カテゴリー Ⅱ 日本建築学会環境系論文集第 80 巻第 718 号, ,2015 年 12 月 J. Environ. Eng., AIJ, Vol. 80 No. 718, , Dec., 2015 DOI

Size: px

Start display at page:

Download "カテゴリー Ⅱ 日本建築学会環境系論文集第 80 巻第 718 号, ,2015 年 12 月 J. Environ. Eng., AIJ, Vol. 80 No. 718, , Dec., 2015 DOI"

阜齐如何
5 years ago
Views:

1 カテゴリー Ⅱ 日本建築学会環境系論文集第巻第 7 号,-, 年月 J. Environ. Eng., AIJ, Vol. No. 7, -, Dec., DOI 有限体積法に基づく Large Eddy Simulation のための対流項離散化スキームの検討建物屋上面からの排ガス拡散への大気安定度の影響その CONVECTION TERM DISCRETIZATION FOR LARGE EDDY SIMLATIONS BASED ON THE FINITE VOLME METHOD Effect of thermal stratification on the dispersion characteristics of rooftop ehaust, Part 小野浩己 *, 瀧本浩史 **, 道岡武信 *** ****, 佐藤歩 Hiroki ONO, Hiroshi TAKIMOTO, Takenobu MICHIOKA and Ayumu SATO Validation of Large Eddy Simulation based on a Finite Volume Method was undertaken using data from our previous eperiment. Four schemes for convection-term discretization were tested on a momentum transport equation, a passive scalar transport equation and an active scalar transport equation. The results support the use of a low-dissipation scheme for the momentum transport equation to avoid overestimating circulation on a rooftop. Such overestimation of circulation can lead to an underestimation of concentrations behind the building. Conversely, a bounded scheme, even in cases that resulted in ecess numerical dissipation, is preferred for each scalar transport equation to maintain a profile and peak concentration value. Keywords : Large Eddy Simulation, Discretization scheme, Numerical diffusion, Finite Volume Method. Large Eddy Simulation (LES) ) ) ) CFD(Computational Fluid Dynamics) ) ) CFD ) LES(Large Eddy Simulation) LES LES 7) LES LES LES * ** *** **** 電力中央研究所特別契約研究員博士 ( 工学 ) 電力中央研究所主任研究員博士 ( 工学 ) 電力中央研究所主任研究員博士 ( 工学 ) ( 現在は近畿大学 ) 電力中央研究所上席研究員博士 ( 工学 ) Visiting Researcher, Central Research Institute of Electric Power Industry, Dr.Eng. Research Scientist, Central Research Institute of Electric Power Industry, Dr.Eng. Research Scientist, Central Research Institute of Electric Power Industry, Dr.Eng. Senior Research Scientist, Central Research Institute of Electric Power Industry, Dr.Eng.

2 . ). () ( ) (Conv.) dv Conv. divu u nds () S S u n (f) f. ().7 (Filtered-) f _ linear f _ upwind f _ filtered linear (7) θ. ) Flow direction D f D DD DD, f f _ linear D f f f ) () ( ) ( ) D f f Target Face f f. (pwind) Fig. Control volume and grid nodes notation if( SlopeSlope > and SlopeSlope < ): then blending pwind f _ upwind () Target Face f. (-pwind) f _ linearupwind f () ). (LST) Fig. Concept of Filtered- scheme Inlet Outlet Thermally stratified chamber Spires Secondary flow suppression plate (mm-high) Γ-type ehaust (Inner Diameter=mm) z Building (mm cube) Angles (mm-high) [m]. [m]. [m]. [m] Fig. Overview of the wind tunnel [m] f _ LST.7 f _ linear. f _ linearupwind () LES = ). f _ f _ upwind r f _ upwind f _ linear r D ) () (=u) Fig. Overview of the mesh Fig. Mesh around the Γ-type ehaust Parabolic Distribution

Spire Secondary flow suppression plate Angle =- PISO ) 9 Fig.

() (9) () () () Fig.7 Mesh around the building. LES.

