Assessing the safety of vaulted masonry structures using thrust network analysis. (December 2021)

• Record Type: Journal Article
• Authors: Maia Avelino, R.; Iannuzzo, A.; Van Mele, T.; Block, P.
• Is Part Of: Computers & structures. Volume 257(2021)
• Publication Date: 2021-12
• Subjects: Masonry assessment -- Thrust network analysis -- Optimisation -- Equilibrium analysis -- Limit analysis -- Geometric safety factor; Structural engineering -- Data processing -- Periodicals; Electronic data processing -- Structures, Theory of -- Periodicals; 624.171
• DOI: 10.1016/j.compstruc.2021.106647
• Languages: English
• ISSNs: 0045-7949

Abstract

Highlights: TNA assessment of masonry vaults by solving nonlinear constrained optimisations. A direct optimisation to find the minimum thickness, and the GSF, of masonry vaults. Robustness of vaulted structures through stability domain analyses. A numerical approach for the stability of vaults with geometry from point clouds. A new SDV-based interactive algorithm to select independent force densities. Abstract: This paper presents a methodology to assess the stability of vaulted masonry structures using Thrust Network Analysis (TNA). It offers a new numerical strategy to compute the Geometric Safety Factor (GSF) of a given structure by directly evaluating its minimum thickness. Moreover, it provides an approach for tracing the vault's stability domain based on its extreme thrust values, which indicates the robustness of the structure. Together, these outcomes represent a proper measure of the safety level of masonry structures. Such results are obtained from constrained nonlinear optimisation problems (NLPs) with appropriate objective functions and constraints enforcing the limit analysis' admissibility criteria. Networks with fixed horizontal projection are considered, for which the spatial geometry is a function of the independent force densities and the height of the support vertices. A faster, interactive procedure is proposed to improve the selection of such independent force densities. The range of applications of the present method includes arbitrary network topologies and different support conditions. Beyond analytically described geometries, the method can deal with geometries obtained numerically (e.g. from point clouds). The presented strategy is implemented in a Python-based package, and relevant applications illustrate the method's potential in assessing the stability of three-dimensional historic vaulted structures.