Stability behavior of multi-celled corrugated-plate concrete filled steel tubular walls

TONG Jingzhong,YU Chaoqun,TONG Genshu,CHEN Ming,ZHANG Lei
DOI: https://doi.org/10.14006/j.jzjgxb.2021.0834
2023-01-01
Abstract:In order to reduce the consumption of steel material, improve the load-resistant efficiency and avoid excessive use of shear connecters in steel-concrete composite walls, the multi-celled corrugated-plate concrete filled steel tubular wall(MCCFSTW) was proposed. An MCCFSTW consists of CFSTs and double-corrugated steel plates filled with concrete. Six MCCFSTW specimens with a height-to-depth ratio of 20.0 or 16.7 were tested to study the stability performance of the walls, in which the corrugated steel plates were trapezoidal or sinusoidal corrugated. The results indicated that the ultimate resistance of all specimens was mainly attributed to their overall stability. The global buckling failure could be observed accompanied by local buckling of steel plates, while the welding seams between the corrugated plate and the steel tube remained undamaged. It was concluded that reducing the height-to-depth ratio of the wall could effectively improve its stability resistance. A refined finite element(FE) model was established, which well predicted the stability performance of MCCFSTWs under axial compression.A parametric study was conducted to investigate the effect of height-to-depth ratio, cross-sectional dimensions, concrete strength and corrugated plate shape on the stability performance of MCCFSTWs. The results show that the effect of steel plate thickness and width of CFST on the ultimate rosistance of MCCFSTWs decreases with the increase of height-to-thickness ratio, while the effect of corrugated cell width and concrete strength on the ultimate resistance of MCCFSTWs become important with large height-to-thickness ratio. Besides, the thickness and shape of corrugated plate have little effect on the ultimate resistance of MCCFSTWs. Finally, a design formula was proposed for practical design, and the ratios of numerical to formula results were basically within 1.00-1.10, which could safely estimated the test results.
What problem does this paper attempt to address?