1. Name of Experiment:
    ------------------
    FNS/JAERI Clean Benchmark Experiment on Tungsten Cylindrical Assembly (1993)


 2. Purpose and Phenomena Tested:
    ----------------------------
    A tungsten cylindrical assembly (diameter = 629 mm, height = 507 mm) was
    irradiated in the D-T neutron source of Fusion Neutronic Source (FNS) 
    facility at JAERI. Neutron spectra down to 5 keV, dosimetry reaction rates,
    gamma-ray spectra and gamma-ray heating rates were measured at 3 positions
    up to 380 mm inside the tungsten assembly.


 3. Description of Source and Experimental Configuration:
    ----------------------------------------------------
    The experimental configuration was basically the same as those used for some
    previous clean benchmark experiments like the Lithium-Oxide, Graphite, Copper,
    Beryllium, Vanadium assemblies.
    The Tungsten assembly was made by stacking bricks of (50.7-50.8 mm) with
    thin aluminium support frame in quasi-cylindrical shape (diameter = 629 mm,
    height = 507 mm). A view of the experimental assembly is shown in Fig. 1.
    The tungsten is not pure, but an alloy with a small amount of nickel and
    copper. The material compositions of the tungsten assembly is given in
    Table 1.

    The experimental assembly was located in front of the neutron source at a
    distance of 200 mm from the target. The tritium-titanium target of ~3.7E11 Bq
    was bombarded by a deuteron beam of 350 keV energy to produce D-T neutrons.
    The number of source D-T neutrons generated during each measurement was
    determined by the alpha-particle detector with an accuracy of 2~3 %. 

    The D-T neutron source can be roughly described as an isotropic point 14-MeV
    neutron source.  The source neutron spectrum and intensity, however, depend
    slightly on the emission angle. The angle-dependent source characteristics
    were investigated experimentally and theoretically in detail, and a source
    subroutine for the Monte Carlo transport code MCNP-4 has been prepared to
    simulate the source condition precisely. This routine is listed in [3]. A
    comparison between the measured and calculated angular distribution is given
    in Fig. 3.2.7. in [4]. As an adequate alternative to the use of the MCNP
    routine it is suggested in [1] that the source spectrum of neutrons emitted
    toward the 0 degree direction with respect to the deuteron beam direction can
    be used for the incident neutron spectrum on the whole front surface of the
    assembly. No quantitative estimation of difference between both sources could
    be though found in the available literature. The 0 degree source neutron
    spectrum is shown in Fig. 2, and input cards of MCNP for description of the
    source spectrum are shown in the MCNP sample input file. The si1 and sp1 cards
    indicate upper neutron energies of the energy bins in MeV and probabilities
    in the bins, respectively. The dir and vec parameters with the sb2 card are
    used for variance reduction with the source biasing method. The weight of a
    source neutron specified by the wgt parameter, 1.1261, is larger than 1.0
    because more D-T neutrons are emitted to the forward direction than in the
    backward direction with respect to the deuteron beam direction.  

    The tritium target region is also a source of gamma-rays, namely, target gamma-
    rays, created by interaction of the source neutrons with structural materials
    of the target. Consideration of the target gamma-rays, however, is not needed
    in calculations of gamma-ray heating rates because contribution of the target
    gamma-ray to the measured heating rates has been already subtracted in the
    experimental data.  On the other hand, the target gamma-ray contribution is
    involved in the measured gamma-ray spectra. The contribution at the detector
    positions deep inside the experimental assemblies is negligible, representing
    at most few percents, because the experimental assembly largely attenuates the
    target gamma-rays. Accordingly, it is not necessary to consider the target
    gamma-rays in transport calculations for the clean benchmark experiments.

    Since the W assemblies were located at least 4 m from the experimental room
    walls and floor, the contribution of the background neutrons and gamma-rays
    coming from the room walls and floor on the measured quantities was negligibly
    small.

 4. Measurement System:
    ------------------
    Three experimental channels for insertion of detectors were set on the lateral
    surface of the assembly, and detectors were placed on the central axis of the
    assembly at four positions: one on the front surface and other three inside
    the assembly at the depths of 76, 228 and 380 mm. The report [1] states that
    the spectra measurements were performed one by one, we conclude therefore
    that the other experimental channels were closed (plugged with tungsten). 
    The dosimetric foils were irradiated together and simultaneously in plural
    detector channels.

