Iterations-CFD

GPU 상주 유한체적 전산유체역학 솔버

주식회사 이터레이션즈 · 협력 및 기여 주식회사 메테오시뮬레이션 · Rust 호스트 + CUDA 커널 · English

시간 적분 루프 전체가 GPU에 머무르도록 설계된 비정렬 격자 유한체적 CFD 솔버입니다. 메쉬와 필드를 한 번 업로드한 뒤에는 시간 루프 안에서 장치 메모리 할당도, 필드 데이터의 호스트 전송도 발생하지 않습니다. 단일 GPU 전용이며, 쓰임새는 비압축성·저마하 유동입니다 — RANS/LES/하이브리드 난류, 부력 플룸, 가변밀도 저마하 유동, 2상 VOF, 켤레 열전달, 면대면 복사, 라그랑주 분무, 팬과 다공성 점프를 갖춘 환기 및 데이터센터 기류. 수치 코어 전체를 공개 문헌으로부터 직접 구현했으며 모든 수식은 rust/SPEC-LIT.md에 원논문 인용과 함께 명세되어 있습니다. 검증은 인위해법(MMS), 해석해, 공개 벤치마크를 사용합니다. 비교가 불충분합니다. 많은 도움 부탁드립니다.

Rust 1.85 호스트에 CUDA C++ 커널, 배정밀도 기본(single 기능으로 단정밀도), 대상은 NVIDIA GPU 한 장. 직접 의존성은 일곱 크레이트 — cudarc, jsonc-parser, schemars, serde, serde_json, serde_path_to_error, thiserror(rust/Cargo.toml의 [dependencies]; 빌드 스크립트에는 anyhow가 하나 더 붙습니다) — 이며, 해석된 그래프는 27개 패키지(이 크레이트에 26개를 더한 수)로 전부 NOTICE에 각자가 실리는 조항과 함께 이름이 올라 있습니다. AMGX는 여전히 선택 기능입니다.

「공공」은 기관이 아니라 연구의 목적을 가리킵니다. 소방·의료·보건·안전·재난·환경 보호를 목적으로 하는 연구는 정부출연연구기관에서 수행하더라도 무료입니다. 반면 특정 제품이나 산업 이전을 목적으로 하는 기술개발 — 전기차, 철도·추진, 항공기 엔진 기술 등 — 은 정출연에서 수행하더라도 상업적 이용입니다. 실무 기준은 하나입니다: 기술료 징수 대상이거나, 기업을 위해 또는 결과를 기업에 이전하는 사업으로 수행되는 연구는 유상 입니다. 대학·학교 및 그 소속 연구소는 교육기관으로서 무료입니다.

30일 시험 기간은 사람이 아니라 회사 단위입니다. 업무 목적으로 쓰신다면 인원 수와 무관하게 회사 전체에 시험 기간이 한 번 주어집니다. CFD는 자기 케이스·자기 격자·자기 하드웨어에서 돌려보지 않고는 도입을 결정할 수 없는 소프트웨어이므로, 이 기간은 평가를 위한 것입니다.

기여를 되돌려 주시는 경우 — Blue Oak 1.0.0, Apache 2.0, MIT, 2-clause BSD 중 하나로 — 그 개발은 상업적 이용으로 계산되지 않습니다.

적용 범위가 불분명한 경우 문의해 주시기 바랍니다. 실질이 상업적이지 않은 이용에 대해서는 무상 또는 감면 라이선스를 검토합니다.

라이선스 문의: simul@msimul.com

저작권자이자 라이선서는 주식회사 이터레이션즈(Iterations Co., Ltd.)이며, 주식회사 메테오시뮬레이션(Meteo Simulation Co., Ltd.)이 협력 및 기여자입니다. 전문은 LICENSE, 상업 라이선스의 범위와 절차는 LICENSING.md, 서드파티 고지는 NOTICE를 참조하십시오.

Prosperity는 비상업 라이선스이지 오픈소스 라이선스가 아닙니다. OSI 승인 라이선스가 아니며, 상업적 이용을 제한하므로 그렇게 부르지 않습니다.

아래 각 줄은 기록용 머신에서 찍힌, 날짜가 붙은 측정값입니다 — NVIDIA GeForce RTX 5070 Ti (sm_120), CUDA 13.3, 배정밀도 — 줄마다 측정 날짜와 기록된 곳을 함께 적어 두었습니다.

cargo test --release   1,774 passed, 0 failed, 6 ignored   (모든 타깃 합계, 18개 스위트; 2026-09-08)
                       1,645 passed, 0 failed, 4 ignored   (lib 크레이트만; 2026-09-08)

ofgpu-validate         1021 / 1026 checks passed           (2026-09-27, SPEC-LIT §95.11)
                       실패하는 다섯 행: Gate 95-A의 네 행(5:1·10:1 캔틸레버, §109.7)과
                       Gate 95-G의 LE10 링 차수 행(§95.11)
                       이어서 판정이 MISSES 또는 OPEN으로 등록된 게이트 전부를 이름으로 부르는 목록

그 목록은 손으로 유지되는 것이 아니라, 게이트가 자신의 판정을 보고하는 바로 그 지점에서 들어가는 레지스트리로부터 생성됩니다(SPEC-LIT §69). 판정을 출력하는 것과 등록하는 것이 같은 호출이므로 등록된 판정은 전부 매 실행 이름으로 불리며, 하나가 더 생기더라도 목록에서 빠질 수 없습니다. ofgpu-validate가 실행하는 1026행 가운데 다섯 행은 실패합니다 — 위 블록이 그 다섯 행의 이름을 적어 둡니다 — 그 외의 모든 행은 통과합니다. 그것은 "이 프로젝트가 비교하는 모든 발표된 벤치마크를 재현한다"와 다른 진술이며, 둘을 혼동해서는 안 됩니다.

설치부터 첫 케이스, 격자, 솔버 선택, 수렴 판독, 결과 보기, 그리고 자주 만나는 오류의 원인까지 — 쓰는 법을 다루는 문서는 docs/GUIDEBOOK.md 하나입니다. 영어판은 docs/GUIDEBOOK.en.md.

가장 많이 틀리는 곳부터 보시려면 6절 「솔버 고르기」입니다 — ofgpu-k-epsilon, ofgpu-k-omega, ofgpu-sa는 속도장을 얼린 채 난류 방정식만 풉니다.

