Elsevier

Engineering Structures

Volume 30, Issue 2, February 2008, Pages 287-299
Engineering Structures

Concrete-filled cold-formed circular steel tubes subjected to variable amplitude cyclic pure bending

https://doi.org/10.1016/j.engstruct.2007.03.025Get rights and content

Abstract

This paper describes an experimental investigation of the cyclic inelastic flexural behaviour of concrete-filled tubes (CFT). Controlled-rotation, symmetrical cyclic bending tests were performed using a variable amplitude loading history on different sizes of CFT with different section slenderness. The CFT beams exhibited stable hysteresis behaviour up to local buckling and then showed considerable degradation in strength and ductility depending on the D/t ratio. Seismic capacity parameters are presented including strength, hysteresis curves and modes of failure for the specimens. Peak moments obtained in the cyclic tests were compared with design moments predicted using a number of steel and concrete specifications. New section slenderness limits suitable for design and construction of seismic resisting structural systems were determined. A comparison is made between these seismic slenderness limits and the limits available in the design codes.

Introduction

The use of composite construction has become more widespread in recent years. Composite structures for buildings often include moment-resisting steel frames or braced frame systems with steel–concrete composite columns to control lateral drift under severe seismic loading. In braced frame systems, the braces may be also composite members [1], [26]. This method of construction combines the advantages of both steel and concrete: namely, the ductility, speed of construction, light weight of the steel, coupled to the inherent mass, stiffness, damping and economy of cheaper concrete. The most popular shape of composite members used as columns, beams, beam–columns, and braces is the concrete-filled steel tube (CFT). The steel tube confines the core concrete and the concrete delays or completely prevents local buckling depending on the diameter-to-thickness (D/t) ratio. The core concrete forces the steel tube to buckle outwards and thus the composite section sustains larger forces and lead to total saving of steel material in comparison to a structural steel system. The strength, ductility and energy absorption under cyclic loading were investigated in the past for circular CFT members [1], [2], [4], [5], [6], [7], [27] and for square CFT members [3], [8], [9]. However, CFT members subjected to cyclic pure bending received little attention particularly those made of slender tubes with D/t ratios larger than about 80.

In related papers, the authors [10], [11] experimentally examined circular CFT beams under both static and constant amplitude cyclic pure bending where 20<D/t<160 and found that: (1) the static strength of the empty tube increases on average by 50% due to concrete filling; (2) cyclic loading has a significant effect on the bending strength of CFT beams made of slender tubes; whereas it has little effect on those CFT beams made of compact and non-compact tubes; (3) concrete filling enhances the Low Cycle Fatigue (LCF) fracture life of circular hollow sections (CHS) beams by at least 2 1/2 times under large deformation cyclic pure bending. The LCF life was determined as the number of cycles to reach fracture of the steel tube.

This paper describes an experimental investigation of the cyclic inelastic flexural behaviour of CFT made of as-received and machined cold-formed CHS and subjected to incrementally increasing cyclic loading (IICL) protocol with diameter-to-thickness ratios in the range of 20<D/t<120. The CFT specimens were tested to destruction under cyclically applied rotations simulating moderate to severe earthquake-type oscillations. A total of ten CFT beams were subjected to cyclic loading protocols as commonly used by Japanese researchers [4], [7], [8]. The rotation capacity was determined and used to derive new fully ductile section slenderness limits suitable for seismic design. A comparison is made between the newly derived seismic slenderness limits, and the available limits in the design codes. The effect of concrete filling is also discussed by comparing the current results with those from empty CHS obtained recently by the authors [24].

In a related paper, the authors have experimentally verified that the plastic slenderness limit λp=188 for circular CFT beams from static pure bending tests [10] where this limit was found to correspond to D/t-limit =110. The steel tubes used to construct the specimens had a nominal and an average measured yield stresses of 350 MPa and 419 MPa, respectively [10]. The present paper attempts to determine new slenderness limits to define Category 1, fully ductile sections (λfd) suitable for seismic design.

It is worth noting that the results obtained in this paper are necessary from a research view point and they are not valid for design purposes as small size tubes were used to construct the specimens. For design purposes, large scale testing on large diameter CHS is required to verify the results presented in this paper and confidently derive new design rules.

