Practical Advanced Design Considering Random Distribution of Initial Geometric Imperfections

• Published in: Advances in structural engineering Vol. 14; no. 3; pp. 379 - 389
• Main Authors: Zhang, Yaochun, Jin, Lu, Shao, Yongsong, Zhao, Jinyou
• Format: Journal Article
• Language: English
• Published: London, England SAGE Publications 01.06.2011
• Subjects: Computer simulation; Defects; Deflection; Design engineering; Frames; Initial geometric imperfections; Monte Carlo methods; Ultimate tensile strength; advanced design; initial geometric imperfection; Monte Carlo method; normal distribution; random combination
• ISSN: 1369-4332; 2048-4011
• DOI: 10.1260/1369-4332.14.3.379

Current advanced analysis and design of steel frames, allows for the most unfavorable combination of initial imperfections. In fact, however, the initial geometric imperfections and their combinations are all random distributions. Recognising this situation, the sway installation tolerances and bow imperfections of both columns and beams are all considered to be the randomly distributed and simulation analyses for imperfections are made using Monte Carlo sampling method in this paper. A large number of parametric studies on twelve-storey single-bay frames with random imperfections have been carried out using the finite element method. Results indicate that both lateral drift values and girder deflections of the frames are distributed normally in the probability statistics terms. The characteristic values of these variables based on many simulation analyses, are extremely close to the results obtained for ideal frames which have on initial imperfections, so that the effects of initial imperfections can be neglected in the advanced design for similar structures. Moreover, a practical advanced design method is proposed, in which the serviceability state of the structure is first investigated under the required characteristic load combinations, and then the ultimate strength capacity is checked under the fundamental load combinations.