Benchmarking of the latest Neutron ... - Nuclear Energy Agency · Benchmarking of the latest Neutron and Gamma Transport Cross Sections for Oxygen, Iron and Uranium in clean Benchmarks

KIT – The Research University in the Helmholtz Association

INSTITUTE for NEUTRON PHYSICS and REACTOR TECHNOLOGY (INR)

www.kit.edu

Benchmarking of the latest Neutron and Gamma Transport Cross Sections for Oxygen, Iron and Uranium in clean Benchmarks driven by D-T, 252Cf and Reactor sources

in collaboration with Experts from CIELO, JEFF, ….:

D. Brown, R. Capote, M. Chadwick, M. Herman, B. Jansky, I. Kodeli,

A. Koning, L. Leal, M. Majerle, A. Plompen, D. Rochman, P. Talou, A. Trkov, …

S.P. Simakov and U. Fischer

2 Long Terma Int. Collaboration, 18 - 21 Dec 2017, IAEA, Vienna

Nuclei Considered:

I. Oxygen: our (not completed yet) validation results also presented as EFFDOC-13XX, NEA Data Bank, Nov 2017

II. Iron: a) continuation of our recent work presented as JEFFDOC-1851 = EFFDOC-1319, Apr 2017; EFFDOC-1303, Dec 2016

b) MC propagation of TENDL-2017 random files to Covariancies

for n & γ-leakage energy spectra first time presented at TALYS/TENDL Workshop, Nov 2017, Prague; then EFFDOC-13XX, Nov 2017

III. Uranium: revisiting our old validation presented as a contribution to ND-2004 AIP Conf. 769(2004)67, Santa Fe, Oct 2004

For these Nuclei following Evaluated Transport Data are under testing:

- ENDF/B-VII.1 and ENDF/B-VIII.0 (= VIII.β4, CIELO)

- JEFF-3.3T4 and “JEFF-4.0T1” (only 56Fe & 54Fe – L. Leal’ work for next JEFF)

- TENDL-2017 evaluation and BMC random files (only Fe stable isotopes)

Tools used: - File processing by NJOY-2012.82 (from LANL web) or .99 (NEA Data Bank)

- MC simulation of considered benchmarks by MCNP-6.1

(for TOF measurements: TOF was simulated then transformed to Energy Spectra)

- dedicated f.95 subroutines for post processing of simulation results

Content

http://www.oecd-nea.org/dbdata/jeff/effdoc.html











http://www.ccfe.ac.uk/easyworkshops.aspx




https://doi.org/10.1063/1.1944958

3

I. Oxygen

Long Terma Int. Collaboration, 18 - 21 Dec 2017, IAEA, Vienna


I. Oxygen (0): overview of

Experiments with Liquid Oxygen involved in present analysis:

Lab

Years

Source:

<Energy>

No. of Spheres:

Wall Thickn., cm

Methoda Detector, Threshold Numerical

Data from

FNS

1991

D-T:

≈ 14 MeV

Cylinder:

thick = 20 cm

Time of Flight

(TOF)

Leaking Neutrons:

NE-213 > 50 keV

Y. Oyama et al.

SINBAD

NEA-1553/61

LLL

1976

D-T:

≈ 14 MeV

1 sphere:

Wall ≈ 10 cm

Time of Flight

(TOF)

Leaking Neutrons:

Pilot B > 1.6 MeV

L. Hansen et al.

NSE 60(1976)27

C. Wong UCRL-

51144(1971)

ORNL

1967

Reactor

Fission :

≈ 2.1 MeV

Cylinder: length =

61, 92 cm (?),

152 cm (available)

Pulse Height

(PH)

Uncollided Neutrons

(pure transmission):

NE-213: 1.9 - 8.6 MeV

C .Clifford et al.

SINBAD

NEA-1517/59

Additionals/Comments:

- liquid Oxygen Transmission experiment (En = 20 - 30 MeV)

in progress at NPI/Rez, M. Majele et al. EFFDOC-1323, Apr 2017

- no gamma-leaking measurements ! (then look what we have for water ?)

https://www.oecd-nea.org/science/wprs/shielding/sinbad/fns_o/fnso-abs.htm



https://www.oecd-nea.org/science/wprs/shielding/sinbad/SDT2/SDT2_A.HTM




I. Oxygen (1): FNS/JAERI, 1991

Neutron scattering benchmark with pure Oxygen, NEA-1553/61:

