Measurement of dose equivalent rate in mixed fast neutron-gamma ray fields.

Abstract

The neutron energy response function of a BC-501A liquid scintillator was measured by a time of flight method, and the energy dependent intrinsic efficiency was estimated. Pulse shape discrimination was used to distinguish neutron and gamma ray interactions. The convolution equation relating the measured neutron pulse height spectrum and the incident neutron energy spectrum was posed as a system of 128 linear equations, with 80 variables. To solve the ill-conditioned system, a truncated singular value decomposition (TSVD) and a modified TSVD (MTSVD) were employed. Pulse height spectra were acquired for a $\sp{252}$Cf and $\sp{239}$PuBe sources and unfolded using the TSVD and MTSVD methods. For the cases studied, a truncation parameter k between 6 and 10 was found to be reasonable for both methods. Unfolded spectra were in good agreement with the best fit to the spectrum possible using the chosen method and truncation parameter. The best possible fit could be improved in a least squares sense by increasing k, but only at the expense of increased noise in the unfolded spectra. The relative error of the unfolded spectra was found to generally decrease with higher numbers of counts, but no clear relationship was found between the relative error and the truncation parameter k. The gamma ray and neutron absorbed dose and dose equivalent were calculated from the measured gamma ray and neutron pulse height spectra of the $\sp{252}$Cf and $\sp{239}$PuBe sources, using the TSVD and MTSVD deconvolution methods. The values of measured neutron dose differed from values calculated for previously reported spectra by 4-12% and 2-8% for the TSVD and MTSVD methods respectively. The values of measured neutron dose equivalent differed from values calculated for previously reported spectra by 5-10% for TSVD method and 9-21% for the MTSVD. For both methods, the neutron dose and dose equivalent calculation were insensitive to the choice of the truncation parameter k for $k > 10.$