NEACRP-L-;tg+

.: .'

International "State of the Art" .- Seminar on "Nuclear Data, Cross Section Libraries and their Application in Nuclear Technology"

Bonn, 1 - 2 October 1985

NEW INTENSE NEUTRON SOURCES AND RELATED NUCLEAR DATA NEEDS

S. Cierjacks

Kernforschungszentru Karlsruhe : Institut fiir Xernphysik

Postfach 3640, 7500 Karlsruhe Federal Republic of Germany

Abstract: Recent developnrents in neutron source design were largely directed towards new applications in radiotherapy, fusion technology and solid state physics. In these fields 14 MeV T(d,n).,'deuteron break-up and spallation neutron sources appear to be the most promising types for providing high source strengths. The main properties of such sources are briefly described with regard to total neutron intensities and angular-dependent spectra. Nuclear data related to the design, operation and application of the sources are discussed in four main areas of interest: Determination of source properties, source and accelerator shielding, safety and maintenance, and radiation damage and dosimetry. The nuclear data aspects including data needs, and availability of experimental data or model predictions are then discussed separately in some detail.

. ‘. ’

-2- .( ‘

1. Introduction

In recent years there has been a growing interest in new intense neutron sources for advanced fields of basic and applied research. Whilesource development overmore than three decades was largely governed by needs in basic nuclear physics and in neutron data measurements for fission reactors,new sources are now largely re- quired for application in radiotherapy, fusion technology and solid state physics. In radiothe,rapy radiobiological studies have led to the hope that neutron'therapy may contribute to some malignant cancer diseases which do not resgond to conventional X-ray trea*Jnent. Neutron sources for radiot.herapy must produce high energy neutrons at high intensity. The high energy is required to provide sufficient penetrations, and the high intensity to provide short treatment times. rad/min (2 1012 n-s-f),

The required intensity is .-, 25 and the average energy should be above

10 MeV. In fusion technoloqy the investigation of radiation damage

a is a severe Problem. Thefirst wall of a DT fusion reactor is subjected to an.exposure of 14 MeV neutrons with a flux of the order of 1Ol4 n-cm-2-s-l. These 74 MeV neutrons have a much higher energy than fission neutrons and are expected to produce more severe radiation damage not only because of their higher energy but also because gas production and elemental transmutation cross sections are much higher at these energies. The i?vestigation of the radiation damage due to gas production (He embrittlement) and elemental transmutations are likely to be cumulative and require fluences of the order of more than 1022 n-cm-2 delivered over volumes of several cm3 within a few months. In solid state physics a large number of investigations involve slow neutron scattering techniques, and many of these are presently carried out at high-flux fission ,reactors. But user-oriented high-flux reactors are oversubscribed by suitably screened experi- ments, so that there is an increasing demand for additional slow neutron sources. Furthermore, the use of time-of-flight instruments and double-or triple-axis spectrometers in this field has also.led

l to an increasing desire for higher peak fluxes of thermal and epi- thermal neutrons. Peak fluxes of Tsre than 1015 n-cmm2-s-' in the thermal range and of more than IO n.cm-2.s-l in the epithermal region would be highly desirable. These fluxes cannot be achieved from steady-state fission reactors. Especially high-flux reactors are presently run at the upper technical limit which is determined by the overall heat deposition released in the fission Process.

2. Neutron Source Reactions

The main types of nuclear reactions which can be used for neutron production are listed in Table I /I/. Also given are the numbers of neutrons per particle or reaction and the heat deposition per source neutron which both represent a measure for the utmost attainable neutron source strengths. In order to judge the suita- bility for application in the intended energy range, the type of source spectrum is also characterized in the last column. From the

‘. .

L

Table I. Neutron Yields and Deposited Heat for Some Neutron Producing Reactions

Reaction

T(d,n) (0.2 MeV)

W(e,n) (100 MeV) Li(d,n) (40 MeV) Fission (235U(n,f) (T,d) fusion Pb spallation (1 GeV) 238~ spallation (1 GeV)

) I - Yield

n/particle or event

8 x lo-Ii n/d 3.2 X low2 n/e

6.7 x 10e2 n/d * 1 n/fission+ * 1 n/fusion

20 n/P

40 n/P

Deposited Reat MeV/n

Type of Spectrum

2500 14 EleV 3100 evaporation

600 d-break-up

200 fission 3 14 MeV

23 evap.+ cascade 50 evap.+ cascade

+ The yield per fission event is 2.4 but % 1.4 neutrons are required to maintair the reaction and compensate for parasitic losses.

I x -4-

data in the table it is obvious that sources based on the fission process or on (e,n)-reactions are poorly suitable for radiotherapy or fusion materials damage studies, because fission or evapo- ration spectra do not provide sufficient intensity of neutrons in the required 'energy range. In radiotherapy the use of the (d,n)-reactions (deuteron break-up) on light nuclei and of the

T(d,n) (0.2 MeV) reaction appearsto be the best choices. For radiation damage studies the most desirable source is a 14 MeV DT source based on either the T(d,n)- or the DT-fusion reaction, respectively. The T(d,n) reaction provides, however, the smallest number of neutrons per particle and causes a very high heat de- position per source neutron in the target. Thus, it cannot provide the required neutron flux densities to study bulk damage effects. Large plasma sources based on the DT fusion reaction cannot be realised except for operating fusion reactors themselves. Neutron sources which produce broad neutron spectra with an average neutron energy near 14 MeV are therefore considered as a reasonable compro- mise. Such spectra are obtained from (d,n)-reactions in the forward direction employing 30-40 MeV deuteron beams. Deuteron break-up in the Coulomb or nuclear field of the target nucleus removes the proton from the projectile and leaves the neutron with an energy of about half that of the incident deuteron. The use of spallation reactions and heavy element targets is most promising to fullfill future needs in solid state physics, Spallation reactions provide by far the highest number of neutrons per proton (up to 40 n/p),and cause except for the DT fusion reaction, the smallest heat,deposition per neutron. Spallation ,reactions, however, produce also broad neutron spectra with contributions from intranuclear cascade reactions (with neutron energies up to those of the incident protons) and from evaporation processes of highly excited comoound nuclei. Therefore, neutron moderation is needed in conjunction with the primary spallation source. Unmoderated spallation sources are also reasonable tools ,for radiation damage studies of fusion materials /i/.

