Hysteretic performance research on high strength circular concrete-filled thin-walled steel tubular columns

Under violent earthquake motions, the severe damage in critical regions of structures could be ascribed to cumulative damage caused by cyclic loading. Using the high strength (HS) materials in concrete-filled steel tubular (CFST) columns is the effective way and popular tendency to promote the seismic behavior in anti-seismic design. In this paper, an experimental study on the hysteretic performance of high strength circular concrete-filled thin-walled steel tubular columns (HCFTST) columns was carried out. A total of six specimens were tested under constant axial compression combining cyclic lateral loading. The tested parameters were the different combinations of diameter-to-thickness (D/t) ratio, axial compression ratio (n) and concrete cylinder compressive strength (fc).The failure modes, load-displacement hysteretic curves, skeleton curves, dissipated energy and stiffness degradation were examined in detail. Through the experiment analysis result, it indicates that the ultimate limit state is reached as the severe local buckling and rupture of the steel tubes accompanying the core concrete crushing occur. Using high strength materials could have a larger elastic deformation capacity and the higher axial compression ratio within test scopes could motivate the potential of HS materials. In brief, the HCFTST columns with ultra-large D/t ratios under reasonable design could perform excellent hysteretic performance, which can be applied in earthquake-prone regions widely.


Introduction
Concrete-filled steel tubular (CFST) columns as a relatively new type of structural member, have been widely applied in engineering construction, not only due to their remarkable mechanical properties for seismic resistance, namely the high strength, large lateral stiffness, high ductility and greater energy dissipation capacity, but also due to its outstanding construction efficiency and economic benefits [1][2][3].Earlier significant experimental researches conducted by many scholars [4][5][6][7][8][9][10][11][12][13][14] do have great importance on the understanding of hysteretic performance of CFST columns.Recently, using the high strength (HS) materials in concretefilled steel tubular (CFST) columns is the effective way and popular tendency to promote the seismic behavior in anti-seismic design [15][16][17][18].
Elremaily and Azizinamini [15] carried out experimental study on cyclic behavior of circular CFST columns with concrete cylinder compressive strength (fc) varying from 40 to 104 MPa.Varma et al. [16,17] examined the seismic behavior of HS square CFST columns filled with HS concrete (fc =110 MPa) and steel yield strength (fy) ranging from 269 to 660 MPa.Skalomenos et al. [18] investigated the hysteretic behavior of circular and square CFST columns with fc varying from 42.4 to 82.3 MPa and fy varying from 387 to 788 MPa.For the CFST columns, the HS steel tube can provide a larger elastic deformation capacity and better confined effect, while the application of HS core concrete can delay local buckling of steel tube.Moreover, using HS concrete and common strength (CS) steel, using HS steel and CS concrete or both materials are HS, can reduce the section size to obtain favorable architecture aesthetics effect and gain further economic benefits under the premise of ensuring the mechanical performance [19].
High strength circular concrete-filled thinwalled steel tubular (HCFTST) columns get rapid development due to the widespread application of HS materials.As for the concretefilled thin-walled steel tubular (CFTST) columns, Zhang et al. [20], Goto et al. [21] and Gajalakshmi [22] investigated the hysteretic behavior of that through quasi-static cyclic loading tests.However, it is noteworthy that no HS materials were adopted in their tests [20][21][22].On the other hand, the aforementioned researches [15][16][17][18] mainly aim at hysteretic behavior of the traditional concrete-filled thickwalled steel tubular columns using HS materials.Therefore, more experimental studies are needed to examine the hysteretic performance of CFTST columns using HS materials, namely the HCFTST columns.
In this paper, an experimental study on the hysteretic performance of HCFTST columns, including six specimens, was carried out under constant axial compression combining cyclic lateral loading.The tested parameters were the different combinations of diameter-to-thickness (D/t) ratio, axial compression ratio (n) and concrete cylinder compressive strength (fc).The failure modes, load-displacement hysteretic curves, skeleton curves, dissipated energy and stiffness degradation were examined in detail.Overall, the HCFTST columns with reasonable design could display favorable hysteretic performance.