요구사항: Rust 1.85 이상, Visual Studio 2022 (C++ 워크로드), CUDA Toolkit 13.x. build.rs가 vcvars64.bat으로 MSVC 환경을 구성하고 모든 .cu를 PTX가 아닌 CUBIN으로 컴파일합니다.

cd rust
cargo build --release
cargo run --release --bin ofgpu-generate-mesh -- channel ..\cases\channel 200 120 1
cargo run --release --bin ofgpu-k-epsilon     -- ..\cases\channel -iters 4000 -check 400

검증 전체는 cargo run --release --bin ofgpu-validate입니다. 나머지 열아홉 개 실행 파일(ofgpu-lowmach, ofgpu-vof, ofgpu-cht, ofgpu-datacentre, ofgpu-decompose, 벤치마크들), 케이스 파일 형식, 명령행 옵션은 사용자 안내서에 있습니다. 케이스는 JSONC 한 파일 또는 OpenFOAM ASCII 케이스 디렉터리로 읽고 씁니다 — 후자는 ParaView·foamToVTK 같은 기존 도구와의 상호운용을 위한 것이며, meteor-cfd는 OpenFOAM의 어떤 부분과도 링크하지 않고 그 소스를 포함하지 않습니다.

형상 하나와 명령 두 줄이 cases/에 들어 있습니다. 격자와 결과는 없습니다 — 합쳐 1 GB가 넘고, 그것을 만드는 것이 이 프로그램이 하는 일입니다.

cd cases
.\racecar.cmd     # 격자 몇 분(모든 코어 병렬) + 솔브 약 40초(GPU)

자세한 것은 cases/racecar.md.

Claude Code, Claude Desktop 등 Model Context Protocol을 말하는 클라이언트가 이 솔버들을 직접 실행하게 할 수 있습니다. 의존성 없는 파일 하나입니다.

claude mcp add ofgpu -- node mcp\server.mjs

사용법과 도구 목록, 무엇을 못 하게 막아 두었는지는 mcp/README.md.

• STEP에서 격자로, 그리고 Fluent로. CAD가 준 STEP 형상은 tools/mesh/step_mesh.py(설정 JSON 하나)로 Gmsh 격자를 만들고, ofgpu-convert-mesh가 그 .msh를 케이스의 constant/polyMesh와 ANSYS Fluent 메쉬로 한 번에 씁니다. 같은 두 단계가 Studio에는 mesh_from_step·mesh_to_fluent 도구로 들어 있습니다 — 가이드북 §5.

• MPI·다중 GPU 없음. 분할·헤일로·분할 불변 축약·분산 PCG/PBiCGStab이 모두 구현되고 게이트되어 있으나(§71–§73) 한 장의 카드 위 한 프로세스에서 돕니다. 통신 라이브러리를 링크하지 않으며 강한 확장성 수치도 발표하지 않습니다.
• 압축성·천음속 없음. 밀도 가중 시간미분은 VOF에서 쓰이지만 압력 방정식은 비압축성입니다.
• 유한율(아레니우스) 화학 없음 — 경직 ODE 적분기도, 야코비안도, 반응 메커니즘도 없습니다.
• Crank–Nicolson은 저완화 방정식과 함께 쓸 수 없습니다 — 묵시적으로 Euler로 대체하지 않고 사유를 오류로 보고합니다.
• 복사는 면대면(surface-to-surface)만 있습니다. §49/§50은 투명 매질을 가로지르는 회색 확산 교환을 풀며, 체적 안에서 흡수·방출·산란하는 것은 없습니다. radiationModel 선택기는 viewFactor(그리고 같은 모델의 고유 표기 s2s)만 인식하고 그 밖의 값은 인식되지 않는 설정으로, 인식되는 집합을 함께 제시하며 거부됩니다 — 조용한 대체는 없습니다(§13.4). constant/radiationProperties가 존재하는데 radiationModel을 이름하지 않는 경우도 같은 이유로 거부됩니다: 그 항목에는 기본값이 없으며, 추측하는 리더는 케이스가 묻지 않은 질문에 답하는 셈이기 때문입니다.
• 면대면 복사·화학종 수송·라그랑주 스프레이는 케이스 형식이 없습니다. 셋 다 라이브러리 API로 명세되고 게이트되어 있으나 어느 드라이버 바이너리도 케이스 파일에서 밀폐공간·화학종 집합·스프레이를 읽지 않습니다(§50.12, §13.4.2).
• ofgpu-generate-mesh가 p와 T를 쓰지 않는 프리셋은 damBreak 하나입니다. 나머지 — channel, cavity, step, big과 부력 쌍 plume/room — 은 블록·컷셀 두 생성 경로 모두 0/에 p와 T까지 쓰므로 운동량을 푸는 드라이버도 갓 생성된 케이스를 그대로 받아 돌립니다. damBreak는 2상 경로라 0/이 alpha.water와 p_rgh를 담고, 여기에 ofgpu-lowmach를 들이대면 여전히 has no p field로 거절합니다. 컷셀 경로가 ofgpu-sa용 nuTilda를 쓰지 않는 것도 그대로입니다 — 그 필드는 블록 격자 경로만 더 씁니다.
• 파셀은 스플래시 자식 액적을 만들지 않고, 막(film) 수송이 없으며, 복사를 흡수하지 않습니다(§78.11, §68.13).
• 적응 세분화는 어떤 솔버에도 연결되어 있지 않습니다 — 면 플럭스를 옮기지 않고, 적응 후 압력 투영도 없습니다.
• 사면체 격자의 다면체 쌍대 변환은 없습니다. 사면체 격자를 읽고 그 위에서 푸는 것은 됩니다 — Gmsh MSH 4.1 리더(io/msh.rs)가 Tet·Hex·Prism·Pyramid를 받고 메쉬 모델이 면 기반이라 읽힌 사면체는 일반 다면체 메쉬로 이미 다뤄집니다. 없는 것은 절점마다 그 둘레의 사면체를 하나의 다면체로 합치는 쌍대 격자 연산으로, Fluent의 다면체 변환과 OpenFOAM의 polyDualMesh가 사면체 격자에서 셀 수를 줄이고 기울기 재구성을 개선하려는 그것입니다 — polyDualMesh·dual_mesh·dualMesh 어느 기호도 트리에 없습니다. 작게 추가할 수 없는 이유는 둘입니다: SPEC-LIT은 이미 모든 면이 평면이고 모든 셀이 볼록하다고 가정하는데("assumes every face planar") 쌍대 격자의 면은 일반적으로 둘 다 아니므로 그 가정을 다시 검토하는 일에서, 그리고 경계 절점의 쌍대 셀이 열려 있다는 경계 처리의 결정에서 작업을 시작해야 변환기가 됩니다.
• AMGX는 기본 비활성이며, 비활성 상태에서도 선택기가 AMGX를 "unavailable"로 명시 보고합니다.
• DES 계열이 발표된 박리 유동 통계를 재현한다고 주장하지 않으며, Spalart-Allmaras의 TMR 평판 게이트는 실행하지 않습니다(§57.12, §56.11).

그리고 빗나가는 게이트 두 개. ofgpu-validate가 매 실행 이름으로 부르며, 아래는 그 이름 그대로입니다.

