HomeUpSearchMail
NEW

Simuleren van wind bij Ren artikel

Simuleren van wind bij Ren artikel by Victor Reijs is licensed under CC BY-NC-SA 4.0

Introductie

• Informatie over de molen omgeving
  
• Eigenschappen van obstakels
• Maken 3D model in SketchUp
  
• Editing 3D model in SIMSCALE
• Configureren van de CFD
• Uitvoering van de CFD
• Conclusies
• Referenties
• Acknowledgements

Sorry for the two langauges gebruikt in deze webpagina!

Vette paarse tekst heeft nog extra aandacht nodig.

Findings by Ren

Information on surroundings

• Ren's representation [2023] of the results don't look to be consistent:
• The color scale in Figure 16 (right top of figure) has a different labelling than in the individual curves in Ren's simulation graphs. For instance middle-dark blue in the color scale is from 2.0 to 3.0m/sec, while in the graph it is from 1.0 to 2.0m/sec:
  
• The curve that starts close to z/H=1 (in Figure 16) shows the speed of 8m/sec, so the title of the graph seems to match the numbers in the graph. So the color scale migth be wrong.
• I have asked feedback (in Dec. 2023 and Dec. 2024) from the authors of [Ren, 2023]; until now no reaction. See also next remark.
  
• In the below the numbers in the graph have been used (and not the color scale).
  
• Comparison measurments and simulations
  
• The definition of u₀ in Figure 13, it looks to be the wind speed in the simulations (Figure 16) at x=-2H_tree and at the same height as where measured behind the obstacle.
• Say we look at the x=3H_tree graph (most right in Figure 13).  In below graph you see dashed yellow line (which were derived from Figure 16) and the black dashed line (which is according to Ren's legend their simulation in Figure 13). So things are comparable. Another proof that the numbers in the graph are correct (and not the colorscale).
  
• The location is where the yellow pin is located (Zjejiang Agriculture and Forestry University in China):
  
• No z₀ was given, but looks to be parkland, bushes; numerous obstacles, so a z₀ = 0.5m is assumed. It used the standard k-epsilon turbulence model [Ren, 2023, section 2.4].
  
• Ren's measurements [2023, section 3.2] were done at 5Bft (8 to 10.7m/sec@10m), so in SIMSCALE simulations an u_Htree=8m/sec is used.
  
• The Ren's tree measured [Ren, 2023, Figure 8] H_tree =6.8m:
  
  Ren's measurements [2023, Figure 13] are below represented in a different graphical way (using the coloring scheme of the below SIMSCALE graphs). As u₀ is not given for the measurements, the wind speed in Figure 16 has been used at x=-2H_tree and at the same height as where measured behind the obstacle:
  
  
  
• Ren's modelled tree [Ren, 2023, 1a]:
  
  If we compare the measured (colored) values [Ren, 2023, Figure 13] with the simulated (black lines; standard k-epsilon and AP=0.52@8m/sec) values [Ren, 2023, Figure 16 (a2) T-S 8m/s]:
  
  The speeds values are not mapping, see also inconsistency above. But the form of the curves and colors looks to map, which is a good sign.
  Remark: This stil needs to be interpreted better! Is the u₀ wrong?
  

Modelling the proxies for a tree

Different tree proxies

Four proxies for a tree have been checked with SIMSCALE:

• A blobbed proxy
  
  A CAD-model available for a leaved tree is used for this blobbed tree proxy. No porocity was added as the model itself already has openings included.
  Crown: diameter 9m, height 5m with a trunk of 1.8m; Realizable k-epsilon; Meshing fineness=9
• A stacked cylinder proxy
  
  
  Crown: diameter 9m, height 5m raised up by 1.8m for its trunk; Realizable k-epsilon; Meshing fineness=5
  The 3D model of this proxy has problems with higher meshing (F=7.5): not getting simulation convergence. So not further pursued.
  
• A single cylinder proxy
  See for an example below.
  A single cylinder from top to bottom and left to right leaves of the crown.
  Crown: diameter 9m, height 5m raised up by 1.8m for its trunk; f=1.2; Realizable k-epsilon; Meshing fineness=9
• An effective single cylinder proxy
  See for an example below.
  A single cylinder from top to bottom and left to right leaves of the crown migth be too wide, so an effective width and heigth is estimated.
  Crown: diameter 8.58m, height 4.25m raised up by 1.73m for its trunk; f=1.4; Realizable k-epsilon; Meshing fineness=9

Working in SketchUp

• 3D model of five 'trees' (which will get different porosities in SIMSCALE) in SketchUp. Stacked cylinder, single cylinder and effective single cylinder models have been implemented. In case of the blobbed proxy only one 'tree' was modelled, as no porosity has been used.
  
