BEHAVIOUR OF THIN-WALLED CONCRETE-FILLED STEEL TUBES IN BENDING AND COMPRESSION

Abstract

A concrete-filled steel tube (CFST) is a structural element that can be used in axial compression, in bending as well as in eccentrically loaded conditions. A CFST element is prepared by filling a hollow steel tube with concrete. Commonly circular, square and rectangular cross-sections of CFST elements are used. Rectangular and square sections are employed to obtain thin-walled concrete-filled steel tubes for testing in axial compression as well as in pure bending. Two different grades of concrete M30 and M50 are used to prepare thin-walled concrete-filled steel tube (TWCFST) elements. M30 and M50 grade concrete is prepared by replacing natural aggregate as 35% and 15% respectively with recycled aggregate. The cylinder strength of M30 and M50 grade concrete was obtained as 33.75 and 53.55 MPa respectively. The steel tubes are prepared with 2 mm thin steel plate of HR3 grade steel. The coupons from the steel plate are prepared according to ASTM E8/E8M-09. The yield strength and ultimate strength of the steel are obtained as 270 MPa and 380 MPa respectively. Different types of core configurations of rectangular sections (90 mm × 120 mm) and square sections (100 mm × 100 mm) were used for the study as: a) Plane core i.e., no stiffener in section, b) A single stiffener of 15 mm size on both larger sides, c) A single stiffener of 20 mm size on both larger sides, d) A single stiffener of 15 mm on each inner face of steel tube, e) A single stiffener of 15 mm in each inner face of steel tube. e) A single stiffener of 20 mm in each inner face of the steel tube, and (f) steel tube inside another steel tube. In case of uniform thin-walled steel tube beam specimens, the section of specimens were 90 mm × 120 mm and 90 mm × 122 mm. In case of non-uniform TWCFST beam specimens, the section was taken as 100 mm × 100 mm and 100 mm × 102 mm. In all cases, the overall length of the beam specimen was taken as 900 mm. To study the behaviour of TWCFST beam elements in pure bending, two types of specimens from the square and rectangular sections are taken; a) Beam specimen having uniformly thin steel on all sides of the section, steel tube thickness 2 mm was taken on all sides of the section. b) Beam specimens having a higher thickness on the bottom side compared to the other three sides. The thickness of the steel tube was taken on the bottom side as 4 mm and on the rest three sides as 2 mm. The parameters for study were shapes of cross-section, grades of in-filled concrete, the thickness of steel tubes on bottom sides, and core configuration. Total 42 TWCFST beam specimens were tested in the laboratory. The structural performance of tested TWCFST beam specimens was investigated for each cross-sectional configuration. A Fibre Element Analysis (FEA) model was used to validate the moment-deflection curve of all tested beam specimens. FEA model is a program in which the stress-strain of steel and concrete are incorporated. Strain value is used as input to find the stress value in both materials simultaneously. These strain values are used to find the force and moments. In the FEA model, a section of beam specimen is divided into many parts, say 100 parts in both directions. Thus, the composite section was divided into 10000 fibers for validation. For division in many more parts, the FEA will provide results more accurately. The test performance of experimental work was validated with the help of the FEA model. The peak value of the experimental moment obtained at any deflection value is compared with the moment value calculated using FEA for the same deflection value. Experimental investigations on bending moment capacity showed that the bending moment capacity of RTWCFST beam specimen increases with increasing the number and size of stiffeners. RTWDSCFST beam specimens failed in a ductile manner during testing in pure bending. The cracked parts of the beam specimen are also displayed under the failure pattern. The thin-walled CFST column specimens with the same configuration as taken for CFST beams are adopted. These TWCFST column specimens were given different surface conditions along with their full height. The sections of TWCFST column specimens are 900 mm × 120 mm and 100 mm × 100 mm. The different surface conditions of these TWCFST column specimens are: a) Continuous plane surface along with the full height, b) A 2 mm cut/notch was provided at midheight of specimens, c) Two cut/notch of 2 mm were provided at a height of one-third distance from both ends of specimens and d) In the case of rectangular section, additional steel plates of 90 mm × 300 mm size were placed on the top side and bottom side on the opposite face of the shorter side. In the case of the square section, the additional steel plates were provided on the 200 mm (50+100+50 mm) perimeter part on the top and bottom side on the opposite face of the specimen. The parameters for experimental studies are section of columns, surface conditions of steel tubes along with the height of the specimen and grades of in-filled concrete. Experimental investigation on the mode of deformations showed that the steel wall of specimens got separated from the concrete core. At the locations of separation, local buckling in the steel wall was triggered. The local buckling may occur on the top side, bottom side, or somewhere along the height of the specimen. The concrete started to be crushed at the location of buckling of the steel wall. The specimen having a single notch normally failed like a plane specimen in most of the cases, the gap in the notch was found as closed. The specimens having notches failed near the location of the notch on the bottom side. The tube in tube specimen, failed sharply after attaining peak load due to failure of the inside tube. The load-displacement relation was also compared for the specimens filled with M30 and M50 concrete. It was observed that after peak load, all the specimens filled with M30 and M50 concrete exhibited ductile behaviour. All TWCFST column specimens behaved successfully up to their peak load and welded joints were intact. The energy absorption for displacement was also calculated for each specimen. The energy absorption capacity of all specimens are compared among themselves. The specimen filled with M30 concrete shows energy absorption variation in a small range as compared with the energy absorption variation filled with M50 concrete. The strength index (SI) for each column specimen was calculated and SI value was correlated with the location of the failure point on the specimen. The post-peak behaviour of the TWCFST column specimen is very important. The specimen should show ductile behaviour during the failure process. The load-displacement curves indicate that the TWCFST columns can take the load at 10 mm and 15 mm displacement upto a satisfactory level.