Eur. Phys. J. Plus (2022) 137: 754
https://doi.org/10.1140/epjp/s13360-022-02968-1

Regular Article

MCCALISO: a californium neutron source strength prediction software based on the Monte Carlo method of propagating probability density functions

¹ Atomic Research Division, Department of Science and Technology - Philippine Nuclear Research Institute (DOST-PNRI), Commonwealth Avenue, Diliman, 1101, Quezon City, Philippines
² Nuclear Services Division, Department of Science and Technology - Philippine Nuclear Research Institute (DOST-PNRI), Commonwealth Avenue, Diliman, 1101, Quezon City, Philippines

^a jmjecong@pnri.dost.gov.ph

Received: 4 April 2022
Accepted: 16 June 2022
Published online: 30 June 2022

Abstract

It is vital to estimate a Californium (Cf) source’s time-dependent strength, B(t), prior to its planned containment, deployment, and exploitation. A novel Windows-based software called MCCALISO (available at https://www.pnri.dost.gov.ph/index.php/downloads/software/mccaliso) was created that applies the Monte Carlo method (MCM) of propagating distributions, to predict the B(t)’s probability density functions (PDFs) at any given time. This prediction software considers ²⁵⁰Cf, ²⁵²Cf, ²⁵⁴Cf, and ²⁴⁸Cm decay models. Its MCM implementation complies with the Guide to the expression of uncertainty in measurement (GUM) Supplement 1, the international guidance on MCM-based PDF propagation for evaluating measurement uncertainty through a mathematical model. All input quantities in the prediction model for B(t) are assigned with either Gaussian or non-Gaussian PDFs. Any required input quantities that are not user-specified are automatically assigned several default PDF approximation models. The percent differences between MCCALISO and conventional method in B(t) predictions reach 1%, 10%, 100%, and 1000%, at source ages 13 years, 24 years, 35 years, and 45 years, respectively. Moreover, MCCALISO’s potential in predicting B(t) even with little known input quantities was shown adequate to cover more precise predictions where extra input quantities were entered. MCCALISO’s predictions agreed with both the GUM uncertainty framework (GUF) method (based on first-order Taylor series approximation) and the MCM via a Python script in the literature.