Characterization of the Neutron Irradiation System for use in the Low-Dose-Rate Irradiation Facility at Sandia National Laboratories

Transcription

1 University of New Mexio UNM Digital Repository Nulear Engineering ETDs Engineering ETDs Charaterization of the Neutron Irradiation System for use in the Low-Dose-Rate Irradiation Faility at Sandia National Laboratories Jr. Manuel Frano Follow this and additional works at: Reommended Citation Frano, Jr. Manuel. "Charaterization of the Neutron Irradiation System for use in the Low-Dose-Rate Irradiation Faility at Sandia National Laboratories." (2015). This Thesis is brought to you for free and open aess by the Engineering ETDs at UNM Digital Repository. It has been aepted for inlusion in Nulear Engineering ETDs by an authorized administrator of UNM Digital Repository. For more information, please ontat

2 Manuel Frano Jr. Candidate Department of Chemial and Nulear Engineering Department This thesis is approved, and it is aeptable in quality and form for publiation: Approved by the Thesis Committee: Dr. Gary Cooper, Chairperson Dr. Maryla A. Wasiolek Dr. Robert D. Bush i

3 CHARACTERIZATION OF THE NEUTRON IRRADIATION SYSTEM FOR USE IN THE LOW-DOSE-RATE IRRADIATION FACILITY AT SANDIA NATIONAL LABORATORIES by MANUEL FRANCO JR. B.S. NUCLEAR ENGINEERING THE UNIVERSITY OF NEW MEXICO, 2011 THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Siene Nulear Engineering The University of New Mexio Albuquerque, New Mexio Deember, 2014 ii

4 ACKNOWLEDGEMENTS I would like to aknowledge the following individuals beause without them this would not be possible. Christina B. Hanson (RSPD) made it known to me that this work needed to be ompleted, and she guided me to the right people. Truly without her this would have never happened. Dr. Maryla Wasiolek (GIF/Researh Advisor) supervised the projet, and she provided tehnial guidane. Dr. Gary Cooper (Graduate advisor) provided guidane, and he agreed to be my graduate advisor. Don Hanson (GIF) moved, onfigured, and assisted in positioning of the equipment and experiment. Dr. Dann Ward (RP) gave tehnial advie on the N-Probe Mirospe and the neutron Bubble Dosimeter Spetrosopy Set. Albert Bendure (RPSD Manager) and Mike Spoerner (TAV Non-Reator Nulear Failities Manager) gave me permission to work on this projet. The radiography group (Kyle Thomspon and Kevin Rolfe) performed the radiography experiment. David Vehar (TAV) provided tehnial advie on the foil ativation. I would like to thank Radiation Protetion Sample Diagnostis for allowing me to irradiate their germanium detetor. I would like to thank everyone from Radiation Protetion for assisting me. There are many others that I would like to thank for their ontribution and assistane in ompleting this work. Thank you all. SAND T iii

5 Charaterization of the Neutron Irradiation System for Use in the Low-Dose-Rate Irradiation Faility at Sandia National Laboratories by Manuel Frano Jr. B.S., Nulear Engineering, University of New Mexio, 2011 M.S., Nulear Engineering, University of New Mexio, 2014 ABSTRACT The objetive of this work was to haraterize the neutron irradiation system onsisting of ameriium-241 beryllium ( 241 AmBe) neutron soures plaed in a polyethylene shielding for use at Sandia National Laboratories (SNL) Low Dose Rate Irradiation Faility (LDRIF). With a total ativity of 0.3 TBq (9 Ci), the soure onsisted of three reyled 241 AmBe soures of different ativities that had been ombined into a single soure. The soure in its polyethylene shielding will be used in neutron irradiation testing of omponents. The haraterization of the soure-shielding system was neessary to evaluate the radiation environment for future experiments. Charaterization of the soure was also neessary beause the doumentation for the three omponent soures and their relative alignment within the Speial Form Capsule (SFC) was inadequate. The system onsisting of the soure and shielding was modeled using Monte Carlo N-Partile transport ode (MCNP). The model was validated by benhmarking it against measurements using multiple tehniques. To haraterize the radiation fields over the full spatial geometry of the irradiation system, it was neessary to use a number of instruments of varying sensitivities. First, the omputed photon radiography assisted in determining orientation of the omponent soures. With the apsule properly oriented inside the shielding, the neutron spetra were measured using a variety of tehniques. A N-probe Mirospe and a neutron Bubble Dosimeter Spetrometer (BDS) set were used to haraterize the neutron spetra/field in several loations. In the third tehnique, neutron foil ativation was used to asertain the neutron spetra. A high purity germanium (HPGe) detetor was used to haraterize the photon spetrum. iv

6 The experimentally measured spetra and the MCNP results ompared well. One the MCNP model was validated to an adequate level of onfidene, parametri analyses was performed on the model to optimize for potential experimental onfigurations and neutron spetra for omponent irradiation. The final produt of this work is a MCNP model validated by measurements, an overall understanding of neutron irradiation system inluding photon/neutron transport and effetive dose rates throughout the system, and possible experimental onfigurations for future irradiation of omponents. v

7 TABLE OF CONTENTS ACKNOWLEDGEMENTS... iii ABSTRACT... iv TABLE OF CONTENTS... vi LIST OF FIGURES... ix LIST OF TABLES... xi NOMENCLATURE... xii CHAPTER 1 INTRODUCTION Purpose Sope...2 CHAPTER 2 BACKGROUND Desription of the System under Evaluation AmBe Soure Speial Form Capsule Aluminum Stand Polyethylene Shielding...9 CHAPTER 3 THEORY Neutron/Photon Prodution in 241 AmBe Soure Computed Photon Radiography Gamma Spetrosopy Photon Interations with HPGe Detetor Neutron Interations with HPGe Detetor Neutron Damage to HPGe Detetor Neutron Spetrosopy Fast Neutron Detetion by Organi Sintillator Fast Neutron Detetion by Bubble Dosimeter Spetrosopy Low Energy Neutron Detetion by 3 He Gas Proportional Counter Neutron Foil Ativation CHAPTER 4 METHODOLOGY AND EXPERIMENT Monte Carlo N-Partile Model Geometry and Materials vi

8 4.1.2 Soure Definition Detetor, Mesh, and Surfae Tallies Instrumentation Used in Model Benhmarking Computed Photon Radiography High Purity Germanium Gamma Spetrometer N-Probe Mirospe Neutron Spetrometer Neutron Bubble Dosimeter Spetrometer Neutron Foil Ativation Experimental Proedures Measurement Loations Proedures CHAPTER 5 RESULTS Computed Photon Radiography Radiography Image Saling Primary Imaging Plate Comparison Seondary Imaging Plate Comparison Neutron Ativation of Phosphor Imaging Plate Monte Carlo N-Partile Neutron Spetral Output Gamma Spetrosopy by HPGe Detetor Gamma Spetrosopy Comparison Identified Peaks in Spetrum Neutron Spetrosopy Using Various Instruments Neutron Calulations N-Probe Mirospe Spetrometer Neutron Bubble Dosimeter Spetrosopy Set Neutron Spetra Comparison for Loation D Neutron Ativation Foils Results Summary of Neutron Flux Density Results Summary of Dosimetri Results Neutron and Photon Flux Map in Neutron Irradiation System Neutron Flux Density Mapping vii

9 5.8.2 Photon Flux Density Mapping CHAPTER 6 NEUTRON SPECTRUM PARAMETRIC ANALYSIS Polyethylene Box with Tested Objet Inside Cadmium Box with Tested Objet Inside Polyethylene Shield between Tested Objet and Neutron Soure Low Energy Neutron Flux Configuration Using Polyethylene Blok CHAPTER 7 ABSORBED DOSE TO VARIOUS MATERIALS CHAPTER 8 CONCLUSIONS Future Work APPENDICIES APPENDIX A: MCNP INPUT DECK A.1 Neutron MCNP Input Code A.2 Radiography Tallies A.3 Bubble Detetor Tallies A.4 Mirospe Tallies A.5 High Purity Germanium Detetor Tallies A.6 Gold Foil Tallies A.7 Polyethylene Dose Tallies A.8 Silion Dose Tallies A.9 Water Dose Tallies REFERENCES viii

10 LIST OF FIGURES Figure 1. LDRIF South Cell Layout...4 Figure 2. NB570 and NB AmBe Soures Photograph and Shemati...6 Figure 3. MRC-AmBe-44 Soure Photograph and Shemati...7 Figure 4. Speial Form Capsule Model II...7 Figure 5. Speial Form Capsule on Aluminium Stand inside the Shielding...8 Figure 6. Polyethylene Chamber for Shielding and Storing the Neutron Soure...9 Figure 7. ISO-8529:1989, 241 AmBe Neutron Energy Distribution Figure 8. Radionulide 241 Am Deay Sheme Figure 9. Photon Absorption Range of Various Phosphors Figure 10. Photo-Stimulated Luminesene Complex Theory Figure 11. Stimulation and Emission Spetra of BaFBr:Eu +2 Imaging Plate Figure Co Spetrum Obtained with P-type HPGe Detetor Figure 13. Neutron Spetra produed by Fast Neutron Irradiation of Ge(Li) Detetor Figure 14. Eletron Energy Light Output as a Funtion of Proton Energy in NE Figure 15. Pulse Height Plots for Organi Sintillator Figure 16. Measured Pulse Height Spetra for Organi Sintillator Figure 17. Neutron Spetral Deonvolution from Mirospe Figure 18. Portable Computed Radiography Sanner Model CR 50P Figure 19. HPGe Gamma Spetrosopy System Figure 20. N-Probe Mirospe Neutron Spetrometer Figure 21. Bubble Detetor Spetrosopy Set and Misellaneous Equipment Figure 22. BDS Reompression Shielding and Reader/Bubble Counting Software Figure 23. Aerial Cross Setional View of LDGIF South Cell Figure 24. Measurement Loations with Door Closed and Open Figure 25. Experimental Imaging Plate obtained by Primary Phosphor IP Figure 26. MCNP Model of the Cylindrial Imaging Plate Figure 27. Horizontal View of the Measured Intensity at Normalization Plane Figure 28. Horizontal View of the Modeled Intensity at Normalization Plane Figure 29. Vertial View of the Measured/Modeled Intensity at Normalization Plane Figure 30. Loation of Eah Component Soure within the SFC ix

11 Figure 31. Confirmatory Phosphor Imaging Plate Modeled/Measured Comparison Figure 32. MCNP Neutron Spetra for Various Loations throughout System Figure 33. MCNP Neutron Spetra for Loations E and F Figure 34. MCNP Neutron Spetra Relative Change Figure 35. Gamma Spetra Summary Energy Range from 10 kev to 2000 kev Figure 36. Gamma Spetra Summary Energy Range from 2 MeV to 5 MeV Figure 37. Neutron Spetra Obtained by N-Probe Mirospe Figure 38. Neutron Spetra Obtained by Bubble Dosimeter Spetrosopy Set Figure 39. Neutron Spetra Comparison at Loation D for BDS and MCNP Figure 40. Neutron Spetra Comparison at Loation D for Mirospe and MCNP Figure 41. Cross Setions for Various Foils as a Funtion of Energy Figure 42. Neutron Flux Density as a Funtion of Distane from the Soure Figure 43. Neutron Effetive Dose Rate as a Funtion of Distane from the Soure Figure 44. Neutron Transport in Neutron Irradiation System Figure 45. Neutron Indued Photon Transport in Neutron Irradiation System Figure 46. Polyethylene Box inside the Polyethylene Shielding Figure 47. Poly-Box Configuration at 91 m Spetra Figure 48. Average Energy as a Funtion of Polyethylene Box Thikness Figure 49. Cadmium-Polyethylene-Soure Configuration Figure 50. Cadmium Box Neutron Spetra Figure 51. Cadmium Box with Varying Thikness Figure 52. Neutron Irradiation System with Polyethylene Shields Figure 53. Neutron Spetra for Varying Number of Poly Slabs Figure 54. Average Neutron Spetra Energy for Varying Number of Poly Slabs Figure 55. Polyethylene Blok Configuration Figure 56. Average Neutron Energy with Varying Polyethylene Blok Thikness Figure 57. Thermal Neutron Spetra with Polyethylene Blok Figure 58. Neutron Dose Conversion Fators for Material of Interest Figure 59. Dose Rates for Various Materials as a Funtion of Distane x

12 LIST OF TABLES Table I. Ativities and Masses for 241 AmBe Component Soures...5 Table II. Dimensions of the Speial Form Capsule Model II...8 Table III. Nulear Properties of 241 AmBe (α,n) Soure Table IV. Major Interations of 241 AmBe Soure Neutrons with HPGe Detetor Table V. Foil Ativation Pathways Table VI. Physial and Nulear Constants for Gold Foils Table VII. Measurement Loations and Methods Inside and Outside the Shielding Table VIII. HPGe Detetor Measurement Details Table IX. N-Probe Mirospe Measurement Table X. BDS Measurements Table XI. Ativation Foil Measurement Time and Dates Table XII. Gamma Statistis Obtained by HPGe Detetor Analyzed by PeakEasy Table XIII. Neutron Ativation Foil Tabulated Results Table XIV. Neutron Flux Density at All Measurements Loations Table XV. Neutron Effetive Dose Rate at All Measurements Loations Table XVI. Ratio of Thermal to Fast Neutrons for Different Box Thiknesses Table XVII. Ratio of Thermal to Fast Neutrons for Different Cadmium Thiknesses Table XVIII. Ratio of Thermal to Fast Neutrons at 91 m verses the Number of Slabs Table XIX. Ratio of Thermal to Fast Neutrons for Varying Blok Width Table XX. Tabulated Dose Rate for Various Material as a Funtion of Distane xi

13 NOMENCLATURE 241 AmBe Ameriium-241 Beryllium AmO 2 Ameriium Oxide BDS Bubble Dosimeter Spetrometer BTI Bubble Tehnology Industries DCFs Dose Conversion Fators DEP Double Esape Peak GIF Gamma Irradiation Faility HPGe High Purity Germanium ICRP International Commission on Radiologial Protetion IPs Imaging Plate(s) ISO International Organization for Standardization LANL Los Alamos National Laboratory LDRIF Low Dose Rate Irradiation Faility MCA Multi-Channel Analyzer MCNP(5) Monte Carlo N-Partile (Version 5) MeVee MeV Eletron Equivalent NIST National Institute of Standards and Tehnology ORNL Oak Ridge National Laboratory OSRP Off-Site Soure Reovery Projet PSLC Photo-Stimulated Luminesene Complex RP Radiation Protetion RPSD Radiation Protetion Sample Diagnostis SEP Single Esape Peak SFC Speial Form Capsule SNL Sandia National Laboratories xii

