TRIUMF: Canada's national laboratory for particle and nuclear physics

Shielding and activation in a 500 MeV H⁻ cyclotron facility Thorson, I. M. 1968

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


51833-TRI-1968-04.pdf [ 32.9MB ]
JSON: 51833-1.0107855.json
JSON-LD: 51833-1.0107855-ld.json
RDF/XML (Pretty): 51833-1.0107855-rdf.xml
RDF/JSON: 51833-1.0107855-rdf.json
Turtle: 51833-1.0107855-turtle.txt
N-Triples: 51833-1.0107855-rdf-ntriples.txt
Original Record: 51833-1.0107855-source.json
Full Text

Full Text

T R I U M FSH IE LD ING  AND A C T IV A T IO N  iN A 5 0 0  MeV H" CYCLOTRON F A C IL I T YI .M .ThorsonAppl ied M a th em a t i c s  Branch  Cha lk  R ive r  N u c l e a r  L a bo ra to r i e s  A tom ic  Energy  of C a n a d a  L im i tedUNIVERSITY OF ALBERTA SIMON FRASER UNIVERSITY UNIVERSITY OF VICTORIA UNIVERSITY OF BRITISH COLUMBIA TRI-68-AVTRI-68-4TRI-UNIVERSITY MESON FACILITYSHIELDING AND ACTIVATION IN A 500 MeV H" CYCLOTRON FACILITYI.M. ThorsonApplied Mathematics Branch Chalk River Nuclear Laboratories Atomic Energy of Canada LimitedPostal Address:TRIUMFUniversity of British Columbia Vancouver 8, B.C.Canada December 1968ABSTRACTThe shielding required for and the residual activity produced in the Tri-University Meson Facility (TRIUMF) are discussed. The facility is based on a 500 MeV H“ isochronous cyclotron capable of accelerating 100 pA of beam current. The 10 kW of beam power lost in the cyclotron by electromagnetic dissociation of the negative hydrogen ions will produce residual radiation fields in the cyclotron of the order of 1 rem hr-1 a few hours after shutdown following a long operating period. The estimated shielding required around the periphery of the cyclotron to achieve radiation fields of 2.5 mrem hr-1 at full power is approximately 9 m of concrete. The primary beam dump designed for use as a neutron irradiation facility can produce a useful neutron flux of 'v 5 x 1012 cm-2 s_1 in the graphite moderator surrounding a 10 cm diameter lead or bismuthC O N T E N T SPage1. INTRODUCTION 12. RADIATION PRODUCTION AND TRANSPORT2.1 Biological Radiation Hazards 22.2 Radiation Energy Transport 22.3 Thermal Neutron Radiation 122.A Residual Activation 133. CYCLOTRON 163.1 General Layout 163.2 Operating and Residual Radiation Fields inthe Main Cyclotron Vault 173.3 Operating Radiation Fields and ActivationConsiderations Outside the Main CyclotronVault 233.4 Operating and Residual Radiation Fields in thethe Cyclotron 264 . EXPERIMENTAL AREAS 334 .1 Primary Beam Tunnel 334 .1.1 Beam Transport System 334 .1.2 Meson Targets 354 .2 Experimental Area P 385. NEUTRON IRRADIATION FACILITY 4O5.1 Neutron Source and Flux 4O5.2 Heat Production 4 l5.3 Shielding 435.4 Activity Production 4A6. DISCUSSION OF ESTIMATES 47Acknowledgements 49References 50FiguresLIST OF TABLESTABLE I TABLE I ITABLE I I I TABLE IVTABLE VTABLE VITABLE VI ITABLE VI I TABLE IXInter-nuclear Cascade Results for Thin Carbon and Lead ShieldsInter-nuclear Cascade Calculation Parameters and Results for C, A 1 , Cu and PbRelaxation Lengths for Various Forms of Rad i at ionResidual y-Radiation Fields from Spallation Products at Concrete Surface of Peripheral Cyclotron ShieldResidual y-Radiation Fields from Spallation Products at the Surface of Thick Slabs of Material at the Outside of Concrete ShieldsResidual Radiation Fields at the Centre of the TRIUMF Cyclotron from Various SourcesCoolant Water Activities for 2 g cm-2 Carbon Targets in the Primary Proton BeamI Secondary Distribution for 500 MeV Protons on 40 cm long Bismuth TargetNeutron Facility Target Activity for Pb and BiLIST OF FIGURESPhoton and Nucleon Flux to Deliver 2.5 mrem h vs EnergyEnergy Flux to Deliver 2.5 mrem h"1 vs EnergyNeutron Inelastic Cross-SectionsCascade Nucleon Collision Energy vs Depth in Carbon for One 500 MeV Proton IncidentCascade Nucleon Collision Energy vs Depth in Lead for One 500 MeV Proton IncidentCascade Neutron Collision Energy per 500 MeV Proton on AluminumCascade Neutron Leakage Spectrum through 1000 g cm Plane Slab for One 500 MeV Proton IncidentAngular Distribution of Secondary Cascade Neutron Energy from 500 MeV Proton CollisionsPlan View of TRIUMF with ShieldingElevations with ShieldingVertical Section of the TRIUMF CyclotronEvaporation Neutron & Heat ProductionINTRODUCTIONThe primary extracted H- ion beam from the TRIUMF isochronous cyclotron will be 100 pA at 500 MeV. Other beams may be extracted from the machine but these will be limited in intensity to 10 pA or—y —^less. Proton beam spill in the cyclotron is mostly due to v x B dissociation of the H“ ions which are then lost tangentially into the vacuum tank wall, magnet yokes, beam transport components or shield­ing at the cyclotron mid-plane. The v x ^ stripping is strongly energy dependent and is expected to spill approximately 10 kW when the cyclotron is operated at 500 MeV and 100 pA beam current, mostly at energies above 450 MeV. The spill will have an azimuthal varia­tion due to the modulation in B around a proton orbit. The modulation has been ignored in making the shielding and activation est i mates.Section 2 of this report deals briefly with some of the more relevant aspects of radiation production and transport and biological effects.In Section 3 estimates are made of the radiation fields and shielding required for various points around the cyclotron when it is operating, and residual activation in the cyclotron and vault after it has been shut down. Estimates are also made of the radiation damage problem at selected points in the cyclotron.Section 4 outlines the estimates of radiation and residual activation and shielding required for the main beam transport system and the other experimental areas.Section 5 outlines the important features of the primary beam dump designed for use as a neutron irradiation facility and estimates approximately the spatial distribution of the heat production in such an assembly.- 2 -2. RADIATION PRODUCTION AND TRANSPORT2.1 Biological Radiation HazardsFigure 1 shows the particle flux as a function of particle energy for neutrons, protons and photons to deliver 2.5 mrem h-1, the average maximum-permissib1e-whole-body dose rate recommended for AO hours occupation per week. Figure 2 shows the energy flux for the same condition as Figure 1. The energy flux for photons of energy greater than 100 keV is fairly constant at 'v 2.5 x 10-10 W cm-2. The range of protons with energies below 50 MeV is sufficiently short - a few g cm-2 - that they do not constitute a significant fraction of radiation leaking through thick shields. The solid line for neutrons is the product of their kinetic energy and flux to deliver 2.5 mrem h_1; thermal or near- thermal neutrons are hazardous not only because of their kinetic energy but also for the y-radiation produced when they are captured. The y-ray energy release when neutrons are captured by protons,2.2 MeV, has been added to their kinetic energy to give the dotted line representing the "total" neutron energy. At thermal energies it approaches the value for y-rays; this is a manifestation of the fact that the hazard from thermal neutrons arises from the y-radiation they produce when captured in organic material. Neutrons of about 1 MeV kinetic energy are the most insidious as far as biological effect is concerned, and if all the radiation leaks out of the shields in this form, an energy flux of 10-11 W cm-2 will deliver 2.5 mrem h-1. Since experience has shown that a significant fraction of the total dose rate leaking through accelerator shieldingpcomes from neutrons in this energy region, we shall adopt 10-11 W cm-2 as a working health tolerance.2.2 Radiation Energy TransportRadiation shielding is essentially an energy degradation process.When high energy (500 MeV) protons are incident on matter, a wide variety of nuclear and atomic interactions are involved in the degradation of the energy, ultimately to heat. The transport of this- 3 -energy through thick shields is generally controlled by the forms this energy takes during the degradation processes that have the largest spatial relaxation lengths. Both the electromagnetic inter­action (ionization) and nuclear collisions are important energy loss mechanisms for the primary protons. The strongly interacting secondary charged particles from the nuclear collisions, such as protons and charged pions, also lose energy by both mechanisms, but the strong inverse relationship of ionization energy loss rate with particle energy quickly makes this mechanism dominant. Table I! shows the range for 100 and 500 MeV protons in C, A 1 , Cu and Pb.The energy carried away by the secondary neutrons produced in the nuclear interactions is degraded by further nuclear collisions only. Thus, virtually all of the power that appears at great depth in shields is transported there by the high energy secondary, and higher order, neutrons. Figure 3 shows the total non-elastic cross-section for four elements as a function of neutron energy. At energies above a few tens of MeV the angular distributions of neutrons suffering elastic scattering are peaked so sharply in the forward direction that it is not a significant interaction mechanism from the shielding point of view. At neutron energies above 100 MeV the experimental data shown in Figure 3 indicate non-elastic cross-sections constant with energy and approximately proportional to A2/3; this dependence can be used to interpolate the data for the non-elastic cross-sections of other elements in this energy region.The relaxation of the radiation energy with distance in structures is dependent not only on cross-sections for the various reactions but also on the angular distribution and spectra of the secondary particles. For non-elastic collisions of high energy nucleons this is particularly important because the angular distributions are highly anisotropic, a significant portion of the spectra extend up to the incident energy, and the two distributions are highly correlated. The experimental information is not comprehensive enough to allow calculation of the energy transport in bulk matter. However, the- b -results from Monte Carlo calculations for various isotopes and for nucleon energies up to 400 MeV have been used to make such estimates.While the agreement between these M.C. results and the experimental1+data is not complete, they are believed to be sufficiently reliable for this purpose. The model used in the M.C. calculations describes the non-elastic collision as a two-step process. Initially, the incident nucleon interacts directly with one or more nucleons in the target nucleus emitting nucleons and/or pions in an intra-nuclear cascade process. At the end of this process the residual nucleus is left in a highly excited state. Most of this excess energy is emitted as kinetic (and separation) energy of "evaporation" nucleons or nuclear fragments. The distribution of the incident nucleon energy among the various secondary particles is dependent on the mass of the target nucleus. At low mass most of the energy is emitted as kinetic energy of the secondary cascade particles; at high mass a substantial fraction of the energy is carried off by the evaporation neutrons.To estimate the radiation