Acoustic emission signal processing.

• Ano de defesa: 2019
• Autor(a) principal: Prete Junior, Carlos Augusto
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Acoustic emission; Compressive sensing; Emissão acústica; Ensaios não destrutivos; Estimação esparsa; Localização de fontes; Monitoramento de saúde de estruturas; Nondestructive testing; Processamento de sinais; Signal processing; Source localization; Sparse estimation; Structural Health Monitoring; Tempo de chegada; Time of arrival
• Link de acesso: http://www.teses.usp.br/teses/disponiveis/3/3142/tde-12022020-151915/

Acoustic Emission is a widely used structure health monitoring (SHM) method for monitoring large structures with high sensitivity. When an acoustic source is active (for example, during the expansion of a crack), it emits an elastic wave that reaches the sensors that are spread along the structure surface. The signals sampled by these sensors are processed and used to estimate the source position. In this work, we study acoustic emission techniques based on the time of arrival (TOA) of the signals received by the sensors and develop methods to improve the source position estimate. More specifically, we derive the probability distribution of the TOA measured by the fixed threshold method, a popular TOA estimation algorithm, as well as an expression for its bias and consequently a TOA debiasing method. Moreover, we derive a nearly-optimal TOA-based source position estimator. Algorithms for anisotropic structures are also investigated. In scenarios where multiple sources are active simultaneously, it is important to group signals (hits) from the same source to avoid using signals emitted by other sources in the localization algorithm. For this reason, we develop hit grouping techniques and compare them with existing methods. We also create a source localization algorithm that directly uses the signals received by the sensors instead of TOAs to estimate the source position. This method takes into account the wave propagation model and also the sparsity of the source signal in a known dictionary to improve the localization performance using sparse reconstruction methods. This work was partially supported by EMBRAER, which provided data of actual acoustic emission tests in complex structures.