Composites are known to display a considerable amount of scatter in their m
aterial properties due to a large number of parameters associated with thei
r fabrication and manufacturing processes. In the present study, the materi
al properties have been modeled as random variables for accurate prediction
of the system behavior. Shear deformation effects have been incorporated i
n the governing equations. First-order perturbation technique has been empl
oyed to obtain the second-order buckling load statistics. The results have
been presented for composite cylindrical panels with all edges simply suppo
rted. These results demonstrate the dependence of scatter in buckling loads
on the basic random variables. The effects of side-to-thickness ratio, asp
ect ratio, curvature-to-side ratio and change in standard deviation of inpu
t random variables have been investigated for cross-ply symmetric and antis
ymmetric laminates. The approach has been validated by a comparison of the
results with those obtained with Monte Carlo simulation. The results for th
e mean buckling load with different shear deformation theories have also be
en compared with those available in the literature. (C) 2001 Elsevier Scien
ce Ltd. All rights reserved.