100 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
Omar Sobh	af19860c04	rtx-cfd k-ε GPU: memcpy_dtoh argument order	2026-09-16 07:16:52 -05:00
Omar Sobh	5e402918bb	rtx-cfd legacy GPU modules: cudarc 0.18 API — unsafe launches, the stream accessor, memcpy_htod argument order; the renamed k-ε constants and ::standard(); get_module clones; CfdConfig test literals take ..Default	2026-09-16 07:16:39 -05:00
Omar Sobh	de435eb91c	rtx-cfd: the legacy GPU modules back in the cuda build (repair in progress)	2026-09-16 07:15:24 -05:00
Omar Sobh	3fd4fb7b8c	rtx-fsi: the interface noise probe reads RTX_FSI2_MG_DEVICE (the device V-cycle's regime gate)	2026-09-16 01:24:14 -05:00
Omar SobhandClaude Fable 5.1	fc556f8a88	PERF-2 P3-ii: the device V-cycle as the CG's preconditioner — poisson/device.rs (one CUDA runtime per process, persistent per-operator buffers, the mg_vcycle.cu kernels at K = 1; upload r, run the V-cycle, download z; the f64 CG unchanged), MultigridParameters::device, Prepared holds the device hierarchy, the CG driver destructures the prepared operator instead of cloning it; EmbeddedPisoSolver::set_poisson_device, overset pass-through, harness knob RTX_FSI2O_MG_DEVICE=1; export_levels factored out; the quarantine's dangling cfg attribute fixed ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-16 01:16:55 -05:00
Omar Sobh	3b7f8fb362	PERF-2 P3 bench: the coarsest level uses the CPU's red-black symmetric ordering (a nearly singular coarsest system settles on a smoother-dependent constant, which the prolongation doubled up every level: the 19 % disagreement)	2026-09-16 01:10:12 -05:00
Omar Sobh	06a60b5fed	PERF-2 P3 bench: per-level comparison of the device V-cycle's work vectors against the CPU's (vcycle_f32_work) to locate the disagreement	2026-09-16 01:08:31 -05:00
Omar Sobh	cb1647c0ef	rtx-cfd: the legacy GPU solver modules (piso_gpu, simple_gpu, d2q9_gpu, d3q19_gpu, smagorinsky_gpu) kept out of the cuda build — broken since the initial commit; the PERF-2 benchmark uses cudarc directly	2026-09-16 01:05:39 -05:00
Omar Sobh	956e9d38e2	rtx-cfd: the cuda feature builds again — the duplicate CudaKernelManager::stream removed, the legacy k-epsilon GPU model (broken since the initial commit) kept out of the build	2026-09-16 01:03:33 -05:00
Omar SobhandClaude Fable 5.1	0dc95a8de5	PERF-2 P3: the batched device V-cycle microbenchmark — mg_vcycle.cu (masked variable-coefficient five-point red-black half-sweeps, residual, CSR restriction in a fixed order, prolongation, coarsest sweeps; [K][n] layout, blockIdx.y = march), LevelExport/export_hierarchy and vcycle_f32_reference on the CPU side, and the ignored cuda-feature test that checks the K = 1 device V-cycle against the CPU f32 reference and times K = 1 / 8 / 16 per march ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-16 01:03:02 -05:00
Omar Sobh	53078c2aa7	rtx-cfd: get_features compiles under --features cuda (the vector must be mutable)	2026-09-16 01:00:34 -05:00
Omar SobhandClaude Fable 5.1	8c130c55c8	PERF-2 P2-c: RTX_FSI2O_INNER_STOP knob — the background Poisson's inner stop as a multiple of the Schwarz tolerance (default 0.1; the band-gated ladder on the CG iteration count) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-16 00:50:52 -05:00
Omar Sobh	822041fb3f	PERF-2 P2: the parallel maps run only on levels with ≥ 8,192 active cells and in tasks of ≥ 2,048 cells (a fork-join over a small level costs more than the map); values unchanged	2026-09-16 00:10:42 -05:00
Omar SobhandClaude Fable 5.1	7f144e0005	PERF-2 P2: the red-black colour maps, the residual and the V-cycle's per-level maps on rayon threads (MultigridParameters::threads; RTX_THREADS in the harness; rtx_cfd::configure_threads) — each colour's values computed from the unchanged other colour into a scratch and written back, the L1 sum in the serial order: bit-identical to the serial red-black (pin: 4 right-hand sides at 4 threads); the coarsest level stays serial; no effect on the lexicographic regime ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 23:59:29 -05:00
Omar Sobh	364766a54a	rtx-fsi: the interface noise probe reads RTX_FSI2_MG_RB itself (its solver is built outside the march's knob site; the first red-black probe run was vacuous)	2026-09-15 23:53:57 -05:00
Omar Sobh	f9d38bee55	rtx-cfd/rtx-fsi: every EmbeddedParameters literal carries the smoother field (all test targets of both crates compile)	2026-09-15 23:50:19 -05:00
Omar Sobh	43d5abde8d	rtx-fsi harness: the embedded harness's EmbeddedParameters carries the smoother field (the previous commit did not compile the rtx-fsi tests)	2026-09-15 23:49:31 -05:00
Omar SobhandClaude Fable 5.1	cb5771fa71	PERF-2: red-black symmetric Gauss–Seidel smoother behind a knob (MgSmoother::RedBlack; default lexicographic = the recorded regime, bit-identical) — red/black cell lists per level, the symmetric pair red,black,black,red, the operator cache keyed on the smoother; EmbeddedParameters::poisson_smoother, harness knobs RTX_FSI2O_MG_RB (overset) and RTX_FSI2_MG_RB (embedded, the noise probe); pin: solves the masked problem to the same stop, agrees with lexicographic to 2e-12, cached = uncached bit for bit ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 23:48:55 -05:00
Omar Sobh	26904e37d8	PERF-2 P1.b: the polyline index prunes from the start with an upper bound from each chunk's first vertex (never the answer; the strict-minimum scan rule is unchanged), so a query near the polygon's far end no longer scans every chunk; pin unchanged, 0 mismatches	2026-09-15 23:35:40 -05:00
Omar SobhandClaude Fable 5.1	e0b5983436	PERF-2 P1.2: the overlap map's cell classification skips the point location for background cells outside the patch's node bounding box (no patch cell can contain them and the body lies inside the patch: Active, the class the search returns); PatchMesh::bounding_box ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 23:34:40 -05:00
Omar SobhandClaude Fable 5.1	79c18def32	PERF-2 P1.1: the multigrid-PCG's prepared operator (hierarchy, f64 fine level, active cells, components) cached on the embedded solver and reused while the operator is bit-identical (an exact key over the coefficient bit patterns, the mask and the hierarchy parameters); solve_multigrid_pcg_cached gives the uncached solve's answer bit for bit (pin poisson_cache_exact.rs: five right-hand sides on one operator, a one-coefficient miss, both precisions); the CG driver split into Prepared::build + run_pcg ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 23:30:53 -05:00
Omar SobhandClaude Fable 5.1	01830e5c8d	PERF-2 P1: exact chunked polyline index for the O-grid's sliding outer ring — PolylineIndex::nearest returns nearest_on_polyline's point bit for bit (same segments, same order, same strict minimum; chunks skipped only when their box is farther than the best by more than 1e-12 relative), used by the Winslow sweep and the two ring projections; pin: 0 mismatches over 39,864 queries on a 3,166-point rounded outline including vertices, mid-points and ties ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 23:08:09 -05:00
Omar Sobh	11f71b832c	rtx-cfd: the regeneration profile block sits above the too_many_arguments attribute, not between it and its function	2026-09-15 22:53:40 -05:00
Omar SobhandClaude Fable 5.1	1547d14ff1	PERF-2 P0-b: CG iterations per Poisson solve in the profile (ms per V-cycle) and sub-timers inside the patch regeneration (outline, hull + offset, ring projection, Winslow sweeps, respace, warm bookkeeping, mesh finalisation) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 22:53:26 -05:00
Omar SobhandClaude Fable 5.1	172a26dee4	PERF-2 P0: intra-step profiler — StepTimers on the overset solver (RTX_PROFILE; overlap build / predictors / patch BiCGSTAB / background Poisson with its setup-vs-iterate split / round exchange / apply / end exchange), PoissonSolution carries setup_ns and iterate_ns, the harness times advance / restore / snapshot / load sampling / force / FEA predictor and prints the wall split of the coupled phase; no clock is read when profiling is off; the reclassified/step progress denominator fixed (was ×4) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 22:25:52 -05:00
Omar SobhandClaude Fable 5.1	2d2fed7181	rtx-fsi harness: 'fixed' coupler (constant relaxation OMEGA0) for the P6-c Robin–Neumann test — at ω = 1 the candidate is the FEA's own output, so the Robin wall's previous-pass datum is the structure's traction at the candidate; design 2's FSI3 divergence is the datum's inconsistency under IQN's ω0 = 0.05 ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 19:22:36 -05:00
Omar SobhandClaude Fable 5.1	96a7f1c700	rtx-cfd: Robin wall sign pin — a still closed annulus under a uniform pressure above a zero datum: the explicit offset recedes into the body at exactly −p₀/(α(1+g)) (3e-16), the implicit compliant term answers with a pressure DROP (mean 0.17 of p₀, max 0.36) that absorbs the recession (net wall flux 6e-19 of 0.51); the steady MMS pin is blind to the sign of both parts ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 11:17:38 -05:00
