Title

CLUSTERING-BASED APPROACH FOR OPTIMAL GRINDING SIZE DETERMINATION

CLUSTERING-BASED APPROACH FOR OPTIMAL GRINDING SIZE DETERMINATION

Authorship

FERREIRA, RODRIGO FINA

FERREIRA, RODRIGO FINA

I am this author

DOI

10.5151/2594-357X-40952

Downloads

Abstract

GRINDING IS A SIGNIFICANT ENERGY-CONSUMING PROCESS IN MINERAL PROCESSING PLANTS. ITS PRIMARY OBJECTIVE IS TO ACHIEVE LIBERATION AT THE LARGEST POSSIBLE PARTICLE SIZE FOR SUBSEQUENT CONCENTRATION. WHEN SUCCESSFUL, THIS NOT ONLY SAVES ENERGY, BUT ALSO SIMPLIFIES AND REDUCES THE COST OF SUBSEQUENT SEPARATION PROCESSES BY MINIMIZING THE PRODUCTION OF FINE PARTICLES. HOWEVER, DETERMINING THE OPTIMAL GRINDING SIZE IS CHALLENGING. THE EXPERIMENTAL DETERMINATION OF THE GRINDING TARGET INVOLVES RESOURCE-INTENSIVE LABORATORY PROCEDURES, PARTICULARLY FOR MINES WITH DIVERSE ORE TEXTURES. COMPREHENSIVE TESTING ON NUMEROUS SAMPLES IS NECESSARY TO ACCURATELY CAPTURE THE BEHAVIOR OF THE ORE BODY. TO REDUCE THE LABORATORY WORKLOAD, A CLUSTERING APPROACH CAN BE EMPLOYED. THIS PAPER PRESENTS A PRACTICAL APPLICATION OF A CLUSTERING METHOD, BASED ON A RECENTLY PROPOSED CLUSTERING CRITERION DERIVED FROM SEM-BASED AUTOMATED MINERALOGY DATA, WHICH CONSIDERS TWO COEFFICIENTS FROM THE EXPONENTIAL CORRELATION EQUATION BETWEEN DEGREE OF LIBERATION AND TOP SIZE OF THE SIZE FRACTION, ALONG WITH THE OVERALL DEGREE OF LIBERATION. EXPERIMENTAL RESULTS DEMONSTRATE THAT THIS METHOD EFFECTIVELY GROUPS SAMPLES WITH SIMILAR MINERAL LIBERATION CHARACTERISTICS. BY SIGNIFICANTLY REDUCING THE LABORATORY EFFORT REQUIRED FOR LIBERATION STUDIES INVOLVING A LARGE NUMBER OF SAMPLES, THIS APPROACH PROVES VALUABLE FOR GEOMETALLURGY.

GRINDING IS A SIGNIFICANT ENERGY-CONSUMING PROCESS IN MINERAL PROCESSING PLANTS. ITS PRIMARY OBJECTIVE IS TO ACHIEVE LIBERATION AT THE LARGEST POSSIBLE PARTICLE SIZE FOR SUBSEQUENT CONCENTRATION. WHEN SUCCESSFUL, THIS NOT ONLY SAVES ENERGY, BUT ALSO SIMPLIFIES AND REDUCES THE COST OF SUBSEQUENT SEPARATION PROCESSES BY MINIMIZING THE PRODUCTION OF FINE PARTICLES. HOWEVER, DETERMINING THE OPTIMAL GRINDING SIZE IS CHALLENGING. THE EXPERIMENTAL DETERMINATION OF THE GRINDING TARGET INVOLVES RESOURCE-INTENSIVE LABORATORY PROCEDURES, PARTICULARLY FOR MINES WITH DIVERSE ORE TEXTURES. COMPREHENSIVE TESTING ON NUMEROUS SAMPLES IS NECESSARY TO ACCURATELY CAPTURE THE BEHAVIOR OF THE ORE BODY. TO REDUCE THE LABORATORY WORKLOAD, A CLUSTERING APPROACH CAN BE EMPLOYED. THIS PAPER PRESENTS A PRACTICAL APPLICATION OF A CLUSTERING METHOD, BASED ON A RECENTLY PROPOSED CLUSTERING CRITERION DERIVED FROM SEM-BASED AUTOMATED MINERALOGY DATA, WHICH CONSIDERS TWO COEFFICIENTS FROM THE EXPONENTIAL CORRELATION EQUATION BETWEEN DEGREE OF LIBERATION AND TOP SIZE OF THE SIZE FRACTION, ALONG WITH THE OVERALL DEGREE OF LIBERATION. EXPERIMENTAL RESULTS DEMONSTRATE THAT THIS METHOD EFFECTIVELY GROUPS SAMPLES WITH SIMILAR MINERAL LIBERATION CHARACTERISTICS. BY SIGNIFICANTLY REDUCING THE LABORATORY EFFORT REQUIRED FOR LIBERATION STUDIES INVOLVING A LARGE NUMBER OF SAMPLES, THIS APPROACH PROVES VALUABLE FOR GEOMETALLURGY.

Keywords

Liberation; Ore mineralogy; Cluster analysis.

Liberation; Ore mineralogy; Cluster analysis.