Community Research and Development Information Service - CORDIS

FP5

Quality assurance of computational tools for dosimetry (QUADOS), Final report, (summary)

Project ID: FIGD-CT-2000-20062
Funded under: FP5-EAECTP C

Abstract

This Concerted Action continued the work of a previous Concerted Action that delivered among other deliverables a review and a questionnaire and held one meeting as a workshop on sensitivity analysis and assessment of uncertainties in numerical dosimetry. The quintessence of what was learnt in these actions was that not only the correct use of computational tools is necessary to meet the needs of modern radiation protection but also a careful sensitivity analysis and, based on this, a consistent and reliable stating of the uncertainties associated with the results of such calculations.
It is astonishing to realise that virtually no scientist will accept experimental results unless the uncertainties associated with these results are stated. But the uncertainties associated with results of calculations are rarely stated! The main tool of computational dosimetry is Monte Carlo transport. Since Monte Carlo techniques use stochastic sampling, uncertainty is associated with all results obtained using them. Especially from the questionnaire it was learnt that users of such codes often do not state these uncertainties and that only very few care about the propagation of the uncertainties that are associated with the input parameters to the computational models that underlie such Monte Carlo calculations. This is insofar surprising as modern transport codes, e.g. MCNP, do provide means for accounting for the major contributions to the combined uncertainty.

Download application/pdf (181089)

Record Number: 8881 / Last updated on: 2007-12-17
Category: PRJS