Estimation of coal density using a 3D imaging system: RhoVol

The results of coal sink and float tests are plotted as washability curves. Analysis these curves indicates the most effective method by which can be cleaned. Sink experiments sampling-dependent, destructive, use toxic chemicals. present research introduces an alternative based on a 3D imaging system called RhoVol, in density individual particles is determined from mass volume measurements. A quantitative error analysis was conducted sample measurements obtained main factors influencing errors identified investigated. show that drawback this technique related to its not being able detect hidden concavities particle, resulting overestimation particle volume. Furthermore, it difficult capture impact porosity cracks silhouette images. This mismatch between RhoVol data test has led attempts apply neural network (NN) linear regression techniques produce fully reliable model for correction estimation density. NN approach offers superior predictive capability over regression, estimated distribution line with analysis. root mean square using less than 0.05 g/cm3.