Download - Combined CFD-Mean energy balance methode to thermal comfort assessment of buildings in a warm tropical climate

8/13/2019 Combined CFD-Mean energy balance methode to thermal comfort assessment of buildings in a warm tropical climate

1/8

1COMBINED CFD-MEAN ENERGY BALANCE METHOD TO THERMAL

COMFORT ASSESSMENT OF BUILDINGS IN A WARM TROPICAL CLIMATE Stphane Sanquer 1, Mohamed Abdesselam 2, Fabien Picgirard 3

1Meteodyn, Nouma, New Caledonia, France2Solener, Lille, France3Ademe, La Runion, France

ABSTRACTThe paper presents combined approaches dedicatedto assess the thermal comfort in a tropical climate.UrbaWind, an automatic computational fluiddynamics code, was developed to model the wind in

urban environments. A module was recently added toassess the buildingss natural cross ventilation. Thesimple thermal method Batipei, allows computing themean overheating inside the building by consideringthe envelope thermal behaviours, the outdoor thermalconditions and the mass flow rate.

Finally, an architectural case of urban renovation ofold buildings in La Reunion Island is presented as anexample of the using of the combined approaches.

INTRODUCTION

Background

In the context of reducing the energy consumption inbuildings for the overseas territories located in warmtropical climate, The French energy agency(ADEME) needed rules and tools for existingbuildings adapting them to such climates, differentfrom the European temperate climate. In the Frenchterritory La Reunion, 17% of the housing is airconditioned. This rate has doubled over the last fiveyears. In this context, the local agency proposes analternative way to improve the comfort withbioclimatic approaches. Architects, designers andproperty developers especially in low-rent buildingsshould know immediately if the buildings could be

made thermally comfortable without cooling.Thermal assessments of buildings are intricate anddetailed in urban areas located on or near complexterrain. They need to carry out their surveys withsimple, easy-to-use methods. Hence, ADEME (LaRunion agency) developed a method BATIPEIdevoted to evaluating the indoor thermal comfort inwarm tropical climates by taking into account theexternal climatic parameters, the internal heat gainsand the air change rate of the room. This method willdetermine buildings or apartments that need airconditioning and those that could be naturallyventilated according to the local climatology.

The natural ventilation of a building is driven by thecombined forces of wind and thermal buoyancy.

However, when openings are large enough and theair change rate is high enough, the indoortemperature is quite homogeneous, and except for tallvolumes, the natural ventilation is mainly driven bythe wind forces. So for the natural ventilationassessment in tropical warm climates, the thermal

forces are generally assumed as negligible comparedto the wind driven forces. The assessment of thecross-ventilation efficiency is completed byconsidering the wind speed levels inside thebuildings, and the flow-rates through the openings orgenerally the air change rate of the volume (ACH).The cross wind flow rate depends on the size of theopening, on their aerodynamic efficiency namelytheir aerodynamic discharge coefficients, on thepressure coefficients on the building envelope and onthe reference wind velocity upstream the building.Aynsley et al. (1977) gave an analytic expressionuseful for a single ventilated volume with twoopenings, but as soon as the number of openings isgreater or the internal volume is divided in subvolumes, a resolution for non-linear system has to beused.All of these characteristics are fundamental anddepend on the external wind pressure at the openings.The pressure on the envelope of the building can bedefined with tables (Liddament 1986, Clarke. J et al.1990 , Eurocode) or by using parametric modelsbased on the analysis of results from wind tunneltests (Grosso 1992). These tables or correlations arepotentially inapplicable when buildings are irregular

and different from the simple shape or when they areflanked by others buildings in a complex urban areawith canyons and corner. In these configurations, thewind fluctuates strongly and the pressure coefficientsand the wind-driven natural airflows could beevaluated with some difficulty. These external windpressures should be evaluated either from windtunnel tests (Kandola, 1990) or from computationalfluid dynamics (CFD) simulations (Ayad 1999,Fahssis et al. 2010). Analytic method such as theAynsley method could not be used for complex flatlayout where pressure loss occurs. Networkaerodynamic tools that include the pressure loss

depending on the internal porosity are necessary toassess the air change rate in such layout. UrbaWind,a CFD software dedicated to modelling the urban


2/8

climatology, was upgraded to deliver the mass flowrate to assess the cross ventilation efficiency(Sanquer et al., 2011).The thermal comfort inside naturally ventilatedbuildings depends on the indoor temperature, the airhumidity and the velocity in the occupied zone.