-* Based on the above considerations, recent development of intense neutron sources has concentrated on the following three types: 14 MeV T(d,n) sources, deuteron break-up sources and spallation neutron sources.

3. Nuclear Data Related to New Intense Neutron Sources

The nuclear data involved in the design and operation of the new intense sources can be grouped mainly into four areas of interest: a. Predictions or Determinations of Source erouerties

The source characteristics are mainly governed by integral and differential neutron yields and spectra. For two types of sources (Li(d,n)and spallation sources) yields and spectra are strongly angular dependent. For this reason the most important quantities

. aL

’ I

b.

l c. Safetv and Maintenance:

8.

-5-

are double, differential neutron production cross sections for protons or deuterons depending on the accelerator type. These differential data should be known for various possible candidates of target materials.

Source and Accelerator Shielding: Shielding of the new sources is complicated not only because of the increased neutron intensities, but also because of the increased energies of source neutrons at which the re- levant neutron cross sections are largely decreased. Shielding calculations typically require energy-dependent total neutron, elastic and non-elastic scattering cross sections for the important shielding materials. For the partial cross sections also detailed angular distributions are needed for energies above a few MeV. Finally, energy dependent y-ray production cross sections must be known for shielding, target and accelerator materials.

At the high intensities and the high energies provided by the new sources not only the target but also experimental equipment can be highly activated. Since beam losses during acceleration cannot be completely avoided, high current proton and deuteron accelerators also accumulate a considerable. activity in the accelerator structure. This.'can cause severe limitations for short-term maintenanc:e and repairment work, To make proper selection of materials,proton-. or deuteron- and neutron-induced activation cross sections are needed for a large variety of technically important materials.

Fusion Dosimetrv and Radiation Damage : Dosimetry plays an important role in damage studies, since it allows to accurately specify the neut.ron exposure of the test samples. Using a large number of activation reactions with thresholds, suitably covering the entire spectral range, allows spectral analyses by the unfolding technique, if the differential cross sections of these reactions are properly known. In addition damage cross secticns allow materials scientists to correlate damage measured in one type of facility with that in another type. The most important (although not only) informations needed in this context are spatial dependent damage parameters such as damage energy, recoil energy displacement, gas production and element trans- mutation,i.e. energy- and angular-dependent displacement, gas production and transmutation cross sections.

Depending on the type of source all mentioned nuclear cross sections should be known over a wide range of energies. This extends from thermal to -50 MeV for deuteron break-up sources.~and from thermal to more than 1 GeV for spallation sources. Evaluated nuclear data from existing libraries (e.g. available for energies below -

ENDF/B-V)/3/ are presently only 20 MeV. But the data above - 3 MeV are

.’ . ’

’ >

(I,

-6-

usually of lower quality than those below, and data above 15 MeV are often only poorly known.

In the following sections the three types of new sources, i.e. 14 MeV sources, deuteron break-up sources and spallation sources are discussed. The scope of individual discussions includes a brief description of the main source properties followed by some examples of presently operating, constructed or planned facilities. Finally, nuclear data aspects related to the four areas of interest defined above are discu,ssed in some detail.

4. 14 MeV T(d,n) Sources

4.1 Design Considerations

Among the neutron producing reactions involving hydrogen isotopes the T(d,n) reaction has by far the largest cross section; it reaches5b at a deuteron bombarding energy of 105 keV. This large cross section at's low.bombarding energy has made this reaction the most widely used reaction for applications in which. a compact intense neutron source is needed. With only a few ex- ceptions deuterons of energy between.100 and 400 keV are stopped in a metalic target in which tritium has been absorbed. The method limits, at present, the maximum neutron source intensenty to - a few times 1013 n/s due to the large inherent heat deposition. connected with this kind of neutron production(see Table I). The use of 14 T(d,n) sources is simplified by the nearly monoenergetic and isotropic neutron field, To measure fluences over short irradiation times of up to a few weeks the 93Nb(n,2n)92mNb reaction has been routinely used,.while for longer irraditions the 54Fe(n,p) 54Mn reaction represented a suitable choice/4/. Particular care must be taken to adequately measure flux gradients in the small experimental volume near the source. This has led to extensive flux mappings /5/ on a very small scale at the RTNS-source (see next section).

4.2 Examnles of 14 MeV T(d,n) Sources

A few examples of presently operating 14 MeV T(d,n) sources and their specifications are given in Table II. Special applications of these sources in activation analysis, radiotherapy and in radiation damage studies are also mentioned. Commercially produced neutron sources typically provide source stren ths of the order of 1012n-s-'. The highest intensity of 2.3 -10 13 could only Abe achieved with the Rotating Target Neutron Source RTNS II at LLNL. /7/. From recent development work at KfX there is alS0 Some hope for achieving a similar intensity with the KfK/H;ifeli sealed neutron source, KARIN, if possible improvements are made /a/. But there is presently little prospect for further improved inten- sity'from 14 MeV T(d,n), sources involving tritium or deuteron / tritium loaded metal targets. Previous studies to produce larger 14 MeV neutron intensities by employing colliding supersonic d- and t-gas jets /9/ have been terminated some time ago, mainly because of the extreme technical problems connected with the design

l l

Table II. Examples of Operating 14 MeV T(d,n) Sources a)

Manufacturer Voltage Current or Laboratory (kV) (mA)

Texas Nuclear

Radiation Dynamic

Philips

Cyclotron Corp.

