REVIEW 3 major objections 5 minor 14 references
Aluminium fast neutron leakage spectrum validation
T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read With a $^{252}$Cf source in an aluminum block, this paper validates ENDF/B-VIII.0, JEFF-3.3, and JENDL-5 fast neutron transport within two sigma except in narrow bands; gold capture is overpredicted at every depth.
desk verdict A useful new aluminum benchmark that needs a fix: the gold capture anomaly is probably a self-shielding artifact and the data aren't public. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing apparatus is a 50.18 cm × 50.12 cm × 54.42 cm aluminum block assembled from eleven plates, with a $^{252}$Cf spontaneous-fission source — emission around $2\times10^8$ neutrons per second, calibrated by a manganese-sulfate bath — at its geometric center, activation foils between plates at depths from 1.18 to 20.95 cm, and two proton-recoil spectrometers outside: a hydrogen-filled proportional detector covering 0.1–1.3 MeV and a stilbene scintillator covering 0.9–13 MeV, with a shielded-cone measurement subtracted as background. The argument-carrying comparison is MCNP6.2 with only the aluminum transport cross sections swapped between ENDF/B-VIII.0, JEFF-3.3, and JENDL-5, while all activation cross sections come from the IRDFF-II dosimetry file; the simulated fluxes are Gaussian-broadened to the measured detector resolution, and every conclusion rests on the $\mathrm{C/E}-1$ ratios inside a two-$\sigma$ uncertainty budget that covers source emission, geometry, detector efficiency and calibration, and counting statistics.
What would settle it
Irradiate paired bare and cadmium-covered (thermal-neutron-blocking) gold foils at the same depths, 1.18 to 16.01 cm, in a replicate block. If the covered foils match simulation while the bare foils keep the growing 15–33 percent overprediction, the bias lies in the predicted thermal and epithermal flux; if both stay discrepant, the $^{197}$Au$(n,\gamma)^{198}$Au cross section in IRDFF-II is the likely cause — either outcome sharpens or overturns the paper's attribution.
Extended reading notes
Core claim
On the paper's own terms: the measured fast neutron leakage spectrum of the $^{252}$Cf-driven aluminum block is matched by MCNP6.2 simulations with all three libraries within two $\sigma$ across most of the 0.1–13 MeV range, the exceptions being JENDL-5 over 1.8–3.4 MeV and ENDF/B-VIII.0, with JEFF-3.3 behaving nearly identically, over 1.06–1.3 MeV. The activation set corroborates the libraries: the threshold reactions $^{58}$Ni$(n,p)^{58}$Co, $^{92}$Mo$(n,p)^{92m}$Nb, natTi$(n,X)^{46}$Sc, natTi$(n,X)^{47}$Sc, and $^{115}$In$(n,n')^{115m}$In, together with the capture reactions $^{63}$Cu$(n,\gamma)^{64}$Cu and $^{181}$Ta$(n,\gamma)^{182}$Ta, fall within two $\sigma$ of experiment, with JENDL-5 slightly closest overall and the indium comparison worsening with aluminum thickness for every library. The single systematic discrepancy is gold capture: the calculated rate exceeds the measured one by about 15 to 33 percent in $\mathrm{C/E}-1$ units, growing with depth and independent of the transport library — a result the paper classifies as unsatisfactory and leaves for further refinement.
Load-bearing premise
The comparison assumes the IRDFF-II activation cross sections are accurate; if the $^{197}$Au$(n,\gamma)^{198}$Au cross section in that dosimetry file is biased, the paper's conclusion that the transport libraries mispredict the gold reaction rate would be wrong.
Editorial extensions
If this is right
- Aluminum transport calculations in reactor and shielding applications can treat ENDF/B-VIII.0 and JEFF-3.3 as nearly interchangeable for fast fluxes, with a shared caveat in the 1.06–1.3 MeV band where both deviate by more than one sigma.
- JENDL-5 gives the closest activation-rate predictions in this benchmark, but its 1.8–3.4 MeV flux deviation marks an aluminum evaluation feature that a future revision would need to correct.
- The gold capture overprediction, growing from about 15 to 33 percent with depth, becomes a quantitative constraint on the next round of gold dosimetry and aluminum transport evaluations alike.
- The measured spectra and reaction rates in this simple slab geometry form a benchmark that future aluminum evaluations can be checked against without a new experiment, and the same plate-stack design carries over to other structural materials.