Neutron Leakage Spectra at 0, 12, 25, 42, 67 deg. from Liquid Oxygen Disc (l = 20 cm × Ø60 cm)

irradiated by pulsed 14 MeV source

Description of Experiment is available in:

- e.g.: Y.Oyama et al., NDST 1991, p.337; Fig.3 =>

- Details & Numerical Data including

author’s MCNP input - in SINBAD

Essential experimental details which impact on modelling and conclusion:

- measured by TOF which was also represented in MCNP simulations

- background was measured/subtracted by filling collimator hole with iron/polyethylene plugs

- efficiency of NE-213 detector was not measured but was MC-calculated by authors

Our validation results of current evaluations on next slide (all O isotopes are included in MC) =>





Findings: - JEFF-3.3T4 = ENDF/B-VII.1 and ENDF/B-VIII.0 are identical

except ≈5% difference only in 3 – 7 MeV, where ENDF/B-VIII.0 trends to unity

I. Oxygen (1): Neutron Leakage from JAERI cylinder with D-T source

0.05 0.2 0.5 2 5 200.1 1 10

10-11

10-10

10-9

10-8

10-7

10-6

10-5

10-4

JEFF-3.3T4

12.2o/10

1

Neutron Energy, MeV

66.8o/10

4

= 0.0o/10

0

ENDF/B-VIII.0

ENDF/B-VII.1

liquid O slab (t=20cm) + T(d,n)

An

gu

lar

Ne

utr

on

Flu

x,

1/s

r/cm

2/M

eV

Experiment:

41.8o/10

3

24.9o/10

2

0.8

1.0

1.2

1.4

JEFF-3.3T4

= 66.8o

ENDF/B-VII.1ENDF/B-VIII.0

= 0.0o

0.8

1.0

1.2

1.4

= 41.8o

= 12.2o

liquid O cylin. slab (t=20cm) + T(d,n)

C/E

fo

r A

ng

ula

r N

eu

tron

Flu

x

0.6

0.8

1.0

1.2 = 24.9o

0.6

0.8

1.0

1.2

0.05 0.2 0.5 2 5 200.1 1 100.4

0.6

0.8

1.0

1.2

Neutron Energy, MeV

?


Observations :

- 1st collision dominates

- ENDF/B-VII.1 is practically identical to ENDF/B-VIII.0

however gives ≈2% more only for 4 - 8 MeV

(similar to what we saw for FNS/JAERI slab)

- both confirm 14 MeV transmission

but overpredict secondary neutrons 4-9 MeV

(it similar as FNS shows at 0 - 25 deg)

I. Oxygen (2): LLL pulsed sphere with D-T source, 1967

0 2 4 6 8 10 12 14 16

0.8

1.0

1.2

1.4 ENDF/B-VIII.0

ENDF/B-VII.1

C/E

Neutron Energy, MeV

10-2

10-1

2x10-3

8x10-1

1st collision

ENDF/B-VII.1

ENDF/B-VIII.0

LLL: R/r = 10.48/0.26 cm

LLL Oxygen sphere with D-T source

Le

akag

e S

pectr

um

, 1/M

eV


I. Oxygen (3): ORNL, Tower Shielding Facility, 1967

Uncollided transmission through Liquid Oxygen

(length: 24in = 60.96 cm & 36in = 91.44 cm; dia: 4in = 10.16 cm)

irradiated by TSR-II reactor spectrum ranged 1.9 to 8.6 MeV

Description of Experiment is available in:

- e.g.: C.Clifford et al., NSE 27(1967)299; E.Straker ORNL-TM-2242(1968), -3868(1972)

- SINBAD NEA-1517/59: description + numerical data (only 60 in. = 152,4 cm), no MCNP model

Our preliminary results on next slide (all O isotopes are included in MC) =>





Observations but with preliminary set-up Model:

- This Reactor Spectrum Transmission Experiment indeed

sensitive only to Uncollided Neutrons, i.e. to σtot :

the fraction of Collided Neutrons seen by detector

amounts only ≈ 5 10-6

- It is insensitive to the change of Oxygen Temperature

from boiling 90.2oK (-183oC) to room one:

σtot changes only below 1 eV

- 152 cm of liquid Oxygen reduce Neutrons by 10-4 !,

Δσ/σ = (n l σ)-1 ΔT/T = 6.5 ΔT/T !

e.g.: 10% change of ΔT/T will demand Δσtot ≈ 0.65 b !

