The Tank Data Matched to 5%. Its Error Bar Did Not Travel.

What this note is: a documentary finding that u_D on the compared result cannot be assembled from what travelled. It attempts Criterion 3 level (c) — uncertainty assembled per reported result, u_D on the compared quantity — for one registered cell of the 2025 K-Site laminar comparison, and reports that level (c) cannot be reached from what travelled. The compared statistic is tank pressure. The claim that governs is the version-of-record sentence in Cryogenics 152 (2025) 104210. Every factual claim about a named document is traceable to a passage in that document, listed in the sources, with the SHA-256 of the hashed capture used for the quote. (“Audit” here means documentary audit, as defined in Note 000.)

It is a documentary Note. It is frozen as v1.0 against LOCK-9. It does not close u_D, u_val, or a validation-uncertainty envelope. Those remain OPEN and are the finding. It is not the quantitative study Notes 000–003 committed. That study remains undone (a later note).

What it is not: a verdict on the laminar model, a claim that the comparison is wrong, or a number invented to fill the hole. Notes 000–003 are not rewritten here. If this Note disagrees with a prior published note, the prior note wins; the disagreement is named.

I. The study those notes committed remains undone

Note 000 stated the comparison-error framing this series uses, and the use that framing will not support without the experimental term:

A validation claim is a comparison between a simulated value S and a reference value D. The comparison error E = S − D supports a quantified accuracy claim only if the uncertainty attached to D is known; without it, E still permits coarser uses — detecting gross discrepancy, comparing trends — but not the percentage-accuracy statements this literature makes. … (Note 000 §I; SHA-256 d4866ff99238085c9ce8aac5ae03ca8b5b4a24051036c134feb8602f720a2ae1)

Note 001 wrote the series’ decomposition, and the independence caveat that blocks treating it as a lookup:

u_val² = u_num² + u_input² + u_D²

where u_num is the numerical (discretization and iteration) uncertainty of the simulation, u_input the uncertainty in parameters and boundary conditions, and u_D the experimental uncertainty of the referent. The quadrature form presumes independent contributions: where the same experimental data inform both the model’s boundary conditions and the compared result — the usual case in this literature — u_input and u_D are coupled and must be treated jointly. Assembling u_val is analysis, not lookup. (Note 001 §II; SHA-256 98dedbc4a0873806af9720d05d01877da9d3674681d874da191882dd0ebb91ce)

Note 000 §IX recorded that the documentary series had argued an engineering consequence and had not demonstrated it:

Demonstrating it would require re-running a published validation with the referent uncertainty included and showing the conclusion move. That is a tractable piece of work and it has not been done. … Until that study exists, the engineering consequence of everything above remains argued, not demonstrated. (Note 000 §IX; SHA-256 d4866ff99238085c9ce8aac5ae03ca8b5b4a24051036c134feb8602f720a2ae1)

The demonstration named there was a quantitative study: reconstruct a defensible result-level u_D for one referent, combine it with the numerical and input uncertainties of a published comparison, covariances included, and report whether that comparison’s validation conclusion moves. In either direction: if nothing moves, that result publishes too — but only after the combination is done. Note 001, Note 002 and Note 003 closed with the same commitment.

This note is not that study. It is a documentary finding that u_D cannot be assembled from what travelled. The quantitative study those notes committed remains undone. “If nothing moves, that result publishes too” only applies after the combination is done.

The six criteria remain those frozen in Note 000. This note does not add a seventh. Canonical wording, v1.0:

Are the raw data published, or only plotted in figures?
Are the boundary conditions complete — heat flux magnitude and distribution, fill level, initial stratification, ullage composition?
Are instrument accuracies stated, and are the derived quantities’ uncertainties propagated?
Is the geometry fully specified?
Is the fluid state specified — for hydrogen, including the ortho/para composition?
Does the quality information the original authors published travel with the data into current use?

Criterion 3 is an ordered scale, not a binary. Note 002 established that by reading the most thoroughly documented referent audited in this series, and Note 000 froze the three levels. The wording is canonical:

Criterion 3 — uncertainty maturity. Three separable levels, because a referent can be exemplary at one and absent at another: (a) uncertainty required in advance, in a test plan or acceptance criteria; (b) uncertainty characterized per instrument, reported for the sensors under the applicable conditions; (c) uncertainty assembled per reported result, propagated into intervals on the headline numbers a modeler would actually use. Level (c) is the one that yields u_D for the derived quantities a validation actually compares, and it is the level most often missing. For a directly measured validation quantity, level (b) may supply the dominant contribution — when the reported figure genuinely characterizes that measurement under the applicable test conditions — but an instrument accuracy specification is not automatically the complete experimental uncertainty: calibration, acquisition, installation, environment, repeatability and averaging can all add to it. (Note 000 §II)

This note attempts level (c) on the compared result of the registered cell. It reports that level (c) cannot be reached from what travelled.

II. The cell, and the claim that governs

Registered cell, frozen before any envelope was asked for: K-Site self-pressurization; fill 49 %; heat flux 3.5 W/m²; model laminar VOF with conjugate heat transfer; statistic object tank pressure. The version-of-record sentence that governs is joint 29 % and 49 %. The two fills are not averaged. The VoR overlays are Cryogenics Fig. 2a (29 %) and Fig. 4a (49 %) under the §5.1 “3.5 W/m2 heat flux” header (capture lines 529–530; discussion 531–536 and 545–550). Flux is in that section header, not in the Fig. 2 / Fig. 4 captions (lines 536, 550). SciTech Fig. 6a is a sibling window; its caption does not print fill or flux (txt line 431). The registered cell is the 49 % fill. This draft does not treat the unlabeled sibling as a joint 29 %-and-49 % window.

