Abstract

Steel beams often require reinforcing while they are under load. This may be due to inappropriate design, defective constructions, structure aging, additional bearing requirements, material deterioration, or accidental damage. A common method for reinforcing steel members is by attaching a steel cover plate onto the existing structure by welding. This strengthening technique can increase the stiffness and strength and can also change the structural behaviour and failure mode of the strengthened steel beam. Very limited research has been conducted on the strengthening of steel beams while they are in service. This research presents a Finite element (FE) based study on steel I-beams welded with steel cover plates while under load. The details of the development of the FE models are presented and the developed models are validated by comparing them against available experimental test results of steel beams reinforced while under load. With the validated FE models, a series of steel I-beams reinforced with steel cover plates at the bottom flanges are analyzed. The considered failures for these beams are cross-section yielding and lateral-torsional buckling limit state flexural resistance. The behaviour and ultimate capacity of simply supported I-beams subjected to positive moment; and continuous-span I-beams subjected to negative moment by welding a cover plate while under load are studied. For the simply supported I-beams, the effects of different parameters, such as residual stress patterns in the I-beam and cover plate, welding residual stress, type of welding patterns, and the difference in steel grades between the I-beam and reinforcing plate, on the behaviour of the steel I-beams reinforced while under load are investigated numerically. FE analysis shows that with increased preload, the capacity of the I-beam reinforced under load reduces when the failure mode of the beam is lateral-torsional buckling (LTB). On the other hand, the variation of the preload has an insignificant effect on the behaviour and ultimate strength of the reinforced beam when the reinforced beam fails in flexural yielding. Moreover, the flexural capacities of reinforced simply supported I-beams with welded cover plates obtained from FE analyses are compared with the capacities predicted by the American (AISC 360-16) and Canadian (CAN/CSA-S16-19) steel design standards. FE analysis shows that AISC 360-16, when the effect of loading height is considered, can reasonably predict the capacity of simply supported I-beams reinforced with welded cover plate at the bottom flange.
In addition, the effects of welding heat, welding sequence, and weld length on the residual welding deformation and behaviour of simply supported steel I-beams reinforced while under load are investigated by considering welding procedure simulation. It is observed that an appropriate welding sequence and weld length can reduce the residual lateral deformations induced from welding a reinforcing plate to the bottom flange of the preloaded I-beam and thus control the unfavorable welding effects. Based on the analyses, a welding segment length of L/9, where L is the length of the beam, is recommended for practical applications. In addition, the effects of initial geometric imperfection and preload level on the welding residual deformation and the behaviour of the reinforced beams are studied numerically. FE analysis shows that the direction and magnitude of initial geometrical imperfection can change the value and direction of the residual deformation resulting from welding.
Finally, the numerical study includes the preloaded steel I-beams reinforced with steel cover plates welded to the compression flanges of the continuous-span beams. Three-point loading condition is considered to simulate continuous span bridges. FE analyses show that adding a cover plate to each span of the beam can increase the ultimate capacity and stiffness of the beam. Also, the reinforcement can prevent the beam from lateral torsional buckling failure mode and the beam can reach its capacity. Furthermore, similar to simply supported beams, the preloading level has an insignificant effect on the behaviour and ultimate capacity of the continuous-span beam reinforced with a cover plate welded to the compression flange.