KARIN, KfK, Haefely

RTNS II, Livermor

200

500

250

175

200

400

I a) From Ref. 6 b) Ref. 8 c) Ref. 7

7

12

10

450

500

00

Source Strength

Neutrons/set

6 'X 10"

3 5 x 10'2 . .

lOI

0 x 10'2

5 x 10'2

(1 x 10'3) b)

2.3 y lOI c)

Special Application

Therapy

Therapy

Therapy

,Therapy and Activation Analysis

Radiation Damage, 14 EleV Cross Sections

’ . : . -a-

D+-Bea

Fig-l: Schematic drawing of the rotating ~target assembly for the high intensity 14-&V T(d,n) source RTNS II /lo/. The shaded elements represent the portion of the system that rotates; the unshaded parts are stationary.

.'

of sucii sources. In order to demonstrate the high sophistication already involved in the target design of tritium metal targets, Fig.1 shows schematically the target assembly used with the RTNS II source /lo/. The neutron producing target is a 50 cm diam. spherially sha?ed water cooled titanian tritide target, which is rotated on a speed of 5000 mm. In addition, the entiretarget assembly is rocked around a pivot ~point, to further distribute, on a time average, the beam power over the whole available radial range.

4.3 Nuclear Data Aspects

The 14 MeV T(d,n) reaction is well understood at the relevant energies and standard fluxes are obtained to an accuracy of -1% /11/. In addition there appear to be nc inherent problems to :the provision of moderated 74 MeV fields other than engineering development and validation /12/. Neutron shielding characteristics can be satisfactorily predicted with existing data libraries established for fission and fusion reactor applications. There have been several calculations of shield a:nd collimator properties, especially for sources applied in radiotherapy,which showed gooc agreement with experimental dose rate measurements, e.g. /13/. Also 14 MeV activation cross sections for various source and shielding materials are sufficiently known and are contained in evaluated files.

.’

: . -9-

5. Deuteron Break-up Sources

5.1 Desiun Considerations

Deuteron break-up cross sections in the forward direction increase with increasing target mass number proportional to -AU3/14/. Nevertheless, forward neutron yields have been found to increase with decreasing mass number of the target nucleus /15/. This is explained by a more rapid increase of the number of target nuclei within the deuteron range. Furthermore, forward angle neutron spectra peak around Ed/2 with half widths of the order of 10 MeV. High intensity deuteron ~break-up sources thus must involve light target elements such as Li or Be. In order to provide neutrons at an average energy of-14 MeV,.deuteron acceleration to 30-40 MeV is sufficient. Typical angular-dependen t neutron spectra measured for 40-MeVdeuteronsbombarding a thick Li target /16/ are shown in Fig.2. The forward neutron yields peak around 14 MeV and exhibit

0 broad continuous distributions with energies extending to more than 50 MeV. Individual neutron spectra change sapidly with emission angle, and the high energy contributions decrease drastically for larger angles.

5.2 ExsmDles of Deuteron Break-uo Sources

Some deuteron break-up sources presently operating or under construction where described in Ref.17.The operating facilities -all represent medium energy cyclo&aons constructed~ originally for low energy nuclear physics research. These typically provide deuterons atan enerqv of - 50 MeV and averaue beam currents of several hundred ~A(UD to 1 mPJ.. /17/. The corresponding source strengths are of the order of 1074 n&s-l, and are not sufficient for large-scale studies of bulk damage effects. For ~the latter

Fig.2: Energy distributions of neutrons emitted at various angles. The data refer to an incident beam of 40 MeV deuterons on a thick Li target /16/.

. ,

: .

-lO-

l

Fig.3: Schematic drawing of the windowless liquid lithium target for use with the FMIT source. The target concept is suitable for operation with 3~ to 10 MW beam power.

purpose a new high intensity source, the Fusion Materials Irradi- ation Test (FMIT) facility /18/, is under c~onstruction in the United States.* of

This facility is designed to produce a source strength 4.1076 n-s-1, and fast neutron fluxes. of up to 1015 n.cm .s-1

over a volume of '- 1 1 /79/. In this facility a 100 mA, 40-MeV deuteron beam will bombard a thick Li target. Detailed ,studies of other target concepts have shown, that solid Be ,targets cannot withstand the heat deposition from the accelerator beam, even if a rapidly rotating target were involved. The design objectives of the FMIT target /2/ are illustrated in Fig.3. The concept involvesa windowless liquid Li target made up by a fast flowing Li jet which moves at a velocity of 15 m-s-l.

5.3 Nuclear Data Aspects

5.3.1 Source prooerties

Yields for 30-40 MeV deuteron bombardment of thick Be- and Li- targets have been measured with high accuracy in recent years (see'bibliography /21/). An example has been shown in Fig.2, and most of the results are available from equivalent data libraries e.g. /22/. There is no urgent need for more accurate data,except perhaps for the low energy ends of the spectra, say below - 2 MeV /.17/, where simultaneous measurements with high energy neutron techniques are more difficult.

9 Unfortunately, funding of this facility Ihas been largely reduced, so that the previous target goal of late 1985 cannot be met. Nevertheless, any final fusion material selection with respect. to bulk damage effects must involveaneutron. source of this kind.