energy transport by the inter-nuclear cascade of high energy nucleons in thick shields of various materials, numerical calculations have been done on the Chalk River G-20 computer using the codes TRANSEC and MARKN. The TRANSEC code calcu­lates the transfer cross-sections for the production of secondary protons and neutrons in 15 energy groups and b angle segments from the parametric fits of Alsmiller et al.5 to Bertini's results. The MARKN code calculates the non-elastic collision probability as a function of energy and position in an infinite plane slab. The probability of a nucleon in the i angle group and the j energy group suffering a k generation collision in the £ layer of the slab is estimated by integrating over all i, j, £ groups for the k-1 generation collisions. The calculation is repeated for successive generations of collisions until the results no longer make a significant contribu­tion to the total non-elastic collision probability. The cascade nucleon leakage currents and spectra are also summed for each generation of collision as are the proton ionization energy loss and3- 5 -the number of protons stopping in each layer. Energy transport by mesons is ignored in the calculations. An ad hoc extrapolation of the multiplicity of secondary cascade nucleons was used for the 400 to 500 MeV energy region; the spectra were estimated from the 400 MeV data by a linear expansion of the energy scale. The inter­actions of the evaporation particles are not treated in detail; only the number produced and their total kinetic energy are tabulated. Virtually all of the evaporation proton energy will be dissipated by i on i zat i on.The results of TRANSEC-MARKN calculations for relatively thin carbon and lead slabs are shown in Table I. For all cases one 500 MeV proton is incident on the slab in the angular interval - relative to the slab normal - shown at the top of the column of results. For the calculations the carbon slabs were divided into 5 equal layers and the lead slab into 10 layers. The number and total kinetic energy of the cascade protons and neutrons leaking out of the slab are shown for the various angular intervals. The number and total kinetic energy of the evaporation particles produced in the slabs are calculated from the collision probabilities for each energy group and Alsmiller's tabulations5 of the evaporation results. The number of protons listed at the bottom of each column includes the incident proton and represents the total number either stopping in the slab or leaking out. The number of neutrons is the sum of the number leaking out as cascade neutrons and the evaporation neutrons produced in the slab; an insignificant number of cascade neutrons remain in the slab at the end of the inter-nuclear cascade calculation.The difference between the total energy accounted for in Table I and the 500 MeV incident proton kinetic energy goes into separation of the neutrons and protons. The binding energy of the neutrons will generally be released as y-radiation or a-particle energy when theyare captured. Thus we see from Table I that while the cascadeparticle energies emerging from the 400 g cm-2 slab of lead (15.4 MeV)are substantially lower than those from the 100 g cm-2 slab ofgraphite (48.8 MeV), the total amounts of energy leaking out of eachTABLE IINTER-NUCLEAR CASCADE IN CARBON AND LEADCARBON LEAD CARBONIncident Angle 30°-60° 30°-60° 60 o_g0°SI ab 100 g cm-2 1*00 g cm-2 50 g cm" 2Th i ckness ^ 50 cm 'v 35 cm 'V 25 cmCascade Proton Leakage No. Energy, MeV No. Energy, MeV No. Energy, MeV0 - 30° .06 2.9 - - .03 1 .230 - 90° .02 1.1 - - .16 8.390 - 180° .09 3.7 .015 1 .0 .65 42.8Cascade Neutron Leakage0 - 30° .09 9.3 .030 2.6 .06 3.730 - 90° .25 23.5 .044 3-9 .22 20.790 - 180° .16 8.3 .18 7.9 .21* 19.8Evaporation ProtonProduct i on 0.7 3.6 .3 3.6 0.6 3.0Evaporation NeutronProduct ion 1.1 3.9 11.1 24 .7 0.8 2.9Proton Ionizationand Stopping 2.1 422 1.1* 376 1.7 377Total 3.0 protons 479 1.7 protons 422 3.1 protons 4801 .6 neutrons 11.1* neutrons 1 .3 neutronsslab, including that carried by the evaporation neutrons, are comparable. Approximately one third to one half of the evaporation neutron kinetic energy would be dissipated by inelastic nuclear collisions in the lead slab, but the remaining energy will be absorbed in 1ight-element moderators outside the slab.The spatial distribution of the total radiation energy suffering a collision in a unit mass of a homogeneous shield can be used to define an effective radiation-energy-removal relaxation length. For high energy nucleons their total energy is adequately approximated by their kinetic energy; for neutrons at lower energies their bind­ing energy in the capturing nuclei should be added to their kinetic energy. Figures 4a and 4b show the cascade proton and neutron collision energy in 1000 g cm-2 slabs of carbon and lead as a function of position in the slab when a 500 MeV proton is incident in the angular interval 0-30° to the normal. At depths beyond the range of the incident protons, nearly all of the energy is carried by the cascade neutrons, and the ratio of the energy carried by the protons and neutrons is nearly constant. In this equilibrium region the exponential relaxation length is always found to be nearly constant with position but is dependent on incident particle, energy and angle as well as shield material.Figure 5 shows the neutron collision energy as a function of depth in a plane slab of aluminum when 500 MeV protons are incident on one face in the angular intervals 0-30°, 30°-60° and 60°-90° from the slab normal. The relaxation length decreases for higher angles of incidence because of the softer spectrum of secondary nucleons at larger angles to the incident proton direction.Table II lists some of the more important parameters and results for inter-nuclear cascade calculation for 500 MeV protons incident at 0-30° on 1000 g cm-2 slabs of carbon, aluminum, copper and lead. The range for 100 and 500 MeV protons and the non-elastic mean free path for nucleons above 100 MeV were used in the MARKN code. The radiation relaxation length Ar was found from Figure 4 for carbon and lead and similar graphs for aluminum and copper. The ratio Ar/Ane is- 7 -- 8 -TABLE I IINTER-NUCLEAR CASCADE PARAMETERS AND RESULTSC A1 Cu PbProton Range, g cm 2100 MeV 500 MeV8.57130.09-94147.811.8170.316.3224.4A , q cm-2 neNon-elastic mean free path above 100 MeV 99 112 143 204A , g cm-2r » aRelaxation length for cascade energy 120 127 151 196A /A r ne 1 .21 1.131.06 0.96Cascade Neutron KE, MeV leaking through 1000 g cm-2 .027 .045 .12 .49Equivalent Cascade Neutron Source Energy MeV110 110 90 80Evaporation Neutron and KE, MeV1 .2 26.1- 9 -a decreasing function of mass, due primarily to the softer secondary cascade nucleon spectra from the heavier elements compared to that from lighter elements. This is further illustrated in Figure 6 where the spectra of cascade neutrons leaking through the 1000 g cm-2 slabs are plotted. For the heavier elements a larger fraction of the energy goes into nuclear excitation and subsequent evaporation processes. Table II lists the energy leaking through the 1000 g cm-2 slabs as cascade neutron kinetic energy as well as the equivalent cascade neutron energy source found by extrapolating these values back to the original side of the shield with the relaxation length Ar.The ratio of the effective source strength to the actual source is often defined as the "build-up" factor. This "build-up" factor for 500 MeV protons incident on the various materials is substantially less than unity, being in the range 0.16 to 0.22 for all of the elements estimated. This low value reflects the relatively large fraction of energy absorbed by ionization from the primary protons and secondary charged particles and the fraction of energy going into neutron kinetic energy in the energy region below ^ 100 MeV, where the non-elastic mean free paths are substantially shorter than at higher energies.Figure 7 shows the angular distribution, estimated from Bertini's results, of the secondary cascade neutron kinetic energy when 500 MeV protons suffer non-elastic collisions with nuclei of various mate­rials. The intensities are less than for the isotropic case for angles greater than ^ 70° to the incident proton direction. The angular distribution resulting from the inter-nuclear cascade in thick shields does not show such a pronounced anisotropy because several generations of collision contribute significantly to the intensity. Figure 7 also shows the effective cascade neutron source energy for radiation leaking through a thick aluminum shield when 500 MeV protons are incident at various angles to the slab normal.The angular distributions for 500 MeV protons were assumed to be the same as Bertini's 400 MeV proton results.Table III shows the characteristic spatial relaxation parameters for- 10 -TABLE I I IRELAXATION LENGTHS FOR VARIOUS FORMS OF RADIATIONXr - 9g cm 2L5 -2 g cm z LT _ 2g cm zA0.5„2 g cm zA3.5_2 g cm zc 120 32 88 11.6 31.1A1 127 M 4 0 43 12.0 30.7Fe 151 9,600 10 12.2 29.3Pb 196 'vl 400 146 6.6 24.6U NI 120 9 3 10.4 27.8S i 0 2 123 ^70 44 11.5 29.8Std. Cone.* 126 35 21 11.5 29.8lim. Cone.* 142 47 8 12.0 29.8* Composition of standard and ilmenite concrete is given in Reference 7-11 -several forms of radiation in various possible shielding materials.Ar is the relaxation length for cascade nucleon energy estimated forC, A 1 , Fe and Pb by the methods outlined above for 500 MeV protonsincident at 0-30°. For the other materials the A values have beenrinterpolated from those for the elements listed on the basis of con­centration and atomic weight. The value for iron is assumed the same as that for copper.Ls and L^ . are the slowing down lengths for the evaporation neutrons and the diffusion length for thermal neutrons, respectively. The values preceded by an ",v" are rough estimates for illustration purposes only. The other values for Lg are mostly based on data for fission neutrons with a correction term added for the harder evaporation spectrum. Aq and A^ ^ are the mean free paths for y-radiation of 0.5 MeV and 3.5 MeV photon energy, respectively.The energy removal formulation of the radiation transport problem in thick shields is only useful where the effective removal relaxation lengths are approximately constant and can be accurately determined. These requirements are satisfied for the thick shields (1000 g cm-2) as can be seen in Figures 4a and 4b, provided that the radiation energy transport is dominated by the cascade neutron component. This condi­tion, in turn, requires that the relaxation lengths for all other forms of radiation accompanying the cascade neutrons be short compared to that for the cascade neutrons. As we see from Table III, this is satisfied directly by water and concrete. Carbon also satisfies the criterion, provided the thermal neutrons can be absorbed in an alpha emitter such as boron, thus depositing the neutron binding energy in the absorber rather than releasing it as y-radiation in (n,y) reactions. The boron need not be distributed in the carbon, provided that complete coverage of the shield volume by the boron is feasible.As is well known, shields for radiation sources producing fast neutrons must contain 1ight-element moderators such as hydrogenous materials. The 'v 0.5% by weight of hydrogen in standard concrete is adequate for accelerator shields where only a few per cent of the initial energy goes into evaporation neutron energy. In iron the hydrogen concentration for minimum shield thickness should be 'v 0.1 toQ0.2% by weight depending on total shield thickness.