14 CHAPTER 1 INTRODUCTION The Gamma Irradiation Faility (GIF), along with its annex, the Low Dose Rate Irradiation Faility (LDRIF) at Sandia National Laboratories (SNL), provides environments for high- to low-dose-rate irradiation for omponents and materials using 60 Co and 137 Cs soures. These apabilities are being expanded to inlude low-dose-rate neutron and mixed neutron-gamma irradiation apabilities. The faility has aquired three reyled ameriium-241 beryllium ( 241 AmBe) soures, with a total ativity of 0.3 TBq (9 Ci). These soures were plaed inside a Speial Form Capsule (SFC), whih was positioned inside a polyethylene shield at the LDRIF. The neutron irradiation system onsisting of the 241 AmBe neutron soures and the polyethylene shielding will provide low-dose-rate neutron and mixed neutron-gamma fields for irradiation of materials, omponents, and equipment, whih omplements high- to low-dose-rate gamma irradiation apabilities provided by the GIF and the LDRIF. Diverse methods were required to understand, haraterize, and model the newly added neutron soure in its polyethylene shielding. First, the orientation of the three soures within the SFC was determined. Then, the predited neutron and photon spetra were obtained by modeling the 241 AmBe soures inside the polyethylene shielding. Finally, the model was validated with data from benhmark experiments. A variety of instruments that were sensitive to different intensities and energies were required to measure the neutron and photon spetra throughout the experimental onfiguration. Multiple independent measurements were obtained with different instruments, and they also served to verify the auray of the measurements. The projet also inluded parametri studies of different soure shielding arrangements for the purpose of neutron spetrum modifiation. From the neutron and photon spetra, it was also possible to determine absorbed dose rates to various materials and effetive dose rates to personnel for omparison with the Radiation Protetion (RP) instrument readings. 1.1 Purpose The purpose of this work was to haraterize the neutron soure and the environment around the polyethylene shielding. This would assist the faility personnel 1

15 to determine neutron and gamma absorbed dose rates in the materials and omponents that may be used in future irradiation experiments. Considerations were inluded of a onfigurable setup that would allow for variation of average spetral energies and intensities to aommodate future experimental needs of the faility ustomers. 1.2 Sope The sope of work inluded MCNP modeling of the irradiation system, MCNP model validation by measurements using different experimental tehniques, parametri studies, and measurements towards potential neutron spetra modifiations, and the use of the validated model to evaluate the absorbed dose rates in various materials as well as effetive dose rates for omparison with the response of radiation protetion instrumentation in neutron fields. The MCNP model was onstruted to predit the neutron field for all measurement loations and to understand the radiation transport in the modeled system. The modeled system onsists of the 241 AmBe soures, the soure support stand, the polyethylene shielding, as well as the spae around the shielding. The model preditions were ompared with the results of measurements at seleted points, and thus validating the model with experimental data. The following experimental tehniques were used to set up the model and, subsequently, to validate it. Computed photon radiography was used to understand the distribution of radioative material within the SFC to appropriately represent the soure in the MCNP model. This tehnique is useful for determining orientation and intensity for bulk radioative material without the need of destroying or violating the integrity of an objet that ontains the radioative material (1). A high-purity germanium (HPGe) detetor was used to determine the gamma spetrum of the soure. It was also used to determine bakground gamma spetrum prior to bringing the soure into the faility. The instrument hosen for this purpose was a oaxial germanium N type detetor manufatured by Canberra Industries, whih is relatively resistant to neutron damage on the rystal (2). The neutron spetrosopy was aomplished using two instruments: the N-probe Mirospe Spetrometer and the Bubble Dosimeter Spetrometer (BDS) Set. These two instruments were used to measure neutron spetra over a wide range of neutron energies (3). Foil ativation was used as a supplemental method to obtain spetral information about the soure (4). 2

16 One the neutron spetra were determined by using a validated model, a parametri study was onduted using MCNP. This was done to develop potential experimental onfigurations for future testing of materials and omponents by exposing them to the soure. The neutron spetra were also used to predit absorbed dose rates in various materials and effetive dose rates to workers. The latter were ompared with readings of instruments routinely used to evaluate neutron dose rates to workers. 3

17 CHAPTER 2 BACKGROUND Chapter two provides a bakground of the neutron irradiation system under evaluation. This setion details the soure omponent design, soure omponent radioativity, and apsule design. The polyethylene hamber and soure loation within the hamber are also desribed. This ombination of soure and hamber is referred to as the neutron irradiation system. 2.1 Desription of the System under Evaluation The system, whih is the subjet of this work, is loated in the Low Dose Rate Irradiation Faility (LDRIF). The LDRIF is a stand-alone onrete bunker, whih inludes two test ells and a setup/data aquisition lab. The LDRIF provides environments for low-dose-rate, long-duration testing of equipment, materials, and omponents. The south ell ontains lead shielding for gamma irradiation experiments using reonfigurable 137 Cs soures and a polyethylene neutron irradiation system. The neutron soure and shielding are the new additions to the faility. The model layout of the south ell is represented in Figure 1. Figure 1. LDRIF South Cell Layout The system onsists of a ustom, one-of-a-kind neutron soure, an aluminum soure stand, polyethylene shielding positioned on a onrete floor, and the environment inside and immediately outside the shielding. The lead aves used for gamma irradiation using 137 Cs sealed soures ontribute low-level gamma bakground inside the south ell. 4

18 AmBe Soure The neutron soure onsists of three well-logging soures, whih in 2012 were plaed inside the Off-Site Soure Reovery Projet (OSRP) Speial Form Capsule (SFC), and they were destined for disposal through OSRP at Los Alamos National Laboratory (LANL). Prior to shipment, the deision to dispose of the soures was re-evaluated, and the soure ownership was transferred to the Gamma Irradiation Faility (GIF) with the intent of using them as a new single soure for the low-dose-rate neutron irradiation system. The information about the omponent soures, based on proess knowledge and doumentation, is provided in Table I. These values and other information were obtained from the disposal request forms at Sandia National Laboratories (SNL). Soure # Serial # Table I. Ativities and Masses for 241 AmBe Component Soures Mass of Ativity at Birth 241 Am (g) a (GBq/Ci) Ativity on 1/19/14 (GBq/Ci) 1 NB (38/1.02) (34/0.92) 2 NB (100/2.7) (93/2.5) 3 MRC-AmBe (230/6.2) (210/5.7) Total SFC 2.9 (370/9.9) (340/9.1) a Values obtained from Disposal Request Form A speifi ativity for 241 Am of ± Ci/g (5) was used for all alulations in this work. It is noted that eah of the omponent neutron soures were reated at different times, and the doumentation of eah was not always omplete. Exat geometri dimensions were not available in the doumentation for omponent soures NB570 and NB540. The dimensions of these two soures were determined by using similar soure designs used at the time of manufature. These two soures were manufatured in the early 1950s. They were modeled in MCNP identially, and the only differenes between the two soures are the mass and the weight frations of the omponent materials. The NB570 and NB540 soures are shown in Figure 2. The photograph on the left of Figure 2 was taken prior to enapsulating the soures in the OSRP SFC, and the other image (to the right) is a shemati drawn using the dimensions from a similar design. 5

19 Figure 2. NB570 and NB AmBe Soures Photograph and Shemati The ratio of ameriium oxide (AmO 2 ) mass to beryllium (Be) mass is not stated in the doumentation for soures NB570 and NB540. Therefore an assumption had to be made based on doumentation from the 1960s and the weight fration alulations from Soure # 3 (MRC-AmBe-44). Based on doumentation from the 1960s, the neutron soure manufaturing industry mixed Be and AmO 2 using ratios from 2:1 to 20:1 by weight (6) (7). For example, ten parts by weight of Be metal was mixed with one part by weight AmO 2. In the example, the weight fration ratio 10:1 was used to avoid self-absorption effets to maximize yield. For soures NB570 and NB540, the ratio of Be to AmO 2 was set at 10:2 (Be-to-AmO 2 ) similar to the MRC-AmBe-44 soure. This assumption was made for NB570 and NB540. The most radioative soure of the three is MRC-AmBe-44. It was manufatured by Monsanto Researh Corporation on January 14, 1964, and there is suffiient information about this soure to haraterize and model it. This soure is shown in Figure 3. The photograph on the left of Figure 3 was taken prior to enapsulating the soures in the OSRP SFC, and the other image (to the right) is a shemati drawn of the soure. 6

20 Figure 3. MRC-AmBe-44 Soure Photograph and Shemati Speial Form Capsule In the late 1980s, it was reognized that a method was needed to qualify suspet or leaking soures as speial form. The SFC was developed and patented by Radiation Servie Organization, In. in It was further developed by LANL, so that it ould be easily sealed in the field and fabriated in different sizes (8). The SFC is designated as Model II, and it was used in this projet to enapsulate the three omponent neutron soures. SFC Model II is shown in Figure 4. Figure 4. Speial Form Capsule Model II (8) The omposite neutron soure used in this projet (RS# 5838) was produed by ombining the three sealed omponent 241 AmBe soures desribed in the previous setion into a Model II OSRP apsule. This produed a single ertified speial form soure. The dimensions of the Model II OSRP apsule are provided in Table II. 7

21 Table II. Dimensions of the Speial Form Capsule Model II Parameter Value/Desription 241 AmBe soure type Certified speial form sealed soure Seal total ativity Capsule type Capsule dimensions (8) (0.34 TBq/9.1 Ci) OSRP speial form apsule, Model II. Outer diameter = 7.6 m Inner diameter = 5.0 m Internal height = 21.6 m Overall length = 29.8 m When the three omponent soures were plaed inside the SFC in 2012, their loation inside the apsule was not well doumented. It was only known that all three soures were staked at the bottom of the SFC, and stainless steel wool was used to keep them from moving around within the SFC. Sine the exat information on the omponent soure geometry was not known, it was not possible to aurately model the soure in MCNP to haraterize the irradiation system Aluminum Stand An aluminum stand was fabriated to aommodate the soure, so it an be plaed at a distane from the onrete floor and the polyethylene roof. The stand also made it easier to reposition the soure stand inside the shielding. The stand and the soure inside the polyethylene shielding are presented in Figure 5. Figure 5. Speial Form Capsule on Aluminium Stand inside the Shielding 8

22 2.1.4 Polyethylene Shielding A speial polyethylene shield was designed and fabriated to aommodate the neutron soure. The interior dimensions of the shielding are m long, m wide, and 91.4 m high. The wall thikness is 26.9 m for the east and the south walls, 24.4 m for the north and the west walls, and 20.7 m for the roof. The soure was plaed in the north-west orner (far right hand side as shown in Figure 6), with its vertial axis at 30 m from the north and the west walls. Cirles were drawn on the floor around the enter of the soure at 30 m inrements from 30 m to 122 m to aid the plaement of the instruments used in this projet and the future experiments. The polyethylene shielding is shown in Figure 6. The door of the irradiation system faes east, and north is to the right of Figure 6. Figure 6. Polyethylene Chamber for Shielding and Storing the Neutron Soure The soure on its aluminum stand was plaed inside this polyethylene shielding, where it is used and stored when not in use. Some of the polyethylene slabs used to onstrut the shielding ontained trae amounts of boron in them, and most of these were plaed on the door. Several 137 Cs soures share the room with the polyethylene shielding, and they are shielded by several inhes of lead. These gamma soures are loated to the south (left in Figure 6) of the irradiation system. 9

23 CHAPTER 3 THEORY Chapter three provides information about the mehanisms of how the measurement values were obtained from the instruments. Basi theory about ameriium-241 beryllium ( 241 AmBe) soures and instruments utilized in the experiment are disussed. The methods of neutron/photon interations and detetion are detailed as well. 3.1 Neutron/Photon Prodution in 241 AmBe Soure Ameriium-241 is an artifiially reated radionulide produed in nulear reators. 241 Am is a deay produt of 241 Pu, and 241 Pu has a half-life of 13.2 years. Ameriium-241 is produed by the following sequene of nulear reations (9). A radioative (α,n) soure an be reated by mixing natural beryllium with 241 Am. This type of neutron soure is alled an Ambee soure. 241 AmBe soures produe neutrons mostly by the 9 Be(α,n) reation, where an alpha partile emitted by 241 Am interats with 9 Be, produing 12 C and a neutron. The neutron emission from an 241 AmBe soure an be aompanied by de-exitation gamma emission from exited 12 C as follows: ( ) Most of the neutron emissions from the above reation are assoiated with a asade of de-exitation gammas whose energies are 4.4 MeV and 7.7 MeV (9) (10) (11). There are other reations ourring within the 241 AmBe soure, but they are not signifiant relative to the alpha-neutron reation shown above. These reations will only be mentioned here for informational purposes, and they are as follows (12) : ( ) ( ) ( ) ( ) 10

24 A brief note about the (n,2n) beryllium reation, this reation inreases the neutron flux density in the system. The flux density inrease is dependent on soure design and loation. The distribution of neutron energies emitted range from approximately 0.01 MeV to 10 MeV. Industry standard, ISO-8529:1989, is used as the known neutron energy distribution throughout the doument (13). The neutron spetrum soure strength of an 241 AmBe soure is plotted as a funtion of energy, and it is presented in Figure 7. Figure 7. ISO-8529:1989, 241 AmBe Neutron Energy Distribution (13) The rate of deay of the 241 AmBe soure is dependent on the deay of 241 Am. The physial properties for 241 AmBe are shown in Table III. Table III. Nulear Properties of 241 (5) (7) (9) (12) AmBe (α,n) Soure Property Half-life of 241 Am Speifi ativity of 241 Am Average neutron energy Max neutron energy Neutron emission Average alpha energy from 241 Am Value ± 0.6 yr Bq/g (3.429 ± Ci/g) 4.16 MeV 11 MeV ~ n/s per TBq (~ n/s per Ci) 5.70 MeV Another signifiant emission of the 241 AmBe soure are the gamma rays emitted by deay of 241 Am. There are many gamma rays emitted by 241 Am, but the most ommon gamma ray energy emitted is at 60 kev. 11

