Strengthening under Load: The Effect of Preload Magnitudes

Abstract

Strengthening existing structures under load by welding plates is often conducted but little research into this subject was presented. Some researchers claim that the column strengthened under load has the same ultimate capacity as the column welded without preloading. Others disagree and recommend safe design not allowing the steel to yield. The paper presents the parametric study of wide flange columns HEA 100 strengthened under load via two plates parallel to flanges conducted on numerical models validated by experiment. The varying parameters are thickness of strengthening flange, column length, initial bow imperfection, preload magnitude and direction of the axis which is pinned while the other axis is fixed. The experimental research consisted of two columns welded without preload and two columns strengthened under preload ratios (preload magnitude divided by base column ultimate strength) equal to 0.5 (200 kN) and 0.75 (300 kN). All columns were 3 m long and the boundary conditions were determined by knife-edge bearings which ensured pinned supports in the direction perpendicular to the strong axis and fixed perpendicular to the weak axis. All strengthening plates were welded to columns with intermittent welds. All columns failed via flexural buckling. The numerical models used for the parametric study were created in ANSYS software. Design values and procedures recommended by EN 1993-1-5, Annex C were used. The most commonly used steel grade S235 was selected for all plates. Shell 181 element type was used for mapped meshing of steel plates. Element birth and death feature was convenient to simulate strengthening under load. Equivalent initial bow imperfection of the first Eigen buckling shape was included, the model was loaded by force and geometrical and material nonlinear analyses were performed. From provided results some conclusions can be reached. The load under which the column is strengthened slightly weakens the column but not by far to the extent of design recommendations used in the Czech Republic. The reduction ratio (ultimate load capacity of column strengthened under load divided by ultimate load capacity of column welded without preloading) decreases with increasing preload magnitude and column slenderness. Surprisingly, the initial geometrical imperfection, while affecting ultimate capacity greatly, has almost no effect on the reduction ratio.Strengthening existing structures under load by welding plates is often conducted but little research into this subject was presented. Some researchers claim that the column strengthened under load has the same ultimate capacity as the column welded without preloading. Others disagree and recommend safe design not allowing the steel to yield. The paper presents the parametric study of wide flange columns HEA 100 strengthened under load via two plates parallel to flanges conducted on numerical models validated by experiment. The varying parameters are thickness of strengthening flange, column length, initial bow imperfection, preload magnitude and direction of the axis which is pinned while the other axis is fixed. The experimental research consisted of two columns welded without preload and two columns strengthened under preload ratios (preload magnitude divided by base column ultimate strength) equal to 0.5 (200 kN) and 0.75 (300 kN). All columns were 3 m long and the boundary conditions were determined by knife-edge bearings which ensured pinned supports in the direction perpendicular to the strong axis and fixed perpendicular to the weak axis. All strengthening plates were welded to columns with intermittent welds. All columns failed via flexural buckling. The numerical models used for the parametric study were created in ANSYS software. Design values and procedures recommended by EN 1993-1-5, Annex C were used. The most commonly used steel grade S235 was selected for all plates. Shell 181 element type was used for mapped meshing of steel plates. Element birth and death feature was convenient to simulate strengthening under load. Equivalent initial bow imperfection of the first Eigen buckling shape was included, the model was loaded by force and geometrical and material nonlinear analyses were performed. From provided results some conclusions can be reached. The load under which the column is strengthened slightly weakens the column but not by far to the extent of design recommendations used in the Czech Republic. The reduction ratio (ultimate load capacity of column strengthened under load divided by ultimate load capacity of column welded without preloading) decreases with increasing preload magnitude and column slenderness. Surprisingly, the initial geometrical imperfection, while affecting ultimate capacity greatly, has almost no effect on the reduction ratio.