Frequency-Based Early Crack Detection and Damage Severity Measure in Structural Glass Members: Application to Beams in Bending

Abstract

As is known, the propagation of cracks in a structural member is associated with local modifications of rigidity and thus possible major effects that can strongly affect its load-bearing capacity and dynamic parameters. As such, among others, the natural frequency control represents a consolidated and efficient approach for damage detection, damage severity assessment, and in situ monitoring for several structural typologies. In this paper, attention is given to typically brittle-in-tension structural glass members and to the analysis of first-crack initiation in terms of natural frequency decrease. The goal is to assess the potential and feasibility of a standardized approach for in situ structural health monitoring (SHM) assessment. To this aim, the investigation takes advantage of a literature analytical model for frequency assessment (i.e., as a function of crack position and depth), and of finite-element (FE) numerical simulations carried out to predict the cracked vibration frequency of various configurations of technical interest. As a first step of this possible methodology assessment, glass beams under an in-plane bending setup are considered, given that they are largely used as stiffeners or fins. It is shown that while glass material is typically brittle in tension and cracks can originate from edges due to several reasons, traditional frequency-based monitoring tools can be efficiently adapted for early detection and to quantify damage in existing glass structures. For the examined configurations, it is shown that frequency reductions up to ≈−20% can be expected due to first-crack initiation. Most importantly, the FE numerical analyses show that crack shape/geometry can further emphasize the expected frequency decrease, and thus additionally enforce the need of specific protocols/performance indicators for SHM purposes in existing structures, as well as for the optimal design of new systems.