25 The deay of 241 Am is given as follows (5) : The 241 Am deay sheme is given in Figure 8. This diagram shows the deay for energies less than 230 kev, showing 60 kev as the most ommon emission (9). Figure 8. Radionulide 241 Am Deay Sheme (9) 3.2 Computed Photon Radiography Figure 7 shows that neutrons emitted from an 241 AmBe soure generate a fast neutron spetrum. These neutrons interat with surrounding material in many ways suh as inelasti/elasti sattering and radiative apture. The neutrons may interat with the material induing gammas rays, and then these gammas rays an interat via the photoeletri effet, Compton satter, and/or pair prodution. Gamma- or x- rays are indiretly produed at a broad range of energies depending on the material (14). The intensity of the rays produed by this fast spetrum is dependent on the neutron soure strength. The nulear reation rates in a neutron field propagating through a material have an inherent gamma- and x- ray field proportional to it. In omputed photon radiography, imaging plates (IPs) are used to detet low energy gamma- and x- rays that ome diretly from the soure or are indued by neutrons. In this experiment, the IPs used are omposed of BaFBr:Eu +2. The IPs are doped with trae amounts of impurities, suh as Eu +2 (also alled ativators), and 12

26 ionization via low energy gamma- and x- rays in the IPs allow for an eletron/hole pairs to raise the ativator to an exited state (15). Photo-stimulable phosphor IPs an trap low energy gamma- or x-ray energy in the rystal struture of the plate for photons having energies of 10 kev to 140 kev depending on IP material. In Figure 9, three different types of phosphor IPs are displayed showing the photon absorption fration as a funtion of inident energy (15). Figure 9. Photon Absorption Range of Various Phosphors (15) If the phosphor IP rystal struture ontaining the trapped energy is stimulated by low energy light of a speified wavelength, it releases the trapped energy in the form of light. This proess is known as photo-stimulated luminesene (16). The low energy gamma- and x- rays absorbed in the phosphor IPs indue the formation of holes and eletrons, and it is best desribed by the Photo-Stimulated Luminesene Complex (PSLC) theory. This theory states that the PSLC is an F-enter in lose proximity to an Eu 3+ -Eu 2+ reombination enter. The holes and eletrons indued by the low energy gamma- and x- rays an either ativate a PSLC by being aptured by an F-Center, or it an form an ative PSLC via formation and reombination of exitons (15). The ative PSLC s reated are proportional to the low energy gamma- and x-ray dose to the phosphor IP. The dose to the IPs is then related to the intensity of the photons. The PSLC theory is illustrated in Figure 10, where the inident photon approahes the IPs from the bottom. The photon interats with the energy band of the phosphor IPs, and then a very low energy photon is emitted via stimulation as shown on the left of Figure

27 Figure 10. Photo-Stimulated Luminesene Complex Theory (15) The IPs are passed through the omputed photon reader. The red laser stimulates the IP ausing emission of blue light, and the low energy photons are shown in Figure 10 to the left oming out at 3.0 ev. This light is then reeived by a photomultiplier tube, whih onverts to a voltage, digitizes, and stores the signal in a digital image matrix. Figure 11 illustrates the stimulation and emission of the IPs for BaFBr:Eu +2 (15). Figure 11. Stimulation and Emission Spetra of BaFBr:Eu +2 Imaging Plate (15) 14

28 3.3 Gamma Spetrosopy This setion desribes basi gamma spetrosopy, neutron interations in a High Purity Germanium (HPGe) detetor relevant to an 241 AmBe soure, and energy resolution degradation due to neutrons in germanium rystals. A gamma spetrum from a germanium detetor exposed to a neutron soure is shown, and the neutron limits determined experimentally by fast neutron degradation of geranium detetors is also shown Photon Interations with HPGe Detetor The 241 AmBe is the soure of neutrons and photons. These partiles interat with the HPGe detetor, i.e., the germanium rystal and the detetor omponents. Photons interat with a material via photoeletri absorption, Compton satter, and pair prodution (16). The photoeletri effet ours when an inident photon transfers its energy to a photoeletron, whih is emitted from one of the atoms eletron shells. A Compton sattering interation ours when an inident photon ollides with an eletron. This reates a reoil eletron and sattered photon sharing the energy between the two, depending on the sattering angle. Pair prodution ours in the eletri field near the nuleus of the absorbing material. This interation results in the reation of an eletron-positron pair in exhange for the inident photon. A minimum gamma-ray energy of 1.02 MeV (two times the rest mass of an eletron) is required for this to energetially be possible (16). The detetor response to photons reflets interations with the material of the detetor, and the photons are diretly related to the energy given to eletrons in the proess, whih is deposited in the detetor. The typial features of the gamma spetrum inlude, but are not limited to the following: baksatter peak, Compton edge, annihilation radiation, single esape peak (SEP), double esape peak (DEP), and summation peaks (16). The 60 Co spetrum showing these features are shown in Figure

29 Figure Co Spetrum Obtained with P-type HPGe Detetor (16) Baksatter peak ours between 0.2 MeV and 0.25 MeV, and it is produed by photons that have first interated via Compton sattering in one or more of the materials surrounding the detetor. The photons baksatter into the detetor at sattering angle of approximately 180 degrees. Annihilation radiation ours when a positron generated in the pair prodution reation annihilates with an eletron and produes two 511 kev photons. A SEP at an energy 511 kev lower than the full energy deposition peak, and it ours when one of the two 511 kev annihilation gamma rays esape while the seond is fully absorbed. A DEP appears at an energy MeV lower than the full energy peak, and it ours when both 511 kev gamma rays esape the detetor. Summation peaks in the spetrum our during oinident detetion of two or more gamma-rays. This phenomenon ours mostly in appliations involving a radionulide that emits multiple asade gamma rays in its deay (16) Neutron Interations with HPGe Detetor In the neutron irradiation system, not only do photons interat with the detetor, but neutrons interat with it as well. To understand what peaks originate from neutron interations, Buntin and Kraushaar olleted a gamma spetrum exposed to monoenergeti neutrons ranging from 1.2 MeV to 16.3 MeV using a Ge(Li) detetor (17). The neutron interations provide an insight to whih peaks originate from neutrons, and these interations are presented in Figure

30 Figure 13. Neutron Spetra produed by Fast Neutron Irradiation of Ge(Li) Detetor (17) There should be an expeted number of peaks in the spetrum due to neutron interations with the detetor. Similar studies on understanding neutron interations in HPGe were performed by I. Abt, et al. (18). In their investigation, they looked at interations in a HPGe detetor exposed to neutrons from an 241 AmBe soure, and they identified signifiant peaks, whih are presented in Table IV. These gamma rays an be used as a guide to identify gamma-ray lines measured at the neutron irradiation system (18). 17

31 Table IV. Major Interations of 241 AmBe Soure Neutrons with HPGe Detetor (18) Energy (kev) Interation Type ( ) ( ) ( ) ( ) DEP of SEP of ( ) 3427 DEP of 4441 kev 3931 SEP of 4441 kev 4441 ( ) Neutron Damage to HPGe Detetor Although the HPGe detetor an indiretly detet neutrons via prodution of gamma-rays, this is not a reommended proedure beause the neutrons ause damage to the germanium rystal. Exessive neutron irradiation to the rystal auses the degradation of energy resolution in HPGe detetor. When irradiating the detetor, it is important to understand the limits of the rystal being irradiated. Due to the nature of the rystal design, N-type detetors are more resistant to neutron damage than P-type detetors. Aording to Knoll (16), exposure of thik planar germanium detetors to a fast neutron fluene of about 10 9 n/m 2 is suffiient to risk measureable hanges in the detetor resolution. Knoll also states that the detetor may beome totally unusable after exposure to a fluene of n/m 2. The IAEA reports (2), the approximate threshold fluenes for HPGe oaxial and planar detetors for early stages of energy resolution degradation when bombarded with fast neutrons to be the following: n/m 2 for P-Type detetors up to 20% in effiieny n/m 2 for P-Type detetors up to 70% in effiieny n/m 2 for N-Type detetors up to 30% in effiieny n/m 2 for N-Type detetors up to 70% in effiieny n/m 2 for planar detetors 18

32 Based on the energy degradation thresholds, it is possible to determine how long a HPGe rystal an be irradiated. 3.4 Neutron Spetrosopy This setion desribes fast neutron detetion via NE213 organi sintillators and Bubble Dosimeter Spetrometers (BDS); BDS utilize superheated vapors. Spetra deonvolution methods for these two detetors will also be disussed. This setion also presents the basi physis of low energy neutron spetrosopy using 3 He gas proportional ounters Fast Neutron Detetion by Organi Sintillator Neutron Interations The NE213 organi sintillator is a proton reoil detetor, and it is omposed mostly of hydrogen and arbon. The soure emits neutrons, and they interat with the elements of the NE213. The neutrons ollide by elasti and inelasti sattering, and they reate harged partiles in the NE213. Most of the energy from the harged partiles is dissipated nonradiatively in the form of lattie vibrations or heat, and a small amount of energy is onverted into fluoresent energy (16). The light yield of organis is dependent on the type of partile interating with it. A speial nomenlature is used to desribe the absolute light yield from harged partiles, and it is used to quantify the light yield on an absolute basis. That unit is alled MeV eletron equivalent (MeVee). A 1 MeV eletron generates 1 MeVee of light in the sintillator. For a given energy, a heavy harged partile, suh as a proton, will generate less light that an eletron of the same energy (16). The light output response of NE213 for eletrons and protons as a funtion of their energy is shown in Figure 14. Sine the light output is proportional to the proton s energy, this relationship is useful for disriminating between different inident neutron energies. 19

33 Figure 14. Eletron Energy Light Output as a Funtion of Proton Energy in NE213 (16) The organi sintillator response to heavy harged partiles, suh as protons, is always less than that for eletrons for equivalent energies. The relationship is linear for energies below 125 kev, and it then beomes nonlinear at higher energies (16). The resultant detetor pulse from Figure 14 is the produt of light output from moleules exited by the passage of protons as olleted by a photomultiplier tube. Figure 15 is obtained from Radiation Detetion and Measurement by Knoll, and it best illustrates organi sintillator detetor pulse response. In Figure 15 (16), part (a) shows a plot of pulse height (H) verses partile energy (E) for a typial organi sintillator. This nonlinear response leads to a distortion of the retangular proton reoil spetrum of part () into the spetrum shown in part (d). Using the spetrum from part (d), a spetral deonvolution using an iterative ode an now be performed on the spetrum to obtain neutron flux density or fluene values from the proton reoil detetor. 20

34 Figure 15. Pulse Height Plots for Organi Sintillator (16) NE213 Detetor Response Liquid sintillators are preferred beause of their pulse shape disrimination apability to suppress gamma-ray response. Organi liquid sintillators are used for the detetion of fast neutrons mainly beause of their high detetion effiieny, good timing resolution, and exellent n- pulse shaped disrimination properties (16). The size of the sintillator is also important. A larger rystal may absorb more energy, but there is a loss of resolution. The energy spetrum of proton reoils detetors will losely approximate the retangular distribution. One would like to keep the sintillator small so that multiple sattering of neutron beomes less likely, and the response is not as omplex for the deonvolution proess. The pulse distribution suffers from edge effets, multiple satter from hydrogen, sattering from arbon, detetor resolution, and ompeting reations at high neutron energies (16). There are many other effets that influene the response of the NE213, but the final spetral produt before deonvolution is shown in Figure 16. This figure gives an example for the overall detetor response funtion for a 2.54 m by 2.54 m ylindrial stilbene rystal for an inident neutron energy of 2.6 MeV as measured by Borman et al (19). 21

35 Figure 16. Measured Pulse Height Spetra for Organi Sintillator (19) NE213 Spetrum Deonvolution Methods One the pulse height spetral information is obtained from an organi sintillator, a speial ode is required to obtain the neutron fluene, and it is also required to understand how the response of the NE213 relates to the neutron flux density. Details for the unfolding algorithm utilized are not provided by the manufaturer for the N-probe Mirospe. A literature searh reveals that the N-probe probably uses FERDOR (20), a well-established ode developed by Oak Ridge National Laboratory (ORNL), or some variation thereof. There are many other odes that ould be used to deonvolve, suh as FORIST, RADAK, and FLYSPEC whih use various methods to deonvolve the spetrum (21). Figure 17, shows the deonvolution of a 3 H(d,n) 4 He produing a 14 MeV neutron beam measured by the N-Probe Mirospe. Figure 17. Neutron Spetral Deonvolution from Mirospe (22) 22

36 3.4.2 Fast Neutron Detetion by Bubble Dosimeter Spetrosopy Neutron-Bubble Interation The Bubble Dosimeter Spetrometer (BDS) set ontains 36 tubes with tiny superheated liquid droplets that are dispersed throughout a firm, transparent, elasti polyarylamide gel. The BDS are reeived from the manufaturer with a layer of low boiling point liquid above the polymer, whih exerts pressure on the elasti polymer. This is done to maintain the detetors in a radiation insensitive state. When the plasti tubes are open and the liquid has dissipated, the BDS is sensitized to neutrons. A typial droplet prior to vaporization is about 25 μm in diameter (23). When the neutrons interat with the detetor gel, the neutrons produe seondary harged partiles. The harged partiles interat with superheated liquid droplets in their viinity ausing them to vaporize. The energy from the harged partiles produe heat in the droplets, and the visible bubbles are formed in the gel. The bubbles are trapped at the loation of the formation, and they are suspended in the elasti polymer. With the bubbles visible and immobilized, this allows for them to be ounted (23). The sensitivity of the BDS an be adjusted by hanging the density or number of droplets per unit volume in the detetor. The material omposition of the BDS an also be adjusted to obtain different neutron energy thresholds. Also hlorine and nitrogen ontribute a thermal neutron sensitivity due to (n,p) reations in the BDS (23). Bubble Theory and Response The mehanism of the formation an be best explained using Seitz s thermal spike model. Aording to Seitz model, the neutron interats with the polymer and transfers its energy to a reoil nuleus. Elasti satter is the primary mehanism for the interation of the inident neutron with the nuleus of an atom. The harged nuleus deposits its energy by interation with eletrons and other nulei via ionization and exitation. The ionization and exitation along the path of the harged partile produes loal heating. This results in the formation of relatively small vapor bubbles (24). One a bubble of radius r is formed, it will remain in a stable state as long as the pressure within the bubble remains equal to the pressure exerted on the bubble from without (24). 23