SGS SGS (νsgs)sgs (α SGS)SGS (γ SGS) Dynamic Smagorinsky ( ) ) SGS

. ( ) Filtered- LST 7) ). (/H) Filtered- LST Filtered- /H=.

3 Spire Secondary flow suppression plate Angle =- PISO ) 9 Fig. Mesh around the roughness elements Table Dynamic Smagorinsky model () (9) () () () Fig.7 Mesh around the building. LES. ( m m m) Γ ( D = mm) (,,) y z ( H = mm ) H.D ( 7). (. - m /s) ( ) Spalding Jayatilleke 9). SGS SGS (νsgs)sgs (α SGS)SGS (γ SGS) Dynamic Smagorinsky ( ) ) SGS (PrSGS)SGS (ScSGS)(=) dynamic. K CFD OpenFOAM(ver.) () () ().. ( ) Filtered- LST 7) ). (/H) Filtered- LST Filtered- /H=. /H ( )LST Filtered- Filtered- LST LST ) /H ( ) Filtered- LST Filtered-

/㻴 Filtered-r LST /H Fig. Profiles of stream-wise mean velocity (altering scheme of the convection term of the momentum transport equation) u'㼞㼙㼟/㻴... Filtered- LST t Fig.

/H Profiles of vertical velocity fluctuation (altering scheme of the convection term of the momentum transport equation) Fig.

4 /㻴 Filtered-r LST /H Fig. Profiles of stream-wise mean velocity (altering scheme of the convection term of the momentum transport equation) u'㼞㼙㼟/㻴... Filtered- LST t Fig.9 /H Profiles of stream-wise velocity fluctuation (altering scheme of the convection term of the momentum transport equation) w'㼞㼙㼟/㻴... Filtered- LST t Fig. /H Profiles of vertical velocity fluctuation (altering scheme of the convection term of the momentum transport equation) Fig. Distributions of stream-wise mean velocity near the roof-top (altering scheme of the convection term of the momentum transport equation) Fig. Distributions of stream-wise mean velocity behind the buildings (altering scheme of the convection term of the momentum transport equation)

