This paper presents a method for computing an info-gap robustness function of
structures, which is regarded as one measure of structural robustness, under
uncertainties associated with the limit load factor. We assume that the external load
in the plastic limit analysis is uncertain around its nominal value. Various
uncertainties are considered for the live, dead, and reference disturbance loads based
on the nonstochastic info-gap uncertainty model. Although the robustness function is
originally defined by using the optimization problem with infinitely many
constraints, we show that the robustness function is obtained as an optimal
value of a linear programming (LP) problem. Hence, we can easily compute
the info-gap robustness function associated with the limit load factor by
solving an LP problem. As the second contribution, we show that the robust
structural optimization problems of trusses and frames can also be reduced to LP
problems. In numerical examples, the robustness functions, as well as the robust
optimal designs, are computed for trusses and framed structures by solving LP
problems.
Keywords
data uncertainty, linear program, plastic limit analysis,
robust optimization, info-gap analysis