NEW LES CAPABILITY OF ADREA-HF Mr. Nektarios Koutsourakis 1 Dr. - PowerPoint PPT Presentation

NEW LES CAPABILITY OF ADREA-HF Mr. Nektarios Koutsourakis 1 Dr. Alexander Venetsanos 2 Prof. John G. Bartzis 1 Mr. Ilias Tolias 2 1 University of West Macedonia Greece 2 NCSR “Demokritos” Greece Harmo 13 – LES of ADREA-HF

ADREA-HF CFD CODE Constantly developing • Latest additions: – LES (Large Eddy Simulation) – Parallel solver – Arbitrary number of species – Combustion – GUI pre and post processor (called EDes) • Advantages: – Robust, powerful and general – Specialization in environmental applications – “ADREA + dispersion”: 29 hits in Scopus Harmo 13 – LES of ADREA-HF

ADREA-HF CFD CODE GUI environment for pre and post processing Harmo 13 – LES of ADREA-HF

ADREA-HF CFD CODE Post processing example with ADREA-HF GUI C1 Flow 2*C1 Sources Harmo 13 – LES of ADREA-HF

LARGE EDDY SIMULATION Between DNS and RANS • Navier-Stokes math. analysis: impossibly difficult • DNS: Fully-resolved NS – Cost  Re 2.75 (for wall-bounded flows  Re 3.5 ) • RANS: Time-averaged NS – Fast , widely tested, usually accurate “enough” • LES: Spatially-filtered NS – Cost near wall:  Re 2.4 ! (at outer layer  Re 0.5 ) – Converges to DNS. As in DNS, LES needs: a) averaging b) demanding boundary conditions – Suggested were RANS fails: transient, separated flow Harmo 13 – LES of ADREA-HF

ENERGY CASCADE LES solves most of the turbulence E(k) Energy containing Inertial Viscous range subrange subrange RANS: All modeled Solved from LES Solved from DNS k  3/4 k=1/l k=1/ =Re /l 0 0 0  : l : 0 Kolmogorov integral scale scale Harmo 13 – LES of ADREA-HF

ENERGY CASCADE LES solves most of the turbulence Universal equilibrium range E(k) ln 2 u l 0 0  -5/3 2/3 E(k)=C k k Energy Very large containing Inertial Viscous scales range subrange subrange range RANS: All modeled Solved from LES Solved from DNS ln(k l ) 0  3/4 k=1/l k=1/ =Re /l 0 0 0   6l l /6 60 8  0 0 : l : 0 Kolmogorov integral 80% of energy 90% of dissipation scale scale Harmo 13 – LES of ADREA-HF

LES EQUATIONS AT ADREA-HF Compressible volume-filtered Navier-Stokes • Filtered equations:     ( u )     i 0 p rT   t x i  2 u          l R       ( u u ) ( ) l k ( u ) p 2 S      i j ij ij ij ij ij i 3 x     k t x x x j i j     u 1 u          j  i R S u u u u     ij 2 x x ij i j i j   j i • Smagorinsky model of the residual stress tensor 1           1 3 Cs=0.1 R 2   2 S ; ( C ) 2 S S V ij kk ij t ij t s ij ij 3 Harmo 13 – LES of ADREA-HF

LES AT ADREA-HF Numerics • ADREA/SIMPLER algorithm • Central differences convection scheme • Solver: Parallel BiCGstab with parallel Schwarz preconditioner. Speedup with 2 CPUs up to 1.7 • For more: – Venetsanos, A. G., E. Papanikolaou and J. G. Bartzis, 2010: The ADREA-HF CFD code for consequence assessment of hydrogen applications. Int. J. Hydrogen Energy, 35, 3908. Harmo 13 – LES of ADREA-HF

LES Boundary and initial conditions • Initial conditions should also create turbulence – “Vortex generator” may be needed – Simpler solution: big over-imposed disturbance • Boundary conditions should also retain turbulence – Difficult to do explicitly – Simpler solution: cyclic BC • Near wall: – Well-resolved LES requires z + =1 – “Wall - function” -type solutions for coarser grids Harmo 13 – LES of ADREA-HF