    No information on the background correction was found in the literature.

    Six techniques were employed to measure neutrons and gamma-rays. The detailed
    description of the experimental techniques is given in [1] (pages 5-20):

    (1) 14 mm-diameter spherical NE213 liquid organic scintillator was used as
        the fast (above 2 MeV) neutron spectrometer. It was inserted into the
        experimental channel hole of 22 mm diameter. 
        Various sources of the NE213 measurement uncertainties are discussed in
        [1] (page 5) and the systematic errors are summarized in Table 2.

    (2) A pair of proton recoil gas proportional counters (PRC) was used to
        measure neutron spectrum in the 5 keV to 1 MeV energy range. The counter
        has a cylindrical shape with the outer diameter of 19 mm and the effective
        length of 127 mm. The counter is inserted into an experimental hole of
        21 mm in diameter with its pre-amplifier. PRC is made of type 304 stainless
        steel of a thickness of 0.41 mm.

        Two types of counters were used to measure neutron spectrum in a wide
        energy range. Hydrogen gas at 0.5677 MPa (5.789 kgf/cm2) with 1 percent
        of CH4 filled counter was used for low-energy neutrons from 3 keV to 150
        keV while for the high-energy neutrons from 150 keV to 1 MeV the counter
        was filled with 50-50 mixture of hydrogen and argon gases with 1.8 percent
        of nitrogen at 0.6102 MPa (6.222 kgf/cm2).

        Error Assessment: Possible error sources are gas pressure (number of
        hydrogen atom), n-p scattering cross section, fitting error for
        differentiation of recoil proton spectrum due to count statistics and
        calibration of recoil proton energy. The fitting error is the largest,
        ~ 3-10 % above 10 keV, while the other errors are expected to be less
        than 1%. Neutron spectra below 10 keV tend to become smaller due to the
        uncertainty of the W-value, which is the average energy loss per ion pair.
        The error due to W-value is not included in the experimental errors.

    (3) Slowing down time (SDT) method for the neutron spectrum from 1 eV to 300 eV.
        A BF3 gas proportional counter with an outer diameter of 14 mm and an 
        effective length of 99 mm, containing 96 % boron-10 (B-10) enriched BF3
        gas at the pressure of 71.5 kPa, was used for neutron detection. Using the
        standard thermal neutron field the effective number of B-10 atoms in the
        counter was determined to be 2.18E20 +- 3 %. The counter was inserted into
        one of the experimental holes of the experimental assembly.

    (4) Dosimetry reaction rate of the Al-27(n,alpha)Na-24, Nb-93(n,2n)Nb-92m,
        In-115(n,n’)In-115m, 186W(n,gamma)187W and Au-197(n,gamma)Au-198 reactions
        were measured by the foil activation method. Table 3 provides the
        characteristics for the reactions used.
        Typical sample size was 10 mm in diameter and 1 mm in thickness for
        activation foils except indium and gold. Indium foils had dimensions of
        10 x 10 x 1 mm3. In order to minimize the self-shielding effect for the
        Au-197(n,gamma) reaction, gold foils with a size of 10 x 10 x 0.001 mm3
        were adopted.

        Experimental Error and Uncertainty: Major sources of the error for the
        reaction rate were the gamma-ray counting statistics (0.1 ~ several %)
        and the detector efficiency (2 ~ 3 %). The error for sum-peak correction
        was estimated less than 2 % depending on the decay mode and fraction of
        multiple gamma-ray cascade. The error for the decay correction was
        reflected from the error of half-life of the activity. If the half-life
        was accurate, the error for the saturation factor should be less than
        1 % even for the short half-life activities.
        The other errors associated with foil weight, gamma-ray self-absorption,
        irradiation time, cooling time and counting time were negligibly small.
        The error for neutron yield was estimated to be 2 %.  The overall error
        for the major part of reaction rate ranged between 3 ~ 6 %.  Some data
        for high threshold reaction in the deep positions suffered from poor
        counting statistics due to low activation rate.

