Year
2008
Abstract
Deadtime corrections for passive neutron coincidence counting are traditionally formulated in terms of the Totals counting rate. The deadtime correction is exponential in form with the effective deadtime being linear in terms of observed Totals rate. The deadtime coefficient for the Reals rate is traditionally fixed at four times that of the Totals rate parameter. When it comes to multiplicity counting, however, more complex expressions are typically used for the Doubles and Triples rates based on mathematical actions to the multiplicity histograms with the Singles (or Trigger) rate being treated rather simplistically. Since the Totals & Singles and Reals & Doubles, respectively, are effectively equivalent physical measures, the difference in deadtime treatment results is an inconsistency – albeit a numerically minor one for most practical applications. Furthermore, and more seriously, additional empirical correction factors are often applied in the case of the multiplicity deadtime corrections and these do not follow from the underlying theoretical framework. The purpose of this paper is to re-examine the semi-empirical deadtime correction expressions from a fresh perspective. We propose to a scheme whereby Totals and Singles are treated equivalently with the correction having the transcendental form of the paralyzable model. The impact of correlations on the Totals deadtime correction is shown to be modest. The deadtime correction factor for Reals and Doubles are again treated similarly also using an exponential form in terms of the corrected Total event rate but with a deadtime parameter which is not fixed ahead of time to be four times that used in the Totals correction. In the case of the Triples correction evaluated from the multiplicity data using a composite expression, the deadtime corrections for the Singles and Doubles are used as appropriate but a new empirical correction, again given in terms of the corrected rate, is introduced. The new correction acts only on the part of the Triples expression which is does not represent the ‘correlated-accidentals’. The new scheme is not based on an elaborate mathematical model for deadtime losses and consequently does not involve a deadtime-function-weighted sum of histogram elements. In that sense it is simpler to understand and implement. Determination of the free deadtime parameters is based on first extracting the deadtime parameter for the Singles for which several options are available. The Doubles deadtime parameter is then based on rendering the corrected Doubles to Singles ratio for Cf-252 invariant to trigger rate. Similarly the Triple deadtime parameter is extracted from Cf-252 data given the Singles (or Trigger) and Doubles deadtime parameters. In this article we briefly lay out the new deadtime correction concept and illustrate its application with experimental Pu data taken using a pair of Passive Scrap Multiplicity Counters