. * --,I-

5.3.2 Source and accelerator shielding:

l

Nuclear data for source and accelerator shielding below -15 MeV are largely available from existing shielding libraries. In the range above 15 MeV total cross section measurements for various shielding materials have been made at KfK ,/23/, ORNL /24/ DC Davis/25/ and at LANL /26/. In addition,, nonelastic and removal cross sections at 40 and 50 MeV have been determined at Davis j25/. What is most needed presently are suitable evaluations of existing shielding data in the high energy range. For this purpose,experimental data must be complemented by results from model calculations in the missing mass and energy regions.

5.3.3 Safetv and maintenance

The imoortant activation reactions involved in the design and operation of the FMIT facility have been discussed in several surveys and contributions to.the 1977 and '1980 Symposia on IO to 50 MeV Neutron.Cross Sections /27,28/, Apart from neutron activation reactions in standard shielding and target materials, safety and maintenance considerations must include neutron-in- duced radioisotope production in accelerator and test cell materials as well as deuteron-induced activation in materials directly exposed to the accelerator beam. A large number of activation reactions of the first type are given in a summary of Carter et.al. /29/, while Johnson et al.. /30/ also,discuss in detail deuteron activation in beam transport system materials. Both surveys list more than fifty important activation reactions for which experimental data are lacking in the range from a few MeV up to 50 MeV. Since current experimental presuppositions do not allow to provide all the required data in due time, ,it is necessary to use model calculations to sup'ply most of the missing data.

5.3.4 Radiation damage and fusion dosimetly

Nucleardata required for the definition of the irradiation environment and the associated damage parameters have been discussed in great detail by Doran and Guinan /31/. Experimental data above 10 MeV centre almost completely around 14 MeV: Total helium production cross sections for various technically important materials were measured y Neff et al. /32/,while Qaim et al./33/ measured~ systematically 13 He-and p-production cross sections. Finally,Haight et al. /34/ measured a large number of proton- and .4He-production cross sections-Apart froln these data at 14 MeV, gas production and element transmutation cross section in the range from lo-50 MeV are -almost completely missing. Threshold activation cross sections desired for passive fusion dosimetry have been ex- tensively discussed by Gold et al. /35/. Adopting.the multiple- foil technique,used successfully in fission reactor dosimetry, up to fifty reactions were considered (24 with high priority) for adequately covering the entire'energy range of the irradiating neutron fields. In addition to (n,f),(n,p), (n,np) and (n,a) thres- hold reactions also (n,xn) reactions with x up to 5 must be in-

-v--

volved t0 provide thresholds in the 10 to 5CMcV range. As for the main activation reactions discussed in the Iprevious subsection, most of the data needs for radiation damage and fusion dosimetry purposes must come from suitable model calculations. The most important models applicable to nuclear data problems in the 10 - 50 MeV range are optical, multistep Hauser Feshbach, direct reaction and intranuclear cascade-evaporation models. Existing model codes of this type have been described in a workshop summary of the 1980 Brookhaven Symposium /36/. The overall accuracy of such calculations above - 10 MeV is, however, not well known and needs to be investigated. This requires detailed intercomparisons of results from different model codes, and validations of individual calculations from existing experimental information at a few energies for some characteristic target nuclei.Such comparisons can ensure that the parameters being used are appropriate for the extended energies and target masses. Furthermore,more systematic e&s *=ort is recruired to comoile,evaluate, document and dissiminate appropriate data for the 10 ' 50 MeV ' range.

6. Spallation Neutron Sources

6.1 Design Considerations

Typicalthick sample neutron yields versus proton energy measured for various target materials are shown in Fig.4 /I/. It can be seen that the number of neutrons per incident proton increases almost linearly with increasing proton energy. Furthermore, the integral yields increase rapidly with increasing target mass.

Fig. 4:

I I .’ ”

/

I

NEUTRON YIELD 10.2 cm x ,6l.O cm

I 50

t

“S

z PROTON ENERGY

k /

0 .5 1.0 I.5 2.0

PROTON ENERGY GeV

Total neutron yields obtained from the bombardement of thick targets by intermediate energy protons. Targets of Sn, Pb and U were cylinders: the first dimension shown being the diameter and the second the length. The Be target was rectangular /I/.

, lo+ zz xi0 . z x0

100 Ibl 10' 10' Neuimn Energy (MeV)

Double differential neutron

production cross sections for eight specified target materials at 900 laboratory angle /48/.

Above - 1.5 GeV neutron yields reach rapidly a saturation value, so that it is not very attractive to aim at higher proton energies. High intensity spallation neutron sources thus involve high energy (0.5 - 1.5 GeV) protons and thick (totally absorbing) heavy element targets. Typical primary neutron spec:tra from spallation reactions will be shown in Se'ct.6.3.1. These exhibit a two compo- nent structure according to the two main types of reaction mechanisms involved: Below ~20 MeV the spectral shape is governed by evaporation neutrons from highly exited nuclei with almost isotropic angular distributions. Neutrons originating from intra- nuclear cascadereactions produce a large frac~tion of high energy neutrons with maximum energies similar to those of the bombarding protons whose angular distribution is s?aongly forward peaked.

6.2 Examples of Spallation Neutron Sources

The characteristics of operating, constructed or planned spallation sources are shown in Table III.This table includes information on proton energies and currents, target materials and-,typical pulse widths together with average and instantaneous neutron intensities. Finally,the present status is given in the last column. It can be seen from Table III that the presently operating spallation sources still produce moderate average neutron intensity. But the slow neutron peak fluxes exeed already those available from fission reactors,particularly in the epithermal range. Also the more advanced

l ‘- .