Omar SobhandClaude Fable 5.1	1c98156653	rtx-cfd: the deformed-tip outline chooses the flat / semicircle branch against the UNDEFORMED half-thickness — at the semicircle setting the deformed tip's radius breathes by nanometres and flipped the topology (143 → 144 points) within the first steps of the P6-b ny 41 cap-12 release at every α; pin: the point count is fixed under ± 1e-6 thickness breathing at h 10 / 6.6 mm and ny 41, semicircle and flat ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 11:13:55 -05:00
Omar SobhandClaude Fable 5.1	c3dbd5040a	P6-b design 2: the Robin wall on the patch's Inner side — RobinWall { alpha, datum } with the wall velocity u_s + (t_f − datum)/α (explicit, damped by the wall's own viscous gain μ/(α d)) and the compliant-wall pressure term |S| p'/α implicit in the projection (the added-mass operator in the fluid's own response); Dirichlet bit for bit when off; MMS pin on the skewed annulus (orders 2.46/2.13 at α = 10 and 100 μ/h, the Dirichlet values; a 1e12 wall reproduces Dirichlet to 2e-6); OversetPisoSolver::patch_mut; harness knob RTX_FSI2O_ROBIN_ALPHA with the previous pass's tractions as the datum, printed in the header ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 10:46:35 -05:00
Omar SobhandClaude Fable 5.1	c144a5f733	P6-b: the compensating load uses the candidate's own Newmark acceleration a(d) = (d − u_pred)/(β Δt²) (consistent with the relaxed candidate the fluid saw; cancels exactly at the fixed point) — the lagged structure-output acceleration of the first try made the loop slower (1.8 → 5.2 subit/step at α 0 → 2) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 08:37:35 -05:00
Omar SobhandClaude Fable 5.1	a2086a59de	P6-b: the fictitious added mass for the partitioned loop — rtx-fea NonlinearDynamicStepper::set_added_lumped_mass (a lumped per-DOF mass in the Newmark inertial residual and effective tangent, never the consistent mass or the rest state; zero = the plain stepper bit for bit) with its pin (compensated step reproduces the plain step to 2.5e-9, uncompensated moves it 11 %); the FSI2 overset harness carries RTX_FSI2O_FICT_MASS=α (α × ρ_f π (c/2)² spread over the wetted nodes) and adds the compensating load M_f ü_k of the previous subiterate to every structure solve (predictor and passes), printed in the header ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL	2026-09-15 08:30:55 -05:00
Omar SobhandClaude Fable 5.1	e54729241d	fsi2 overset harness: RTX_FSI2O_CASE=FSI3 (the FSI3 constants with FEATFLOW level-4 references in the summary), RTX_FSI2O_SAVE_FROM (instants only from a time on — a dense last period for the viewer) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01LzcjQX7tvgn87CQCyg9Cfr	2026-09-14 06:36:45 -05:00
Omar SobhandClaude Fable 5.1	415c32ab04	overset mesh: the flat-tip outline (P5-3 option B) — cylinder_flag_outline_deformed_tip / cylinder_flag_patch_deformed{,_from}_tip take the tip's corner radius (t = the recorded semicircle, bit for bit; smaller = the benchmark's flat face between corner arcs sampled at the wall spacing, point count fixed across the deformation, the tip's corners kept off the hull source); pins: corner t reproduces the recorded outline, the 2.5 mm outline has the exact area excess with no collapsed spacing, the O-grid builds straight and at ± 80 mm cold and warm at ny 41/62; harness knob RTX_FSI2O_TIP_CORNER (default 0.01) on the cold and warm builds, printed in both headers; pinned-toolchain fmt on touched crates ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01LzcjQX7tvgn87CQCyg9Cfr	2026-09-13 22:22:08 -05:00
Omar SobhandClaude Fable 5.1	7224521309	fsi2 overset harness: reference constants from FEATFLOW's FSI2 level-4 table (uy 1.24 ± 81.7 mm at 1.93 Hz, ux −14.87 ± 12.73, drag 215.18 ± 77.78, lift 0.87 ± 238.0) — the 2006 table's 80.6 / 2.0 was level 2 with the frequency rounded; the summary line prints the full reference ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01LzcjQX7tvgn87CQCyg9Cfr	2026-09-13 19:43:17 -05:00
Omar SobhandClaude Fable 5.1	5aa05f140d	fsi2 replay: RTX_FSI2O_MAX_ROUNDS knob (3) on the prescribed-motion instrument, printed in its header — the coupled march cannot run at a tighter Schwarz cap (release dies), so the fluid side of the rounds question is measured on the replay ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01LzcjQX7tvgn87CQCyg9Cfr	2026-09-13 19:00:02 -05:00
Omar SobhandClaude Fable 5.1	0645565bd8	fsi2 harness: RTX_FSI2O_FILLET knob (the outline's fillet radius, 5 mm = t/2 in every recorded run: a half-round tip and 5 mm root fillets vs the reference's sharp rectangle; printed in both headers) — P5-3's first problem-definition probe after the h-term was traced to the wake's strength on the background ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01LzcjQX7tvgn87CQCyg9Cfr	2026-09-13 01:38:45 -05:00
Omar SobhandClaude Fable 5.1	c63115053b	fsi2 harness: RTX_FSI2O_PATCH_STRETCH knob (the patch's across stretch, 4.0 in every recorded run; printed in both headers) — P5-3's wall-normal resolution test at fixed h: the rows ladder does not build (the overlap needs outer patch cells ≈ h/2 or larger), the stretch refines the wall cell 0.227h → 0.144h → 0.088h at 4/8/16 while the outer cells grow ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01LzcjQX7tvgn87CQCyg9Cfr	2026-09-13 00:16:20 -05:00
Omar SobhandClaude Fable 5.1	c7d2870580	fsi2 harness: RTX_FSI2O_BG_CONVECTION knob (tvd default / upwind) on the background, printed in both headers beside the patch's — the patch scheme was excluded as the h-term's carrier (upwind ladder +1.52 vs TVD +1.73 W/m from ny 41 to 62), the background's wake dissipation is the registered next probe ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01LzcjQX7tvgn87CQCyg9Cfr	2026-09-12 23:01:57 -05:00
Omar SobhandClaude Fable 5.1	ff7cc2d739	fsi2 harness: RTX_FSI2O_PATCH_CONVECTION knob (tvd default / upwind / none) on the overset patch, printed in the march's coupling header and the replay's PRESCRIBED header — P5-3's near-wake dispersion test on the ny 62 fixed-motion replay; harness files under the pinned toolchain's cargo fmt ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01LzcjQX7tvgn87CQCyg9Cfr	2026-09-12 21:18:03 -05:00
Omar SobhandClaude Fable 5.1	754dcd3679	overset: the energy pin (P5-3) — the rigid cylinder + flag outline pitching about the cylinder's centre in still fluid on the moving patch at the FSI2 fluid and grid; per-cycle net fluid work on the body (must be ≤ 0 at every h), gross +/− split, and the last-two-period moment fit → added inertia + rotational damping; knobs RTX_OVERSET_NY / _EP_PERIODS / _EP_TIP / _EP_F; ny 41 one-period smoke: net −5.62 J/m, damping 56.9 N·m·s/m (π c θ0² ω = 5.6 J, self-consistent), 21 s ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01LzcjQX7tvgn87CQCyg9Cfr	2026-09-12 20:57:58 -05:00
Omar SobhandClaude Fable 5.1	914d369efe	overset: the stamp-energy instrument (P5-3's energy-source measurement) — OversetResult gains stamp_energy (½ρ(u_stamped² − u_before²)·dx·dy summed over the fringe faces after the flux balance on a reclassification step) and reclass_energy (the faces touching a cell whose class changed); the replay CSV logs both plus the reclassified-cell count per step — the comparator showed the plate loads per unit motion agree with ours to ~10 %, so the discrepancy is our limit-cycle amplitude, an h-growing energy source, and the hole boundary's reclassification is the candidate ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-12 17:32:05 -05:00
Omar SobhandClaude Opus 5	796b8487ad	rtx-backend-cuda: native index_select / index_add — HetGAT training 2.5x, inference 4.3x ... The Backend trait's index_select and index_add have default bodies that round-trip through host memory. That is correct everywhere and was the only implementation CUDA had. Graph message passing is made of these two ops, so dg-gnn's HetGAT paid a device->host->device copy per layer per pass and the RTX 5060 Ti sat at ~10% utilisation during training. Design follows rtx-backend-metal's ops::index: gather is one thread per output element; scatter-add walks the CSR of the adjoint selection matrix S^T, built host-side by counting sort, so it needs NO atomics and is deterministic with duplicate indices — the training loss is bit-identical to the host reference. Device index buffers are cached per thread keyed by the exact index list, so a static graph topology uploads once. Two small NVRTC kernels; no cuSPARSE. Measured on dg-gnn, Harris 42,955 links, v8 recipe, RTX 5060 Ti: training batch 8 9,042 -> 3,562 ms/step (2.5x) inference single p50 55.4 -> 12.9 ms (4.3x) inference batch 8 257 -> 20 ms/scen (13x; batching helps again) GPU utilisation median 10% -> 21%, p90 17% -> 43% Tests: gather with repeats, scatter-add with duplicates and untouched rows, the adjoint identity <S x, y> == <x, S^T y> (what autograd relies on), a hub-heavy pattern against the host reference, and the range-check panic. rtx-backend-cuda --features cuda: 60 + 16 passed, 0 failed. Co-Authored-By: Claude Opus 5 <[email protected]>	2026-09-11 23:09:01 -05:00