- JEFF-3.3T4 is practically identical to ENDF/B-VII.1

- ENDF/B-VIII.0 gives 10% less then VII.1 only for 5-8 MeV

or similar trend which we observe

for FNS/JAERI slab and LLL sphere

I. Oxygen (3): ORNL, Tower Shielding Facility, 1967 (cont.)

1 2 3 4 5 6 7 8 9 10 11 1212

0.4

0.6

0.8

1.0

1.2

1.4

1.6

JEFF-3.3T4ENDF/B-VII.1

ENDF/B-VIII.0

C/E

Neutron Energy, MeV

10-6

10-5

10-4

10-3

10-2

10-1

(tot) ENDF/B-VIII.0 & VII.1

Trasmission

Liquid O broomstick (60in=152cm) + TSR-II

Source: TSR-II spectrum/10

4

ENDF/B-VIII.0

ENDF/B-VII.1

Ne

utr

on

Flu

x, 1

/cm

2/m

in/M

eV

/W

100

101

102

103

104

100

101

102

103

104

(t

ot)

, b

10

II. Iron, the first part:

Secondary Neutron Validation



II. Iron (0): overview of

Experiments involved in the present analysis:

Lab

Years

Source:

<Energy>

No. of Spheres:

Wall Thickn., cm

Method Detector, Threshold Numerical

Data from

IPPE

90-98

D-T:

≈ 14 MeV

5 spheres:

Wall = 2.5, 7.5,

10.0, 18.1, 28 cm

Time of Flight

(TOF)

Neutrons:

p-Ter.scint. > 50 keV

S.Simakov et al.

SINBAD Fusion:

NEA-1553/75

IPPE

1985

252Cf:

≈ 2.1 MeV

6 spheres: Wall =

4, 9, 19, 23, 29, 34

cm

Pulse Height

(PH)

Neutrons:

H-prop. 5 – 700 keV

Stilben 0.2 - 17 MeV

Gammas:


L.Trykov et al.

ICSBEP/DICE:

ALARM-CF_FE-

SHIELD-001

LLL

1976

D-T:

≈ 14 MeV

3 spheres:

Wall ≈ 4, 13, 22

cm

Time of Flight

(TOF)

Neutrons:

NE-213 > 1.6 MeV

L.Hansen et al.

NSE 60(1976)27

C.Wong UCRL-

51144(1971)

LLL

1990

D-T:

≈ 14 MeV

1 sphere:

Wall ≈ 4 cm

Time of Flight

(TOF)

Neutrons:

NE-213 > 1.6 MeV

Gammas (e-Recoils):

NE-213 > 0.255 MeV

E.Goldberg et al.

NSE105(1990)319

VNIITF

1991

D-T:

≈ 14 MeV

1 sphere &

1 semi-sphere:

Wall = 5 cm

Time of Flight

(TOF)

Neutrons:

Stilben > 0.25 MeV

Gammas (e-Recoils):

Stilben > 0.35 MeV

A.Saukov et al.

SINBAD Fusion:

NEA-1517/74

http://www.oecd-nea.org/tools/abstract/detail/nea-1553



https://www.oecd-nea.org/science/wpncs/icsbep/dice.html

https://www.oecd-nea.org/science/wpncs/icsbep/dice.html

http://www.oecd-nea.org/tools/abstract/detail/nea-1517/74




II. Iron (1): Results for IPPE D-T pulsed sphere (Wall=28cm), Neutron leakage

Observations for 14 MeV pulsed

and thickest (Wall = 28 cm) sphere:

- JEFF-4.0T1 (Geel) is different from EFF-3.3T4

but comparable with it from view abs(C/E -1)

- ENDF/B-VIII.0 looks better than ENDF/VII.1 0.05 0.2 0.5 2 50.1 1 10 20

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

JEFF-3.3T4JEFF-4.0

ENDF/B-VIII.

ENDF/B-VII.1

C/E

Neutron Energy, MeV

10-3

10-2

10-1

100

Experiment

ENDF/B-VIII. ENDF/B-VII.1

, b

Fe-sphere (R/r=30/2cm) + T(d,n)

n-L

ea

ka

ge

Sp

ectr

um

, 1

/MeV

tot(nat

Fe)

100

101

102


II. Iron (2): Comparison of IPPE (28cm) and LLL (22cm) spheres with D-T pulsed source: Neutron leakage

Observations from comparison of

LLL (Wall = 22cm) and IPPE (Wall= 28 cm):

- LLL: Neutrons > 1.6 MeV (for convenience of

comparison I converted original ToF spectra

in Energy ones and folded with Gauss)

