Passive neutron multiplicity counting is widely used as a nondestructive assay technique to quantify mass of plutonium material. One goal of this technique is to achieve good precision in a short measurement time. In this paper, we describe a procedure to derive mass assay variance for multiplicity counting based on the threeparameter model, and analytical equations are established using the measured neutron multiplicity distribution.Monte Carlo simulations are performed to evaluate precision versus plutonium mass under a fixed measurement time with the equations. Experimental data of seven weapons-grade plutonium samples are presented to test the expected performance. This variance analysis has been used for the counter design and optimal gate-width setting at Institute of Nuclear Physics and Chemistry.
A perturbation method is proposed to obtain the effective delayed neutron fraction βeff of a cylindrical highly enriched uranium reactor. Based on reactivity measurements with and without a sample at a specified position using the positive period technique, the reactor reactivity perturbation Ap of the sample in βeff units is measured. Simulations of the perturbation experiments are performed using the MCNP program. The PERT card is used to provide the difference dk of effective neutron multiplication factors with and without the sample inside the reactor. Based on the relationship between the effective multiplication factor and the reactivity, the equation βeff=dk/△ρ is derived. In this paper, the reactivity perturbations of 13 metal samples at the designable position of the reactor are measured and calculated. The average βeff value of the reactor is given as 0.00645, and the standard uncertainty is 3.0%. Additionally, the perturbation experiments for fleer can be used to evaluate the reliabilities of the delayed neutron parameters. This work shows that the delayed neutron data of 235U and 23SU froin G.R. Keepin＇s publication are more reliable than those from ENDF-B6.0, ENDF-B7.0, JENDL3.3 and CENDL2.2.