37 This is desribed as the effetive surfae pressure: ( ) One the vapor bubble expands, it reahes its ritial radius defined as: ( ) (1) (2) (T) is the temperature dependent surfae tension of the liquid. The pressure differene aross the bubble surfae is given as: ( ) (3) P o is vapor pressure outside of the bubble, and P v (T) is the vapor pressure within the bubble at temperature T. Equation 2 indiates that if a bubble grows to a size where its radius is as large as or exeeds r that bubble beomes thermodynamially unstable. At this point, it will ontinue to grow by onsuming or boiling the entire super-heated liquid droplet (24). BDS Neutron Spetrum Deonvolution Sine there are not many energy bins available for the BDS, the neutron spetral deonvolution is relatively simple to perform, and it is performed by spetral stripping. In spetral stripping, the spetrum is unpeeled by starting with the largest energy bin. The orresponding response funtion for the BDS is multiplied by a variable fator, and then subtrated from the oded spetrum. The next lower energy is then hosen, and the subtration proess is repeated (16). The deonvoluted spetrum looks similar to the spetrum shown in Figure 17, exept it will have fewer energy bins Low Energy Neutron Detetion by 3 He Gas Proportional Counter Neutron Interations The seond method of detetion in the N-Probe Mirospe is the 3 He gas proportional ounter. The Mirospe measures neutron energies less than about 1.5 MeV using the following reation: ( ) The ross setion for this reation follows the well-known 1/v dependene with the thermal neutron ross setion at 5330 barns. The Q-value of this reation is 24

38 0.764 MeV. Conservation of energy and momentum ditates the proton will be given an energy of MeV and the 3 He an energy of MeV (16). A literature searh reveals that several authors have suessfully unfolded the inident neutron spetrum from 3 He (25). The alulations of 3 He detetor response funtions are omplex, and the alulations an lead to large unertainties, whih an distort the unfolded spetrum. Also, low energy neutrons reate pile-up whih obsures details in the fast neutron spetrum (26). Bubble Tehnology Industries (BTI) resolved these problems assigning three energy bins in the unfolded spetrum: thermal to 10 kev, 10 kev to 500 kev, and 500 kev to 1 MeV. The highest energy group overlaps with the lowest NE213 group. Information about how the unfolding algorithm determines the neutron fluene at the ross-over bin is not known. BTI also used a speial boron shield plaed around the 3 He detetor, so the over-response to low energy neutrons in the lower energy bin is dereased. Aording to BTI (25), these tehniques have been arefully designed to optimize the detetors energy response by the inlusion of partial and full penetrations through the shield. 3.5 Neutron Foil Ativation Many materials (often in the form of foils) an be ativated by neutrons by a variety of neutron reations, some of whih may have energy thresholds. By exposing enough different foils having a suitable range of reation thresholds to a neutron soure and measuring the resulting ativities, the spetrum of the neutron soure an be unfolded and the flux density determined. Table V shows the reations of interest for all the foils used in the experiment. Sine gold was the only foil to atually yield useful data, it will be the only one disussed in detail here. Reation of Interest Table V. Foil Ativation Pathways Half-Life (days) (27) (28) Effetive Threshold Energy (MeV) Energy of Partile Emitted (MeV) ( ) (beta) ( ) (gamma) ( ) 2.7 Thermal 0.41 (gamma) The determination of thermal neutron flux density by gold and admium overed gold was obtained using the following methodology. This method is aepted in industry, 25

39 and it was disovered as a viable method dating as early as the 1960s in an unlassified doument (4). In general, there an be many orretion fators that an be applied to the gold foil methodology to improve auray of the measurement, and it is left up to the experimenter to make that determination. The speifi methodology used to determine the gold foil results in this work is as follows. Assuming a Maxwellian distribution at temperature T and given the ross-setion of natural gold, the group averaged ross-setion is given by (29) : ( ) (4) The number of target nulei is defined by N T as: (5) m Mass of sample (g) w Abundane of target isotope (%) A Mass number of target isotope (amu) N a Avogadro s number (amu/g) The marosopi absorption ross-setion is defined as: ( ) ( ) (6) The admium orretion fator is used to orret for absorption of neutrons in the admium overed gold foil. The equation for this fator is given as: ( ) (7) The following equation orrets for build-up in the gold foil during irradiation: ( ) (8) λ Deay onstant of The following equation orrets for deay between the removal of the foil from field and the start of the ount time: ( ) (9) 26

40 The following equation orrets for deay during the ount time: ( ) (10) The ativity indued in the foil by thermal neutron apture is given as: ( ) (11) The self-shielding orretion fator of the gold foils was determined by modeling the gold foil in MCNP. The ratio of the perturbed neutron flux to the unperturbed neutron flux was determined by modeling and it was found to be The following equation was used: (12) - Average Perturbed Thermal Neutron Flux - Unperturbed Thermal Neutron Flux The thermal neutron flux density equation an be obtained using the following equation (4) (28) : ( ) ( ) (13) Equation 13 only quantifies neutrons for energies less than 0.5 ev, the admium utoff from the admium ross-setion. For onveniene, Table VI provides the physial and nulear onstants for gold foils. Table VI. Physial and Nulear Constants for Gold Foils # Property Symbol Value 1 Density ρ g/ 2 Isotopi abundane of 100% 3 Average atomi weight A amu 4 Half-life of days 5 Deay onstant λ /days 6 Gamma ray of interest 412 kev 7 Absorption mirosopi ross-setion 98.8 b (thermal) 27

41 CHAPTER 4 METHODOLOGY AND EXPERIMENT The objetive of this work is to develop the ability to predit the neutron flux density, neutron spetrum, and assoiated gamma-ray spetrum as a funtion of loation in the neutron irradiation system for any desired shielding or sattering onfiguration. The methodology used to aomplish this goal will be to model the neutron irradiation system using Monte Carlo N-Partile Code (MCNP), Version 5; to verify the soure onfiguration using omputed photon radiography; and to measure the neutron flux density and spetrum at a variety of loations using a variety of instruments and tehniques inluding the N-probe Mirospe neutron spetrometer, the Bubble Dosimeter Spetrometer, foil ativation; and to measure the gamma-ray spetrum with a High Purity Germanium (HPGe) detetor. This hapter will desribe how eah of these tehniques and instruments will help ahieve the objetive of this work. 4.1 Monte Carlo N-Partile Model The model of the neutron irradiation system was onstruted using the MCNP. MCNP an be used for neutron, photon, eletron, or oupled neutron/photon/eletron transport. It an perform alulations for many appliations inluding, but not limited to, radiation protetion, radiation shielding, radiography, nulear ritiality safety, reator design, et. MCNP was used to determine the radiation transport throughout the neutron irradiation system, and it was used to verify omponent orientation inside the Speial Form Capsule (SFC). One validated, the MCNP ode an be used to quikly estimate the neutron flux density and spetrum at any point in the system. By use of a validated model, one avoids the neessity of experimentally measuring these loations, whih would be labor intensive and require many man-hours to omplete. The MCNP model was also used for the parametri analysis whose goal was to find experimental onfigurations inside the shielding that result in speifi average spetral energies and intensities. Details of the onstrution of the MCNP model follow Geometry and Materials The geometry of the neutron irradiation system is relatively simple to model in MCNP. Figure 1 (Chapter 2) was generated using MCNPX Visual Editor (Vised) from the MCNP model of the Low Dose Rate Irradiation Faility (LDRIF) room. The entire 28

42 room was inluded in the model to determine the neutron radiation field not only inside the neutron shielding but also inside the room. This was done to better simulate the overall environment and to have the apability to evaluate neutron ativation of items outside the shielding, if neessary. The polyethylene shielding was modeled as five slabs. The door material was modeled as borated high density polyethylene, and the remainder of the hamber was modeled high density polyethylene. An iterative method was used to align the MCNP model with the physial representation of the neutron soure setup based on the radiography images, whih are disussed in Setion The radiography gave guidane on the loation of the strongest soure, but it did not provide enough information about the other two soures. An assumption was made that that the strongest of the three soures and one of the other soures are at the bottom of the SFC. The remaining soure was modeled as staked on top of these two soures. The metal mount of the shielding door was not modeled in the MCNP model. Transformation ards were used to allow the flexibility to represent the shielding with the door open, losed, or partially open (45 and 60 ). The material ompositions for the MCNP model were taken from Paifi Northwest National Laboratory doument titled, Compendium of Material Composition Data for Radiation Transport Modeling. (30) This referene is a tool useful for determining omposition of material in MCNP. The alternative soure of material ompositions was the National Institute of Standards and Tehnology (NIST) website (31) Soure Definition For eah ameriium-241 beryllium ( 241 AmBe) omponent soure, the MCNP model ativities were taken from the last olumn of Table I. This information was used to simulate the probability of neutron emission from eah soure. The neutron energy distribution of the 241 AmBe was modeled using the ISO spetrum (Figure 7). The soure definition was formulated for eah omponent soure, and eah soure was modeled as ylinder. Representing eah of the omponent soures separately in the MCNP model (rather than modeling the SFC as the radioative soure) allows the radiation transport within the SFC to be modeled more aurately. Then using MCNP, it is possible to determine the emission rate of the SFC if needed, with the understanding that the soure material is not homogenous inside the SFC. From the MCNP model, it 29

43 was possible to obtain the net neutron multipliation from (n,2n) reations in the 241 AmBe soures, and it was determined to be indiating a 2 % inrease in the neutron population Detetor, Mesh, and Surfae Tallies The total emission rate for all omposite soures within the SFC is n/s. This value was used as a multiplier for all the tallies in the MCNP model of the neutron irradiation system. An ativity-to-emission rate onversion fator of n/s/ci (7) (32) was used to determine the emission rate for eah omponent soure. All neutron flux density tallies, in partiular F4 tallies, were modeled for a sphere with 1 m radius to simulate the flux density at point near the measurement loations. A F5 tally was used to determine the self-shielding orretion fator of gold foils (Equation 12) for the neutron foil ativation tehnique desribed in Setion 3.5. It was also used to determine the neutron flux density inident on the foil, and this is the flux density found in Table XIII. For the dose tallies, the DF and DE ards were used to assign the dose onversion fators (DCFs) to the output. For the biologial dose rates (effetive dose rates), ICRP-74 H*(10) neutron and photon DCF oeffiients were used (33). The energy binning used to represent the distribution of flux density throughout the alulations followed the energy binning from ICRP-74. The effetive and absorbed dose rates were also determined using the DF and DE ards. The FMESH ard allows the user to define a mesh tally superimposed over the problem geometry. The FMESH ard was used to determine the neutron and photon flux density throughout the polyethylene shielding up to 30 m from the door and the side walls. Tally ards FIR and FIC were used for the radiography, and they are further disussed in the radiography results setion. 4.2 Instrumentation Used in Model Benhmarking The instrumentation used in the experiment is presented in this setion. Brief desription, design, and limitations are provided for eah instrument used for the haraterization. All hardware and software used are also listed, and purpose of use for the instrument is also stated. 30

44 4.2.1 Computed Photon Radiography To appropriately represent the soure in the MCNP model, omputed photon radiography was used to understand the distribution of radioative material within the SFC. The Portable Computed Radiography Sanner, model CR 50P from GE Inspetion Tehnologies with speial flexible phosphor imaging plates (IPs) was used to provide rapid images for radiography analysis. The flexible phosphor IPs allow for any needed geometry of analysis, suh as wrapping around a ylindrial soure. The IPs ability to instantaneously erase the image and to reuse them was benefiial to the haraterization of the soure. The Rhythm software provided rapid data analysis by quikly ontrasting and adjusting images to bring out important features of the radiograph. Combination of these features made omputed radiography a tool useful for the haraterization of the soure, speifially the soure omponents. Two IPs were used in the measurements. The primary IP had dimensions of 19.7 m 26.7 m with a plate thikness of 605 µm. This IP was used for most of the measurements by wrapping it around the soure. A seondary IP made of similar material and of the same thiknesses was also used. Its dimensions were 40.7 m 35.0 m. The seondary IP was plaed next to the soure as onfirmatory measurement to verify the image of the primary IP. The photon radiography system used is shown in Figure 18 with one of the IP loaded on the sanner. Figure 18. Portable Computed Radiography Sanner Model CR 50P (34) 31

45 The following radiography equipment was used to determine the soure orientation: GE Inspetion Tehnologies, Pegasus TM CR50P omputed radiography imaging plate sanner Laptop with GE Rhythm Aquire, version and GE Rhythm Review, Version 2.2 software Kodax Industrex Flex GP S-170 phosphor IPs High Purity Germanium Gamma Spetrometer A HPGe detetor was used to determine the bakground gamma radiation spetrum from the neighboring gamma shielding, whih is o-loated with the neutron irradiation system. The gamma spetrum of the neutron soure was also obtained by the HPGe detetor. The instrument hosen for this purpose was Canberra Industries oaxial germanium N-type detetor, whih is relatively resistant to neutron damage on the rystal. In general, N-type detetors are more resistant to neutron damage than P-type germanium detetors, therefore, this detetor allowed for longer irradiation time at a given loation, resulting in redued unertainties. The gamma-ray spetrosopy system onsisted of the detetor and the Canberra InSpetor 2000 Multihannel Analyzer (MCA). With an InSpetor 2000, the digital MCA allows for dead time orretion up to 50 % with auray up to 5 % at that system dead time (35). Also this HPGe detetor was hosen over sodium iodide (NaI) or any other gamma instrument due to the higher resolution. HPGe detetors have a 20 to 30 times better resolution than NaI detetors (36). The MCA provides a broad-window range of photon visibility, up to 64 thousand hannels, and it also allows adjustment of gamma-ray energy window to a relatively high energy. The gamma-ray spetrosopy system used in this projet is shown in Figure

46 Figure 19. HPGe Gamma Spetrosopy System The following Canberra equipment was used to determine the bakground and neutron soure photon spetra: HPGe N-type detetor with 50 % effiieny rystal o Crystal Dimensions 64 mm 39 mm (diameter length) InSpetor 2000 Digital Signal Proessing Composite Detetor Cables o Preamp power (9-pin D), Energy, HV Inhibit, BNCs, High Voltage (SHV) Computer Cables o RS-232, USB, MCS Cable Laptop with spetrosopy software Genie 2000 Version N-Probe Mirospe Neutron Spetrometer The N-probe Mirospe neutron spetrometer was used to determine neutron spetra of the soure at distanes of 122 m and greater from the soure. The N-Probe Mirospe an only be exposed to up to 200 μsv/hr (20 mrem/hr) without suffering too muh dead time; for greater dose rates the auray of the measurement would derease (37). This instrument was manufatured by Bubble Tehnology Industries (BTI). The N-Probe Mirospe uses a ombination of two neutron detetors: a liquid sintillator, NE213 onsisting mostly of hydrogen and arbon, and a 3 He spherial gas ounter 33