5 C/C Filtered- LST Fig. /H 7 Profiles of mean concentration (altering scheme of the convection term of the momentum transport equation) Fig. Distributions of mean concentration near the roof-top Filtered- Oscillation Fig. Instantaneous velocity distributions at = ᅗ 9㸪 ᘓ ᶍᆺ ᚋ ࡅ ὶ᪉ ᗘኚ 㖄 ᪉ ࢡ ᗘ 㐣ᑠホ౯ࡋ ࠋ LST ᵝ /H = ᗘኚ ᖹᆒᖹ᪉ (u rms/h w rms /H ) ศᕸ ࢡ 㧗 ࡉ 㧗 ࡃ ホ ౯ ࡍ ࠋ /H = ࡍࠋ Filtered- 㛫 ᕪ ࡃ㸪ᐇ Filtered- ᯝ ࡋ 㸪ᐇ ᯝ ᑐᛂ ᯝ ᴫ ࡋ ࠋ/H = 㹼 ᘓ 㧗ࡉ ప ࡀ㸪/H = ௨㝆 ࢡ ᗘ 㐣ᑠホ౯ࡋ ࠋLST 㡿ᇦ 㸪LST Filtered- ẚ ᘓ ᒇ 㠃 ᚠ ὶࡀ Filtered- ẚ ࡁࡃ 㸪ᐇ ࡁࡃ㸪 ࢫᨺฟᆅⅬ ᒇ 㠃 Wake ᚰ ࡍ ࠋ ᯝ ẚ 㐣ホ౯ࡍ ഴ ࠋ ᯝ ᴫ Filtered- ࡋ 㸪 ᐇ ᯝ ᑐᛂࡀⰋ ࠋ๓ 㸪LST ᘓ ᒇ 㠃 ᚠ ὶ ࡁࡃ ホ ౯ࡋ ࠋ ࡑ ࡗ ࡗ 㸪ᅗ ࡍ ᨺฟ ᚋ 㧗 ᗘ ᒇ 㠃㢼 ከࡃ ᗘࡀ ࡉ ࡓ 㸪ᘓ ᶍᆺᚋ᪉ ᗘ ᑠ ࡉࡃホ౯ࡋ ࠋ Filtered- ẚ ᒇ 㠃 Wake እ ࡅ ௨ 㸪ᮏࢣ ࢫ ࡅ ᘓ ᶍᆺ ᚋ Ẽὶᛶ ᗘᣑᩓ ࡀ ࡁࡃ 㸪 ࡉ ࡓ ࡀ Wake ࡗ ࡉ 㸪ᘓ ἣ ࡍ 㸪㐠 ᪉ ᘧ ᑐὶ㡯ࢫ ᚋ 㐣ホ౯ࡋ ࠋ ࡓ Filtered- ࡀ㐺ࡋ ࠋࡋ ࡋ㸪 ᛶࡀࡇ ᢚ ࡀ㸪LST ᯝࡀ ᛶ ᅗ ࡍ 㸪 ᨭᣢ㒊 ᘓ ᶍᆺ 㢼 ࢥ ᅇ 㸪㐣ホ౯ ࡘ ࡀࡗ ࠋᅗ ᘓ Ⅼ ࡋࡓ ࡁ ࡀ ࡍ ࠋࡇ ᱁Ꮚ 㸪 ᶍᆺᚋ᪉ ࡅ 㛫ᖹᆒ ᗘ(C/C㸪C ฟᆅⅬ ᗘ) ศ 㛫 ࡇ ᢚไ ࡁ ᛶࡀ ࡀ㸪ィ ࢯ ࢫ ᕸ ࡍࠋ Filtered- ᯝ ᴫ ࡋ 㸪せồ ቑ ᣍࡁ㸪ᐇ ゎỴἲ ゝ ࠋᑐࡋ ᯝ ᑐ ᛂ Ⰻ ࠋ LST /H = Filtered- 㸪㢼 ࢥ ࡎ ศᕸࡀ Filtered- ẚ ࢡ㧗ࡉࡀ 㧗 ࠋࡇ ᒇ 㠃 ᚠ ṧ 㸪 ẚ ᖜ 㝖ཤࡍ ࡇ ᡂ ὶ ࡁࡃண ࡋ ࠋ ࡗ 㸪Filtered- 㐠 ᪉ ᘧ ᑐὶ㡯 ᐇ ࡋ ࡇ 㛵㐃ࡋ ࠋ ࡓ㸪 /H = ௨㝆㸪Filtered- ᐇ ᯝ ẚ 㞳ᩓ ࢫ ࡋ ゝ ࠋ 7

6 C/C Filtered- LST Fig. /㻴 /H Profiles of stream-wise mean velocity (altering scheme of the Fig. /H Profiles of stream-wise velocity fluctuation (altering scheme T [deg.c.] LST n LST n of the convection term of the temperature transport equation) w'㼞㼙㼟/㻴... LST n convection term of the temperature transport equation) /H Profiles of vertical velocity fluctuation (altering scheme of the Fig. convection term of the temperature transport equation) /H Profiles of mean Temperature (altering scheme of the convection term of the temperature transport equation) C/C LST 7 Fig.9 u'㼞㼙㼟/㻴... Fig.7 LST n Profiles of mean concentration (altering scheme of the convection term of the concentration transport equation) /H /H 7 Fig. Profiles of mean concentration (altering scheme of the convection term of the temperature transport equation)