The claim that governs is Kartuzova, Kassemi and Hauser, Cryogenics 152 (2025) 104210. Quotes are taken from a ScienceDirect HTML page capture (no publisher PDF is in the repo; no printed-page concordance exists for any Cryogenics line cited). The hashed capture

D:\CMPS Elarion\sources\refs\KSITE_cryogenics2025_104210_fulltext.txt

SHA-256 831a5f076b016f8d7f89dc666fc9f77ecb96f013a814017fc75903799c300b02.

In the conclusion, immediately after a sentence that the effect of heat leaks through instrumentation penetrating the tank lid was investigated, the capture prints:

At the high heat flux value of 3.5 W/m2, the laminar model predicted tank pressure within 5 % of measured values for the 29 % and 49 % fill level cases, while the RANS model underpredicted pressures by up to 19 %. … (capture line 617; the line continues: IDDES “to 10 % or less”; at 83 % fill, all models “up to 25 %”.)

That is the sentence on which this note attempts Criterion 3 (c). The “within 5 %” is joint: 29 % and 49 %. It is not a 49 %-only claim. “within 5” occurs once in the hashed capture (line 617).

The article presents lid heat as a separate study (§5.1.5; capture lines 587–598) and names Fig. 10: “Effect of additional heat leak through the tank lid on predicted pressure rise rate at different fill levels with a base heat flux of 3.5 W/m2” (line 592). It reports that effect at line 620 separately from line 617, so line 617 is presented as the base-heat overlays Fig. 2a / Fig. 4a. Residual, scoped OPEN: §5.1.1 and §5.1.2 print no agreement percentage, and the run-provenance of the number is never printed in those subsections. Cite lines 587–598, 617, 620. It is not closed by averaging, and it is not closed by guessing. The unscoped reading “5 % may be base or lid” is not kept.

The same conclusion qualifies the laminar overlay and records that the data cannot be shared:

While the laminar VOF model showed the best agreement with experimental data for tank pressure and vapor temperatures, making it acceptable for certain engineering applications, its use is not appropriate for a turbulent situation from a rigorous scientific point of view. (capture line 621)

The authors do not have permission to share data. (capture lines 628–629)

Digitization uncertainty is a missing contribution if plots are used; this pass does not digitize.

The RANS “up to 19 %” in the same sentence is a different model. It is not this cell’s laminar claim, and it is not averaged with 5 %.

III. The sibling, mapped, not averaged

A sibling container of the same study does not govern.

Kartuzova, Kassemi and Hauser, AIAA SciTech Forum, January 2025, NTRS 20240016283. Two objects, not mixed:

PDF, manifest / NTRS file: SHA-256 2801c5695f2253d69d0f0be41b65fa9febf8f83ff01004b63642602f59fdef6d
Extracted text used for quotes: KSITE_scitech2025_full_20240016283.txt SHA-256 baa92a71944046351686730eb6faec24d3a600ee43e74d8bb709c5ab330e0f19

The quote object is the .txt. PDF printed-page concordance of the quoted laminar percentages (txt lines 519 and 534–535) was recovered from the hashed PDF SHA-256 2801c5695f2253d69d0f0be41b65fa9febf8f83ff01004b63642602f59fdef6d as printed page 14 (file page 14). Other SciTech PDF-page mappings remain UNVERIFIED.

The SciTech paper selects the 49 % fill and two heat fluxes (txt lines 19–20, 70–71). It does not make the joint 29 % and 49 % claim. In the conclusion the capture prints:

Predicted tank pressures are within 1% of the measured ones for the 3.5 W/m2 heat flux case and within 3.5% for the 2 W/m2 heat flux case. (txt lines 534–535)

The immediately preceding sentence names the laminar VOF with conjugate heat transfer model (txt line 533). The “within 1%” is therefore a 49 %-only sibling of the VoR’s joint “within 5 %”, at the same 3.5 W/m². Same study, two containers. They are mapped. They are not averaged.

Census of printed laminar percentages in the sibling .txt, not averaged with one another or with the VoR 5 %:

Body, 2 W/m²: the laminar model predictions are “within 3%” of the measured values (txt line 519).
Conclusion: “within 1%” at 3.5 W/m² and “within 3.5%” at 2 W/m² (txt lines 534–535).
Conclusion, no fill or flux named: “within 3% of the measured values with the laminar model” (txt lines 561–563).

The 3 % (txt line 519) and the 3.5 % (txt lines 534–535) are both printed for 2 W/m². They are kept as a census. They are not called a contradiction. Compatibility is not invented. They are not a third sibling of the VoR 5 %. None of these percentages is averaged with any other.

SciTech Figure 6a is a sibling window — laminar versus turbulence, tank pressure. Its caption does not print fill or flux (txt line 431). The VoR names Fig. 4 / Fig. 4a for the 49 % overlay (capture lines 545–550). The sibling window, as the SciTech capture prints it:

The effects of turbulence modeling on the tank pressure and interfacial phase change rate are presented in Figures 6a and 6b (txt lines 422–423)

Figure 6: The effect of turbulence modeling on tank pressure (a) and phase change rate (b) (txt line 431)

Figure 5a is a pressure overlay. SciTech txt lines 394–396 present tank pressures (with vapor temperatures and phase-change rates) for the uniform “thin” wall, the variable tank wall, and the no-wall models, “compared with the experimental data in Fig. 5.” Line 413 captions Figure 5 as “the effect of modeling tank wall thickness/geometry on tank pressure (a), vapor temperatures (b), and phase change rate (c).” It is the wall-thickness / geometry study’s pressure overlay, not the laminar-versus-turbulence window (that window is Fig. 6a). This draft does not digitize SciTech Figure 6a, and it does not digitize VoR Fig. 4 / Fig. 4a.

