Mean value first order second moment analysis of buckling of axially loaded thin plates with random geometrical imperfections

Abstract

Buckling strength of thin plate structures under axial compression is more dominantly affected by the initial geometric imperfections than the other types of imperfections present in them. Since these initial geometric imperfections are random in nature, the collapse strength distribution will also be random. Hence a probabilistic approach is required for reliable design of
these thin plate structures. In this paper, by keeping the variance of imperfections of all the models at assumed manufacturing tolerance of 1.71 mm and maintaining the maximum amplitude of imperfections within ±8 mm, 1024 random geometrical imperfect plate models are generated by the linear combination of first 10 eigen affine mode shapes using 2k factorial design. These imperfect models are analysed using ANSYS non-linear FE buckling analysis including both geometrical and material non-linearities. From these FE analysis results, the strength distribution of the plate is obtained and reliability analysis is carried out using Mean Value First Order Second Moment (MVFOSM) method.

Keywords: Buckling strength, Thin plates, Geometrical imperfections, Random modeling, Reliability based design, MVFOSM.