Omar SobhandClaude Opus 5	4f7b350d96	rtx-backend-cuda: transpose of tall tensors failed to launch, and the generic permute was an identity copy ... Found by running dg-gnn's HetGAT training on an RTX 5060 Ti: the first backward pass panicked in transpose_2d with CUDA_ERROR_INVALID_VALUE. The tensor was a [4,211,136 x 1] gradient. The tiled kernel maps rows onto grid.y, and CUDA caps grid.y at 65,535 blocks, so anything past ~2.1M rows was rejected at launch. Inference never hit it — the forward pass has no transposes. Three fixes, each tested on the GPU: - A row or column vector transposes to itself in memory ([N,1] and [1,N] are the same N floats). swap_dims now shares the buffer and swaps the shape; no kernel. This is the case that actually failed. - transpose_2d_gpu hands matrices with rows/32 > 65,535 to the 1-D generic permute, whose grid.x allows 2^31-1 blocks. - transpose_generic_gpu passed the INPUT's strides (swapped) as the kernel's output_strides — but the kernel DECODES each flat output index with those, so they must be the contiguous strides of the output shape. For a 2-D input that was [1, cols], which decodes every index to itself: the "permute" was a plain copy. Verified against the original code — a [2,3,4] swap_dims(0,2) returned b[1][0][0] = 12.0 where 1.0 is correct. Every higher-D dim swap went through this path; nothing had checked it numerically. Also stamps kernel outputs with the contiguous strides of their new shape rather than the input's strides swapped. Nothing outside ops/shape.rs reads strides today, so that was latent, but is_contiguous() now tells the truth. rtx-backend-cuda --features cuda: 55 + 16 passed, 0 failed (was 51 + 16; four new tests, one of which fails on the original code for each bug above). Co-Authored-By: Claude Opus 5 <[email protected]>	2026-09-11 21:42:27 -05:00
Omar SobhandClaude Fable 5.1	33534ee688	overset added-mass pin (P5-3): a no-slip cylinder on the moving patch oscillating in a closed box of still fluid, the force fitted over the last two periods — C_m = 1.069 (n = 32) / 1.097 (n = 64) against Stokes's 1.064 plus box blockage: the composite's unsteady pressure response is right, the FSI2 lift excess is not the unsteady term's scale ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-09 18:01:53 -07:00
Omar SobhandClaude Fable 5.1	aeb70eb507	prescribed-motion replay: per-step Schwarz rounds, convergence, background residual, patch divergence and mass defects in the CSV — so the next burst is seen forming ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-09 17:57:06 -07:00
Omar SobhandClaude Fable 5.1	1562b7cbb5	prescribed-motion replay: the fluid's power on the flag split face by face (normal vs tangential traction; root / middle / tip thirds) beside the nodal Σ F·ḋ, and replay instants (RTX_FSI2O_SAVE + SAVE_EVERY or SAVE_AT=t1,t2,… — the step that first reaches each time) for the per-region pressure audit at one physical instant across h ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-09 15:49:27 -07:00
Omar SobhandClaude Fable 5.1	f61c427b6d	P5-3 energy discriminator: prescribed motion on the overset fluid alone — fsi2_harness/replay.rs (a march's saved instants as cubic-Hermite kinematics with the saved velocities as slopes, the prescribed wall velocity the interpolant's exact derivative; a smooth ramp) with a knot/derivative pin, and fsi2_overset_prescribed_motion (RTX_FSI2O_REPLAY=dir: replay from RTX_FSI2O_REPLAY_T0 on RTX_FSI2O_LOAD's rigid state at any ny; per-step CSV of tip uy, drag, lift, the fluid's power on the flag Σ F·ḋ, wall flux vs area rate; window summary) — at a fixed motion, does the fluid's work per cycle grow with h the way the coupled amplitude ladder (91.2 → 94.8 → 98.0 mm) does? ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-09 13:13:59 -07:00
Omar SobhandClaude Fable 5.1	7e13d14e96	FSI2 overset harness: RTX_FSI2O_DT_SCALE scales the CFL-derived fluid step — P5-3's dt-convergence probe of the first-order Euler fluid steps (the band's placement is excluded: 94.0 mm at a 60 mm patch vs 94.8 at 40 mm; the s = 8 pair moved the amplitude 95 → 60 mm) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-09 06:41:08 -07:00
Omar SobhandClaude Fable 5.1	75ab6e77b4	FSI2 overset march: print every coupling setting once (coupler, floor, rtol, stall accept, max subit, predictor, s, patch offset × rows) and add RTX_FSI2O_LOAD_SETTLE (rigid steps after a loaded state, clock put back; default 20) — P5-3 lost a day to the default floor 2e-4 against the overnight marches' 1e-6: at 2e-4 every s = 1 release at ny = 62 rings at ± 1000 N/m (the flag's thickness breathing), at 1e-6 all are clean, thick patches included ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-08 14:52:02 -07:00
Omar SobhandClaude Fable 5.1	3e019d1491	P5-3 instruments: per-region wall loads on saved instants (top / bottom / tip arc / fillets+cylinder — drag, lift, normal-traction mean/min/max; RTX_FSI2O_AUDIT_FACES, RTX_FSI2O_AUDIT_ONLY) and the patch-thickness knobs (RTX_FSI2O_PATCH_OFFSET / _PATCH_ROWS, RTX_OVERSET_PATCH_OFFSET / _PATCH_ROWS; save tags carry the offset) — the overlap band sat at a fixed number of cells (6 h) and moved inward in metres with refinement ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-08 05:35:14 -07:00
Omar SobhandClaude Fable 5.1	2c1143076e	P5 audit exact on the moving path: the composite rebuilt around a saved instant carries the march's dt (from the mesh the march started on — the warm base under RTX_FSI2O_WARM_SWEEPS — whatever patch the instant holds; 0.2 % of dt read as 0.3–5 N/m of solved-face residual), and instants save dt; over 131 instants of the FSI2 march to t = 16 s the solved-face residual is 1.7e-10 at every one ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 12:32:38 -07:00
Omar SobhandClaude Fable 5.1	06229dd9a2	rtx-cfd: the deformed cylinder–flag build's hull source carries the deformed outline — the cylinder + tip disc alone give the chord under a bent flag's convex side (P5-2 died at tip −35 mm with the patch 4.4 h thick where 6 h was asked, an acceptor's donor on a prescribed face); the death reproduced from the last saved instant by extrapolating the interface (fsi2_overset_probe_death_from_instant, with the failing acceptor's ray geometry and the overlap depths that build) and gone with the fix at every τ; RTX_FSI2O_ROWS; the zero-displacement mesh gate at 1e-5 (the pipeline's), the outline gate 1e-12 unchanged ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 10:25:46 -07:00
Omar SobhandClaude Fable 5.1	3c2add3f32	P5 audit: the overset march saves the composite at every N-th committed step (RTX_FSI2O_SAVE_EVERY: background, the deformed patch's nodes, its vectors, d, ddot, time), and fsi2_overset_audit_of_saved_instants (RTX_FSI2O_AUDIT=dir) rebuilds the composite around each saved patch (build_case_with an initial mesh, so the overlap is the instant's) and prints the solver-metric chain — the CFD23 instrument on the moving patch ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 09:13:07 -07:00
Omar SobhandClaude Fable 5.1	3a68048934	rtx-cfd: cylinder_flag_patch_deformed_from — the warm-started regeneration (P5's cost lever): the previous patch's interior and its outer ring projected onto the new offset polygon as the start, and an IDEMPOTENT build (one Winslow sweep alternating with the ray re-spacing; the P4-0 one-shot is not a fixed point — measured as the interior moving 0.05 h per build at rest), so a base converged as that map's fixed point stays put (9e-11 h at rest) and each build tracks the outline (interior/wall motion 0.99–1.14, worst angle 74.7°, wall row 0.37 h) 4.5× faster than the cold build; the rigid generator untouched (pins pass); fsi2 overset harness: RTX_FSI2O_WARM_SWEEPS with the converged base as the initial patch and the committed step's mesh as each step's warm start ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 09:10:05 -07:00
Omar SobhandClaude Fable 5.1	62f1d4bf9b	P5-1: RTX_FSI2O_NEWMARK_GAMMA — Newmark dissipation on the flag (β = (γ + ½)²/4); at γ = 0.9 the s = 1 overset march holds 0.3 s from the cached state (789 steps, 1.8 passes, no death) with the physical onset, where γ = ½ ran away in 90 steps: the flag's thickness breathing, resolved at s = 1 and felt by the body-fitted wall, damped at the source ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 08:45:45 -07:00
Omar SobhandClaude Fable 5.1	565215baf9	P5-1 instruments and discriminators: RTX_FSI2O_NO_WALL_VEL, _FREEZE_PATCH, _VOLUME_PRESERVE (the imposed wall velocity's net flux removed uniformly along the wall), the pass trace carries the fluid's power on the flag, the wall-pressure roughness, the patch's pressure level, the wall's net volume flux against the FEA polygon's area rate, and the mass defects — the s = 1 runaway is the flag's thickness breathing (ν = 0.4 Quad8) seen by the body-fitted wall as a volume flux equal to the polygon's area rate, answered by the projection's pressure level at ρ/dt ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 08:37:34 -07:00