.

Table III: Characteristics of Some Spatlation Neutron Sources’

Facility Location

Pro ton Target Pulse Rep P&k Avg. Peak Avg. Energy Material Widths Rate Current Current Intens. Intens.

status Ref.

(v see) W) (f-9 W) (n/s) (n/s)

UNR,US 800 w lo8100a) 1’20 2: .003 l*lol* l.1015 operating

KENS, Japan 500 W,U

3: tab)

20 .006 ,.,0’8d) 2.,0’(’ operating

IPNS I.IJS 500 u 5 80 .012 2. ,p d) 3. lOI operating

TNF,Canada 450 Cont.

50 -1

.I00 3.10’5 operating

SNS ,,UK 800, u 3, rob) .200 3.1020 4.1016 operating

SINq,Switz. 600 Pb-Bi Cont. 1. 6*10’6 under cons tr .

SNQ. FRG i 100 Pb,U 250’) 100 100 5. 4.1019 l.io’8 aborted

1371

1381

12,39/ 1401 I

s: 1411 I

142/

1431

URR/PSR,US 800 U 3,iob) 12 2000 .I 2. ,020 d) i.1016 under constr. 1441

TKF ,Canada 2000 unspec. 3 I 30000 .I 9. ,021 e) 3.10’6 planned I451

GEMINI, Jap. 800 unspec. 330 6. 101’ e, 1.10” planned /46/

ASPIJN,US 1 too unspec. 3,rob) ‘T :: i’5 9.,022 d,e) 3.io’7 planned 1471

a) For thermal and cold neutrons b) For epithermal and thermal neutrons c) Without compressor ring

d) For lowest pulse width e) Assuming a U target

. I ’

i .

-15-

0

conceptslike SNQ or ASPUN do not merely exploit the extreme capability .owing to the spallation process; 'such facilities do not involve other than presently available technology. For all of the presently operating spallation sources there'exist already proposals for greatly improved replacement facilities; so, the WNR/PSR, the TKF, the GEMINI and the ASPUN projects are either upgradings or future replacement co:ncepts for WNR, TNF KENS and IPNS, respectively. For the TRIUMF Kaon Facility (TKFj the production of neutron beams is,in fact , not one of the prime motivationsfor constructing this facility. But the high proton currents, in particular from the booster stage of accele- ration, also make it a unique source of spallation neutrons. For the SNQ, it should be mentioned that it has recently not been funded, so that it will not be built.

6~. 3 Nuclear Data Asnects

6.3.1 Source pronerties

There are a few systematic measurements of double differential neutron production cross sections for proton bombarding energies above 450 MeV /48,49,50/. The 900 data at !585 MeV proton energy measured by the KfK group /48/ for eight target materials are shown in Fig.5. The thin target data exhib:Lt an overall spectral shape similar to that of thick sample yields. The cross sections increase rapidly with increasing target mass. In contrast,the fraction of cascade neutrons above 15 MeVincreases .smoothly for lighter target elements. Similar data have been measured in the same study for two additional laboratory angles of 30' and 150°. A special feature of the angular distributions for heavy and medium weight nuclei is that neutron emission in .the evaporation region is almost isotropic, while emission in the cascade regions is strongly forward peaked. This is illustrated in Fig.6 which shows the corresponding results for. lead at the three emission angles of 3Oo, 90° and 1500.

All presently available neutron production cross section measure- ments,which quote typical uncertainties between 10 - 15%, e.xhibit considerable agreement in many instances,~ but there exist also severe discrepancies in a few cases. A particular case of disagree- ment is shown in Fig.7 which displays 300 neutron production cross sections for a few target elements commonly studied by the LAEL, the KfK and the ORAL groupsat some different proton energies. While below - 15 MeV there is reasonable agreement between the LANL and the KfK data(except Al) serious di.screpancies are found in the high energy region especially above 50 MeV. Apart from significant differences in the spectral shapes, absolute cross sections measured by the LANL group are up to a factor of five higher than the KfK data. For Al the KfK data above 120 MeV are, however, in good agreement with the ORNL data for 450 MeV protons. Recently a KFA/LANL Collaboration has measured 350 and 800 McV neutron production cross sections for U, Pb, and C at a few extreme

. ,*

’

l L

1

10 10' 10’

Neulrnn Energy (MN)

-16-

Fig.7: Comparison of measured double differential neutron production cross section for 30°. The proton bombarding

10

energies were 585 MeV in the KfK and 800 MeV in the LANL measurement. - 10 The shaded ORNL results 5 /SO/ refer to 450 MeV + proton energy. t

: 10' = s .z z co10

I

10’

Fig.6: Double differential neutron production cross sections for a typical heavy element target. Data are shown for three laboratory angles. of 300, 900 and 150° /48/.

z: E

10.

10‘

A : Pb(KfK) -zPb(LIINL) 0 =In (KfK)

--:ln (LANL) * : Fe (KfK)

---:Cu (LANL) o-Al IKfKl

---Al (LAiL) z=Al (ORNL)

I P 100 10’ ‘10’ 1

Neutmn Energy (MeV)

, ”

. . -17-

forward angles /52/. Particularly the results at 30' will be very useful, because they can help to solve the existing dis- crepancies in the earlier LANL and the KfK results at this angle. A more comprehensive intercomparison of available diffe- rential production cross section measurements is given by Cierjacks et al, /51/. What is urgently needed more than new measurements are extensive c.omparisons with calculations from intranuclear cascade evapo- ration models in order to check the overall reliability ~of presently used model parameters. Model calculations are especially needed for inter- and extrapolations into <energy and angle regions in which experimental results are :not available. A first step into this direction was made by Filge,s ,et al. /53/ who compared BETC/KFA-1 model calculations /54,/ with the 585 MeV KfK neutron data in great detail. A typical ex;ample from this work is shown in Fig.8. For the 900 data of lead, it can be seen that the XFA-code suitably predicts the absolutes evaporation cross sections below - 15 MeV. But it underpredicts the high&energy data increasingly with increasing neutron knergy. This 'trend is even more pronounced for larger emission ang.les as can be judged from Fig.9. In this figure the ratios of ca~lculated. to measured cross.sections are plotted versus neutron energy. Underoredictions of more than a factor of ten at the very high energies clearly indicate that model parameter modifications are urgently needed.

