Image analysis tools for the assessment of carbon anodes : a thesis presented in fulfilment of the requirements for the degree of Master of Technology in Manufacturing and Industrial Technology at Massey University
The energy efficiency and performance of an aluminium smelter depends critically on the quality and consistency of properties of the carbon electrodes that are consumed during the normal operation of the electrolytic cells or "pots". Unfortunately, although a small number of experts are able to assess anode quality by examining l0x images of samples, no objective method exists for making quality determinations. This thesis is about a project that has the goal of developing such an objective method. This thesis describes methods that have been developed for the characterization of the microstructure of carbon anodes. As a result of the process by which they are manufactured, carbon anodes contain pores or voids caused by out-gasing. In this continuing project we have concentrated on developing means for characterising the size and spatial distributions of these voids. Some of the methods used to characterise the spatial distribution include order neighbour analysis (a method used in geographical studies), and statistical texture analysis. These methods and the analysis described in this thesis are of general application.