Scatter in KIC-results can often be quite extensive, and to make reliable interpretations of the results it is of great importance to understand the nature of it. Cleavage fracture in steels is of a statistical nature and therefore the scatter in KIC-results will behave similarly.
Two different approaches, one based on a microstructural statistical model and an other based on the Weibull distribution are applied to evaluate the theoretical scatter in KIC-results. With both methods it is shown that the theoretical value of the relative scatter described through the Weibull slope factor is constant and equal to four.
The reason for the discrepancy between the theoretical value and the experimentally determined values of the slope factor is shown to be caused by inadequate number of experimental KIC-measurements. The existence of a lower limiting Kmin value is verified and a simple procedure for conservative estimation of the KIc-mean and lower bound values is presented.