- 12 -2.3 Thermal Neutron RadiationThe evaporation and lower energy cascade neutrons lose energy in the shields and machine components by inelastic collisions at energies 9* 1 MeV and by elastic collisions at lower energies. These neutrons generally approach or achieve thermal energies before they are captured. The complicated geometries involved effectively preclude accurate estimates of the absolute level, spectra and spatial distributions of such neutrons. Since these neutrons cannot be ignored in making estimates of activation and operating radiation fields, recourse must be made to empirical evidence, which at least definesupper limits for the fields.yThe empirical relation for the thermal neutron flux is based on experimental measurements in a fairly small (101 cube) concrete wall vault containing a Po-Be neutron source. The thermal flux is given byF = 1-3 Q/A cm-2 s_1where Q is the fast neutron source strength and A is the surface area of the interior vault walls. This relation should hold for larger vaults until the linear dimensions exceed the thermal neutron diffusion length in air of 'v 20 m.For tunnels some unpublished work at SLAC10 indicates that for distances up to 1001 the thermal neutron flux from a centrally located point source of Q neutrons s-1 in a 101 x ll1 concrete wall tunnel is given byxF(x) = 1.5 x 10“6 Q e 90 cm-2 s"1where x is the distance along the tunnel in cm from the source. Integration of this expression over x to estimate the result of a continuous line source givesF = 1.5 x 10-3 q cm-2 s-1where q is the source strength per cm of tunnel length. Generalizing this expression for tunnels of other sizes on the same basis as for- 13 -the completely enclosed vaults givesF = 1.9 q/P cm-2 s"1 where P is the perimeter area per unit length of tunnel.For concrete containing elemental boron in concentrations of10 mg cm-3 the thermal neutron flux in concrete tunnels is reduced10by a factor 'v 15.2.4 Residual ActivationThe most serious potential problem arising from beam spill in theTRIUMF cyclotron and beam transport system is the residualactivity produced. Barbier and Cooper11 and Barbier12 have tabu­lated both experimental data and calculated estimates for activityproduction by nucleons in the energy range of interest for TRIUMF.1 3They used Rudstam’s empirical formula and the known decay data toestimate the residual radiation fields for all elements between12Mg and 83Bi. The agreement between the experimental results and estimates is quite good, both as to relative intensity between elements and time dependence of the radioactive decay. Notable exceptions to this situation are Co and Ni, which are underestimated by a factor 'v 3, and Pb, which is overestimated by a factor 'v 5, after normalizing to the aluminum data. Part of the discrepancy in the Pb results may be due to self-shielding of the residual y-rad i at ions.The residual activity in elements below 12Mg can be estimated fromthe experimental data for the more important products. For carbon the dominant products are C 11 and Be7. The C12 (n ,2n)C11 cross- section is fairly constant at approximately 25 mb in the neutron energy region 50 MeV to 500 MeV, falling at lower energies to the threshold at 22 MeV.14 The C12(p,pn)Cn  cross-sections have athreshold at 19 MeV with values of 80 mb at 50 MeV, 60 mb at 100 MeV and 30 mb at 500 MeV.15 The C12(p,3p3n)Be7 cross-section varies smoothly from 14 mb at 60 MeV incident proton energy to 8 mb at 320 MeV.15 Using these data as a guide we shall adopt 10 mb as the- 14 -Be7 production cross-section by neutrons or protons in the energy interval 50 to 500 MeV. The oxygen cross-sections of most interest are those for the 0 15(n,p)N16 reaction, 'v 30 mb above the threshold energy of 11 MeV17 and the N13, C 11 and Be7 production cross- sections by high energy nucleons. Extrapolation from experimental results at higher energies than those encountered in TRIUMF suggest values of 6, 12 and 10 mb, respectively, for these reactions.16The cascade neutron spectrum leaking through thick concrete shields should be similar to that shown in Figure 6 for aluminum. To make estimates of the spallation activation outside thick concrete shields from the data in reference (11), the cascade neutron spectrum was divided into 3 energy regions. The particle flux in aluminum below 75 MeV is approximately 30% of the total, between 75 MeV and 200 MeV,50% and above 200 MeV.20%.From the MARKN calculations for the 1000 g Cm”2 slab of aluminumthe total leakage energy of .045 MeV per incident proton correspondsto a total leakage current of 'v 3-5 x 10-1+ neutrons. Thus the normalization factor relating the power and neutron flux, assuming a cosine angular distribution of the neutron leakage current, is0.045 x 106 , , ,  T— Z TTTT---- 2 o----, a1~8~ = 10 u W s  n L .2 x 3.5 x 10 H x 6.2 x 1010Activation by thermal or near-thermal neutrons varies greatly with the particular element. It can always be suppressed in hardware components by the addition of non-activating absorbers such as cadmium or boron. Air activation by neutron capture in Ar40 produc­ing the 1.8 hour half-life Ar41 can, however, only be controlled by general suppress ion of the neutron flux.The macroscopic cross-section for producing Ar41 by thermal neutrons in air is 1.5 x 10”7 cm-1. The maximum permissible concentration18 of Ar41 for 40-hour-per-week occupation is2yCi/m. This concentra­tion would be achieved at equilibrium in a thermal neutron flux of 'v 5 x 10 5 cm-2 s-1.- 15 -The principal activity produced in concrete by thermal neutrons is the 15 hour half-life Na2l+ activity. For concrete containing 2%(by weight) of sodium and no significant amounts of absorber such as B10, a thermal neutron flux of 5 x 105 cm-2 s"1 will produce saturated residual radiation fields of approximately 15 mrem h-i at the surface of a thick concrete slab. A concentration of 0.3% (by weight) of Mn in the concrete will make a similar contribution to the saturated residual fields; however, the Mn56 half-life is only 2.6 hours. For a given neutron source the activity production rate by thermal neutrons in concrete containing 10 mg cm-3 of elemental boron is less by a factor 40 than in concrete containing no boron.- 16 -3. CYCLOTRON3.1 General LayoutA plan view of TRIUMF at the level of the beam line is shown in Figure 8. Several elevation sections as indicated in Figure 8 are shown in Figure 9. These show the gross features of the shielding proposed for the cyclotron. Figure 10 is a section of the cyclotron showing the more important machine components with regard to shielding, activation and operating radiation fields.The H" ions are accelerated in a very narrow band in the horizontal mid-plane of the vacuum tank between the poles of the sector-focus magnet system. The ratio of the betatron oscillation frequency to the azimuthal ion frequency is 'v 1, and the amplitude of the betatron oscillation is ^ 2 cm. Thus the neutral hydrogen atoms, lost tangentially from the focusing system when the extra electron is stripped by the v x ^ electric field, are incident on a band approximately 2 cm wide at the circumference of the vacuum tank. Tominimize the activation in machine components such as end walls on the accelerating dee-structure and vacuum tank wall, the amount of material at the mid-plane inside any shielding must be kept to a minimum. At this point the paramount consideration in the selection of shielding material is low residual activation; thus a ring of graphite of basic cross-section dimension 'v 50 cm by 50 cm should be installed between the main magnet windings, as shown in Figure 10.A 3.7 m (121) thick concrete shield is shown around the main accelerating structure in Figure 10 and sections F-F and H-H of Figure 9. This shielding reduces the radiation intensity at the out­side of the shield sufficiently that induced residual activation in the main vault area does not significantly impede maintenance operat i ons.The A.9 m (16’) thick concrete shield or equivalent earth fill, on all sides of the cyclotron vault, reduces the radiation fields when- 17 -the machine is operating to the maximum permissible level for 40-hour-per-week occupation.The radiation field immediately above the 4.9 m thick removable concrete roof will be substantially higher than that for normal working occupation. It should not, however, cause radiation fields that restrict occupation of local control areas or any other normally inhabited areas.3.2 Operating and Residual Radiation Fields in the Main Cyclotron VaultThe effective cascade neutron source energy when 500 MeV protons are incident on carbon, as shown in Table II, is 110 MeV. This estimate is based on incidence of the primary protons in the angular interval 0 to 30° to the plane slab normal. The actual angle of incidence of the primary protons on the graphite ring shield shown in Figure 10 depends on the point where the H“ ions are stripped and hence on ion energy. The angle of incidence varies from approximately 55° at 400 MeV to 67° at 500 MeV. The assumption of isotropic scattering of the radiation energy in estimating the radial leakage will thus be conservative; the actual reduction due to anisotropy has not been estimated precisely and will be used only as a factor of safety.Assuming a graphite density of 1.6 g cm  ^ and 50 cm radial thickness in the horizontal plane, the graphite ring shield would transmit approximately 50% of the effective cascade neutron energy from a radial source. From Table I I I the estimated relaxation length for cascade neutron energy in standard concrete is 126 g cm-2. Thus the 3.7 m concrete shield at the circumference of the main machine will attenuate the energy by a factor- 370x2.4 e ^  = .00086in the radial direction. Assuming an isotropic angular distribution of the cascade neutron energy emerging from the graphite shield, the- 18 -solid angle geometry factor at the outside of the 3-7 m concrete shield isVERSTY O NFFERSTALRTY B RST O M CRH - 6T H CThis expression can be generalized for field points in the horizontal mid-plane of the machine at a distance r from the graphite ring toG(r) = 2F7T(r+830) ’Thus, on the isotropic assumption, the cascade neutron energy current at the outside of the shield for a total power loss in the cyclotron of 10 kW is,0 x 103 x H f  x 0.50 x 0.00086 x 0.36 x 10"6= 3 . 