Reliability assessment of existing transmission line towers considering mechanical model uncertainties

Resumo

Recent studies have shown the importance of the bolt slippage effect on the structural behavior of
transmission line (TL) towers (CIGRE-187 (2009), Ramalingam e Jayachandran (2016), Jiang et al. (2017), just
to name a few). However, these effects are not included among the studies regarding the structural reliability of
these structures considering model uncertainties. Additionally, the bolt slippage effect has also been disregarded
in the structural design, which is carried out through commercial computational packages that employ a linear or
geometrically nonlinear elastic analysis. Moreover, the magnitude of these effects is highly dependent on the tower
topology, a definition influenced by the engineer experience. Thus, neglecting those effects can lead to a wrong
choice of topology configuration. In these cases, the tower elements that are originally designed to resist to axial
forces determined through structural analysis procedures that do not include the bolt slippage effect, indeed, can be
actually subjected to higher values due to this inaccurate modeling (i.e., they present smaller cross sections that
they should have). This fact is even more relevant when old structures are considered, which were designed when
studies on this subject were not well established. Consequently, these structures are especially prone to present
topologies which increases the influence of the bolt slippage effect, resulting on a structural behavior further away
from a linear elastic hypothesis, originally employed on its design. Therefore, given the high number of old towers
still operating in Brazil and its relevance for the entire electric system, the main goal of this paper is to assess the
structural reliability of existing transmission towers considering mechanical model uncertainties, on a design region
(for pre-collapse loads). For this, the behavior of the connections and modeling assumptions are be considered as
uncertainties. Thus, allowing, for example, to analyze the reliability of these structures considering different levels
of modeling complexity.