Material properties
The reserved concrete blocks for compressive strength test were curing under the same condition with test columns.The actual cylinder strengths (fc) are listed in Table 1.The Q690 HS steel was used to make the steel tubes.Tensile coupon test shown in Fig. 1, according to "Metallic materials tensile testing at ambient temperature" (GB/T 228-2002), was carried out to determine the material properties.The test result shows that the yield strength fy is 722.64 MPa and ultimate strength fu is 765.37 MPa.The elastic modulus Es is 222010 Mpa and the Poisson's ratio vs is 0.281.

Specimen design
A summary of specimens is presented in Table 2, in which the D/t ratios adopted exceeds the current limitations specified by various codes [22] to examine the effect of it.The axial compression ratio (n) in this paper is defined as follows: where P is the axial load applied on the HCFTST columns and P0 is the nominal squash load calculated by the equation as follows: 0 0.85 where As and Ac are the cross-sectional areas of the steel tube and concrete core respectively.
The designed details of all aforementioned test specimens are shown in Fig. 2. The height of the HCFTST columns is 525 mm and a reinforced concrete (RC) foundation was built for the convenience of applying cyclic loads.The thickness of all end plates is 20 mm, especially for end plates II with bolt holes welded at the top of specimen for loading.The cross section sizes of the RC foundation are 1200(length) 450(width) 420(height)  mm and 1200 500 420  mm, where the former was adopt by the HCFTST columns with diameters 140 mm and 180mm, and the later was applied to the other specimens.HS steel bars (fy=1000MPa) were used to prevent the RC foundations out of failure during test.Furthermore, stiffening ribs with a thickness of 10 mm were also applied by welding to strengthen the foundations.

Test setup and procedures
The loading apparatus adopted for the test shown in Fig. 3 enables the application of the constant axial compressive load combining cyclic lateral load.The axial compression was imposed on the column attached with the triangular box connector through a hydraulic jack of 2000 kN capacity and the 1000 kN loading capacity MTS 244.51 actuator was used to apply cyclic lateral load.In addition, the connector, test columns and actuator were fixed together by HS bolts Grade 12.9 M28.Furthermore, to ensure the stability of test specimen, the RC foundation was attached to the ground tightly with pressure beams and anchor stocks while two adjustable seats were installed to prevent the lateral sliding of the specimen.The horizontal load and displacement were automatically recorded by the MTS actuator acting on the test columns.The axial force was monitored by the pressure transducers fixed on the jack.During the test, preloading procedure was conducted to ensure the equipment ran normally.If all was in order, the axial load was applied firstly to the targeted value and then kept constant during testing.Afterwards the lateral force was cycled under the displacement control mode as shown in Fig. 4. Four single cycles with peak drift ratios of / L  0.10%, 0.25%, 0.50% and 0.75% were initially applied, where '  ' indicates the lateral displacement and 'L' stands for the column length.Then three cycles were employed to the specimen at each drift ratio level from 1% to 8%.What's more, if there was no evident failure occurring after applying drift ratio 8%, the test would continue by applying drift ratio 10% until the specimen failed.The test ended when the specimen was unable to sustain the targeted axial force due to the severe fracture and crushing of the steel tube and core concrete or the lateral load degraded over than 35% of the maximum experienced load.

Failure modes
The typical events including slight local buckling, severe buckling before fracture, crack appearing without penetrating tube and serious accompanied with rupture almost penetrating cross section occurred during the test.A summary of failure modes for HCFTST columns is shown in Fig. 5.As is shown, the severe fracture and local buckling of thin-walled steel tubes could be observed apparently in the end.For all test columns, the cracking height is about 20 mm away from the RC foundation surface.After the test, the specimen CFST-3-2 with the severest cross section failure was cut off to examine the failure mode of core concrete.Apparently, the core concrete experienced serious damage under the cyclic loading.Moreover, the confined effect decreased gradually due to the local buckling occurring, for which the HS core concrete of the test column finally underwent an adverse shear failure mode with a smooth fracture surface.

