<p>This paper develops a unified method to predict the compressive strength of fixed-ended lipped channel columns. Using numerical data from different authors, novel design curves are calibrated for isolated buckling modes failure—local (L), distortional (D|) and global (G) buckling – as well as L-D and L-D-G interactive failure modes. The basis of the method is similar to the Direct Strength Method’s L-G treatment, in which the resistance related to each buckling mode affects the slenderness rate of the next mode to be verified. The main difference are the order of the modes for the calculations, from the mode with most post-buckling reserve (L) to the mode with least post-buckling reserve (G), and the inclusion of the distortional mode in the process. The proposed design curves are then compared to a large data of failure loads from numerical simulations. The preliminary results for 1493 numerical columns show that the proposed set of equations are reliable (average results for <i>f</i><sub><i>u, numerical</i></sub><i>/f</i><sub><i>proposed</i></sub> are 1.014 ± 0.067, resistance factor <i> ϕ</i> = 0.914) and provide a unified procedure to obtain the ultimate strength of columns undergoing different failures modes.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Towards a General Design Method for Thin-walled Lipped Channel Columns

  • Guilherme C. de Salles

摘要

This paper develops a unified method to predict the compressive strength of fixed-ended lipped channel columns. Using numerical data from different authors, novel design curves are calibrated for isolated buckling modes failure—local (L), distortional (D|) and global (G) buckling – as well as L-D and L-D-G interactive failure modes. The basis of the method is similar to the Direct Strength Method’s L-G treatment, in which the resistance related to each buckling mode affects the slenderness rate of the next mode to be verified. The main difference are the order of the modes for the calculations, from the mode with most post-buckling reserve (L) to the mode with least post-buckling reserve (G), and the inclusion of the distortional mode in the process. The proposed design curves are then compared to a large data of failure loads from numerical simulations. The preliminary results for 1493 numerical columns show that the proposed set of equations are reliable (average results for fu, numerical/fproposed are 1.014 ± 0.067, resistance factor ϕ = 0.914) and provide a unified procedure to obtain the ultimate strength of columns undergoing different failures modes.