Givoni defined comfort curves (Krger et al., 2010)as areas in the wet air diagram that depend on theindoor velocity. This diagram allows designers toknow if an indoor situation defined with velocity,temperature and humidity is comfortable.. The figure1 shows that when indoor velocity is close to unity,the comfort area (continuous black line) is larger thanin a lower velocity (no indoor motion in dotted line,velocity=0.5 m/s in dashed line). Points in thediagram were determined by dynamic simulations(Garde R., 2006). Some are in the comfort domainbounded by polygonal curves (comfortablesituations) whereas the others remain outside the area(too warm and wet or too cold) indicatinguncomfortable indoor climates.

Figure 1 : Thermal dynamic simulation in an officeat La Runion Island during the summer (Garde R.,

2006)A thermally comfortable building which is welldesigned for a tropical climate is described as having:

An indoor mean temperature close to theoutdoor temperature. That means thedesigners properly managed the heatbalance : solar radiation flux, internal heatgains, heat extraction with the ventilation

A maximum temperature smaller than theacceptable comfort temperature according tothe humidity and the indoor air velocity,depending on air motion dynamics.

The indoor temperature depends on the air changerate and the thermal characteristics of the buildingsenvelope (conductance, specific heat). Dynamic

Thermal Modelling is developed with absoluteprecision to simulate the instantaneous physicalbehaviours of a building, while delivering much data.

Knowledge of building heat transfer and airflow isrequired to use the software and to assess the thermalefficiency of the buildings. Developments andimprovements of the thermal behaviours of thebuildings are not simple. Variations of any inputparameter give whole variations of the thermal

equilibrium of the building. Models complexity andmany interactions avoid the explicit knowledge of thethermal mechanisms. The contribution of variousparts of the building into the energy balance shouldbe clearly highlighted to allow users to findimprovements in the comfort level. Unfortunately,Dynamic Thermal Modelling can be expensive andtime consuming, and cannot be used for smallerprojects or for the initial step of creating a proposalwhere budgets are limited. Such a simple approachcannot be based on dynamic responses of thebuilding but rather on average energy balanceequation.

Therefore, new simple model, firstly pedagogical,based on an average method, was developed forwarm tropical climates by Abdesselam (1997). Theeffects of each heat gain and loss were evaluated onthe temperature of the room. Indoor overheating incomparison with the outdoor temperature iscomputed. The comfort criteria is targeted in a placewhere the temperature criteria keeps the indoorclimate comfortable.

PurposesThe first objective of this report is to present amethod of assessment for the indoor thermal comfort

in a warm tropical climate. The method is composedof two separate tools: an aerodynamic model thatestimates the air exchange through the buildingwalls, roof and openings, and a thermal model thatcomputes the indoor temperature.The aerodynamic tool is a CFD-Network softwarenamed UrbaWind which performs calculation of theoutdoor wind combined with a macroscopic networkmethod for the evaluation of the mass flow ratesthrough the openings. Results depend on the externalwind conditions, taking into account the local windclimatology.

The thermal tool named BATIPEI is a simple thermalmethod, which computes the mean overheating insidethe building. The inputs of the tool are the buildingsthermal characteristics, the outdoor thermalconditions, including temperature and solarradiations and the mass flow rate delivered byaerodynamic tools.The second part of this article addresses the urbanrenovation of old buildings overheated during thesummer at the city of Saint-Denis (La reunion Island)carried out with this method. The main objective forthe apartments owners is to define new thermalbehaviours of buildings in order to improve indoor

thermal comfort and to minimize air conditioning.


3/8

NUMERICAL APPROACHES

Aerodynamic ToolThe CFD method of UrbaWind consists of solvingthe Reynolds-Averaged Navier-Stokes equations onan unstructured rectangular grid with automatic

refinement of the mesh near obstacles. The CFD tooldelivers the tri-dimensional mean velocity field andthe mean pressure for each point in the domain.When the airflow is steady and the fluidincompressible, the mean equations for the meanvelocity components contains unknown quantities,the turbulent fluxes that can be solved by a one-equation model. The transport equation for theturbulent kinetic energy k contains a dissipation term deduced from the mixing-length theory.