Load-displacement hysteretic curves
The load-displacement hysteretic curves of all test columns are shown in the plots as follows.It indicates that there is an initial elastic response for all specimens and then the columns enter the elastic-plastic process accompanied by the increasing stiffness degradation and energy dissipation gradually.The initial local buckling points and fracture of steel tube are marked in the curves.Generally the initial local buckling occurred at drift ratio 2%3% for most specimens and the fracture of steel tube emerged until reaching the 7%8% drift.A gradual drop in strength can be seen when maximum lateral load is reached.The hysteretic curves are not strictly symmetrical because during the reciprocating loading cumulative damage develops preferentially due to the initial imperfections of the HS material and obvious Bauschinger effect of HS steel.For the other HCFTST columns, with the increasing of drift level, there are obvious softening platforms existing in hysteretic curves when the lateral load turns from unloading to reloading.For HCFTST columns with ultra-large D/t ratio exceeding current codes, increasing axial compression ratio is an efficient way than improving concrete strength to motivate the potential of HS materials for offering an excellent confined action to limit the propagation of the concrete shear cracks in antiseismic design.

Energy dissipation capacity
The energy dissipation analysis was conducted by the area integral based on the hysteretic curves, as shown in Fig. 9, from which it indicates that the loop energy of different drift levels amplifies gradually with the increasing of lateral deflection level while within the same drift deformation the loop energy decreases gradually due to the cumulative damage caused by cyclic loading.when entering the failure stage, the evident descending of loop energy could be observed revealing that the tested specimens would fail soon later.As for the accumulated energy ratio, it can be expressed as follows: 1 where Ei is the accumulated energy until the cyclic loop n and ET is the total ultimate energy dissipation.Fig. 10 shows the result of accumulatived energy ratio analysis, which display that the cumulative damage develops in a nonlinear way with starting slowly and then converts to a burgeoning mode after the appearance of initial local buckling.The energy dissipation capacity could also be evaluated by the equivalent viscous damping coefficient shown in Fig. 11, which can be calculated through the equation:

Skeleton curves
The skeleton curves of tested specimens shown in Fig. 13 were constructed by connecting maximum load point at each displacement level according to the load-displacement hysteretic curves.From the skeleton curves, improving the concrete strength and axial compression ratio within the test scope, can promote the the ultimate bearing capacity of HCFTST columns.For the columns with ultra-large D/t ratio, improving the axial load level could motivate the potential of the HS materials due to the better workability of confined effect.

Stiffness degradation
In this test, linear fit method was employed to calculate the tangent stiffness for all loops at different drift ratios of all tested columns during the unloading.The summary of unified tangent stiffness for all tested specimens is shown in Fig. 14, in which the gradual degradation of stiffness could be observed apparently due to the local buckling and concrete crushing.All that describes the basic propagation of cumulative damage under reciprocating incremental lateral load mode.

Conclusions
This paper presents the experimental study on hysteretic performance of HCFTST columns.Based on the experimental results, the following conclusions can be drawn within the scope of the current study: (1) The failure mode of HCFTST columns was reached as the severe local buckling and fracture of steel tube accompanied the core concrete crushing at 7%8% drift ratio.For HCFTST columns with ultra-large D/t ratio exceeding current codes, increasing axial compression ratio is an efficient way than improving concrete strength to motivate the potential of HS materials for offering an excellent confined action to limit the propagation of the concrete shear cracks in antiseismic design.
(2) The evident descending of energy dissipation could be observed in failure stage, indicating forthcoming severe failure phenomenon, namely the complete fracture of steel tube and severe crushing of core concrete.
(3) The HCFTST columns with reasonable design could display favorable hysteretic performance and can be expected to have a widespread application in earthquake-prone regions.

Fig. 9 .Fig. 10 .
Fig. 9. Energy dissipation analysis is the area of triangles.The analysis of equivalent viscous damping ratio is displayed in Fig.12.The result reiterates the fact that the energy dissipation capacity of the HCFTST columns decreases obviously in the failure stage due to the rupture of the thinwallled steel tubes and concrete crushing.

Table 1 .
Actual strength of concrete