Additonal authors: Toro, C.A.. Book title: Proceedings of the 58th Conference of Metallurgists Hosting Copper 2019. Chapter: . Chapter title:
Torres, S. N.
This work presents visible-near infrared (VIS-NIR) spectroscopy measurements, and signal analytical techniques to characterize minerals and relevant pyrometallurgical process variables for the copper extractive industry. Data signal processing techniques such as baseline correction and numerical analysis are used to estimate the continuous and discontinuous spectral components. Such spectral components are used to estimate the process temperature, and patterns such as the integrated radiation. Also, metal combustion emissivity calculation are depicted. Our findings suggest that VIS-NIR spectroscopy may be a suitable technique to assess the process condition state, appearing as an interesting sensing alternative.
INTRODUCTION
Although the flash smelting process is one key stage in the copper extractive industry, representing about 40% of copper matte production worldwide (Cantallopts, 2017), few attention is dedicated to develop sensing solutions to assess process variables, such as mineralogical composition and temperature, in the feeding and bleeding stages, compared to the steel industry (Strakowski, et al., 2014). In this process and other industries, e.g. the thermoelectric power generation one, combustion and particularly the generated flame are the production chain’s heart. Since the flame is a highly radiative emitting process where chemical and physical transformations take place, many works had been dedicated to characterize the combustion state by sensing its radiation using optical and photonic technologies (Ballester & García-Armingol, 2010). In this context, this work aims to contribute with visible-near infrared (VIS-NIR) spectroscopy and pyrometry techniques to retrieve relevant information from a laboratory scale flash smelting process.
On the other hand, since photonic instrumentation allow us to perform contactless measurements of emitted, absorbed, transmitted or scattered radiation, it appears as a suitable technique for sensing process properties in harsh environments, e.g. high temperature processes. One of these techniques is the Visible- Near Infrared (VIS-NIR) spectroscopy, which provides information about the spectral radiation intensity distribution, in this work, the emitted flame’s radiation. It is well known in combustion research that the spectral emission from a flame is composed by a black body continuous radiation background, molecular emissions, elemental emissions and additive noise features (Arias et al., 2011). To make sense of such information signal processing techniques are needed, here we propose to analyse the continuous and discontinuous spectral components separately.