    (5) Prompt gamma-ray spectrum were measured by a 40 mm diameter spherical
        BC537 liquid organic scintillation counter. Outer diameter of the detector
        is 48 mm and length including a photomultiplier assembly is 262 mm.
        Experimental uncertainty: As explained in (6) below, typical experimental
        uncertainties of the gamma-ray heating rate measured by TLDs range
        between 7 ~ 15 %. The gamma-ray spectra are normalized to the gamma-ray
        heating rates. Uncertainties of about 10 % is introduced by the
        normalization procedure. All of the rest of experimental uncertainties,
        such as statistical errors, uncertainties of the response functions,
        determination of source D-T neutrons and subtraction of decay gamma-rays,
        are less than 5 %.  Therefore, total experimental uncertainties are
        approximately 15 ~ 20 %. Note that only the statistical erorrs are
        included in the Table 7.

    (6) Gamma-ray heating rate were measured by thermoluminescent dosimeters
        (TLDs) in combination with the atomic number interpolation method.
        Gamma-ray heating rates of vanadium were deduced by interpolating the
        gamma-ray heating rates measured by three types of TLDs, Mg2SiO4 (MSO,
        effective atomic number Zeff = 11.1), Sr2SiO4 (SSO, Zeff = 32.5) and
        Ba2SiO4 (BSO, Zeff = 49.9).

        Sources of error in the measured gamma-ray heating rates are as follows:

        Statistical deviation of four TLDs      5 - 15 %
        Number of neutrons generated            2 - 3  %
        Calibration of the TLD reader             5    %

        Due to the subtraction of target gamma-rays and neutron response, the
        following uncertainties are to be added to the above errors according to
        the quadratic propagation of error.

        Response functions for neutrons    30 %
        Neutron energy spectra             10 %
        Target gamma-ray                   20 %

        Overall errors for the obtained gamma-ray heating rates of vanadium are
        ~ 10%, except at the positions in the front surface of the assemblies
        where they are 20 ~ 25 %.


 5. Description of Results and Analysis:
    -----------------------------------
    The measured neutron spectra by the NE213 and PRC at the depths of 76 mm,
    228 mm and 380 mm are shown in Figs.3, 4 and 5, respectively. The numerical
    data of the spectra measured by the NE213 and PRC are given in Tables 4 and 5,
    respectively. A measurement of neutron spectra below 10 keV by the SDT method
    was attempted. However, it could not be made because of too low neutron flux
    in the energy range. The measured dosimetry reaction rates are shown in Table6.
    The measured gamma-ray spectra at the three positions in the tungsten assembly
    are shown in Figs.6, 7 and 8. The numerical data of the spectra are given in
    Table 7. The Table 8 summarises the measured gamma-ray heating rates in tungsten.
    Most of the observed gamma-rays at the front surface of the assembly, 0 mm, 
    are produced by high energy neutron of > 1 MeV. At 76 mm, both halves of 
    observed gamma-rays are produced by neutrons of > 1 MeV and 0.01-1 MeV.
    At 228 and 380 mm, neutrons with energy between 1 keV - 1 MeV contribute
    predominantly to produce secondary gamma-rays. 

    Example of Experiment Analysis:
    The sample input for the MCNP-4A code is given in file mcnp-w.inp. The
    input was taken from [1], except that the 0 degree source was used instead
    of the source subroutine (not provided in the document [1]).

    Transport calculations with the MCNP-4A code are presented in [2]. The
    calculations were performed using the JENDL-3.2, JENDL-FF and FENDL/E-1.0
    nuclear data libraries.

    The calculations were performed at the OECD/NEA Data Bank [5] using the
    included MCNP inputs and the FENDL-2 (=JENDL-FF) and ENDF/B-VI.8 cross section
    evaluations. The neutron spectra and the corresponding C/E ratios are
    presented in Figures 3, 4 and 5. The measured fusion peak is wider than the
    one calculated with the provided MCNP input. The gamma spectra obtained using
    the ENDF/B-VI.3 data are shown in Figures 6, 7 and 8. They indicate reasonable
    agreement between the experiment and the calculation.