- IPPE: Neutrons > 0.1 MeV

- both: show similar C/E in overlapping energies,

e.g. 4 - 10 MeV is overestimated by ENDF/VII.1,

where ENDF/B-VIII.0 performs better 0 2 4 6 8 10 12 14 160.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

ENDF/B-VIII.0

ENDF/B-VII.1

C/E

Neutron Energy, MeV

10-3

10-2

10-1

ENDF/B-VII.1

ENDF/B-VIII.0

LLL: R/r = 22.3/0.5cm

IPPE: R/r = 30.0/2.0cm

IPPE & LLL Fe spheres with D-T source

Le

akag

e S

pectr

um

, 1/M

eV


II. Iron (3): Results for IPPE Fe sphere (Wall=29 cm) + 252Cf(s.f.), Neutron Leakage

Observations from C/E comparison:

- JEFF-3.3T4 looks even more better than others

- possible reasons for underestimation

(σtot, angular distribution …) by ENDF/B-VIII.0

were studied 0.03 0.3 30.01 0.1 1 10 20

0.7

0.8

0.9

1.0

1.1

1.2

0.7

0.8

0.9

1.0

1.1

1.2JEFF-4.0T1 JEFF-3.3T4

ENDF/B-VII.1

ENDF/B-VIII.0

C/E

n-L

ea

ka

ge

Neutron Energy, MeV

10-5

10-3

10-1

101

ENDF/B-VII.1

ENDF/B-VIII.0

tot

(nat

Fe)

ENDF/B-VIII.0

ENDF/B-VII.1

Fe-sphere (R/r=30/1cm) +252

Cf

n-L

ea

ka

ge

, 1

/MeV

/n-S

ou

rce

252Cf TFNS

0.2

1

10

50

Before MCNP simulation we modified MCNP

input given in ICSBEP/Alarm-Cf_Fe-Shield-001:

- 252Cf(s.f.) source neutron spectrum (given as

Watt with Froehner parameters) was replaced

by PFNS Standard (Mannhart evaluation),

νp = 3.7590

- and we added delayed neutrons (νd = 0.0086)

with DFNS spectrum from ENDF/B-VII.1


II. Iron (4): IPPE Fe thckest sphere (Wall=34 cm) with 252Cf source, Neutron Leakage

0.03 0.3 30.01 0.1 1 10 20

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

JEFF-4.0T156

Fe - Perey56

Fe - Plompen

ENDF/B-VII.1 ENDF/B-VIII.0

C/E

Neutron Energy, MeV

10-7

10-5

10-3

10-1

101

JEFF-4.0T1

ENDF/B-VIII.0

ENDF/B-VII.1


Cf

n-L

eakage, 1/M

eV

/n-S

ourc

e

Observations:

JEFF-4.0T1 which adopts 56Fe(n,n’) either

Perey’71 (ORNL) or Plompen’13 (Geel)

are hardly distinguished by this benchmark

16

II. Iron, the second part:

- Secondary Gammas Validation and

- problem of L.Trykov’ Benchmark (ALARM-CF_FE-SHIELD-001)


L. Trykov et al. Atom Energy, 98(2005)50 (highlighting by couloirs is my):

“The experimental fluxes of >0.8 MeV neutrons are, for the most part, greater than

the computed values, and the fluxes of >0.35 MeV photons are also greater than

the computed values. The differences reach a factor of 2–3“:

What could be a reason for differences between L.Trykov experiment and calculations

by factor 2–3 for leaking Gammas, when for leaking neutrons - it is within 10% (???)


II. Iron (1): IPPE bare 252Cf source of Gammas vs. others

252Cf(s.f.) Gamma Multiplicity: L.Trykov (ICSBEP, …) vs. other known data

Observations on Gamma to Neutron Multiplicities ratio Mg/Mn for 252Cf(s.f.), Eγ > 100 keV:

- all known measurements (Cf fission chamber) and theory ≈ ( 8 - 10)/3.7676 = (2.1 - 2.7) – it’s truth !

- Trykov/Alarm-Cf-Fe-Schield-001 (encapsulated Cf) ≈ (18 - 20)/3.7676 = (4.7 - 5.3) – why ?