6.3.2 Source and accelerator shielding

The present nuclear data status in this field has been described in detail by Armstrong et al. /55/(see also contributed paper to this Seminar). A large fraction of data needed below - 20 MeV is available through the ENDF/B-V library /3/. On a largely extended energy scale two act elerator shie:!ding libraries have been generated at ORNL /56/ and L&YL /57/, which contain neutron data up to 400 I?eV and y-data below 20 MeV,. and only neutron data

up to 800 MeV, respectively. The most important shortcomings for use with intense spallation neutron sources are: 1. The lack of high energy data up to 1100 MeV (the design objective of several proposals for new sources). 2. 'Ithe lack of y-data (LANL library) or of y-data above 20 MeV (ORNL library) and 3. The lack of Legendre expansions abcve P3 and P respectively, for the corresponding angular distributions. Tze'latter deficiency was particularly stressed in the shielding paper of Armstrong et al. /55/ who showed that calculated neutron cross sections at high energies cannot be adequately reproduced by neither P3 nor P5 expansions. In* their example rather a Pig expansion is most appropriate. In Ref. 55 the authors also pointed out that the present data contained in the two shielding libraries need revision.

This is demonstrated in Fig.10, which displavs a number of measured, and calculated elastic scattering cross sections of iron together with data used in the LANL and in the ORNL (HILO) libraries.

- 18-

PB207 90 DEG. E IMEVI

Fig.8: Comparison of measured and calculated ~neutron'production cross sections for lead at 900 /53/. For the calculations the HETC/KFA-1 code /54/ was used.

?--l--m ‘. PB207 E., I ItEV 1

Fig.9: Ratios of calculated to measured neutron production cross sections for Pb versus energy. The three curves refer to data at emission angles of 30°, 90° and 150° /53/.

. . ’

-s .

- 19 -

l

2

10-l . 10’ 2 5 102 2 5 103 2 5 IO‘ 2

Neutron Energy [MeV]

Fig.10: Comparison of various measured and calculated neutron elastic scattering cross sections with data contained in present accelerator shielding libraries:

This comparison clearly shows the need for reevaluation. There appears to be also an urgent need of benchmark tests for deep penetration shielding problems /55/. Even irhouqh first attempts are being made at XFA for improving existing data libraries and tc validate various transport calculation, it is felt that such tasks need large-scale inter-laboratory effort, in order to be most efficient.

6.3.3. Safety ,and maintenance

The most comprehensive attempt to measure radioisotope production from proton bombardment of thick lead and uranium targets was recently made by Amian et al. /58/. These authors measured reaction rates for nearly 50 radioactive spallatibn and fission products by y-spectroscopy with a large Ge(Li) detector for each of the two targets at two proton bombarding enerqies of 600 and 1100 MeV. Back-up calculations of the corresponding reaction rates em- ploying the HETC/XFA-1 code were made by Fi.lges et-al,. /58/. Comparisons of measured and calculated reaction rates for uranium at 1100MeV are shown in Fig.11. There is a surprisingly good agreement of the shapes of both mass distributions. Unfortunately, low mass spectra are shifted by ~225 mass units against each other, and this shift is not yet well understood. Thus radioisotope production is another stressing case for suitable modifications of model parameters in the IIETC code, before it can be used for

- 20 -

Fig.11: Comparison of measured and calculated isotope production reaction rates versus mass of spallation and fission products. Experimental results are from Amian et al. /58/. Model calculations were made by Armstrong et al., using the HETC/KFA-1 code.

reliable predictions of lacking safety and maintenance data. Presently' almost all neutron and proton-induced activation cross sections above - 15 MeV must come from model calculations alone,

since experimental data are extremely. scarce.

6.3.4 Radiation damage and dosimetrv

Experimental information of spatial dependent materials damage parameters, such as damage energy, recoil energy displacement, gas production and transmutations above 20 MeV is extremely rare. At 0.6 and 3 GeV Kruger and Heymann have measured hydrogen and helium production cross'sections for C,O and Si /49/. Moreover, Kwiatkowski et al. /!iO/ studied mass- , energy- and angular-distri- butions of reaction products formed in collisions of 180 MeV protons with aluminium target nuclei. Such data are especially valuable for comparisons with model predictions which must provide the majority of high-energy data. Comparisons of the data from Ref. 50 showed that there is evidence for an enhanced energy de- position relative to model predictions from the intranuclear cascade- evaporation model. In contrast, preequilibrium calculations provide stronger energy damping, more consistent with the experi- mental results. Similar comparisons with existing experimental data and predictions from other model codes, were made by Amian et al. (see contributed paper) using the HETC/KFA-1 code. Their comparison shows partly also very poor agreement of the calculations with experimental data and predictions from other model codes.

