Computational fluid dynamics (CFD) simulations were used to analyze natural ventilation and ventilative cooling in buildings. Reduced-scale experiments with particle image velocimetry (PIV) were conducted to validate the CFD models. The CFD simulations analyzed transitional indoor airflow and the effects of ventilation louver slat angle on air exchange efficiency and heat removal effectiveness.
2. PAGE 2
Contents
1. CFD: what and why?
2. Natural ventilation: indoor airflow
o Computational Fluid Dynamics simulations (+ experiments)
3. Natural ventilation: ventilative cooling
o Computational Fluid Dynamics simulations
4. Natural ventilation: stadium study
o Computational Fluid Dynamics simulations (+ experiments)
5. Closing remarks
3. PAGE 3
CFD: what and why?
• What:
o CFD = Computational fluid dynamics
o CFD is “solving fluid flow problems numerically”
o "CFD is the art of replacing the integrals or the partial derivatives
(as the case may be) in the Navier-Stokes equations by
discretized algebraic forms, which in turn are solved to obtain
numbers for the flow field values at discrete points in time and/or
space.”
(John D. Anderson, Jr. 1995).
o Computational Fluid Dynamics is a tool that allows us to solve
flow problems that do not have known analytical solutions and
cannot be solved in any other way.
4. PAGE 4
CFD: what and why?
• Why:
o Understanding & interpreting (“numerical experiments” to provide
information if experiments are not possible, or in addition to
experiments)
o Designing (“future objects” on which experiments are not yet
possible, simply because they do not yet exist)
5. PAGE 5
CFD: what and why?
• Advantages
o Relatively inexpensive and fast (computational costs decrease as a function
of time)
o CFD provides “complete” information (all relevant variables in the whole
domain)
o Easily allows parametric studies (= important in design)
o No similarity constraints (simulations can be performed at full scale)
o Allows numerical experiments (e.g. study of explosions, failures, … which
you do not want to reproduce in reality)
• Disadvantages
o Accuracy and reliability are major concerns
o Results are very sensitive to large number of parameters to be set by the user
o Verification and validation are imperative (and validation requires
experiments)
6. PAGE 6
• Modeling approach: RANS vs. LES
o Reynolds-averaged Navier-Stokes: time-averaged
solution.
• Mean flow is ‘solved’, all eddies are ‘modeled’.
o Large Eddy Simulation: time-dependent solution.
• Large eddies are ‘solved’, smaller eddies are ‘modeled’.
• More accurate, but computationally far more expensive.
CFD: what and why?
7. PAGE 7
• Modeling approach: RANS vs. LES
o Reynolds-averaged Navier-Stokes: time-averaged
solution.
• Mean flow is ‘solved’, all eddies are ‘modeled’.
o Large Eddy Simulation: time-dependent solution.
• Large eddies are ‘solved’, smaller eddies are ‘modeled’.
• More accurate, but computationally far more expensive.
CFD: what and why?
8. PAGE 8
2. Natural ventilation: Indoor airflow
Dr.ir. Twan van Hooff Leuven University / TU/e
Prof.dr.ir. Bert Blocken TU/e – Leuven University
Prof.dr.ir. GertJan van Heijst TU/e
Prof.dr.ir. Tine Baelmans Leuven University
Prof.dr.ir. Johan Meyers Leuven University
Prof.dr.ir. Jan Carmeliet ETH Zurich / EMPA
Dr.ir. Thijs Defraeye ETH Zurich / EMPA
9. PAGE 9
• Mixing ventilation
o Previous studies conducted mostly for fully turbulent flows
(high Reslot values) (e.g. Nielsen 1974, Chen 1995).
o Lack of studies on transitional flows (low Reslot).
Indoor airflow
Reslot = U0h/ν
10. PAGE 10
• Transitional flow
Indoor airflow
Flow
Flow
Re = UL/ν
U = characteristic velocity [m/s]
L = characteristic length scale [m]
ν = kinematic viscosity [m2/s]
Fully turbulent
Modified after Gogineni and Shih 1997.