Omar SobhandClaude Fable 5.1	2ed36c3a8c	P5-1 instruments: the overset fluid saves/loads its rigid-phase state (RTX_FSI2O_SAVE / _LOAD — a coupling experiment then costs seconds, not the 12-minute rigid march), dumps the interface edges when the composite refuses a patch (RTX_FSI2O_DUMP_DIR, at the advance where the overlap is rebuilt), the release time is the fluid's clock; patch_cylinder_flag_deformed: cells inside the flag are holes at small bends and both sweep counts (the cantilever shape does not reproduce the P5-1 death — the CSVs show a coupling runaway), and a dumped-edges reproduction test ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 08:09:56 -07:00
Omar SobhandClaude Fable 5.1	d74571e21a	rtx-cfd: the deformed cylinder–flag outline's edge fractions come from the UNDEFORMED straight edge (x_tip_ref), so the patch topology is the mesh family's and not the deformation's — the P5-1 gate run died at the first bend with set_mesh refusing 143 → 144 wall cells; the P5-0 test now gates ns across amplitudes (143 at −80 / +40 / +80 mm) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 07:34:26 -07:00
Omar SobhandClaude Fable 5.1	f0b2563bf8	P5-1: FSI2 with the fluid on the overset — fsi2_harness/overset.rs (the background without a body, the cylinder–flag patch regenerated around the deformed flag every pass via set_patch_mesh, the wall velocity from the interface velocities along the wetted polygon, the load from the patch's wall faces into WettedSurface::transfer_load — no probes, no clamp, no smoothing), fsi2_harness/overset_march.rs (the harness's rigid phase / release / subiterated coupling with its acceptance rule, no rescue machinery, death returned not panicked), tests/turek_hron_fsi2_overset.rs (RTX_FSI2O_* knobs); rtx-cfd: CurvilinearPisoSolver::wall_tractions (per-face pressure + full-stress traction, surface_force sums the same terms bit-identically); Interface::edges (bottom/tip/top for the generator); cylinder_flag_mms RTX_CF_BEND (P5-0 gate iv: Stokes orders 2.14 / 2.09 on the flag bent to 80 mm) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 07:16:53 -07:00
Omar SobhandClaude Fable 5.1	c2451fbacf	rtx-cfd: P5-0 — cylinder_flag_patch_deformed / cylinder_flag_outline_deformed: the O-grid around the cylinder and the FEA's deformed flag edges (FlagEdges bottom / tip / top), edges resampled at the rigid outline's graded arclength spacing, root fillets from the rigid construction, the tip semicircle from the deformed tip's centre and axis; the O-grid body factored out (o_grid_from_outline, hull_source) and the rigid generator kept bit-identical (its pins pass); gates: zero displacement reproduces the rigid outline to 6e-17 (mesh 2.5e-8 through the hull/Winslow pipeline, recorded), ±80 mm cantilever shapes valid at 80° with wall row 0.43 h, along-body 0.196 h, wall nodes on the edges to 6e-17, ny = 41 overlap builds ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 07:10:01 -07:00
Omar SobhandClaude Fable 5.1	cee9adef26	rtx-cfd: overset_cfd23 RTX_OVERSET_CFD23_PHASE=N — the momentum chain's stages sampled every N steps and averaged over the window; over whole shedding periods the unsteady terms vanish and the averaged chain is exact (the patch balance's unstored unsteady term included) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 04:14:55 -07:00
Omar SobhandClaude Fable 5.1	1a8f4aaebe	rtx-cfd: OversetPisoSolver::set_time; overset_cfd23 resumes a march from a saved field (RTX_OVERSET_CFD23_RESUME_T, _T_START) — CFD3 ny = 82's wake is still growing at t = 6–9 s (half-window amplitudes 187 / 339), the settled amplitude needs t = 15 ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-07 00:08:27 -07:00
Omar SobhandClaude Fable 5.1	ec17217e9e	rtx-cfd: overset_cfd23 RTX_OVERSET_CFD23_ND=explicit — the patch's across-diffusion explicit (dt limit then includes the n-spacing) for the 10 mm damping-floor probe; field tags carry it ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-06 20:44:07 -07:00
Omar SobhandClaude Fable 5.1	11d001fe48	rtx-cfd: solver_metric_force carries the scheme's limited face correction with the predictor's far-upwind choice, so the flux form is the solver's under TVD too (CFD2 ny = 41: three boxes spread 1.8 N/m → 5e-13; the upwind-only form read 120.2 for a solver box of 126.0) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-06 17:42:47 -07:00
Omar SobhandClaude Fable 5.1	ce3cbcce57	rtx-cfd: overset_cfd23 — Turek–Hron CFD2/CFD3 on the overset (the harness's ramped inflow from rest, TVD background and patch, 3-round cap), loads by the patch wall stress with the solver-flux-form box and the CV formula beside it as window statistics, CFD3 frequency from lift crossings, the solver-metric momentum chain at the final state (solved-face residual is the pin), save/load of the fields by case; momentum_residual checks the far-upwind neighbours explicitly under a limited background (the van Albada limiter swallows a NaN silently) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-06 17:10:18 -07:00
Omar SobhandClaude Fable 5.1	36af178980	rtx-cfd: overset_cfd1 RTX_OVERSET_ROWS — the overlap depth (patch rows kept non-hole below the acceptor row, default 4) as a knob for the §5.11 band probe; saved-field tags carry it; rows ≤ 3 are refused by the map at ny = 41 (acceptor donors reach the hole at the flag tip) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_0116sg1Qz1gMv9hdcKP1XUam	2026-09-06 15:30:31 -07:00
Omar SobhandClaude Fable 5.1	4c000266d9	rtx-cfd: overset momentum residual carries its pieces per ring face (time, convective, diffusive, pressure); overset_cfd1 offline probes on the saved fields — RTX_OVERSET_CFD1_FRONT dumps the cylinder-front faces by piece, RTX_OVERSET_CFD1_RAW re-evaluates the ring on the raw interpolated stamping without balance_fringe_fluxes (the balance is not the ring's source: +0.241 raw vs +0.247 balanced at ny = 41 upwind) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-06 14:37:08 -07:00
Omar SobhandClaude Fable 5.1	0215c7d6a5	rtx-cfd: CurvilinearPisoSolver::momentum_balance — the patch's own momentum balance on its solved cells in the scheme's fluxes (outward ρFu_f with the predictor's face value, Laplacian-form μ∇u·S on the solved/acceptor interface and the wall, the least-squares pressure volume sum vs the face-pressure integrals); flux_force, wall_force, pressure_defect δP; overset_cfd1 prints it and the acceptor band's mismatch, saves the patch flux, and RTX_OVERSET_CFD1_LOAD=dir runs the diagnostics offline on saved fields ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-06 14:13:04 -07:00
Omar SobhandClaude Fable 5.1	e31576d543	rtx-cfd: OversetPisoSolver::solver_metric_force — the box force in the predictor's own flux form (upwind convective, diffusive, cell pressure on the momentum control volumes' faces, minus the unsteady term); box-independent to rounding in the active region (five boxes spread 2e-12 at step 5 vs the CV formula's ±0.5 %); overset_cfd1 prints wall − box = ring Σr + rest and saves the settled fields under RTX_OVERSET_CFD1_SAVE ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-06 13:47:43 -07:00
Omar SobhandClaude Fable 5.1	6f9b0d43b2	rtx-cfd: OversetPisoSolver::momentum_residual — the background predictor's own staggered stencil (u_rhs/v_rhs, factored out of the predictor bit-identically) evaluated on every face of a NaN-masked field; solved faces read rounding, the active–fringe interface reads the composite's pressure level offset δ·h (cancels in the sum), prescribed fringe–fringe / fringe–hole faces read the stamping's momentum injection; hole ghosts (p, u, v) from a band widened three rows into the hole make every ring face evaluable; overset_cfd1 prints the buckets, the ring x-bands and δ at the settled state; pin: residual vanishes on the solved faces ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-06 13:25:14 -07:00
Omar SobhandClaude Fable 5.1	cbec40b999	rtx-cfd: overset_cfd1 prints five CV boxes (all outside the fringe ring), the wall pressure/viscous split and the fringe extents at the settled state ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-06 10:14:43 -07:00
Omar SobhandClaude Fable 5.1	134ac03870	rtx-cfd: OverlapMap::region_force — momentum flux into a background region (control-volume face formula, one-sided next to holes); overset_cfd1 prints the four momentum routes (CV box, ring outer, hole boundary, wall) and their defects at the settled state ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-06 09:08:01 -07:00
Omar SobhandClaude Fable 5.1	5f780447de	rtx-cfd: PatchConvection::TvdVanAlbada — van Albada deferred correction on the curvilinear predictor (downwind-side linear weight, gradient-ratio r over the face d lengths, far-upwind across the opposite face, boundary faces upwind); annulus MMS orders 2.10/1.69 at 0.24× upwind; cylinder-flag MMS orders 1.98/1.97 (1.06× upwind — diffusion-dominated, recorded); knobs RTX_OVERSET_CFD1_TVD, RTX_OVERSET_MAX_ROUNDS, RTX_CF_SCHEME ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-06 00:21:48 -07:00
Omar SobhandClaude Fable 5.1	02c855b9c4	rtx-cfd: overset Schwarz stall rule guarded (fires only within 10× the tolerance) — unguarded it cut CFD1's transient at round 3 every step and the coupled march diverged at ny=62 (pressure 4e4 → 1e140 by step 450); CFD1 harness knobs (stall, step cap, trace) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-05 21:01:57 -07:00