Section snippets

Section slenderness limits

The section slenderness parameter λs is often used in the classification of bare steel members under static and cyclic loading, and they are defined in AS4100 [20] and NZS 3404 [21] as λs=(Dt)(σy250).

The non-dimensional buckling parameter α used in [10], [11] is α=(Esσy)(Dt) where D is the measured outside diameter of the tube, t is the measured wall thickness of the tube, and σy and Es is the measured yield stress and elastic modulus of the steel tube.

The slenderness limits are widely used in

Test specimens

Ten cyclic bending tests were performed on CFT uncapped specimens made from as-received cold-formed circular hollow section (CHS) and machined CHS to achieve larger D/t ratios. Table 1 summarises the key mechanical properties of the specimens. The nominal end-to-end and mid-span pure bending lengths of the specimens were L0=1500mm and Lc=800mm, respectively as shown in Appendix A. Note, in Fig. A.2 that the length LAB after deformation is less than LAB at the start of the test because the

Failure modes

Fig. 1 shows a schematic of the typical failure mode that occurred during the IICL tests. The specimens after testing and those presented in [10] are shown in Appendix C together with CHS specimens subjected to cyclic bending [1], [24]. In the early cycles, the steel tube formed plastic ripples which continued to grow with cycling to form the ring mode (Fig. 1(a)) until the tube fractured (Fig. 1(b)) at the critical section with the dominant buckle which was accompanied with concrete fracture

Strength of CFT beams

A simplified rigid plastic approach was used in Elchalakani et al. [11] to determine the flexural capacity of circular CFT. It is worth noting that no slip was observed in the previous static or the present cyclic tests. Thus, assuming full bond between the steel tube and the concrete core, and that the section modulus of the steel tube is not affected by the formation of plastic ripples, the ultimate flexural strength of CFT was determined as [11]Mtheory=23σcRi3cos3γ0+4σyRm2tcosγ0 where γ0 is

Fully ductile slenderness limits

Fig. 8 shows the results of the CFT cyclic tests together with previous results from static bending tests on CHS [31] and CFT beams [11]. The results of previous cyclic tests on CHS are also shown [24]. This plot shows the variation of the rotation capacity [R=(θmaxθy)/θy] with the section slenderness λs. θmax is the inelastic rotation at the forming of the plastic hinge for the CHS and at fracture for the CFT, where θy was defined in Eqs. (5), (6) for the CFT. For those specimens

Conclusions

The IICL loading protocol was found to have a considerable effect on the static strength of CFT beams, particularly for Class 3 and 4 CFT sections with large section slenderness. Class 1 CFT section did not ripple or fracture up to nf=1718 cycles. Class 2 CFT section rippled and fractured in the range of nf=1015 cycles, whereas Class 3 CFT section rippled and fractured in the range of nf=810 cycles. Class 4 CFT section rippled and fractured in the range of nf6 cycles. It can also be

Acknowledgements

The writers are grateful to the Australian Research Council and Monash University for their financial assistance for the project. Thanks are given to Palmer Tube Mills and OneSteel Market Mills for providing the steel tubes. The technical assistance of Mr. Graham Rundle, the late laboratory manager, Mr. Kevin Nievaart and Mr. Geoff Doddrell is gratefully acknowledged. Special thanks go to Prof. Jerome Hajjar of the University of Minnesota and A/Prof. Sherif M. El-Tawil of the University of

References (34)

  • H.G.L. Prion et al.

    Beam–column behaviour of steel tubes filled with high strength concrete

    Canadian J Civil Eng

    (1993)
  • Toshiyuki F, Noguchi T, Mori O. Evaluation of deformation capacity of concrete filled tubular beam–columns. In: 3rd...
  • T. Usami et al.

    An experimental study on elastic–plastic behaviour of compression members under cyclic loading-effect of loading program

    Japanese J Struct Eng

    (1993)
  • X.L. Zhao et al.

    Void-filled SHS subjected to large deformation cyclic bending

    J Struct Eng ASCE

    (1999)
  • R. Bergmann et al.
    (1995)
  • AS 1163. Standards Association of Australia. In: Structural steel hollow sections. Sydney (Australia);...
  • AS 1391. Standards Association of Australia. In: Methods for tensile testing of metals. Sydney (Australia);...
  • Cited by (75)

    View all citing articles on Scopus
    View full text