11. PAGE 11
• Transitional ventilation flow
o General
• Supply jet region (low velocity/turbulence intensity)
• Near buoyant plumes
• Corners of enclosure
o Operating theatres
o Airplane cabins
Indoor airflow
Aircraft cabin ventilation
(http://www.travelfreak.com/2012/09/18/airplane‐health‐tips‐
the‐doctors/)
Operating theatre ventilation
(http://www.ikv.uk.com/Medical‐Air‐Technologies.html)
Office ventilation
(http://images.businessweek.com/ss/08/02/0213_dag
her_engineering/source/9.htm)
12. PAGE 12
• Reduced-scale experiments
Indoor airflow
van Hooff T, Blocken B, Defraeye T, Carmeliet J, van Heijst GJF, 2012. PIV measurements and analysis
of transitional flow in a reduced‐scale model: ventilation by a free plane jet with Coanda effect.
Building and Environment 56: 301‐313.
13. PAGE 13
• Reduced-scale experiments
Indoor airflow
van Hooff T, Blocken B, Defraeye T, Carmeliet J, van Heijst GJF, 2012. PIV measurements and analysis
of transitional flow in a reduced‐scale model: ventilation by a free plane jet with Coanda effect.
Building and Environment 56: 301‐313.
14. PAGE 14
• Reduced-scale experiments
o Flow visualizations
Indoor airflow
van Hooff T, Blocken B, Defraeye T, Carmeliet J, van Heijst GJF, 2012. PIV measurements of a plane wall
jet in a confined space at transitional slot Reynolds numbers. Experiments in Fluids 53(2): 499‐517.
15. PAGE 15
Outlet
Indoor airflow
van Hooff T, Blocken B, Defraeye T, Carmeliet J, van Heijst GJF, 2012. PIV measurements of a plane wall jet
in a confined space at transitional slot Reynolds numbers. Experiments in Fluids 53(2): 499‐517.
16. PAGE 16
• Reduced-scale experiments
o Flow visualizations in vertical center plane
Indoor airflow
van Hooff T, Blocken B, Defraeye T, Carmeliet J, van Heijst GJF, 2012. PIV measurements of a plane wall jet in a confined
space at transitional slot Reynolds numbers. Experiments in Fluids 53(2): 499‐517.
17. PAGE 17
• Reduced-scale experiments
o PIV measurements in vertical center plane
Indoor airflow
van Hooff T, Blocken B, Defraeye T, Carmeliet J, van Heijst GJF, 2012. PIV measurements of a
plane wall jet in a confined space at transitional slot Reynolds numbers. Experiments in Fluids
53(2): 499‐517.
18. PAGE 18
• Reduced-scale experiments
o PIV measurements in vertical center plane
• Setup
• 2D PIV system Nd:Yag (532 nm) double-cavity laser (2 x 200 mJ,
repetition rate < 10 Hz)
• CCD (Charge Coupled Device) camera (1376 x 1040 pixel, 10 fps)
positioned perpendicular to the water cube
• Seeding by hollow glass micro spheres (3M; type K1); ds = 30–115
μm
• Software package Davis 7.1 (LaVision)
• Time-averaged velocities and turbulence intensities calculated from
360 uncorrelated samples measured at 2 Hz
Indoor airflow
19. PAGE 19
• Reduced-scale experiments
o PIV measurements in vertical center plane
• Time-averaged velocity vector fields in vertical center plane
Indoor airflow
van Hooff T, Blocken B, Defraeye T, Carmeliet J, van Heijst GJF, 2012. PIV measurements of
a plane wall jet in a confined space at transitional slot Reynolds numbers. Experiments in
Fluids 53(2): 499‐517.
20. PAGE 20
• Reduced-scale experiments
o PIV measurements in vertical center plane
• Time-averaged velocity vector fields in vertical center plane
Indoor airflow
van Hooff T, Blocken B, Defraeye T, Carmeliet J, van Heijst GJF, 2012. PIV measurements of
a plane wall jet in a confined space at transitional slot Reynolds numbers. Experiments in
Fluids 53(2): 499‐517.
21. PAGE 21
• CFD simulations
o Steady 3D Reynolds-Averaged Navier-Stokes equations
• Four different turbulence models
• RNG k-ε model (Yakhot et al. 1992)
• Low-Re k-ε model (Chang et al. 1995)
• SST k-ω model (Menter 1994)
• Low-Re stress-omega Reynolds Stress Model (Wilcox 1998)
o Low-Reynolds number modeling
• Solving the flow all the way down to the wall, including the
thin laminar sublayer
• y* preferably around 1
Indoor airflow
22. PAGE 22
• CFD simulations
o Model and grid
Indoor airflow
van Hooff T, Blocken B, van Heijst GJF, 2013. On the suitability of steady RANS CFD for forced
mixing ventilation at transitional slot Reynolds numbers. Indoor Air 23(3): 236‐249.