+

=

j

j

i

j

j

iT

ik

T i

i x

u

x

u

xu

xk

k u x

(1)

The turbulence viscosity t is considered as theproduct of a length scale with a speed scale, whichare both characteristic lengths of the turbulentfluctuations.

T t Lk 2 / 1 = (2)

09.0andand2 / 3

== C L

k C

T

(3)

The turbulent length scale L T varies linearly with thedistance at the nearest wall (ground and buildings).This methodology, along with the use of a veryefficient multi-grid coupled solver allows for a fastconvergence for every kind of geometry.Boundary conditions are automatically generated.The mean velocity profile at the computation domaininlet is determined by the logarithmic law in thesurface layer, and by the Ekman function. ABlasius-type ground law is implemented to modelfrictions (velocity components and turbulent kineticenergy) at the surfaces (ground and buildings).Typically CFD computation domains cover an areaof maximum 1000 m x 1000 m, taking into accountthe effects of neighbouring buildings. The meshrefinement at an obstacle can reach 5 cm. If needed,

it is possible to also compute the flow inside thebuilding, but this can be time consuming andexpensive.For the simple thermal assessment in suchapplications, computing the indoor domains is notnecessary. For rooms with more than two openings, anetwork approach delivers the mass flow ratecrossing each opening (Fhassi et al. , 2010). Theinternal pressure is not known and the flow ratethrough openings are solved by an iterative process.Firstly, the indoor pressure P i is initialized as theaverage of the outer surface pressures P k at theopening k, weighted by the square of the freeaerodynamic area A k . The values of the P k areobtained by the CFD calculation with UrbaWind.

Then, after several steps, the computation isachieved. At the iteration n, the flow rates at eachopening k are computed by:

/ 2 nik k nk PP AQ = (4)

The outflow rates are updated by solving an equation(5) with a Newton-Raphson iterative method.

0=k

nk Q (5)

The internal pressure at step n+1 is updated and theflow rates are adjusted with the new internalpressure. The iteration stops when convergence isachieved. In the case of a flat with several rooms, thefree aerodynamic area of the opening k is replaced byan equivalent aerodynamic surface, taking intoaccount the door aerodynamic surface Adoo r of thesecondary room (Sanquer et al. , 2011).

22

11 / 1 door k

eqk A A A += (6)

Thermal toolThe maximum temperature T max and the meanoverheating depend on thermal inertia of thebuildings. To assess simply the comfort, one solutionconsists of replacing the requirement on T max with arequirement on T, the mean daily overheating.Abdesselam (1997) developed a new method namedBATIPEI for the French tropical island La Runion.As soon as the air change rate is high enough (morethan 10 vol/h) and the temperature daily range is

moderate (6 C) as in such a warm tropical climate,the contribution of the thermal inertia in theevaluation of the indoor temperature and of thecomfort level is relatively low. During thedevelopment of the method, it was shown thatwhatever the thermal inertia, comfort was achievedfor such a oceanic tropical climate with a meanoverheating T that didnt exceed 2C. The influenceof inertia is lower than half a degree (figure 2) andbecomes a second order parameter compared to theventilation rate in the energy balance equation.

Figure 2 : Indoor temperature for various inertiawith an air change rate equal to 15 vol/h

This method was developed as a preliminary andcomplementary thermal approach for designers,especially for renovated buildings that are overheated

during the summer. The mean overheating is relevantthermal criteria and suitable to design the thermalcharacteristics of buildings because:


4/8

It is a representation of tand also of the excess heatdesign could eliminate orparameters (convective traand shade coefficients, air c

It can be computed wapproach allowing separatthe overheating. Such auseful to assess the nbuildings with a simple to worksheets.

The main aspects of the method cothe gains and losses of enthalpyequation in a periodic and stationenergy balance written on a daily cy

The sun radiation gains recindoor air from the wal( E sun ), transmitted directopenings ( E o), and the inte

The losses given to theconduction through the envthe air change rate ( E ach ).

