• No results found

1. Lid driven flow in a cavity [Time: 1 h 30 ]

N/A
N/A
Protected

Academic year: 2021

Share "1. Lid driven flow in a cavity [Time: 1 h 30 ]"

Copied!
10
0
0

Loading.... (view fulltext now)

Full text

(1)

Hands on computer session:

1. Lid driven flow in a cavity

[Time: 1 h 30’]

Objects

• choice of computational domain, problem adimensionalization and definition of boundary con-ditions;

• influence of the mesh refinement (uniform mesh vs. near wall refined mesh, increasingly refined meshes);

• evaluation of different discretization schemes: accurancy and dissipative effects;

• identification of reference data, calculation of quantities to verify simulation accuracy (stream-function, velocity profiles along vertical and horizontal midlines);

• Richardson extrapolation as a discretization error extimator.

Physical problem

A square cavity filled with flow; the top wall moves with a uniform velocity U: • isothermal laminar incompressible problem;

• primary central vortex and secondary vortices at the corners; • complex flow patterns at increasing Reynolds number; • stability bounds for Central Difference Scheme.

(2)

top wall velocity U = 1

cavity width L= 2

cavity height H= 2

Reynolds number Re= 1000 andRe= 10000 fluid adimensional density rho= 1

fluid adimensional viscosity mu= 2/Re

Worksheet

Mesh generation • mesh.4a: uniform mesh 20x20x1 cells;

• mesh.4b: near wall refined mesh 20x20x1 cells; • mesh.4c: near wall refined mesh 40x40x1 cells; • mesh.4d: near wall refined mesh 80x80x1 cells;

→ create a subdomain (a quarter domain) with block structure grid → create near wall refinement (uniform space factor)

→ use CGEN-REFLECT to replicate subdomain regions Boundary conditions → use symmetry B.C. to set up a bidimensional simulation

→ use Wall B.C.(slip option) at the top wall

→ use Wall B.C.(no-slip option) at the bottom and side walls

Solution method • steady state simulation at Re=1000 (mesh 4a,4b,4c), solve forV, P • steady state simulation at Re=10000 (mesh 4c,4d), solve forV, P → use of differencing scheme (UD,CD,MARS)

Results considerations over velocity field (velocity vectors and stream function con-tour plots)

extraction of u and v velocity component profiles along the vertical and the horizontal midlines

comparison to the literature results Richardson extrapolation

References

Ghia, U., Ghia, K. N., and Shin C.T., (1982), “High-Re solutions for incompressible flows using the Navier-Stokes equations and a multigrid method”, J.Comput.Phys , 48, 387-411.

Shon, J.L., (1988), “Evaluation of FIDAP on some classical laminar and turbolent bench-marks”, Int. J. Numer. Meth. Fluids, 8, 1469-1490.

Nonino, C., and Croce, G., (1997), “An equal-order velocity-pressure algorithm for incompressible thermal flows, part 2: validation”,Numer. Heat Transfer, 32, 17-35.

(3)

Geometry and Results

Figure 1: Refined grid 40x40

Figure 2: Streamfunction plot: right, Re=1000; left, Re=10000 [Ref.3]

(4)

Re1000: u and v velocity components along vertical and horizontal midlines

Figure 4: mesh with 20x20 cells: left,uniform mesh; right: near wall refined mesh.Square[Ref.3]

(5)

Re10000: u and v velocity components along vertical and horizontal midlines

(6)

Reference values

u components along vertical midline

Re=1000 Re=10000 y u u 1.0 1.0 1.0 0.9532 0.65928 0.47221 0.9376 0.57492 0.47783 0.9218 0.51117 0.4807 0.9062 0.46604 0.47804 0.7032 0.33304 0.34635 0.4688 0.18719 0.20673 0.2344 0.05702 0.08344 0.0 -0.0608 0.03111 -0.0938 -0.10648 -0.0754 -0.4374 -0.27805 -0.23186 -0.6562 -0.38289 -0.32709 -0.7968 -0.2973 -0.38 -0.8594 -0.2222 -0.41657 -0.875 -0.20196 -0.42537 -0.8906 -0.18109 -0.42735 -1.0 0.0 0.0

v components along horizontal midline

Re=1000 Re=10000 x v v 1.0 0.0 0.0 0.9376 -0.21388 -0.54302 0.9218 -0.27669 -0.52987 0.9062 -0.33714 -0.49099 0.8906 -0.39188 -0.45863 0.8126 -0.5155 -0.41496 0.7188 -0.42665 -0.36737 0.6094 -0.31966 -0.30719 0.0 0.02526 0.00831 -0.5312 0.32235 0.27224 -0.5468 0.33075 0.28003 -0.6874 0.37095 0.3507 -0.8124 0.32627 0.41487 -0.8438 0.30353 0.43124

(7)

