Desenvolvimento de um modelo empírico para controle de backbreak gerado por desmontes em uma mina a céu aberto

• Ano de defesa: 2024
• Autor(a) principal: Flávio Loyola Tavares
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; ENG - DEPARTAMENTO DE ENGENHARIA MINAS; Programa de Pós-Graduação em Engenharia Metalúrgica, Materiais e de Minas - Mestrado Profissional; UFMG
• Palavras-chave em Português: Desmonte por explosivos; Simulação; Modelagem matemática; Otimização; Engenharia de minas; Tecnologia mineral; Minas e mineração; Mina a céu aberto; Detonação; Rochas; Explosivos; Aderência
• Link de acesso: http://hdl.handle.net/1843/73278

Blasting rocks with explosives is a method widely employed in numerous mining operations to achieve rock mass fragmentation. However, associated with this process are undesired effects that can compromise overall blasting outcomes, such as backbreak. This phenomenon has the potential to generate negative impacts on subsequent processes, particularly concerning safety, fragmentation quality, long-term mining planning, and geotechnics. Recognizing the relevance of this topic, this dissertation aims to: analyze a database from 37 blasting in an open-pit mine to evaluate the correlation between different controllable blasting parameters and backbreak, develop a predictive model based on multiple linear regression to estimate average overbreak, and analyze its applicability for the area under study. For this purpose, the study considers input data such as average burden, average spacing, average depth, maximum charge per delay in production holes, maximum charge per delay in the last row of holes, loading ratio, number of decks in production holes, number of decks in the last row of holes, number of rows, P80, and top size. Simple linear regressions indicated that the blasting parameters with the highest correlation to backbreak are: maximum charge per delay in production holes, maximum charge per delay in the last row of holes, loading ratio, number of decks in production holes, and number of decks in the last row. Furthermore, multiple linear regression analysis shows that these aforementioned variables exhibit a strong correlation with average backbreak when considered collectively. Finally, the developed predictive model demonstrated a strong correlation between the measured and simulated average damage. Additionally, when evaluating a second scenario by excluding blasts with the highest maximum charges per delay, it was observed that the presence of these data positively influences the correlation of parameters with backbreak. However, it is essential to highlight that the model does not consider intrinsic rock mass characteristics, which may negatively impact correlation with certain parameters.