Omar SobhandClaude Fable 5.1	00f73eb70a	rtx-cfd: overset CFD1 harness — the steady-march stall rule (stall_rounds 2) actually enabled (the previous commit's edit had missed its anchor) ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-05 20:28:25 -07:00
Omar SobhandClaude Fable 5.1	9761cf2319	rtx-cfd: overset P4 — CFD1 on the composite (tests/overset_cfd1.rs): ny=41 wall drag 15.2156 (+6.46%), lift 1.0879 (−2.78%), CV drag 15.528 (+8.64%), routes 2.0% apart (staircase +10%); 28550 steps, 2099 s at dt 4.57e-4 with 5.0 Schwarz rounds mean (stall rule off) — steady-march stall rule on for the next rungs ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-05 20:27:52 -07:00
Omar SobhandClaude Fable 5.1	45ff34da8f	rtx-cfd: overset P4-0 — the O-grid around the Turek–Hron rigid body (cylinder + flag), gated ... patch_gen::{cylinder_flag_outline, cylinder_flag_patch, winslow_smooth, respace_rays} (+ convex_hull, offset_convex_polygon, nearest_on_polyline): the outline CCW then reversed to clockwise (tip semicircle 16 cells, junction fillets of a FIXED radius with 3 cells, straights graded 0.3 h → h, cylinder arc at h); the outer ring the 6 h normal offset of the body's convex hull; initial pairing by the inner point's normal offset projected onto the hull offset (an arclength-proportional pairing folded the transfinite grid at the tip: rays crossed where the curvatures differ); Winslow (TTM) smoothing of the interior with the outer nodes SLIDING along the hull offset (each re-placed at the nearest point to the extrapolated ray), then re-spacing along the smoothed rays to the across stretch. Gates (tests/patch_cylinder_flag.rs): ny = 41/62/82 → 143×12 / 183×12 / 225×12 cells, positive, wall row 0.23 h (fillet max 0.37 / 0.43 / 0.50 h), worst non-orthogonality 76.6 / 69.1 / 63.3° at the concave fillets (structural: a concave arc's normals converge at its centre), classification of the benchmark background with both donor invariants. P0 MMS on these meshes (tests/cylinder_flag_mms.rs, exact acceptors, line-implicit): Stokes orders 2.17 / 2.13, upwind 2.00 / 1.86 (cell Péclet ≈ 0.1), divergence ≤ 9e-14 — the fillet skew costs nothing measurable. Rule: a refinement ladder's geometry must be fixed in physical units — with fillet = h/2 the Stokes orders read 1.74 → 1.31, the O(h) boundary perturbation masquerading as a scheme defect; the fillet is a parameter (5 mm across the ladder). Also: the P3b knock-outs H1/H2 on the balanced default (no effect), the P3 §5.10 record in the falsifier's header. Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-05 19:50:14 -07:00
Omar SobhandClaude Fable 5.1	62df6bd628	rtx-cfd: overset P3b — the reclassification impulse located (the fringe ring is a staircase of the interpolated velocities' mass defect) and removed by a converged fringe flux balance (default on): falsifier max spike 594 → 5.50 N/m at the FSI2 step (staircase 6490), 10.95 / 16.79 at dt/2 / dt/4 (12600 / 25600), rms spike 0.07% of the force, far probe 6 (7900), KE per event 4.9e-3 J/m falling with Δt (2.6 fixed) ... OverlapMap::balance_fringe_fluxes: Gauss–Seidel through the prescribed faces of every fringe cell to 1e-12 of the prescribed flux scale (≤ 50 sweeps), after every fringe stamping (3 fixed sweeps 101 N/m, 10 sweeps 5.5 — converged is the rule). OversetParameters: fringe_flux_balance (default on, RTX_OVERSET_NO_BALANCE off), fringe_balance_tolerance, refill_turned_active (measured no effect: 593.7 → 593.8; kept as the record), stall_rounds opt-in. P3b locating trace RTX_OVERSET_TRACE_SP (continuity source by class change in cell volumes/step, stored-pressure jump of turned-active cells): the flipped cells' mass source ≤ 6e-3 cell volumes/step, their stored pressure 5–10% of the range off their neighbours (4.4% on the static MMS — the meshes' discretization disagreement). Knock-outs refuted (RTX_OVERSET_H1 keep own face velocities, H4 no warm start, pressure refill): 593–597 N/m each. S4 MMS with the balance: velocity errors within 0.1% of the pinned values, the background's overlap mass defect 1e-13 by construction, pressure errors unchanged. overset_mms prints pressure diagnostics; overset_falsifier records the balanced ladder (regression guard 20 N/m at dt; RTX_OVERSET_FALSIFIER_STRICT asserts the registered gates — (ii) holds at dt, misses at dt/2, dt/4; (iv) fails: residual ∝ 1/Δt^0.8). Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-05 19:16:16 -07:00
Omar SobhandClaude Fable 5.1	6b8837301f	rtx-cfd: overset Schwarz stall detection is opt-in (stall_rounds, default 0): right for a steady march (n=64: 2.26 rounds mean, no cap hits, L2 identical), wrong for a transient (falsifier max spike 594 → 4044 N/m when on); the MMS harness sets 2 ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-05 18:35:06 -07:00
Omar SobhandClaude Fable 5.1	e2edff9b1d	rtx-cfd: overset A-P3 — the falsifier plate on the overset (FAILS the registered gates by one order less than the staircase); wall force; composite pressure-level pin; Schwarz stall detection ... patch_gen::{stadium, graded_fractions}: the falsifier plate as a stadium O-grid (semicircular ends r = half-thickness; 16 cells per end arc, straights graded 0.30 h -> h at 1.15, offset 6 h, 12 rows stretched 4x; 148x12 cells, every ray a normal, worst non-orthogonality 4 deg). CurvilinearPisoSolver::surface_force (+ PatchLoad): F = sum(-p_f S_f + mu (grad u + grad u^T)_f . S_f) on the wall faces with the wall cell's LSQ gradients (wall Dirichlet in the velocity fit); HELD on the phantom circle against the exact stress integral: 1.3e-2 / 6.3e-3 / 3.6e-3 at n = 32/64/128 (orders 1.05 / 0.81), 22x the staircase's accuracy. OversetPisoSolver: the composite p' level pinned to zero mean over the active cells every round (the coupled problem is pure Neumann; the temporal warm start handed each step's level to the next — background pressure 1e7 growing 5e4 per step on the falsifier; an unpinned level also inflated the relative Schwarz stop); stall detection (no progress over three rounds = the inner solvers' noise floor; 6560 of 150k steps burned the 20-round cap at n = 64, a 7.5 h n = 128 march); schwarz_stalled in the result. tests/overset_falsifier.rs (records; RTX_OVERSET_FALSIFIER_STRICT asserts the registered gates, _LADDER runs dt/2 and dt/4, _TRACE the top-12 spike steps): max spike 594 / 981 / 1720 N/m at dt / dt/2 / dt/4 (staircase 6490 / 12600 / 25600), rms spike 61-89 (810), far probe 502-1509 (7900), KE injection 0.16-0.21 J/m per event on the common cell set (2.6) — every large spike a ~104-cell full-row reclassification; exponent -0.77 (-1.0). The registered 5% gate (8.75 N/m) is missed 68x: the overset's own reclassification impulse is the finding (omni-cortex overset_metal_campaign.md §5.10); P3b = locate per cell, then the fringe flux balance. tests/patch_stadium.rs, curvilinear_loads.rs, overset_common::plate_patch. Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-05 04:32:43 -07:00
Omar SobhandClaude Fable 5.1	afd1bff6ee	rtx-cfd: overset A-P2 — the patch overlaps the background (OversetPisoSolver), gated S1–S5 ... Background = the embedded solver with a mask from the overlap classification (embedded/{mod,projection}.rs: module split, projection's solve/apply halves, set_overlap, fringe p' Dirichlet by elimination into extra_diag/rhs, anchor dropped, set_inner_stop_factor, phase API begin_step/solve_correction/ apply_correction/end_step; advance rebuilt on the phases — every suite digit- identical, FSI2 default line-for-line). Patch = the curvilinear solver with an acceptor ring (set_side_velocity; set_acceptor_ring/stamp_acceptors/ set_acceptor_correction; acceptor Dirichlet by elimination into PressureSystem.links so the BiCGSTAB stop stays in flux units — identity rows measured unconverged at 2431 iterations; same phase API). overset/overlap.rs: OverlapMap — hole/fringe/active from the patch's own indices (hole = body or k <= nn-1-overlap_rows, DEFAULT_OVERLAP_ROWS = 4 from the 2.9 h depth budget), dual-quad inverse-bilinear donors patch→fringe, lattice donors →acceptors, both invariants asserted, mass-defect measures. overset/mod.rs: OversetPisoSolver — advance (exchange rebuilt BEFORE the predictors from the previous corrected field), alternating Schwarz on the acceptor p' vector with Anderson(3) (plain Schwarz measured 0.82/round: floating patch, Neumann wall) and the previous step's vector as warm start (1 round/corrector at steady state), stop relative to the STEP's p' scale (the MG absolute stop is 1e-9/dt² in pressure — the whole second correction), set_patch_mesh, snapshot/restore carrying the warm-start vector. Gates: overlap linear-exact 1e-13, quadratic orders 1.96/1.99 (acceptors), 1.40/1.91 (fringe); half-couplings: patch with exact acceptors Stokes 2.07/1.98 + 2.08/1.98, upwind 0.84/0.84, background with exact fringe 7.86e-3/2.90e-3/ 1.09e-3 (1.44/1.41); two-mesh MMS n=32/64: background 8.717e-3/4.207e-3 (1.03x/ 0.97x the embedded circle), patch 1.322e-2/6.904e-3 (1.5-1.6x), orders 1.05/ 0.94, patch div <= 5e-13, overlap mass defect 3.6e-3 -> 8.2e-4 of the overlap flux (under the registered 1e-3 from n=64; disclosed at 32); motion: stationary patch through set_patch_mesh bit-identical, snapshot/restore with a pending mesh bit-identical, translating phantom circle 1.22x/1.19x the static level over 4.5 cells. Inherited, disclosed: poisson_equivalence's no-body multigrid pin fails by 3.9e-9 at d46fb0b (M1's commit; verified in a clean worktree). Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-04 19:24:13 -07:00