Bare 252Cf source encapsulated

in 3.2mm thick Fe + 2.3mm Cu:

- Trykov’85 (FEI-1730) = 3.8 g/n

then *3.7676 = 14.3 g/f

252Cf from ABBN’93 = 5.287 g/n

then *3.7676 = 19.9 g/f

Measurements which employ

252Cf fission chambers

and coincidence/counting FF

as well as Theory/Modelling 0 50 100 150 200 250 300 350 400 450

6

8

10

12

14

16

18

20

22

24

Bo

wm

an

'58

Sm

ith

'56

Va

lsk

ii'6

5S

ka

rsv

ag

'70

Ple

as

on

ton

'72

Ve

rbin

sk

i'7

3

Bru

ns

on

'82

Ch

yzh

'12

Ch

yzh

'14O

be

rste

dt'

15

Try

ko

v'8

5

Try

ko

v'7

6Talou'16 (< 3 ns)

Talou'16 (< 5 mks)

Litaize'14(< 10 ns)

Litaize'14(< 1 ms)

Litaize'14(< 1 ms)

ABBN93(total)

252Cf in capsule (Trykov) and ABBN-93)

gMultiplicity - Experiment

gMultiplicity - Theory

Gam

ma M

ultip

licty

, g/fis

sio

n

Gamma Energy Threshold, keV

252Cf(s.f.)

252Cf fission chamber Experim and Theory


Looking on these spectral data I did following:

1. from comparison of g-spectra (Eg = 0.5 - 3 MeV)

from L.Trykov’ bare Cf source

(corrected for attenuation in CuFe capsule)

and known PFGS

there was found Normaliz. Factor = 0.33 ± 20%

2. this Factor was applied to all Iron g-leakage

spectra measured by L.Trykov

(all slides on next slides show his data

after correction)

3. in MCNP input file the ABBN-93 g-spectrum

was replaced by one derived from independent

measured PFGS (< 1 MeV ) and

P.Talou’ modelling (> 1 MeV)

NB. Additional verifications of such correction

against independent measurements

of g-leakage from encapsulated Cf source

and Fe spheres are highly needed

II. Iron (1): IPPE bare 252Cf source of Gammas vs. others (cont.)

252Cf(s.f.) Gamma Energy Spectra:

Trykov (ICSBEP, …) vs. other known data

10-5

10-4

10-3

10-2

10-1

100

101

Sokol'91(p)

t(>0.1MeV) = 8.32 Pandit'10(p)

Smith'56(p)

Chyzh'13(p)

d(>0.25 s ) = 0.03

AB

BN

-93(t

)

Oberstedt'15(p)

JEFF-3.1.1(p)

Verbinski'73(p)

252Cf(s.f.) P(T)FGS

Gam

ma

Yie

ld, /

MeV

/s.f.

p(>0.1MeV) = 8.29+/-0.13

Ajitanand'69(p)

Trykov'79(t)

0.05 0.3 0.5 3 50.1 1 10

0.5

2.5

1.0

Sokol'91(p)

Kwan'12

Oberstedt'15

AB

BN

-93

(t)

Talo

u'1

7(p

)

Verbinskiy'73

Chyzh'13

Ra

tio

of Y

ield

to

8.2

9*

(Ma

xw

ells

)

Gamma Energy, MeV


Observations/Comments on:

1. Distortion of

Cf(s.f.) Gamma Spectrum in Fe/Cu capsule:

- below 7 MeV – attenuation by ≈ 10%

- above 7 MeV – increasing by up to 50%

due to (n,γ) on Fe and Cu

2. above 5 MeV the Cf(s.f.) γ-spectrum or

L.Trykov’ bare source data seem

are not so reliable - there we used

P.Talou‘ modelling since existing experimental

PFGS data are controversial

II. Iron (1): Results for IPPE bare 252Cf(s.f.) source: Gamma Spectrum

0.3 0.5 0.7 2 3 5 71 10

1.0

1.5

2.0

2.5

0.7

Distortion of Cf(s.f.) in Capsule(3.2mm Fe + 2.3mm Cu)

ENDF/B-VII.1

ENDF/B-VIII.0

C/E

Gamma Energy, MeV

10-5

10-4

10-3

10-2

10-1

100

101

Trykov

ENDF/B-VIII.0

ENDF/B-VII.1

252Cf(s.f.) source in Capsule

S

pe

ctr

um

, 1

/Me

V/n

-So

urc

e


Observations/Comments for Gamma leakage:

- No essential difference between

JEFF-4.0T1, ENDF/B-VII.1 or ENDF/B-VIII.0

- but JEFF-3.3T4 lacks of high energy gammas

from “thermal” (n,γ) on Fe-56 and Fe-54

as TENDL-2015/2017 do

II. Iron (1): Results for IPPE Fe sphere (Wall=29 cm) with 252Cf(s.f.): Gamma Leakage

0.5 0.7 2 3 5 71 100.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

JEFF-4T1 (Geel)

