forest-sphere¶
英語 | 日本語
Flow around a sphere with AMR -- the AMR-coupling force-QoI validation on the FOREST (block-AMR) solver, the OpenRheo analogue of waLBerla's FlowAroundSphere. Uniform free-stream flow past a solid sphere of diameter D at Re = U D / nu; refine a band/box hugging the sphere to max_level and show the time-mean drag coefficient Cd from AMR matches BOTH the uniform-fine result AND the Schiller-Naumann correlation Cd = (24/Re)(1+0.15 Re^0.687), proving the coarse/fine coupling does not corrupt a precise force QoI.
概要: 同じ球を近傍 AMR 帯付きで解く。一様細格子と AMR の Cd 一致度を見ることで、粗細界面が力積分を汚していないかを確認する。
Geometry¶

Fields¶
この計算が最後の出力ステップで書き出した場をそのまま描いたもの。3D ケースは中央断面。


Plots¶
この計算が書き出した QoI の時系列・分布。参照値がある場合は水平線で重ねている。

Build & run¶
このページの図を作った計算のコスト(Miyabi-G Grace CPU (72 cores, -stdpar=multicore)): 48.4 s wall。
Validation¶
| Tier | T2 — 単点比較 |
| QoI | Cd |
| Reference | Schiller-Naumann / Clift-Gauvin |
| Gate | — ゲートなし |
Source & run details¶
| Item | Value |
|---|---|
| Target | orheo-forest-sphere |
| Source | native/benchmarks/forest_sphere.cpp |
| Lattice | — |
| GPU twin | yes (orheo-forest-sphere-gpu) |
| Modules | ahmed_solver.hpp, forest_regrid.hpp, forest_step.hpp, forest_surface_force.hpp, sphere_geom.hpp |
| Gallery run | 2026-07-27 @ fc5fbb30 |
| Args | 8 8000 0.05 100 0.0 1 2 0 2 5 2 . |
| Wall time | 48.42 s |
| CPU time | 1845.38 s |
| Device | Miyabi-G Grace CPU (72 cores, -stdpar=multicore) |
Computation log¶
この計算の実際の出力(先頭と末尾)。
=== forest sphere (flow around a sphere, AMR force-QoI) ===
nd=8 steps=8000 ulat=0.0500 (Mach=0.087) re=100.0 cs=0.000 max_level=1 band=2 box=0.00
D=1.000m dx0=0.12500m domain blocks=14x8x8 (kNb=4) nu_lat=4.0000e-03 tau=0.5120
Schiller-Naumann reference Cd(Re=100) = 1.0917
=== sphere surface-following forest ===
forest: 2408 leaves, max_level reached=1 (requested 1)
=== sphere wall data (solid mask + Bouzidi qlink) ===
solid cells=7192 (of 520128 pool cells)
BC: uniform free-stream inlet u=0.0500; anti-bounce-back pressure (anchor) outlet; ALL lateral/top/bottom faces = specular free-slip far-field
=== time-stepping ===
step 0/8000 t_conv=0.000 Cd=7.8988 (SN 1.092) Cl=-0.0000 umax=0.093 | 0.0s 0 step/s 0 MLUPS | Mdrift=+0.0 ppm
step 1000/8000 t_conv=0.893 Cd=1.2340 (SN 1.092) Cl=0.0000 umax=0.060 | 6.0s 166 step/s 26 MLUPS | Mdrift=+2197.7 ppm
step 2000/8000 t_conv=1.786 Cd=1.2706 (SN 1.092) Cl=-0.0000 umax=0.060 | 12.6s 158 step/s 24 MLUPS | Mdrift=+2212.0 ppm
step 3000/8000 t_conv=2.679 Cd=1.2692 (SN 1.092) Cl=-0.0000 umax=0.060 | 19.1s 157 step/s 24 MLUPS | Mdrift=+2212.6 ppm