ON LINE CONTROL OF APPARENT DENSITY - COMPUTER-SIMULATION AND ANALYSIS

A description is given of a new approach for the quality control of me tal powder production which employs computer based simulation and imag e analysis of irregular peptides. In order to predict the packing dens ity for particles with a given degree of irregularity computer simulat ions ore employed to develop models describing the morphology and pack ing behaviour of irregular particles. Vision system analysis of partic le images generated microscopically is utilized for evaluation of part icle irregularity for a wide range of powders, Examples include partic les produced by the water atomization of iron, These values for irregu larity are then substituted into powder packing models to generate pre dicted values for powder packing density, A facility for on-line real time monitoring of powder density has been formalised and designated a s simulation and vision analysis, in order to validate the practical u tility of the proposed methodology, it was applied to samples of comme rcial iron powder and the results compared to densities measured exper imentally for the powders, The regression correlation coefficient obta ined (0.96) provides confirmation of the validity of analysis for pred icting the packing densities of certain types of irregular powders, Th is facility, in combination with knowledge based systems and statistic al process control methods offers potential for intelligent control of powder production.