47 enased within a 10 B shield. These two separate detetors over a neutron energy range from 10-8 MeV to 20 MeV. The N-Probe ontains a 5 m 5 m NE213 type liquid sintillator, whih measures energies from 800 kev to 20 MeV. The speial 3 He gas proportional ounter enased within a 10 B shield measures energies from thermal up to 800 kev. Sophistiated pulse-shape disrimination rejets unwanted photons up to a gamma-ray to neutron rejetion ratio of approximately 1000:1 (37). The software for neutron spetral deonvolution is proprietary. The NE213 output is binned into 13 energy groups, and the 3 He is binned into four energy groups. Thus, the Mirospe has a total of 17 energy bins (25). The instrument manual states the dose rate sensitivity range is from <1 to 200 μsv/hr (<0.1 to 20 mrem/hr). The N-Probe software has the ability to determine the kerma, dose equivalent NCRP-38 (38), and ambient dose equivalent, H*(10) ICRP-74 (33). The dose rate feature of the Mirospe is useful for relating the measured dose rate to the MCNP results and to other instrument readings. The Mirospe has proven to be an instrument useful for the haraterization of the soure, and it is shown in Figure 20. The N-probe Mirospe is in the top left of the figure. Figure 20. N-Probe Mirospe Neutron Spetrometer The following is a list of equipment used to determine the neutron spetrum using the N-Probe Mirospe: 34

48 Analyzer o Mirrospe 2 o Palmtop omputer with Nview Software o Flash ard N-probe o High voltage module o Organi liquid sintillator with photomultiplier tube and 3 He gas ounter with boron shield and preamplifier o Pulse shape disrimination eletronis o Signal multiplex eletronis Signal Cable o Retratile able with a XLR onnetor for the analyzer o A Bendix onnetor for the probe Charger for o Analyzer module batteries o The N-Probe batteries Neutron Bubble Dosimeter Spetrometer The Bubble Dosimeter Spetrometer (BDS) set was used to determine neutron spetra at various loations within the polyethylene shielding out to a distane of 122 m from the soure. Thus, the BDS measurements omplemented those obtained with N-probe Mirospe, whih ould not be used in lose proximity to the soure due to its sensitivity limitations, and inluded one overlapping point at 122 m. The BDS was also manufatured by BTI. BDS ontain gels with superheated vapors, whih are engineered to be sensitive to different neutron energies. The BDS tubes of various sensitivities an be used as one spetrometer set. There are six detetors of eah of six different energy thresholds for a total of 36 tubes in the set. The different energy thresholds provided are 10, 100, 600, 1000, 2500, and kev. The sensitivities of these tubes are between 0.1 bubbles/μsv and 0.2 bubbles/μsv (1 bubbles/mrem to 2 bubbles/mrem) (39). 35

49 During irradiation the BDS must be kept at a temperature of 20 ± 0.5 C, otherwise the measurements are biased due to temperature sensitivities. One the tubes are irradiated, the bubbles must be ounted to determine the neutron spetrum. Following the ount, the bubbles inside the BDS are removed using the bubble reompression hambers (to zero the detetor set). During storage, the tubes must be kept at 6 C, and their reuse is limited to three months aording to the manufaturer. The algorithm used by this instrument is Exel spreadsheet-based, and it is provided by BTI, whih applies spetral stripping to deonvolve the spetrum (39). The total ount of eah the bin is determined, and it is iteratively subtrated as the alulation progresses to the lower energy bins. The BDS set to be used in this projet is shown in Figure 21 (image to the left). The bubble reompression hambers and ooler are presented in Figure 21 as well (image to the right). The other equipment used in onjuntion with the BDS set is shown in Figure 22. Figure 21. Bubble Detetor Spetrosopy Set and Misellaneous Equipment 36

50 Figure 22. BDS Reompression Shielding and Reader/Bubble Counting Software The following equipment was used to determine the neutron spetrum. All the equipment is from Bubble Tehnology Industries. BDS TM Bubble Detetor Spetrometer Set RC-18 Reompression Chambers BDR-III Automati Bubble Reader Bubble Detetor Bubble Counter Software Unfolding Bubble Detetor Software Distilled ionized water Cooler apable of providing the 20 C environment Neutron Foil Ativation Foil ativation was used as a supplemental method to obtain spetral information about the neutron soure. Three different foils were initially seleted: nikel, sulfur, and gold. An additional gold foil with a admium layer was also used to filter the thermal neutrons. These foils were seleted to give a oarse evaluation of the neutron energy spetrum from thermal energies to several MeV. All of the foils were m thik, and they all have a radius of 0.75 m. Preditions of expeted indued ativities suggest that the neutron flux density may not be suffiiently strong to ativate the foils to measureable levels. 37

51 4.3 Experimental Proedures This setion desribes measurement loations and instrument usage at speified loations. The dates and times the experiments were performed are provided. Details on how the measurements were taken and any anomalies during the measurements are also desribed Measurement Loations This setion desribes the experimental setup and measurement loations to be used in the benhmarking of the MCNP model. Figure 23 presents a ross-setional view of the south ell at the LDRIF, showing the loation of the polyethylene shielding with the neutron soure and the gamma shielding. The experimental loations inside and outside the shielding (with shielding door losed and open) are identified in Figure 24. These loations indiate where the N-Probe Mirospe, BDS set, and ativation foils were plaed to obtain measurements for benhmarking the MCNP model. It was not possible to plae all instruments at all loations. Rather, they were plaed as their sensitivity limits permitted. For instane, the HPGe detetor ould only be used outside the shielding with the door open due to dead time limitations and the desire to minimize potential neutron damage to the detetor. For eah lettered loation, Table VII desribes the distane from the soure and the detetion methods used at eah respetive loation. Figure 23. Aerial Cross Setional View of LDGIF South Cell 38

52 Figure 24. Measurement Loations with Door Closed and Open Table VII. Measurement Loations and Methods Inside and Outside the Shielding Loation Distane from Soure (m) Detetion Method A (inside) 30 Ativation Foils/ Bubble Detetors B (inside) 61 Bubble Detetors C (inside) 91 Bubble Detetors D (inside) 122 Bubble Detetors/ Mirospe E (inside) 61 Bubble Detetors F (outside) 145 Mirospe G (outside) 183 Mirospe/ HPGe Detetor Proedures Radiography A total of six separate exposures were taken on the IPs for the radiography on 11/21/2013. The primary IP was wrapped around the soure. The first exposure was 60 seonds in duration, and it produed an image with intensity levels just above bakground noise. Subsequent exposures were taken for about 10 minutes, whih produed useful results. During the last three exposures, a larger flat IP was plaed behind the primary IP. It was entered on the soure hotspot whih was identified in the previous exposures. This seond plate onfirmed that the image of the hotspot on the primary IP was not an anomaly. The last image produed the best results, and it was used to mark the hot range on the soure apsule. 39

53 In the last trial, the primary IP was positioned so that the bottom edge would be aligned with the bottom of the soure apsule. The images were exposed for 11 minutes. Lead numbers were taped around the soure to identify diretion on the primary IP. The images were used to obtain an understanding of the diretionality and positioning of the 241 AmBe soure omponents inside the SFC. The images assisted in an iterative proess of adjusting the soure within the MCNP model to best simulate the atual soure geometry. Gamma Spetrosopy The HPGe detetor was plaed outside the shielding at loation G (1.83 m from the soure) faing the hot spot. Before the HPGe detetor was used, preliminary measurements were taken using Portable Survey Neutron REM Meter Model E-600/NRD and the Mirospe to determine the dose rate and the neutron flux density. This was done to ensure that the HPGe detetor plaement was suh as to minimize rystal degradation from the fast neutrons. At loation G, the REM Meter measured 125μSv/hr (12.5 mrem/hr), whih was less than the upper threshold limit of the Mirospe. The Mirospe data showed the neutron flux density at loation G to be 120 n/m 2 /s. This intensity was adequate for the gamma measurement. The gamma spetrosopy ount was taken for a total of seonds, or live-time seonds. The initial ount time duration was 300 seonds, live-time seonds, but fear of over exposure to the germanium lead to an early termination of the measurement. Dead time for this measurement was at about 41 %. For the dead time orretion, Canberra states (35) the auray of referene peak area is 5 % (3 % typial) at up to 50 % system dead time with a setting of 5.6 μs rise time and 0.8 μs flat top. The total fluene on the germanium rystal for fast neutrons (energy greater than 10 kev) was n/m 2 based on the Mirospe measurement. This is muh less than the fluene threshold before energy resolution degradation ours (2), whih is ~ n/m 2. The summary of the HPGe detetor measurement details is presented in Table VIII. 40

54 Measurement Number 1 12/11/13 Mirospe Neutron Spetrosopy Table VIII. HPGe Detetor Measurement Details Date Duration of Exposure Distane Loation s (live time) s (real time) 183 m G (door open) The N-Probe Mirospe was plaed in two different loations with the door losed and one loation with the door open. The loations with the door losed are D and F, 122 m and 144 m from the soure, respetively. Loation D was the farthest point from the soure in the shielding, enterline to the hotspot. It was deided to make loation D the pivotal point for the neutron measurements. The measurement at loation F was made to determine how the neutrons propagated through the polyethylene wall. The third loation was outside of the shielding with the door open at point G, 182 m from the soure, between the door and the opening of the irradiation system. A summary of the N-Probe Mirospe measurement details is presented in Table IX. The measurement number five at loation F was the longest in duration beause it was outside of the irradiation system behind a polyethylene wall, and additional time was needed for good statistis. Measurement Number Table IX. N-Probe Mirospe Measurement Date Duration (s) Distane (m) Loation 1 12/11/ D (door losed) 2 12/11/ D (door open) 3 12/11/ G (door open) 4 5/2/ D (door losed) 5 5/2/14 3, F (door losed) 6 5/2/ D (door losed) Bubble Detetor Spetrosopy Beause the BDS are not as sensitive as the Mirospe, they were plaed loser to the soure. The BDS were plaed every 30 m (loations A through E) from 30 m to 122 m from the soure. During irradiations, the BDS were kept inside the ooler at a onstant temperature of 20 C. The bubble detetors had to be irradiated long enough to 41

55 generate more than 75 bubbles but not more than 200 bubbles in eah of the tubes for all energy ranges. This was neessary to obtain good statistis for the deonvolution software. The numbers of bubbles were determined using the bubble reader and ounting software. After the number of the bubbles was determined, the tubes had to immediately be reompressed to avoid the expansion of the bubbles and potential ripping of the gel. The BDS were reompressed using the reompression hamber. The BDS were reompressed using pressures up to 4.14 MPa (600 psi) for at least 45 minutes. If all of the bubbles did not reompress, longer times were used. This was done weekly, regardless of the BDS use, as reommended by BTI. The details of the BDS measurements are provided in Table X. Measurement number five at loation E was performed on the side meaning diretly south of the soure, rather than southeast. Measurement Number Foil Ativation Date Table X. BDS Measurements Irradiation Time Distane from Soure (m) Loation 1 1/29/14 20 ± 0.25 min 30 A 2 1/29/ hr ± 30 s 122 D 3 2/5/ hr 122 D 4 2/19/14 50 ± 1 min 61 B 5 2/26/14 1 hr ± 30 s 61 (on side) E 6 2/26/14 2 hr 5 min ± 30 s 91 C 7 3/11/14 2 hr and 5 min ± 30 s 91 C Four different foils were plaed at position A, whih was 30 m from the soure enterline, to determine if foil ativation was a viable method for neutron field haraterization. The four foils onsisted of a gold foil, a admium-overed gold foil, a nikel foil, and sulfur pellet. After one week of irradiation, they were pulled out, and they were ounted via gamma spetrosopy and gross alpha/beta. The times and dates of ations performed on the foil measurements are presented in Table XI. 42

56 Table XI. Ativation Foil Measurement Time and Dates Event Dates and Times Irradiation start time 1/15/2014, 15:00 Irradiation end time 1/22/2014, 13:00 Gold foil ount start time 1/23/2014, 15:32 Gold foil ount end time 1/24/2014, 5:26 Cadmium-overed gold foil ount start time 1/23/2014, 16:11 Cadmium-overed gold foil ount end time 1/24/2014, 6:04 43

57 CHAPTER 5 RESULTS Chapter five presents the modeling and experimental results, whih were obtained using multiple measurement tehniques. Throughout the hapter the measurements are ompared with the orresponding quantities obtained from the Monte Carlo N-Partile (MCNP) model of the neutron irradiation system (the ombined soure and polyethylene shielding) for model validation. 5.1 Computed Photon Radiography Computed photon radiography was used to determine the loation of the omponent soures inside the Speial Form Capsule (SFC) and the diretionality of the soure. This tehnique relies on photons produing an image on the image plates (IPs). The photon radiation emitted from the soure was used as a surrogate of the neutron radiation, whih was assumed to have the same spatial distribution Radiography Image Saling MATLAB was used to perform image proessing on the radiography data from the tally results from MCNP and on the experimental images obtained from omputed photon radiography. The intensity of the radiographs was normalized to highest intensity value (100%). The plane ontaining the highest intensity point is referred to hereafter as the normalization plane. The length and width of the images were saled to the atual physial dimensions of the IPs. All the images were saled aordingly, so that proper omparison ould be done Primary Imaging Plate Comparison It was expeted that the radiography would give suffiient resolution to determine the orientation of eah omponent soure inside the SFC. Some of the resolution was lost due to the photon attenuation in the stainless steel, but the intensity was suffiient to produe adequate images. Computed photon radiography is a tehnique used for photon imaging, and the magnitude of the neutron interation with the IPs, if any, is not known. However, it an be assumed that the strongest neutron emission would be in the same region as the strongest photon emission. For the primary IP, the bottom of the IP was aligned with the bottom of the apsule. Lead numbers were taped around the apsule 44