IV. What would have been compared, and what was not computed

The comparison error this series uses is E = S − D. For this cell, E(t) = p_S(t) − p_D(t). Three relative forms were frozen in LOCK (SHA-256 0dd5aeaa0dbcd43c6ed78d2d66b7249a15f6d07fbb1c988bdf2df7e92dec2eeb) and remain defined in LOCK-2 (SHA-256 bf73939368a5b930decf00e01d4514455f858b01b4703a949aab1d1cd05813e9) as the primary set before any of them was evaluated:

E_end = |E(t_end)| / p_D(t_end)
E_rise = |E(t_end)| / [p_D(t_end) − p_D(t_0)]
E_max = max_t |E(t)| / p_D(t)

Window, as LOCK-2 restates it: VoR overlays are Cryogenics Fig. 2a (29 %) and Fig. 4a (49 %) under the §5.1 “3.5 W/m2 heat flux” header (capture lines 529–530, 531–536, 545–550); flux is in the header, not the caption. SciTech Fig. 6a is a sibling window whose caption does not print fill or flux (txt line 431). t = 0 included. TM-105411 records that at the 49 percent fill level, much of the increase in pressure results from the initial transient period (extract lines 270–272; SHA-256 1211ee29fada9b728a9e5c896233776d9f0cb5933a389bf00e71168b70f899f4). If only one of the three forms were ≤ the VoR’s 5 %, that fact would be reported. None of the three is computed in this draft. That is disclosed. It is not a silent skip of a closed gate.

The question those three numbers would answer is LOCK-2 L60: does the observed E that the VoR calls “within 5 %” support 5 % predictive accuracy on those cells, or only agreement finer than the envelope can discriminate? That question stays on the board. It is not answered this pass. Note 000 §I is a bound: without u_D, percentage-accuracy statements are not supported. That bound does not close L60 by computing E. The licensed object of this file is the documentary finding that u_D on the compared tank-pressure result cannot be assembled from what travelled. E_end, E_rise, and E_max, frozen in the LOCK, were not computed in this draft. That is disclosed.

V. D0 stays OPEN

Printed “±X” is not a standard uncertainty.

TM-105411 states instrument performance as “accurate to” and, once, “maximum error”. The hashed extract of that report (SHA-256 1211ee29fada9b728a9e5c896233776d9f0cb5933a389bf00e71168b70f899f4) does not print k or a coverage probability for those figures. Note 001 restates the same paragraph from page images, not from the extract, because the ± symbols do not survive extraction from this scan (Note 001 §I). The sibling extract of TM-103804 (SHA-256 a9e5f78e2ad8a8c6650476316e04058ddd8d0f079760fff6db00f1ca37489570) preserves most ± signs and the same “accurate to” / “maximum error” wording (extract lines 147–156). Neither primary states a coverage factor.

Note 003’s printed sentence on this point is binding. The words that follow are Note 003 L61, not a paraphrase (SHA-256 b44694ec6fc4098a7414c4f35b0158088be25b7ca7560b77fa0f33ae9bf5f266):

The instrument accuracies are stated. Ullage pressure P4 is an MKS Baratron 0–666 kPa absolute transducer “with an accuracy of ±0.02 percent”; the pressure-control loop used a second Baratron (0–133 kPa, ±0.02 percent) plus a 1-torr-head differential transducer at ±0.04 percent (P4 on p. 6; the control-loop pair on p. 60). … [Scoped at v1.0, Erratum 26: the specification does not state percent-of-reading versus percent-of-full-scale — at 133 kPa, 0.02 % of reading would be ±0.027 kPa, larger than the control band — an ambiguity that must be resolved before this figure can serve as a u_D contribution.]

The MHTB case is not this cell. The formalization “do not RSS as 1σ by habit” is LOCK/PLAN language (LOCK L94; BUDGET L89 already records that Note 003 does not contain the words “RSS” or “1σ”). An unspecified % is not a coverage factor.

D0 therefore remains OPEN. The printed half-widths recovered below are carried as half-widths of unknown coverage. They are not RSS’d. They are not treated as standard uncertainties. ASME V&V 20 is the conceptual source of the E = S − D framing in Note 000 and Note 001; the standard itself was not read for this series (Note 001, Sources B). An unread standard does not close D0. Notes 000–003 did not license an unassembled envelope as the demonstration they named.

VI. What travelled: instrument statements, not assembled u_D

Note 001 recorded that four of K-Site’s five published instrument-accuracy categories reappear in the 2025 full-length papers, essentially verbatim, and that the fifth — the boil-off flowmeters — does not (Note 001 §III). A search of the two hashed 2025 extracts used here, for SCMH, flowmeter, flow meter, 0.030, 0.089, in situ, in-situ, 0.1 K, 120 to 150, and single vertical, returned no matches. (SciTech txt line 498 contains “0.1” as a vapor volume fraction, not a temperature accuracy.)

The four restored statements are Criterion 3 level (b) material. They contribute to u_D. They are not assembled u_D. “Calibration not priced” is scoped: TM-105411 extract lines 171–174 already report an in-situ calibration that increases liquid-vapor temperature accuracy to 0.1 K. Other calibration, acquisition, installation, environment, repeatability and averaging terms remain unpriced (Note 000 §II, Criterion 3). Note 001’s wording is the one this note uses:

The instrument accuracy contributes to u_D — the uncertainty of the measurement as a measurement. (Note 001 §II)

Tank pressure (D1)

Compared statistic: tank pressure, measured in the vent line.

