Optimization of the radiation shielding capabilities of bismuth-borate glasses using the genetic algorithm

Glasses in the series, 49B(2)O(3)-25SiO(2)-(26-x)Bi2O3-xBaO (with x between 2 and 8 wt%) are considered for radiation shielding applications based on the gamma-ray half-value layer (HVL) and mass neutron removal cross-section. Theoretical data for the calculation of the HVL is obtained from the NIST XCOM software. The calculated HVL shows good correspondence to the experimentally determined values. The mass neutron removal cross-section and the half-value layer of the glass system are optimized using the genetic algorithm. The genetic algorithm is used alongside regression analysis to determine the relative concentration of each oxide needed to obtain the best radiation shielding parameter values. The use of numerical methods is employed to ascertain simple functions that can be used instead of traditional analytical functions in modeling the behavior of glasses. The auto-Zeff software is used to generate a plot of the effective atomic number over a spectrum of photon energies, while spectroscopic techniques are used to examine the structure and optical properties to determine what advantages the glasses have over concrete or water. This approach had fast convergence due to increased numerical stability. The optimal solution is equal to barite concrete in gamma-ray shielding and limonite concrete in neutron shielding.