JEFF-3.3T4

ENDF/B-VII.1

ENDF/B-VIII

C/E

Leaking Gamma Energy, MeV

10-5

10-4

10-3

10-2

10-1

JEFF-3.3T4

direct from 252

Cf(s.f.) from54

Fe(n,)

from56

Fe(n,)

ENDF/B-VIII.0

ENDF/B-VII.1

Fe sphere (R/r=30/1cm) +252

Cf(s.f.)

le

aka

ge

, 1

/MeV

/n-S

ou

rce


Observations:

- Leaking Neutrons

0.1 - 1,5 MeV – both ENDF/B-VII.1 and

ENDF/B-VIII.0

are doing well

1.5 – 10 MeV – both lack removal XS

= 60 mb (ENDF/B-VII.1)

= 120 mb (ENDF/B-VIII.0

- Leaking Gammas

status is worse than for neutrons

II. Iron (1): Results for IPPE Fe sphere with 252Cf(s.f.), Neutron and Gamma Leakage vs. Wall Thickness

5 15 25 350 10 20 300.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

E = 1.1 - 5.2 MeV

E = 5.2 - 9.1 MeV

Leaking Gammas

E = 0.50 -1.1 MeV

--- ENDF/B-VII.1

--- ENDF/B-VIII.0

C/E

fo

r G

am

ma

s

Sphere Wall Thickness (R - r), cm

0.7

0.8

0.9

1.0

1.1

1.2

= 60 mbEn = 1.5 -10 MeV

En = 0.1 -1.5 MeV

--- ENDF/B-VIII.0

--- ENDF/B-VII.1

Fe spheres with 252

Cf(s.f.) source

C/E

fo

r N

eu

tro

ns

25

2C

f so

urc

e

Leaking Neutrons

= 120 mb


Observation for differential Gamma spectra:

- Differential Experimental γ-ray production spectra for Fe(n,xγ) reaction

at neutron energies typical for Cf(s.f.) transmission benchmark

do support the latest evaluations ENDF/B-VII.1, -VIII.0 and TENDL-2017

II. Iron (2): Independent DDX Measurements: Fe(n,x) γ-production spectra at ≈2 and ≈7 MeV

0.7 2 3 5 71 1010

-5

10-4

10-3

10-2

10-1

100

TENDl-2017

Drake'70

ENDF/B-VII.1

ENDF/B-VIII.0Chapman'76

Tanabe'94

En = 6.0 - 9.0 MeV

En = 2.0 - 2.5 MeV

natFe(n,xg)

d/d

E/d

, b/M

eV

/sr

Secondary Gamma Energy, MeV


II. Iron (3): Independent Experiment for Gamma leakage: VNIITF hemi- and spheres with D-T source

Observations for VNIITF’ Fe-sphere (Wall = 5 cm),

Leaking Gammas:

- whole sphere: ENDF/VII.1 looks almost perfect

and better than ENDF/B-VIII.0

- back hemi-sphere: both ENDF/VII.1 and

ENDF/B-VIII.0 overpredict by 30-150% 0.3 0.5 0.7 2 3 5 71

0.5

1.0

1.5

2.0

ENDF/B-VII.1

ENDF/B-VIII.0

Fe sphere:

Fe back-hemisphere:

ENDF/B-VII.1

ENDF/B-VIII.0

C/E


10-2

10-1

100

back-hemi-sphere

sphere

ENDF/B-VIII.0: sphere (); hemi-sphere (---)

ENDF/B-VII.1: sphere (): hemi-sphre (---)

Fe (R/r = 10/5 cm) + T(d,n)

le

akag

e,

1/M

eV

/n

24

II. Iron, the third part:

straightforward Monte-Carlo propagation of

Cross Section Uncertainties to

Clean Benchmark Response Covariances



Propagation Method implemented:

- 1 + 500 TENDL-2017 Bayesian Monte Carlo (BMC) random files for

each stable Fe isotope were provided by A.Koning and D.Rochman

- all were processed to ACE files by NJOY2012.82

- MCNP-6 simulation of neutron-gamma transport was re-run 1 + 500 times

each run (1.E+9 events to reduce statistical uncertainty well below 1%)

takes ≈ 3 min on 300 CPUs of KIT server

total Wall time consumed was ≈ 16 hours (i.e., it is a feasible task !)

