Cyclic performance and strengthening of built-up battened columns

Abstract

Built-up battened columns have been widely used in steel structures mainly because of providing a higher moment of inertia than solid sections with a similar weight. Despite wide application in steel constructions, the seismic design of these columns has not been well addressed in the literature, and seismic design codes do not provide a specific seismic design guideline for them. On the other hand, past earthquakes have shown that built-up battened columns have been vulnerable against seismic actions mainly because of the plastic deformation in battens, fracture of battens, global buckling of columns, local buckling of web and flanges, and formation of plastic hinges at their bottom panel. Therefore, their seismic behaviour should be investigated, and an efficient strengthening method should be proposed. In this study, experimental works and numerical simulations were conducted to investigate the governing failure modes of built-up batten columns. Besides, the effect of batten's thickness, battens spacing, chord distance and axial load on the ultimate load, ductility ratio, stiffness degradation rate and energy dissipation capacity of built-up battened columns were investigated through quasi-static cyclic loading. This study also proposed a strengthening method through the filling of chords with grout and wrapping it with carbon fibre reinforced polymer (CFRP). Experimental works included four unstrengthen and four strengthened columns with different battens spacing and chord distances. Besides, 210 built-up battened columns with different batten thicknesses, battens spacing, chord distances, axial forces, number of CFRPs layers and number of strengthened panels were simulated in ABAQUS software and subjected to cyclic loading. The obtained results indicated that the bulging of chord webs together with the local buckling of chord flanges were the main reason for the failure of columns. Moreover, built-up columns did not reach their plastic moment capacity because of local buckling in flanges. Furthermore, the columns with 62 mm batten spacing showed a 30% larger ultimate load than that of the column with 550 mm batten spacing. The results also indicated that the columns with 62 mm batten spacing reached 95.91% of their theoretical bending capacity. It was shown that design codes" requirements for batten spacing was not conservative and did not result in an identical safety margin for the bending moment capacity of built-up columns. An increase in the chord distance from 50 mm to 150 mm enhanced the lateral strength of the column by 35%. On the other hand, an increase in the axial force from 0.1Fy to 0.4Fy decreased the lateral strength and ultimate displacement by 24% and 36%, respectively. The displacement ductility ratios of the unstrengthen built-up battened columns were less than two even when subjected to an axial compression ratio smaller than 0.2. The results indicated that CFRP application delays/shifts the local buckling of flanges and bulging of the web to the upper un-retrofitted panels; however, an increase in the number of CFRPs layers did not show any pronounced effect. The retrofitting of columns resulted in a significant increase in the lateral strength and corresponding displacement by 32.15% and 39.34%, respectively, as compared to the un-retrofitted columns. The energy dissipation capacity of retrofitted columns was 66.39% higher than that of the un-retrofitted columns. The retrofitted columns lost 27%, while the un-retrofitted columns lost 52% of their initial lateral stiffness at a drift ratio of 5.0%. In addition, the retrofitted columns were also able to reach their plastic moment capacity and had a displacement ductility ratio larger than two. The outcome of this study helps practice engineers to understand the seismic behavior of built-up battened columns better and provides an efficient retrofitting method for these columns.