58 about 2.5 m apart from eah other as indiators, and then the IP was wrapped around the apsule. The lead numbers were used to mark the orientation of the IP relative to the soure to determine the soure diretionality. Figure 25 shows the image from the primary IP. The lead numbers are also visible in the image. Figure 25. Experimental Imaging Plate obtained by Primary Phosphor IP A white vertial line an be identified at about 160 degrees. This is the edge of the IP. Lead numbers from one to four are barely visible, and the remaining lead numbers are smeared. Therefore, the numbers were marked below the smeared numbers. The darkest regions in Figure 25 are between the numbers one and two, and the numbers eight and nine. These two regions of the highest intensity indiate whih part of the soure has the highest photon (and neutron) emissions. This range is between 90 degrees and 250 degrees, or between numbers eight and three respetively. This orresponds to about 160 degrees. The height of the hot spot is approximately 5.5 m from the base of the IP and the soure. An MCNP radiography model was generated using the FIC (Flux Image on Cylinder) tally. A ylindrial radiography image was simulated to align the experiment with the model. In the model, the image was wrapped around the SFC at the bottom, similar to how the experiment was performed. The omponent soures in the model were moved around inside the SFC, in an iterative manner, to obtain the best math with the 45

59 experimental radiography image until the model mathed the experiment. The modeled radiography image is shown in Figure 26. Figure 26. MCNP Model of the Cylindrial Imaging Plate In the MCNP radiography model, a hot spot range an be seen between 110 degrees and 230 degrees. This is approximately 120 degrees. The height of the hot spot is approximately 5.5 m from the bottom of the modeled SFC. The modeled image mathes the experimental image well. Further analysis of the images reveals more details about the omponent soures. Image proessing reveals some insight about the orientation, as shown in the horizontal view aross the normalization plane in Figure 27. Figure 27. Horizontal View of the Measured Intensity at Normalization Plane 46

60 Figure 27 is the experimental radiography image intensity aross the horizontal normalization plane. Again, the IP saturation range an be seen from 90 to 250 degrees, and the disontinuity from the IPs an be seen at 160 degrees. Also, the intensity suppression from one of lead numbers ours at 125 degrees. It an be onluded from the experimental radiography image that the areas of the highest intensity are not spots, but rather they extend over approximately 180 degree range. From this image, it an be onluded that the strongest soure is loated at the bottom of the SFC tuked on one side, sharing half of the spae in the SFC with one of the other omponent soures. The radiography image from the model an be analyzed in a similar fashion. Figure 28 shows the horizontal normalization plane of the modeled radiography image. Figure 28. Horizontal View of the Modeled Intensity at Normalization Plane The distribution of intensity aross the normalization plane in MCNP is illustrated in Figure 28. The measured image from the IP shows the saturation of the phosphor, whih is absent from the MCNP image. Therefore, the maximum was entered at the midpoint of the hot range from the experimental IP, whih is at 160 degrees. Vertial ross setions analysis of the IPs verifies that the strongest soure was plaed at the bottom of the apsule. The apsule has a layer of stainless steel on the bottom with at a thikness of 4.3 m, and the vertial hot spot is at 5.5 m. The vertial ross-setions are shown in Figure

61 Figure 29. Vertial View of the Measured/Modeled Intensity at Normalization Plane The images in Figure 25 through Figure 29 onfirm the orientation and vertial loation of the strongest 0.3 MBq (6 Ci) soure within the SFC. The strongest soure was plaed at the bottom of the SFC tuked on one side, whih is learly visible from the images. One of the lower ativity soures was plaed next to it, whih an be seen in Figure 25. Suh onfiguration explains the higher intensity in half of Figure 25, and the lower intensity in the other half, whih was exposed to one of the other weaker omponent soures. The MCNP model was oriented aording to the atual experiment. The hot spot was marked on the SFC, and the spot was oriented toward the enter of the door for all measurements desribed in this work. Figure 30 illustrates where eah omponent ameriium-241 beryllium ( 241 AmBe) soure was plaed. Neutron soure MRC-AmBe-44 and NB540 were modeled at the bottom inside the SFC. NB570, the weakest soure of the three, was modeled above the srew of the MRC-AmBe-44. This was the optimal onfiguration for best math between the modeled radiography and the experimental radiography as shown below. 48

62 Figure 30. Loation of Eah Component Soure within the SFC Seondary Imaging Plate Comparison A seond IP was plaed in front of the hot spot for onfirmatory measurement. This image not only onfirmed that the primary image was produing orret results, but it also onfirmed the height of the strongest soure. In addition, it onfirmed that there was not any radioative material inside of the SFC at the top. The MCNP radiography model was obtained using the FIR (Flux Image on Radiograph) tally. The model and the onfirmatory IPs results have good agreement. The images from the seondary plate and the orresponding modeled MCNP images are presented in Figure 31. The experimental radiography IP is to the left of the figure, while the MCNP model is to the right. Figure 31. Confirmatory Phosphor Imaging Plate Modeled/Measured Comparison 49

63 The experimental radiography image to the left shows that all of the radioative material is loated at the bottom of the SFC, and it also shows that there is nothing else above it. The MCNP model image to the right, after aligning the ylindrial radiography, produed similar results to the experiment. There was no need to iterate the seond model radiography image to onverge on the experimental image Neutron Ativation of Phosphor Imaging Plate Following the irradiation of the primary IP, neutron ativation of the phosphor was observed. The radionulides produed as a result of neutron ativation were onsistent with the omposition of the phosphor. The IP is omprised of a layer of bariumfluorobromioiodide doped with divalent europium (BaFBr/I:Eu2+). Gamma spetrosopy analysis of the plate immediately after the irradiation showed short-lived radionulides suh as 128 I (produed from 127 I), 139 Ba (produed from 128 Ba), 80 Br, and 82 Br. (produed from 79 Br and 81 Br, respetively), whih are all produed via (n, ) apture reations. All of these radionulides were short-lived, an all of them deayed after about two weeks. 5.2 Monte Carlo N-Partile Neutron Spetral Output Preditions of the neutron spetra at the experimental loations, A through G, were first obtained by MCNP modeling of the neutron irradiation system. The spetra predited by the model are represented in Figure 32 and Figure 33 for the loations inside and outside the shielding. The shape of the original neutron spetrum, as provided by ISO 8529, is disernable for all measurement loations. 50

64 Figure 32. MCNP Neutron Spetra for Various Loations throughout System Figure 32 shows the neutron spetra for loations A through D, and loation G. All spetra inlude the fast neutrons in the energy range onsistent with ISO 8529, from 2 MeV to 10 MeV. Also visible is a low energy peak ranging from 10-8 MeV to 10-6 MeV, whih results from neutrons sattering from the polyethylene walls and onrete. Also, there is another neutron peak ranging from 0.5 MeV to 1 MeV, whih is a result of neutron satter in the SFC. The SFC is mostly omposed of iron and nikel, and neutron sattering from them reates that distribution. The neutron distribution aused by iron and nikel an be seen as low as 10-2 MeV. A small fration of the fast neutrons have elastially sattered multiple times in the SFC before exiting, and they have lost more energy than neutrons that have sattered one. This is why there is a gradual inrease in the flux density from 10-2 MeV to 1 MeV. These three neutron distributions are dominant throughout the neutron irradiation system. For loations A through D, as the distane from the soure inreases, the intensity of the fast neutron omponent dereases. The intensity of the thermal peak does not vary muh throughout the inside of the irradiation system. At loation G (outside the shielding with the door open), both the fast neutron and thermal neutron intensities derease. The thermal peak signifiantly dereases relative to loations A through D beause of the redued neutron satter due to the absene of some of the sattering surfaes, so neutrons 51

65 do not have the opportunity moderate beause some of the neutrons esape through the door opening. Figure 33. MCNP Neutron Spetra for Loations E and F Neutron spetra for loations E and F are shown in Figure 33. Both of these loations are south of the soure, and both spetra were modeled with the door losed. The loation E is 61 m from the soure, diretly south. The spetrum at loation E is similar in distribution and intensity to the other measurement at 61 m, at loation B, whih is southeast of the soure. The measurement loation F is outside the shielding, and the neutron flux density at that loation dereases substantially as the neutrons are moderated through the wall of the polyethylene shielding. To better ompare reation rates and to bring out other effets in the neutron irradiation system, the area under the plots was normalized. In Figure 34, the area under the urve is onstant for all spetra from A through D and G. The area under the urve has been set to one, and all the spetra are normalized to spetrum A, whih is the strongest. Figure 34 shows the relative ontributions of different energy neutrons for other loations with respet to loation A in the shielding. 52

66 Figure 34. MCNP Neutron Spetra Relative Change As the distane from the soure inreases, the ratio of the fast neutrons in the spetrum to thermal neutrons dereases. This trend an be seen in spetra B through G. The fast neutrons satter and moderate from the high energy peak to the thermal peak. In loation G, there are more fast neutrons than at loation D beause the door is open and the neutrons are not thermalized as effiiently. The open door also allows the neutrons to esape the system rather than thermalize, so leaving the door open redues the thermal ontribution. The thermal peak is redued by leaving the door open. The MCNP model and detetor tallies are provided in APPENDIX A: MCNP INPUT DECK. 5.3 Gamma Spetrosopy by HPGe Detetor Gamma spetrosopy was used to haraterize the gamma radiation field assoiated with the neutron irradiation system. The photons ontributing to that field may originate from the neutron soure, but they may also be produed as a result of neutron interations with the materials around the soure. A bakground measurement was taken prior to plaing the neutron soure in the irradiation system to determine the natural bakground omponent as well as the ontribution from the 137 Cs gamma irradiation hambers loated in the same room. The MCNP model was also used to determine gamma emissions from the neutron soure via elasti and inelasti sattering. 53

67 5.3.1 Gamma Spetrosopy Comparison The gamma radiation bakground was obtained with the High Purity Germanium (HPGe) detetor plaed inside of the shielding with the door losed when the soure was not present. The gamma bakground spetrum is shown in Figure 35 and Figure 36, identified by the red line. After the soure was moved into plae within the shielding, the gamma spetrometry ount was taken outside of the irradiation system with the door open, at 1.22 m from the soure. Maintaining the distane from the soure was neessary to minimize the degradation of the germanium rystal by the fast neutrons. The gamma spetrum produed by the soure is identified in Figure 35 and Figure 36 by the blue line, and it does not orret for bakground radiation. The gamma spetrum of the soure obtained from the MCNP model is identified in the same figures by the blak line. Some of the disernable peaks in the gamma spetrum from the soure are due to inelasti and elasti sattering of neutrons from germanium, hromium, and aluminum from the stand or the detetor. There is also small ontribution from neighboring 137 Cs in the spetrum. A 2.2 MeV gamma-ray is of interest in the gamma spetrum beause it is not produed diretly from the soure, and it is emitted from the thermal neutron apture in hydrogen. This reation ours in the onrete (the floor) and also in the walls of the polyethylene shielding. The 2.2 MeV gamma-rays ause single and double esape peaks in the detetor (SE pk and DE pk in Figure 35). In the bakground spetrum, there is a prominent 137 Cs peak from nearby soures. There are also terrestrial radiation peaks from 40 K and from radionulides in the natural deay series. The bakground spetrum was olleted to identify and disriminate the bakground omponent of the gamma spetrum from the neutron soure gamma spetrum. The gamma spetrum obtained from the MCNP model only aounts for materials defined in the model, and it does not aount for potentially ativated omponents in the detetor. Modeling the gamma spetrum in MCNP assists in identifying lines that stritly ome from the modeled onfiguration. For example, the spetrum inludes the gamma-rays from neutron sattering with aluminum (from the aluminum stand). Also present are 2.2 MeV photons from the thermal neutron apture reation in hydrogen. The MCNP model was expanded up to 5 MeV to see if there are any other interations 54

68 resulting in gamma emissions. As a result, three additional peaks were identified. These three peaks ome from the neutron apture in aluminum (the stand), the polyethylene walls, and from the 241 AmBe soure (de-exitation) gamma following the (α, n) reation in beryllium. PeakEasy (40) software was utilized to identify the peaks, and it was used to determine where these peaks originated. The gamma spetrum results for the bakground, the soure, and the MCNP model are presented in Table XII. 55

69 56 Figure 35. Gamma Spetra Summary Energy Range from 10 kev to 2000 kev

70 57 Figure 36. Gamma Spetra Summary Energy Range from 2 MeV to 5 MeV

71 Identified Peaks in Spetrum ROI ID# Centroid (kev) ROI Total Counts Table XII. Gamma Statistis Obtained by HPGe Detetor Analyzed by PeakEasy Net Area (Counts) Net Area Unertainty (Counts) Peak (CPS) FWHM (kev) FWHM (%) Redued Chi Sqr Reation E E E E % 1.34 Bak satter peak E E E E % 2.1 ( ) E E E E % 2.39 ( ) E E E E % 1.66 Annihilation Radiation E E E E % 1.8 ( ) E E E E % E E E E % 0.77 ( )/ E E E E % 1.63 ( ) ( ) E E E E % 1.19 ( ) E E E E % 1.01 ( ) E E E E % 0.88 Double esape peak E E E E % 1.05 ( ) E E E E % 0.87 ( ) E E E E % E E E E % 1.07 Single esape peak E E E E % 0.9 ( ) E E E E % 6.19 ( ) ( ) ( ) ( ) Errors Reported at one sigma (68% CI)

72 5.4 Neutron Spetrosopy Using Various Instruments This setion presents the results of neutron spetra measurements, whih are then ompared with the modeled spetra. The first subsetion explains how the plots presented throughout Setion 5.4 were generated. The following subsetion presents the results obtained by the neutron spetrometers: the N-probe Mirospe and the Bubble Dosimeter Spetrometer set. Then, the ativation foil results used to determine neutron flux density are presented and disussed. Finally, by summarizing all the neutron measurements, it is possible to validate the MCNP model. It is possible to determine how the neutron population behaves and moves throughout the neutron irradiation system onfiguration. From a validated MCNP model, the effetive dose rates to personnel an be alulated Neutron Calulations Neutron Spetra Plotting When plotting funtion ( ) on the ordinate and variable on the absissa over a broad range of energies, it is neessary to plot on a logarithmi sale rather than a linear sale to bring out signifiant effets in the spetrum. When hanging the absissa to logarithmi sale, the ordinate and the absissa hange aordingly ( ) ( ( )) and ( ). This subtends the plot area by the urve between any pair of absissas log(e 1 ) and log(e 2 ). The logarithmi plot is now proportional to the integral of the distribution between E 1 and E 2. Therefore, any group bin is desribed by it lower boundary (E i ) and its upper boundary (E i+1 ). There are any number of bins based on instrument design and limitations. The group flux density bin is given as: ( ) (14) As the energy in the spetra range from thermal to fast energies, the absissa will be plotted on a logarithmi sale. All neutron spetra ordinates in this report are plotted using the following equation: ( ) (15) 59