각 검증 절이 어떤 종류의 검증이고 무엇에 견준 것인지(논문·벤치마크 수치·상관식·공개 데이터·저장소의 기록), 그리고 그 자료의 접속 주소(DOI/URL)는 docs/GUIDEBOOK.md §11 「검증 항목 하나하나」의 표에 있습니다.

판정이 OPEN인 것이 여섯 개 더 있으며 같은 목록의 둘째 그룹으로 출력됩니다. 셋은 §32.4의 평판 채널을 측정값이 아니라 상관식(Gnielinski 1976)에 견준 것이고, 넷째 78-D는 발표된 두 스플래시 기준이 Weber 수로 4.78배 서로 어긋나기 때문이며, 다섯째 §88.10 Gate 88-T는 T3A 평판의 측정된 개시 Re_x를 구하지 못해 비교를 닫지 못했기 때문이고, 여섯째 §90.10 Gate 90-T는 2015 모형의 T3A 개시 위치로서 88-T와 같은 이유입니다.

수치 기법과 모형의 출처입니다. 절 번호는 SPEC-LIT 기준이며, 각 항목은 SPEC-LIT이 실제로 인쇄한 서지사항만 담습니다 — 제목·호수·면수가 없는 곳은 SPEC-LIT이 그것 없이 인용했기 때문입니다. 읽지 않은 출처(유료·접근 불가·의도적으로 열지 않은 것)와 ofgpu-validate가 이름으로 인쇄하는 판정은 해당 줄에 표시했습니다.

• Jasak, H. (1996). Error Analysis and Estimation for the Finite Volume Method with Applications to Fluid Flows. PhD thesis, Imperial College London. http://hdl.handle.net/10044/1/8335 — §2, §3, §74, §82
• Moukalled, F., Mangani, L., & Darwish, M. (2016). The Finite Volume Method in Computational Fluid Dynamics. Springer. — §2, §3, §11, §74, §82
• Ferziger, J. H., & Perić, M. Computational Methods for Fluid Dynamics, 3rd ed. Springer (2002). — §2.4, §3.3, §11.1, §11.5, §74, §82
• Patankar, S. V. (1980). Numerical Heat Transfer and Fluid Flow. Hemisphere. ISBN 0-89116-522-3. — §3.4, §5.2, §18, §46, §50, §56, §68

• Warming, R. F., & Beam, R. M. (1976). AIAA Journal, 14, 1241–1249. — §11.2
• Leonard, B. P. (1979). Computer Methods in Applied Mechanics and Engineering, 19, 59–98. — §11.3
• Leonard, B. P. (1991). Computer Methods in Applied Mechanics and Engineering, 88, 17–74. — §7 (the NVD framework)
• Khosla, P. K., & Rubin, S. G. (1974). Computers & Fluids, 2, 207–209. — §11.1
• Jasak, H., Weller, H. G., & Gosman, A. D. (1999). International Journal for Numerical Methods in Fluids, 31, 431–449. — §11.6
• Sweby, P. K. (1984). SIAM Journal on Numerical Analysis, 21, 995–1011. — §7 (the TVD framework)
• van Leer, B. (1977). Journal of Computational Physics, 23. — §7 (the limiter)
• van Leer, B. (1979). — §7 (the MUSCL limiter)
• van Albada, G. D., van Leer, B., & Roberts, W. W. (1982). Astronomy and Astrophysics, 108. — §7
• Roe, P. L. (1986). — §7 (minmod and Superbee)
• Darwish, M., & Moukalled, F. (2003). International Journal of Heat and Mass Transfer, 46, 599–611. — §7 (the gradient ratio on an unstructured mesh)

• Barth, T. J., & Jespersen, D. C. (1989). The design and application of upwind schemes on unstructured meshes. 27th Aerospace Sciences Meeting, AIAA 89-0366. DOI 10.2514/6.1989-366 — §12.2, §75
• Venkatakrishnan, V. (1993). AIAA Paper 93-0880. — §12.2 (the smooth variant)

• Crank, J., & Nicolson, P. (1947). Proceedings of the Cambridge Philosophical Society, 43, 50–67. — §13.1

• Patankar, S. V., & Spalding, D. B. (1972). — §5.2 (SIMPLE)
• Van Doormaal, J. P., & Raithby, G. D. (1984). — §5.3 (SIMPLEC)
• Issa, R. I. (1986). Journal of Computational Physics, 62, 40–65. — §5.4, §14
• Rhie, C. M., & Chow, W. L. (1983). AIAA Journal, 21, 1525–1532. — §5.1, §52

• Launder, B. E., & Spalding, D. B. (1974). Computer Methods in Applied Mechanics and Engineering, 3, 269–289. — §6.1, §6.4
• Wilcox, D. C. Turbulence Modeling for CFD. DCW Industries. — §6.2 (the 1988 form); §5.4 there is the source of the Favre-averaged dilatation terms
• Menter, F. R. (1994). AIAA Journal, 32, 1598–1605. — §6.3
• Menter, F. R., Kuntz, M., & Langtry, R. (2003). Turbulence, Heat and Mass Transfer, 4. — §6.3 (the 2003 revision)
• Launder, B. E., & Sharma, B. I. (1974). Letters in Heat and Mass Transfer, 1, 131–138. — §33
• Patel, V. C., Rodi, W., & Scheuerer, G. (1985). AIAA Journal, 23, 1308. — §33 (the review of the low-Reynolds-number family)
• Shih, T.-H., Liou, W. W., Shabbir, A., Yang, Z., & Zhu, J. (1995). Computers & Fluids, 24, 227–238. Read as NASA TM-106721 / ICOMP-94-21 (August 1994), https://ntrs.nasa.gov/citations/19950005029, a US government work in the public domain; the journal version is paywalled and was not read. — §40
• Yakhot, V., Orszag, S. A., Thangam, S., Gatski, T. B., & Speziale, C. G. (1992). Physics of Fluids A, 4, 1510–1520. Read as ICASE Report 91-65 / NASA CR-187611 (1991), https://ntrs.nasa.gov/citations/19910021152, US government-sponsored, public domain via NTRS. — §41
• Yakhot, V., & Orszag, S. A. (1986). Journal of Scientific Computing, 1, 3–51. — §41 (the original renormalisation-group derivation)
• Reynolds, W. C. AGARD Report 755 (1987). — §40 (the realizability constraints — positivity of the normal stresses, the Schwarz inequality — that the variable C_mu is constructed to satisfy)
• Lumley, J. L. (1978). Advances in Applied Mechanics, 18, 123–176. — §40 (realizability as a modelling principle)
• Pope, S. B. Turbulent Flows (2000), §10.4. — §40
• Spalart, P. R., & Allmaras, S. R. AIAA Paper 92-0439 (1992); also La Recherche Aérospatiale, 1 (1994), 5–21. — §56 (the original)
• Allmaras, S. R., Johnson, F. T., & Spalart, P. R. (2012). Modifications and Clarifications for the Implementation of the Spalart-Allmaras Turbulence Model. ICCFD7-1902. https://www.iccfd.org/iccfd7/assets/pdf/papers/ICCFD7-1902_paper.pdf — a freely distributed conference paper, the copy actually read, and the implementation reference. — §56
• NASA / Turbulence Modeling Benchmarking Working Group. Turbulence Modeling Resource — The Spalart-Allmaras Turbulence Model. https://tmbwg.github.io/turbmodels/spalart.html — US government-authored DOCUMENTATION, not source; quoted to the printed digit. — §56
• Rumsey, C. L., & Spalart, P. R. (2009). AIAA Journal, 47, 982–993. — §56 (why the free-stream nu~/nu matters)
• Menter, F. R., Smirnov, P. E., Liu, T., & Avancha, R. (2015). Flow, Turbulence and Combustion, 95, 583–619. DOI 10.1007/s10494-015-9622-4. Paywalled and NOT read; every digit comes from the NASA/TMBWG Turbulence Modeling Resource page below. — §90
• NASA / Turbulence Modeling Benchmarking Working Group. Turbulence Modeling Resource — SST-2003-Menter-Gamma-2015. https://tmbwg.github.io/turbmodels/menter_gamma_3eqn.html — US government-authored DOCUMENTATION, not source; fetched and read 2026-09-05, quoted to the printed digit, including the constant it prints in no equation (§90.3). — §90

