1、382 2010 ASHRAEABSTRACTA mechanically ventilated, photovoltaic-integrated,multiple skin facade, consisting of a between-the-panes Vene-tian blind layer in the upper section and a between-the-panesphotovoltaic array in the lower section was considered. Thecombined photovoltaic and shading system can
2、produce elec-tricity and thermal energy in the form of preheated fresh air,and allow for adjustable daylighting. Numerical simulationswere developed to predict the velocity of air and glazing andshading temperatures only in the upper section of the system(around the Venetian blind slats). A total of
3、 nine experimentalscenarios were designed and tested at three blind slat angles( = 0, 45, and 75) and three fan speeds (10, 20, and 30 Hz)in an outdoor test room. The simulations predicted the averageblind and indoor glazing layer temperatures to within 2.3 and5.4C (4.1 and 9.7F), respectively, of t
4、he experiment andpredicted the air peak streamwise velocities to within 18% ofParticle Image Velocimetry measurements. The present studyintroduce, for the first time, a detailed account of numericalsimulation development of ventilated windows with between-the-panes Venetian blinds. Future studies wi
5、ll focus on deriv-ing between-the-panes convective heat transfer correlationsand developing simple one-dimensional models that can beintegrated into building energy simulation software.INTRODUCTIONShading devices such as Venetian blinds are commonlyused in building applications to control daylight,
6、reduce glare,and to control the fenestration total thermal transmission (U-value) and the Solar Heat Gain Coefficient (SHGC). They arecommonly placed on the interior side of a glazing unit, but canalso be placed in between a double or multiple skin glazingarrangement. There are several advantages of
7、 placing a blindlayer in a between-the-panes arrangement. Other than portray-ing an aesthetically pleasing look, it facilitates the control andautomation of the drive systems, because it provides a goodlocation to safely place delicate mechanisms. It also bringsadditional solar-thermal savings if am
8、bient air is drawn fromthe outdoor environment, forced to flow over the blind layer,and then collected and supplied to the heating, ventilation, andair conditioning (HVAC) unit of building. This is what isreferred to as a ventilated window in a supply-air mode.The presence of a blind layer inside a
9、ventilated windowcan significantly alter the thermal characteristics of a venti-lated window. For example, the direct transmission gains (theconvective and long-wave radiative heat flux from the inner-most pane to the indoor environment) and the thermal gains(the convective heat transfer from the gl
10、azing and blindsurfaces to the channel air) depend on the slat angle, spacings,slat material (thermophysical and radiative) properties, andairflow characteristics.Mechanically ventilated windows can be operated in anumber of modes, depending on the position of the inlet andoutlet with respect to the
11、 indoor or outdoor environment, andalso the direction of airflow. Two of the most important modesof their operation are “supply-air” and “exhaust-air” becausethey can directly and effectively reduce the heating and cool-ing loads of a building, respectively. A ventilated window, infact, must be capa
12、ble of operation under various modes toensure overall year-round benefits (Saelens et al. 2003).During the heating season, a supply-air arrangement forms aninsulating air layer between the interior and exterior and alsobrings heated air inside the building. This compensates for thespace heating load
13、 of the HVAC unit. During the coolingExperimental and Numerical Investigation of a Mechanically Ventilated, Multiple Skin Faade with Between-the-Panes Venetian BlindsOmid Nemati Michael R. Collins, PhDStudent Member ASHRAE Associate Member ASHRAELuis Candanedo Andreas Athienitis, PhD, PEStudent Memb
14、er ASHRAE Member ASHRAEOmid Nemati is a student and Michael R. Collins is an assistant professor in the Department of Mechanical and Mechatronic Engineering,University of Waterloo, Ontario, Canada. Luis Candanedo is a student and Andreas Athienitis is a professor in the Department of Building,Civil
15、and Environmental Engineering, Concordia University, Quebec, Canada.OR-10-041 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution,
16、 or transmission in either print or digital form is not permitted without ASHRAEs prior written permission. ASHRAE Transactions 383season, an exhaust-air arrangement can be used to alleviateoverheating problems by taking the air inside the building,forcing it to flow upwards or downwards through the
17、 windowand then exhausting it to the exterior. Thus, surface tempera-tures inside the ventilated window will remain lower, andresult in lower overall transmission gains during the summer.Two vertical, faade-integrated solar-thermal systems,consisting of upper shading sections and lower PV sections,w