Note 001 page-image restatement: tank pressure ± 0.01 kPa.

TM-105411 extract line 175: “Tank pressure measurements are accurate to rt 0.01 kPa.”

SciTech txt lines 110–111: “Tank pressure is measured by pressure transducers located in the vent line with accuracy of ± 0.01 kPa.”

Cryogenics capture line 119: “Tank pressure is measured by pressure transducers located in the vent line with an accuracy of ±0.01 kPa.”

Travels. Bound-only. “Accurate to”; coverage unknown; not 1σ; not assembled u_D.

Fluid and wall temperature (D1b, and the uncalibrated pair)

Note 001 page-image restatement: liquid and vapor temperatures ± 0.3 K, improved to ± 0.1 K by in-situ calibration against known saturation conditions; wall temperatures ± 0.6 K.

TM-105411 extract lines 170–174: “accurate to k 0.3 K” (fluid), “accurate to rt 0.6 K” (wall), in situ calibration “to rf: 0.1 K” against known saturation conditions.

SciTech txt lines 108–109 restore uncalibrated fluid ± 0.3 K and wall ± 0.6 K. Cryogenics capture line 119 the same. Neither hashed 2025 extract restores the in-situ ± 0.1 K.

The in-situ adjustment is against known saturation conditions, which are a pressure–temperature pair. That is a possible common-mode path between D1b and D1. Covariance and independence are both unreported. Possible common-mode (e.g. saturation used to set initial pressure) remains OPEN. This draft does not say flatly that D1b is coupled to D1. Wall ± 0.6 K is not the compared statistic. The compared statistic is tank pressure.

D1b: bound in the 1992 primary; missing from the 2025 restorations; possible common-mode with D1 unreported; OPEN.

Fill (D3)

Note 001 page-image restatement: capacitance level probe ± 1.9 cm, a maximum of ± 1.5 % fill at the 50 % level.

TM-105411 extract lines 176–180: “accurate to rt 1.9 cm, translating to a maximum error of f 1.5 percent fill at the 50 percent fill level (by volume).” The page break “2” sits between the centimetre clause and “translating…”.

SciTech txt lines 106–107: “capacitance probe with accuracy of ± 1.9 cm, which is equivalent to ± 1.5% fill at the 50% fill level by volume.”

Cryogenics capture line 119: “capacitance probe with an accuracy of ± 1.9 cm, which is equivalent to ±1.5 % fill by volume at the 50 % fill level.”

The TM’s own words are maximum error. The registered fill is 49 %, next to a 50 % statement. Coverage is not stated. The compared statistic here is pressure, not fill; D3 may contribute to u_input (ullage volume / fill as a boundary condition) and is not, in this cell, a u_D on the pressure reading. Bound-only.

Boil-off flowmeters (D4) — did not travel

Note 001 page-image restatement: boil-off flow ± 0.030 and ± 0.089 SCMH on the two meters.

TM-105411 extract lines 181–184: “Boil-off flow measurements are accurate to f 0.030 and f 0.089 standard m 3 h (SCMH) for the 2.83 and 8.49 SCMH meters, respectively.”

TM-103804 (sibling 83 % / low-heat-flux TM; SHA-256 a9e5f78e2ad8a8c6650476316e04058ddd8d0f079760fff6db00f1ca37489570) prints the same two meters, 2.83 and 8.49 SCMH (extract lines 154–156), and boil-off rates 0.45, 2.7, and 4.7 SCMH corresponding to 0.35, 2.0, and 3.5 W/m² (extract lines 258–264). A 4.7 SCMH rate cannot use the 2.83 SCMH meter. The meter is not explicitly named; it is uniquely implied as 8.49 SCMH in that TM. Whether that meter ran the registered 49 % TM-105411 cell stays OPEN. Those sibling boil-off tests are at 95 % fill (TM-103804 extract line 259), so the implication does not automatically travel to the 49 % cell.

Stochl and Knoll, on the same facility’s boil-off metering, print a different statement. Hashed extract SHA-256 8929612aef9c48eda77ee6cb7cdb8ab2b7cef9e65925544b997ed9f00469d966, lines 287–293:

The evaporation rate from the test tank was metered by one of a series of four volume flowmeters. These meters had full scale ranges of 10, 100, 300, and 1500 standard cubic feet of gaseous hydrogen per hour with an uncertainty of ±0.5 percent of full scale.

The series’ A1dep record (SHA-256 da7c77833cf8cd1630c327c2c5b02873558f4dd619ebf4cfaf31a2abd51baba3; en dataset copy, lines 79–82) prints 1.06 % FS (2.83 SCMH meter) and 1.05 % FS (8.49 SCMH meter) beside S&K’s ±0.5 % FS on the 100 / 300 SCFH meters. Both A1dep figures are printed here. They are not collapsed. Note 000 §IV.6 already named the factor-of-two companion discrepancy; it did not print these two percentages. Neither document explains the difference. The conflict is a discrete ambiguity, not an average.

D4 may contribute to u_input (heat-input magnitude). It is not the compared pressure reading. It does not reappear in the 2025 extracts searched above. Status: OPEN.

The 3.5 W/m² figure itself is not an instrument ±. TM-105411 extract lines 200–203 (sentence starts at line 200): “The average wall heat flux values of 2.0 and 3.5 W/m2 were obtained from shroud operating temperatures of 294 and 350 K, respectively3.”