• Smagorinsky, J. (1963). Monthly Weather Review, 91, 99–164. — §6.5
• Deardorff, J. W. (1970). Journal of Fluid Mechanics, 41, 453–480. — §16.1
• Deardorff, J. W. (1980). Boundary-Layer Meteorology, 18, 495–527. — §6.5 (the model FDS uses)
• Nicoud, F., & Ducros, F. (1999). Flow, Turbulence and Combustion, 62, 183–200. — §6.5 (WALE)
• van Driest, E. R. (1956). Journal of the Aeronautical Sciences, 23, 1007–1011. — §16.4
• Scotti, A., Meneveau, C., & Lilly, D. K. (1993). Physics of Fluids A, 5, 2306–2308. — §16.3

• Spalart, P. R., Jou, W.-H., Strelets, M., & Allmaras, S. R. (1997). Comments on the feasibility of LES for wings, and on a hybrid RANS/LES approach. In Advances in DNS/LES, Greyden Press, 137–147. — §57 (DES97)
• Shur, M., Spalart, P. R., Strelets, M., & Travin, A. (1999). Engineering Turbulence Modelling and Experiments 4, 669–678. — §57 (the C_DES = 0.65 calibration on the SA background, at Delta = h_max)
• Strelets, M. AIAA Paper 2001-0879. — §57 (SST-DES, the k-equation dissipation form)
• Spalart, P. R., Deck, S., Shur, M., Squires, K. D., Strelets, M., & Travin, A. (2006). Theoretical and Computational Fluid Dynamics, 20, 181–195. — §57 (DDES: r_d, f_d, and the grid-induced separation they fix)
• Shur, M., Spalart, P. R., Strelets, M., & Travin, A. (2008). International Journal of Heat and Fluid Flow, 29, 1638–1649 — IDDES. Paywalled and NOT read; §57's IDDES equations come from the two open-access restatements below. — §57
• Gritskevich, M. S., Garbaruk, A. V., Schütze, J., & Menter, F. R. (2012). Flow, Turbulence and Combustion, 88, 431–449 — the SST-background recalibration. Paywalled and NOT read: C_dt1 = 20, c_t = 1.87 and c_l = 5.0 are carried from a design note's reading of it, defaulted, printed in the banner, and not independently verified. — §57
• Herr, F., Radespiel, R., & Probst, A. (2023). Improved Delayed Detached Eddy Simulation with Reynolds-Stress Background Modeling. arXiv:2301.07223v2; published in Computers & Fluids, 265 (2023) 106014. Appendix A is a complete restatement of the IDDES formulation and is where (57.9)–(57.16) come from, equation by equation. Open access, fetched and read in full. — §57
• Savino, A., Griffin, K., Lee, S., Vijayakumar, G., Wu, S., & Sprague, M. (2026). Improving boundary-layer separation prediction by an IDDES turbulence model using a pressure-gradient sensor. arXiv:2603.08875, arXiv non-exclusive distribution licence. Section 2 states SST-IDDES and is where C_DES1 = 0.78, C_DES2 = 0.61, C_w = 0.15 and the simplified filter width (57.18) come from. Open access, read in full. — §57
• Nikitin, N. V., Nicoud, F., Wasistho, B., Squires, K. D., & Spalart, P. R. (2000). Physics of Fluids, 12, 1629–1632. — §57 (the log-layer mismatch f_e exists to remove)
• Spalart, P. R. (2009). Annual Review of Fluid Mechanics, 41, 181–202. — §57 (the review)
• Fröhlich, J., Mellen, C. P., Rodi, W., Temmerman, L., & Leschziner, M. A. (2005). Journal of Fluid Mechanics, 526, 19–66. — §57.12, the periodic-hill gate at Re_b = 10 595: named and NOT run

• Spalding, D. B. (1961). Journal of Applied Mechanics, 28, 455–458. — §6.4, §15.1
• Cebeci, T., & Bradshaw, P. Momentum Transfer in Boundary Layers, Hemisphere (1977). — §15.3 (rough-wall boundary layers; Nikuradse's sand-grain data underlies the constants)
• Jayatilleke, C. L. V. (1969). Progress in Heat and Mass Transfer, 1, 193–330. — §29.3 (the sublayer-resistance correction to the thermal log law)
• Werner, H., & Wengle, H. (1991). Large-eddy simulation of turbulent flow over and around a cube in a plate channel. 8th Symposium on Turbulent Shear Flows. — §30.1
• Tucker, P. G. (1998). Applied Mathematical Modelling, 22, 293–305. — §6.6 (the Poisson wall-distance approach)
• Dittus, F. W., & Boelter, L. M. K. (1930). University of California Publications in Engineering, 2, 443, reprinted in International Communications in Heat and Mass Transfer, 12 (1985) 3. — §32.3. Conventionally quoted at ±20–25 %.
• Gnielinski, V. (1976). International Chemical Engineering, 16, 359. — §32.3, ±10 %. OPEN × 3. §32.4's three channel verdicts are held against this correlation and none closes at the shipped default: verdict 1 (absolute prediction, resolved leg, at the Petukhov smooth-pipe f), and verdict 2 (Reynolds analogy, at each leg's own MEASURED f) on both the wall-function leg and the resolved leg. Every such statement must name which f Gnielinski was evaluated at — §32.3's own rule.
• Kays, W. M. (1994). ASME Journal of Heat Transfer, 116, 284–295. — §32.5, §37 (that Pr_t rises towards a wall: a named hypothesis with a mechanism and a direction for the §32.4 verdicts, and nothing here has measured it)