Omar SobhandClaude Fable 5.1	d46fb0b7a7	rtx-cfd: overset A-P1 GATED — the curvilinear patch moves and deforms under an exact 2-D DGCL ... StepGeometry (motion.rs): time-averaged face vectors S̄_f = ½(S^n + S^{n+1}) and swept volumes δV_f = S̄_f·δc_f — exact for linear node motion on any quad, so Σ sign δV_f = V^{n+1} − V^n is algebra (1.8e-14 measured; the EndOfStep control 2.1e-3). CurvilinearPisoSolver::set_mesh(next) names the end-of-step geometry; advance swaps it in, rebuilds operators + pressure matrix on it (L_f, LSQ gradients, no mesh-velocity term in the projection), keeps the old mesh for V^n and the explicit boundary data; predictor in the conservative ALE form V^{n+1} û = V^n u^n + dt(−Σ sign (F − δV/dt) u_f + ν D + f V^n), written as u·(V^n/V^{n+1}) + … so a stationary mesh is bitwise the static path; fluxes on S̄_f; snapshot carries both meshes; swept_face_rule knob (Trapezoidal default, EndOfStep = negative control). Stokes limit keeps the mesh flux (was dropped with convection) and centres it (upwinding it cost an order: 1.06/1.00). Gates (tests/curvilinear_ale.rs, 11 tests, 83 s): uniform flow on a wiggling AND bending annulus 4.44e-15 over 400 steps, p exactly 0, 0 pressure iterations; control deviates 4.9e-5; stationary mesh through the moving path bit-identical (both diffusion variants); snapshot/restore on the moving mesh bit-identical; Taylor–Green orders unchanged — upwind 0.995/0.976 vs fixed 0.987/0.975 at 1.05× error, Stokes 1.92/1.97 vs 1.94/1.99 at 2.4×, moving annulus 2.11/2.02; linear-field falsifier 1.95/1.92 (annulus), 1.91/1.43 (square, sliding wall nodes). P0 ladders re-run identical to every digit. Rule from the diagnosis: start a moving run ON the t = 0 mesh and sweep less than a cell per step — a first step that jumped 2–4 cells imprinted an O(displacement) error no refinement removed (dt- and motion-independent). Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-04 17:00:04 -07:00
Omar SobhandClaude Fable 5.1	c63d79c300	rtx-cfd/rtx-fsi: overset A-P0 GATED + M1 precision probe — curvilinear collocated PISO: relative-reduction pressure stop (the absolute stop floored |du/dt| at 2e-4 on 64²), line-implicit-n sign fix, adjustPhi; gates: Cartesian reduction 1.37–1.40x the staggered error at orders 0.83/0.90; skewed stretched periodic annulus Stokes orders 2.30/2.06 (explicit and line-implicit), upwind 1.08/0.80; Poiseuille exact to 1e-9 on Cartesian and affine-sheared periodic channels (both diffusion variants), varying-skew channel order 2.02 (v 1.9), cell mass 1e-14; divergence ≤ 1e-11 relative every step; snapshot/restore bit-identical. M1: poisson.rs multigrid hierarchy generic over MgScalar (f32/f64), f64 CG keeps its own fine level; MgPrecision on MultigridParameters/EmbeddedParameters/PisoParameters, set_poisson_precision, harness RTX_FSI2_POISSON_F32 (march + noise probe, printed marker); f64 arm bit-identical in vivo (FSI2 default line-for-line with 08-31), f32 arm holds the noise floor and stall pins and the FSI2 band; poisson_equivalence f32 arm ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-04 12:40:43 -07:00
Omar SobhandClaude Fable 5.1	52da75a3a9	rtx-cfd: curvilinear collocated PISO on a structured patch (overset A-P0, WIP) — PatchMesh (right-handed s,n; periodic seam with shift; face metrics), patch generators (TFI, skewed annulus, sheared/varying-skew channels), CSR + Jacobi-BiCGSTAB, the Zang–Street–Koseff incremental step with the node-based 9-point L_f, LSQ gradients, explicit and line-implicit-n predictors, adjustPhi; tests: mesh metrics (5 green), operators exact on linear fields incl. the seam (green), sparse (2 green), MMS ladder (Cartesian 16/32: 1.37–1.39x the staggered error, order 0.83; n=64 stalls at a |du/dt| floor 2e-4 — open, tolerance-scaling hypothesis), annulus/Poiseuille not yet run ... Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-04 05:00:08 -07:00
Omar SobhandClaude Fable 5.1	1347bc6772	test(rtx-cfd): fresh-cell falsifier — circle body, far-field probe, kinetic energy per step, speed knob, and the per-flip kinetic-energy split by face class (interior / fresh / dead) ... The split locates the flip's velocity change: 71% (plate) / 61% (circle) on the DEAD faces — fluid faces the wall reaches, overwritten with the wall-side reconstruction in one step (a ~0.25 U_b jump per face) — 27–38% in the interior response, ~1% on fresh faces. Five treatments leave the per-flip impulse/energy unchanged on this falsifier (all measured, none kept unless already committed): substeps, the swept-volume source (40x worse), the fresh-face field extension, face apertures + wall-relative divergence in the projection (-12%, and it breaks the static invariants), and imposing the t^{n+1} wall before the predictor. The impulse is intrinsic to a binary staircase wall that stands still between flips and jumps a cell; the record is in omni-cortex docs/fresh_cell_gcl_campaign.md. Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-03 11:56:04 -07:00
Omar SobhandClaude Fable 5.1	c9492d3e1e	test(rtx-cfd): fresh-cell falsifier extended (circle body, far-field probe, kinetic energy, speed knob) + print-only divergence trace; two candidate fixes REFUTED on it (swept-volume source 40x worse at either sign; fresh-face field extension no effect), both kept default-off with their verdicts ... Phase 1 of omni-cortex docs/fresh_cell_gcl_campaign.md. The mechanism of the moving-body force spikes is measured from four directions: per-flip force amplitude ∝ 1/dt, kinetic energy injected per flipped cell 0.048 J/m independent of dt and body shape (plate row vs circle), felt at a far-field pressure probe, and ∝ U^2 (2.60 / 0.64 / 0.15 J/m at U = 1 / 0.5 / 0.25). A binary mask's wall position jumps by one cell at every flip and the fluid answers with a fixed impulse. Neither the swept-volume source (the wall faces already carry the swept volume — the source double-counts it) nor the fresh-face velocity is where it lives. Next: the virtual cut cell in the projection (apertures + the wall-relative divergence), registered in the campaign doc. Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-03 11:40:28 -07:00
Omar SobhandClaude Fable 5.1	23eb996a9f	test(rtx-cfd): the fresh-cell falsifier — an oscillating flag-sized plate on the FSI2 grid and dt; force spikes track fresh cells (90% at dt) and their per-event amplitude scales as dt^-1.0 (rms dt^-0.44) ... Phase 0 of omni-cortex docs/fresh_cell_gcl_campaign.md (Seo & Mittal's oscillating-body test on this embedded solver). Plate 0.35 x 0.02 m, 1 m/s peak, 80 mm amplitude, still fluid, h = 1/152, dt = 3.24e-4 and its halvings; force sampled as the coupling samples it. Measured: at rest exactly zero; moving, 648 fresh cells per sweep at every dt, rms spike 810 / 1080 / 1490 N/m and MAX spike 6.5e3 / 1.26e4 / 2.56e4 at dt, dt/2, dt/4 — the per-event impulse doubles with each halving (a fixed whole-cell volume error per fresh cell delivered in one step, the raw (dV/dt)|1 - CFL_b| source), against a physical added-mass force of ~1.3e3. RTX_FRESHCELL_LADDER=1 runs the ladder, RTX_FRESHCELL_CSV=<dir> dumps per-step records. Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-03 11:24:24 -07:00
Omar SobhandClaude Fable 5.1	19d8307ae4	test(rtx-fsi): closure schedule by tip speed (CRESCUE_SPEED=<f>), default off ... Route 1 of omni-cortex docs/coupling_rescue_campaign.md §12, registered in docs/closure_scheduling_campaign.md: whenever the last committed tip jump exceeds f x its trailing-2000 peak, open a coarse episode (the s=2 interpolated closure, CRESCUE_COARSE=<M> steps, re-evaluated at each end) BEFORE the step — preventive, because the s=1 closure incubates the crossing instability inside the healthy envelope and no signature-triggered rescue acts early enough (rungs A, C, A' refuted). Speed episodes are expected twice per period: not counted against the runaway-episode cap; the coarse fraction is printed at the end of the march. Verified: FSI2 committed default digit-identical knob-off vs the same-day baseline; fmt + clippy clean. Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-03 09:57:40 -07:00
Omar SobhandClaude Fable 5.1	b66f73aeef	test(rtx-fsi): coupling rescue rung A' — increment trigger (CRESCUE_INC=<K>), default off ... Trigger (c) of omni-cortex docs/coupling_rescue_campaign.md §11: a predictor increment above K x its trailing-2000 median (non-rescued steps; median refreshed every 10 steps) opens a coarse episode (rung C) BEFORE the step's first pass — no pass is spent on a step the episode replaces. Calibrated on the INCTRACE dumps: the anchor u=1.00 r=1 stays under 1.72x for its whole march (K = 2..8 never fire); the (1.27, 2.0) death crosses 3x at 145 steps before its panic and never during resonant growth (the tip-jump trigger fires at 54). Registered K = 3, fallback 4. Needs CRESCUE=1 and CRESCUE_COARSE=<M>. Verified: FSI2 committed default digit-identical knob-off vs the same-day baseline; fmt + clippy clean on the touched files. Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-02 23:01:12 -07:00