VII. Heat-input distribution is not the 1991 allocation

Criterion 2 asks for heat flux magnitude and distribution. Stochl and Knoll locate 13 to 17 percent of total heat in discrete penetrations — 12 fiberglass support struts, tank plumbing, and instrumentation lines — not in the insulation (txt lines 47–51 and 605–609; SHA-256 8929612aef9c48eda77ee6cb7cdb8ab2b7cef9e65925544b997ed9f00469d966). TM-105411’s own parenthesis: the uniform heat flux assumption is reasonable, since about 85 percent of the heat leak is uniformly distributed and the remaining heat leak paths are located in both the upper and lower halves of the tank (extract lines 694–698). That is Erratum 25’s surviving route. It still travels as context.

The 2025 recovery is not that allocation. Cryogenics capture line 516: heat fluxes of 2.0 and 3.5 W/m2 “are uniformly distributed along the tank wall.” Capture line 588: additional heat from Stochl and Knoll for plumbing, ducts and electrical wiring “was applied uniformly across the top surface of the tank lid.” SciTech txt lines 268–269: heat fluxes, “uniformly distributed along tank wall, of 2 and 3.5 W/m2 are considered.” Note 001 §III already recorded the partial return: the Stochl–Knoll dependency is cited and lid heat is studied; the original spatial information has not travelled whole.

Lid-only 2025 heat is not the original 13–17 % allocation. Lid attribution is scoped as in §II (§5.1.5 / Fig. 10 / lines 587–598, 617, 620): L617 is presented as the base-heat overlays; residual OPEN is run-provenance, not an unscoped “base or lid.”

VIII. Terms that are not u_D

D7 — energy residual and the single-axis rake

TM-105411, page-image-checked in Note 001, reports energy-balance errors from 120 to 150 percent, increasing with fill level, and questions the uniform radial temperature assumption because an exact energy balance would require a liquid level error of 2 to 3.5 times the measurement uncertainty (extract lines 625–644; Note 001 §I). Liquid and vapor temperatures were limited to a single vertical axis near the tank centerline; the one-dimensional analysis assumes constant radial temperatures (extract lines 606–611).

Note 001’s distinction stands:

The interpretive caveat is something else, and arguably worse: it says that the quantity measured — temperature at sensor locations on a single vertical rake — is not the quantity a bulk-energy comparison needs. That is not measurement uncertainty; it is a mismatch between the measured variable and the validation variable, and no error bar on the sensor repairs it. (Note 001 §II)

D7 is out of u_D. Attribution of the residual is OPEN. If unsupported, the envelope on a bulk-energy reading of those temperatures is unbounded. The compared statistic in this cell is tank pressure, not bulk liquid temperature. The caveat still did not travel: the same search of the 2025 extracts that found no flowmeter accuracies found no 120-to-150 % residual and no single-vertical-axis limitation. Note 001 §III recorded that absence. This draft does not stuff D7 into a sensor bar.

D2 — TM-105411 names Fig. 2; OPEN is whether 2025 digitized it

TM-105411 extract lines 228–230 (SHA-256 1211ee29fada9b728a9e5c896233776d9f0cb5933a389bf00e71168b70f899f4) print:

Pressure rise measurements at fill levels of 29,49, and 83 percent with heat flux equal to 2.0 and 3.5 W/m2 are shown in Fig. 2.

The same extract labels panel (a) “q = 2.0 Wlm2” (line 252) and panel (b) “q = 3.5 Wlm2” (line 258); caption “Figure 2. - Effect of Fill Level on Pressure Rise Rate.” (lines 259–260). For the 3.5 W/m² cell the panel is Fig. 2(b). “D2 unnamed” is not a reason Criterion 3 (c) fails. OPEN only whether the 2025 overlays digitize that TM figure. The two 2025 txts do not print “Van Dresar Fig. 2.”

SciTech validates “against tank self-pressurization experiments with a 4.89 m3 liquid hydrogen storage tank, as reported by Hasan et al. in the 1991 paper1” and then selects “Test cases with 49% fill level” (txt lines 68–71). Reference [1] in that paper is Hasan, Lin and Van Dresar, NASA TM-103804, 1991 (txt lines 568–569). TM-103804 is the 83–84 % campaign (Note 000, registry A1). The 49 % data live in TM-105411 (Note 000, registry A2; SciTech’s own VoR sibling cites Van Dresar, Lin and Hasan, NASA TM-105411, as [1] — Cryogenics capture lines 632–634). That sibling IC cite is not a reason the TM source figure of p(t) is unnamed.

The VoR overlays it plots are Fig. 4 / Fig. 4a (49 % fill tank-pressure overlay under §5.1.2; capture lines 545–550) and Fig. 2a (29 % under §5.1.1; capture lines 531–536), both under the §5.1 3.5 W/m² header (lines 529–530). SciTech Fig. 6a is a sibling window; its caption does not print fill or flux (txt line 431). This pass does not digitize TM Fig. 2 / Fig. 2(b), and it does not digitize the named VoR overlay.

The VoR names Van Dresar et al. [1] Fig. 4 and Fig. 5 for the initial conditions it applies (capture line 518). SciTech applies “measured initial ullage and tank wall temperature profiles reported by Hasan et al.1” (txt line 272) — the TM-103804 cite. The sibling’s IC source remains that cite; the VoR’s IC figures are the ones just named.

