ACTIVITY, EXPOSURE RATE AND GAMMA SPECTRUM PREDICTION FOR NEUTRON IRRADIATED MATERIALS AT RADIATION SCIENCE AND ENGINEERING CENTER

Open Access
- Author:
- Sahin, Dagistan
- Graduate Program:
- Nuclear Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- None
- Committee Members:
- Kenan Unlu, Thesis Advisor/Co-Advisor
Kenan Unlu, Thesis Advisor/Co-Advisor - Keywords:
- prediction
java
Monte Carlo
Geant-4
neutron
activity
irradiation
gamma
spectroscopy
neutron activation analysis - Abstract:
- Neutron activation analysis (NAA) is a highly precise and accurate method to determine material composition [1]. NAA has been used widely for research and industrial applications. When a target isotope is irradiated with a neutron flux, then a reaction occurs and secondary products may be emitted from the product isotope if the product isotope is radioactive. These secondary products can be detected using suitable detectors. In order to determine the content of an isotope using NAA, knowledge of reactions that occur when the isotope is irradiated, neutron flux and the gamma ray energies that may be released are required. If this is known, the product isotope, and its content can be determined. It will be beneficial to know the activity, exposure rate in air and the resulting gamma spectrum before performing an irradiation experiment in order to prevent unnecessary radiation exposure. With this study NAA experiments that are performed in Penn State Breazeale Reactor (PSBR) at Radiation Science and Engineering Center (RSEC) were modeled. The programs developed in this study can predict the activities and exposure rates of the irradiated samples, prepare necessary activity prediction reports, and predict spectra that would be obtained using the spectroscopy system. The irradiations of the samples take place at dry irradiation tubes of PSBR. The flux values at these locations are measured using cadmium ratio method with gold-aluminum wires and cadmium tubing. The maximum thermal neutron flux was 9.83x10E13 n/cm2-s and the maximum resonance flux was 5.03x10E11 n/cm2-s. A JAVA [TM] [2] software has been written, called Activity Predictor, which analytically calculates the activity and exposure rates of irradiated samples at a given time after irradiation. The photo peak efficiency and Full Width at Half Maximum (FWHM) was measured for PSBR-RSEC NAA system that utilizes a High Purity Germanium (HPGe) detector. A National Institute of Standards and Technology (NIST) certified standard reference material was used for those measurements. Activity Predictor uses the experimental flux values, half-life and decay data from Lund/LBNL Nuclear Data Search [3] and energy absorption coefficient data for air from Shultis et al., [4]. This software generates a rough gamma spectrum prediction using a very fast quasi Monte Carlo method. Analytically calculated activity values and gamma cross section values from XCOM Photon Cross Sections Database [5] are used for spectrum prediction. Integral counts of the photo peaks are obtained with calculated activity, efficiency of the NAA system, measured neutron flux and nuclear data. Normal distribution with a FWHM dependent variance is applied to generate photo peaks. Integral counts in the Compton continuum are determined with energy dependent Compton scattering to photoelectric absorption cross section ratio for germanium. The energies of the Compton scattered photons are sampled via the Klein-Nishina method. The software was validated with hand calculations and a series of irradiation experiments. Iron, nickel and copper samples were irradiated at dry irradiation tube 1 (DT1) and gamma spectrums were collected. It has been found that Activity Predictor calculates the activities with 0.08-0.4% error, and exposure rates with 2.4-4.7% error compared to hand calculations. The predicted spectra agreed partially in visual manners with 4.8% to 51.3% error in photo peak net area values compared to experimental spectra. To generate a more realistic spectrum, Geant-4, code was used to model the spectroscopy system in detail. The features of the HPGe detector and surrounding lead shielding are modeled. The model was validated with a NIST certified Co-60 source. The energy and efficiency of the spectroscopy system was measured with Co-60 source. The experimental efficiency of the detector did not fit the Geant-4 modeled detector efficiency values at low gamma energies (<300 keV). Therefore, the dead layer thickness of the detector was adjusted until agreement was achieved. After adjustment, the experimental spectrum for Co-60 agreed very well with predicted spectrum and had 8.6-9.3% error in photo peak net area values. Iron, nickel and copper sample spectra was used for further validation of the model. The predicted spectra agreed very well with experimental spectra in visual manners with 12.1-33.6% error in photo peak net area values. Along with the previously developed Activity Predictor software this new model in Geant-4 provides a more realistic spectra prediction for NAA experiments. Improvements for the Geant-4 model can be achieved by more accurately determining detector dimensions and inactive regions at the front and sides of the detector. Also the normal distribution generation used for spectrum predictions can be improved with introduction of a new variance variable depending on energy and number of counts.