MACROS

!**************** ! Geometry.MAC * !****************

!——————————————————

!*** complete mesh generation by symmetry along x and y !——————————————————

get vmax mxve cset all

cgen,2,vmax,cset,cset,1,vref,1,2 cset all

cplot

get vmaxnew mxve

cgen,2,vmaxnew,cset,cset,1,vref,1,1 cset all cplot vmer all vcomp all ccomp all cset all cdirection,1,0,1,1,1 restructure,50000,,default, $y vset news vran 0,49000,1 vcomp all $y cset all cplot !**************** !Umidplane-x0-GU.MAC * !**************** !————————————————————————————————–

!*** y coordinate and u velocity component extraction on vertical midline to screen and to file.usr !————————————————————————————————–

vset all

vset subset gran -0.0001 0.0001,,, -0.01 0.01 greset

y

getv, None ,SU gpost,vset,,,next frame,1,xreg,init,4 frame,1,xtitle,4.3,1.5 Y LOCATION frame,1,yreg,init,9 frame,1,ytitle,0.5,5.5 SU term,,X gdraw,1 rlabel,9,on

(8)

bart,9,9,90,1 gredraw oper,getv,x,1,1 oper,getv,y,1,2 oper,getv,z,1,3 oper,getv,su,4 savu,file.usr,both,user,vset I9,6X,4G16.9 close file.usr !**************** ! Vmidplane-y0-GU.MAC * !**************** !————————————————————————————————–

!*** x coordinate and v velocity component extraction on horizontal midline to screen and to file.usr !————————————————————————————————–

vset all

vset subset gran ,,, -0.0001 0.0001 -0.01 0.01 greset y getv, None ,SV gpost,vset,,,next frame,1,xreg,init,3 frame,1,xtitle,4.3,1.5 X LOCATION frame,1,yreg,init,9 frame,1,ytitle,0.5,5.5 SV term,,X gdraw,1 rlabel,9,on SV lint,9,9,3,1 symt,9,9,8,1 $rtab,9,y,y bart,9,9,90,1 gredraw oper,getv,x,1,1 oper,getv,y,1,2 oper,getv,z,1,3 oper,getv,sv,4 savu,file.usr,both,user,vset I9,6X,4G16.9 close file.usr

(9)

!**************** ! U-V-midplane.MAC * !****************

!———————————————————————

!*** comparison of u and v velocity profile from different file.pst

!———————————————————————-load exe4-UD.pst greset

y vset all

vset subset gran -0.0001 0.0001,,, -0.01 0.01 greset

y

numreg,100 getv, None ,SU gpost,vset,,,1 sens dele all vset all

vset subset gran ,,, -0.0001 0.0001 -0.01 0.01 getv, None ,SV gpost,vset,,,12 frame,1,xreg,init,14,9 frame,1,yreg,init,20,4 frame,1,xrange,-1,1 frame,1,xtitle,4.94,1.6 U frame,1,yrange,-1,1 frame,1,ytitle,0.71,5.51 V rlabel,20,on SV lint,20,11,3,1 symt,20,20,8,1 $rtab,20,y,n bart,20,20,90,1 rlabel,4,on Y LOCATION lint,4,4,3,1 symt,4,4,8,1 $rtab,4,y,n bart,4,4,90,1 term,,x gdraw,1 load exe4MARS.pst vset all

vset subset gran -0.0001 0.0001,,, -0.01 0.01 getv, None ,SU

gpost,vset,,,24 sens dele all

(10)

frame,1,xrange,-1,1 frame,1,xtitle,4.94,1.6 U frame,1,yrange,-1,1 frame,1,ytitle,0.71,5.51 V rlabel,44,on SV lint,44,3,3,1 symt,44,44,8,1 $rtab,44,y,n bart,44,44,90,1 rlabel,27,on Y LOCATION lint,27,27,3,1 symt,27,27,8,1 $rtab,27,y,n bart,27,27,90,1 term,,x gdraw,1

References

Related documents

We carry out a direct search for bar-like non-circular flows in intermediate-inclination, gas-rich disk galaxies with a range of morphological types and photometric bar

The eircom F2000 eFibre Modem is a next generation voice gateway that supports very-high-data-rate digital subscriber line 2 (VDSL2) uplink, one Giga Ethernet uplink

When a Member is receiving concurrent benefits from Medicare, the PERS Select Supplemental Plan payments for covered charges are provided according to whether the provider

For example, if most individuals have strong preferences for organ donation and families observe an uninformative indicator of donation preferences, a presumed consent law may create

Our algorithms are very efficient (logarithmic rounds or better with just polylogarithmic communication work per node per round) whenever the combinatorial dimension of the

Prevalence of metabolic syndrome diagnosed using three different definitions and risk of ischemic heart disease among Kaunas adult population.. Medicina

Hypothesis: we assumed that the introduction of special exercises with a partner into the training process, aimed at the developing of motor qualities, would allow bottom acrobats

Recasting his economic theory of discrimination in the context of the mortgage lending market, Becker argues that bigoted lenders are willing to sacrifice the profit they might earn