On 8–10 May, the University of Minho (Guimarães, Portugal) held the 20th International Probabilistic Workshop (IPW-2024), which was attended by Nikolay Kozak, PhD of Engineering Sciences, senior lecturer at the SPbGASU Department of Roads, Bridges and Tunnels. He presented the results of his recent research on numerical assessment of the reliability of bridge spans taking into account the accumulation of fatigue damage in the elements of the slab and beam combination.

Nikolay Kozak started working on this topic in 2021 during his PhD internship at the University of Minho, and then continued in 2023 with the support of SPbGASU as part of a grant competition for research and teaching staff.

• In total, the conference featured 50 reports and lectures from participants from 17 countries. Most of the participants were from the host country - Portugal, with Italy and Germany also having significant representation. We asked the scientist to talk about his participation in the conference and his work.

– A distinctive feature of the workshop is the inclusion in the program of “key lectures” on basic fundamental topics from recognized experts in the field. This was the case at IPW 2024: in addition to the standard presentation sessions where participants shared the results of their research, there were four “key lectures” from recognized professors. Considering that there were many young researchers (masters, PhD students, postdoctoral students) at the conference, this was very useful. Feedback is also important - professors had the opportunity to discuss the reports of young researchers, give their recommendations and parting words, said Nikolay Kozak.

– What are probabilistic methods?

– We are talking about methods and tools that make it possible to numerically assess the reliability of a structure, taking into account all the initial unknowns. For example, if we know the distribution law of possible load values on a beam and the distribution law of possible material resistance, then we can numerically estimate the probability of structural failure equal to, say, 1%. But, of course, the issue of selecting adequate models for the distribution of loads and resistances is very complex - this was one of the topics of discussion at the workshop. For building structures, where the probabilities of failure are small, the issue of calculations itself is also important, since the simplest method (Monte Carlo simulation) requires a very large number of calculations and is therefore ineffective.

– How significant is this line of research?

– The use of probabilistic methods provides a more flexible approach to determining reliability coefficients when calculating structures, which, in turn, makes it possible to increase the efficiency of using materials and at the same time be more confident in the reliability of the designed structures. Despite significant developments in this area by Soviet researchers (N.S. Streletsky, A.R. Rzhanitsyn), who, in general, predetermined all modern methods used, in our country these tools have not received significant distribution in the engineering environment. Although in reality their potential is very great! My dissertation work was devoted to the related topic of structural endurance, and now, in its development, I am working on the use of probabilistic methods to solve problems of taking into account the fatigue of building structures.