GEOMATICS SYNTHESIS PROJECT SYMPOSIUM · Introduction 2. Background 3. Site Description 4. LASER/F 5. Field Campaign 6. Simulation Results 7. Computational Fluid Dynamics 8. Conclusions

GEOMATICS SYNTHESIS PROJECTSYMPOSIUM

Thursday 8 November, 2012 Oostserre, BKCity

in association with

OVERVIEW

1. Introduction2. Background3. Site Description4. LASER/F5. Field Campaign6. Simulation Results 7. Computational Fluid Dynamics8. Conclusions and Recommendations


The effect of 3D geometry complexity on simulating radiative, conductive, and convective fluxes in an urban canyon.

URBANHEATINTRODUCTION

1

INTRODUCTION THE GSP TEAM + OBJECTIVES

2

INTRODUCTION PARTNERS

Teledetection, Radiometrie et Imagerie Optique (TRIO)(Laboratoire des Sciences de l’Image, de l’Informatique et de la Télédétection (LSIIT), University of Strasbourg)

Transport Phenomena Group (Chemical Engineering, TU Delft)

+

3

INTRODUCTION RESEARCH QUESTION AND HYPOTHESIS

We wondered...What level of effort with 3D modeling is required to accurately simulate heat flux?

4


Research question:How does 3D geometry complexity affect the accuracy of simulating heat flux in an urban canyon?


4


Research question:How does 3D geometry complexity affect the accuracy of simulating heat flux in an urban canyon?


Hypothesis:Increasing the level-of-detail increases accuracy of the simulation.

4

INTRODUCTION RESEARCH APPROACH

Phase I + IIData acquisition, pre-processing, and sensitivity tests.

5


Complex 3D geometry

3D geometry scenarios Forcing file

LASER/F

Meteorological data Thermal imagery Phase I + IIData acquisition, pre-processing, and sensitivity tests.

5


Complex 3D geometry


LASER/F

Meteorological data Thermal imagery

Output Project output onto image


Phase IIIComparison of the results

Compare results

5


Complex 3D geometry


LASER/F

Meteorological data Thermal imagery

Output Project output onto image

Compare results

Wind profileConclusions and

recommendationsCFD MODEL


Phase IIIComparison of the results

Phase IVAnalysis and conclusions

5

BACKGROUND URBAN CANYON SCALE STUDIES

6

Geometry of an urban canyoni. orientationii. aspect ratio (H/W)iii. sky view factori ii iii


7

Character of an urban canyoni. geometry solar illuminationii. surface thermal properties fluxiii. streetscape elements mech. turbulence

Geometry of an urban canyoni. orientationii. aspect ratio (H/W)iii. sky view factor

aspect ratio and wind flow


i ii iii

Character of an urban canyoni. geometry solar illuminationii. surface thermal properties fluxiii. streetscape elements mech. turbulence

7

SITE DESCRIPTION RUE DE L’ARGONNE

8

170m 20m wide 15-20m tall AR = 0.75 - 1

SITE DESCRIPTION BUILDINGS OF INTEREST

1 Rue de l’Argonne 4 Rue de l’Argonne

9

LASER/F MODEL

triangulated mesh

heat flux calculated for each triangle mesh

surface temperaturederived from energy balance equation

10

LASER/F 3D GEOMETRY

Inhomogeneous complexity in the model

Incomplete building envelopes

How should it be simplified?

11

LASER/F BASELINE DETERMINATION

12

1

LASER/F SEVEN SIMULATION SCENARIOS

12


1

2

33

12


1

2

333 44

12


1

2

333 44

55

12


1

2

333 44

55

6612


77

1

2

333 44

55

6612

FIELD CAMPAIGN STANDARD METEOROLOGIC STATION

Measurements for the forcing file:1. air temperature2. air pressure3. relative humidity4. wind speed5. wind direction6. incoming direct solar radiation7. incoming diffuse solar radiation8. outgoing longwave radiation

13

FIELD CAMPAIGN THERMAL IMAGING

24hr measurements1 Rue de l’Argonne9 images x 20 epochs

radiant exitance (reflected and emitted energy)

1

2

3

MP5 MP6 MP7

14

1

2

3

4

MP1 MP2 MP3 MP4

FIELD CAMPAIGN THERMAL IMAGING

24hr measurements4 Rue de l’Argonne16 images x 20 epochs

15

SIMULATION RESULTS COMPARISON METHOD

16


Measured (FLIR camera)Simulated (LASER/F)

17


Measured (FLIR camera)1. Filter out sky and windows

Simulated (LASER/F)1. Filter out sky and windows

17


Measured (FLIR camera)1. Filter out sky and windows2. Generalize measurements


17



Measured (FLIR camera)1. Filter out sky and windows2. Generalize measurements

17



Measured (FLIR camera)1. Filter out sky and windows2. Generalize measurements3. Compare results

17

RESULTS COOLER SIMULATED VALUES

1 Rue de l’Argonne

mean exitance

18

RESULTS COOLER SIMULATED VALUES


mean exitancemean exitancemean exitance

19

RESULTS RADIANT EXITANCE BEHAVIOUR

RMSE

20

behaviour/shape warming period effect of balconies


RESULTS RADIANT EXITANCE BEHAVIOUR


RMSERMSERMSE

21

behaviour/shape warming period

COMPUTATION FLUID DYNAMICS WIND PROFILE

static simulation results22

Recall A systematic tendency of simulating cooler

temperatures.

Why? More convective cooling above podium

Generally 0.5 ms-1 higher wind speeds in LASER/F compared to the CFD model results.

COMPUTATION FLUID DYNAMICS WIND PROFILE

LASER/F

CFD

0 0.5- 0.5 1.0 1.5 2.0 2.5

wind speed (ms-1)

heig

ht (m

)

510

1520

25 rooftop

1st Floor

23

Increased 3D Complexity: Does increase simulation time.

CONCLUSIONS AND RECOMMENDATIONS

24



24



24


Does not necessarily increase simulation accuracy.

Study shows: Different radiant exitance behaviour over time

between the two facades - accuracy is affected by daily solar illumination pattern.


25


Does not necessarily increase simulation accuracy.

Study shows: Different radiant exitance behaviour over time

between the two facades - accuracy is affected by daily solar illumination pattern.

Systematically lower temperatures in LASER/F could be explained by over-simplified wind profile.


LASER/F

CFD

0 0.5- 0.5 1.0 1.5 2.0 2.5

wind speed (ms-1)

heig

ht (m

)

510

1520

25

26

RESEARCH CONTRIBUTION

Gives an order of magnitude on the effort needed to model urban heat at the canyon scale.


27



RECOMMENDATIONS

Improve vertical wind profile in LASER/F


27



RECOMMENDATIONS


Better modeling of building volumes and materials


27



RECOMMENDATIONS



Use a model with consistent complexity


27



RECOMMENDATIONS



Use a model with consistent complexity

Thermal image measurements (define ground control points and more overlap be-tween images)


27




FUTURE WORK?

Compare two similar facades

Compare two or more different canyons

Run simulation of urban neighbourhood or district

28


Thursday 8 November, 2012 Oostserre, BKCity

in association with

Documents

GEOMATICS SYNTHESIS PROJECT SYMPOSIUM · Introduction 2. Background 3. Site Description 4. LASER/F 5. Field Campaign 6. Simulation Results 7. Computational Fluid Dynamics 8. Conclusions