MLALookUP – a new software for geometallurgical and process-oriented analysis of MLA data

A new software, “MLALookUP”, for the analysis of geometallurgical data is demonstrated. It provides advanced visualization, integration of data from multiple sources, virtual process emulation, and data comparison.
Advanced visualization includes modal mineralogy, element deportment, virtual sieving, recovery curves, grade curves, etc. for various processing models. In this way promising processing machines and reasonable property thresholds can be identified.
The software allows combining data from various sources: automated mineralogy (e.g. from Mineral Liberation Analysis (MLA)), mineralogical compositions (e.g. from X-Ray powder diffraction) and elemental compositions (e.g. from wet geochemistry).
MLALookUP also allows to virtually “process” the material by emulating various processing machines. The emulation can separate particles based on various properties like size, density, magnetic and shape properties, and surface liberation. In this way it can emulate various processes from classification to flotation. It can handle ideal property based separations as well as imperfect separations, with property dependent recovery probabilities. A simple milling model allows including comminution.
Various processing machines can be combined in a sequence to model the effect of complete processing plants. In this way we can predict recovery, grade and mass streams, before actual processing experiments are need. This allows to play with processing parameters and to define appropriate machines, separation parameters milling target sizes, and promising processing routes. As MLALookUP can compute approximate liberation and particle properties the geometallurgical properties of the material is taken into account during this virtual processing.
Finally the software allows comparing all the different datasets in parallel graphics. This e.g. allows comparing emulation result with experimental data, which might either be given as another MLA dataset or as compositional data from other sources.
It is intended to provide more accurate simulation methods in later versions of the software.