7 . ࢩ ࢫ ( ᗘ) ᑐὶ㡯 ᑐࡍ 㞳ᩓ ࢫ ᯝ ௨ ࠋ ) 㐠 ᪉ ᘧ ᑐὶ㡯㞳ᩓ 㸪ࡑ ࢫ 㔞 ẚ. ウࢣ ࢫ ẼᏳᐃᗘࡀ ሙ ᐇ㦂 ᯝ ẚ ᑐ ࡋ 㸪 ᗘ ᯝ ࡁࡃᙳ㡪ࡋ㸪 ᣑᩓࡀᑡ ᪉ࡀᐇ ᑐᛂ Ⰻࡃ ᪉ ᘧ ᑐὶ㡯㞳ᩓ ࢫ ኚ ࡉࡏ㸪ィ ᯝ ࠋࡋ ࡋ㸪 ᣑᩓ ᑡ ᙧ 㛫() ᙳ㡪ウࡋࡓࠋ ࡓ㞳ᩓ ࢫ Filtered-㸪LST㸪 ࡀ ࡋ 㠀 ᐇ ゎ ᛶࡀ ࠋࡑࡇ 㢮㸪 ࢣ ࢫ 㐠 ᪉ ᘧ ᑐὶ㡯㑅ᢥ ᢚไࡍ 㢼 (Filtered-) ࢫ Filtered- ࡓࠋ ࡇ ᗘ ᣢࡋࡘࡘ ᢚไࡀ ࠋ ) 㐠 ᪉ ᘧ ᑐὶ㡯ࢫ Filtered- ࡓሙ 㸪. ウ ᯝ ᅗ ᘓ ᶍᆺ ᚋ ࡅ 㛫ᖹᆒ ᗘ(C/C ) ศᕸ ࡍࠋ ᗘ ᪉ ᘧ ᑐὶ㡯ࢫ ኚ ࡉࡏࡓࡇ ᕪ ᑠ ᘓ ᶍᆺ ᚋ /H = 㞳ᩓ ࢫ ᕪ ࡉࡃ㸪 ࢫ ⴭࡋࡃண ࡀ 㸪 ᘓ ᶍ ᆺ 㞳 ࡘ ᕪ ࡀ ࡌ 㸪 LST ࡗࡓࡀ㸪 ᛶ ࡓࡍ ࢫ ࡀẚ Ⰻ ᗘࡀప ࡍ ࡇ Filtered- ẚ ࡎ ࡀᑠࡉࡃホ౯ࡍ ዲ ᯝ ࡗࡓࠋ ἣ ࡗ ᛶ ࡓࡉ ࢫ ഴ ࠋLeonard ) ࢫᆺ ศᕸ ࡘࢫ (㸪Filtered-) ศᕸᙧ ṇࡋࡃホ౯ ࡁ ὶ 㛵ࡋ 㸪㞳ᩓ ㄗᕪ せ㡯ࡀᣑᩓ ᚰᕪศ ࢫ ᛶ 㸪ィ ࡀ Ᏻᐃ ᛶࡀ ࡓ 㸪ὀពࡀᚲせ ࠋ ) ศᕸᙧ ᣢ ࡁࡎ ࢡ 㧗ࡉ ࡁ ࡇ ᣦ ࡋ ࠋ ࡓ㸪Jasak ) NVD ࢫ ࡇ 㸪 ࢡ 㧗ࡉ 㐣ᑠホ౯ࡍ ᗘ 㸪ᾋຊ㡯 ࡋ 㐠 ᪉ ᘧ ᐦ᥋ 㛵 ࡓ 㸪ィ Ᏻᐃᛶ ࡁࡃᕥ ࡍ ࠋࡋࡓࡀࡗ 㸪 ᗘ ᪉ ᘧ ᑐὶ㡯 Ᏻᐃᛶ ࡓ ࢫ ࡀ 㑅ᢥ ࠋ )㐠 ᪉ ᘧ ᑐὶ㡯 Filtered-㸪 ᗘ ศᕸᙧ ṇ ᤊ ࡇ ࡋ ࠋᮏሗ ィ ᯝ ᗘ ᚩ ࡇ ࡋ ࡀ㸪㐠 ᪉ ᘧ ᑐὶ㡯ࢫ ᪉ ᘧ ᑐὶ㡯 ࢫ ࡓィ ᯝ 㸪 ẼᏳᐃᗘ ኚ ࡉࡏࡓሙ ẚ ィ ᯝ ᙳ㡪 ᑠࡉ ࠋ㐠 ᪉ ࡀᏳᐃ ᪉ ࢣ ࢫ ᐇ ᘧ ᑐὶ㡯ࢫ Filtered- ሙ 㸪Ⅼ 㸪ᑡ ࡃ ᮏࢣ ࢫ ࢩ ࢩ ᗘᣑᩓ ᑐࡍ ᗘ ᪉ ᘧ ᑐὶ㡯ࢫ ࡀ 㸪ࡇ ࢫ ࡏ ࡓ 㝈య ἲ LES ẚ ࠋ ởᰁ ᣑᩓ ィ㸪ᐇ ศ ᗘ ࡋ ゝ ᯝ ᗘ ࡃ ࡋ ࠋ. ࢡ ࢫ ( ᗘ) ᑐὶ㡯 ᑐࡍ 㞳ᩓ ࢫ ཧ. ウࢣ ࢫ ẼᏳᐃᗘࡀᏳᐃ ሙ ᐇ㦂 ᯝ ẚ ᑐ ࡋ 㸪 ᗘ ᪉ ᘧ ᑐὶ㡯㞳ᩓ ࢫ ኚ ࡉࡏ㸪ィ ᯝ ᙳ㡪ウࡋࡓࠋ ࡓ㞳ᩓ ࢫ 㸪Filtered-㸪 ) ከ 㸪㐲 ⰾಙ㸪 Ụ㱟㑻㸪ᣢ ⅉ㸪 ⱥஅ㸸㧗ᒙᘓ ᚋὶᙅ 㢼ᇦ ࡅ ࢫᣑᩓᛶ 㛵ࡍ LES Durbin ᆺ k-ȝ ẚ ) ᐩỌ 㸪㯮 ᗈ 㸸༢య ᪉య ởᰁ ᣑᩓ ᑐ ࡋࡓ 㸪 ᮏᘓ Ꮫ ቃ ㄽ 㞟 7㸪7㸪pp.