TEST CASES SETUP a) Channel flow • 1D fully-developed channel flow: the classic LES test – DNS of Moser et al, 1999 Re = 8000 – 90000 cells – Cyclic BC – z + = 1 – CFL max < 0.3 – C s = 0.065 – Mass flow correction to control Re – Time 50s = 160 passes Averaging from 20s Harmo 13 – LES of ADREA-HF

TEST CASES SETUP b) Street canyon • 2D street canyon: the most basic urban case – Water channel of Li et al, 2009 Re = 12000 – 300000 cells – Cyclic BC: sequence of identical canyons – z + = 1 – Time 1000s = 250H/U Averaging from 400s Harmo 13 – LES of ADREA-HF

FULLY DEVELOPED CHANNEL FLOW Pressure isosurfaces and velocity vectors Harmo 13 – LES of ADREA-HF

FULLY DEVELOPED CHANNEL FLOW Comparison with DNS and other LES 14 DNS ADREA LES DNS u'u'/u*2 20 STAR-CD LES 12 DNS w'w'/u*2 STAR-CD DNS v'v'/u*2 DNS u'w'/u*2 DNS STAR-CD 10 ADREA u'u'/u*2 ADREA w'w'/u*2 15 Reynolds stress/u*2 ADREA v'v'/u*2 ADREA 8 ADREA u'w'/u*2 STAR-CD u'u'/u*2 STAR-CD w'w'/u*2 U+ ADREA 6 STAR-CD v'v'/u*2 10 STAR-CD u'w'/u*2 DNS 4 2 5 0 0 -2 0 2 4 6 0 0.5 1 ln(z+) z/H Harmo 13 – LES of ADREA-HF

STREET CANYON Canyon 1: Building width=2H. Real time movie Harmo 13 – LES of ADREA-HF

STREET CANYON Canyon 2: Building =1H. Comparison with exp. x=0.25H x=0.25H 1.2 1.2 Li et al 2008 LES ADREA 1 1 Experiment Li et al Experiment 0.8 0.8 z/H z/H ADREA LES Li et al fine-LES 0.6 0.6 u’ profile near 0.4 U profile near 0.4 upwind wall upwind wall 0.2 0.2 U/Uref u'/Uref 0 0 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 0 0.05 0.1 0.15 0.2 0.25 Harmo 13 – LES of ADREA-HF

STREET CANYON Canyon 2: Building =1H. Comparison with exp. x=0.25H x=0.75H x=0.25H x=0.75H 1.2 1.2 1.2 1.2 1 1 1 1 0.8 0.8 0.8 0.8 z/H z/H z/H z/H 0.6 0.6 0.6 0.6 0.4 0.4 0.4 0.4 0.2 0.2 0.2 0.2 U/Uref U/Uref W/Uref W/Uref 0 0 0 0 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 -0.2 0 0.2 -0.2 0 0.2 x=0.25H x=0.75H x=0.25H x=0.75H 1.2 1.2 1.2 1.2 1 1 1 1 Li et al Experiment 0.8 0.8 0.8 0.8 z/H ADREA LES z/H z/H z/H Li et al fine-LES 0.6 0.6 0.6 0.6 0.4 0.4 0.4 0.4 0.2 0.2 0.2 0.2 u'/Uref u'/Uref w'/Uref w'/Uref 0 0 0 0 0 0.05 0.1 0.15 0.2 0.25 0 0.05 0.1 0.15 0.2 0.25 0 0.05 0.1 0.15 0.2 0.25 0 0.05 0.1 0.15 0.2 0.25 Harmo 13 – LES of ADREA-HF

DISCUSSION LES of ADREA-HF is competitive • ADREA performs well against exps and other LESs • Channel: C s correction needed near wall • General tendency of Reynolds stresses predicted • Canyon: Performance very close to fine-grid LES • Profiles similar shape to experimental ones • Experiment stronger vortex with more turbulence: – In experiment vortex generators were used – In experiment 3D effects were present • RANS (not shown) was also good in these tests Harmo 13 – LES of ADREA-HF

CONCLUSIONS LES is not for everyday use • ADREA-HF has now tested LES • LES is not RANS++ • LES orders of magnitude more expensive than RANS • Near wall treatment a key point for real-world cases • What now: – Refine ADREA LES – Add and test more LES options Harmo 13 – LES of ADREA-HF

THE END … is also a beginning For any suggestions: nk@ipta.demokritos.gr Harmo 13 – LES of ADREA-HF