1.25 million cells
23. PAGE 23
• CFD simulations
o Boundary conditions
• Velocity based on Reynolds number
• Turbulence intensity based on PIV measurements
• Zero static pressure at outlet
• Walls modeled as no-slip walls
Indoor airflow
van Hooff T, Blocken B, van Heijst GJF, 2013. On the suitability of steady RANS CFD for forced
mixing ventilation at transitional slot Reynolds numbers. Indoor Air 23(3): 236‐249.
24. PAGE 24
• CFD simulations
o Results: Re ≈ 1,000
Indoor airflow
van Hooff T, Blocken B, van Heijst GJF, 2013. On the suitability of steady RANS CFD for forced
mixing ventilation at transitional slot Reynolds numbers. Indoor Air 23(3): 236‐249.
25. PAGE 25
• CFD simulations
o Results: Re ≈ 1,000
Indoor airflow
van Hooff T, Blocken B, van Heijst GJF, 2013. On the suitability of steady RANS CFD for forced
mixing ventilation at transitional slot Reynolds numbers. Indoor Air 23(3): 236‐249.
26. PAGE 26
• CFD simulations
o Results: Re ≈ 1,000
Indoor airflow
van Hooff T, Blocken B, van Heijst GJF, 2013. On the suitability of steady RANS CFD for forced
mixing ventilation at transitional slot Reynolds numbers. Indoor Air 23(3): 236‐249.
27. PAGE 27
• CFD simulations – non-isothermal (preliminary)
o Results: Re ≈ 2,800
Indoor airflow
o L = 2 m
o Inlet height = h/L = 0.025
o Floor heating
o Other walls adiabatic
o Supply temperature -7°C
28. PAGE 28
• CFD simulations – non-isothermal (preliminary)
o Results: Re ≈ 2,800
Indoor airflow
o L = 2 m
o Inlet height = h/L = 0.025
o Floor heating
o Other walls adiabatic
o Supply temperature -7°C
29. PAGE 29
3. Natural ventilation: Ventilative cooling
Ir. Katarina Kosutova TU/e
Prof.dr.ir. Bert Blocken TU/e – Leuven University
Prof.dr.ir. Jan Hensen TU/e
Dr.ir. Twan van Hooff Leuven University
Kosutova K, van Hooff T, Blocken B, Hensen JLM, 2015. CFD analysis of ventilative cooling in a generic
isolated building equipped with ventilation louvers. Healthy Buildings Europe 2015, 18-20 May 2015,
Eindhoven, The Netherlands.
30. PAGE 30
• PhD research: Katarina Kosutova
o Multi-scale computational assessment of ventilative cooling as an
energy-efficient measure to avoid indoor overheating
Ventilative cooling
31. PAGE 31
Ventilative cooling refers to the use of natural or mechanical ventilation strategies
to cool indoor spaces [1].
.
Why ventilative cooling?
Sustainable and energy efficient solution to reduce the cooling demand of the
building.
Helps to prevent indoor overheating.
Helps to maintain healthy indoor environment.
Objective of this sub-research
To investigate the influence of ventilation louver slat angle on the air exchange
efficiency and heat removal effectiveness in a generic isolated building.
[1] www.venticool.eu, 01.05.2015
Ventilative cooling
32. PAGE 32
www.glasbau-hahn.com
Energy efficient
Possible to control ventilation due to operable slats
Can be opened during rain to allow fresh washed air
to enter the house
Seal tightly when closed
Mostly used in warm climates
Ventilative cooling
• Ventilation louvers
34. PAGE 34
[2] Karava P, Stathopoulos T, Athienitis AK. 2011. Airflow assessment in cross-ventilated
buildings with operable façade elements. Build. Environ. 46(1): 266-279.
[m]
Ventilative cooling
• Building model
• Generic isolated building was chosen in order to validate the CFD
simulations with the wind tunnel PIV measurements[2].