TM-105411 extract lines 487–490 print that at the 29 percent fill level “the test was terminated (due to tank pressure safety limits) before steady temperatures were clearly established,” and that “it may be argued that the data in Table 1 for 29 percent fill is not quasi-steady.” That attaches to Table 1 rates. It does not automatically attach to the p(t) overlay. It still belongs on a note whose VoR sentence is joint 29 % + 49 %.

u_num — not reported in the two audited 2025 captures

SciTech’s grid-independence study “was conducted with the IDDES turbulence model” (txt lines 321–323). Medium and fine meshes result in similar tank pressures; the coarse mesh under-predicts (txt lines 327–328). No numeric Δp and no GCI are recovered from that paragraph. Cryogenics capture line 527: mesh sizes “based on the results of a previously completed grid independence study, as reported by the authors in an earlier work [15]” — that earlier work is the SciTech IDDES study. Laminar u_num is not reported in the two audited 2025 captures. It is missing, not zero.

Covariance

TM-105411 lists distinct instruments (extract lines 158–184): a capacitance probe for fill, silicon diodes for liquid, vapor and wall temperature, vent-line pressure transducers for tank pressure, and thermal-dispersion flowmeters for boil-off. This draft does not say that the same measurements supply fill, flux and compared pressure. Covariance and independence are both unreported. Possible common-mode (e.g. saturation used to set initial pressure) remains OPEN. u_input and u_D are not added in quadrature in this draft, because they are not assembled, and because quadrature is not licensed.

IX. Why level (c) cannot be reached

No u_D on tank pressure for this cell is claimed. No u_val. No envelope number. The reasons are grouped. They are not ranked so that one of them can be repaired in isolation to produce a number the others still forbid. D4 stays u_input, not a u_D-on-pressure blocker: the compared claim is tank pressure, not boil-off (VoR capture line 617). Those facts stay in §VI.

(A) Experimental-pressure u_D blockers. D0 is OPEN. Printed ± is not 1σ. TM-105411 never states k or coverage. RSS-as-1σ is forbidden by LOCK/PLAN (BUDGET L89) and by D0; that formalization is not a Note 003 quotation. Instrument ± contributes to u_D; it is not assembled u_D. Criterion 3 level (b) is not level (c). In-situ fluid-temperature calibration to 0.1 K is already in TM-105411 (extract lines 171–174); other calibration, installation and environment terms remain unpriced.

(B) Comparison / metric reconstruction. The TM source figure of the experimental p(t) is named: TM-105411 Fig. 2, panel (b) at 3.5 W/m² (extract lines 228–230, 258–260). OPEN only whether 2025 digitized that figure. The 2025 txts do not print “Van Dresar Fig. 2.” SciTech cites the 83 % campaign TM for a 49 % comparison (sibling IC). The VoR overlays are Fig. 2a (29 %) and Fig. 4a (49 %) under the §5.1 3.5 W/m² header; SciTech Fig. 6a is sibling. Not digitized. TM-105411 lines 487–490: the 29 % test was terminated before steady temperatures were clearly established; Table 1 at 29 % may not be quasi-steady. That attaches to Table 1 rates, not automatically to the p(t) overlay; it still belongs here because the VoR sentence is joint 29 % + 49 %. Lid attribution is scoped as in §II (Fig. 10; lines 587–598, 617, 620): L617 is presented as the base-heat overlays; residual OPEN is that §5.1.1 / §5.1.2 never print the run-provenance of the number. The SciTech 1 % is not averaged with the VoR 5 %. The 3 % / 3.5 % census is kept; it is not a contradiction; compatibility is not invented.

(C) u_val blockers. Laminar u_num is not reported in the two audited 2025 captures (the grid study is IDDES). u_input includes D4 (flowmeters; meter identity for 3.5 W/m² OPEN) and heat-input magnitude / distribution. Covariance and independence are both unreported. Possible common-mode (e.g. saturation used to set initial pressure) remains OPEN. Quadrature is not licensed.

(D) Contextual, outside u_D. D7 is out of u_D. The 120–150 % residual and the single-axis caveat require attribution first. Sensor bars do not repair a variable mismatch. Lid-only 2025 heat is not the original 13–17 % allocation in both halves.

Criterion 3 (c) is the level this note attempted. It cannot be reached from what travelled. Level (b) is present / bound-only for tank pressure: ±0.01 kPa travelled (TM-105411 extract line 175; SciTech txt lines 110–111; Cryogenics capture line 119). Coverage unknown; not 1σ; not assembled u_D. Flowmeters stay u_input / the fifth category that did not travel; they are not a reason to call (b) partial. D0 remains OPEN. In-situ T calibration is priced in the TM and not restored in 2025; other calibration, installation and environment terms remain unpriced. That is a position on the scale. It is not a yes, and it is not a closed u_D.

X. The finding, not upgraded

The licensed object of this file is the documentary finding that u_D on the 2025 K-Site compared tank-pressure result cannot be assembled from what travelled. Note 000 §I is a bound: without u_D, percentage-accuracy statements are not supported. That bound does not close LOCK-2 L60 (E versus envelope). E_end, E_rise, and E_max were not computed. LOCK-9 binds that object. LOCK-8 is the parent freeze and is not rewritten here.

That hole is the result. This file is the documentary precursor. The committed quantitative study remains undone.

It does not mean the laminar model is wrong. It does not mean the laminar model is right. Note 000 already separated those: a model can be excellent while its validation is unquantifiable, if the referent’s uncertainty was never carried along. This draft does not reopen that separation. It is a documentary finding that u_D cannot be assembled from what travelled. The sentence remains a published comparison. “If nothing moves, that result publishes too” only applies after the combination is done.

XI. What this draft does not do

It does not compute E_end, E_rise, or E_max. It does not digitize SciTech Fig. 6a, VoR Fig. 2a / Fig. 4a, or TM-105411 Fig. 2 / Fig. 2(b). It does not invent σ. It does not RSS. It does not average 1 % with 5 %. It does not treat TM-105411 “accurate to” or “maximum error” as a coverage factor. It does not stuff D7 into u_D. It does not sell lid-only 2025 heat as the original 13–17 % spatial allocation. It does not rewrite Notes 000–003. Errata append later; this draft is not rewritten in place after the fact.