6 x 10-7 W cm-2.The operating radiation field at the outside of the shield is then approximately2 x 3-6 x 1]°~- x .0025 = 180 R h"1 10_11due to radiation leaking through the shield, assuming a factor 2 for conversion of energy current to energy flux.Based on the normalization factor relating cascade neutron energy andparticle currents (2 x 10-11 W s n-1) and the leakage spectrum of1 1 ,  ,. ,Figure 6 for aluminum, the data of Barbier and Cooper for Si have been used to estimate the residual radiation fields from the concrete. The residual activity field estimated in reference (11) are tabulated in two forms. One form is the tabulation of unshielded field strengths for specified geometry and irradiation and decay histories. The other form is tabulation of danger parameters for specified irradiation and decay histories. The danger parameters involve the self-shielding properties of the irradiated material and are defined to be the residual field strength in a cavity in an infinite mass of uniformly irradiated material for various decay periods. The correct,- 19 -self-shielded residual y-radiation fields for other geometries are estimated by multiplying the danger parameter values by the fraction of 4 it solid angle subtended at the field point by the source. For example, the solid angle factor for an infinite plane slab is 1/2. Table IV shows the danger parameter values from reference (11) for various decay periods following a very long irradiation of silicon in a nucleon flux of 106 cm-2 s-1, for nucleon energies of 50, 100 and 600 MeV. The fractions f. of the cascade neutron flux in eachienergy group for the spectrum leaking through a thick aluminum shield are also listed in Table IV. The residual y-ray field strength values (RGF) are estimated by the expression3-6 x 10-7 1 . _ \RGF(mR) - „ 8 - „ 6 -  2 x ]0_„ J -  ? f, x(D.P.)rThe factor 0.8 is an allowance for the increased cascade neutron flux at a depth of 30 g cm-2 inside the concrete shield, the factor 0.6 converts the leakage cascade neutron current to flux and the factor 1/2 is the infinite-plane geometry factor for the danger parameter coefficients.The other constituents do not appreciably change the residual1 2activity properties of the concrete; the self-shielding afforded by the oxygen presumably compensates for the higher specific activa­tion in the other species.Table V shows the y-radiation fields after a long irradiation at the surface of thick slabs of concrete, aluminum, iron, copper and lead. The estimates are also based on Barbier and Cooper's11 data and the same cascade neutron flux as for Table IV. To estimate the residual field for shorter irradiation periods, the field strength at the irradiation plus decay period is subtracted from the field strength for the decay period. For example, the field at a concrete surface after a one-hour decay period following a 23-hour full power operating period would be1.4 - 0.6 = 0.8 mR hr-1.- 20 -TABLE IV84DG50e" . h8e5GeoGrs nG4"5D n8rp Dle""eoGrs l8r50ioD eo irsi84o4 D08nei4 rn l48Gld48e" iai"ro8rs DdG4"5DECAY TIME DANGER PARAMETER FOR Si RESIDUAL FIELD mR h -1E , MeV n 25-75(50)75-200(100)200-500(600)Spectrum Fract i on 0.3 0.5 0.21 hour 45 90 66 1 .41 day 17 37 31 0.61 week 2.4 9 13 0.151 month 2.3 9 12 O.l4TABLE V84DG50e" . h8e5GeoGrs nG4"5D n8rp Dle""eoGrs l8r50ioD eo od4 D08nei4 rn odGit D"emD rn peo48Ge" eo od4 r0oDG54 rn irsi84o4 DdG4"5DDECAY TIME RESIDUAL FIELD mR r 1Concrete A1 Fe Cu Pb1 hou r 1 .4 2.6 1-13 1.95 0.881 day 0.6 1.1 1.07 0.72 0.211 week 0.15 0.250.88 0.68 0.0751 month 0.14 0.25 0.69 0.63 0.023- 21 -At field points away from the main shield surface the y-radiation intensity will decrease for geometrical reasons by a factor which is always less than (r^/r)2 , where rQ is the shield radius and r the field point radius, both measured from the centre of the cyclotron. Attenuation by the air in the vault is not significant for either the cascade neutrons or residual y-radiation.The evaporation neutron production near the outside surface of the peripheral shield surrounding the cyclotron will depend on the materials in the vicinity. Assuming that the inter-nuclear cascade calculations for aluminum are representative of concrete, the total evaporation neutron production within one slowing down length of the concrete shield surfaces is approximately 0.3 x 10"3 per 500 MeV proton lost in the cyclotron. Thus for a spill rate of 20 yA the effective source strength from theconcrete shield would be0.3 x 10“3 x 20 x 6 x 1012 = b x 1010 s_1.This is not, however, the largest potential source of evaporation neutrons. The evaporation neutron production in the graphite ring shield is approximately 0.9 neutrons per spilled proton. Because of the complicated geometry of the magnet assemblies in this region, it is very difficult to estimate the probability of these neutrons escaping into the main vault.Most of these neutrons will escape the graphite ring either as thermal or epithermal neutrons. Approximately 0.2 neutrons per incident proton will escape back into the cyclotron structure through the interior ring wall with epithermal energy. Another 'v 0.2 neutrons will leak back through the interior ring surface as thermal neutrons maintaining a thermal neutron flux in the immediate vicinity of the order of 108 cm-2 s"1. These neutrons must be absorbed in a non-activating material such as cadmium or boron. If they are incident directly on standard concrete, for example, the saturated Na21+ activity y-radiation field at the- 22 -concrete surface will be 'v 5 rem h- 1 . Absorption in boron is preferable, to avoid the neutron capture y-radiation fields of the order of 1000 rad h-1 in the immediate vicinity of a cadmium absorber.Thermal neutrons will be absorbed with a relaxation greater than one decade per 0.06 g cm-2 layer of boron or 0.18 g cm-2 layer of cadmium; three or four decades of attenuation should effectively suppress any problems arising from the thermal neutrons in the graphi te.Assuming that 1/2 of the neutrons leaking back towards the centre of the machine emerge eventually as epithermal neutrons into the main vault, the source strength in this region is thenj x  0.2 x 20 x 6 x 1012 = 1.2 x 1013 s"1.This component would completely dominate the source of evaporation neutrons at the outside of the concrete shield and would, according to the empirical recipe cited above for concrete vaults,3 sustain a thermal flux level of1.3 x 1.2 x 1013/ 2 .9 x 107 = 5 x 105 cm-2 s-1in the vault. At saturation in such a flux the 2.56 hour Mn56 residual y-radiation activity at the surface of a thick slab of steel containing 1% manganese would be approximately 10 mR h _1.The longer-lived residual activity components in stainless steel would give y-radiation fields between one and two orders of magnitude lower than this. The saturation level Ar1*1 content of the air in the vault in such a thermal neutron flux is just equal to the maximum permissible concentration for **0 hours per week occupation. The capture y-radiation fields at the surface of a thick stainless steel slab would be 'v 2 R h"1 in the 5 x 105 cm-2 s-1 thermal neutron f 1 ux.- 23 -3 .3 Operating Radiation Fields and Activation Consideration Outside the Main Cyclotron VaultThe radiation field outside the main vault at the plane of the beam line can be estimated by extrapolating the radiation energy current of 3.6 x 10-7 W cm-2 at the outside of the main cyclotron shield to the outside of the main vault shield. The energy current will have decreased by a factor370 x 1200 _895 x 1725 "due to geometry and_ 490 x 2.4 e ^  = 0.85 x 10_1+due to attenuation by the 4.9 m (16') concrete shield. Thus the radiation power through the shield is3.6 x 10"7 x 0.29 x 0.85 x }0~h= 0.9 x 10-11 W cm-2.Assuming a factor 2 for current-to-f1ux conversion this is approxi­mately double the maximum-permissible-whole-body dose rate for 40 hours per week occupation if all the radiation energy leaks out as 'v 1 MeV neutrons.The residual activation level outside the 4.9 m main vault concrete shield will not be significant. The induced activity in the sand or earth-fi11 shielding outside the 2-1/21 thick concrete walls on the two external sides of the main cyclotron vault can be estimated from Barbier and Cooper's data.11 Assuming a 1/e relaxation length in the earth-fill of 126 g cm-2, as in the concrete, and a cascade neutron energy current of 3 x 10-8 W cm-1, then the total equilibrium activity level of species with lifetimes longer than 1 hour is 'v 10 millicuries of y-radiation and 'v 5 millicuries of g-radiation. Most of this activity will be fixed in the soil, and in any case at the maximum concentration of 0.05 yCi kg-1 should not present any hazard to ground water.- 24 -The radiation field above the 4.9 m (16') roof shield over the main cyclotron vault is more difficult to estimate accurately because of the irregular geometry of magnet pole assemblies in the region of the magnet windings and graphite ring shield. From Table I, the cascade neutron energy reflected back into the ^ 2tt space inside the carbon ring when a 500 MeV proton is incident in the angular interval 30°-60° to the ring normal is 8.3 MeV. For protons incident at 60°-90° to the ring normal, the reflected cascade neutron energy is 19-8 MeV.Since most of the protons are incident on the graphite ring shield at 'v 60° to the normal, an average effective return leakage of 15 MeV per proton was interpolated for 500 MeV protons incident on the ring shield. This current is not isotropic, being concentrated around the mid-plane of the cyclotron. The angle between a field point at the centre of and above the concrete vault roof shield is 'v 90°, and the assumption of isotropic or average cascade neutron kinetic energy in this direction can be made with confidence. Thus, assuming no effec­tive shielding by the magnet pole assemblies, the cascade neutronenergy current through the roof shield at the centre would be1Ql+ x 5^ x 0-5 X 10_1+ X 0,5 X 10"7= 75 x 10-11 W cm"2for a total proton spill power of 10 kW. The factor 0.5 x 10-i+ is the attenuation factor for the cascade neutron energy by the roof shielding, and the factor 0.5 x 10-7 is the isotropic geometry factor at 12 m distance from the radiation source at the graphite ring shield.Approximately 80% of the magnet windings are covered by magnet pole iron affording an attenuation factor < 0.1 for the cascade neutronenergy; thus the estimated power through the top shield is75 x 10-11 (1 - 0.8) = 15 x 10"11 W cm-2or a radiation field of approximately 75 mR h -1, again assuming afactor 2 for current-to-f1ux conversion.