 6. Special Features:
    ----------------
    None


 7. Author/Organizer:
    ----------------
    Experiment and analysis:
    F. Maekawa, C. Konno, Y. Kasugai, Y. Oyama, Y. Ikeda
    Japan Atomic Energy Research Institute
    Tokai-mura, Naka-gun, Ibaraki-ken, 319-1195 Japan
    Phone: 81-292-82-6809 (for Y.O.) or -6015 (for H.M.)
    Fax: 81-292-82-5996 (for Y.O.) or -6365 (for H.M.)
    E-mail: oyama@cens.tokai.jaeri.go.jp
    or      fujio@cens.tokai.jaeri.go.jp

    Compiler of data for Sinbad:
    S. Kitsos
    OECD/NEA, 12 bd des Iles, 92130 Issy les Moulineaux, France
    E-mail: stavros.kitsos@free.fr

    Reviewer of compiled data:
    I. Kodeli
    OECD/NEA, 12 bd des Iles, 92130 Issy les Moulineaux, France
    e-mail: ivo.kodeli@oecd.org


 8. Availability:
    ------------
    Unrestricted


 9. References:
    ----------

    [1] F. Maekawa, C. Konno, Y. Kasugai, Y. Oyama, Y. Ikeda: 
        “Data Collection of Fusion Neutronics Benchmark Experiment Conducted at
        FNS/JAERI”, JAERI-Data/Code 98-021 (1998).
    [2] F. Maekawa, M. Wada, C. Ichihara, Y. Makita, A.Takahashi, Y. Oyama: 
        “Compilation of Benchmark Results for Fusion Related Nuclear Data ”,
        JAERI-Data/Code 98-024 (1998).
    [3] F. Maekawa, C. Konno, K. Kosako, Y. Oyama, Y. Ikeda and H. Maekawa: 
        “Bulk Shielding Experiments on Large SS316 Assemblies Bombarded by 
        D-T Neutrons, Volume II: Analysis“, JAERI-Research 94-044 (1994).
    [4] Y. Oyama: “Experimental Study of Angular Neutron Flux Spectra on a Slab
        Surface to Assess Nuclear Data and Calculational Methods for a Fusion
        Reactor Design“, JAERI-M 88-101 (1988).
    [5] I. Kodeli, Recent Progress in the SINBAD Project, EFFDOC-866, EFF
        Meeting, Issy-les-Moulinaux (April 2003)

    References to other useful documents can be found in the above reports.


10. Data and Format:
    ---------------

    DETAILED FILE DESCRIPTIONS
    --------------------------
        Filename     Size[bytes]   Content
    ---------------- ----------- -------------
  1 fnsw-abs.htm       20,069  This information file
  2 fnsw-exp.htm       30,854  Description of experiment
  3 mcnp-w.inp         12,010  Input data for MCNP-4A calculations
  4 fnsw-1.gif         39,827  Fig. 1: Experimental assembly of tungsten
  5 fnsw-2.gif         11,332  Fig. 2: Source neutron spectrum emitted towards the
                                       0 degree from the target.
  6 fnsw-3.gif         27,839  Fig. 3: Neutron spectra at z=76 mm in W (ref. [5])
  7 fnsw-4.gif         29,378  Fig. 4: Neutron spectra at z=228 mm in W (ref. [5])
  8 fnsw-5.gif         28,549  Fig. 5: Neutron spectra at z=380 mm in W (ref. [5])
  9 fnsw-6.jpg         37,080  Fig. 6: Gamma-ray spectra at z=76 mm in W (ref. [5])
 10 fnsw-7.jpg         36,395  Fig. 7: Gamma-ray spectra at z=228 mm in W (ref. [5])
 11 fnsw-8.jpg         36,642  Fig. 8: Gamma-ray spectra at z=380 mm in W (ref. [5])
 12 j98-021.pdf     6,886,161  Reference JAERI-Data/Code 98-021
 13 j98-024.pdf    10,339,840  Reference JAERI-Data/Code 98-024
  

    File fnsw-exp.htm contains the following tables:

      Table 1:      Material specification
      Table 2:      Systematic errors for neutron spectra measurements by NE213
      Table 3:      Dosimetry Reactions used in the foil activation method
      Tables 4-5:   Measured neutron spectra
      Tables 6:     Measured dosimetry reaction rates
      Table 7:      Gamma-ray spectra
      Table 8:      TLD Measurements

   Figures are included in GIF and JPG formats.

SINBAD Benchmark Generation Date: 05/2003
SINBAD Benchmark Last Update: 05/2003