Gonzalez, Yesenia Amanda (2016-12). Deconvoluting an Americium-Beryllium Neutron Spectrum from a Proton Recoil Detector. Master's Thesis. Thesis uri icon

abstract

  • A proton recoil detector was acquired from LND, Inc. to be used in neutron counting experiments with an 241AmBe source. Neutrons elastically scatter off of and ionize hydrogen atoms creating protons that can be detected. A single neutron can undergo multiple scattering events and this can result in several pulses being registered by the detector. This method of detection creates a detector response that changes with varying neutron fields, and unless both the incident neutron field and detector response to the neutron field are known, then it is not possible to visually determine what the detector is measuring. In order to solve this problem, we attempted to unfold the measured spectrum using MAXED. MAXED requires the measured spectrum, as well as a file containing response functions. Since there was no access to monoenergetic neutron sources to experimentally determine these response functions, the response functions were calculated using MCNPX. In addition, the detector response to several different 241AmBe neutron spectra were calculated and compared to the experimental data. Using MCNPX, the detector response was able to be modeled. Unfolding the experimental data was not successful, but the MCNPX results are consistent with the detector measuring an 241AmBe spectrum. The actual 241AmBe neutron spectrum can vary from source to source, so the spectrum used in MCNPX is not necessarily the spectrum emitted by the 241AmBe source. For future experiments and simulations using this specific source, the results of the MCNPX simulation can serve as a starting point for the neutron energy distribution. There will still be uncertainty associated with the source in simulations; however, with the number of different 241AmBe source spectra available, using a spectrum that produces results that match experiments can help reduce the error in future simulations.
  • A proton recoil detector was acquired from LND, Inc. to be used in neutron counting experiments with an 241AmBe source. Neutrons elastically scatter off of and ionize hydrogen atoms creating protons that can be detected. A single neutron can undergo multiple scattering events and this can result in several pulses being registered by the detector. This method of detection creates a detector response that changes with varying neutron fields, and unless both the incident neutron field and detector response to the neutron field are known, then it is not possible to visually determine what the detector is measuring.

    In order to solve this problem, we attempted to unfold the measured spectrum using MAXED. MAXED requires the measured spectrum, as well as a file containing response functions. Since there was no access to monoenergetic neutron sources to experimentally determine these response functions, the response functions were calculated using MCNPX. In addition, the detector response to several different 241AmBe neutron spectra were calculated and compared to the experimental data.

    Using MCNPX, the detector response was able to be modeled. Unfolding the experimental data was not successful, but the MCNPX results are consistent with the detector measuring an 241AmBe spectrum. The actual 241AmBe neutron spectrum can vary from source to source, so the spectrum used in MCNPX is not necessarily the spectrum emitted by the 241AmBe source. For future experiments and simulations using this specific source, the results of the MCNPX simulation can serve as a starting point for the neutron energy distribution. There will still be uncertainty associated with the source in simulations; however, with the number of different 241AmBe source spectra available, using a spectrum that produces results that match experiments can help reduce the error in future simulations.

publication date

  • December 2016