- neutron and gamma leakage spectra have been read-in from 501 MCNP

output files to compute Covariancies Matrices

Example of Application:

- IPPE Iron sphere (R/r = 30/1cm, dia = 60 cm) with 252Cf(s.f.) source:

results - on the next 2 slides =>

II. Iron: from TENDL-2017 random files to Covariancies

for Energy Spectra of Neutrons and Gammas

leaking from Iron Spheres fed by 252Cf(s.f.) source


II. Iron: TENDL-2017 and Covariance propagation

for IPPE Fe sphere with Cf source: Neutron Leakage

0.03 0.3 30.01 0.1 1 10 20

0.7

0.8

0.9

1.0

1.1

1.2

Exp.Uncert.

TENDL-2017 and Uncert.

ENDF/B-VII.1

ENDF/B-VIII.0

C/E

n-L

eakag

e

Neutron Energy, MeV

10-5

10-3

10-1

101

TENDL-2017

ENDF/B-VII.1

ENDF/B-VIII.0

tot

(nat

Fe)

ENDF/B-VIII.0

ENDF/B-VII.1


Cf

n-L

eakage, 1/M

eV

/n-S

ourc

e

252Cf TFNS

0.2

1

10

50

← n Leakage and Uncertainties ≈ 2 - 20%

↓ n E-E Correlation Matrix: from strong

correlation (≈ 1) to anti-correlation (≈ -1) NB: Please note: original values are discrete points,

colours are continuum interpolation between them!

Findings for n-Leakage: - Uncertainties (≈ 2-20%) from TENDL-2017 are comparable with Experimental

- E-E correlations will play role below 100 keV, above 1 MeV – no correlation

0.03 0.3 30.01 0.1 1 10

0.03

0.3

3

0.01

0.1

1

10

Neutr

on E

nerg

y, M

eV

Neutron Energy, MeV

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

CorrelationNeutron Leakage from dia.60cm Fe-sphere with 252

Cf


II. Iron: TENDL-2017 and Covariance propagation

for IPPE Fe sphere with Cf source: Gamma Leakage

Findings for γ-Leakage: - Uncertainties (≈ 20-40%) from TENDL-2017 are comparable with Experimental

- TENDL-2017 underestimates Measurements: seem it lacks (n,γ) to ground state

- E-E correlations will play role in whole energy range 100 keV - 10 MeV

← γ Leakage and Uncertainties ≈ 20 - 40%

↓ γ E-E Correlation Matrix: practically

everywhere strong correlation (≈ 1)

NB: Please note: original values are discrete points,

colours are continuum interpolation between them!

0.5 0.7 2 3 5 71 100.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

TENDL-2017

with Uncert.

ENDF/B-VII.1

ENDF/B-VIII.0

C/E

and U

ncert

ain

ties


Exp.Uncert.

10-5

10-4

10-3

10-2

10-1

TENDL-2017

direct from 252

Cf(s.f.) from

54Fe(n,)

from56

Fe(n,)

ENDF/B-VIII.0

ENDF/B-VII.1

Fe sphere (R/r=30/1cm) +252

Cf(s.f.)

Leakage, 1/M

eV

/n-S

ourc

e

0.3 31 10

0.3

3

1

10

Gam

man E

nerg

y, M

eV

Gamma Energy, MeV

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

CorrelationGamma Leakage from dia.60cm Fe-sphere with

252Cf

28

III. Uranium:

Secondary Neutrons and Gammas Validation



III. Uranium (0) – overview of

Long Terma Int. Collaboartion, 18 - 21 Dec 2017, IAEA, Vienna

Experiments involved in the present analysis:

Lab

Years

Source:

<Energy>

No. of Spheres:

Wall Thickn., cm

Method Detector, Threshold Numerical

Data from

IPPE

90-91

D-T:

≈ 14 MeV

1 sphere:

Wall = 8.0 cm

Time of Flight

(TOF)

Neutrons:

Stilben > 0.5 MeV

S.Simakov et al.

AIP Conf.

769(2004)67

IPPE

1998

252Cf:

≈ 2.1 MeV

1 sphere:

Wall = 8.0 cm

Time of Flight

(TOF)

Neutrons:


S.Simakov et al.

AIP Conf.

769(2004)67

LLL

1976

D-T:

≈ 14 MeV

2 spheres:

Wall ≈ 3.2, 10.4

cm

Time of Flight

(TOF)

Neutrons:

NE-213 > 0.010 MeV

Li-glass 10keV - 1MeV

L.Hansen et al.