Reading between the lines
- Editorial inference: the depth trend in the gold discrepancy — $\mathrm{C/E}-1$ climbing from about 14.7 percent at 1.18 cm to about 33 percent at 16.01 cm — is the sharpest clue the paper leaves unanalyzed; a bias that grows with depth points to the accumulated thermal and epithermal flux rather than a single resonance, and paired bare and cadmium-covered gold foils would separate those contribut
- Editorial inference: because every library shares the same IRDFF-II dosimetry file, the paper's verdict on aluminum transport is conditional on that file's gold evaluation; recomputing the same measured rates with an updated $^{197}$Au$(n,\gamma)$ cross section would settle whether the transport libraries or the dosimetry standard carries the bias.
- Editorial inference: the indium agreement tends to worsen with thickness (JENDL-5 moves from $\mathrm{C/E}-1$ of $-2.3\%$ at 1.18 cm to $-10.2\%$ at 20.95 cm), hinting that the fast-neutron population at depth is slightly underpredicted; adding reactions tied to the $^{27}$Al$(n,\alpha)$ or $^{27}$Al$(n,2n)$ channels at the deepest foils would localize a possible inelastic-source error.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript reports integral validation measurements for aluminum neutron transport libraries. A 252Cf source was placed at the center of an aluminum block, and the authors measured the fast neutron leakage spectrum with hydrogen-filled proportional and stilbene detectors (0.1–1.3 MeV and 0.9–13 MeV, respectively) and activation reaction rates for 115In(n,n'), 197Au(n,γ), 63Cu(n,γ), 181Ta(n,γ), 58Ni(n,p), 92Mo(n,p), and natTi(n,x) at depths from 1.18 to 25.91 cm. MCNP6.2 simulations with ENDF/B-VIII.0, JEFF-3.3, and JENDL-5 transport data, using IRDFF-II activation cross sections, are compared with measurements via C/E ratios. The main claims are that all libraries agree with the measured fast leakage spectrum within two sigma except JENDL-5 in 1.8–3.4 MeV and ENDF/B-VIII.0/JEFF-3.3 in 1.06–1.3 MeV, and that activation results are reasonable except for the 197Au(n,γ) reaction, which is discrepant for all libraries and depths.
Significance. The experiment is a useful integral benchmark: the geometry is simple, the 252Cf source emission was calibrated at NPL, multiple activation reactions cover both threshold and capture responses, and three modern evaluated libraries are compared. The paper reports forward calculations with no fitted parameters, and the C/E presentation is honest, including regions that exceed one sigma. The unresolved gold discrepancy and the modeling question around it limit the current strength of the validation claim, but the dataset itself is valuable for nuclear data validation. The manuscript would be substantially strengthened by including the raw spectra, input files, and a quantitative uncertainty budget for the leakage spectrum.
major comments (3)
- [Monte Carlo transport simulations / Table 4] The conclusion that 197Au(n,γ) is the only inconsistent reaction is not supported because the simulation method as described may bias the gold capture rate. The paper states: 'The reaction rates were calculated using the neutron flux in the activation detector multiplied by the activation cross section of the dosimeter.' That is, an F4 track-length flux tally in the foil region was multiplied by the IRDFF-II cross section, without an explicit transport-based reaction-rate estimator that accounts for the foil's self-shielding and flux depression. Gold has a large thermal and epithermal capture cross section, and the foils (0.05–0.06 mm thick) are not infinitesimally thin. The systematic increase of C/E from 14.4% at 1.18 cm to 33.7% at 16.01 cm is qualitatively consistent with a growing self-shielding/flux-perturbation bias as the spectrum softens. The authors should demonstrate, e.g., by comparing with an explicit FM4 reaction-rate tally or a detailed foil model with correct geometry and material, that the F4-flux-times-cross-section approach is unbiased for gold. Without this, the gold discrepancy cannot be attributed to the transport libraries or to IRDFF-II.
- [Results / Figure 8] The headline two-sigma statement is not quantitatively supported. The paper presents a one-sigma uncertainty band in Figure 8, but it does not itemize the components (detector efficiency, energy calibration, source emission, positioning, statistics) or state how the band was propagated from the measurements. Since the central validation claim depends on whether C/E departs by more than two sigma, the authors should report the numerical uncertainty budget for the spectrum, or at least reference the specific equations or tables in reference [8] that define it. Without this, the deviation of JENDL-5 in the 1.8–3.4 MeV region cannot be distinguished from an underestimated systematic uncertainty.