I . ’ - 21-

“’ _

An outstanding observation from this work is that neither the KFA calculations nor the LANL calculations for damage energy and displacement cross sections agree with ENDF data at 20 MeV. Also the agreement between the measured recoil spectra from Ref. 50 and corresponding XFA calculations is rather poor, with differences of the order of more than ~a factor two. For dosi- metry purposes with the Radiation Effect Facility (REF) at Argonne, Birtcher et al. /61/ have produced a data library for 32 threshold reactions which employs.cross: Sections extended to 44 MeV using available data and calculations. Spectrum measure- ments employing the multifoil activation method and unfolding techniques were estimated to determine hicrh energy neutron fields with an accuracy of less than 15% ;'62/. According to the authors this is mainly due to the fact; that contributions from energies above 44 MeV are almost negligible. This may not be true for other spallation sources which produce neutrons on a largely extended energy scale.

l 7. Conclusions

The proper and economic design and operation of advanced intense neutron sources ~requires a large amount of new nuclear data. On the basis of the discussed problem areas a wide extension of existing evaluated data to the energy range from about 10 MeV to 1 GeV seems to be necessary. In this range very few measure- ments have been made, and there is little prospect for a drastical change of this situation. There exist; however, many nuclearmodel codes which are, in principle, suitable to calculate the lacking data. But the employed nuclearmodels are often based on assump- tions that cannot be fully justified theoretically. Therefore, model predictions need verification from comparisons with suitable ex?erimental'data. Ofcourse,.existing new intense neutron sources have been builton the basis of presently available data, and are properly running. But future conce+s aim at largely increased intensity of more energetic neutrons. These will certainly cause more severe,problems concerning shielding, maintenance, radiation damage and,dosimetry aspects. Two stressing examples of this kind are.the maintenance problems experienced with the Operating~LAMPF facility, and the shielding difficulties anticipated for a spallation source like SXQ. In the first case, the LAMPF accelerator has been designed for operation with an average beam current of 1 mA, and the present accelerator specifications are suitable for this kind of operation. Nonetheless, LAXPF has been operated in the past only with less than half of this value, because of maintenance problems. A 1 mA operation would produce a too high radiation level in the accelerator structure to .allow for short-term maintenance and repairment work. In the other case, deep penetration shielding of a spallation source like SNQ requires 17 decades of attenuation,i.e. an equi- valent of 6 m of iron shield. Only a 10% error inXatt is equivalent to about 1 m of iron or 1 - 2 orders of magnitude in radiation dose. 1 m of additional iron shield would increase the total amount of shielding material by more than :30:, and thus cause a major increase in investment costs. Puthcrmore, such an additional shield would also reduce the available space for experimental equipment.

- 22 - , *

L

l .

l

References

/2/

/3/

/4/

/5/

/6/

/7/

/a/ -/g/

/13/

/14/

/15/

/16/

J.S.Fraser and G.A.BartholomeW, "Spallation Neutron Sou in Neutron Sources for Basic Physics and Applications, Ed. S.Cierjacks, Pergamon Press, 1983, p.217ff

M.A.Kirk, Proc. Int. Coil. on Advanced Neutron Sources, ICANS VII, Chalk River Nucl. Laboratories, 1984, AECL-8,

ENDF(102) 2nd Edition, Brookhaven National Laboratory, BNL-NCS-50496

L.R.Greenwood, Symp. on Neutron Cross Sections from 10 : 50 MeV, Eds. M.R. Bhat and S.Pearlstein, Brookhaven Nat; Laboratory, 1980, BNL-NCS-51245, p-75

H.Farrar IV, D.W.Kneff, R.A.Britten and R.R.Heinrich, SymP. on Neutron Cross Sections from 10 to 40 MeV, Eds~. M.R.Bhat and S.Pearlstein, Brookhaven Nat. Laboratory, 1977, BNL-NCS-50681, p.175

H.H.Barschall, "14 MeV D-T Sources", in Neutron Sources Basic Physics and Applications, Pergamon Press, 1983, p*

H.H.Barschall, Neutron Induced Reactions, ,Physics and Ap cations, Vol~.lO, Institut of Physics EPRC, Slowak Academ of Sciences, 1982, p.279

X.A.Schmidt, private communication, 1985

M.C.Cline and C.R.Emigh, Proc. 1st ANS. Topical Meeting 0: the Technology of Controlled Nuclear Fusion, San Diego, 1974, p.313

R.Bdoth et al., Nucl. Instr. Meth. ,145 (1977)25

M.Drosg, Proc. IAEA Consultants Meeting on Neutron Sourc Properties, INDC(NDS), 1980

A.B.Smith, Symp. on Neutron Cross Sections from 16 to 50: Brookhaven Nat.Laboratory 1980, BNL-NCS-51245, p.61

X.E.Scheer, K.A. Schmidt and K.H. Hover, Int. Conf. on tt teractions of Neutrons with Nuclei, Lowell 1976, CONF-76C

R.Serber, Phys. Rev. ~(1947)1008

G.W.Schweimer, Nucl. Phys. AlC)O(1969)537 -- M.J.Saltmarsh C.A.Ludemann, Nucl. Instr. Mith.

C.B.Fulmer 145(1977) 81

and R.C.Styles, -

M.A.Lone and C.B.Bigham, "Intense Deuteron and Proton Ind~ (CW) Sources", in Neutron Sources for Basic Physics and Applications, Pergamon Press, 1983, p.133ff

A.LaMar Trego, Hanford Enqineering Development Laboratory: Report, HEDL-SA-1919FP, 1380

.