Omar SobhandClaude Fable 5.1	7e0f159097	test(rtx-fsi): coupling rescue rung C (CRESCUE_COARSE, burst-local s=2 coarsening) + INCTRACE per-step increment dump — both default off ... Coupling-rescue campaign, continued (omni-cortex docs/coupling_rescue_campaign.md §11). Rung A was refuted 4/4 by mechanism (the substeps reproduce the rejected motion — the runaway is in the converged coupled load at the crossing). Diagnostics on the same death: SUBCYCLE=2 marches GREEN to t=16 (zero bursts), HYST=0.25 dies EARLIER. So: - RTX_{prefix}_CRESCUE_COARSE=<M> (with CRESCUE=1): on a trigger, reject the step and take 2dt coupled steps with the fluid subcycled at 2x (fluid dt unchanged = the s=2 interpolated closure) for M coupled steps, then resume; episodes counted, cap 5 per second of march (loud). rescue.rs: coarse_step / attempt_with generalisation; march loop is now a while loop (a coarse step consumes two indices, the series carries a linear midpoint). VERDICT: refuted 2/2 — the coarse steps themselves cannot close once the state is 10x wild; both rungs act too late (the kinematic trigger is the limitation). - RTX_{prefix}_INCTRACE=<csv>: reporting-only per-step dump (step, t, predictor increment, tol_step, passes, residual, stalled, tip jump) — rung A''s calibration data (healthy anchor vs death). Verified: the FSI2 committed default digit-identical knob-off after each change (same-day baseline); fmt + clippy clean on the touched files. Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-02 22:09:49 -07:00
Omar SobhandClaude Fable 5.1	e76271ac67	test(rtx-fsi): coupling-level rescue rung A (RTX_FSI{2,3}_CRESCUE, default off) + TRACE_FROM autopsy window ... Coupling-rescue campaign (omni-cortex docs/coupling_rescue_campaign.md). Every recorded FSI2 s=1 coupling death is the END of a multi-step runaway (74-march burst scan: green marches never exceed 2.4x their own p95 tip jump; 27/28 deaths burst 5-37 steps first). The traced autopsy of the (1.27, 2.0) death (replay digit-identical, CSV byte-identical) shows a growing period-2 instability of the CONVERGED coupled scheme at the tip's max-velocity crossing: increment x33 and converged load 1e4 -> 1e6 N over 200 steps with the coupling converging on 111 of the first 112 steps. - march.rs: RTX_{prefix}_TRACE_FROM (print-only autopsy window: per-pass residual/load + one line per step with increment, tol_step, retry_at, acceptable, outcome, committed tip jump); RTX_{prefix}_CRESCUE (default off): on a fatal stall or a committed tip jump > 3x the running p95 (trailing 2000 non-rescued steps), reject the step and repeat the interval as 2/4/8/16/32 coupled substeps of dt/n (Mayr-Wall-Gee reduced-step repetition, five repetitions); per-rescue record printed, MarchResult.coupling_rescues/_failures/rescue_records, rate cap 20 rescues per second of march (dies loudly). - rescue.rs (new): the substep ladder — each substep a complete coupled step at dt/n with its own predictor, fresh coupler, C1 velocity chaining, tolerances at the substep's increment; the march's own step path is NOT routed through it (digit identity by construction). - rtx-fea NonlinearDynamicStepper::step_with_dt (step at an explicit dt; step() delegates float-for-float); Fsi2Harness::advance_subcycled_with (explicit fluid dt; advance_subcycled delegates). Verified: FSI2 + FSI3 committed defaults and the noise probe digit-identical knob-off vs same-day / 2026-08-31 baselines; knob ON on the anchor u=1.00 r=1: green, zero rescues, CSV byte-identical to TWIN-1's u1.00.csv; fmt + clippy clean on the touched files. Verdict of rung A on the provocation set: REFUTED 3/3 by mechanism — every rescued interval was carried but reproduced the rejected step's motion (dt/2..dt/32 give the same jump), so the runaway lives in the coupled LOAD at the crossing, not in the time integration; the SUBCYCLE=2 closure marches through the same crossing. The knob stays, default off, as the instrument that measured this. Co-Authored-By: Claude Fable 5.1 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-02 21:36:06 -07:00
Omar SobhandClaude Fable 5.1	0f578087ce	feat(rtx-interpret): drift-gated decoder renormalisation (SAEConfig::normalize_gate) ... `normalize_gate: Option<(lo, hi)>` — when set, a decoder column is rescaled to unit norm only if its norm has left the band; columns inside are left exactly as the gradient step made them (divisor 1.0). `None` keeps per-step renormalisation, bit for bit. Why (omni-cortex D629/D630): per-step rescaling was measured doing two opposite things on the same 32-unit SAE. With it off, two runs descended cleanly to floors 3-7x LOWER than with it on — it was fighting descent. Two other runs (lr 0.01, seeds 7 and 99) diverged outright without it — it was also the clamp holding an unstable rate finite, turning a blow-up into a slow oscillation that looked like a healthy dictionary drifting. The band keeps the second role and drops the first; D630 measures whether it does both. The cold-start exemption is applied after the gate, unchanged. Test pins: in-band columns bit-identical before/after, out-of-band pulled to unit, gate None == per-step. Fixture norms sit strictly off the band edge — a hand-scaled 2.0 came out 2.0000002 in f32 and was, correctly, treated as outside. Co-Authored-By: Claude Fable 5.1 <[email protected]>	2026-09-02 20:30:25 -07:00
Omar SobhandClaude Opus 5	9575b84803	style: clear the fmt gate and two lib clippy warnings ... Deferred deliberately while the TWIN-2B/2C campaign had live marches: each march is a fresh `cargo test` invocation, so reformatting `turek_hron_fsi2.rs` mid-campaign would have forced a test-binary rebuild and cost comparability for a cosmetic gate. The family closed, so this is now free. - `cargo fmt --all` across 8 files that had drifted (including the FSI2/FSI3 harnesses touched by the UMEAN/ES override commits). - `rtx-feature-store/tests/integration_tests.rs` had trailing whitespace rustfmt refused to format around ("left behind trailing whitespace" internal error), so the whole file was being skipped; stripped it and the file formats now. - Two `unnecessary_parentheses` warnings in the rtx-transformers lib (`continual/progressive.rs`, `curriculum/mod.rs`) — these were the only rustytorch warnings surfacing through omni-cortex's workspace clippy gate, which is how they were found. No behaviour change. rtx-fsi test binaries still build. Co-Authored-By: Claude Opus 5 (1M context) <[email protected]> Claude-Session: https://claude.ai/code/session_01B1feFAQxjbCRHePUdxuNra	2026-09-02 19:15:53 -07:00
Omar SobhandClaude Fable 5	045e145962	test(rtx-fsi): RTX_FSI{2,3}_ES stiffness override for the TWIN-2 parameter sweep ... The second knob of the TWIN-2 campaign (omni-cortex docs/twin2_stiffness_campaign.md): case_from_env now also reads RTX_{prefix}_ES over the case's benchmark Young's modulus, mirroring UMEAN exactly — digit-identical with the knob unset (verified in vivo on BOTH committed defaults against same-day baselines: FSI2 and FSI3 physics lines diff-clean, only compile/wall timings moved), a loud override line naming the benchmark value and E/E0 when set. The benchmark-inflow physics-band guard in both turek_hron tests widens to a benchmark-CASE guard (u_mean AND e_s bit-identical) so no measured band ever asserts off-benchmark; machinery invariants stay asserted at every inflow and stiffness. Co-Authored-By: Claude Fable 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-09-01 20:50:58 -07:00
Omar SobhandClaude Fable 5	c08abbdc4f	test(rtx-fsi): RTX_FSI{2,3}_UMEAN inflow override for the TWIN-1 parameter sweep ... case_from_env applies the knob over the case's benchmark inflow; unset, the same f64 flows and both committed defaults are verified digit-identical in vivo (physics lines diff-clean vs same-day baselines; only wall-split timing percentages moved). Physics bands in both tests are guarded to benchmark inflow — an off-benchmark march (the sweep) asserts machinery invariants only; its bands live in the TWIN-1 composition harness (omni-cortex docs/twin_composition_campaign.md). The override prints loudly so sweep logs are self-describing. Co-Authored-By: Claude Fable 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-08-31 21:25:30 -07:00