  

• Include at the back (50m or 100m from the trees) a small (dummy) box (which is needed for being able to simulate things in CFD in case all other objects will have porosity)
  
• Sluit af als klaar: 'Save' en 'Download' -> 'STL'
• Effort: 0.5 hours for each proxy
  

Editing 3D-model in SIMSCALE

We are making use of SIMSCALE. Make sure you make a Community Plan account. This Community Plan allows only for 10 simulations (but one can do many more; but with more manual work). Anyway make sure you have a good CAD-model and that you train/educate yourself around SIMSCALE.
The following steps can be done:

• The different proxy simulations are here:
• Blobbed proxy
• Stacked, Single and Effective Single cylinder proxies
  
• Import het STL model
• Edit this (not yet rotated) model by: Edit a copy
  
• Make a Flow Volume -> External waarbij de poreuze poxies (bomen: 'Excluded parts') niet deelnemen aan het 'External flow volume'.
  
  Zie voor richtlijnen mbt vrije afstanden rondom het model hier [Franke, 2007]:
  
  • Xmin, Xmax: minimaal 5*hoogte/grootte hoogste obstakel [Franke, 2007, page 17] extra aan linker en rechter kant van 3D-model. In dit geval ~5*~6.8m (fence): ~ 40m
  • Ymin: Distance in front of model (in Y-direction) minimaal 8*hoogte/grootte hoogste obstakel [Franke, 2007, page 18] extra aan voorkant van 3D-model. In dit geval ~8*~6.8m (fence): ~ 60m
  • Ymax: Distance at back of model (in Y-direction) minimaal 15*hoogte/grootte hoogste obstakel [Franke, 2007, page 18] extra aan achterkant van 3D-model. In dit geval ~15*~6.8m (fence): ~ 100m
  • Zmin: Ground (in Z-direction) minimaal om invloed van modelerings afrondingen (gaps) te voorkomen:  ~ 0.05m
  • Zmax: Height (in Z-direction) minimaal 6*hoogte/grootte hoogste obstakel [Franke, 2007, page 17]. In dit geval ~6*~6.8m (fence): ~ 50m
  • There is also a directional blockage ratio defined (mainly imortant for the lateral direction) [Blocken, 2015, formula (13)): being BR_L=L_building/L_flowvolume <=17%
    For this model this BR_L is observered, as BR_L=9/89=10%
    
  • By the way, the default values in SIMSCALE were always somewhat larger than the above, so its defaults are also ok. Except for Zmin; I would put that always around 0.05m.
• Delete the dummy box, but not the tree(s)
  
• Save
• Effort: 0.25 hours for each proxy.

Configureren van de CFD

Take the following steps (if not default, it is explicitly shown in below steps)

• Goto SIMULATIONS +
• Incompressible -> Turbulence-model -> Realizable k-epsilon [Franke, 2007, page 14] [Franke, 2007, section B.2.1 for PWC]
  Standard k-epsilon and k-omega SST will also be used to provide insight and make comparisons.
  
  
• Materials -> Air -> Apply
  Assigned Volumes -> Flow region
  
• Boundary conditions
• Velocity Inlet
  Assigned Faces -> the wind side
  (U) Velocity -> U_y -> ABL Formula (9.08m/sec at 10m and z₀=0.5m, the wind direction is 180deg [S])
  (0.41*9.08/log((10+0.5)/0.5))/0.41*log((z+0.5)/0.5)
  
  Turbulence -> Fixed value
  (k) Turb. kinetic energy -> ABL derived Formula
  (0.41*9.08/log((10+0.5)/0.5))/(0.41*(0.09)^0.5)*1/(z+0.5)
  
  (ω) Specific dissipation rate -> ABL derived Formula
  (0.41*9.08/log((10+0.5)/0.5))/(0.41*(0.09)^0.5)*1/(z+0.5)
  
  (ε) Dissipation rate -> ABL derived Formula
  (0.41*9.08/log((10+0.5)/0.5))^3/(0.41)/(z+0.5)
  
  
  Save
  
  
• Pressure outlet
  Assigned Faces -> opposite inlet side
  Save
  
  
• Wall
  Assigned Faces -> two sides and top
  (U) Velocity -> Slip
  Save
  
  
• Custom (ground)
  Assigned Faces -> bottom side
  Wall roughness -> On
  Roughness height -> 5m (k_S; 5m (~10*z₀), [Blocken, 2015, formula (15)]. as SIMSCALE is based on OPENFOAM and C_s=1)
  Roughness constant -> 0.5
  Save
  
  
• Mesh
  25m fence: Fineness = 7.5 (1.5Mcells)
  
  
• Mesh -> Refinements -> Inflate boundary layer
  
• Mesh -> Refinements -> Volume custom sizing
  
• Mesh -> Geometry primitives
  
  
• Advanced concepts -> Porous media -> Darcy-Forchheimer Medium
  Forchheimer coefficent f=1.4, 1.2, 1, 0.8, 0.6
  
• Effort: 0.75 hour for each proxy.

