No abstract

The paper considers an approach to the probabilistic analysis of the reliability of flat trusses based on p-boxes (probability boxes, p-boxes). Modeling of stochastic parameters in the form of p-blocks is justified for building pavement structures due to significant variability of climatic loads, variations in the physical and mechanical properties of coating materials, installation tolerances and other uncertainties. The advantage of this method is the possibility of using it with incomplete (limited) statistical information - when it is difficult to determine the probability distribution law or the parameters of a random variable. Variants of constructing p-blocks are illustrated for various types of incompleteness of statistical information: for an unknown distribution law using Chebyshev's inequality, for interval estimates of the parameters of random variables, etc. Information is given on the possibility of performing algebraic operations on p-blocks. The probability of no-failure operation with such approaches will be presented as an interval of values. If the interval is too wide (uninformative), the quality of statistical information should be improved by conducting additional tests. The paper presents mathematical models of limiting states taking into account the variability of the basic random variables. The possibility of using the proposed approach in the framework of most practical problems in the construction industry for assessing the safety of statically definable farms is shown. As a result, a formula is given for assessing the reliability of a truss as a conditional mechanical sequential system (in terms of the theory of reliability), taking into account the lack of information about the dependence of its elements. The algorithm for analyzing reliability is considered on a numerical example. The developed approach can be used for other types of statically definable hinge-rod systems.

No abstract