- 25 -At the inside of the concrete roof shield the cascade neutron current is approximately 5 x 105 cm-2 s-1 on the basis as outlined above for estimating the neutron leakage current through the roof shield. The residual activation produced in the concrete by this flux will lead to y-radiation fields ranging from 12 mR h_1 one hour after shutdown to 1.3 mR after one week over most of the vault at the beam line level. The fairly massive steel support structure above the machine will also intercept some of the cascade nucleons leaking between the magnet poles. Spallation activation in these components will probably contribute y-radiation fields near the top of the cyclotron that are comparable to and might be a few times higher than those from the concrete roof for decay periods of one hour. The radioactivity in the iron will decay by a factor of only "v 0.7 for the interval after shutdown of one hour to one week.Some of the radiation leaking out through the roof shield over the cyclotron vault will be scattered by the air and roof into nearby local control areas. To estimate the air scattering, the calculated kernels for an infinite air medium relating the flux to the directed 14 MeV neutron source1  ^ were used. For the 15 x 10-11 W cm 2 radiation leakage over the 8 x 106 cm2 effective roof area, the field intensity at the adjacent local control room floor would be approximately 0.5 x 10"11 W cm-2 or 1 mrem h _1 from air scattering without any roof. The building roof, of average thickness 'v 20 g cm-2, will scatter approximately 20% of the radiation incident on it. Assuming, conservatively, that this radiation is scattered isotropical1y and without attenuation by the building roof, the field intensity at the local control area floor adjacent to the vault shield would be ^ 1 x 10-11 W cm-2 or one working health tolerance.To reduce the radiation leaking through the vault roof and the internal residual activation fields coming from the vault roof after shutdown, it is recommended that provision be made for installing shielding in the vicinity of the magnet windings between the magnet- 26 -pole assemblies. This shielding should extend into at least the position of the 500 MeV orbit. It should be a fairly close fit tothe vacuum tank wall and be of sufficient thickness to provide thesame attenuation to the cascade nucleon energy as the minimum provided by the magnet pole assemblies.3.4 Operating and Residual Radiation Fields in the CyclotronThe dose rates in the vicinity of the graphite ring when the cyclotron is operated at full power can be estimated approximately from the data listed in Tables I and II. Most of the energy leaking out of the graphite ring shield has one of the following forms:1. Cascade protons2. Cascade neutrons3. Partially degraded evaporation neutronsThe evaporation proton kinetic energy will be mostly dissipated in the graphite ring and the y-radiation energy released by various reactions in the graphite will be small compared to the items listed above. Since the angle of incidence for the primary protons on the graphite ring is 'v 60°, average values for the two carbon cases shown in Table I will be used. Approximately 6% of the incident energy emerges from the plane slabs of carbon as cascade proton energy and approximately 8.5% as cascade neutron kinetic energy. The fraction emerging as cascade proton energy is very strongly dependent on the angle of incidence. At large angles of incidence a relatively small scattering angle is required to reflect the energy back into the 2 cF half-space of origin. This phenomenom makes the estimation of radiation fields at the inside of the graphite ring very difficult, and a substantial safety margin should be applied to the final results. For protons emerging from surfaces other than that on which they are incident, the data in Table I indi­cate that only about 1.5% of the incident energy emerges as cascade proton energy.- 27 -Assuming that 10% of incident energy leaks isotropica11y from the carbon ring in the form of cascade nucleon energy and that this energy is dissipated in the surrounding media with the relaxation lengths listed in Table II, then the energy absorption rate at the graphite-ring, magnet-winding interface in carbon can be estimatedfrom the following considerations: For a total power loss of10 kW by v stripping in the cyclotron, the power absorbed per cm in the 16.6 m diameter graphite ring is10 x  1 0 3 _  i Q w , .m- lr~7 7 T7vZ^  ' • 9 ™ •TT 16.6 X 10^Assuming, conservatively, that the 10% of the incident energy carried out of the 50 cm square graphite ring has an isotropic angular distribution, the energy current at the graphite surface is1 '? X r'rV'" = 0.96 mW cm"2 .1) x 50For an energy absorption mean free path of 120 g cm-2 the power dissipation is0.96 mW c n T j = 3 3 pW g-l.120 g cm-2By definition, 10 pW g_1 = 1 rad s-1 so the dose rate is 0.8 rad s"1 or 25 Mrad per full power year in carbon.The dose rate in most other materials will be comparable. The absorption coefficient for the cascade nucleon energy varies from 1/120 g"1 cm2 in carbon to 1/196 g_1 cm2 in lead. Only a small fraction of the primary proton energy ('v 0.5%) is converted into evaporation neutron kinetic energy in the graphite ring, and most of this will be dissipated in the graphite. If the primary protons were incident on a lead shield, approximately 5% of the energy would be converted to evaporation neutron kinetic energy and up to 15% of the primary proton energy released when these neutrons were captured. In that case the dose rate in carbon at the edge of the ring shield would be higher by a factor 2 due to increased radiation- 28 -leakage, and a factor 4 due to the decreased radiation energy relaxation length, compared to that quoted above for a homogeneous carbon shield, i.e. 200 Mrad per full power year. The dose rate in hydrogenous materials at the edge of a lead ring shield would be higher by an overall factor of approximately 12, compared to that for the carbon shield, due primarily to the shorter relaxation length for the evaporation neutron kinetic energy.The dose rate estimated above is not applicable to components penetrated by the primary protons. For these the dose rate is very much higher due to direct ionization by the primary protons. For example, the dose rate at the vacuum tank wall from ionization by the primary protons is ^ 1.3 x 10^ Mrad per full power year, assuming a uniform 20 yA beam spill over a 2 cm wide strip at the tank wall.The dose rate at the centre of the cyclotron can be estimated by considering the cascade nucleon energy reflected back from the graphite ring. Assuming, conservatively for points at the centre of the cyclotron, that the 'v 37 MeV per incident proton reflected back into the 'v 2FF space inside the graphite ring has an isotropic distri­bution, then the radiation current at the centre of the machine is3 .7  X 106 X ?.p , 1P-A ,  ,  7 „  w c m - 2 .2 tt 8 3 0 zFor a relaxation length of 120 g cm-2 appropriate to carbon or a similar light element, the dose rate is then1 ’ 7 1 2010~^ X 105 = ° ’ ] k  rad S_1 or 4.5 Mrad per full power year.For points off-centre in the machine but in the median plane, the geometry factor relative to that for the centre point for radiation originating at the circumference is- 29 -where R is the radius of the source (graphite ring) and r is the radius of the field point; thus, g(0.707 R)=2 and g(0.895R)=5» i-e- the radiation field ^ 2/3 of the way from the centre to the edge is double the radiation field at the centre. The radiation fields at points away from the median plane have a much more complicated position dependence because of the asymmetrical magnet shapes. Forpoints on the vertical axis of the machine but far removed from it,the field point views 'v 4/5 of the total circumference of thegraphite shield and tank through the thick magnet poles. Thus the effective source strength to such points is approximately 1/5 of the total.On the same basis as for estimating the cascade nucleon current at the inside of the main vault roof, the current at the floor in the service space below the cyclotron is approximately 0.7 x 106 cm-2 s-1 at the centre rising to 1.0 x 10® cm-2 s-  ^ directly below the graphite ring. The residual activation in the concrete floor would produce y-radiation fields of approximately 70 mR h-1 directly below the magnet windings after a decay period of one hour following a long irradiation. This field will decay to approximately 8 mR h_1 after one week.The 1-1/2" stainless steel rods used to anchor the vacuum tank wall to the floor will also contribute residual activity fields in this service space. Based on an average cascade nucleon flux of 106 cm-2 s-1 in this region, the y-radiation field from the 'v 300 rods will be approximately 45 mR h-1 after one hour decay falling to approximately 30 mR h 1 after one week. The estimate is for a site at the mid-point between four nearest neighbours at the centre of an 0.84 m spaced square lattice. The field is nearly uniform. No allowance was made for self-shielding. The self-shielding of the iron rods will not significantly affect the fields. Other massive components such as vacuum pumps and the centre support column will also be activated. Assuming such iron components to be thick in terms of mean free paths of the residual y-radiation,-  30  -the surface fields will be 30 mR h-1 after one hour, decaying to 20 mR h”1 after one week.On the same basis as above for the main vault region, approximately 0.1 neutrons per spilled proton will leak back into the service space with epithermal energy. The thermal neutrons leaking out of the graphite will all be captured in the immediate vicinity, either in boron placed there for that purpose or other structural components such as the stainless steel tank wall. The thermal flux in the 19 m diameter, 4.5 m high space, based on the recipe for concrete vaults, is thus1 .3 x 0.1 x 20 x 6.2 x 1012FFENk E 1900/2)2 + 450 x 1900)= 2 x 1 0 ®  cm-2 s"1.This is probably an overestimate because of the heavier absorption in the service space compared to that in the bare concrete vault for which the relation applies. This thermal flux level is a factor 4 greater than that estimated above for the main vault region. The various operating and residual radiation fields will also scale by a factor 4. Thus, for example, the equilibrium Ar1*1 concentration is 4 times the maximum permissible concentration for working occupation, and the saturation level Mn56 activity in a steel containing II Mn would be approximately 40 mR h-1.The above estimates of radiation fields in the service space assume that the graphite ring is unshielded to this area except by the magnet poles. If shielding is provided between the poles in the vicinity of the graphite ring, the radiation fields and residual activation can probably be reduced by a factor of several. It would also decrease the fields from the residual spallation activity in the stainless steel vacuum tank wall discussed below. Graphite would be the best shield material for this purpose.- 31 -The dominant residual activity in the cyclotron will be produced by spallation reactions of the high energy protons lost, by stripping, into the circumferential vacuum tank wall and graphite ring shield. Assuming that 20 pA of beam current is lost at 500 MeV, Table VI shows the residual y-radiation fields at the centre of the cyclotron after various decay periods following a very long irradiation. The effective angles of incidence on the various components were based on tangential loss from a 500 MeV orbit. The estimates for the lead shield indicate substantially higher fields than from the graphite. Other factors also militate against using lead for the shield between the magnet windings. The large fraction of energy going into evaporation neutron production could produce high operating dose rates in the magnet winding insulation,and the continuous decay of the lead activity, while an advantage after shutdown, indicates a fairly rapid build-up of the activity for short operating periods.For all decay periods the residual field is dominated by the component from the stainless steel vacuum tank wall. Further reduc­tions in the thickness of this wall would correspondingly decrease the residual activity fields.For carbon, the residual activity estimate is based on the experimental data14’15 for the production of C11 and Be7 , the only significant contributors to y-radiation fields. For the copper and stainless steel walls no allowance is made for self-shielding. For the graphite and lead ring-shield cases self-shielding has been taken into account.The residual radiation field at other points in the mid-plane of the cyclotron can be estimated from Table VI and the expression quoted above for g(r), the geometry factor for a ring source for non-central points relative to that for the centre of the ring. For other points, off the mid-plane of the machine, the radiation fields can be estimated by consideration of the 1ine-of-sight to the various sources listed in the table.