Omar SobhandClaude Fable 5	4aec4e589d	rtx-cfd/rtx-fsi: the FlowField->clawview exporter — real FSI fields through the viewer, end to end ... The march gains RTX_FSI{2,3}_FFLD (MarchConfig::ffld_dir, off by default): every snap_every committed steps (10 when SNAPEVERY is 0), dump the committed FlowField (FlowField::save, bit-exact), the solver's own fluid-cell mask and the interface polygon as text sidecars, plus an index.csv. Reporting-only after acceptance; the FSI2 default with the knob off reproduces every printed digit of the warm-start baseline (uy 3.4921 +- 3.5109, conservation 8.25e-12). The exporter (rtx-cfd examples/ffld_to_vtk): a dump directory -> clawview-readable legacy VTK — per-snapshot 2D triangle meshes (one selectable point scalar --field p|umag|vort, the 0/1 fluid mask as integer CELL_DATA) and an optional space-time volume (--spacetime: frames stacked along z = time, prisms split to tets, POINT_DATA phi), which clawview's slice-plane animation plays as a transient movie. Verified end to end with REAL fields, not synthetic: a 29-frame FSI3 release-transient dump (t 4.0 -> 4.1, 64 MB) exported to 29 snapshots + a 984,312-tet space-time volume; the clawview server loaded both (176,320 nodes volume; 23,877-node snapshots) and served live cross-sections (time slices), vorticity isosurfaces and contours from them. The check caught two real viewer-contract constraints now encoded in the exporter: clawview's legacy-VTK path parses CELL_DATA scalars as INTEGER markers only, and supports exactly ONE point scalar (all POINT_DATA blocks append into `phi`) — multi-field snapshots need the .clwv route, out of scope here. Closes the top open thread of the fifteenth-session handoff (the FSNP/FlowField->viewer exporter); the mesh-repo push and the claw-gds cargo feature remain clawview-side items. Co-Authored-By: Claude Fable 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-08-30 14:45:54 -05:00
Omar SobhandClaude Fable 5	4d2cede7bc	rtx-cfd: warm-start the first corrector's pressure-correction solve — measured 2.5x fewer PCG iterations where it counts ... The MG smoother (neighbour_sum + PCG, 34-40% of the fluid) was the remaining fluid cost. Measured in vivo (FSI2 default, temporary iteration counters): 2.7-3.0 PCG iterations/solve from a zero initial guess. The correction field is temporally correlated step to step, so project() now seeds the FIRST corrector's solve from the previous step's p' (current fluid cells only — the p'=0 invariant on non-fluid cells survives the copy-back); later correctors and the SOR fallback start from zero exactly as before. The corrector restriction is measured, not guessed: the all-correctors draft cut the rigid phase 2.8 -> 1.27 iters/solve but cost 3.9/solve in the coupled phase (baseline 2.56) — corrector 2 solves for a much smaller correction and corrector 1's full-magnitude p' is a WORSE guess than zero there. First-corrector-only: rigid 2.8 -> ~1.1 (best of the three variants), coupled 2.62 ~= baseline. Total PCG iterations on the FSI2 default: 55,088 -> 32,836 (1.68x fewer). Wall: FSI2 default 225 -> 172 s, FSI3 default 343 -> 249 s. Session cumulative (banded LU + indexed SDF + warm start): FSI2 524 -> 172 s (3.0x), FSI3 944 -> 249 s (3.8x). This is a TOLERANCE-LEVEL solver-path change (each projection reaches the same true-residual stop from a different start), and the FSI3 release-window pins fired for the third and fourth time across the drafts — completing the picture: EVERY windowed observable of the [4.0, 4.2] release transient is branch-sensitive (four measured branches now recorded in the test: uy mid 10.77/6.02/2.91/8.88, amp 23.6/25.2/24.6/19.9, ux mid -2.90/-2.91/-2.56/-1.82, retries 2/0/1/1). The release bands are re-pinned as gross-physics tripwires around the measured scatter; the load-bearing regression pins for solver changes are the settled-cycle study bands, whose re-verification under this change is launched (verdicts to solver_status.md). Protocol: FSI2 default green (uy 3.4921 in-band), FSI3 default green under the re-pinned release bands (deterministic across two runs), FSI1 green, noise-probe floors identical, rtx-cfd suite + rtx-fsi quick tests green. Co-Authored-By: Claude Fable 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-08-30 08:54:27 -05:00
Omar SobhandClaude Fable 5	328f65233e	rtx-interpret: decoder cold-start window for reinitialized SAE units ... reinitialize_encoder_neuron_cold / SAETrainer::reinitialize_neuron_cold exempt a freshly reset unit's decoder column from per-step normalize_decoder for cold_steps training steps, so its small random init is not blown up to unit norm before it has learned anything — the mechanism omni-cortex's D605 refutation left as the prime suspect, now testable (D620 downstream). Runtime-only state, not carried through checkpoints (documented); cold_steps 0 is exactly the plain reinit, and existing entry points delegate with 0. Co-Authored-By: Claude Fable 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01B1feFAQxjbCRHePUdxuNra	2026-08-30 08:31:08 -05:00
Omar SobhandClaude Fable 5	6c48e53998	rtx-cfd: indexed polygon SDF — bit-identical queries, the fluid's measured hot function cut ... The 2026-08-30 fluid profile (symbolized samples, rigid AND coupled phases of the FSI3 default) attributed the fluid step to the function: polygon_signed_distance 51% rigid / 35% coupled — the embedded mask rebuild and its ghost reconstruction walk every edge of the ~150-vertex interface polygon for every cell-centre and face query, every step. (Also measured, refuting the parked consolidation: Level::new — the MG hierarchy build — is 0.5-0.7% in BOTH phases; caching it would buy nothing. The MG smoother at 34-40% is the honest remaining fluid cost.) PolygonSdf (solvers/incompressible/polygon_sdf.rs): a binned edge index whose query is BIT-IDENTICAL to polygon_signed_distance by construction — per-edge distances use the same float ops, the ring search provably visits a superset of the argmin (convex-projection lower bound sqrt(d_out^2 + ((r-1)b)^2)), and parity XORs the same ray tests over exactly the straddling edges (y-binned). Equality is ASSERTED, not assumed: tests compare to_bits against the brute force over ~40k adversarial points (flag-like walks, random polygons with degenerate zero-length edges, horizontal-edge/vertex-y rays). Wired into EmbeddedBody::polygon and the FSI harness's shared geometry (rebuilt per set_geometry, ~microseconds for 150 edges). Verification — the bar for a bit-exact change is digit identity, and it holds: FSI2 and FSI3 committed defaults reproduce EVERY printed digit of the banded-LU baseline logs (uy 3.7732±3.7920 / 6.0229± 25.2190 mm, conservation 8.26e-12 / 1.49e-12, rigid drags 121.4 / 426.9); rtx-cfd full suite 0 failures; rtx-fsi lib/piston/transfer/ FSI1 green. The study pins need no re-run: the trajectories are unchanged by construction and confirmed by measurement. Wall clock: FSI2 rigid 323 -> 167 s (1.93x), whole default 400 -> 225 s; FSI3 rigid 420 -> 250 s (1.68x), whole default 539 -> 343 s. Cumulative with the banded LU this session: FSI3 default 944 -> 343 s (2.75x), FSI2 524 -> 225 s (2.33x). Co-Authored-By: Claude Fable 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-08-30 07:47:42 -05:00
Omar SobhandClaude Fable 5	0b4f306ed1	rtx-fea: reduced Newmark (mor::dynamic) + the phase-4a offline replay — the ≥10x gate is REFUTED by measurement at the validated resolution ... The dynamic layer over ReducedNonlinearModel: reduced consistent mass V'MV (full element sum, never ECSW-sampled — ECSW weights are trained on internal-force virtual work and would conserve the wrong inertia), reduced_force_and_jacobian exposed (solve() refactored onto it), and ReducedNewmark mirroring NonlinearDynamicStepper::newmark_newton in reduced coordinates (same predictor, residual, tangent shape; no rescue ladder by design — a reduced Newton death is a finding). TDD (tests/reduced_newmark.rs): identity-basis march reproduces the full stepper to 2.4e-14 over 15 steps (both Newton loops tightened to 1e-10 so only solver rounding separates them); rigid-translation reduced mass = rho*A to 1e-9; a 6-mode POD basis tracks its training trajectory at 4.2e-4 rms against a 1.0e-4 projection floor. Phase 4a (fsi3_ecsw_offline.rs, fsi3_reduced_newmark_replay, env-gated): reduced Newmark replay of the harvested FSI3 trajectory at record cadence (dt_rec = 5x march dt), driven by the recorded end-of-step loads. Measured, m=12/20: - COST (dt-independent, the verdict): 3,068/3,580 us/step at 4.6/5.0 Newton iters — 2.0-2.3x the banded full-order structural step (7,200 us/pass, bandedlu_fsi3_ny62_t85). The >=10x gate needs <=720 us/step; one reduced eval alone costs ~640 us because phase 2 refuted hyperreduction (every eval loops all 70 elements). The gate arithmetic is closed: reduced Newton needs >=2 evals, capping the ROM at ~5x. THE CAMPAIGN GATE (pinned cycle bands at >=10x structural speedup) CANNOT BE MET at the validated resolution. - TRACKING at record cadence diverges in the release transient (dies t=4.35-4.45) — and the RTX_REPLAY_IDENTITY control dies EARLIER (t=4.13) in the exact subspace: the death is the 5x-coarse integration + aliased loads, NOT the reduction. The record-cadence replay cannot judge subspace dynamics; the projection floor (1.1e-3 at m=12) remains the honest subspace statement. Campaign verdict to be recorded in omni-cortex in the pre-registered words. Co-Authored-By: Claude Fable 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01X2GmJXeQ2njUecEKiJZ1G2	2026-08-30 07:07:49 -05:00