N.Tech 51(1980)70

C.Wong UCRL-

76263(1971)

LLL

1990

D-T:

≈ 14 MeV

1 sphere:

Wall ≈ 3.2 cm

Time of Flight

(TOF)

Neutrons:

NE-213 > 1.2 MeV

Gammas (as e-recoils):

NE-213 > 0.255 MeV

E.Goldberg et al.

NSE105(1990)319

VNIITF

1991

D-T:

≈ 14 MeV

1 sphere &

1 semi-sphere:

Wall = 5 cm

Time of Flight

(TOF)

Neutrons:

Stilben > 0.25 MeV

Gammas:

Stilben > 0.35 MeV

A.Saukov et al.

SINBAD Fusion:

NEA-1517/74

https://doi.org/10.1063/1.1944958

https://doi.org/10.1063/1.1944958

https://doi.org/10.1063/1.1944958

https://doi.org/10.1063/1.1944958





Observations:

Both Spherical Benchmark and SED (dσ/dE)

discover very similar trend:

- ENDF/B-VIII.0 underestimate 6-12 MeV interval

- ENDF/B-VII.1 looks better

III. U (1): Results – Neutron Leakage from IPPE Sphere with D-T source

0.5 2 31 105 200.6

0.7

0.8

0.9

1.0

1.1

1.2

ENDF/B-VIII.0

ENDF/B-VII.1

C/E

Neutron Energy, MeV

10-2

10-1

100

ENDF/B-VIII.0

ENDF/B-VII.1

U sphere (R/r = 12/4cm) + T(d,n)

Le

aka

ge

Sp

ectr

um

, 1

/Me

V

Experiment

Neutron Leakage at 14 MeV dσ/dE at 14 MeV

0.5 0.7 2 3 5 71 100.3 20

102

103

104

Simakov: (n,xn)-(n,nel)

Baba: (n,xn)

ENDF/B-VII.1

ENDF/B-VIII.0

natU(n,xn), E = 14.1 MeV

d(E

)/dE

, m

b/M

eV

Neutron Energy, MeV


III. U (2): Results - Neutron leakage from IPPE sphere

with 252Cf source

Observations:

- ENDF/B-VIII.0 and ENDF/B-VII.1

perform similarly well

10-4

10-3

10-2

10-1

100

0.2 0.5 20.1 1 105 180.7

0.8

0.9

1.0

1.1

1.2

1.3

ENDF/B-VIII.0

Experiment

ENDF/B-VII.1

C/E

Neutron Energy, MeV

ENDF/B-VIII.0

ENDF/B-VII.1

U sphere (R/r =12/4cm) +252

CfL

ea

ka

ge

Sp

ectr

um

, 1

/Me

V


III. U (3): Results - Gamma leakage from VNIITF hemi- and whole

spheres with D-T source

Observations:

- first results – no conclusion yet,

the experiment set-up models

has to be checked 0.3 0.5 0.7 2 3 5 71 10

0.5

1.0

1.5

2.0

ENDF/B-VII.1

ENDF/B-VIII.0

U sphere:

U back-hemisphere:

ENDF/B-VII.1

ENDF/B-VIII.0

C/E


10-3

10-2

10-1

U back-hemisphere

U sphere

ENDF/B-VIII.0

ENDF/B-VII.1

U (R/r=10/5 cm) + T(d,n)

le

aka

ge

, 1

/MeV

/DT

n


latest versions of ENDF, JEFF, TENDL evaluations for O, Fe and U

are currently under validation against known clean benchmarks

driven by neutron sources D-T, Cf and fast fields created at reactors

quantities analyzed/tested are neutron and gamma leakage, uncollided

neutron transmission

relevant experimental benchmarks are available in SINBAD, ICSBEP or

literature; however in some cases the complete information about

measurements are not presented in these resources that prevent

the correct modelling and concluding nowadays

the main purposes of this work was to check whether independent

benchmarks as well as differential XS indicate the similar trend for potential

corrections/adjustments

we generally saw this however excepting the case of L.Trykov’ measurements

of γ-leakage from Iron spheres with Cf-source: γ-spectra seem have to be

downscaled by factor ≈3 to reach agreement with other differential and

integral measurements (? - still need confirmation ! - additional independent

measurements will be decisive)

XS uncertainties (presented as TENDL-2017 randoms) were MC propagated

to Covariancies of n- and γ-leakage spectra from Fe sphere fed by 252Cf

General Summary (details for each Nuclei - in proper sections)

Documents

Benchmarking of the latest Neutron ... - Nuclear Energy Agency · Benchmarking of the latest Neutron and Gamma Transport Cross Sections for Oxygen, Iron and Uranium in clean Benchmarks