Paper EFFECT OF WALL THICKNESS ON MEASUREMENT OF DOSE FOR HIGH ENERGY NEUTRONS Delia Perez-Nunez and Leslie A. Braby* Abstract—Neutrons produced from the interaction between galactic cosmic rays and spacecraft materials are responsible for a very important portion of the dose received by astro- nauts. The neutron energy spectrum depends on the incident charged particle spectrum and the scattering environment but generally extends to beyond 100 MeV. Tissue-equivalent pro- portional counters (TEPC) are used to measure the dose during the space mission, but their weight and size are very important factors for their design and construction. To achieve ideal neutron dosimetry, the wall thickness should be at least the range of a proton having the maximum energy of the neutrons to be monitored. This proton range is 0.1 cm for 10 MeV neutrons and 7.6 cm for 100 MeV neutrons. A 7.6 cm wall thickness TEPC would provide charged particle equilibrium (CPE) for neutrons up to 100 MeV, but for space applications it would not be reasonable in terms of weight and size. In order to estimate the errors in measured dose due to absence of CPE, MCNPX simulations of energy deposited by 10 MeV and 100 MeV neutrons in sites with wall thickness between 0.1 cm and 8.5 cm were performed. The results for 100 MeV neutrons show that energy deposition per incident neutron approaches a plateau as the wall thickness approaches 7.6 cm. For the 10 MeV neutrons, energy deposition per incident neutron de- creases as the wall thickness increases above 0.1 cm due to attenuation. Health Phys. 98(1):37– 41; 2010 Key words: dose, absorbed; dosimetry; Monte Carlo; neutrons INTRODUCTION THE DESIGN of a TEPC for dosimetry of indirectly ionizing radiation, such as neutrons, as well as the high energy charged particles found in space is complicated. The design is influenced by detector size and weight as well as the desirability of secondary particle equilibrium in the detector wall. When secondary particle equilibrium cannot be achieved, it is important to have an understand- ing of the magnitude of the errors that are introduced. Evaluation of the error caused by a thin wall requires data on the energy deposited in the cavity as a function of wall thickness. These data were obtained through Monte Carlo simulations of energy imparted by recoil particles in a fixed-size gas volume surrounded by walls of different thickness and atomic composition. The accuracy of such estimates is limited by the accuracy of the available neutron cross section data and the methods for following recoil protons, but is adequate for guiding the detector design. In order to achieve ideal neutron dosimetry, it is important to achieve an optimal wall thickness. The main consideration is to satisfy the secondary charged particle equilibrium (CPE) condition to make dose equal to kerma. The objective is for the dose in the wall to represent dose at a point in an infinite uniform medium like the human body. In order to comply with the CPE condition, the wall thickness should be at least as thick as the range of a proton having the maximum energy of the neutrons to be monitored. However, thick walls will also attenuate low energy neutrons, resulting in an underesti- mate of their contribution to the total dose. Monte Carlo calculations were used to evaluate energy deposition per incident neutron in simulated low pressure propane-filled proportional counters as a function of the wall thickness. MONTE CARLO SIMULATION MCNPX version 2.4.0 by Los Alamos National Laboratory was used for the simulations. The program was set up to calculate the track length estimate of energy deposition in a 0.9 cm radius sphere filled with propane with density of 2.59  10 -5 g cm -3 , corresponding to a pressure of 10 torr, and resulting in a simulated site diameter of 0.47 m in unit density tissue. The simula- tions were conducted for a monoenergetic and monodi- rectional 12 cm diameter plane disk neutron source located 50 cm from the center of the propane sphere. For the 100 MeV neutrons, the mean free path in tissue-equivalent material is more than 10 cm (Zaider and Rossi 2001), significantly more than the simulated detector wall thicknesses. Thus any problem stemming * Department of Nuclear Engineering, Texas A&M University, 3133 TAMU, College Station, TX 77843-3133. For correspondence contact Delia Perez-Nunez at the above address or email at deperez@tamu.edu. (Manuscript accepted 22 July 2009) 0017-9078/10/0 Copyright © 2009 Health Physics Society DOI: 10.1097/HP.0b013e3181b8d032 37