The energy loaded by the buildincycle is zero. The dynamic paramthe balance equation. Hence,

In stationary periodic regime, theventilation terms can be written durequivalent warm days as follows:

With j the wall index, T the periowall conductance, A j the wall j arindoor and outdoor temperatures, C of air, ACH the Air Change Rate, V The analysis is based with mean vmean overheating of the room conductance of the room .The energy loss can be determinedoverheating as follows:

A The energy gains can be written aftfollows:

1

E wj is the sun radiation energy absowith an area of A j. E soi is the suncrossing the opening i up to the wa

the convective heat transfer cooutdoor and indoor walls.

e mean comfortthat a successful

avoid with staticnsfer, absorptionhange rate).

ith an analyticing the origin ofmethod becomesatural ventilatedl made of Excel

nsist in clarifyingwith a balance

ary regime. Thiscle separates:

ived by the rooml by conductionly through thenal gains ( E int )environment bylope ( E c) or with

s during a dailyters disappear in

(7)

conductive anding a sequence of

89

duration, U j thea, T, and T o thethe specific heat

the room volume.alues namely the

and the mean

y the mean daily

(10)er calculations as

(11)

(12)

rbed by the wall j radiation energy

ll j. hoj and h ij are

fficients of the

The mean overheating of thethe sun radiation contributionintroducing a term dependingand another that corresponds twithout ventilation.

, Where D is specific mass flow 0 This is the temperature reacwithout ventilation, namely thof the energy gains are extracthrough the walls and the roonly on the characteristics offaces and characterize:

The building perforsun protection. In fa

E soi /T are the mean datransmitted through tdepending on the solfaces (colors, sun scdaily solar radiations,

The repartition of theto solar exposure of t

In theory, the optimized layoulowest conductance namely tare the most exposed faces to

Rc parameter defined the ratiwith the conduction as:

, The curves of RC versus theshow on figure 2:

The more insulatedvalue of U m), thechange rate is.

For medium size(2< U m


5/8

Figure 5 : Block geometry model in UrbaWind

EXAMPLE OF NATURALVENTILATION AND THERMALASSESSMENT OF BUILDINGS IN ANURBAN AREAAt the Runion Island, trade winds are moderate onthe north side of the island. The climatologycharacteristics estimated at 40 m above the ground atthe site is shown on figure 3. The average velocity isclose to 3.5 m/s at 40 m. So according the log-lawprofile, the mean velocity is expected to be close to 2m/s at 10 metres above the ground and the buildingsmay be naturally ventilated at this point.

Figure 3 : Climatology of the site (h=40 m)Due to the neighbouring buildings, the wind shouldbe computed, and the mass flow rates entering theflat, should be assessed.

Figure 4 : Chateau Morange block at Saint-Denis(La Runion Island)

The layout of the Chateau Morange block, which islocated close to the north coast of La Runion islandat Saint-Denis, is shown on Figure 4. The projectconsists of the construction of new buildings (red

ones), of and the demolition (blue ones) orrenovation of (white ones) the existing buildings.Figure 5 shows the layout with the CFD softwareUrbawind where the neighbouring buildings andtrees were also defined.The computational domain covered 500 m by 500 marea and 100 m in height The horizontal and verticalrefinement for the cells close to the walls, the groundand all the result points are defined at 1 meter givinga total number of 1 million cells. The wind flowsimulations were conducted for each wind incidencefrom 0 to 330 with a step of 45. The time ofcomputation was 1 hour for each run.Figure 6 shows an example of the mean speed upfactor field for trade winds and their direction in thisarea (Alizs trade winds flow in the sector 120).The speed up factor is defined as the ratio of the

mean speed at the result points to the reference meanvelocity measured close to the computational domainentrance at 100 meters above the ground.

Figure 6 : Speed up factor field at 2 above theground (wind direction 120)

Figure 7 shows an example of a mean pressurecoefficient on buildings of interest. The pressurecoefficient is defined as the ratio of the meanpressure on the building walls to the referencedynamic pressure of the incoming wind measured


6/8

close to the computational domain entrance at 10meters above the ground.

Figure 7: mean pressure coefficient on the buildingswalls (wind direction 120)

For each wind direction giving a pressure field on theoutside walls, the flow rate was evaluated followingthe macroscopic method described previously for anapartment located at the top level of the buildingshown with a circle on figure 7. This apartment waschosen because of its exposition to the sun radiation,(roof and east sidewall) making it the worst case withthe most expensive cost to renovate.The flat is a medium sized flat with a main room andtwo bedrooms as shown on figure 8. The surface andvolume of the space are respectivly 65 m and 160m3.

