A STUDY OF OPTIMIZING GEOMETRICAL DIMENSIONS OF THE MARINELLI BEAKER USING A DIFFERENTIAL EVOLUTION ALGORITHM COUPLED WITH MCNP4C2 COMPUTATIONAL CODE

Abstract

The Marinelli beakers with large volume have been widely used to contain samples in measurements of low-activity environmental samples by gamma spectrometers using high purity germanium (HPGe) detectors due to the solid angle of gamma beam from the radiation source to the detector to be wide. In this work, we have developed a computational methodology for optimizing geometrical dimensions of the Marinelli beaker based on the differential evolution algorithm couple with MCNP4C2 code. The selected sample was the radiation solution of ¹³¹I having the energy line of 0.364MeV and the main component to be water with volume of 450cm³ and density of 1g/cm³. The differential evolution algorithm was used to search the geometrical dimensions of the Marinelli beaker including the upper radius, r₁ and the lower height, h₂ through maximizing the detector efficiency computed by the MCNP4C2 code for that geometrical dimensions, respectively. After over 100 generations, the search algorithm was focused with the maximum efficiency of 0.041185 for the Marinelli beaker model having the upper radius, r₁ equal to 5.4941cm and the lower height, h₂ equal to 6.1042cm. Comparing with the genetic algorithm, the differential evolution algorithm showed that there was the outstanding advantage in optimization searching. Within the error range of measuring instrument, the optimization geometrical dimensions obtained from this work agrees with our previous results.