35. PAGE 35
Building geometry (without louvers)
Generic isolated building
H = 4 m (height of the building)
Generic isolated building
H = 4 m (height of the building)
Computational domain
Ventilative cooling
• Building model
36. PAGE 36
Building geometry (without louvers)
Generic isolated building
Configurations considered:
Window without any louvers
Windows equipped with ventilation louvers
with slat angles: 0°, 30° and 45°
Generic isolated building
Configurations considered:
Window without any louvers
Windows equipped with ventilation louvers
with slat angles: 0°, 30° and 45°
Building geometry (with louvers)
Ventilative cooling
• Building geometry and mesh
37. PAGE 37
Building geometry (without louvers)
Generic isolated building
Configurations considered:
Window without any louvers
Windows equipped with ventilation louvers
with slat angles: 0°, 30° and 45°
Generic isolated building
Configurations considered:
Window without any louvers
Windows equipped with ventilation louvers
with slat angles: 0°, 30° and 45°
Building geometry (with louvers)
Grid created in Gambit
Size of the grid based on the grid-sensitivity
analysis
Basic grid: 6,766,578 cells
Hexahedral cells
Grid created in Gambit
Size of the grid based on the grid-sensitivity
analysis
Basic grid: 6,766,578 cells
Hexahedral cells
Ventilative cooling
• Building geometry and mesh
38. PAGE 38
Logarithmic velocity profile imposed at the inlet
Inlet turbulent kinetic energy (k)
Inlet specific dissipation rate (ω)
Logarithmic velocity profile imposed at the inlet
Inlet turbulent kinetic energy (k)
Inlet specific dissipation rate (ω)
k a UIU
2
* 3
ABL
0
u
(y y )
U y
uABL
*
ln
y y0
y0
Ck
,
κ = 0.42
u* = 0.363 m/s
y0 = 0.00125 m
Cμ = 0.09
a = 1[3]
assuming σu = (σv + σw)
[3] Ramponi R, Blocken B. 2012. CFD simulation of cross-ventilation for a generic isolated
building: Impact of computational parameters. Build Environ 53: 34-48.
Ventilative cooling
• CFD: boundary conditions
39. PAGE 39
Inlet: velocity inlet, θi = 20°C
Outlet: pressure outlet
Side and top planes: symmetry
Ground : wall, ks = 0.0125 m
Cs = 0.979
Building surfaces: wall, θw = 30°C
Inlet: velocity inlet, θi = 20°C
Outlet: pressure outlet
Side and top planes: symmetry
Ground : wall, ks = 0.0125 m
Cs = 0.979
Building surfaces: wall, θw = 30°C
s
s
C
y
k 0
793
.
9
[4]
[4] Blocken B, Stathopoulos T, Carmeliet J. 2007. CFD simulation of the atmospheric
boundary layer: wall function problems. Atmos Environ 41(2): 238-252.
Ventilative cooling
• CFD: boundary conditions
40. PAGE 40
3D steady RANS equations
SST k-ω turbulence model [5]
SIMPLE for pressure-velocity coupling
PRESTO! for pressure interpolation
Second-order discretization schemes for momentum, k, ω and energy
3D steady RANS equations
SST k-ω turbulence model [5]
SIMPLE for pressure-velocity coupling
PRESTO! for pressure interpolation
Second-order discretization schemes for momentum, k, ω and energy
[5] Menter FR. 1994. Two-equation eddy viscosity turbulence models for engineering
applications, AIAA J, 32: 1598-1605.
Ventilative cooling
• CFD: parameters and settings
41. PAGE 41
Contours of dimensionless velocity (|V|/Uref) in the vertical
center plane, Uref = 6.97 m/s
Ventilative cooling
Geometry of the louvers
• CFD: Results
42. PAGE 42
Contours of dimensionless temperature (θ/θref) in the vertical
center plane, θref = 20°C
Ventilative cooling
• CFD: Results
Geometry of the louvers
43. PAGE 43
Air exchange efficiency
Geometry of the louvers
100
out
(2av)
Ventilative cooling
45. PAGE 45
CFD simulations were performed for a relatively high reference wind speed
(Uref = 6.97 m/s).
