Uncertainty of measurement and lower detection limit of track etched detector systems: experimental verification and consequences for intercomparison experiments

Instead of using three times the empirical standard deviation, SO, of the background density, this approach makes use of the random uncertainty of measurement relative to track density and the definition of the lower limit of detection, HLLD. In addition to SO values, this takes into account (a) the corresponding single counting areas of the background detectors, (b) the total counting area of the background as well as field detectors used for the actual measurement of the quantity of interest, (c) the desired confidence level, and (d) the response of the detector for the quantity of interest. In agreement with experimental results, the S value also includes contributions from systematic uncertainties. In order to reduce the HLLD value of an etched track detector system, the so-called ''counting factor'' introduced here may be used to optimise the number of background and field detectors. The ''detector quality factor'' was found to provide a quantitative criterion for the quality control of CR-39 systems. This approach was utilised for the Eurados-Cendos Intercomparison 1990 of different CR-39 systems. The HLLD values used here for a normalised counting area of 1 cm2 improve the comparison of systems and are recommended for future intercomparisons. In addition, the coefficient of variation relative to neutron dose is presented for typical KfK systems using CR-39 and Makrofol plastics. (Author)