In the matrix simulation of size reduction processes it is important to select the size ratio of intervals in matrices in a proper manner.
Some examples demonstrating how the inappropriate selection of size ratio of intervals affects the calculated results of the size distribution of products were presented.
The error in the estimation of size distribution due to the inappropriate selection of size ratio of intervals was analyzed, divided into the “error of representation” and the “exponential error”, and examined from the standpoint that how they are influenced by the parameters in the selection function or in the distribution function.
That the size ratio of intervals in matrices should be chosen small enough according to the parameters in the selection function or in the distribution function was found to be especially important when the exponent of matrices is considerably large or when “error of representation” has the same sign as “exponential error”.
The adoption of 2
-1 as the size ratio of intervals in matrices has proved very dangerous in most cases, whereas the adoption of values less than or equal to 2
-1/4 much desirable when a high accuracy of estimation is required.
View full abstract