• Rehm, R. G., & Baum, H. R. (1978). The equations of motion for thermally driven, buoyant flows. Journal of Research of the National Bureau of Standards, 83, 297–308. — §9, §25, §77
• Majda, A., & Sethian, J. (1985). Combustion Science and Technology, 42, 185. — §25
• Spiegel, E. A., & Veronis, G. (1960). Astrophysical Journal, 131, 442. — §9 (the ΔT/T << 1 requirement, which a strongly heated plume does not meet)
• Rodi, W. (1987). Journal of Geophysical Research, 92, 5305–5328. — §17
• Henkes, R. A. W. M., van der Vlugt, F. F., & Hoogendoorn, C. J. (1991). International Journal of Heat and Mass Transfer, 34, 377–388. — §17

• Carslaw, H. S., & Jaeger, J. C. Conduction of Heat in Solids, 2nd ed., Oxford University Press (1959), ch. I. ISBN 0-19-853368-3. — §46 (the anisotropic solid, and the affine transformation that reduces div(K grad T) to lap T)
• Aavatsmark, I. (2002). An introduction to multipoint flux approximations for quadrilateral grids. Computational Geosciences, 6, 405–432. DOI 10.1023/A:1021291114475 — §46.4 (the rigorous full-tensor treatment, and therefore the reason §46.4 refuses rather than approximating)
• Lipnikov, K., Shashkov, M., Svyatskiy, D., & Vassilevski, Yu. (2007). Journal of Computational Physics, 227, 492–512. DOI 10.1016/j.jcp.2007.08.008 — §46.4 (the nonlinear monotone alternative, named in the same refusal)
• Yovanovich, M. M. (2005). IEEE Transactions on Components and Packaging Technologies, 28, 182–206. DOI 10.1109/TCAPT.2005.848483 — §46.3 (the layered-stack conductivities the Wiener pair homogenises), §47.12 (the review, and the gas-gap and elastic regimes §47.12 omits)
• Giles, M. B. (1997). International Journal for Numerical Methods in Fluids, 25, 421–436. DOI 10.1002/(SICI)1097-0363(19970830)25:4<421::AID-FLD557>3.0.CO;2-J — §47 (the Godunov–Ryabenkii normal-mode analysis behind the classical "Dirichlet on the fluid, Neumann on the solid" rule)
• Meng, F., Banks, J. W., Henshaw, W. D., & Schwendeman, D. W. (2017). A stable and accurate partitioned algorithm for conjugate heat transfer. Journal of Computational Physics, 344, 51–85. DOI 10.1016/j.jcp.2017.04.052 — §47.7, Theorem 1: the amplification factor that is the reason Dirichlet–Neumann partitioning is not implemented here
• Henshaw, W. D., & Chand, K. K. (2009). Journal of Computational Physics, 228, 3708–3741. DOI 10.1016/j.jcp.2009.02.007 — §47 (Robin coefficients can always be chosen so the sub-time-step iteration converges)
• Verstraete, T., & Scholl, S. (2016). International Journal of Heat and Mass Transfer, 101, 852–869. DOI 10.1016/j.ijheatmasstransfer.2016.05.041 — §47 (the numerical Biot number, and FFTB's instability above Bi = 1)
• Gander, M. J. (2006). Optimized Schwarz methods. SIAM Journal on Numerical Analysis, 44, 699–731. DOI 10.1137/S0036142903425409 — §47 (the physical series conductance is the zeroth-order optimised-Schwarz weight; the optimal weight is a non-local operator)
• Cooper, M. G., Mikic, B. B., & Yovanovich, M. M. (1969). Thermal contact conductance. International Journal of Heat and Mass Transfer, 12, 279–300. DOI 10.1016/0017-9310(69)90011-8 — §47.12 (the plastic-deformation contact conductance correlation)
• de Vahl Davis, G. (1983). Natural convection of air in a square cavity: a bench mark numerical solution. International Journal for Numerical Methods in Fluids, 3, 249–264. DOI 10.1002/fld.1650030305 — §59.8, the fluid-only anchor run first, because a conjugate answer built on an unvalidated buoyant solver measures nothing. The primary is paywalled; its four numbers are quoted from Qi et al., Nanoscale Research Letters, 8 (2013) 56, DOI 10.1186/1556-276X-8-56, Table 3 (open access).
• Kaminski, D. A., & Prakash, C. (1986). International Journal of Heat and Mass Transfer, 29(12), 1979–1988. DOI 10.1016/0017-9310(86)90017-7 — §47.12's Gate 5, the configuration §60.5 runs. Paywalled; no open-access copy was found and the primary table was never read, so no title is asserted for it here either. MISSES — Gate 5.
• Belazizia, A., Benissaad, S., & Abboudi, S. (2012). Effect of wall conductivity on conjugate natural convection in a square enclosure with finite vertical wall thickness. Advanced Theoretical and Applied Mechanics, 5, no. 4, 179–190. Open access at m-hikari.com/atam/atam2012/atam1-4-2012/ — an independent published solution of the Kaminski–Prakash configuration, itself validated against it. The SECONDARY source Gate 5 actually compares against. — §60.5
• Qu, W., & Mudawar, I. (2002). Analysis of three-dimensional heat transfer in micro-channel heat sinks. International Journal of Heat and Mass Transfer, 45, 3973–3985. DOI 10.1016/S0017-9310(02)00101-1 — §47.12's Gate 6, the semiconductor gate. Read in full from the authors' own public copy. §79.12 runs it and it PASSES: both substrate temperatures fall inside the experimental uncertainty the paper draws. — §79
• Kawano, K., Minakami, K., Iwasaki, H., & Ishizuka, M. Micro channel heat exchanger for cooling electrical equipment. ASME HTD-361-3/PID-3 (1998) 173–180. — the inlet and outlet thermal-resistance measurements Gate 6 is held against. NOT OBTAINED (an ASME conference volume; no copy found), so the comparison is a digitisation of Qu & Mudawar's Fig. 4, which §79.12's Disclosure 1 states. — §79