-㸪. RANS LES ẚ 㸪CFD ᘓ ഐ ởᰁ ᣑᩓ ண ᗘ 㛵ࡍ ࡑ 㸪 ᮏᘓ Ꮫ ቃ ㄽ 㞟 7㸪㸪pp.7-㸪 LST㸪 ࠋ㐠 ᪉ ᘧ ᑐὶ㡯 Filtered-㸪 ᗘ ᪉ ᘧ ᑐὶ㡯 ࡓࠋ. ウ ᯝ 9. ) Dispersion in a Spatially-Developing Turbulent Boundary Layer Flow, Computational Simulations and Applications, InTech, ᗘ 㞳ᩓ ࢫ ࡓ Filtered- ࡋࡓሙ 㸪ィ㛤ጞ ᗘ ᩓࡋ ศ ᖹᆒ 㛫ࡀᚓ ࡗࡓࠋࡑ ࡓ 㸪LST㸪 ẚ ᑐ ࡋࡓࠋ pp.-,. ) ᗘ T ศᕸ ࡍࠋLST ィ ᯝ ᕪ ࡀ ࡃ㸪ᐇ ᯝ ࡋ ࠋࡓࡔࡋ㸪. ) /H = 㹼 ࡎ ࢡ ᗘ ᕪ ࡀ ࡌ 㸪. ) ᯝ ᑐᛂ Ⰻ ࠋ௨ 7) ẼᏳᐃᗘ ᙳ㡪 ࡑ 㸪 ᮏᘓ Ꮫ ቃ ㄽ 㞟㸪79㸪7㸪 pp.7-㸪.7 ᗘ ᪉ ᘧ ᑐὶ㡯ࢫ ィ Ᏻᐃᛶ ィ ᗘ ୧㠃 ࡀ ࠋ 7. ᖺ ࡀቑຍࡋ 㝈య ἲ ᇶ ࡓ LES 㝿㸪㐺 ᑐὶ㡯㞳ᩓ ࢫ ࡘ ウ ࡗࡓࠋ ᑠ㔝ᾈᕫ㸪 ᮏᾈ 㸪㐨ᒸṊಙ㸪బ Ṍ㸸Ᏻᐃᡂᒙὶ ࡓ༢ ᘓ ᶍᆺ ởᰁ ᣑᩓ 㢼Ὕᐇ㦂㸪ᘓ ᒇ 㠃 ࢫᣑᩓ 㸪㐠 ᪉ ᘧ ᑐὶ㡯ࢫ Filtered- ࡓ ሙ 㸪 ࢡ ࢫ すὒᖹ㸸㠀ᅽ ᛶὶయゎᯒ ࡅ ᕪศࢫ ಖᏑ ᛶ㸪 ሗ ࢫࢱ ࢥ ࢣ ᱁Ꮚ ᕪศࢫ 㸪 ᮏᶵ Ꮫ ㄽ 㞟 B 㸪()㸪pp.9-㸪99. య LST 㛫 ᕪ ᑠࡉ ࠋ 㸪㐲᪉ ࢡ ᗘ 㐣ᑠホ౯ࡍ 㸪ᐇ ᯘఙ 㸪ᆏ 㸪ᐩỌ 㸪 Ἴ 㸪ᮏ 㸸LES ఫᏯ 㢼ᛶ 㛵ࡍ 㸪ࡑ ᶆ k-ȝ LES ఫᏯ Ẽὶᛶ ẚ 㸪 ᮏᘓ Ꮫ Ꮫ ㅮ₇ ᴫ㞟 D-㸪pp.-㸪 u rms /H㸪w rms/h ᴟഹ LST ᪉ࡀ ࡁ ࠋ ᅗ ᘓ ᶍᆺ ᚋ ࡅ 㛫ᖹᆒ ᗘ(C/C ) ศᕸ ࡍࠋ ᮏ㞞ᙪ㸪 ༤㐨㸪ᑠ ಙஓ㸪す Ẏ㸪 ཪ 㸪ᒾぢ㐩ஓ㸸ᐦ㞟ᕷ ᆅ ࡅ Ẽ 㢼ᛶ ホ౯ 㛤㸪ࡑ ὶయィ㒊 ศ 㛤㸪 ᮏᘓ Ꮫ Ꮫ ㅮ ₇ ᴫ 㞟 D-㸪 pp.79-77㸪 ᅗ 7~ ᘓ ᶍᆺ ᚋ ࡅ /H㸪u rms/h㸪w rms /H㸪㛫ᖹᆒ H. Nakayama: Large-Eddy Simulation of Turbulent Flow and Plume ) J. H. Ferziger, M. Peric: Computational Methods for Fluid Dynamics, rd Edition, Springer, 9) C. L. V. Jayatilleke: The influence of Prandtl Number and Surface Roughness on the Resistance of the Laminar Sublayer to Momentum and Heat Transfer, Prog. Heat Mass Transfer, pp.9-9, 99 ) D.K. Lily: A proposed modification of the Germano subgrid scale 9

