foundation of China (Grant no. 52268048) and the Guangxi Science and Technology Major Project of China (Gui Ke AA22068066).

College of Civil Engineering and Architecture, Guangxi University, Nanning, 530004, China

Nianchun Deng, Haoxu Li, Jingyao Ni, Xiuning Peng & Guohua lv

Key Laboratory of Disaster Prevention and Structural Safety of Ministry of Education, Guangxi University, Nanning, 530004, China

Nianchun Deng

Guangxi Road and Bridge Engineering Group Co., Ltd., Nanning, 530200, China

Zhaotao Chen

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

N. D.: Conceptualization, Funding acquisition, Formal analysis, Writing-original draft & editing. H. L.: Supervision, Validation, Methodology, Writing-Review & Editing. J. N.: Data curation, Investigation, Visualization, Writing-original draft. X. P.: Methodology, Formal analysis. G. l.v.: Data curation, Formal analysis. Z. C.: Data curation, Formal analysis.

Correspondence to Haoxu Li or Xiuning Peng.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Deng, N., Li, H., Ni, J. et al. Effect of vibration on the interface properties of welded steel joints and filled concrete in steel pipes. Sci Rep 14, 20391 (2024). https://doi.org/10.1038/s41598-024-68186-0

Download citation

Received: 19 April 2024

Accepted: 22 July 2024

Published: 02 September 2024

DOI: https://doi.org/10.1038/s41598-024-68186-0

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative