Axial behavior of precast segmental hollow bridge columns confined by grouted steel tubes

Abstract

To avoid the problems of the conventional precast segmental bridge columns (PSBCs) such as the concrete crushing at column ends, joint openings between segments, and torsional deformation under seismic attacks, a novel type of precast segmental bridge column confined by a grouted steel tube (TPSBC) was proposed, and herein the axial compressive behavior of the TPSBCs was investigated by testing of 14 specimens. The main parameters of the tested specimens included the diameter-thickness ratios of the steel tubes, the number of the precast segments, reinforcement arrangements in the precast hollow segments, initial prestress levels, sandwich concrete strengths, and loading forms. The effects of the parameters on the failure modes, loading bearing capacity, ductility, and steel strain development of the specimens were evaluated. The test results indicate that the proposed TPSBCs with proper configurations exhibited desirable compressive behavior with significantly improved load bearing capacity and deformability compared with the PSBCs. Based on the analysis of the confinement effects, analytical models were developed for predicting the axial bearing capacity of the TPSBCs under the different loading forms.

Keywords

Precast segmental bridge columns axial compressive behavior grouted tube prestress level loading forms

Introduction

In order to shorten the construction period, reduce the impact of the bridge construction process on traffic and the surrounding environment, and facilitate the construction of bridges under complex conditions such as sea and mountainous areas, the precast segment assembly method as shown in Figure 1(a) for accelerated bridge construction has attracted increasing attention in recent decades (Billington et al., 1999; Bu et al., 2016; Cai et al., 2021; Jia et al., 2020; Li et al., 2019; Mashal and Palermo, 2019; Ou et al., 2018). Meanwhile, to prevent the bridge columns from being demolished due to excessive residual deformation after an earthquake (Kawashima et al., 1998), unbonded post-tensioned prestressing tendons (UPPTs) applied to the precast segmental bridge columns (PSBCs) have been extensively studied to improve their self-centering ability (Chou and Chen, 2006; Guerrini et al., 2015; Hewes, 2007; Ichikawa et al., 2016; Marriott et al., 2009; Shen et al., 2021; Wang et al., 2018b; Zhang et al., 2021). It has been proved that the PSBC assembled through UPPTs can undergo a large lateral drift with small residual displacement. Concrete spalling or crushing, however, was observed at the base of the bottom segments or near the joints above the bottom segments (Cai et al., 2021; Chou and Chen, 2006; Hewes, 2007; Jia et al., 2020; Li et al., 2019). Joint openings could also be found above the bottom segments, especially when the longitudinal steel bars or outer steel tubes were not continuous at the joints (Bu et al., 2016; Chou and Chen, 2006; Chou et al., 2013; Ichikawa et al., 2016; Wang et al., 2018a, 2018b). The joint openings will result in the significant reduction of the lateral stiffness of the columns and the potential corrosion at the segmental joints. Moreover, previous studies also indicated that conventional PSBCs suffered significant torsion under multi-directional loading (Ichikawa et al., 2016; Jia et al., 2020; Li et al., 2019). The conventional segmental joints were insufficient to resist eccentric lateral force applied on the columns. Therefore, to extend the application of PSBCs to earthquake zones, it is of great importance to improve both the lateral stiffness of the segmental columns and the bearing strength of column bases.

Figure 1.

Sketches of the PSBC and the proposed TPSBC: (a) PSBC, and (b) TPSBC.

Considering the desirable behavior of concrete filled steel tube segments (Chou and Chen, 2006; Guerrini et al., 2015; Han et al., 2005; Shen et al., 2021; Xiao et al., 2005), a novel type of precast segmental bridge column confined by a grouted steel tube (TPSBC) was proposed. As shown in Figure 1(b), the core part of the TPSBC is the same as the conventional PSBC, and the grouted steel tube consists of an outer steel tube and cast-in-situ sandwich concrete. The outer steel tube does not extend into the footing. Energy dissipation (ED) steel bars (Jia et al., 2020; Shen et al., 2021; Wang et al., 2018b) or replaceable external energy dissipation devices (Chou and Chen, 2006; Guerrini et al., 2015; Marriott et al., 2009) can be further used to enhance the lateral stiffness and energy dissipation capacity of the TPSBCs. The proposed TPSBC is expected to have the following advantages. The segments will be prefabricated by centrifugal method with high production efficiency and better quality control (Zhao et al., 2019). The hollow section of the inner PSBC reduces the amount of concrete and the weight of the component. The grouted steel tube is employed to improve the integrity and torsional resistance of the PSBC under multi-directional loading. The grouted steel tube can also provide confinement to concrete at column ends and delay the concrete crushing there. The assembled inner PSBC can successfully carry the gravity load, and thus the construction of superstructures and grouted steel tube can be carried out separately after the assembly of the precast segments. The construction of the grouted steel tubes has little effects on the construction of superstructures.

Prior to the study of the seismic behavior of the TPSBCs, their axial behavior was to be studied to better understand the confinement effects of the grouted steel tubes on the segmental columns. 14 specimens were tested under axial loads to achieve the following objectives: (1) to evaluate the effects of the initial prestress levels, loading forms, outer steel tubes, the number of the precast segments, reinforcement arrangements of precast segments, and sandwich concrete strengths on the behavior of the axially loaded specimens; (2) to develop analytical models for predicting the ultimate compressive bearing capacities of the proposed TPSBCs under the different loading forms.

Experimental program

Test specimens and specimens construction

A total of 14 stub columns, including two conventional PSBC specimens and 12 TPSBC specimens, were designed and tested under axial loads. The PSBC specimens were scaled down based on a bridge pier prototype with a diameter of 1.22 m (Li et al., 2019). A geometry scale factor of 1/4 was selected for PSBC specimens considering the loading capacity of the testing machine and the feasibility of scaled model construction. The size effects of the stub columns have been studied by Jin et al. (2019), Zhao et al. (2020), Chen et al. (2020), and were not studied herein. The main parameters of the specimens included the diameter-thickness ratios of the outer steel tubes, the number of the precast segments, reinforcement arrangements in the precast segments, initial prestress levels and its application sequence, sandwich concrete strengths, and loading forms. The details of the tested specimens are shown in Table 1 and Figure 2. D_so, D_si, and t in Table 1 are the outer diameter, the inner diameter, and the wall thickness of the steel tube, respectively; n is the number of the precast segments. The specimen labels are illustrated in Figure 3. Taking specimen TP12-S3B2P1-T as an example, it was a TPSBC specimen with a bigger diameter-thickness ratio of the steel tube compared to other TPSBC specimens, and its three precast segments with both inner and outer reinforcements were assembled to an inner PSBC through unbonded post-tensioned prestressing bar (UPPB). For brevity, only the first part of the label was referenced in some places.

Table 1.

Details of the specimens.

Specimen labels	D_so (mm)	t (mm)	D_si(mm)	D_so/t	n	Outer steel bars	Inner steel bars	Spiral stirrups	P_o(kN)	$f_{c u, k}^{p}$ (MPa)	$f_{c u, k}^{c}$ (MPa)	Loading forms
P2–S1B2P1	—	—	—	—	1	6Φ10	6Φ10	Φ6@50	257	53.7	38.1	A
P3–S3B2P1	—	—	—	—	3	6Φ10	6Φ10	Φ6@50	254		38.1	A
TP1-S1B0P1	376	3.9	368.2	96	1	—	—	—	384		38.1	A
TP2-S1B2P1	376	3.9	368.2	96	1	6Φ10	6Φ10	Φ6@50	430		38.1	A
TP3-S3B0P0	376	3.9	368.2	96	3	—	—	—	0		38.1	A
TP4-S3B0P1	376	3.9	368.2	96	3	—	—	—	432		38.1	A
TP5-S3B0P1–B	376	3.9	368.2	96	3	—	—	—	453		38.1	B
TP6-S3B2P0	376	3.9	368.2	96	3	6Φ10	6Φ10	Φ6@50	0		38.1	A
TP7-S3B2P1	376	3.9	368.2	96	3	6Φ10	6Φ10	Φ6@50	526		38.1	A
TP8-S3B2P1–B	376	3.9	368.2	96	3	6Φ10	6Φ10	Φ6@50	482		38.1	B
TP9-S3B2P1–P	376	3.9	368.2	96	3	6Φ10	6Φ10	Φ6@50	501		38.1	A
TP10-S3B2P1–C	376	3.9	368.2	96	3	6Φ10	6Φ10	Φ6@50	452		65.5	A
TP11-S3B1P1	376	3.9	368.2	96	3	—	6Φ14	Φ6@50	457		38.1	A
TP12-S3B2P1-T	376	2.9	370.2	130	3	6Φ10	6Φ10	Φ6@50	474		38.1	A

Figure 2.

Details of the specimens: (a) typical PSBC specimen and (b) typical TPSBC specimen.

Figure 3.

Illustration of the specimen labels.

The specimens had the same nominal height of 1.2 m, and the slenderness ratios of the PSBCs and the TPSBCs were 4.0 and 3.2, respectively. The hollow ratio ψ_p (D_ci/D_co) of the precast segment was 0.3, and the hollow ratio ψ_c (D_ci/D_si) of the internal concrete was 0.24. D_co and D_ci are the outer diameter and the inner diameter of the precast hollow segment, respectively.

Specimens P2 and P3 were circular hollow reinforced concrete columns as shown in Figure 2(a). These two columns were not confined by the steel tubes for comparison purposes. The only difference between these two specimens was the number of the segments. For the proposed TPSBC, an outer steel tube with a diameter of 376 mm was employed to confine the precast concrete core. Under the condition of meeting the requirement of concrete construction, the void for the post-poured sandwich concrete was designed as small as possible to reduce the onsite labor and shorten the construction time. The outer steel tube in Figure 1(b) was placed on the footing and attached to the sandwich concrete and the precast segments through bond and friction at the interfaces. In this study, no welding rings or shear studs were welded onto the inner surface of the steel tube. Note that the use of shear studs is suggested for the columns with large diameters to ensure good long-time behavior. According to GB 50936-2014 (2014), for the circular steel tube with a diameter larger than 2.0 m or the rectangular steel tube with a cross-sectional width larger than 1.5 m, the use of shear studs is suggested. Except specimens TP1 and TP2 with one monolithic precast segment, the other TPSBC specimens had the precast concrete core consisting of three segments, as shown in Figure 2(b). Three types of reinforcement arrangements were used in the segments, as shown in Table 1 and Figure 3.

In the segmental columns, the prestress technique is usually used to improve the strength, lateral stiffness, and self-centering ability of the columns (Hewes, 2007; Zhang et al., 2021). Herein the effects of the unbond prestress on the axial behavior of the proposed TPSBC were evaluated. Specimens TP3 and TP6 with a label of P0 had no initial prestress applied to the columns, while all other specimens had the prestress about 10% of the simple superposed bearing capacity of the tested specimens (Bu et al., 2012; Cai et al., 2021). The detailed prestress (P₀) applied to each specimen is listed in Table 1. For each prestressed specimen, one UPPB with a diameter of 32 mm was located at the center of the cross section. The UPPB had a nominal yield strength of 1160 MPa, giving a tensile capacity of 932 kN. To investigate the effect of the prestress application sequence, the prestress of specimen TP9 was applied after the curing of the sandwich concrete, while the prestress of all other specimens under the loading form A was applied before the sandwich concrete casting.

The main construction procedures of the specimens are shown in Figure 4. Taking specimen TP7-S3B2P1 as an example, self-consolidating concrete was used to match the surfaces of the adjacent segments and adjust the verticality of the segments. The prestress was applied to the segments by tensioning the UPPB using a hollow hydraulic jack, as shown in Figure 4(e). In the process of applying the prestress, it was monitored by a calibrated load cell, and the stress-strain curves of each UPPB can be calibrated by four strain gauges affixed on the prestressing bar. Finally, a small amount of concrete was poured into the void between the steel tube and the inner PSBC. A TPSBC specimen containing an outer steel tube, sandwich concrete, and precast segments was formed.

Figure 4.

Construction of the specimens: (a) reinforcement arrangements and formwork of the segments, (b) concrete casting of the segments, (c) concrete curing of the segments, (d) segments assembly, (e) prestress applying, and (f) sandwich concrete curing.

Material properties

The longitudinal steel bars and the spiral stirrups used in the segments were hot-rolled ribbed bars (HRBs) and hot-rolled plain bars (HPBs), respectively. Two types of welded circular steel tubes were used. Three tensile coupons were taken from each type of steel bar and steel tube. The tested yield strength (f_y), ultimate strength (f_u), Young’s modulus (E_s), and percentage elongation after fracture (PEF) are presented in Table 2.

Table 2.

Properties of the steel.

Category	Size or wall thickness	f_y (MPa)	f_u (MPa)	E_s (GPa)	PEF (%)
HRB steel bars	Φ10	442.0	639.9	200	26.7
HRB steel bars	Φ14	414.2	628.8	200	24.3
HPB stirrup	Φ6	423.1	558.6	210	27.8
Steel tubes	2.9 mm	345.2	477.6	206	28.3
Steel tubes	3.9 mm	343.9	484.2	206	27.4

Commercial concrete was used for all segments and the sandwich concrete in all specimens except specimen TP10, the sandwich concrete of which was the self-consolidating concrete with higher strength. The cubic compressive strength of the segments concrete ( $f_{c u, k}^{p}$ ) and sandwich concrete ( $f_{c u, k}^{c}$ ) are summarized in Table 1. The prism axial compressive strengths of the concrete for segments, sandwich concrete, and self-consolidating concrete were taken as 34.6 MPa, 25.5Mpa, and 41.8 MPa in the following sections, respectively (GB 50010-2010, 2015).

Experimental methods

Two types of loading forms were studied. The loading form A shown in Table 1 represented that both the prestress and the axial load were only applied to the precast segments, which was consistent with the actual local compression of the top end of the TPSBC. The loading form B represented that both the prestress and axial load were applied to the full cross-section of the TPSBC to obtain its fully loaded bearing capacity.

The specimens were axially loaded by a 7000 kN multifunctional testing machine. For the specimen with initial prestress, the UPPB protruded at two ends of the specimen. Two hollow rigid steel stub columns were specially designed and placed at the two ends of the specimen to help the application of the axial load, as shown in Figure 5.

Figure 5.

Test setup and instrumentations.

Each specimen was loaded at a rate of 0.12 mm/min before the peak load, and the loading rate gradually increased to 0.6 mm/min after the peak load. The axial loads were synchronously recorded by the data acquisition system of the testing machine. Six linear variable differential transformers (LVDTs) shown in Figure 5 were arranged along the periphery of each specimen. Four of them (Group 1) were used to measure the axial deformation from the top of the specimen to the top of the rigid steel base, and the other two LVDTs (Group 2) were used to measure the deformation of the bottom rigid steel stub column. Vertical and hoop strain gauges were affixed at the circumferential quarters of the steel tubes at sections A, B, and C. Strain gauges were also affixed on the surfaces of the steel bars and stirrups to monitor the strain development of the reinforcements.

Test results and discussion

General behavior and failure modes

Conventional PSBC specimens

The final failure of the specimens P2 and P3 was triggered by the concrete crushing, as shown in Figure 6(a)∼(c). The concrete crushing of the three-segment specimen P3 developed along the whole height of the column, while the concrete crushing of the monolithic column P2 mainly occurred at the two ends of the specimen. The different failure modes were mainly induced by the discontinuity of the reinforcements and concrete at the segmental joints in specimen P3. Different from the solid sections, the hollow sections suffered concrete crushing at both the external and the internal surfaces, and the concrete crushing near the external surfaces of the specimens was more serious than that near the internal surfaces.

Figure 6.

Failure modes of the PSBC and TPSBC specimens under the loading form B: (a) P2–S1B2P1, (b) P3–S3B2P1, (c) inward bulges of the PSBCs, (d) TP5-S3B0P1–B, (e) internal surface of TP5, (f) shear failure of internal concrete in TP5, (g) TP8-S3B2P1–B, (h) internal surface of TP8, and (i) sandwich concrete crushing of TP8.

Proposed TPSBC specimens under loading form B

Specimens TP5 and TP8 were tested under the loading form B. Before the peak loads, there was no obvious change in the surfaces of the steel tubes. After the peak loads, the lateral deformations of the TPSBCs increased rapidly. For specimen TP5 without reinforcement, an inclined shear failure plane was finally formed and the inclined shear failure plane passed through the segments, sandwich concrete, and segment joint. The internal concrete was divided into two wedges, as shown in Figure 6(f). For specimen TP8 with both inner and outer reinforcements, the shear failure was successfully prevented and the steel tube locally buckled at the top segmental joint, indicating the serious concrete crushing occurred there. The failure modes of specimen TP8 are shown in Figure 6(g) ∼ (i).

Proposed TPSBC specimens under loading form A

The failure modes of the proposed TPSBC specimens under the loading form A are presented in Figure 7. For these specimens, the failure was triggered by the serious intrusion of the steel bearing plates into concrete core due to the local loading. The intrusion of the bearing plates resulted in a lot of cracks distributed along the radial direction in the sandwich concrete. At the column ends, the steel tubes bulged outward in a bud shape within the length of its diameter. The concrete damage near the internal surfaces of the hollow segments was usually more serious than that of the conventional PSBCs since the grouted steel tubes inhibited the outward expansion of the segments. Comparison of the failure processes of the specimens with different reinforcement types, it can be found the concrete crushing of the TPSBCs with both inner and outer reinforcements delayed, and the specimens exhibited slower strength deterioration than the TPSBCs without or with only inner reinforcement. It should be noted that the testing of specimen TP7 was interrupted at about 90% of its peak load due to some problems of the testing machine, and the specimen was reloaded after the machine was repaired.

Figure 7.

Failure modes of the TPSBC specimens under the loading form A: (a) TP1-S1B0P1, (b) internal surface of TP1, (c) interface of TP1, (d) TP2-S1B2P1, (e) TP3-S3B0P0, (f) TP4-S3B0P1, (g) TP6-S3B2P0, (h) end plane of TP6, (i) end part of sandwich concrete in TP6, (j) TP7-S3B2P1, (k) TP9-S3B2P1–P, (l) TP10-S3B2P1–C, (m) end part of sandwich concrete in TP10, (n) interface of TP10, (o) TP11-S3B1P1, (p) end plane of TP11, (q) segment concrete crushing of TP11, and (r) TP12-S3B2P1-T.

After removing the steel tubes, it can be found that, for all TPSBC specimens regardless of under the loading form A or B, the segments were well confined by the grouted steel tubes, and no visible separation between segments and sandwich concrete occurred. The proposed TPSBC specimens showed much more ductile failure than the conventional PSBC specimens did.

Effects of the parameters on the axial behavior

The axial load (N) versus average strain (ε) curves given in Figure 8 and the normalized N/N_ue-ε/ε_ue curves given in Figure 9 are used to evaluate the effects of the parameters on the axial compressive behavior of the specimens, where N_ue is the experimental peak load (i.e. experimental maximum bearing capacity) of the specimen, and ε_ue is the strain corresponding to the peak load. The average axial strain was obtained by dividing the total axial deformation (D) by the specimen length. The N_ue and ε_ue are shown in Table 3. In Table 3, the strength index expressed as SI=N_ue/N_uc is defined to evaluate the confinement efficiency of the steel tubes and the spiral stirrups (Su et al., 2021), where N_uc is the simple superposed bearing capacity obtained by:

N_{u c} = f_{y t} A_{s t} + f_{y b} A_{s b} + f_{c}^{p} A_{c p} + f_{c}^{c} A_{c c}

(1)

in which A_st, A_sb, A_cp, and A_cc are the cross-sectional areas of the steel tube, longitudinal steel bars, segment concrete, and sandwich concrete, respectively; f_yt and f_yb are the yield strength of the steel tube and steel bars, respectively;

f_{c}^{p}

and

f_{c}^{c}

are the axial compressive strength of the segment concrete and the sandwich concrete, respectively.

Figure 8.

Axial load versus average strain curves.

Figure 9.

N/N_ue-ε/ε_ue curves.

Table 3.

Test results.

Specimen labels	P₀ (kN)	N_y (kN)	ε_y (ε)	N_ue (kN)	ε_ue (ε)	N_f (kN)	ε_f (ε)	μ	N_uc (kN)	SI	N_u (kN)	N_ue/N_u
P2–S1B2P1	257	2507.4	0.279	3027.2	0.387	2573.1	0.576	2.06	2624.4	1.15	2624.4	1.15
P3–S3B2P1	254	2540.4	0.286	2845.2	0.424	2418.4	0.612	2.14	2624.4	1.08	2624.4	1.08
TP1-S1B0P1	384	4026.9	0.413	5264.6	0.710	4474.9	1.071	2.59	4703.8	1.12	4876.3	1.08
TP2-S1B2P1	430	5266.3	0.519	5759.1	0.808	5357.3	3.719	7.17	5103.6	1.13	5569.9	1.03
TP3-S3B0P0	0	4103.7	0.398	5174.2	0.655	4398.1	0.912	2.29	4703.8	1.10	4876.3	1.06
TP4-S3B0P1	432	3648.0	0.397	5417.9	0.749	4605.2	1.108	2.79	4703.8	1.15	4876.3	1.11
TP5-S3B0P1–B	453	5149.7	0.434	6281.5	0.644	5339.3	0.879	2.03	4703.8	1.34	5779.4	1.09
TP6-S3B2P0	0	5398.7	0.555	5947.0	0.950	5185.4	3.719	6.70	5103.6	1.17	5569.9	1.07
TP7-S3B2P1	526	5093.6	0.447	6120.8	0.833	5202.7	1.541	3.45	5103.6	1.20	5569.9	1.10
TP8-S3B2P1–B	482	5146.9	0.445	6473.1	0.711	5502.2	2.008	4.51	5103.6	1.27	6527.3	0.99
TP9-S3B2P1–P	501	5171.9	0.570	6033.1	0.926	5128.1	2.985	5.24	5103.6	1.18	5569.9	1.08
TP10-S3B2P1–C	452	5240.2	0.412	6366.0	0.680	6428.7	3.692	8.96	5686.7	1.12	6037.3	1.05
TP11-S3B1P1	457	4403.4	0.396	5803.8	0.706	4933.2	1.259	3.18	5071.0	1.14	5243.6	1.11
TP12-S3B2P1-T	474	4814.1	0.605	5598.4	1.115	4758.6	2.038	3.37	4738.9	1.18	4974.1	1.13

Note: For the PSBC specimens, the N_u is $N_{u}^{p}$ ; for the TPSBC specimens under the loading form A, the N_u is N_u,l.

As shown in Figure 8, the N-ε curves of all specimens were basically linear in the initial stage. Subsequently, with the concrete cracking of the PSBCs or the steel tubes yielding and concrete crushing of the TPSBCs, the curves began to deviate from the initial straight lines, and the specimens entered the elastic-plastic state. The bearing capacity of specimens P2 and P3 decreased rapidly after the peak loads, which attributed to the inward deformations of the concrete due to the hollow sections. Compared with the PSBC specimens, the maximum bearing capacity of the TPSBC specimens was significantly improved. However, the bearing capacity of the TPSBC specimens without reinforcement or with only inner reinforcement decreased rapidly after the peak loads, while the TPSBC specimens with both inner and outer reinforcements showed good ductile responses.

The ductility factor (μ) is used to quantify the deformation capacity of the specimens after their yield points, and its expression is given as:

μ = ε_{f} / ε_{y}

(2)

where ε_f and ε_y are the strains at the failure points and yield points, respectively.

The farthest point method proposed by Feng et al. (2017) is used to determine the yield point of a specimen. The failure point is defined as the point when the axial load drops to 85% of its peak load. If the bearing capacity of the specimen decreases moderately after its peak load, the point whose deformation reaches 45 mm is taken as the failure point of the specimen. The yield loads (N_y) and yield strains corresponding to the yield points, the failure loads (N_f) and failure strains corresponding to the failure points, and the μ of the specimens are summarized in Table 3.

Effects of the prestress levels

Table 3 shows the initial prestress (P₀) of the specimens determined by the measured strains of the UPPBs before axial loading. Figure 10 shows the deterioration of the prestress in several typical specimens. Because the free ends of the UPPBs can deformed freely without restriction due to the specially designed hollow rigid steel stub columns, it can be seen that the prestress on the specimens decreased with the increase of the applied loads and was unloaded to zero before the peak loads, indicating that the UPPBs had withdrawn from work. Therefore, the effects of the initial prestress and its application sequence on the axial compressive behavior of the specimens can be ignored. This fact can be evidenced by the comparisons shown in Table 3. As shown in Table 3, for the specimens without reinforcement (with both inner and outer reinforcements), the difference of the N_ue between the prestressed specimen TP4 (TP7) and the non-prestressed specimen TP3 (TP6) was less than 5%. The N_ue of specimens TP9 and TP7 with different prestress application sequence was basically the same. The N/N_ue-ε/ε_ue curves of specimens TP4 and TP3 without reinforcement were basically coincident. The N/N_ue-ε/ε_ue curves of specimens TP9 and TP6 with both inner and outer reinforcements were also similar, except that the bearing capacity of specimen TP7 may decrease rapidly after the peak load due to retest.

Figure 10.

Deterioration of the prestress.

Effects of the loading forms

According to the test results of specimens TP4 (TP7) and TP5 (TP8), it can be found that the N_ue of the TPSBCs under the loading form A was smaller than that of the TPSBCs under the loading form B in the case of the segments without reinforcement or with both inner and outer reinforcements. This result was attributed to the adverse local intrusion of the bearing plate, which resulted in the premature crushing of the concrete at the column ends. The comparison of the SI in Table 3 also indicated that specimens TP5 and TP8 had the SI values equal to 1.34 and 1.27, respectively, which were larger than those of other specimens under the loading form A. This result indicated that grouted steel tube provided more effective confinement to the internal concrete under the loading form B.

The ductility of the TPSBCs under the loading form A was generally better than that of the TPSBCs under the loading form B. As shown in Table 3, the μ of specimen TP4 under the loading form A was 37.4% higher than that of specimen TP5 under the loading form B, and the average μ of specimens TP6 and TP9 under the loading form A was 32.4% higher than the μ of specimen TP8 under the loading form B. Moreover, the ductility of the TPSBCs under both of the two loading forms was significantly improved when the segments were reinforced with both inner and outer reinforcements, indicating that the TPSBCs can achieve satisfactory ductility under both of the two loading forms when the reinforcements were properly arranged in the segments.

Effects of the grouted steel tubes

The effects of the grouted steel tubes were evaluated by comparison of the test results of specimens P2 (P3) and TP2 (TP6, TP7, TP9, TP12). It can be seen from Figure 8 that the grouted steel tubes significantly improved the N_ue of the tested specimens. The N_ue of the TPSPCs under the loading form A was 1.9–2.2 times that of the PSBCs. For the specimens with one precast segment, the μ of specimen TP2 was 3.5 times that of specimen P2. For the specimens with three precast segments, the μ of specimen TP12 was 1.6 times that of specimen P3, and the average μ of specimens TP6 and TP9 was 2.8 times that of specimen P3. Moreover, the N_ue and μ of the TPSBCs increased with the decrease of the diameter-thickness ratios of the outer steel tubes. The N_ue of specimen TP7 with a steel tube diameter-thickness ratio of 96 was 9.3% higher than that of specimen TP12 with a steel tube diameter-thickness ratio of 130.

Effects of the sandwich concrete strengths

The effects of the sandwich concrete strengths were evaluated by the test results of TPSBC specimens TP10 and TP7. As shown in Table 3, the SI value of specimen TP7 was 1.20, while specimen TP10 with higher sandwich concrete strength had a lower SI value equal to 1.12. The reduction of the SI value of specimen TP10 can be explained by the confinement index ( $θ$ ) that used to describe the “composite action” between the steel tube and internal concrete (Han et al., 2014). The confinement index is expressed as:

θ = \frac{f_{y t} A_{s t}}{f_{c} A_{c g}}

(3)

in which f_c and A_cg are the axial compressive strength and the gross cross-sectional area of the confined concrete. As seen from Equation (3), the confinement index decreases with increase of the concrete strength. The SI values of TPSBCs that reflect the confinement efficiency of grouted steel tubes thus decrease with increase of the sandwich concrete strengths. However, the increase of the sandwich concrete strength resulted in the improvement of both the N_ue and the ductility of specimen TP10. Among all the tested specimens, specimen TP10 had the largest ductility factor, which was equal to 8.96.

Effects of the number of the precast segments

The effects of the number of the precast segments were evaluated by the test results of PSBC specimens P2 and P3, TPSBC specimens TP1 (TP2) and TP4 (TP6, TP7, TP9) shown in Figure 8 and Figure 9. The number of the segments had no obvious effect on the N_ue and ductility of the PSBCs and the TPSBCs without reinforcement, as comparison of the N-ε curves and the N/N_ue-ε/ε_ue curves of specimens P3 and P2 as well as those of TP4 and TP1. For the TPSBCs with both inner and outer reinforcements, the N_ue of specimen TP2 with one monolithic segment was similar to the average N_ue of specimens TP6, TP7 and TP9 with three segments. However, the number of the segments showed some adverse effects on the ductility of the TPSBCs with reinforcements. The μ of specimen TP6 and TP9 with three segments was 6.6% and 26.9% less than that of specimen TP2 with one segment, respectively. It should be noted that both specimens TP6 and TP9 exhibited ductile behavior with the ductility factor of 6.70 and 5.24, respectively. Therefore, the adverse effects of the number of the segments on the axial compressive behavior of the specimens can be ignored. Similar conclusions were drawn by the study performed by Wu et al. (2018).

Effects of the reinforcements in the precast segments

The use of the reinforcements in the segments improved the N_ue of the TPSBCs under the same loading forms, as shown in Table 3. The effect of the inner spiral stirrups was evaluated by comparison of the test results of specimens TP4 and TP11. The N_ue of specimen TP11 was approximately equal to the sum of the N_ue of specimen TP4 and the compressive capacity of six longitudinal steel bars with a diameter of 14 mm, indicating that the inner spiral stirrups can hardly provide hoop confinement to the hollow segments. The outer spiral stirrups, however, can effectively confine the internal concrete under the loading form A. Specimens TP2 with outer spiral stirrups had the N_ue greater than the sum of the N_ue of specimens TP1 without reinforcement and the compressive capacity of 12 longitudinal steel bars with a diameter of 10 mm. The same conclusion can be drawn from the results of specimens TP6 and TP7. However, the confinement effects of the outer spiral stirrups reduced when the specimens were under the loading form B, as evidenced by the comparison of the test results of specimensTP8 and TP5. This may be attributed to the premature concrete crushing near its segment joint.

As shown in Table 3, for the TPSBCs with one monolithic segment under the loading form A, the μ of specimen TP2 with both inner and outer reinforcements was 2.8 times that of specimen TP1 without reinforcement. For the TPSBCs with three segments under the loading form A, the μ of specimen TP6 with both inner and outer reinforcements was 2.6 times the average μ of specimens TP3 and TP4 without reinforcement. For the TPSBCs with three segments under the loading form B, the μ of specimen TP8 with both inner and outer reinforcements was 2.2 times that of specimen TP5 without reinforcement. The μ of specimen TP11 with only inner reinforcement was only 14.0% higher than that of specimen TP4 without reinforcement, and the μ of the TPSBCs without reinforcement was close to that of the PSBCs. The above results showed that it was necessary to arrange the inner and outer reinforcements in the hollow segments in order to achieve a ductile manner for the TPSBCs.

Axial load versus steel strain analysis

When PSBC specimens were loaded to about 85% of their peak loads, cracks in the concrete cover and yielding of the longitudinal steel bars were observed. However, the bearing capacity increased owing to the increased confinement of the outer spiral stirrups.

The N-ε curves and the N–U curves of the steel tubes for typical TPSBCs are depicted in Figure 11, where ε_vA (ε_vB, ε_vC) and ε_hA (ε_hB, ε_hC) are the average vertical and hoop strains at section A (B, C), respectively, and the transverse deformation coefficient U is the absolute value of the ratio of the average hoop strain ε_h to vertical strain ε_v at the same section.

Figure 11.

N-ε curves and N–U curves of the outer steel tubes for typical TPSBCs: (a) N-ε curves of TP2-S1B2P1, (b) N–U curves of TP2-S1B2P1, (c) N-ε curves of TP8-S3B2P1–B, (d) N–U curves of TP8-S3B2P1–B, (e) N-ε curves of TP10-S3B2P1–C, and (f) N–U curves of TP10-S3B2P1–C.

For the two types of steel tubes with diameter-thickness ratios of 96 and 130, the compressive stress p on the surfaces of the steel tubes was about 2.1% and 1.5% of the hoop stress σ_h in the walls of the steel tubes, respectively. Therefore, the yield point f_yt of the steel tube can be determined according to Von Mises criterion as follows (Liu and Zhou, 2010):

σ_{v} = (\frac{E_{s}}{1 - ν_{s}^{2}}) (ε_{v} + ν_{s} ε_{h})

(4)

σ_{h} = (\frac{E_{s}}{1 - ν_{s}^{2}}) (ν_{s} ε_{v} + ε_{h})

(5)

f_{y t} = \sqrt{\frac{1}{2} [σ_{v}^{2} + {(σ_{v} - σ_{h})}^{2} + σ_{h}^{2}]}

(6)

in which σ_v is the vertical stress of the steel tube; v_s is the Poisson’s ratio of the steel tube.

As shown in Figure 11, the axial load versus the vertical and hoop strain curves at each section of the TPSBCs increased linearly with the increase of the axial load in the initial stage. It can be seen from Figure 8 that the steel tubes of the specimens yielded firstly near the peak loads. After the peak loads, the strain development of the TPSBCs under the different loading forms was quite different. Specimen TP8 under the loading form B had large strain at the middle section B, while the large strain occurred at the column ends of the specimens under the loading form A due to the intrusion of the bearing plates. The intrusion of the bearing plates resulted in the transverse deformation coefficients at sections A and C greater than 1.0 at the peak loads. For specimen TP8 under the loading form B, the transverse deformation coefficients were less than 1.0 at the peak load because the steel tube resisted the vertical load directly together with the internal concrete.

The development of the transverse deformation coefficients at the middle section B of the TPSBCs under the loading form A was basically the same as that of specimen TP8, which meant that the middle part of the TPSBCs had a similar stress state under the loading forms A and B due to good load transferring between the internal concrete and the steel tube by the bond and friction. According to the failure modes of the specimens and the lateral deformation coefficient development of the steel tubes, the reason why the N_ue of the TPSBC under the loading form A was less than that of the TPSBC under the loading form B can be well explained.

Analytical models

For the TPSBC under the loading form B, its axial compressive behavior is similar to that of the concrete-filled steel tubular columns (Ci et al., 2022; Han et al., 2008, 2014; Li et al., 2017; Zhao et al., 2019). Three types of confined concrete are considered due to the different strengths of the segment concrete and sandwich concrete, and different lateral confinements, as shown in Figure 12. The strength of the concrete $f_{c}^{*}$ confined by steel tube or spiral stirrup can be calculated as (Wu et al., 2018; Cai, 2007):

f_{c}^{*} = f_{c} + 2 f_{c} θ

(7)

Figure 12.

The cross section divisions.

For the concrete confined by the steel tube:

θ = θ_{s t} = \frac{f_{y t} A_{s t}}{f_{c} A_{c g}}

(8)

For the concrete confined by the spiral stirrup:

θ = θ_{s s} = \frac{f_{y v} V_{s s}}{f_{c} V_{c g}} = \frac{f_{y v} A_{s s} π D_{s i}}{f_{c} s π D_{s i}^{2} / 4} = \frac{4 f_{y v} A_{s s}}{f_{c} s D_{s i}}

(9)

in which V_ss and V_cg are the volume of the stirrup and the gross volume of the internal concrete, respectively; A_ss and s are the cross-sectional area of the stirrup and the spacing of the stirrups, respectively; f_yv is the yield strength of the stirrups.

According to the related research (Liu and Zhou, 2010; Wang et al., 2019) and the above analysis, the compressive capacity of the longitudinal steel bars can be directly added to the compressive bearing capacity of the composite column. Therefore, the ultimate bearing capacity of the fully loaded TPSBC can be calculated as:

N_{u} = η [f_{c 1}^{*} (A_{c n 1} - A_{c n 2}) + f_{c 2}^{*} (A_{c n 2} - A_{c n 3}) + f_{c 3}^{*} A_{c n 3}] + f_{y b} A_{s b}

(10)

By introducing equations (7) ∼ (9), the equation (10) becomes:

N_{u} = η [f_{c 1} (A_{c n 1} - A_{c n 2}) + f_{c 2} (A_{c n 2} - A_{c n 3}) + f_{c 3} A_{c n 3} + \frac{2 f_{y t} A_{s t}}{A_{c g 1}} A_{c n 1} + \frac{2 f_{y v} V_{s s}}{V_{c g 3}} A_{c n 3}] + f_{y b} A_{s b}

(11)

in which η is a reduction factor introduced to consider the asynchronous deformations corresponding to the confined concrete strength in different areas (Cai, 2007), and the value of η is recommended to be 0.95 according to the test results in this study. A_cn1, A_cn2, and A_cn3 are the net cross-sectional areas of the concrete surrounded by the outer steel tube, the virtual steel tube, and the inner edge of the stirrup, respectively.