It does not start a new research programme. The six criteria are unchanged. The registered cell is unchanged.

Budget incompleteness, named. The inventory’s recovered-objects table hashes Notes 000, 001 and 003 as published in publications\ and does not hash Note 002. Note 002 as published in publications\002-built-to-travel.md hashes to SHA-256 e2227f063b411700bbcbfc3f128dda6f2b4e1846f106706d747d36479e96f0b6. This note binds that file. The omitted row is not a licence to flatten Criterion 3, which Note 002 is the source of. The published note wins.

A1dep copy split, named. The en dataset copy cited above hashes to da7c77833cf8cd1630c327c2c5b02873558f4dd619ebf4cfaf31a2abd51baba3. The working copy at referents\A1dep_STOCHL_KNOLL_boundary_condition.md hashes differently (f7e5019b612106187d328457bc0f4e24e02c0c7564ad3c98914293c0f2337370). This note cites the dataset copy the inventory named. It does not silently prefer the other.

XII. Limitations

Absence in the published modelling papers is not absence in practice. Note 000 §IX (SHA-256 d4866ff99238085c9ce8aac5ae03ca8b5b4a24051036c134feb8602f720a2ae1, line 257) prints the heading “Absence in conference papers is not absence in practice” and the body “the search was for statements in published modelling papers.” This Note names that divergence and uses the broader body wording because the governing source is a journal article. The search was for statements in published modelling papers. Programme-level verification and validation documentation, credibility assessments under NASA-STD-7009, internal review records and reviewer correspondence were not examined, and any of them could carry the context that the papers do not. The VoR’s “The authors do not have permission to share data” (capture line 629) shows material exists outside what travelled.
One auditor, working with AI assistance. No independent human reviewer has yet examined this draft. Under the project’s own authority scheme this is exploratory (A0) output, and it is labeled as such. The LOCK’s validation gate — two internal audits and one external Cursor audit — has not closed on this prose. The inventory is marked UNVERIFIED. This draft inherits that mark.
Competing interest, declared. ElarionX CPMS develops and evaluates engineering models for cryogenic propulsion systems, and the machine-checked referent registry this series argues for — Phase 0 — is the project’s own planned product. The series’ conclusions therefore favor something its author intends to build, and a reader should weight them knowing that. The mitigation offered is not neutrality but verifiability: every claim carries the provenance needed to check it without trusting the author. (Note 000 §IX; still declared.)
Text extraction is OCR-limited for the scanned 1990s documents. Quantitative claims about TM-105411’s instrument paragraph follow Note 001’s page-image restatement; the extract is quoted as extracted, with its OCR tokens (k, rt, rf:, f) visible. TM-103804’s sibling extract is the cleaner OCR of the same instrument paragraph and is not the 49 % primary.
The VoR quote object is a ScienceDirect page capture. There is no Cryogenics PDF in the repo and no printed-page concordance for any Cryogenics line cited.
SciTech quotes are from the .txt. Printed-page concordance of txt lines 519 and 534–535 was recovered from hashed PDF 2801c5695f2253d69d0f0be41b65fa9febf8f83ff01004b63642602f59fdef6d as printed page 14. Other PDF-page mappings remain UNVERIFIED.
This draft does not read ASME V&V 20 in the original. D0 stays OPEN rather than being closed by an unread standard. Notes 000–003 did not license an unassembled envelope as the demonstration they named.
The quantity Notes 000–003 named did not close, and the study that would close it remains undone. Assembled u_D on this cell cannot be produced from what travelled. Note 000 §I is a bound: without u_D, percentage-accuracy statements are not supported. That bound does not close LOCK-2 L60. E_end, E_rise, and E_max were not computed. What a later pass would need, at shortest, is listed in the inventory and is not performed here: page-image confirmation of TM-105411’s instrument paragraph and energy-balance passage (Note 001 §I: pages 2–3 and page 7); SciTech PDF-page concordance beyond printed page 14 of the laminar percentages; whether the 2025 overlays digitize TM-105411 Fig. 2 / Fig. 2(b) (the VoR overlays are Fig. 2a / Fig. 4a under §5.1); whether the 8.49 SCMH meter implied in TM-103804 ran the registered 49 % TM-105411 cell; laminar u_num as reported or an explicit keep-unpublished from the two 2025 captures; then the combination Notes 000–003 committed — u_D with u_num and u_input, covariances included — and a report of whether the published conclusion moves. None of those is a closed envelope in disguise. “If nothing moves, that result publishes too” only applies after that combination is done.

On verification. Quotations from hashed captures were checked against those captures at the line numbers given. Anything found wrong after this draft produces a new version carrying a visible erratum — the superseded draft remains preserved — including when the correction weakens a conclusion already stated. This file is not rewritten in place.