Future research
CFD simulations for lower wind speed
CFD simulations for different values of y0
More realistic building geometry
Influence of the urban surrounding on the air
exchange efficiency and heat removal effectiveness
CFD simulations for lower wind speed
CFD simulations for different values of y0
More realistic building geometry
Influence of the urban surrounding on the air
exchange efficiency and heat removal effectiveness
Ventilative cooling
• Discussion
46. PAGE 46
Contours of dimensionless velocity (|V|/Uref) in the vertical
center plane, Uref = 1.15 m/s
Ventilative cooling
• CFD: Results
Geometry of the louvers
47. PAGE 47
Contours of dimensionless temperature (θ/θref) in the vertical
center plane, θref = 20°C
Ventilative cooling
Geometry of the louvers
• CFD: Results
48. PAGE 48
4. Natural ventilation: Stadium study
Dr.ir. Twan van Hooff Leuven University – TU/e
Prof.dr.ir. Bert Blocken TU/e – Leuven University
49. PAGE 49
• Stadium description
o Amsterdam ‘ArenA’
• Completed in 1995 in Amsterdam
• Multifunctional stadium
• Capacity of 51,628 spectators
• Retractable (semi-)transparent roof
• No HVAC Services
• Natural ventilation through the roof and
openings in the building facade
Stadium study
51. PAGE 51
• Stadium description
o Natural ventilation through the roof, corners of stadium and
relatively small openings in the building facade
Ventilation opening Surface area (m2)
Roof 4,400
Four openings in corners of stadium 166
Opening between stand and roof construction 130
Opening between fixed roof and movable roof 85
Stadium study
52. PAGE 52
• Full-scale measurements (1)
o During summer:
• Temperature
• Relative humidity
• Air speed
• Globe temperature
• CO2 concentration
(All measured on 4 positions inside the stadium. T en RH also measured outside the stadium)
• Irradiance of the sky
Stadium study
53. PAGE 53
• Full-scale measurements (1)
o Measurement positions
Measuring positions for the air temperature, relative humidity, CO2 concentration and air speed () inside the stadium
with (a) positions in a horizontal plane; (b) positions in a vertical plane.
Stadium study
54. PAGE 54
• Full-scale measurements (1)
van Hooff T, Blocken B, 2012. Full‐scale measurements of indoor environmental conditions and natural
ventilation in a large semi‐enclosed stadium: possibilities and limitations for CFD validation. Journal of
Wind Engineering and Industrial Aerodynamics 104‐106: 330‐341.
Stadium study
55. PAGE 55
• Full-scale measurements (2)
o Wind speed measurements inside and around the stadium
Measurement with ultrasonic anemometer outside the ArenA
Stadium study
Measurement with ultrasonic anemometer in the corners of the ArenA
56. PAGE 56
• CFD simulations
o 3D steady state CFD simulations
• Domain: 2,900 x 2,900 x 908.5 m3 (LxWxH)
• Hybrid grid (5.5 million cells)
• Realizable k-ε turbulence model (Shih et al. 1995)
• Standard wall functions (Launder and Spalding 1974) with sand-grain
based roughness modification (Cebeci and Bradshaw 1977).
• Logarithmic wind speed profile (U10 = 5 m/s, y0 = 0.5 m or 1.0 m)
• Several wind directions
• Estimated surface temperatures imposed to take into account solar
irradiation
Stadium study
58. PAGE 58
• CFD simulations
o Grid
Stadium study
5.5 million cells
59. PAGE 59
• CFD simulations
o Grid-sensitivity analysis
Stadium study
60. PAGE 60
• CFD simulations
o Validation using wind speed measurements
Stadium study
61. PAGE 61
Stadium study
Van Hooff T, Blocken B, 2010. On the effect of wind direction and urban surroundings on natural
ventilation of a large semi‐enclosed stadium. Computers & Fluids 39, 1146‐1155.
• CFD simulations
o Results
62. PAGE 62
• CFD simulations
o Results
Van Hooff T, Blocken B, 2010. On the effect of wind direction and urban surroundings on natural
ventilation of a large semi‐enclosed stadium. Computers & Fluids 39, 1146‐1155.
Stadium study
63. PAGE 63
• CFD simulations
o Results
Stadium study
Van Hooff T, Blocken B, 2010. On the effect of wind direction and urban surroundings on natural
ventilation of a large semi‐enclosed stadium. Computers & Fluids 39, 1146‐1155.
64. PAGE 64
Contours of wind speed ratio
U/U10 in four horizontal planes,
for φ = 196° (SSW) and U10 = 5
m/s; at (a) 10 m; (b) 20 m; (c)
40 m and (d) 60 m above the
ArenA deck.
Stadium study
66. PAGE 66
• CFD simulations can provide high-resolution data on
natural ventilation flows.
• Both basic and applied research is needed.
o Understanding the flow
o Optimizing ventilation flows for practical situations
• Experimental studies are imperative.
o To obtain valuable insights in physical processes
o For validation purposes
Closing remarks
67. CFD-simulaties van natuurlijke
ventilatie
Bedankt voor uw aandacht!
Dr.ir. Twan van Hooff
FWO postdoctoral research fellow
KU Leuven
twan.vanhooff@bwk.kuleuven.be
Prof.dr.ir. Bert Blocken
TU Eindhoven / KU Leuven