-  32 -TABLE VIRESIDUAL RADIATION FIELDS AT CENTRE OF TRIUMF CYCLOTRONContributingPointRadiation Field after decay, mR/hour1 hour 1 day 1 week 1 month1.6 mm Cu resonator wall at 73° inc.25% of circumference 55 43 40 376.3 mm SS vacuum tank wal1 at 67° inc. 670 540 430 290Thi ck graphi te shield at 60° inc.380 60 54 40Thick lead shield at 60° inc. 1650 580 170 70-  33 -4. EXPERIMENTAL AREAS4.1 Primary Beam Tunnel4.1.1 Beam Transport SystemIf the beam transport system has a uniform spill rate of 1 nA m-1 of 500 MeV protons, the residual activity radiation fields will be 50 mR h"1 one hour after shutdown from a long operating period.The estimate is made for an unshielded field point one meter from a copper beam line that is bombarded at grazing incidence, i.e. all protons suffer a non-elastic nuclear collision in the beam tube,which is assumed to be thin to the residual activity radiation. The fields from copper will decay to 40 mR h-1 in one day and 30 mR h-1 in one week.This procedure overestimates the activity production by the primary protons in the beam tube by a factor 'v 2 because other mechanisms not leading to residual activity, such as elastic nuclear scattering and coulomb scattering, would remove some of the protons from the tube. At high nucleon energies the elastic and non­elastic nuclear cross-sections are comparable. The lost protons, and the secondary cascade nucleons from the non-elastic collisions, do, however, produce further activation outside the beam tube. Graphite or other low activation material should be used for this shielding to minimize the activation in components near the beam line. The detailed design of such shields will depend, of course, on the expected beam spill distribution.Assuming a uniform spill of 1 nA m-1 of 500 MeV protons from an infinite unshielded beam line, the energy current through a hypothetical 6 m radius cylindrical surface centred at the beam 1i ne wou1d be500 x 106 x 1 x 1 0_9/2FF 600 x 100 = 1 . 3 x 1 0-6 W cm-2.On the assumption that the radiation energy is emitted isotropically, which is conservative for field points behind a shield parallel to- 34 -the beam line, and taking from Table II the effective cascade neutron energy source of 90 MeV per 500 MeV proton incident on copper, the attenuation factor A required to achieve an energy flux of 10"11 W cm"2 at 6 m from the beam line is_ 10~n  . _90_ . J_A " 1.3 x 10“6 500 2- 2 x 10"5where a factor 2 is used for the current-to-flux conversion.The attenuation for a line source in the approximation of removal by one "effective" collision is given byA (x/A ) = ——  r A r K1 (x/Ar) - Seci (x/Ar)Seci (x/A^) for x/Ar >> 110xx3 _where Seci (x) = e"x sec6 d6 is shown in graphical form in Appendix 5 of reference (7)- A = 2 x 10-5 for x/Ar = 9-9; thus, the standard concrete shield thickness x to achieve dose rates of2.5 mrem h “1 is9-92V 26 = 520 cm = 17 feet-This is the shield thickness shown around the primary beam tunnel in Figures 8 and 9.For the assumed spill rate of 1 nA m"1 the evaporation neutron production in a copper beam tube is approximately6 .2  x i 1>Q9 x ,5  .  3 x 107 n cm- i  s- iwhere the factor of 5 evaporation neutrons per incident proton allows for the evaporation neutron production by secondary cascade neutrons. The peripheral surface area of the 11' x 13' primary beam tunnel is 1.5 x 103 cm2 per cm of tunnel length. Thus, according to the expression deduced above, the estimated thermal- 35 -neutron flux in such a tunnel is1.9 x 3 x 107/ 1.5 x 103 = 4 x 10*+ cm-2 s"1.This flux will not cause any serious activation problems.4.1.2 Meson TargetsThe major sources of radiation in the primary beam tunnel are the targets placed in the main beam for meson production.Approximately 2% of the primary beam is removed by non-elastic nuclear collisions and another 2% is removed by elastic nuclear collisions in a 2 g cm-2 carbon target. Thus, for a beam power of 50 kW, the total power removed from the beam by nuclear collisions in such a target is 2 kW. Only a small fraction of this energy,< 5%, is deposited in the target as heat and is less than 20% ofthe 0.5 kW energy deposited by ionization by the 100 pA beam. Mostof the 2 kW scattered by nuclear collisions is dissipated in the surrounding shielding and apparatus near the beam line.Assuming that the carbon targets are operated long enough to saturate the 54-day half-life Be7 activity, the equilibrium source strength is10 100 x 6.2 x 1012 r!X .02 X --- — —-------— Tfi  C l200 3 . 7  x 1010= 17 Ci of Be7where 10/200 is the fraction of non-elastic collisions producing Be7 nuclei in a 2 g cm-2 carbon target. The high energy secondary nucleons and pions will also suffer non-elastic nuclear collisions; the inter-nuclear cascade calculations for carbon indicate that approximately two non-elastic collisions will result in a thick carbon shield for each proton incident. Thus, the total equilibrium_ Asource strength of Be7 in a carbon shield surrounding the 2 g cm carbon targets will be3 x 17 = 51 Ci of Be7- 36 -where 1/3 arises from the secondary nucleons from the non-elastic collisions in the target and 2/3 from the protons removed from the beam by elastic nuclear collisions. The radiation field at 1 meter distance from the unshielded 17 Ci source of Be7 is0.5 R h"1. The equilibrium C 11 source strength in carbon targets is approximately 2.5 times that for the Be7 ; the field from the 20-minute half-life positron annihilation y-radiation is approximately 43 times that from the Be7 , i.e. 22 R h-1 at 1 meter. The total activity and radiation fields from 20 g cm-2 carbon targets will be a factor 10 higher, assuming that all of the beam not suffering a nuclear collision in the target can be cleanly removed to the beam dump.If copper targets are used for meson production, a variety of spallation products will accumulate in the target. For a six-month irradiation in a 100 pA beam the activity in a 2 g cm-2 copper target after a decay period of one hour would produce radiation fields of 24 R h-1 at 1 meter from the target. This source is equivalent to 45 Ci of 1 MeV y-radiation. The radiation field at 1 meter would decay to approximately 14 R h-1 after one day and 12 R h-1 after one week.The cooling water for the meson production targets will be activated by the 0 16(n,p)N16 reaction and by spallation reactions of the primary protons, assuming they pass through the coolant, and high energy secondary particles from the target. Assuming that both the primary protons and the secondary nucleons have an effective path length in the coolant water channel of 2 mm, the equilibrium activities produced in the water are shown in Table VII. The estimate is for a 2 g cm-2 carbon target in a 100 pA proton beam.-  37  -TABLE VI ICOOLANT WATER ACTIVITIES FOR 2 g cm"2 CARBON TARGETS IN THE PRIMARY PROTON BEAMEqui1. Activity R h"1 at 1 m ±-1 feN 16 .087 Ci .23 7.5 secN13 1.2 Ci 0.6 10 mi nC11 2.4 Ci 1.3 20.5 mi nBe7 2.2 Ci .067 54 daysThe third column shows the equilibrium radiation field at one meter from the entire cooling water supply making no allowance for self­shielding. The N16 estimate assumes the production of 1.3 neutrons in the carbon target by each non-elastic proton collision, all with energy greater than the 11 MeV threshold for the 0 16(n,p)N16 reaction, and a reaction cross-section of 30 mb per oxygen nucleus. The spallation reactions cross-section used are based on experimental data at higher proton energies but probably are accurate to a factor 2 for 500 MeV protons. The spallation activity is essential­ly that produced by a 100 pA beam passing through a 0.4 g cm-2 water slab. For a copper target the spallation activity from the oxygen in the cooling water will be unchanged but the N16 activity will be greater by up to a factor 'v 2 due to the increased secondary neutron mult i pii c i ty.To achieve operating fields of one health tolerance in the experi­mental areas la and lb and on local control room floor above the primary beam tunnel, additional attenuation by a factor of 3 x 10"3 is required for the radiation from a 2 g cm-2 meson production target. This can be achieved by 1 m of graphite and 35 cm of lead. Because of the asymmetry in the radiation current, the shield will be thicker in forward direction; the assumption of isotropy should be-  38  -conservative for directions perpendicular to the beam. Approxi­mately 50 cm of graphite should be used immediately surrounding target assemblies to reduce the residual activation. The 'v 35 cm of lead should be followed by an additional layer of graphite to thermalize evaporation neutrons produced in the lead.The cascade neutron flux at the outside of this graphite-1ead- graphite shield would be ^ 3 x 105 cm-2 s_1. The residual activa­tion at the surface of a thick iron slab 1 hour after shutdown from a long irradiation in such a flux would be approximately 10 mR h"1.The total cascade neutron leakage through this shield into the primary beam tunnel would be approximately 5 x 1010 s_1; while evaporation neutron production and leakage depends on the detailed design of the target assembly, a very conservative estimate can be made by scaling the cascade component by the ratio of the evaporation to cascade neutron multiplicities in lead, 'v 6. Thus a total epithermal leakage of 3 x 1011 s-1 will, on the basis of the experimental evidence quoted above,10 support a thermal neutron flux of 'v 5 x 105 cm-2 s"1 in the immediate vicinity of the targets.The residual activation fields at the surface of a concrete slab in such a flux is 'v 15 mR h-1 from the saturated Na2t+ activity, as noted above.The operating radiation field at outside of the graphite-lead-graphite shield will be approximately 250 rad h-1 based on a radiation energy current of 11 pW cm-2 and a relaxation length of 30 g cm-2.4.2 Experimental Area PThis area is divided into two sub-areas; one surrounded by a thick, removable-block, concrete shield and the other a relatively lightly shielded experimental area. The area shielded by 17 1 of concrete blocks can take proton beams of the same order as that removed by the meson production targets in the primary proton beam line. An approximate upper limit is probably 10 pA of 500 MeV protons. It is set by the necessity of dumping this beam locally rather than in a-  39 -separately located beam dump. To achieve conditions outside this beam dump comparable to those outside the 2 g cm-2 meson production targets in the primary beam line, an additional 16 cm of lead shielding is required to supplement the shielding described above for them. This estimate assumes 17' of concrete block shielding on all sides of the area.If the 5 kW of beam power were dumped into a large tank of water, the equilibrium activity content of the water would be approximately 50 times that listed in Table VII for N 16 and 500 times that for the other oxygen spallation activities. This can be taken as the upper limit on the cooling water activity. If copper plates are used in the tank to absorb most of the proton energy, the coolant activity will be reduced. The actual coolant activity will depend strongly on the detailed design of such a dump,but activity production of the same order as those quoted in Table VII can probably be achieved by minimizing the coolant volume in the region of the primary proton i nteractions.The main part of experimental area P has roof shielding of 3* of concrete. Proton beams of the order of 1 nA intensity can be accommodated in this area. If such a beam is incident on a thick, unshielded target, the radiation field at the floor above the area would be approximately 400 mR h-1. Thus, local shielding of approximately 9' of concrete will be required around the dump for such a beam to achieve a dose rate of 2.5 mR h-3 on the floor above.