73 Average Neutron Spetral Energy The average neutron spetral energy is alulated by: ( ( ) ) (16) The average neutron spetral energy is useful in determining the average energy of the neutron population at any given loation in the irradiation system. Ratio of Thermal to Fast Neutrons The thermal neutron fration is defined as: ( ) (17) The fast neutron fration is defined as: ( ) (18) The ratio between thermal and fast neutrons is defined as: (19) The ratio between thermal and fast neutrons is useful for determining how many baksatter neutrons ompared with fast neutrons oming from the 241 AmBe soure are present at any given loation. For omponent irradiation, whether if the average neutron energy dereases or inreases, the ratio will inform if there are more thermal neutrons rossing through a point in the hamber; it is useful for parametri analysis of the irradiation system N-Probe Mirospe Spetrometer Due to its sensitivity limitation, the N-probe Mirospe was plaed at measurement loations D, F, and G, whih are relatively far from the soure (1.22 m, 1.45 m, and 1.83 m respetively). The measurement loations are shown in Figure 24, and the neutron spetra obtained at these three loations are shown in Figure 37. The 60

74 energy range is from 10-8 MeV to 20 MeV, and there are 17 energy bins for the Mirospe. Figure 37. Neutron Spetra Obtained by N-Probe Mirospe In Figure 37, the Mirospe measurements display a neutron energy distribution from about 0.5 MeV to 10 MeV. This is onsistent with the MCNP model. The 241 AmBe neutron energy distribution is shown between 2 MeV and 10 MeV, and the neutron spetral shift aused by neutron interations with the iron in the SFC expands the energy distribution down to 0.5 MeV for all loations. Unfortunately, the resolution in the Mirospe does not bring out the full peak distribution from the iron, but it does broaden the 241 AmBe distribution. This is a key indiator that the neutron satter off of iron is seen by the Mirospe. This broadening auses the average neutron energy emitted from the soure to derease. At measurement loation D, the door is losed in the experimental onfiguration, and the instrument is 1.22 m from the soure. The overall neutron intensity is the highest for this loation for the Mirospe beause it is the losest to the soure from the three loations. It is worth noting the thermal bin (10-8 MeV to 10-2 MeV) for loation D is the highest among the three spetra. This indiates that for loation D, there are relatively more neutrons in the thermal bin than for the other two loations. This is due to the ontribution of the low energy neutrons from the polyethylene, whih ats as a moderating medium. The Mirospe deonvolution software plaes all the thermal and 61

75 low energy neutrons in one bin. Therefore, this bin should only be used as an indiator of neutron moderation. If the 241 AmBe soures were suspended in air with no moderator nearby, the low energy bin would derease rather than inrease. At measurement loation G, the onfiguration has the door open, and the instrument was plaed at a farther distane from the soure than loation D (1.83 m). At loation G, there is a derease in the overall neutron intensity due to inreased distane from the soure. The 241 AmBe energy spetrum does not hange signifiantly with distane at loation G, and it is still similar to ISO 8529 at this distane. The thermal bin intensity is lower at loation G with the door open. This derease in low energy neutrons is beause neutrons esape the hamber with the door open instead of being moderated by the door and refleting bak into the hamber. The measurement at loation F was taken with the door losed, and the instrument was plaed outside of the shielding on the side, diretly south, and at 1.45 m from the soure. The fast neutron omponent of the 241 AmBe energy distribution is still disernable outside of the polyethylene shielding. The intensity has dropped by several orders of magnitude, and at this point the neutrons have passed through approximately 25.4 m of polyethylene Neutron Bubble Dosimeter Spetrosopy Set The neutron Bubble Dosimeter Spetrosopy (BDS) set was plaed at loations A though D inside the shielding. The loations used for the BDS measurements were loser to the soure than the Mirospe beause of the lower sensitivity of the BDS. The energy bins measured by the BDS are wider, and the BDS do not measure energies lower than 10-2 MeV. These are also a few limitations of the BDS, whih are desribed below. The spetra obtained with the BDS for all measurement loations are shown in Figure

76 Figure 38. Neutron Spetra Obtained by Bubble Dosimeter Spetrosopy Set At measurement loation A, the BDS was 30 m from the soure. The spetrum for this loation is haraterized by a broad distribution of neutron energies up to 10 MeV, whih is onsistent with the referene distribution of an 241 AmBe soure, as defined in ISO The BDS at loation A also detet the neutrons that have lost a fration of their energy by the iron. At loation B, the detetor set was plaed 61 m from the soure. Although the neutron flux density is lower in all energy bins than in loation A, the general shape of the spetrum is similar. For the loation C, whih was 91 m from the soure, the spetral deonvolution software begins to over-strip at the lower bin beause of the low bubble ount in the lower energy bins. However, the energy distribution for the higher energies is onsistent with the referene spetrum. For measurement loation D, at 122 m from the soure, the spetral deonvolution software eliminates another bin, but the 241 AmBe distribution in the high energy range is still onsistent with the referene spetrum. The BDS deonvolution software uses an underdetermined matrix to obtain the flux density. The BDS produe a spetrum that is rather rude, yet the results are still valid. For the loations A through D, the flux density shows a derease in neutron intensity with inreasing distane from the soure. 63

77 5.4.4 Neutron Spetra Comparison for Loation D The Mirospe and the BDS were both used to measure neutron spetra at loation D, and a MCNP spetrum was obtained as well. In Figure 39, the BDS and MCNP spetra are plotted together to see how they ompare. For the bins available in the BDS, they ompared reasonably well with the MCNP model. The BDS lost two bins due to over stripping from the deonvolution algorithm. All of the low energy neutrons in the MCNP model were binned in the lowest bin. If the deonvolution algorithm would have preserved the spetrum for the lower energy bin of the BDS, it would have been lower than the MCNP representation. Figure 39. Neutron Spetra Comparison at Loation D for BDS and MCNP The Mirospe and the MCNP neutron spetra at loation D are shown in Figure 40. The Mirospe ompares well with the MCNP model for the fast spetrum, but it does not ompare well for energies lower than 1 MeV. It begins to deviate at about 2 MeV. The three lower energy bins belong to the 3 He gas proportional ounter, and the ross over between the two instruments our at 1 MeV. The Mirospe low energy bin ompared reasonably well with the MCNP model. 64

78 Figure 40. Neutron Spetra Comparison at Loation D for Mirospe and MCNP The Mirospe uses the 3 He gas proportional ounter to measure lower energy neutrons whih has a very oarse resolution, and the software uses one bin for thermal neutrons. The auray of the N-probe spetral data in the lower energy groups was questioned by Williams (25). In his report he ompared an ISO referene for an 241 AmBe to the N-Probe Mirospe. In the example spetrum in his report, the high-energy portion ompared fairly, but the lower energy bins disagreed. Regardless, for pratial appliations in this experiment, the Mirospe low neutron energy bins an serve as indiators of low energy neutrons in the field, whih an help with the haraterization of the neutron energy distributions. Eah analytial method identified the 241 AmBe energy distribution from the soure, and the methods also indiretly measured the neutron energy shift aused by the iron in the SFC. 5.5 Neutron Ativation Foils Results Ativation foils were used as one of the methods of neutron spetra haraterization, but they were partially suessful. There was no detetable ativity from the nikel foil or the sulfur pellet. There was some ativity from both the admium-overed gold foil and the bare gold foil. The neutron flux density obtained from the gold foils is shown in Table XIII. The foils were exposed at loation A, enterline with the hot range, for seven days. 65

79 Table XIII. Neutron Ativation Foil Tabulated Results Foil Method Neutron Flux Density (n/m 2 /s) Experiment MCNP Nikel Sulfur Gold (E < 0.5 ev) (620 ± 40) (740 ± 22) *Errors Reported at one sigma The gold foils were irradiated to about 85 % of the saturation ativity. The neutron flux density alulated from the experiment is ompared with the results predited by a MCNP alulation. The MCNP flux density at the foil loation is higher, but the method to quantify the experimental flux density from the gold foil exposure ould be improved by using a more preise ross-setion rather than assuming a Maxwellian distribution. This might bring the experimental value loser to the MCNP value. The samples ould also have been plaed loser to the soure to signifiantly inrease the flux rate to inrease the ativity of foils. The following figure illustrates the apture ross setions for various foils as a funtion of energy, obtained from ENDFPLOT. Figure 41. Cross Setions for Various Foils as a Funtion of Energy (41) The sulfur pellet and the nikel foil did not get ativated to detetable levels due to the relatively short irradiation time ompared with the half-life (refer to Table V) of the ativation produts and the relatively low neutron flux density. The 2,200 m/s mirosopi absorption ross setion for gold is 98.8 barns (29), and the sulfur and nikel apture ross setions for fast energy neutrons reah a maximum value at approximately 66

80 0.5 barns as shown in Figure 41. If the gold foil was the only foil to undergo ativation, based on the apture ross setions, the other two foils have signifiantly less of a hane of absorbing neutrons. Fortunately, enough gold was ativated to be deteted. Still, the experimental value obtained from the neutron ativation of gold foils and the value obtained from MCNP ompared relatively well. 5.6 Summary of Neutron Flux Density Results The neutron flux density spetral information from all the neutron instruments and the MCNP model are summarized in Table XIV. The loation and distane from the soure are provided in the first and seond olumns. In the third olumn, the total flux density for the entire spetrum from MCNP is provided. In the fourth olumn only energies greater than 10 kev from MCNP are given. Displaying olumn three and four in suh matter allows for an understanding of how many neutrons have been slowed down to less than 10 kev. In the fifth and sixth olumns the total flux density from the Mirospe and BDS measurements are provided, respetively, so they an be ompared. The average energy of the neutron spetrum at the loation is provided in the last olumn, and it is based on the MCNP model as defined by Equation 16. Loation Table XIV. Neutron Flux Density at All Measurements Loations Distane from Soure (m) Neutron Flux Density (n/m 2 /s) All Energy E > 10keV E > 10keV E > 10keV MCNP MCNP Mirospe BDS 67 Average Energy (MeV) A ± B ± C ± D ± ± G ± F ± E ± Errors Reported at one sigma From Table XIV, it is possible to draw many onlusions about the neutron irradiation system. In omparing the MCNP results for all the neutrons and the neutrons with energies greater than 10 kev, it is obvious that initially most of the neutrons are fast neutrons. As the distane from the soure inreases, there are signifiantly more low

81 energy neutrons, even at loation F whih is outside the shielding. In omparing all three results from MCNP, Mirospe, and BDS for energies greater than 10 kev, they all ompare well given the unertainties. The unertainty in the BDS measurements ould be improved by repeating them. The best results are for loation D, where all instruments results align with the model. The average energy inside the shielding with the door losed is the highest loser to the soure, and the average energy dereases as a funtion of distane due to the diminishing ontribution from fast neutrons. Outside the shielding, the ontribution from the moderated neutrons dereases, whih results in the overall inrease of the average neutron energy. The neutron flux density is signifiantly lower for the loations outside the shielding beause the shielding attenuates many neutrons. The flux densities for neutrons with energies greater than 10 kev were plotted for loations A though D (with door losed) and G (with door open) as a funtion of distane to demonstrate relationship between these quantities. The flux density does not derease as fast as a 1/r 2 point soure, but instead it is loser to a 1/r 1.7 relationship. This is shown in Figure 42 for those loations throughout the irradiation system. All the MCNP results shown in the figure are within one standard deviation of the measurement results. The MCNP theoretial flux density and the experimental flux density logarithmi fits are also plotted, and they ompare well. For the experimental fit, loations A through D and G only were aounted for in the power funtion fit. Figure 42. Neutron Flux Density as a Funtion of Distane from the Soure 68

82 5.7 Summary of Dosimetri Results From the neutron spetral information, it is possible to alulate the effetive dose to a human body. The values in Table XV were alulated by onverting the neutron flux densities to the effetive dose rates using the ICRP-74 neutron dose onversion fators. The olumns in Table XV are displayed in a similar fashion to Table XIV along with the measurements obtained with the Portable Survey Neutron REM Meter Model E-600/NRD. Loation Table XV. Neutron Effetive Dose Rate at All Measurements Loations Distane (m) All Energy Effetive Dose Rate (mrem/hr) E > 10keV E > 10keV E > 10keV MCNP MCNP Mirospe BDS E-600/NRD A ± 60 B ± 20 C ± 14 D ± 5 40 ± G ± F ± E ± 20 Errors Reported at one sigma As expeted, the effetive dose rates dereases with distane from the soure. It an be seen from the MCNP values that the effetive dose rate ontribution from the neutrons with energies less than 10 kev is negligible ompared with those with energies greater than 10 kev. Overall, for most loations, the effetive dose rates ompare well among the instruments within the range of unertainty. At loation D, the bubble detetor dose rate deviates from the MCNP model and the Mirospe measurements. This is due to the spetral loss of the lower energy bins by over stripping in the neutron spetrum. Similar to Figure 42, the effetive dose rates were plotted as a funtion of distane from the soure for eah instrument, and both MCNP model values and experimental data show an approximate 1/r 1.7 relationship for the effetive dose rate

83 Figure 43. Neutron Effetive Dose Rate as a Funtion of Distane from the Soure 5.8 Neutron and Photon Flux Map in Neutron Irradiation System This setion presents the three dimensional spatial neutron and photon distribution in a steady state ase with the polyethylene shielding door losed. Based on the suessful MCNP model validation presented in the previous setions, the model an be utilized to determine other properties of the neutron irradiation system without the need of additional measurements Neutron Flux Density Mapping The neutron distribution within the neutron irradiation system was represented visually using the Fmesh feature of the MCNP software. Figure 44 visually demonstrates how the neutrons propagate throughout the shielding, and it shows how strong the field is at any given loation. In the figure, the soure is loated toward the bak of the shielding away from the door 30 m form the west and the north walls. The spatial domain for the MCNP neutron radiation transport modeling extends from the far right hand side orner of the shielding out to 30 m from the shielding s outer wall and door, and up 30 m from its top. 70