• McGrattan, K., Hostikka, S., McDermott, R., Floyd, J., Weinschenk, C., Overholt, K., Vanella, M., et al. Fire Dynamics Simulator Technical Reference Guide, NIST SP 1018-1. NIST, US public domain; read locally from reference/fds/Manuals/ with reference/fds/LICENSE.md read verbatim. No FDS source code was read for these sections. — §25, §66, §68, §76
• Modest, M. F. Radiative Heat Transfer, 3rd ed., Academic Press (2013), ch. 5. — §50
• Hottel, H. C., & Sarofim, A. F. Radiative Transfer, McGraw-Hill (1967), ch. 3, 5. — §50 (the net-radiation exchange method; the method of images, named in §50.9's refusal)
• Walton, G. N. Calculation of Obstructed View Factors by Adaptive Integration. NISTIR 6925, National Institute of Standards and Technology, November 2002. https://nvlpubs.nist.gov/nistpubs/Legacy/IR/nistir6925.pdf — US Government, public domain. §49 (the double area integral and its dot-product form, the obstruction-elimination tests, the row-sum figure of merit, and the BB104 benchmark)
• Shapiro, A. B. FACET — A Radiation View Factor Computer Code for Axisymmetric, 2D Planar and 3D Geometries with Shadowing. UCID-19887, Lawrence Livermore National Laboratory, 1983. DOI 10.2172/5607653 — US DOE, public domain. §49.8 (the shadowed configuration F_12 = 0.115621)
• Howell, J. R. A Catalog of Radiation Heat Transfer Configuration Factors, 3rd ed. https://www.thermalradiation.net/ — entries C-11 (identical parallel directly-opposed rectangles) and C-14 (two rectangles of equal length sharing an edge at 90°), both tracing to Hottel (1931) and Hamilton & Morgan (1952). §49.8 (the two analytic view-factor gates)
• Gebhart, B. (1961). International Journal of Heat and Mass Transfer, 3(4), 341–346. DOI 10.1016/0017-9310(61)90048-5 — §50.2, the absorption-factor alternative: named and not used
• Balaji, C., & Venkateshan, S. P. International Journal of Heat and Fluid Flow, 14(3) (1993) 260–267, DOI 10.1016/0142-727X(93)90057-T, and 15(3) (1994) 249–251, DOI 10.1016/0142-727X(94)90046-9; Akiyama, M., & Chong, Q. P. Numerical Heat Transfer A, 32(4) (1997) 419–433, DOI 10.1080/10407789708913899 — the coupled convection-plus-surface-radiation cavity gate. — §50.12, NOT run: the tabulated Nu_conv/Nu_rad are paywalled and the fluid side has no case format for a radiating enclosure.

• Ostwald, W. (1925). Kolloid-Zeitschrift, 36, 99–117; de Waele, A. (1923). — §38 (the power law)
• Cross, M. M. (1965). Journal of Colloid Science, 20, 417–437. — §38
• Carreau, P. J. (1972). Transactions of the Society of Rheology, 16, 99–127; Yasuda, K., Armstrong, R. C., & Cohen, R. E. (1981). Rheologica Acta, 20, 163–178. — §38 (one formula serves both)
• Herschel, W. H., & Bulkley, R. (1926). Kolloid-Zeitschrift, 39, 291–300. — §38
• Casson, N. In Mill (ed.), Rheology of Disperse Systems, Pergamon (1959), 84–104. — §38
• Papanastasiou, T. C. (1987). Journal of Rheology, 31, 385–404. — §38.3 (the regularisation, in the product form)
• Bercovier, M., & Engelman, M. (1980). Journal of Computational Physics, 36, 313–326. — §38.3 (the alternative regularisation)
• Frigaard, I. A., & Nouar, C. (2005). Journal of Non-Newtonian Fluid Mechanics, 127, 1–26. — §38.3 (what regularisation costs)
• Bird, R. B., Armstrong, R. C., & Hassager, O. Dynamics of Polymeric Liquids, vol. 1, 2nd ed., Wiley (1987). — §38 (the family)
• Chhabra, R. P., & Richardson, J. F. Non-Newtonian Flow and Applied Rheology, 2nd ed. (2008). — §38.9 (Buckingham–Reiner)
• Young, T. (1805). Philosophical Transactions of the Royal Society, 95, 65–87. — §39 (the equilibrium angle)
• Huh, C., & Scriven, L. E. (1971). Journal of Colloid and Interface Science, 35, 85–101. — §39 (the moving contact-line singularity)
• Voinov, O. V. (1976). Fluid Dynamics, 11, 714–721; Cox, R. G. (1986). Journal of Fluid Mechanics, 168, 169–194. — §39.4 (the asymptotic matching)
• Hoffman, R. L. (1975). Journal of Colloid and Interface Science, 50, 228–241. — §39.4 (the master curve)
• Jiang, T.-S., Oh, S.-G., & Slattery, J. C. (1979). Journal of Colloid and Interface Science, 69, 74–77. — §39.4 (the explicit correlation used here; Kistler's fit is deliberately absent, its four constants coming from a book chapter this project has not read)
• Afkhami, S., Zaleski, S., & Bussmann, M. (2009). Journal of Computational Physics, 228, 5370–5389 — the mesh-dependent (numerical-slip) angle. §39.8, named and deliberately not implemented until the gate that would show it works exists
• Sui, Y., Ding, H., & Spelt, P. D. M. (2014). Annual Review of Fluid Mechanics, 46, 97–119. — §39 (the review)
• Washburn, E. W. (1921). Physical Review, 17, 273–283. — §39.7 (capillary rise)