For the TPSBC under the loading form A, a strength reduction factor was introduced to consider the local compression effect on the concrete and it can be conservatively taken as ${(A_{l} / A_{c n})}^{0.5}$ (Cai, 2007). The strength of the concrete $f_{c, l}^{*}$ confined by the steel tube or spiral stirrup under local compression can be calculated as:

f_{c, l}^{*} = (f_{c} + 2 f_{c} θ) \sqrt{\frac{A_{l}}{A_{c n}}}

(12)

in which A_l is the local compression area.

The ultimate bearing capacity of the locally loaded TPSBC can be calculated as:

N_{u, l} = η [f_{c 1, l}^{*} (A_{c n 1} - A_{c n 2}) + f_{c 2, l}^{*} (A_{c n 2} - A_{c n 3}) + f_{c 3, l}^{*} A_{c n 3}] + f_{y b} A_{s b}

(13)

For the convenience of analysis and superposition, the reduction factor ${(A_{l} / A_{c n})}^{0.5}$ is simplified as ${(A_{l} / A_{c n 1})}^{0.5}$ in equation (12).

By introducing equation (12), the equation (13) becomes:

N_{u, l} = η [f_{c 1} (A_{c n 1} - A_{c n 2}) + f_{c 2} (A_{c n 2} - A_{c n 3}) + f_{c 3} A_{c n 3} + \frac{2 f_{y t} A_{s t}}{A_{c g 1}} A_{c n 1} + \frac{2 f_{y v} V_{s s}}{V_{c g 3}} A_{c n 3}] \sqrt{\frac{A_{l}}{A_{c n 1}}} + f_{y b} A_{s b}

(14)

The ultimate bearing capacity of the PSBCs can be calculated as:

N_{u}^{P} = f_{c} A_{c n} (1 + 2 θ_{s s}) + f_{y b} A_{s b}

(15)

Table 3 presents the ultimate bearing capacity of all specimens calculated by the proposed analytical models. As shown in Table 3, the calculated bearing capacity is less than the experimental peak load (N_ue) in almost of all specimens except specimen TP8, the calculated bearing capacity of which is slightly greater than its N_ue due to the premature concrete crushing near the segment joint. The performance of the analytical models is further evaluated by the mean value (MV) and the standard deviation (SD) of the ratios of the N_ue to the calculated values. The MVs of the PSBC specimens and TPSBC specimens under the loading form B are 1.119 and 1.039, respectively. The MV and SD of the TPSBC specimens under the loading form B are 1.082 and 0.027, respectively. The proposed analytical models give a relatively rational prediction of the ultimate bearing capacity of the PSBC and TPSBC specimens. However, due to lack of enough experimental results, the size effects of the proposed TPSBC columns were not considered in the analytical models. Therefore, further research is still necessary to study the size effects of hollow TPSBC columns with reinforcements, and more parametric studies are still needed to validate the feasibility of the proposed analytical models.

Conclusions

The axial compressive behavior of the proposed TPSBC was investigated by testing of 14 specimens. Based on the experimental results and discussion presented in this paper, the following conclusions can be drawn:

1. The final failure of the conventional PSBC specimens was induced by the concrete crushing, and the concrete damage was more serious near segment joints due to the discontinuity of the reinforcements and concrete. The proposed TPSBC achieved a desirable ductile failure due to the use of grouted steel tubes.

2. The loading forms and the reinforcement arrangements in the precast hollow segments affected the failure modes of the proposed TPSBC specimens. For the specimens under the loading form B (full compression), the inclined shear failure finally occurred in the specimens without reinforcement, while the specimens with both inner and outer reinforcements failed with concrete crushing at the segment joints and outward local buckling of the steel tube. For the specimens under the loading from A (local compression), the steel tube of each TPSBC bulged outward in a bud shape within the length of its diameter, and the concrete inside the steel tubes underwent large plastic deformation and was finally crushed.

3. The loading forms affected the maximum bearing capacity and the ductility of the TPSBCs with hollow sections. The maximum bearing capacity decreased with the decrease of the compression area; however, the ductility factors of the specimens under local compression were higher from a range between 32.4% and 37.4% than that of the specimens under full compression.

4. The prestress used to clamp the precast segments together and the number of the precast segments had little effect on the maximum bearing capacity and ductility of the TPSBC specimens.

5. The use of grouted steel tubes significantly improved the axial compressive behavior of the specimens. The maximum bearing capacity and the ductility factors of the TPSBCs were more than 1.9 and 1.6 times those of the PSBCs, respectively. The increase of the sandwich concrete strengths significantly improved the ductility of the specimens under local compression, while the increase of the diameter-thickness ratios of the outer steel tubes deteriorated the maximum bearing capacity and the ductility of the TPSBCs.

6. With proper consideration of the loading forms, and the combined confinement effects of the steel tube and the outer spiral stirrups, analytical models were developed for predicting the ultimate bearing capacity of the TPSBCs. The calculated results were in good agreement with the experimental peak loads.

The good behavior of the proposed TPSBC under axial compression has already been verified with tests of 14 stub columns. The authors have also verified the desirable seismic performance of the proposed TPSBC, and the test results will be reported later. To extend the application of TPSBC in seismic zones, further research is still necessary to validate the developed analytical models with more experimental results and investigate the behavior of the columns under complicated loading conditions.

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: The research reported was carried out under the support of the National Natural Science Foundation of China (No. 51978656) and Jiangsu Key Laboratory of Environmental Impact and Structural Safety in Engineering, China University of Mining & Technology (No. KFJJ202004).

ORCID iDs

Tao Zhou

Xian Li

References

Billington

Barnes

Breen

(1999) A precast segmental substructure system for standard bridges. PCI Journal 44(4): 56–73.