8 closure method, Phys. Fluids A, -, 99 ) B. P. Leonard and H.S. Nikhafs: Sharp Monotonic Resolution Of Discontinuties Without Clipping Of Narrow Etrema, Computers & Fluids, Vol.9, No., pp.-, 99 ) H. Jasak, H.G. Weller and A.D. Gosman: High Resolution NVD Differencing Scheme For Aabitrarily nstructured Meshes, Int. J. Numer. Meth. Fluids, pp. 9,999 ) M.S. Darwish, F. Moukalled: schemes for unstructured grids, Int. J. Heat and Mass Transfer, pp.99-, ) P. K. Sweby: High Resolution Schemes sing Flu Limiters for Hyperbolic Conservation Laws, SIAM J. Numer. Anal., (), pp.99, 9 ), f, Df( D - ) OpenFOAM f ) D D D ) i divergence dv dv ds i i i i i n div All surfaces V V V i i S surf. surf. surf. V i ) OpenFOAM LST(-pwind Stabilized Transport) ) ) D r D D D ) ) θ - DD ( D) b/a. b/a θ. OpenFOAM filtered([., ]) D DD Target Face f b. a a mac, d b min, DD D c D d mad, DD D D DD D D Sup. Fig. pwind blending factor of Filtered- DD Filtered- ( pwind) ) - H( =.7m/s=.79m/s)±% Filtered- LST Sup. Fig. Vertical profiles of the ehaust point without the building - CFD CFD LST T [deg.c] at inlet T [deg.c] at the ehaust point Sup. Fig. Vertical profiles of temperature at inlet and the ehaust point 7) Limited- Sweby ) - -pwind:φ=r nd order region Φ(r) :Φ= pwind:φ= r Φ(r) Limited- r Sup. Fig. Limited- on Sweby diagram ) LES CFD LES