-  40  -5. NEUTRON IRRADIATION FACILITY5• 1 Neutron Source and FluxThe primary beam dump can be used as a neutron irradiation facility.The total neutron source strength for various targets can be esti-2 0mated from the data in the ING proposal. If a lead or bismuth target, 10 cm in diameter, were used to absorb a 100 yA beam of 500 MeV protons, the source strength would be100 x 6.2b x 1012 x 8 = 5-0 x 1015 n s_1.If this source were imbedded in a 2 .b m diameter tank of 020, the unperturbed thermal flux at a point 25 cm from the centre line and in the mid-plane of the target-moderator assembly [see Fig. V I 1.9 of reference (20)] would be5 * ] f 5 x 2 x 1016 = 1013 cm'2 s"1.1019For a graphite moderator surrounding the lead target the thermal neutron flux near the target will be approximately 1/2 that in a D20 moderator. This estimate is based on a two energy group, diffusion approximation for the thermal flux from a hollow sphericalshell source of fast neutrons surrounded by an infinite moderating2 1medium. If l_t  ^ and l_s are the diffusion lengths for the thermal and fast neutrons respectively and £t|_ is the thermal neutron transport mean free path, all for the moderating medium, the thermal flux at t*ie source radius R for a total source strength ofQ neutrons per second is given byA (R) = 30. _J------___  _!_ Ls Lrh»th' W  £tr Lth+Ls • Ls • R+Ls • R+Lth ‘Now £ = 2.6 cm for both D.O and graphite, L, = 100 cm for D„0 andtr 2 3 K ’ th 25b cm for graphite,and Lg = 13 cm for D20 and 20 cm for graphite, where the values for Ls are slightly larger than those normally quoted for a fission source because of the harder evaporation spectrum. Choosing R=10 cm as representative of the source, the- 41 -ratio of the thermal flux in the graphite to that in the D20 is0.55. The thermal flux will have a steeper radial dependence in the graphite than in the D20; to a first approximation the dependence will scale in proportion to the thermal diffusion length,1.e. a factor 2 shorter for graphite than for D20. Thus the volume over which the flux is developed is reduced by a factor 23 = 8.5.2 Heat ProductionTable VIII shows the partial energy distributions resulting from one 500 MeV proton incident on bismuth targets 10 cm and 20 cm diameter. The numbers in brackets are the total power distribution for a100 pA beam. The calculations were done using the same Monte Carlo2 2code as was used for the ING estimate. A 2 cm diameter, parallel proton beam was assumed incident on the axis of bO cm long bismuth targets for the two target diameters shown.The total heat production in the target is nearly constant as a function of target diameter. The energy of the escaping cascade nucleons decreases by about 22% from 49 to 38 MeV per incident proton. The number of evaporation neutrons escaping the target increases by 'v 1% from 8.3 to 8.9, but the extra kinetic energy of the evaporation neutrons is mostly dissipated by increased inelastic y-ray p roduct i o n .Figure 11 shows the evaporation neutron current from the sides of the 10 cm diameter bismuth target and the heat production per primary 500 MeV proton as a function of x, the distance from the front face of the target. The scatter in the points arises from the statistical inaccuracies in the Monte Carlo calculations. The solid point x=0 represents the total evaporation neutron leakage out the front face of the target re-normalized to the same unit area as a 1 cm circum­ferential strip. The heat production changes by less than a factor 3 at distances up to the range of the primary protons; beyond this distance the heat production drops by a factor 'v 40, illustrating the-  42 -TABLE VI I ISECONDARY ENERGY DISTRIBUTION FOR 500 MeV PROTONS ON 40 cm LONG BISMUTH TARGETTarget Diameter 10 cm 20 cmTotal Target Heat, MeV (kW) 370 (37.0) 375 (37.5)Escaping Case. Neutrons NumberEnergy, MeV (kW).66 48 (4.8).50 38 (3.8)Escaping Case. Protons NumberEnergy, MeV (kW).0181.7 (-17).005 0.3 (.03)Escaping Evap. Neutrons NumberEnergy, MeV (kW)8.3 28 (2.8)8.9 28 (2.8)Inelastic y-Radiation MeV (kW) 5.3 (.53) 8.9 (.89)-  43  -dominance of ionization as the important energy loss mechanism for 500 MeV protons.For a 100 pA, 500 MeV proton beam on a 10 cm diameter bismuth target at the centre of a 1.2 m diameter graphite moderator assembly, approximately 38 kW of heat must be removed from the target. This estimate includes 0.5 kW of inelastic y-ray power listed in Table VIII but does not include the y-radiation from thermal neutron capture in the target.For a 1.2 m diameter graphite assembly essentially all of the kinetic energy of the evaporation neutrons escaping the target will be released as heat in the graphite. Up to 501 of the thermalized evaporation neutrons may be captured in the graphite releasing 5-0 MeV per capture. Thus, assuming no net transport of y-radiation into or out of the graphite, the power production in the graphite arising from evaporation neutrons is 4.9 kW. Approximately 50% of the energy carried out of the bismuth target by the cascade nucleons will be deposited in the graphite. Thus the total heat production in the graphite will be 'v 7.5 kW.Assuming that the remaining 50% of the thermalized evaporation neutrons are captured in iron shields outside the graphite, with no y-ray leakage, the total heat production in the shields will be 'v 5 . 6 kW.5.3 Sh ield i ngAssuming that the 2.4 kW of cascade neutron energy emerging from the graphite assembly is distributed isotropically, the unshielded energy current at a point 5 m from the assembly would bex ]°3 = 0.76 x 1CT3 W cm"2 .i xx Q w w _Thus, to achieve an energy flux of 10-11 W cm-2 corresponding to- 44 -2 . 5  mrem h _1 , attenuation by a factor10"11 ---- — zr rt— 3" = 0 . 6 6 x 1 0 - 82 x 0 . 7 6 x 10 0is required, where the factor 2 is used for the current-to-f1 ux conversion. Using the relaxation lengths quoted in Table II, the estimated shielding requirements are thus 23 7 0 g cm - 2  of concrete or 2840 g cm - 2  of iron or 3690 g cm - 2  of lead to achieve the maximum permissible biological tolerance. The linear thicknesses are 10.0 m for concrete, 3-6 m for iron and 3-3 m for lead. If iron or lead is used, approximately 1 0% of the total shield volume must be water or a similar hydrogenous material to thermalize the evaporation neutrons produced in the shield. Thus the linear thicknesses of the composite iron-water or lead-water shields would be 10% greater than those quoted above.The Monte Carlo calculations indicate that approximately 6 x 10- 3cascade neutrons per incident proton, with a mean energy of 'v 25 MeV,emerge from the front face of the bismuth target into one steradian of solid angle in the direction opposite to the incident proton beam. For the same conditions and geometry approximately 0.36 evap­oration neutrons with mean energy 3*3 MeV emerge. Assuming a constant collimation of 10 cm diameter back to a point 3 m ahead of the target front face, the total neutron current out of this port, for a 100 pA proton beam, is 2 x 101 1 s_1 carrying a total power of 0.12 W. At the same 3 m point the neutron radiation power suffering grazing collisions with the 10 cm diameter beam pipe wall is4.6 x 10- 7  W cm-1. Thus, assuming that all of this power is absorbed in a 5 cm layer of water-equivalent hydrogenous material surrounding the beam pipe, the energy deposition rate is 4.5 erg s" 1 g " 1 or1.4 Mrad per full power year.5.4 Activity ProductionTable IX shows the g- and y-activities present in lead and bismuth targets due to spallation reactions at various times following the- 45 -TABLE IXNEUTRON FACILITY TARGET ACTIVITY Infinite Irradiation by 100 yA Proton BeamDecay T i meg-Activi ty kC iy-Act i vi ty kCi VMeVPb Bi Pb Bi Pb Bi1 hour 0 . 3 0 0.16 3.4 4.0 0.42 0.631 day 0.27 0.14 2 . 6 2 . 6 0 . 2 2 0.671 week 0.23 0 . 1 2 0.63 1.5 0.33 0.751 month 0 . 2 1 0 . 1 0 0.24 0.70 0 . 3 2 0.716 months 0 . 1 7 0 . 0 8 0.07 0.43 0.24 0.70- 46 -end of a very long irradiation by a 100 pA, 500 MeV proton beam. Barbier and Cooper's thin target data 1 1 were used with a factor 2.5 for inter-nuclear cascading within the target. The equivalent y-radiation energy is deduced from their y-radiation field strength tabulations using an equivalence factor 550 MeV cm - 2  s - 1  = 1 mrem h 1. For a lead target the activities listed in Table IX should represent essentially all of the induced activity. For a 10 cm diameter,40 cm long bismuth target, thermal neutron capture will produce a0.2 kCi equilibrium source of Bi210. This activity B-decays with a 5 day half-life to Po2 1 0 which decays with a 138 day half-life by a-emission. Thus, after full power operating periods of several months, a 0.2 kCi source of Po21 0 will exist in a bismuth target.The activity production in the cooling water for the thermal neutronfacility target depends on the details of the assembly. For ratherconservative assumptions, e.g. one neutron per incident proton with energy greater than 11 MeV traversing a 0.3 cm H20 cooling channel at an angle to give a path length in the water of 0 . 9 cm, the equilibrium, full power N 16 activity would be approximately 15 Ci. Assuming a total water flow rate of 1 & s--*-, the unshielded radiation field from the N 16 activity one meter from a long, 2 cm diameter coolant line at a point 10 sec down stream from the target would be100 rem h " 1. The C 1 1 and Be7 activities in the cooling water aremore dependent on the target design than the N16; they are unlikely to be more than 20 to 30% of N 16 activity even under the worst conditions. The radiation level at the boundary between the bulk concrete shielding surrounding the thermal neutron facility and earth fill can be much higher than biological tolerance. The limit is set by the activation in the soil. Assuming a specific activity production in soil of 0 . 