- [Conclusions / Table 4] The manuscript does not attempt to diagnose the gold discrepancy, yet the conclusion states that all libraries generally offer good agreement except for 197Au(n,γ). This is too strong given that the origin is undetermined. At minimum, the authors should compare the C/E for gold with the C/E for the other capture reactions (63Cu and 181Ta) at the same depths, and decompose the gold response into thermal, epithermal, and resonance contributions using the calculated flux spectra and the IRDFF-II group cross sections. A decomposition would clarify whether the issue is the gold resonance integral, the thermal component, or a fluence/self-shielding effect. The paper's own text acknowledges that the gold result is 'unsatisfactory' regardless of library, but it stops short of the diagnostic analysis needed to support a library-validation claim.
minor comments (5)
- [Abstract and general text] There are several typographical errors, including 'achived' in the abstract, 'worser' in the Results section, and a broken cross-reference 'Eq. Chyba! Nenalezen zdroj odkazů.' immediately before Eq. (1). These should be corrected in a revised manuscript.
- [Monte Carlo transport simulations] The description of the resolution broadening is incomplete: the text mentions 'Gaussian broadening' but does not specify the kernel parameters or whether the same broadening was applied to both the HPD and stilbene calculations. Please state the energy-resolution function used for each detector.
- [Table 4] The reaction label '92Mo(n,p)92*Nb' should be corrected to '92Mo(n,p)92mNb' or the asterisk notation should be defined, since the text and Table 2 use '92Mo(n,p)92mNb'.
- [Figure 7 and Figure 8] The captions and legends should be more explicit: Figure 7 should state what the curves and bands represent (e.g., Gaussian-broadened calculations, measured points, one-sigma uncertainty), and Figure 8 should identify the energy regions where the C/E exceeds the plotted one-sigma band rather than leaving the reader to estimate them from the graph.
- [Data availability] The paper does not state whether the raw experimental spectra, activation data, or MCNP input files are available. Given that the paper's validation claims rest on reproducibility, an explicit data availability statement would be valuable.
Circularity Check
No circularity: forward transport validation with independent libraries and no fitted parameters.
full rationale
The paper performs forward MCNP6.2 calculations using independently evaluated transport libraries (ENDF/B-VIII.0, JEFF-3.3, JENDL-5), the standard IRDFF-II 252Cf source spectrum, and IRDFF-II activation cross sections. No parameters are fitted to the measured reaction rates or leakage spectra, and no predicted quantity is defined in terms of the measured data. Activation reaction rates are computed as the F4 track-length flux in the foil region multiplied by the IRDFF-II cross section, while the experimental rate q from Eq. (1) is obtained independently from gamma spectrometry. The conclusions are C/E comparisons, so the derivation chain is input libraries to transport to predicted rates and spectra to comparison. The self-citations [1,3,8] describe measurement and unfolding methodology, but the validation conclusion does not rest on an unverified uniqueness theorem, a fitted parameter, or an ansatz smuggled in by citation. The 197Au(n,γ)198Au disagreement is reported as an anomaly, not as a prediction forced by construction; whether it originates in foil self-shielding or cross-section bias is a correctness concern, not circularity. No self-definitional, fitted-input, or self-citation-load-bearing step is present.
Assumptions & free parameters
assumptions (5)
- domain assumption The 252Cf(sf) neutron spectrum from IRDFF-II is accurate.
- domain assumption The IRDFF-II activation cross sections for the seven dosimetry reactions are accurate.
- domain assumption The MCNP6.2 geometry model matches the experimental arrangement.
- domain assumption The HPGe detector efficiency calibration via MCNP is correct.
- domain assumption The aluminum composition from XRF is correct within stated uncertainties.
