Thesis INFLUENCIA DE LA REOLOGÍA DE UN RELAVE EN LA DISTANCIA PELIGROSA MEDIANTE HERRAMIENTAS DE FLUIDODINÁMICA COMPUTACIONAL (CFD)
Date
2021-01
Authors
VERGARA BARRÍA, ÁLVARO FERNANDO
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
La industria minera genera diariamente una gran cantidad de desechos húmedos, conocidos como relaves, los cuales son almacenados en depósitos que deben cumplir la normativa de seguridad para mantener un óptimo control y monitoreo de su estabilidad. Esto debido a que los depósitos de relaves son obras potenciales de fallar, con la capacidad de generar catástrofes medioambientales, económicas y sociales de alto impacto una vez colapsan. En la presente memoria se realiza el modelamiento computacional de relaves como fluidos no newtonianos, utilizando herramientas CFD, en particular FLUENT®, con el objetivo de modelar el comportamiento de estos sistemas como un acercamiento preliminar al estudio de la distancia peligrosa a partir de las características físicas y reológicas de los relaves. Para ello se realiza el análisis dimensional del sistema, considerando las variables más relevantes en el estudio, con el fin de determinar los parámetros adimensionales de interés que permiten describir de manera simple las relaciones funcionales entre dichas variables. Se realiza la simplificación del dominio físico a partir de la creación de un canal rectangular de 30 metros de longitud, sobre el cual el fluido escurre libremente en base a las condiciones de inicialización del modelo. Esta geometría es discretizada en 475.408 celdas poliédricas donde en cada una de ellas se encontrarán las soluciones aproximadas a las ecuaciones de conservación de masa y momentum en los distintos intervalos de tiempo. El flujo se modela en régimen laminar, siguiendo la estructura del modelo de viscosidad de Herschel-Bulkley utilizado para la descripción física de los relaves. Los resultados muestran una muy buena correlación entre las variables, permitiendo deducir una alta influencia de la tensión de fluencia y la pendiente del terreno, reflejados tanto en el número de Reynolds como el coeficiente de fluencia. A partir de ellos, se propone un modelo matemático simple obtenido del análisis multidimensional y un ábaco acotado para intervalos regulares, que apoye como complemento adicional a los modelos simples ya existentes para la estimación de la distancia peligrosa. Con ello, se incentiva el uso de las herramientas CFD para perfeccionar el complejo estudio que conllevan los fluidos no newtonianos.
The mining industry generates a large amount of wet mining waste daily, known as tailings, which are stored in deposits that must comply with safety regulations to maintain optimal control and monitoring of their stability. This given that tailings deposits are potential works to fail, with the capacity to generate high-impact environmental, economic, and social catastrophes once they collapse. In this study, the computational modeling of tailings as non-Newtonian fluids is carried out, using CFD tools, provided by FLUENT®, with the aim of modeling the behavior of these systems as a preliminary approach to the study of the runout distance from the physical and rheological characteristics of the tailings. For this, the dimensional analysis of the system is carried out, considering the most relevant variables in the study, to determine the dimensionless parameters of interest that allow the functional relationships between these variables to be described in a simple way. The simplification of the physical domain is carried out from the creation of a rectangular channel 30 meters long, over which the fluid flows freely from the initialization conditions of the model. This geometry is discretized in 475,408 polyhedral cells where in each one of them the approximate solutions to the mass and momentum conservation equations in the different time intervals will be found. The flow is modeled at a laminar regime, following the structure of the Herschel-Bulkley viscosity model used for the physical description of the tailings. The results show a very good correlation between the variables, allowing to deduce a high influence of the yield stress and the slope, reflected both in the modified Reynolds number and the yield coefficient. Based on them, a simple mathematical model obtained from multidimensional analysis and a bounded abacus for regular intervals is proposed, which supports as an additional complement to the existing simple models for estimating the runout distance. This encourages the use of CFD tools to perfect the complex study involved in non-Newtonian fluids.
The mining industry generates a large amount of wet mining waste daily, known as tailings, which are stored in deposits that must comply with safety regulations to maintain optimal control and monitoring of their stability. This given that tailings deposits are potential works to fail, with the capacity to generate high-impact environmental, economic, and social catastrophes once they collapse. In this study, the computational modeling of tailings as non-Newtonian fluids is carried out, using CFD tools, provided by FLUENT®, with the aim of modeling the behavior of these systems as a preliminary approach to the study of the runout distance from the physical and rheological characteristics of the tailings. For this, the dimensional analysis of the system is carried out, considering the most relevant variables in the study, to determine the dimensionless parameters of interest that allow the functional relationships between these variables to be described in a simple way. The simplification of the physical domain is carried out from the creation of a rectangular channel 30 meters long, over which the fluid flows freely from the initialization conditions of the model. This geometry is discretized in 475,408 polyhedral cells where in each one of them the approximate solutions to the mass and momentum conservation equations in the different time intervals will be found. The flow is modeled at a laminar regime, following the structure of the Herschel-Bulkley viscosity model used for the physical description of the tailings. The results show a very good correlation between the variables, allowing to deduce a high influence of the yield stress and the slope, reflected both in the modified Reynolds number and the yield coefficient. Based on them, a simple mathematical model obtained from multidimensional analysis and a bounded abacus for regular intervals is proposed, which supports as an additional complement to the existing simple models for estimating the runout distance. This encourages the use of CFD tools to perfect the complex study involved in non-Newtonian fluids.
Description
Keywords
REOLOGIA , DINAMICA DE FLUIDOS , FLUIDOS NO NEWTONIANOS