Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 91
PROCEEDINGS OF THE TWELFTH INTERNATIONAL CONFERENCE ON CIVIL, STRUCTURAL AND ENVIRONMENTAL ENGINEERING COMPUTING
Edited by: B.H.V. Topping, L.F. Costa Neves and R.C. Barros
Paper 15

Finite Element Simulation of Slender Thin-Walled Box Columns by Implementing Real Initial Conditions

L. Pavlovcic1, U. Kuhlmann2, B. Froschmeier2 and D. Beg1

1Faculty of Civil and Geodetic Engineering, University of Ljubljana, Slovenia
2Institute for Structural Design, University of Stuttgart, Germany

Full Bibliographic Reference for this paper
L. Pavlovcic, U. Kuhlmann, B. Froschmeier, D. Beg, "Finite Element Simulation of Slender Thin-Walled Box Columns by Implementing Real Initial Conditions", in B.H.V. Topping, L.F. Costa Neves, R.C. Barros, (Editors), "Proceedings of the Twelfth International Conference on Civil, Structural and Environmental Engineering Computing", Civil-Comp Press, Stirlingshire, UK, Paper 15, 2009. doi:10.4203/ccp.91.15
Keywords: slender columns, stability, buckling, tests, finite element analysis, numerical simulations, initial imperfections, residual stresses.

Summary
This paper deals with the numerical simulation of tests on slender thin-walled box columns, susceptible to instability of both types: to global Euler buckling as well as to local buckling of steel plates. Results of full-scale tests served for the verification of numerical models, which allowed detailed study of different initial imperfections and their combinations.

In the first part the paper briefly presents different tests and their results. Eight full-scale tests were carried out with different global slenderness of welded and cold-formed columns subjected to centric and eccentric compression. The columns failed as planned: shorter centrically loaded columns in combined local and global buckling and all the rest in more pronounced global buckling mode. For the purpose of profound numerical simulations, material properties were also tested, and the initial column geometry and residual stresses were carefully measured. It is worth pointing out that very sensitive measurement of residual stresses appeared to be successful, as it resulted in clear stress distribution for welded and for cold-formed type of profiles.

The finite element simulation of tests was carried out according to nonlinear large displacement theory using the ABAQUS software [1]. Two types of elastic-plastic material were modelled based on measured material properties. Special attention was paid to careful implementation of all types of measured initial imperfections: residual stresses, geometric imperfections and column inclination. The results of the finite element simulations show good agreement with the test results: agreement in load-carrying capacities was reached within only 6% and the deformations agree in both shape and amplitudes.

On verified numerical models the sensitivity of column resistance to each particular measured imperfection and their combinations was studied. Among the different imperfections the residual stresses appeared to have prevailing impact, reducing column capacity up to 37%. The major effect of geometric imperfections consists of global column bow rather than of local imperfections of plates. In the case of centric tests the complete imperfection combination reduces the resistance up to 45% regarding initially perfect column. With the increase of designed load eccentricity, the imperfections have a smaller influence (reduction by 11 to 19%), since the load eccentricity, i.e. the introduction of additional bending moment, reduces the column elastic resistance markedly already by itself. From these results it can be concluded that for the accurate numerical simulation of elements in compression it is unavoidable to implement different initial imperfections very carefully based on accurate data, provided from different accompanying tests and measurements.

References