• Dukowicz, J. K. A particle-fluid numerical model for liquid sprays. Journal of Computational Physics, 35 (1980) 229–253. DOI 10.1016/0021-9991(80)90087-X — §66 (the discrete droplet model: the parcel, and the real-valued weight n_p)
• Crowe, C. T., Sharma, M. P., & Stock, D. E. The particle-source-in-cell (PSI-CELL) model for gas-droplet flows. Journal of Fluids Engineering, 99 (1977) 325. DOI 10.1115/1.3448756 — §67, §68 (the per-cell sum every coupled source is)
• Crowe, C., Sommerfeld, M., & Tsuji, Y. Multiphase Flows with Droplets and Particles, CRC Press (1998). ISBN 0-8493-9469-4 — §66.2 (the equation of motion, and the regime argument for which of its terms survive)
• Maxey, M. R., & Riley, J. J. Physics of Fluids, 26 (1983) 883. DOI 10.1063/1.864230 — §66, §68 (the equation of motion: the added-mass coefficient, and the drag term §68 returns to the gas)
• Schiller, L., & Naumann, A. Zeitschrift des Vereines Deutscher Ingenieure, 77 (1933) 318, in the form compiled by Clift, R., Grace, J. R., & Weber, M. E. Bubbles, Drops, and Particles, Academic Press (1978). ISBN 0-12-176950-X — §66.3 (the drag correlation)
• Macpherson, G. B., Nordin, N., & Weller, H. G. Communications in Numerical Methods in Engineering, 25 (2009) 263. DOI 10.1002/cnm.1128 — §66.6, barycentric tracking: the paper was read and it is not implemented — the one case the face-crossing walk cannot do. Its OpenFOAM implementation is GPL-3.0 and was not opened.
• Elghobashi, S. On predicting particle-laden turbulent flows. Applied Scientific Research, 52 (1994) 309. DOI 10.1007/BF00936835 — §67, §68 (the coupling map: below alpha_p ~ 1e-6 one-way coupling suffices, 1e-6–1e-3 needs §68, above 1e-3 collisions matter and are not here)
• Satish, N., Harris, M., & Garland, M. Designing efficient sorting algorithms for manycore GPUs. IEEE IPDPS 2009. DOI 10.1109/IPDPS.2009.5161005 — the paper was read; no implementation of it was opened. §67.4 (the three-phase radix pass)
• Merrill, D., & Grimshaw, A. Parallel scan for stream architectures. University of Virginia Technical Report CS2009-14. — §67.2 (the reduce-then-scan decomposition)
• Blelloch, G. E. Prefix sums and their applications. CMU-CS-90-190 (1990). — §67.2 (the exclusive scan and its work-efficiency argument)
• Hillis, W. D., & Steele, G. L., Jr. Communications of the ACM, 29(12) (1986) 1170. DOI 10.1145/7902.7903 — §67.2 (the in-block scan network, chosen for a property that is not speed)
• Steele, G. L., Jr., Lea, D., & Flood, C. H. Fast splittable pseudorandom number generators. OOPSLA 2014, ACM SIGPLAN Notices, 49(10) 453. DOI 10.1145/2660193.2660195 — §66.9 (the SplitMix64 finalising mix, used as a bijection and not as a generator)
• Ranz, W. E., & Marshall, W. R. Evaporation from drops. Chemical Engineering Progress, 48 (1952) 141–146 (Part I) and 173–180 (Part II). — §68.5 (the sensible-heat half of Nu_0 = 2 + 0.6 Re^(1/2) Pr^(1/3)), §76 (the mass-transfer half, and the 56 suspended-droplet experiments §76.12's first gate measures)
• Spalding, D. B. The combustion of liquid fuels. 4th Symposium (International) on Combustion (1953) 847–864; and Convective Mass Transfer: An Introduction, Edward Arnold (1963). — §76.6 (B_M, and the Stefan-flow rate)
• Godsave, G. A. E. Studies of the combustion of drops in a fuel spray. 4th Symposium (International) on Combustion (1953) 818–830. — §76.9 (the heat-limited rate at the boiling point)
• Abramzon, B., & Sirignano, W. A. Droplet vaporization model for spray combustion calculations. International Journal of Heat and Mass Transfer, 32 (1989) 1605–1618. DOI 10.1016/0017-9310(89)90043-4 — §76.6 (B_T = (1 + B_M)^φ − 1, the default)
• Sazhin, S. S. Advanced models of fuel droplet heating and evaporation. Progress in Energy and Combustion Science, 32 (2006) 162–214. DOI 10.1016/j.pecs.2005.11.001 — §76
• Watson, K. M. Thermodynamics of the liquid state. Industrial & Engineering Chemistry, 35 (1943) 398–406. — §76.4 (h_v(T))
• Marrero, T. R., & Mason, E. A. Gaseous diffusion coefficients. Journal of Physical and Chemical Reference Data, 1 (1972) 3–118. DOI 10.1063/1.3253094 — §76.4
• Lewis, W. K. The evaporation of a liquid into a gas. Transactions of the ASME, 44 (1922) 325–340. — §76.13
• NIST Chemistry WebBook, SRD 69. US government, public domain. — §76.4 (the water-vapour specific heat and the critical constants)
• Theobald, R. C. The effect of nozzle design on the stability and performance of turbulent water jets. Fire Safety Journal, 4 (1981) 1–13. — §68.12, about 90 hose-stream experiments. MISSES — Gate 68-C, with the gas held at rest.
• Bai, C., & Gosman, A. D. SAE 950283 (1995). — §78.1, §78.4 (the impact regime map, and the alternative splash threshold)
• Mundo, C., Sommerfeld, M., & Tropea, C. International Journal of Multiphase Flow, 21 (1995) 151. — §78.4 (K = Oh Re^1.25, K_crit = 57.7, the default). The experimental data itself was not transcribed, which is why Gate 78-D is open. OPEN — Gate 78-D: the two published splash criteria disagree by a measured factor in Weber number for the same droplet.
• Yarin, A. L. Annual Review of Fluid Mechanics, 38 (2006) 159. — §78.1 (the review that explains why neither threshold is better than approximately right)
• IAPWS R1-76 (2014). — §78.2 (sigma = B tau^mu (1 + b tau), implemented verbatim and gated against the release's own table)

• Hirt, C. W., & Nichols, B. D. Journal of Computational Physics, 39 (1981) 201–225. — §20.1
• Zalesak, S. T. Journal of Computational Physics, 31 (1979) 335–362. — §20.2 (and §22's rotating-slotted-disc boundedness check)
• Brackbill, J. U., Kothe, D. B., & Zemach, C. Journal of Computational Physics, 100 (1992) 335–354. — §20.4, §87 (the continuum-surface-force regularisation)
• Ubbink, O. PhD thesis, Imperial College London (1997). — §20.1 (the interface-compressed finite-volume form on unstructured meshes)
• Rusche, H. PhD thesis, Imperial College London (2002). — §20.1 (the same)

• Saad, Y. Iterative Methods for Sparse Linear Systems, 2nd ed. SIAM (2003). DOI 10.1137/1.9780898718003 — §8, §21 (§6.7 PCG, §7.4.2 BiCGStab, §12.4 multicolour ILU, ch. 14 block Jacobi and additive Schwarz)
• van der Vorst, H. A. Bi-CGSTAB: A Fast and Smoothly Converging Variant of Bi-CG for the Solution of Nonsymmetric Linear Systems. SIAM Journal on Scientific and Statistical Computing, 13(2), 631–644 (1992). DOI 10.1137/0913035 — §8.1
• Hestenes, M. R., & Stiefel, E. Methods of conjugate gradients for solving linear systems. Journal of Research of the National Bureau of Standards, 49(6), 409 (1952). DOI 10.6028/jres.049.044 — §8.2. A US Government work, public domain.
• Swarztrauber, P. N. SIAM Review, 19 (1977) 490–501. — §8.5
• Stüben, K. Journal of Computational and Applied Mathematics, 128 (2001) 281–309; Ruge & Stüben (1987). — §8.3. Provided here by AMGX (BSD-3-Clause), not reimplemented.

