Dynamic image analysis is one of the novel approaches for measurment of particle size and is complementing more established methods like sieve analysis, static image analysis, laser diffraction and scanning electronic microscope. The purpose of this master thesis is to evaluate the ability of dynamic image analysis in determining pellet film coat thickness and comparing the results of before mentioned methods. With image analysis we are analysing three-dimensional objects with two-dimensional projections from which different size parameters can be calculated. Evaluation of particle shape on different size parameters showed no significant difference between them when analysing spherical particles. However, analysing irregular shaped particles minimal Ferrets diameter and minimal chord length gave best results with lowest relative standard deviation compared to diameter of circle with equivalent area and maximum Ferrets diameter. Secondary sampling with chute sampler and sampling with spatula after mixing the primary sample, showed best representativness of the sample. For determining the thickness of pellet film coating, larger samples within 4 g and 10 g resulted in more representative samples with relative standard deviation of less than 0,7 %, while smaller sized samples of 200 mg are enough for size estimation. Thickness of film coating, in theory, is equal to difference in size of starting cores and end pellets. However this simple mathematical approach is suitable for determination of thick coatings. For thinner coatings a model with linear regression of samples with different weight gain is a better approach. Results of the model showed differences ranging between 0,6 μm and 2,6 μm for coat thickness of less than 17 μm compared to results of scanning electron microscope. The model takes into account primary sampling error, secondary sampling error and error of measurement device, therefore it is superior to the simple approach mentioned above. The analysis results of selected samples with sieve analysis, static image analysis, laser diffraction and scanning electronic microscope were comparable to dynamic image analysis. However, for determining pellet film coating thickness only laser diffraction and scanning electronic microscope were accurate enough.
|