Read-only demonstration. Four sanitised or hypothetical engineering projects. Each states its own validation status; nothing shown is engineer approved.

MECHANICAL — CFD AND ENERGY · CASE STUDY

Data Centre CFD Analysis — Simulations

NOT ENGINEERING VALIDATED — native CFD converged; validation gates partly failed

Simulations

Final CFD runs

Result	Origin	Software · date	Verdict	Status	Source · evidence hash	Engineering approval
DC00B-R6 baseline CFD (final)CFD runFinal baseline: DC00B-R5 field resumed with the same boundaries; all 7 residuals at the 1e-5 termination residual after 2,599 outer iterations; rack inlets 19.09-21.62 °C (max 21.62 °C), within 0.02 K of R5. Convergence only - the study is NOT ENGINEERING VALIDATED.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	PASS	exploratory	cfd-DC00-20261024-r2/cfd/DC00B-R6/HX_DC00_M1.logsha256 b03ee3d19604…	OUTSTANDING — no engineer review or approval recorded
DC00O-R1 optimised CFD (final)CFD runFinal optimised design (supply 1.05 x IT airflow, two hot-aisle extracts): all 7 residuals at 1e-5 after 653 outer iterations. Convergence only - not engineering validated.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	PASS	exploratory	cfd-DC00-20261024-r2/cfd/DC00O-R1/HX_DC00_MO.logsha256 5f26864180c4…	OUTSTANDING — no engineer review or approval recorded
DC00O-R1b rack-inlet monitor read-outCFD runConverged DC00O field resumed 1 iteration with 11 cell monitors: rack inlets 19.02-21.49 °C (max 21.49 °C); represented rack heat 9.85-9.97 kW per 10 kW rack. Values are the converged state; no stability window.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	PASS	exploratory	cfd-DC00-20261024-r2/cfd/DC00O-R1b/HX_DC00_MO.logsha256 6c564f5ac841…	OUTSTANDING — no engineer review or approval recorded

Mesh study

Result	Origin	Software · date	Verdict	Status	Source · evidence hash	Engineering approval
DC00O-C1 coarse-mesh CFDMesh studyCoarse grid 0.133 m (44 x 35 x 22 = 33,880 cells, aspect 2:1) converged after 517 iterations; third grid for the GCI study.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	PASS	exploratory	cfd-DC00-20261024-r2/cfd/DC00O-C1/HX_DC00_MOC.logsha256 a6ed0cc09169…	OUTSTANDING — no engineer review or approval recorded
DC00O-F1 refined-mesh CFDMesh studyRefined grid 0.075 m (80 x 63 x 39 = 196,560 cells) converged after 830 iterations; rack inlets within +0.08 K of the base grid; max inlet 21.53 °C.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	PASS	exploratory	cfd-DC00-20261024-r2/cfd/DC00O-F1/HX_DC00_MOF.logsha256 86bc4b430b60…	OUTSTANDING — no engineer review or approval recorded
Mesh sensitivity, optimised design (two-grid and three-grid GCI)Mesh studyTwo-grid comparison (86,400 vs 196,560 cells) within the ASSUMED tolerances (\|dT\| <= 0.2 K, speed <= 10 %). Three-grid GCI (Celik et al. 2008): low and mid inlets monotonic with GCI <= 0.1 K; top inlets not monotonic (spread <= 0.08 K). The gate check is FAIL. The baseline was solved on one grid only.	Hexanta offline	Hexanta CFD engine (hexanta_api.cfd) branch feat/cfd-dc01-engine2026-10-24	PARTIAL	exploratory	cfd-DC00-20261024-r2/study-manifest.study.mdsha256 3a1534d76fb3…	OUTSTANDING — no engineer review or approval recorded

Mesh study only partly passed

Low and mid inlets are monotonic over three grids (GCI <= 0.1 K); the top inlets are not monotonic (spread <= 0.08 K) and the gate check is FAIL. The baseline was solved on one grid only.

Superseded and aborted runs

Result	Origin	Software · date	Verdict	Status	Source · evidence hash	Engineering approval
DC00B-R5 baseline CFD (32 °C exhaust)CFD runNOT CONVERGED to the strict target (t 1.73e-5, e 1.65e-5 vs 1e-5) at 2,500 iterations; monitors steady. Source field of the baseline figures; numbers superseded by R6.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	PARTIAL	superseded	cfd-DC00-20261024-r2/cfd/DC00B-R5/HX_DC00_M1.logsha256 413274e10830…	OUTSTANDING — no engineer review or approval recorded
DC00B-R4 baseline CFD (resumed)CFD runConverged, but the edited 32 °C exhaust was not applied (a resumed run keeps the old boundaries); superseded.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	FAIL	superseded	cfd-DC00-20261024-r2/cfd/DC00B-R4/HX_DC00_M1.logsha256 4c6fda4725f9…	OUTSTANDING — no engineer review or approval recorded
DC00B-R3 baseline CFD (regenerated)CFD runNOT CONVERGED at 1,500 iterations; airflow orientation verified; represented rack heat -16 % at 30 °C exhaust, which led to the 32 °C exhaust iteration.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	FAIL	superseded	cfd-DC00-20261024-r2/cfd/DC00B-R3/HX_DC00_M1.logsha256 eb1e153f6a32…	OUTSTANDING — no engineer review or approval recorded
DC00B-R2 baseline CFDCFD runNOT CONVERGED (iteration limit; resumed the aborted R1 field; 30 °C exhaust). Its early figures are not used.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	FAIL	superseded	cfd-DC00-20261024-r2/cfd/DC00B-R2/HX_DC00_M1.logsha256 cae7587d6d5f…	OUTSTANDING — no engineer review or approval recorded
DC00B-R1 baseline CFDCFD runABORTED after about 270 iterations by an automation fault (the watcher locked the solver log). Led to the rule: never read solver files during a solve.	Native tool output	IESVE (MicroFlo CFD) 2026 (trial licence, internal evaluation)2026-10-24	FAIL	exploratory	raw solver output store (not in the sealed evidence folder)	OUTSTANDING — no engineer review or approval recorded

**Convergence history (Hexanta plot)**Largest of the 7 residuals per outer iteration, read from the solver residual logs of DC00B-R6 (2,599 iterations) and DC00O-R1 (653). Convergence PASS does not make the study valid.Hexanta plot of evidence values · source: cfd-DC00-20261024-r2/cfd/DC00B-R6/HX_DC00_M1.log; cfd-DC00-20261024-r2/cfd/DC00O-R1/HX_DC00_MO.log

**Solver monitor - DC00B-R6 set-up and residual history**Native solver monitor dialog: residual history of u, v, w, t, k, e, m, solver settings (k-epsilon, inner iterations 3, false time step 0.0987, relaxation 1, termination residual 1e-05) and the cell-monitor table.Screenshot of the native tool (static image) · source: cfd-DC00-20261024-r2/figures/m45-DC00B-R6-monitor.png (cropped)

**Grid statistics - refined mesh DC00O-F1**Native grid statistics dialog: 80 x 63 x 39 = 196,560 cells, maximum cell aspect ratio 1:1 (memory figures are those of the test machine).Screenshot of the native tool (static image) · source: cfd-DC00-20261024-r2/figures/m45-DC00O-F1-gridstats.png (cropped)

Sanitised derivative of preserved evidence (cfd-DC00-20261024-r2, cfd-DC01-N1-DSM-20261024). Nothing was re-simulated. Screenshot crops are static images of the native viewer; plots and the report contain only values read from the evidence. · Pack cfd-data-centre-r2 (built 2026-10-25); every file is checked against its SHA-256 before it is shown.