9 CONVECTION CONVECTION TERM DISCRETIZATION TERM DISCRETIZATION FOR LARGE FOR EDDY LARGE SIMLATIONS EDDY SIMLATIONS BASED ON THE BASED ON FINITE THE VOLME FINITE VOLME METHOD METHOD Effect of of thermal stratification on the dispersion characteristics of rooftop ehaust, Part Hiroki ONO * Hiroki Hiroshi ONO TAKIMOTO *, Hiroshi TAKIMOTO ** Takenobu **, MICHIOKA Takenobu MICHIOKA *** and Ayumu *** SATO **** and Ayumu SATO **** * Visiting Researcher, Central Research Institute of Electric Power Industry, Dr. Eng. ** * Research Visiting Researcher, Scientist, Central Research Institute of of Electric Power Industry, Dr.Eng. ** Research Scientist, Central Research Institute of Electric Power Industry, Dr.Eng. *** Research Scientist, *** Central Research Institute of Electric Power Industry, Dr. Eng. (Kinki niversity at the present) Research Scientist, Central Research Institute of Electric Power Industry, Dr.Eng. ****Senior **** Senior Research Research Scientist, Scientist, Central Central Research Institute of of Electric Power Power Industry, Dr. Dr.Eng. Large Eddy Simulation (LES) is considered to be the most likely method to calculate concentration dispersion in a situation involving comple flow. Validations of the LES based on a Finite Difference Method (FDM) with staggered grid have carried with wind tunnel eperiments. Currently, computational fluid dynamics (CFD) codes based on a Finite Volume Method (FVM) with collocation grids are popular because of their ability to handle comple geometries such as curved surfaces and geographical features. However, it has been pointed out that accuracy between the structured collocation grid and the structured staggered grid cannot be discussed on the same contet because the structured collocation grid violates the conservation of kinetic energy. So, validation of the LES with unstructured collocation grids is significant even if previous works with structured staggered grids are available on a target field. It is important for the LES to minimize numerical diffusion caused by discretization of the convection term. Most previous work based on the FDM has used a second-order central differencing scheme (for a momentum transport equation) and/or combined with a high-order upwind differencing scheme (for a concentration transport equation). Conversely, blending of first- and second-order schemes is often used in the FVM, because it is difficult to use a thirdor higher-order scheme without spoiling the versatility of the FVM based CFD code. However, the applicability of these blended schemes for the LES is not apparent. Therefore, we carried validation of the LES with these schemes on our previous wind tunnel eperiment. From the results, the choice of scheme for the convection term of the momentum transport equation is observed to be much more sensitive than in other transport equations. A linear-interpolation scheme, which has the least numerical diffusion, has the best agreement with measurements; however, unphysical oscillations are introduced with this scheme. A filtered-linear scheme, which is based on linear interpolation blended with small amounts of upwind only if oscillation appears, is the best method from the point of view of accuracy and boundedness. Conversely, a total variation diminishing () scheme is effective for strictly bounded scalar quantities such as concentration. An unbounded scheme cannot reproduce an eact profile of such scalars. The scheme is also effective for the temperature transport equation, because it is closely related to the momentum transport equation and can influence calculation stability. In conclusion, the FVM based LES with the filtered linear scheme (for the momentum transport equation) and the scheme (for the scalar transport equation) scheme is practical as it can present measurements with high accuracy regardless of atmospheric stability. ( 年 9 月日原稿受理, 年 9 月日採用決定 )

< BDF91E58E9F90A291E38AEE94D595F18D908F E696E6464>

< BDF91E58E9F90A291E38AEE94D595F18D908F E696E6464> 近畿大学次世代基盤技術研究所報告 Vol. 9(2018)55-59 Additive Manufacturing *1 *2 *3 The Recent Trend on Additive Manufacturing Technology: A Review Hideki KYOGOKU *1, Toshi-Taka IKESHOJI, Makiko YONEHARA The recent trend