0 5 yCi kg  ^ as was estimated above for the outside walls of the cyclotron, the shield thickness required between the thermal neutron facility and the soil is 5-5 m of concrete.-  47  -6 . DISCUSSION OF ESTIMATESShielding requirements, where they have been explicitly estimated, tend to be conservative. This situation has occurred for three reasons: ignorance of how to make the radiation transport andactivity production estimates accurately; ignorance of the level of radiation power loss; and the relatively small fraction of the cost for shielding low intensity facilities. The last two reasons do not apply to TRIUMF.The radiation power leaking through thick shields generally decreases exponentially with shield thickness. Thus relatively modest errors in the radiation relaxation length can produce large errors in the radiation leakage, which in turn can be reduced by relatively modest increases in the shield thickness. These considerations indicate a need for conservatism wherever the incremental shielding costs are relatively small, e.g. bulk shielding, requiring form-work, that has modest shielding support costs. Wherever possible, the final demountable shielding requirements should be estimated from radiation intensity measurements made at low power after the cyclotron is completed.Several of the parameters used in estimating the radiation energy transport are probably conservative, but not excessively so. The choice of 10- 1 1  W cm - 2  as the radiation power level corresponding to2 . 5  mrem h - 1  is reasonable in view of the past accelerator experience and the lack of explicit consideration of the evaporation neutron transport and therma1ization in making the estimates. The cascade neutron relaxation lengths used were those estimated from the calculations for protons incident on the shields near normal direc­tions. The results shown in Figure 5 indicate that for the situation where the field points are at 'v 90° to the proton beam direction, the relaxation length is smaller by 6% and, from Figure 7, the effective cascade neutron source energy less by a factor 3- The roof shielding required over the cyclotron vault would decrease by 30 cm if the change in relaxation length were applied and by 60 cm for the change in source strength. These would appear to be reasonable safety marg i ns.-  48  -The accuracy of the residual radiation field estimates is more difficult to assess because they usually involve radiation transport as well as activity production. The basic activation data are probably accurate to a factor 2 except for those cases discussed explicitly above; thus the overall accuracy of the residual radiationfield estimates inside the cyclotron is probably a factor 3-Residual field estimates for the main cyclotron vault and the service space below the cyclotron are probably less accurate, say a factor 5- Because assumptions believed to be conservative were used, where necessary, in making these estimates, they are more likely to be high than low.The choice of 2.5 mrem h -1 as the acceptable operating radiation field strength in inhabited areas is more difficult to defend. It probably should be used only for areas where it can be demonstrated that occupation averaged over one week will never be required for more than 20% of the time, or 8 hours. Where this cannot be shown, aworking level of 0.5 mrem h -1 is recommended.- 49 -I wish to thank Mr. J.J. Burgerjon and Drs. B.D. Pate and J.B. Warren of TRIUMF and Drs. D.G. Hurst, A.M. Aikin,C.H. Millar and M.F. Duret of CRNL for the opportunity to participate in a mutually useful collaboration between Chalk River Nuclear Laboratories and a Canadian university research group.I should like to thank Dr. S.A. Kushneriuk for his guidance in making the thermal flux estimate in the graphite moderator of the neutron irradiation facility and Dr. J.S. Fraser for the Monte Carlo calculations of the neutron production and leakage from a bismuth target.I should also like to express my appreciation to Miss P. Menzies at CRNL and Miss A. Strathdee of TRIUMF at UBC for their patient typing and retyping of the various drafts of this report.ACKNOWLEDGEMENTS-  50  -REFERENCES1. TRIUMF Proposal and Cost Estimate, edited by E.W. Vogt andJ.J. Burgerjon, November 19662. Proceedings of the USAEC First Symposium on AcceleratorRadiation Dosimetry and Experience, BNL, November 3"5, 1965; papers by H. Wade Patterson and others3. H.W. Bertini, Phys. Rev. 131, 1801 (1963), Phys. Rev. 138,AB2 (1965) and 0RNL-33834. R.W. Peele, T.A. Love, N.W. Hill and R.T. Santoro, Phys. Rev. 167, 981 (1968)5. R.G. Alsmiller, M. Leimdorfer and J. Barish, 0RNL-4046 19676. Reactor Physics Constants, ANL-58OO, Second Edition (19 6 3)7. W.J. Henderson and A.C. Whittier, Handbook of Shielding and Heat Production Calculations for the NRU Reactor, AECL-403 (1955)8. B.S. Sychev et al., Passage of High Energy Neutrons in I ron- Aqueous Mixtures, JINR-P-2359 (1965)9. H.W. Patterson and R. Wallace, UCRL-8359 (1958)10. H. De Staebler and T. Jenkins, SLAC-TN-65-24 (1965)11. M. Barbier and A. Cooper, Estimates of Induced Radioactivity in Accelerators, CERN 65-34 (1965)12. M. Barbier, Radioactivity Induced in Building Materials, CERN-67-25 (1967)13. G.Z. Rudstam, f .Naturforsch, 21a, 1027 (1966)14. Philippe Tardy-Joybert, Thesis, 1966; also available as CEA-R-2975, Saclay Nuclear Research Centre (1966) and in translation as ORNL-tr-1368 (1967)15. E. Bruninx, High Energy Nuclear Cross-Sections, CERN reports CERN-61-1 , CERN-62-9 and CERN-64-1716. P.A. Benioff, Phys. Rev. 119, 316 (I960)17. Neutron Cross-Section Compilation, BNL-325, Vols. I — I I I, Second Edition, Supp. #2 (1964-1965)18. ICRP Publication #2, 1959, Pergamon Press-  51 -19- Reactor Handbook, Vol. II, Part B, Second Edition, E.P.Blizard, Editor, p. 246, Interscience Publishers (1962)20. The AECL Study for an Intense Neutron Generator (Technical Details), AECL-2600, edited by G.A. Bartholomew andP.R. Tunnicliffe (1966)21. Private Communication, S.A. Kushneriuk, CRNL, 196822. Private Communication, J.S. Fraser, CRNL, 196823. R.G.P. Voss and R. Wilson, Proc. Roy. Soc. (London), A 236, 41 (1956)24. W.P. Ball, Nuclear Scattering of 300 MeV Neutrons,UCRL-1938 (1952)25. M.H. MacGregor et al., Phys. Rev. Ill, 1155 (1958)26. T. Coor et al., Phys. Rev. 98_, 1369 (1955)27. F.F. Chen et al., Phys. Rev. 99, 857 (1955)28. J.R. Oakey, G.E. Crippen & Associates Ltd., Building Concept with Cost Estimates, Site Exploration and Engineering Data for a 500 MeV H" Cyclotron Facility, TRI-68-2 (1968)PHOTON FLUX cm'10 10 10 10 10ENERGY MeVPHOTONS (L.H.S.)NEUTRONS(R.H.S.)PRO T ONS(R.H.S.)FIGURE 1The photon, proton and neutron flux to deliver a biological dose of 2.5 mrem h_1 is shown as a function of particle energy.NUCLEON FLUX cmsENERGY FLUX MeVFIGURE 2The energy flux of radiation in the form of photons, protons and neutrons to deliver a biological dose of 2.5 mrem h_1 is shown as a function of particle energy. The dotted line for neutrons represents the kinetic energy of the neutrons plus the 2.5 MeV bind­ing energy released as y-radiation when they are captured by protons.ENERGY FLUX WcmA.E.C.L. Ref. # A- 3291-DENERGY -  MeVFIGURE 3The non-elastic cross-sections for C, A 1, Cu and Pb as a function of nucleon energy. The solid data points are for neutrons from references 2 3 , 2b, 25 and 26, and the open points are for protons from reference 2 7 .t- g cm'2FIGURE 4aFigure 4a shows the estimated energy of the cascade neutrons gnd protons of ail orders suffering non-elastic nuclear collisions in a one g cm-2 layer as a function of position in a 1000 g cm-2 thick infinite slab of carbon when one 500 MeV proton is incident on one side in the angular interval 0-30° to the slab normal.Figure 4b shows the same dependence for a lead slab.MeV cmt - g cm'2FIGURE kb5 0 0  0 1000t g cm_zFIGURE 5The estimated cascade neutron energy suffering non-elastic nuclear collisions of all orders are shown as a function of position in a 1000 g cm"2 slab of aluminum when one proton is incident on one face of the infinite slab in the various angular intervals shown.En MeVFIGURE 6The estimated spectra are shown of cascade neutrons leaking through 1000 g cm-2 thick infinite slabs of carbon, aluminum, copper and lead when one 500 MeV proton is incident in the angular interval 0-3 0° to the slab normal.FIGURE 7The estimated angular distribution of the secondary cascade neutron energy is shown for carbon, aluminum, copper and lead. The points show the effective cascade neutron energy source strength for cascade neutrons leaking through a 1000 g cm-2 infinite slab of aluminum when one 500 MeV proton is incident in the various angular intervals relative to the slab normal.LEGENDJI ST STAGE C ONCRETEACCELERATOR SUPPORT CONCRETE, NOT INCLUDED IN FACILITIES ESTIMATEITEM I CAST-IN-PLACE SEMI - P E R M A N E N T  SHIELDINGITEM IA ACCELERATOR VAULT R O O F  B E A M SnITEM 2 T U N N E L  SHIELDING BLOCKSITEM 3 REMOVABLE STRUCTURAL B E A M S  AND C O L U M N SITEM 4 REMOVABLE SHIELDING BLOCKS AND LOCAL CONTROL R O O M  FLOORN O T E S/. F o r  r c P o r c r c c  c / r - o y v i n g s  D w g  < ?.2. For forts Dwg 740Scats in Feet50ifIua■zMADE5aMl * n N -T R I -U N I V E R S I T Y  M E S O N  F A C I L I T YFIGURE 8Plan View of TRIUMF with Shielding□. E. CRIPPEN &  ASSO CIATES LTD . - NORTH VA N C O U VER , B.C.3ECT/ON F-FLEGENDElevator <5 Fan room housingEt. 287.0EL 274.0xtzEL 259.0-77 oP F B *osr> Control ShoPt-~ Original ground surface~ Future serviceson +unnal■it.'.*-'-.'2:Service tunnel-3ECT/ON G-GI ST STAGE CONCRETE-Primary Beam TunnelCable trays <5 pipes - ^ ■Service funnelSECTION H -H SE CT I ON d - dACCELERATOR SUPPORT CONCRETE. NOT INCLUDED IN FACILITIES ESTIMATEITEM I CAST-IN - PLACE SEMI - P E R M A N E N T  SHIELDINGITEM IA ACCELERATOR VAULT R O O F  B E A M SITEM 2 T U N N E L  SHIELDING BLOCKSITEM 3 REMOVABLE STRUCTURAL B E A M S  A N D  C O L U M N SITEM 4  REMOVABLE SHIELOING BLOCKS AND LOCAL CONTROL R O O M  FLOORNOTES1. For re fe re n c e  draw ings s e e  Dvjg S.2. For location o f  se c tio n s  s e e  Dv/gs S & 6.SO  to o £Scale in Feet-Low Level Chem LobsStee / $h ieldingSECTION K -KSiK E r  PE AN5P*5SI MADEIn N - aT R I - U N I V E R S I T Y  M E S O N  F A C I L I T YFIGURE 9 Elevations with Shielding0 . E. CRIPPEN *  A S SO C IATE S LTD . -  N O RTH VANC O U VER * B.C.GRAPH ITE  R ING  SH IELDMAIN M AGNET  CO ILUPPER  YOKE UPPER  PO LETANKVERT ICA LLOWER YOKE LOWER POLE/V / / / / / / / /YOKEEND WALL(f^  COPPER)FIGURE 10A vertical section of the TRIUMF cyclotron is shown. The inset shows the geometry assumed in making the estimates of the residual activation radiation fields.FIGURE 11The evaporation neutron leakage from and the heat production in a 10 cm diameter, bO cm long bismuth target is shown for 500 MeV protons incident in a 2 cm diameter beam. The scatter in the points is from statistical inaccuracies in the Monte Carlo calculations.The solid point at x=0 on the neutron leakage curve shows the leak­age out the end of the target renormalized to the same area as a 1 cm wide circumferential strip. Nearly all of the heat production occurs within the range in bismuth of 500 MeV protons, 'v 23 cm.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items