Figure 8 : mean pressure coefficient on the buildingswalls (wind direction 120)

Before renovation, three windows (1.2x1.2m) arefitted on the north side and one window on the southside. To increase the cross ventilation rate, tworenovations were made: enlarging the doors in orderto increase the indoor permeability, and addinganother window on the south side (a louver on thebalcony wall with an area of 1m or 2 m). Thebedroom doors were expected to be opened but aconfiguration with closed doors and a jalousie

window (0.5 m) above the door was tested.

The ventilation was estimated according to theclimate around the block by taking into account theoccurrence of both the velocity and incidence (Figure9).

Figure 9: Air change rate distribution (yearly data)The yearly average Air Change Rate andother criteria are shown in table 1 for sixconfigurations before (the first line) and afterrenovations (the five following lines).P(ACH


7/8

improved up to 40% of the average value and 65% for the lower value of ACH. By enlargingthe second window to 2 m this allowed double thethe amount of cross ventilation in the apartment.Increasing the internal porosity is an efficient way toimprove the cross ventilation. The ratio of north wall

to south wall porosity is too large before therenovation. The flow of air has to go through thesouth bedroom to cross ventilate the flat. Opening upthe south wall of the main room should berecommended. More over, to open the wall above thebedroom doors with a small louver maintains aircirculation even when the door is closed.Following the aerodynamic assessment, designersshould check to ensure the air change rate is highenough to allow the extraction of heat. Thetemperature variation at Saint-Denis is shown onfigure 10. January and February are the most difficultperiods to achieve the indoor thermal comfort. Theaverage daily temperature can rise up to 30C.According to the Givonis Diagram (figure 1),overheating of the indoor air should not exceed 2Cin this area in order to not exceed 32C inside thebuildings.

Figure 10 : mean of the extreme temperature atSaint-Denis (from Meteo France)

The apartment is located on the top of the buildingand is exposed to sun radiation. The heat flux withthe apartment on the floor below was neglected. Thewalls and ceiling are made of concrete with aninternal layer of wall plaster. The northeast windowsare opened and sun screens reduce the sun radiationby about 30%. The external walls are clear and thesolar absorption coefficient is 0.4.The computation of the mean thermal balance wascarried out with the method BATIPEI previouslydescribed. The fluxes associated with the highestsolar radiation at Saint-Denis (Austral summer) areused as input data in the simulation as shown infigure 11.The computation was carried out by writing thesurface of the walls, roofs and openings area, andtheir thermal coefficients (convective transfer,absorption and shade coefficients) in tables for each

apartment. The mass flow rate defined as an inputwas calculated with UrbaWind. Nevertheless, furthercomputations were carried out by varying the ACH

in order to evaluate sensibility of the system to the airexchange rate.

Figure 11 : Solar fluxes on the roof and the flat walls(Austral summer)

Figure 12 : Overheating temperature versus the airchange rate for reference layout and improved

layouts with insulation.Without additional insulation, the convective transfercoefficient U wall=U ceiling =2.8 W/m/K. As shown onfigures 12, the mean overheating is in the range [4-6]C for ACH in the range [10-20] volume/hourexceeding the 2C limit. The main part of the heattransfer is associated to the roof (70%) as shown onfigure 13.

Figure 13 : Overheating temperature for everybuilding part without ventilation.

Hence, the external treatment of the roof with a

mineral wool (thickness= 0.1m, thermal conductivity =0.04 W/m/K) was necessary. The overheatingbecame close to 2C with ACH between 10 to 20

0

5

10

15

20

25

30

35

T e m p e r a

t u r e

( C )

Tmax

Tmin

0

100

200

300

400

500

600

700

800

900

1000

0 3 6 9 12 15 18 21 24

S o l a r f l u x ( W

/ m )

Hour

North East wall

South east wall

North west wall

Horizontal Roof

0

1

2

3

4

5

6

7

8

9

10

0 10 20 30 40 50

T ( i n d o o r ) - T ( o u t d o o r ) ( C )

ACH (1/h)

Roof without insulation

Roof insulation

Roof and NE wall insulation

0

2

4

6

8

10

12

T w i t h o u t v e n t i l a t i o n ( )

Without insulation

Roof with insulation

Roof and NE insulation


8/8

Vol/h. The overheating is mainly due to the sunexposure of the northeast wall and can be reducedwith an additional outside insulation, for instancewith a thin polyurethane layer about 20 mm.Finally, the overheating is lower than 2C for therange of ACH expected to be reach during summer.