84 Figure 44. Neutron Transport in Neutron Irradiation System It an be seen where the wall of the polyethylene shielding begins and ends beause this is where the neutron intensity dereases substantially. As expeted, the neutron intensity dereases as a funtion of distane from the soure, and it also dereases substantially as it propagates through the polyethylene shielding. There is minor leakage through the raks of the polyethylene door. The figure shows that the radioative material in the soure behaves similar, but not exatly, to a point soure Photon Flux Density Mapping Similarly, the photon omponent an be modeled using MCNP. Figure 45 is set up in a similar fashion to Figure 44. In this figure, the only photons aounted for are photons indued by neutrons. The SFC was modeled using material 304 stainless steel, whih ontains 69 % iron, 19 % hromium, 9.3 % nikel, 2 % manganese, and 0.7 % other materials by weight. The 2,200 m/s absorption ross setions, not inluding other materials, for eah of the elements are 2.55 barns, 3.1 barns, 4.43 barns, and 13.3 barns, respetively. The satter ross setions are 10.9 barns, 3.8 barns, 17.3 barns, and 2.1 barns (29), respetively (not inluding other material ). This information is provided to make a judgment as to whih element is more likely to apture neutrons, and whih element is likely to satter a neutron. 71

85 The other materials are arbon, silion, phosphorus, and sulfur, and the probability of a gamma emission or neutron satter from them is highly unlikely sine they represent less than 1 % of the material by weight. Most of the other materials do not generate a signifiant amount of photons, relatively speaking. There are many nulear reations and proesses ourring in the SFC that ontribute to the large prodution of photons, and only the top two most abundant materials in the SFC will be disussed. The SFC is omposed mostly of iron, and iron is a likely andidate for produing a signifiant number photons. In the gamma spetrum from Table XII, there are two possible inelasti sattering peaks identified by the HPGe detetor. Inelasti sattering of iron, 56 Fe(n,n ), was deteted at 1.24 MeV and possibly at 0.84 MeV. It is diffiult to quantify whether all or some of the peaks originated in the SFC or in the HPGe detetor enapsulation. Although 56 Fe is a stable nulide, neutron apture by 56 Fe also emits prompt gamma rays. There are two possible nulides from natural iron that may apture neutrons, and they are 54 Fe and 58 Fe. 58 Fe annot be a large ontributor to gamma emission beause it is less than 0.3 % abundant in natural iron. As for 54 Fe, although it may be 5.8% abundant, the neutron apture produes 55 Fe whih does not emit gammas. The seond most abundant element in the SFC is the hromium, and it generates photons in the SFC. The HPGe detetor deteted a peak at 1.4 MeV whih was identified as inelasti satter of hromium, 52 Cr(n,n ). In regards to neutron apture, there are two possible stable nulides, and they are 50 Cr and 54 Cr. The natural isotopi abundane of eah is 4.3 % and 2.4 %, respetfully. One the hromium stable nulides absorb neutrons, they reate 51 Cr and 55 Cr whih emit gammas. These are a few examples illustrating how photons are generated in the SFC. 72

86 Figure 45. Neutron Indued Photon Transport in Neutron Irradiation System There is no ontribution from the 241 Am, the neighboring 137 Cs soures, or any natural gamma emitter in the MCNP model and the MCNP image. In the ase of 241 Am, most of the low energy photons have been attenuated, but loser look at the gamma spetrum in Table XII results reveals a peak at 0.72 MeV. This is a relatively dominant peak in 241 Am when the gamma emission is suffiiently strong enough to be deteted. This gamma energy may be from the 241 Am, but it is diffiult to determine. It ould be oming from inelasti sattering in the borated polyethylene. Further modeling is required to make this determination, and it will be left for future work. In the photon flux density image, it is possible to see the inrease of the gamma intensity along the polyethylene wall whih is due to the radiative thermal neutron apture in hydrogen induing the emission of 2.2 MeV gamma-rays. In summary, the photon flux density image shows signifiant self-absorption in the SFC. This is one reason why the low energy gamma rays of 241 Am were not deteted by the HPGe detetor. One the neutrons are emitted from the omposite neutron soures, they interat within the SFC. The neutrons elastially satter, inelastially satter, and undergo radiative apture generating gamma- and x- rays and hanged partiles around the SFC. As expeted, the photon flux density dereases as a funtion of distane. Most of the photons produed in the system originate from neutron interations in the 241 AmBe soures and SFC. 73

87 CHAPTER 6 NEUTRON SPECTRUM PARAMETRIC ANALYSIS Chapter six desribes parametri analysis of the neutron irradiation system for various onfigurations. In the parametri analysis, the validated MCNP model is used to study pratial, experimental setups for irradiation of materials and omponents. The model is also used to examine how simple geometri hanges in the setup an be used to modify the energies of neutrons inident on materials. Four onfigurations were investigated to determine how the neutron spetra hange inside the polyethylene hamber. The four onfigurations are the following: a polyethylene box with a tested objet inside, a admium box with the tested objet inside, polyethylene shields between the soure and the tested objet, and thik polyethylene blok near the wall between soure and the tested objet. All MCNP tallies were measured at 91 m from the soure. 6.1 Polyethylene Box with Tested Objet Inside In this parametri analysis, a polyethylene box, with a test objet to be studied inside, was added to the MCNP model of the neutron irradiation system at 91 m from the neutron soure. The onfiguration is shown in Figure 46. The thikness of the box was varied from 2.5 m to 25.4 m. The neutron soure was plaed in the far right hand orner, and the polyethylene box with tested objet was entered at 91 m from the soure with the door losed. The purpose of this analysis was to determine how the spetra and the average spetral energy hange inside the polyethylene box as a funtion of its thikness. Figure 46. Polyethylene Box inside the Polyethylene Shielding 74

88 Figure 47, shows the hanges and effets of the poly-box onfiguration for the neutron spetra at 91 m from the soure. From the parametri analysis, it was found that as the polyethylene box thikness inreases, the intensity of the fast neutrons dereases, as expeted. A strong thermal peak remains inside the poly-box. For a poly-box thikness less than about 5 m, the thermal neutron peak is not signifiantly impated. However, for poly-box thikness greater than about 5 m, the thermal peak dereases proportionally with the fast neutrons. This is an indiator that the low energy neutrons in the spetra are dependent on the fast neutrons in the system. Figure 47. Poly-Box Configuration at 91 m Spetra The average neutron energy for various poly-box thiknesses brings out an interesting relationship. There is an optimal minimum energy at a thikness of about 6 m of polyethylene (see Figure 48). If the polyethylene box is any thiker, the spetrum beomes faster due to the loss of low energy neutrons, as they begin to be absorbed in the polyethylene. The average energy as a funtion of polyethylene box thikness is shown in Figure 48, and the minimum average energy obtainable in this onfiguration is about 6 m of polyethylene thikness at an average energy of 0.45 MeV. 75

89 Figure 48. Average Energy as a Funtion of Polyethylene Box Thikness From this parametri analysis, it is onluded that to derease the average energy inside the box using this method, only a few entimeters of polyethylene are reommended. Even though the average spetral energy dereases, the overall neutron intensity is not signifiantly impated ompared to the original spetrum. The ratio of thermal to fast neutrons as defined in Setion was alulated for this parametri analysis. Without a box, the ratio is 2.22, and as the thikness of the box inreases the ratio osillates around 4.5 as shown in Table XVI. Even though the total flux density dereases with inreasing thikness, the relative ontribution between thermal and fast neutrons does not inrease nor derease. The ratio of thermal to fast neutrons is given in Table XVI. In this table thermal neutrons are defined as neutrons with energies between 10-8 MeV and 10-6 MeV, and the fast neutrons as having energies between 10-1 MeV and 10 1 MeV. Thus, even though the average neutron energy is lowest at a box thikness of 6.35 m, there are atually somewhat more fast neutrons than thermal ones at 6.35 m than for a box thiknesses of 2.54 m and 12.5 m. 76

90 Table XVI. Ratio of Thermal to Fast Neutrons for Different Box Thiknesses Thikness (m) R Cadmium Box with Tested Objet Inside In this parametri analysis, a admium box with a tested objet was added to the MCNP model of the neutron irradiation system at 91 m from the neutron soure. The onfiguration is shown in Figure 49. The thikness of the box was varied from 0.1 m to 1 m. The neutron soure was plaed in the far right hand orner, and the admium box with tested objet was entered at 91 m from the soure with the door losed. The purpose of this analysis was to determine how the spetra and the average spetral energy hange inside the admium box a funtion of its thikness. Figure 49. Cadmium-Polyethylene-Soure Configuration From the parametri analysis, it was observed that a very thin layer of admium will ompletely eliminate the thermal peak from the neutron spetrum as shown in Figure 50. It is also shown that the fast neutron spetra are not signifiantly affeted by the admium. 77

91 Figure 50. Cadmium Box Neutron Spetra Using a admium box is a simple and pratial method of exposing objets to a faster spetrum while inside the polyethylene shielding. The neutron spetrum inside the polyethylene shielding has a signifiant low-energy omponent, whih results from the fast neutrons interating with the wall and then sattering bak into the irradiation shielding. The admium box effetively eliminates this omponent. In the analysis, the average neutron spetral energy inreases from just under 0.8 MeV to almost 2 MeV with just 0.1 m of admium. For the admium thiknesses greater than 0.1 m, the fast spetra begin to gradually lose neutrons due to the resonanes and attenuation in the admium. The average energy for the neutron spetra is shown in Figure 51. As before, the ratio of thermal flux density to fast flux density as a funtion of admium wall thikness is given in Table XVII. 78

92 Figure 51. Cadmium Box with Varying Thikness The ratio between thermal to fast neutrons dereases substantially with the admium box as shown in Table XVII. With 1 mm of admium, the ratio dereases by a ouple of order of magnitudes, and it does not hange by muh thereafter. This explains the large gain in the neutron average energy. Table XVII. Ratio of Thermal to Fast Neutrons for Different Cadmium Thiknesses Thikness (m) R It is onluded that a admium box of a very small thikness will signifiantly inrease the average neutron energy, provided the admium box is thin enough to not absorb fast neutrons. 6.3 Polyethylene Shield between Tested Objet and Neutron Soure In this parametri analysis, several polyethylene slabs 2.54 m thik were added to the MCNP model of the neutron irradiation system. The results were evaluated at 91 m from the neutron soure. The number of slabs was varied from zero to ten slabs. Another thing to note is that 2.54 m of spaing were added between eah slab. The purpose of this setup was to derease the average neutron spetral energy by eliminating 79

93 the fast omponent from the soure and preserving the low-energy omponent from the wall inident on the irradiated objets. In Figure 52, showing the setup, the neutron soure is plaed in the far right hand orner of the hamber, and the polyethylene slabs are plaed between the soure and the 91 m detetor tally, and the hamber door is losed. Figure 52. Neutron Irradiation System with Polyethylene Shields The spetral analysis for this setup was plotted on Figure 53. The analysis showed that as the number of slabs inreased, the intensity of the fast neutron ontribution from the soure dereased. The low-energy peak was affeted to a muh lesser degree, indiating that the response from the hamber was influened less than the fast neutron portion of the spetrum by plaing the slabs in the hamber. The effet of the first four slabs on the neutron spetrum is greater than that ause by adding more slabs. 80

94 Figure 53. Neutron Spetra for Varying Number of Poly Slabs As expeted, plaing a shield between the soure and the tested objets dereases the average neutron energy. In Figure 53, it is noted that the slabs have the most effet for the first five slabs. Similar effets an be observed for the average neutron energy. The first five slabs are the most effiient in reduing the average neutron energy. In Figure 54, the average spetral energy is plotted as a funtion of number of slabs 2.54 m thik at 91 m and 122 m from the soure. At 91 m with the first five slabs, the energy dereases from 0.79 MeV to 0.31 MeV, about 0.48 MeV differene. Adding another five slabs derease the average energy from 0.3 MeV to 0.15 MeV, for about 0.15 MeV differene. Looking at this trend another way, the first five slabs derease the average energy by 60 %, while the seond five slabs only derease the average energy by 50 %. 81

95 Figure 54. Average Neutron Spetra Energy for Varying Number of Poly Slabs In Figure 54, the neutron tally was taken not only at 91 m, but also at 122 m to demonstrate that plaing the tested objet farther away from the neutron soure will also derease the average energy at the objet s loation. It is onluded that this method will derease the average neutron energy inident on the tested objets. The ratio between thermal and fast was determined at 91 m for varying slabs. The ratio inreases as the number of slabs inrease. Table XVIII shows the ratio of thermal to fast neutron flux densities as a funtion of the number of slabs. This table shows that the ratio of thermal neutron flux density to fast neutron flux density varies at approximately the same rate with the number of slabs as does the average neutron energy. 82

96 Table XVIII. Ratio of Thermal to Fast Neutrons at 91 m verses the Number of Slabs # of Slabs Thikness (m) R Low Energy Neutron Flux Configuration Using Polyethylene Blok In this parametri analysis, a blok of polyethylene was added to the MCNP model. It was plaed up against the north polyethylene wall. The tally was alulated at 91 m away from the neutron soure, and the thikness of the polyethylene blok was varied. The thikness of poly-blok was varied from 12.7 m to 63.5 m. The purpose of this analysis was to evaluate if plaing a poly-blok in the hamber will derease the average neutron spetral energy by reduing the fast omponent from the soure and preserving the low-energy omponent from the wall at the objet s loation. This arrangement is a variation of the poly-shield setup presented in Setion 6.3, but loation of the poly-blok inside the hamber is the key in reduing the average neutron energy spetrum. The arrangement is shown in Figure 55. The neutron soure is plaed in the far right hand orner, and the polyethylene blok is plaed on the northern wall between the soure and the 91 m detetor tally; the hamber door is losed. A small gap was left between the poly-blok and the top of the shielding, so that more neutrons oming from the hamber walls an arrive at the loation of the tested objet. 83

97 Figure 55. Polyethylene Blok Configuration Figure 56 shows a plot of the average neutron spetral energy as a funtion of polyethylene blok thikness. It an be seen that plaing the polyethylene blok in this onfiguration signifiantly dereased the average neutron spetral energy inident on a tested objet at 91 m from the neutron soure. This is an effetive method for reduing the fast neutron omponent of the spetrum. The energy at that loation with no blok present is 0.79 MeV, and with 30 m of polyethylene blok thikness, the average energy is 0.12 MeV, about 0.67 MeV differene. If another 30 m of polyethylene are added, totaling 60 m, this further dereases the average energy from 0.12 MeV to 0.06 MeV, about 0.06 MeV differene (a derease of another fator of two). Although the first 30 m is the most effetive in reduing the average neutron energy, this onfiguration makes it possible to bring the average neutron energy at the loation of the tested objet to as low as 60 kev. 84