- 23 -

/19/ F.M.Mann, F.Schmittroth and L.L.Carter, Symp. on Neutron Cross Sections from 10 to 50 Me'?, 3NL-KS-51245, 1980, 9.517

/20/ P.Grand and A.N.Goland, Nucl. Instr. Meth. 145 (1977)49 --

/21/ T.Burrows and G.Wyant, "The Bibliography of Integral Charged Particle Nucl.Data", 4th Edit., Brookhaven Nat. Laboratory, 1981

/22/ LLNL Nuclear Data Libraries.. Lawrence Livermore Nat. Laboratory Report Series, CCRL-i400

/23/ S.Cierjacks et al., KS-1000, 1968

Kernforschungszentru Xarlsruhe Report,

/24/ D.C.Larson, Symp. on Neutron Cross Secticns from 1 to 50 HeV 3NL-NCS-5145, 1980, p.277

/ZS/ C.r.zanelli et al., ibid p.313

e /26/ P.W.Lisowski et al.,ibid pi301

/27/ Proc. Symp. one Neutron Cross Sections from~lo'to 40 MeV, 1977, Eds. icr,.R.Bhat and S. BrookhaGen Nat. Laboratory,

Pearlstein, SNL-NCS-50681

/28/ Proc. Symp. on Neutron Cross Brookhaven Nat. Laboratory, Pearlstein, 3NL-KS-51245

Sections from,SO to 50 XeV 1980, Eds. X.R.Shat and S.

/29/ L.L.Carter, R.J.Xorford and A.D.Wilcox, ibid p.431

/30/ D.L.Johnson, F.M.Xann and R.E.Schen+*r . . --- , ibid p.495

/31/ D.G.Port? and ti.X.Guizan, ikid p.:59

/32/ D.W.Neff et al., ibid 9.289

/33/ S.K.Qaim, S.!Zhatur. azd R.X6lfl.e, ibid 3.539

/34/ R.C.Zri,-ht and S.X.Grimes, ibid 3.245

/35/ R-Gold et al., ibid p.553

/36/ "Xoricshcp Resort on Xuclear xodei Cedes and Data Evaluation", ibidp.43

/37/ G.J.Russel, P.W.Lisowski, S.D.Sowe, N.S.?.Kinc; and M.iPl.:Xeyer Proc. Int. Conf. on Nuclear Data for Sc:ience and Technolcgy,. Ant-derp 1982, 0.831

/38/ H.Sakai, Proc. Int. Coil. on Advanced Neutron Sources, ICANS VII, Chalk River >Jucl. Laboratories, 1984, p.15

- 24 -

~ ” .p

‘

- I .

. /39/ G.H.Lander et al.,ibid p.40

. /40/ I.M.Thorson et al., ibid p-31

/41/ G.,IYanning j'ibidp.34

/42/ F.Atchison., ibidp.23

/43/ G.S.Bauer, ibid p.2

/44/ C..D.Bowman, ibid p.9

/45/ E.W.Slackmore, ibid p-55

/46/ iI.Sa!!ai, ibid p.50

/47/ Th.Roe and R.L.Kustom,.ibid p.46

/48/ S.Cierjacks, Y.Hino, S.D.,Eowe, F.Rau?p and L.Suth, Proc. Int. Conf. on Nuclear Data for Science and Tech?olo5y,

a Antwerp 1982, p.383

/49/ S-D-Howe, PhD Thesis, Kansas State University,f 980

/SO/ J.W.Wachter, W.A.Gibbor, and W.R.Burrus, Phys. .Rev.,C6(:972) 1496

-~

/ST/ S.Cierjac!ks, Y..Hino, F.Raum, S.D.Sowe, L.&t:?, D.?ilges, P,Cloth and. T.~.X~strong~,-"Slstenatics of -Angular-Cesendent Xeutrcn Production by 585 Me-V Proton on Targets wit,? 12 (A < 238"~, prepared for publicatioxz

/52/ M.Eeyer et al., int. Con?. on Nucl. Dat,a. fcr Basic and Applied Physics, Santa Fe, Nay 1985, Contrib. JC 23

/S3/ D .,Filces, S . Cier j a&s, Y . Eino , T.:i..L~ast.ronq~ ar.d ?.Cloth, icer3forsc:?~-,gsa~-~a~e JUL 7 c:i asd. :ZernZor:schuncp zentr-xiz Karls rche Report, JGl-1960, XfX-3779, i9a4

a /54/ T.w.Armstrong, s.c1ot.i, 3.Colbo:n a..d f).ci1q2s, "zTC/xF'.A-l", KerzfcrschW.gsanlage JElich Report; ZX.-Sgez-196 ,19:3

/55/ T.:~.Armstzo;lq, ?.Clot4, D.Filges and E.Sehaal, ~Proceedinqs ICAXS VII, Challc Riv2r Laboratories, 1934, p.208

/56/ EILO Data ?ac!cage, "66 Neutron., 21 Gaima Ray Gzcu? Cross Sections for Radiation Transport for Neutron Energies q to 400 KeV", DCL-a7,, RSIC

/57/ W.B.Wilson, "Nuclear Data Development and Shield Design for Neutrons Below 60 MeV, Los Alamos Report, LA-7159-T, 197,8; and private comnunication thereafter

‘_ .\ . . :, -2s-

= e .

. /58/ W.Amian, N.F.Peek, D.J.Shadan and G.Sterzenbach,

Proceedings ICAiiS VII, Chalk River Nucl. Laboratories, . 1984, p.62

/S9/ S,T.Kruger and D.Heymann, ?hys.Rev: =(1973)2179

/60/ K.Xwiatkowski et al., Phys.Rev.Lett. 50(1983)1648 -.

/61/ R.C.Birtcher, M.A.Xirk, T.M.Blewett and R.C.Greenwood, Proceedings ICANS VI, Argonne Nat. Laboratory, 1962 , p.407

/62/ L.R.Greenwood and R.J.Popek, ibid p.606