Note for the summer months, statistically speaking,ACH is lower than 10 volume/hour 5 hours permonth and lower than 20 volume/hour 17 hours permonth. In others words, the mean overheating of therenovated flat could only exceed 1C one day asummer month and 2C for only a few hours.

CONCLUSIONThe indoor thermal comfort in a tropical climate ofnaturally ventilated buildings depends both on thethermal behaviours of the buildings and the airexchange through the openings. Natural ventilation isless effective in urban environments and is not easilydesigned because of the complexity of wind velocitybehaviours. Hence air exchanges across thebuildings openings are functions of the wind speedin the urban areas, of the building shape and on theefficiency of the openings. Designers and architectsalike, need tools to swiftly validate the coolingstrategy for the building design. The openings(positions, dimensions and efficiency) and thethermal definition of the envelope (insulation layer,sunscreen, windows glass) need to be defined withsimple tools that give results which are easilyunderstood.That being said, this article presented a method toassess the indoor thermal comfort in a warm tropicalclimate. The method is made of two separateaerodynamic and thermal tools. First, the assessmentis carried out from an aerodynamic point of view. AirChange Rates are evaluated for each apartment fromthe buildings in this project. Statistics of ventilationrate are then extracted from the local climatology.The second step consists of computing the thermalresponse of a building to the climatic parameters:wind, temperature, humidity and solar radiation.Mean overheating inside the building is calculated

and compared with the maximum indoor temperatureacceptable according to the comfort criteria, forinstance with the Givonis diagram.Finally, a case of urban renovation of on olderbuilding at the city of Saint-Denis (La reunionIsland), is presented here as an example of the use ofcombined approaches devoted to designers. The toolsallow upgrading the strategy of cross ventilation byintroducing new openings in order to obtain an airchange rate high enough to allow heat extraction. Theenvelope is also upgraded according to the thermaltool resulting in the improvement of the insulation ofthe roof and of external sidewall source of heatingflux.

ACKNOWLEDGEMENTAuthors are grateful to ADEME (La Reunion agency)that supported this research, and to IMAGEEN andSocite Immobilire du Dpartement de la Runion that allow using the results produced.

REFERENCESAbdesselam M., 1997, Contribution l'tude

analytique du problme du dimensionnementthermique des btiments : application laconception thermique des btiments en payschauds. PhD Thsis, Ecole des Mines de Paris.

Ayad S.S., 1999, Computational study of naturalventilation, Journal of Wind Engineering andIndustrial Aerodynamics, 82, pp 49-68

Aynsley R.M, Melbourne. W., Vickery B.J., 1977,Architectural aerodynamics, Applied sciencePublishers London

Clarke J, 1990, Hand J, Strachan P, A building andplant energy simulation system. EnergySimulation Resarch Unit, Department ofMechanical Engineering, University ofStrathclyde.

Eurocode 1 : Actions on structure Part 1-4 : generalactions Wind Actions, EN 1991-1-4, CEN/TC250

Fahssis K., Dupont G., Sanquer S., Leyronnas P.,2010, Integration of the natural cross ventilationin the CFD software UrbaWind, International

Conference on Applied Energy, 21-23th of April2010, Singapor.

Garde F., 2006, Ralisation dun btiment nergiepositive : De la phase programme la phaseconception, Journe SFT-IPBSA INESChambry 21 mars 2006

Grosso M, 1992 Wind pressure distribution aroundbuildings : A Parametrical Model, Energy andBuilding, 18: 101-131.

Kandola B.S., 1990, Effects of atmospheric Wind onflows through natural convection roof vents, Firetechnology, 05-1990, pp106-120

Kruger E., Cruz E.G., and Givoni. B., 2010Effectiveness of indirect evaporative cooling andthermal mass in a hot arid climate, Building andEnvironment, n45, pp 1422-1433

Liddament, M. W., 1986, Air Infiltration CalculationTechniques, An Applications Guide, AIVC.

Sanquer S., Caniot G., Li W., Delaunay D., 2011, Acombined CFD-Network method for the cross-ventilation assessment in buildings, 13 th International Conference on Wind Engineering,Amsterdam.

Download - Combined CFD-Mean energy balance methode to thermal comfort assessment of buildings in a warm tropical climate

Top Related