Sources

A. Documents obtained and read for this note

Kartuzova, O., Kassemi, M., Hauser, D. — CFD validation of k-site tank self-pressurization under varying fill levels and heat fluxes with different turbulence models. Cryogenics 152 (2025) 104210. doi:10.1016/j.cryogenics.2025.104210. Quote object: sources\refs\KSITE_cryogenics2025_104210_fulltext.txt, SHA-256 831a5f076b016f8d7f89dc666fc9f77ecb96f013a814017fc75903799c300b02. The file is a ScienceDirect <main> scrape (capture L1–L3), not a publisher PDF. No printed-page concordance exists for any Cryogenics line cited. Governing “within 5 %” sentence: capture line 617.
Kartuzova, O., Kassemi, M., Hauser, D. — Validation of a Two-Phase CFD Model for Predicting Tank Self-Pressurization in the Ground-Based K-Site Experiment. AIAA SciTech Forum, January 2025. NTRS 20240016283. Quote object: sources\refs\KSITE_scitech2025_full_20240016283.txt, SHA-256 baa92a71944046351686730eb6faec24d3a600ee43e74d8bb709c5ab330e0f19. Sibling “within 1%” sentence: txt lines 534–535; hashed PDF 2801c5695f2253d69d0f0be41b65fa9febf8f83ff01004b63642602f59fdef6d printed page 14. PDF is not the quote object.
Van Dresar, N. T., Lin, C.-S., Hasan, M. M. — Self-Pressurization of a Flightweight Liquid Hydrogen Tank: Effects of Fill Level at Low Wall Heat Flux. NASA TM-105411 / AIAA-92-0818. NTRS 19920009200. PDF SHA-256 91775f3185f280847c5fc27cc086517cc7d8fd8db32d5bfd5deda0a79f109e78. Quote object for the extract: sources\refs\KSITE_filllevel_19920009200.txt, SHA-256 1211ee29fada9b728a9e5c896233776d9f0cb5933a389bf00e71168b70f899f4. Instrument-paragraph ± values as used here follow Note 001’s page-image restatement, not this extract.
Hasan, M. M., Lin, C. S., Van Dresar, N. T. — Self-Pressurization of a Flightweight Liquid Hydrogen Storage Tank Subjected to Low Heat Flux. NASA TM-103804, 1991. NTRS 19910011011. PDF SHA-256 b08568a71c7678dfd63c93ac93c35b2737df48cedd36f4147fd6c009ce351513. Quote object: sources\refs\KSITE_lowheatflux_19910011011.txt, SHA-256 a9e5f78e2ad8a8c6650476316e04058ddd8d0f079760fff6db00f1ca37489570. Sibling extract of the instrument paragraph; not the 49 % primary.
Stochl, R. J., Knoll, R. H. — Thermal Performance of a Liquid Hydrogen Tank Multilayer Insulation System at Warm Boundary Temperatures of 630, 530 and 152 °R. NASA TM-104476 / AIAA 91-2400. NTRS 19910015845. PDF SHA-256 87250c7124e5642e3774bbfb4747cf34a7e235a1451aa7061287444d0d2dea99. Quote object: sources\refs\STOCHL_KNOLL_MLI_19910015845.txt, SHA-256 8929612aef9c48eda77ee6cb7cdb8ab2b7cef9e65925544b997ed9f00469d966.
ElarionX CPMS, referent record A1dep — Stochl & Knoll: the K-Site boundary condition, en dataset edition. SHA-256 da7c77833cf8cd1630c327c2c5b02873558f4dd619ebf4cfaf31a2abd51baba3. Source of the two uncollapsed A1 flowmeter % FS figures. Companion dataset: doi:10.5281/zenodo.21895803.

B. Prior notes of this series — bound as published

ElarionX CPMS, Technical Note 000 — The Cryogenic Referent Registry (doi:10.5281/zenodo.21895568). publications\000-the-referent-registry.md, SHA-256 d4866ff99238085c9ce8aac5ae03ca8b5b4a24051036c134feb8602f720a2ae1. Canonical six criteria; Criterion 3 scale; competing-interest declaration; A0 mark; the commitment this note does not discharge.
ElarionX CPMS, Technical Note 001 — The Accuracies Travelled Back. The Warning Did Not. K-Site, 1992–2025 (doi:10.5281/zenodo.21895605). publications\001-the-vanishing-referent.md, SHA-256 98dedbc4a0873806af9720d05d01877da9d3674681d874da191882dd0ebb91ce. Page-image instrument restatement; u_val sentence; travel of four accuracies; D7 out of u_D.
ElarionX CPMS, Technical Note 002 — 286 Pages of Candor, and Still No Assembled u_D (doi:10.5281/zenodo.21895647). publications\002-built-to-travel.md, SHA-256 e2227f063b411700bbcbfc3f128dda6f2b4e1846f106706d747d36479e96f0b6. Criterion 3 as an ordered scale. Not hashed in the inventory’s recovered-objects table; bound from the published file.
ElarionX CPMS, Technical Note 003 — The MHTB Primary Is a TVS Performance Study, Not a Self-Pressurization Characterization (doi:10.5281/zenodo.21895743). publications\003-the-referent-beneath.md, SHA-256 b44694ec6fc4098a7414c4f35b0158088be25b7ca7560b77fa0f33ae9bf5f266. L61 as quoted in §V. “Do not RSS as 1σ by habit” is LOCK/PLAN, not a Note 003 sentence (BUDGET L89).

C. Framework, not read in the original

ASME V&V 20-2009 (R2021), Standard for Verification and Validation in Computational Fluid Dynamics and Heat Transfer. Conceptual source of the comparison-error framing, as already scoped in Note 000 and Note 001. Not read for this draft. An unread standard does not close D0, and Notes 000–003 did not license an unassembled envelope as the demonstration they named.

ElarionX CPMS develops and evaluates engineering models for cryogenic propulsion systems, and publishes what it finds — including when the finding is about its own work. Competing interest: declared in Note 000 §IX and restated in §XII. This draft is exploratory (A0): it states that u_D on the compared tank-pressure result cannot be assembled from what travelled. It is not a validated result. Corrections and correspondence: Luis.emc2@elarionx.com