Cite this review
Pith. "Pith review of Aluminium fast neutron leakage spectrum validation." pith.science (2026). https://pith.science/paper/FXJR53QE
@misc{pith2026250619861,
author = {Pith},
title = {Pith review of: Aluminium fast neutron leakage spectrum validation},
year = {2026},
howpublished = {\url{https://pith.science/paper/FXJR53QE}},
note = {Machine review of arXiv:2506.19861}
}
read the original abstract
Aluminum is a crucial material in the nuclear industry, valued for its ability to perform reliably over time. The manuscript focuses on validating aluminum neutron transport libraries. The validation was conducted by activating samples in different positions, and measuring the fast neutron spectrum in the energy range of 0.1 MeV to 1.3 MeV by means of a proportional detector filled with hydrogen and, in the energy range of 1-12 MeV using a scintillation stilbene detector. Validation experiments were performed on the aluminium block with a central hole which was assembled from smaller aluminum plates. The dosimetric reactions studied for validation purposes were 58Ni(n,p)58Co, 197Au(n,g)198Au, 63Cu(n,g)64Cu, 181Ta(n,g)182Ta, 92Mo(n,p)92mNb, natTi(n,X)46Sc, and natTi(n,X)47Sc. All experimental data were compared to MCNP6.2 simulations using the ENDF/B-VIII.0, JEFF-3.3, and JENDL-5 neutron transport libraries. Activation cross sections were taken from IRDFF-II library. Concerning activation reactions results, unsatisfactory results are achieved for 197Au(n,g)198Au reaction regardless of thickness and library. All other results are reasonable regardless on library. JEFF-3.3 and ENDF/B-VIII.0 fast neutron flux calculations are similar.
Figures
Reference graph
Works this paper leans on
- [8]
- [1]
-
[2]
M. Schulc, M. Košťál, E. Novák, J. Šimon, Copper neutron transport libraries validation by means of a 252Cf standard neutron source, Nuclear Engineering and Technology, https://doi.org/10.1016/j.net.2021.04.029
-
[3]
B. J ánský, E. Novák, Z. Turzík, J. Kyncl, F. Cvachovec, L.A. Trykov and V.S. Volkov, Neutron and gamma spectra measurements and calculations in benchmark spherical iron assemblies with 252Cf neutron source in the centre, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment , Volume...
work page 2002
-
[4]
IRDFF-II: An Updated Neutron Metrology Library,
A. Trkov, P.J. Griffin, S.P. Simakov, L.R. Greenwood, K.I. Zolotarev, R. Capote et al., “IRDFF-II: An Updated Neutron Metrology Library,” Nuclear Data Sheets 163, 1-108, (2020)
work page 2020
- [5]
-
[6]
Z. Matěj, Digitalization of Spectrometric System for Mixed Field of Radiation, LAP LAMBERT Academic Publishing (2014)
work page 2014
-
[7]
Z. Matěj, F. Mravec, A. Jančar, M. Kostal, F. Kucera, F. Cvachovec, V. Prenosil, Z. Kopecky, J. Culen, O. Pecak , Comparison of neutron -Gamma separation qualities of various organic scintillation materials and liquid scintillator LSB-200, J. Nucl. Eng. Radiat. Sci., 7 (2) (2021), p. 24502
work page 2021
Show all 14 references
-
[9]
C. J. Werner, J. S. Bull, C. J. Solomon., F. B. Brown, G. W. McKinney, M. E. Rising et al.,
-
[10]
ENDF/B -VIII.0: The 8th Major Release of the Nuclear Reaction Data Library with CIELO-project Cross Sections, New Standards and Thermal Scattering Data
D.A. Brown, M.B. Chadwick, R. Capote et al, "ENDF/B -VIII.0: The 8th Major Release of the Nuclear Reaction Data Library with CIELO-project Cross Sections, New Standards and Thermal Scattering Data", Nucl. Data Sheets, 148 (2018), pp. 1–142
2018
-
[11]
A. J. M. Plompen, O. Cabellos, C. De Saint Jean, M. Fleming, A. Algora, M. Angelone et al., The joint evaluated fission and fusion nuclear data library, JEFF-3.3. The European Physical Journal A, 56(7) (2020)
2020
-
[12]
Status of JENDL
O. Iwamoto, N. Iwamoto, K. Shibata, A. Ichihara, S. Kunieda, F. Minato, and S. Nakayama, "Status of JENDL", EPJ Web of Conferences, 239, 09002, 1-6 (2020)
2020
-
[13]
Status of the Evaluation of the Neutron Spectrum of 252Cf(sf)
W. Mannhart, “Status of the Evaluation of the Neutron Spectrum of 252Cf(sf)”, IAEA Technical Report INDC(NDS) -0540, IAEA, Vienna (2008). Presentation available at www - nds.iaea.org/standards-cm-oct-2008/6.PDF
2008
-
[2018]
LA -UR-18-20808)
MCNP Version 6.2 Release Notes (No. LA -UR-18-20808). Los Alamos, NM (United States). https://doi.org/10.2172/1419730
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.