The In-Plane Stability Carrying Capacity Analysis Of Beam-Columns Considering The Post-Buckling Strength Of The Web

The effective width-methods for beam-columns considering post-buckling strength of the web differ greatly in present Chinese codes GB50017-2003 and CECS102:2002. The method in GB50017-2003 doesn't consider the differences of the height-to-thickness ratio of the web, and has no relation to the stress distribution on the sections. So, it seems too roughly. But the one in CECS 102:2002 considers the influences of many parameters, so the method is rather delicate than that in GB50017-2003. It has been found that the ultimate loads from CECS 102:2002 is closer to those calculated by dynamic flab and numerical integral methods than that from GB50017-2003. But the study didn't consider the shearing force, which is not in accordance with some real fact. Besides, the effective width-methods are combined with the formula of stability, but the study doesn't distinguish the differences between the formulas of stability in both GB50017-2003 and CECS102:2002, so there may be some errors finally.To consider the above questions, this paper takes account of simple supported I-section beam-columns with unequal bending moments at two ends as the calculation model. Meanwhile, the formula of stability in GB50017-2003 is adopted only. Moreover, the analysis for the in-plane ultimate load-carrying capacity of beam-columns is carried out by a finite element method. With various parameters of web height-to-thickness ratio, flange width-to-thickness ratio, slenderness ratio of column and load eccentricity, the ultimate loads by analysis results are compared with those by the two approaches. It is found that the effective width-method in GB50017-2003 is rather conservative in most cases, especially when the load eccentricity is smaller, and the method in CECS102:2002 is more rational. For members with large web height-to-thickness ratio, the post-buckling strength varies slightly, so the web should be broadened. With large flange width-to-thickness ratio and slenderness ratio of column, the web doesn't be supported enough and overall instability occurs later, so it has no meaning to consider the post-buckling strength. As the result, it is suggested that when web height-to-thickness ratio exceeds the limit in GB50017-2003 and flange width-thickness ratio and slenderness ratio of column are smaller, the method in CECS102:2002 is more economic and rational to consider the post-buckling strength of the web of I-section beam-columns.