Ding

Chen

, et al. (2012) Investigation of the seismic performance of precast segmental tall bridge columns. Structural Engineering and Mechanics 43(3): 287–309.

Song

, et al. (2016) Cyclic loading test of unbonded and bonded posttensioned precast segmental bridge columns with circular section. Journal of Bridge Engineering 21(2): 04015043.

Cai

(2007) Modern Steel Tube Confined Concrete Structures. Revised edition. Beijing: China Communications Press. (in Chinese).

Cai

Yuan

Wang

, et al. (2021) Seismic behavior of precast segmental bridge columns reinforced with hybrid FRP-steel bars. Engineering Structures 228: 111484.

Chen

Wang

Zhang

(2020) Size effect prediction on axial compression strength of circular CFST columns. Journal of Constructional Steel Research 172: 106221.

Chou

Chen

(2006) Cyclic tests of post-tensioned precast CFT segmental bridge columns with unbonded strands. Earthquake Engineering and Structural Dynamics 35(2): 159–175.

Chou

Chang

Hewes

(2013) Two-plastic-hinge and two dimensional finite element models for post-tensioned precast concrete segmental bridge columns. Engineering Structures 46: 205–217.

Ahmed

Tran

, et al. (2022) Axial compressive behavior of circular concrete-filled double steel tubular short columns. Advances in Structural Engineering 25(2): 259–276.

10.

Feng

Qiang

(2017) Discussion and definition on yield points of materials, members and structures. Engineering Mechanics 34(3): 36–46. (in Chinese).

11.

GB 50010–2010 . (2015) Code for Design of Concrete Structures. (in Chinese).

12.

GB 50936-2014 . (2014) Technical Code for Concrete Filled Steel Tubular Structures. (in Chinese).

13.

Guerrini

Restrepo

Massari

, et al. (2015) Seismic behavior of posttensioned self-centering precast concrete dual-shell steel columns. Journal of Structural Engineering 141(4): 04014115.

14.

Han

Huo

Wang

(2005) Compressive and flexural behaviour of concrete filled steel tubes after exposure to standard fire. Journal of Constructional Steel Research 61(7): 882–901.

15.

Han

Liu

Yang

(2008) Behavior of thin walled steel tube confined concrete stub columns subjected to axial local compression. Thin-Walled Structures 46(2): 155–164.

16.

Han

Bjorhovde

(2014) Developments and advanced applications of concrete-filled steel tubular (CFST) structures: members. Journal of Constructional Steel Research 100: 211–228.

17.

Hewes

(2007) Seismic tests on precast segmental concrete columns with unbonded tendons. Bridge Structures 3(3–4): 215–227.

18.

Ichikawa

Matsuzaki

Moustafa

, et al. (2016) Seismic-resistant bridge columns with ultrahigh-performance concrete segments. Journal of Bridge Engineering 21(9): 04016049.

19.

Jia

Zhang

, et al. (2020) Seismic performance of self-centering precast segmental bridge columns under different lateral loading directions. Engineering Structures 221: 111037.

20.

Jin

Fan

, et al. (2019) Size effect of axial-loaded concrete-filled steel tubular columns with different confinement coefficients. Engineering Structures 198: 109503.

21.

Kawashima

MacRae

Hoshikuma

, et al. (1998) Residual displacement response spectrum. Journal of Structural Engineering 124(5): 523–530.

22.

Wang

Han

(2017) Behaviour of grout-filled double skin steel tubes under compression and bending: experiments. Thin-Walled Structures 116: 307–319.

23.

Hao

(2019) Seismic performances of precast segmental column under bidirectional earthquake motions: shake table test and numerical evaluation. Engineering Structures 187: 314–328.

24.

Liu

Zhou

(2010) Behavior and strength of tubed RC stub columns under axial compression. Journal of Constructional Steel Research 66(1): 28–36.

25.

Marriott

Pampanin

Palermo

(2009) Quasi-static and pseudo-dynamic testing of unbonded post-tensioned rocking bridge piers with external replaceable dissipaters. Earthquake Engineering and Structural Dynamics 38(3): 331–354.

26.

Mashal

Palermo

(2019) Low-damage seismic design for accelerated bridge construction. Journal of Bridge Engineering 24(7): 04019066.

27.

Pratiwi

Song

(2018) Pseudodynamic testing and inelastic displacement ratios of self-centering precast concrete segmental bridge columns. Journal of Structural Engineering 144(9): 04018158.

28.

Shen

Liu

, et al. (2021) Cyclic tests of precast post-tensioned concrete filled steel tubular (PCFT) columns with internal energy-dissipating bars. Engineering Structures 229: 111651.

29.

Zhong

, et al. (2021) Axial behavior of novel CFDST columns with outer welded corrugated steel tubes. Structures 34: 2708–2720.

30.

Wang

Tang

, et al. (2018a) Cyclic loading test of self-centering precast segmental unbonded posttensioned UHPFRC bridge columns. Bulletin of Earthquake Engineering 16(11): 5227–5255.

31.

Wang

Tang

, et al. (2018b) Seismic behavior of precast segmental UHPC bridge columns with replaceable external cover plates and internal dissipaters. Engineering Structures 177: 540–555.

32.

Wang

Yang

, et al. (2019) Behaviour of concrete-filled corrugated steel tubes under axial compression. Engineering Structures 183: 475–495.

33.

Peng

Zhao

, et al. (2018) Axial loading tests of thin-walled circular steel tubes infilled with cast-in-place concrete and precast segments containing DCLs. Thin-Walled Structures 127: 275–289.

34.

Xiao

Choi

(2005) Confined concrete-filled tubular columns. Journal of Structural Engineering 131(3): 488–497.

35.

Zhang

(2021) Energy dissipation and resilience of precast segmented concrete-filled steel tube self-centering column: parameter study and design methodology. Engineering Structures 244: 112747.

36.

Zhao

Yan

Lin

(2019) Compressive strength of axially loaded circular hollow centrifugal concrete-filled steel tubular short columns. Engineering Structures 201: 109828.

37.

Zhao

Liang

, et al. (2020) Size effect of square steel tube and sandwiched concrete jacketed circular RC columns under axial compression. Journal of Constructional Steel Research 166: 105912.