Uitvoering van de CFD

• Simulation Runs +
• In all graphs the magnitude of the wind speed is provided.

• Proxy	Tubulence model	Mesh fineness	Porosity	Output (x and y-axis and in steps of Htree [=6.8m]). Gray contours are Ren's simulation at uHtree=8m/sec
  Ren's measurement and model	k-epsilon	N/A	AP=0.52
  Effective single cylinder	Realizable  k-epsilon	9	f=1.4
  	k-epsilon	9	f=1.4
  	k-omega SST	9	f=1.4
  Blobbed	Realizable  k-epsilon	9	N/A
  	k-epsilon	9	N/A
  	k-omega SST	9	N/A
  Single cylinder	Realizable  k-epsilon	9	f=1.2
  	k-epsilon	9	f=1.2
  	k-omega SST	9	f=1.2
  Stacked cylinder	Realizable  k-epsilon	5	f=1.2

• Ren's measurements and simulation don't match! There is some 2m/sec between the contours. This is possibly related to the inconsistancy mentioned earlier.
  
• Realized k-epsilon (recommended by Franke [2007]) and standard k-epsilon (used by Ren [2023]) provide quite similar results.
• k-omega SST always provided a higher wind profile for leeway positions. Remember k-omega SST is not a recommendated CFD for urban situations [Franke, 2007].
• The Blobbed proxy provides a wind profile closest to Ren's wind profile, but it has the problem that the porosity is not explictily defined, so it makes adjusting to other trees species or summer/winter effect more difficult.
• Meshing sizing recommendations [Green Mark department, 2007]:
  
  
• When increasing the number of meshing cells (by using a Cartesian Box around the proxy towards the end of the flow volume), at a certain size the wind profile becomes (very) different. With porous medium (white, yellow background) this non-convergence happens quite explictly (0.2m), with a solid obstacle (green background) this happens perhaps at smaller mesh default size (0.15m).
  So what is the best mesh size: 0.3m or is 0.2m (with a lot of likely detail?) the best????
  

  Proxy	Turbulence model	Mesh fineness F is for whole external flow volume DS is for a Cartesian box over the tree towards end	Porosity	Output (x and y-axis and in steps of Htree [=6.8m]). Gray contours are Ren's simulation at uHtree=8m/sec	Mesh size
  Effective single cylinder	Realizable  k-epsilon	F=5.1 & DS=0.2m*	f=1.4		21M
  		F=5.1 & DS=0.3m			4.8M
  		F=5.1 & DS=0.4m			4.3M
  		F=5.1 & DS=0.5m			1.6M
  		F=9			6.6M
  		F=5.1 & DS=0.3m	f=0.9		4.8M
  		F=5 & DS=0.15m*	NA		19M
  		F=5 & DS=0.2m	NA		18M
  		F=5 & DS=0.3m	NA		4.2M

  * For small mesh default size, the porous medium's (DS=0.2m) and solid obstacle's (DS=0.15m) wind profile results seem not to converge.
  
• At a leeway distance of 10H_tree all turbulence models, sensible mesh sizes and sensible porosites look to have similar wind profiles. This is expected as the influence of an obstacle becomes much smaller after 10H, so the sensitivity is not that large for leeways of 10H_tree. For nearby distances there is more variation. 
• My simulated solid obstable's wind profile looks very close to Ren's simulations (of a porous medium)!
  Remark: In some way a strange result.
  
• Effort: 5 hours per proxy (incl. sensitivity analysis).

Conclusions

• Most default values of SIMSCALE look ok, except:
• Use Realizable k-epsilon
• Zmin should be always 0.05m
• The higher the Fineness the better, somewhere between 7.5 and 9 (instead of default of 5), or better: use your own mesh configuration.
• Thoughs about the meshing:
  
• All mesh metrics stay within their defined min and max range; at least for Fineness tested between 5 and 9.2 and this 3D-model.
  All residuals are smaller than 10^-3.
  
• hgsajkf kfjhajkfh
  
• An overview of outstanding issues/questions/etc is here.

References

Acknowledgements

Disclaimer and Copyright
HomeUpSearchMail