Reproducibility of TLD systems

With respect to an application in the lower dose range, the statistical error of dose measurement has been investigated as a function of exposure by separating detector and reader dependent parameters such as PM dark current and zero dose reading, irradiation and annealing history of the dosimeter batch, batch uniformity, as well as the evaluation technique applied. The standard deviation vs. exposure curve can be interpreted on the basis of a complex two-parameter fit. Experimental results are discussed with respect to batch and reader dependent parameters. (orig.)