• Ward, J. C. Turbulent flow in porous media. Journal of the Hydraulics Division, ASCE, 90(5) (1964) 1–12. DOI 10.1061/JYCEAJ.0001096 — §18, §53 (the same Darcy–Forchheimer law integrated through a slab instead of over a cell)
• Idelchik, I. E. Handbook of Hydraulic Resistance, 4th ed., Begell House (2007). ISBN 978-1-56700-251-5, Diagrams 8-1 to 8-6 — perforated plates and screens, the source of K(sigma). Not opened for §53; the thin-plate form used is the one published in the open literature, and §53.7 gates it against its own limits. — §53
• Karki, K. C., Radmehr, A., & Patankar, S. V. Use of computational fluid dynamics for calculating flow rates through perforated tiles in raised-floor data centers. HVAC&R Research, 9(2) (2003) 153–166. DOI 10.1080/10789669.2003.10391062 — §53.8, the per-tile flow-rate gate: NOT run — the paper was not reachable from this environment.
• Karki, K. C., & Patankar, S. V. Airflow distribution through perforated tiles in raised-floor data centers. Building and Environment, 41(6) (2006) 734–744. DOI 10.1016/j.buildenv.2005.03.005 — §53

• AMCA 210 / ASHRAE 51, Laboratory Methods of Testing Fans for Certified Aerodynamic Performance Rating. — §52 (what a manufacturer's curve is — a static-pressure rise against volumetric flow at a stated density and shaft speed — and therefore why §52.5 carries a density and a speed correction rather than treating the table as absolute)
• NIST, Fire Dynamics Simulator 6 verification suite, Verification/HVAC/fan_test.fds and qfan_test.fds with their published .csv reference values — US Government public domain. The case files and their reference numbers are the external cross-check of §52.12 Gate 52-B. Source/hvac.f90 was read for the DISCIPLINE only: that a fan curve is scaled by rho/rho_curve at every evaluation, and that its tabulated branch resolves the operating point by a bisection with a data-dependent trip count, which is correct for a CPU code and uncapturable here (§52.7).
• Buzbee, B. L., Dorr, F. W., George, J. A., & Golub, G. H. The direct solution of the discrete Poisson equation on irregular regions. SIAM Journal on Numerical Analysis, 8(4) (1971) 722–736. DOI 10.1137/0708066 — §52.9, the capacitance-matrix path: named and refused
• ASHRAE. ASHRAE Handbook—Fundamentals, Chapter 1, "Psychrometrics", ASHRAE (2021). — §54.2 (whose equation numbering is used), §54.8 (Table 2, the external comparison, at 101.325 kPa)
• Hyland, R. W., & Wexler, A. Formulations for the thermodynamic properties of the saturated phases of H2O from 173.15 K to 473.15 K. ASHRAE Transactions, 89(2A) (1983) 500–519. — §54.2 (the C1–C13 coefficients), §76.4 (reused rather than re-fitted)
• Herrmann, S., Kretzschmar, H.-J., & Gatley, D. P. Thermodynamic properties of real moist air, dry air, steam, water, and ice (RP-1485). HVAC&R Research, 15(5) (2009) 961–986. DOI 10.1080/10789669.2009.10390874 — §54.3, named and not implemented; the enhancement factor it carries is what makes the ideal relations 0.44 % low in W_s at 25 °C, which §54.8 prints rather than tolerates.
• Gatley, D. P., Herrmann, S., & Kretzschmar, H.-J. A twenty-first century molar mass for dry air. HVAC&R Research, 14(5) (2008) 655–662. DOI 10.1080/10789669.2008.10391032 — §54 (where M_a = 28.966 g/mol, and hence eps = 0.621945, come from)
• Herrlin, M. K. Rack cooling effectiveness in data centers and telecom central offices: the Rack Cooling Index (RCI). ASHRAE Transactions, 111(2) (2005) 725–731. (ASHRAE Transactions of this vintage carries no DOI; stable record https://www.semanticscholar.org/paper/99b942df4aa448a1e06f77d36b48d5d52a40c6e0.) — §55.1
• Herrlin, M. K. Airflow and cooling performance of data centers: two performance metrics. ASHRAE Transactions, 114(2) (2008) 182–187. (No DOI; same caveat.) — §55.2 (RTI)
• Sharma, R. K., Bash, C. E., & Patel, C. D. Dimensionless parameters for evaluation of thermal design and performance of large-scale data centers. AIAA 2002-3091 (2002). DOI 10.2514/6.2002-3091 — §55.3 (SHI and RHI)
• ASHRAE Technical Committee 9.9. Thermal Guidelines for Data Processing Environments, 5th ed., ASHRAE (2021). ISBN 978-1-947192-90-4 — §55.1 (the Class A1–A4 recommended, 18–27 °C, and allowable envelopes RCI is measured against)
• Wibron, E., Ljung, A.-L., & Lundström, T. S. Energies, 12(8) (2019) 1473. DOI 10.3390/en12081473 — CC-BY-4.0, licence verified live through the Crossref REST API, but the full text was not reachable from this environment, so §55.8's six-configuration ranking gate is NOT run and only the one relation the abstract states is gated. — §55.8

• Ghia, U., Ghia, K. N., & Shin, C. T. Journal of Computational Physics, 48 (1982) 387–411. — the lid-driven cavity's tabulated centreline profiles
• Moser, R. D., Kim, J., & Mansour, N. N. Physics of Fluids, 11 (1999) 943. — DNS channel profiles at Re_tau 180 / 395 / 590, and the sublayer k+ ≈ C_v (y+)^2 with C_v ≈ 0.07 of §15.2. The key files' host states no redistribution terms, so they are not distributed; gated by SPEC-LIT §110.2 Gate 110-A
• Driver, D. M., & Seegmiller, H. L. AIAA Journal, 23 (1985) 163–171. — the backward-facing step's reattachment length, x_r/h = 6.26 ± 0.10. Gated by SPEC-LIT §110.3 Gate 110-B — the "NOT run" of §40.7 is retired: the step is now carved out of one block
• McCaffrey, B. J. NBSIR 79-1910 (1979). — the buoyant plume's centreline temperature and velocity decay correlations, ΔT ~ z^{−5/3} in the plume region. US Government work, public domain. Gated by SPEC-LIT §110.4 Gate 110-C
• Martin, J. C., & Moyce, W. J. Philosophical Transactions of the Royal Society A, 244 (1952) 312. — the dam break's surge-front position against time

simul@msimul.com · 주식회사 이터레이션즈 / Iterations Co., Ltd. 협력 및 기여: 주식회사 메테오시뮬레이션 / Meteo Simulation Co., Ltd.

교육과 학술 연구는 무료입니다. 기업 연구개발, 학교에 속하지 않은 연구기관, 수탁 및 상업적 이용은 별도 라이선스가 필요합니다 — LICENSE 제2·3절. 적용 범위가 불분명한 경우 문의해 주시기 바랍니다.