18、ere constructed on a roof-top test room at Concordia Uni-versity to study the potentials of these systems, model theirthermal performance and perhaps design and optimize theiroperation (Figure 1). These two, so-called Building-Inte-grated Photovoltaic (BIPV/T) configurations, are mechani-cally venti
19、lated and can replace an entire south facing facade.Previous works on these two configurations at Concordia Uni-versity by Liao et al. (2007) and Charron and Athienitis (2006)reveal that both configurations are capable of reducing theheating demand during the heating season (supply-air mode).Also, a
20、lthough not tested yet, the two configurations arethought to be able to reduce the cooling demand during thecooling season also (return-air mode). Their measurementsshowed that when the PV panel was placed in a between-the-panes arrangement (configuration II) the PV section efficiencywas as much as
21、25% higher than when the PV panel wasplaced on the exterior of the setup (configuration I). However,the generated electricity of configuration II could be 21%lower because of reflection from the surface of the exteriorpane, especially at high incidence angles. The overall com-bined PV and window sec
22、tion efficiency of both configura-tions could reach 70%.In this study only the upper section of one of the sections(a ventilated window with between-the-panes Venetianblinds) has been considered. The PV panel has been integratedinto the ventilated window system to bring additional savingsbut the ven
23、tilated window section can be studied separately.The PV panel however affects the inlet (just below the lowestslat and just above the PV) air velocity and temperaturecompared to the case if the ventilated window was a stand-alone system. The PV panel causes the inlet velocity andtemperature profiles
24、 to be quite asymmetric. Careful model-ling of the inlet conditions is therefore essential in accuratemodelling of the multiple skin facades (Saelens et al. 2003).Numerical modelling of ventilated windows is a challeng-ing task due to the various modes of heat transfer and the cou-pled nature of the
25、 heat transfer mechanisms. Free and forcedconvection, long-wave and short-wave radiation, and conduc-tion through solids must all be modeled. In addition, simpleconventional U- and SHGC-type models do not adequatelyrepresent the entire dynamics of the system because they donot take into account the
26、enthalpy change of the air. In otherwords, the direction of heat transfer is not only in the horizon-tal direction, but is also in the vertical direction to the air. Forthese reasons, to numerically model ventilated windows com-prehensive Computational Fluid Dynamics (CFD) simula-tions, with as litt
27、le simplifying assumptions as possible, mustbe considered as a first step in analyzing and optimizing BIPV/T systems. There is very little information available howeveron the modelling of mechanically ventilated windows withbetween-the-panes Venetian blinds. There is only one study bySafer et al. (2
28、005) which presents numerical modelling of amechanically ventilated, double-skin window with between-the-panes Venetian blinds developed in a typical commercialCFD package. In that study Safer et al. (2005) simulated theflow field (not the temperature field) using a steady-state sim-ulation and a tu
29、rbulent k- model. Safer et al. (2005) believesthat such simulations can predict the air velocity field reason-ably well.The present work introduces, for the first time, acombined numerical and experimental study of a ventilatedwindow with between-the-panes Venetian blinds undermechanical ventilation
30、. Detailed numerical simulation devel-opment (including the temperature field) and validation havebeen presented. The simulations were validated againsttemperature measurements and Particle Image VelocimetryFigure 1 The two BIPV/T configurations at Concordia Uni-versity (a) and the schematic of conf
31、iguration II (b). 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not perm
32、itted without ASHRAEs prior written permission. 384 ASHRAE Transactions(PIV) measurements. Once a validated CFD model is in place,it can be used to obtain between-the-panes average heat trans-fer coefficients and pave way for simple one-dimensional(1-D) modelling of the system where every glazing or
33、 shadinglayer is represented by just one node. Such simple models willbe very helpful because they can be used to quickly and easilyinvestigate the effect of a design change when integrated intobuilding energy simulation software. For example, they can beused to quickly find the effect of a geometry
34、 change on the heatflux distribution inside the system. Nemati et al. (2009) showsthat despite the complexity of the heat transfer interactionsinside the window and despite high layer vertical temperaturestratifications, a simple 1-D model, based on isothermal layerassumption and average heat transf
35、er coefficients, can predictthe area-averaged blind layer temperature to within 2.5C(4.5F) of CFD and within 3.0C (5.4F) of experiment andthe average indoor glazing (outdoor pane) layer temperature towithin 3.3C (5.9F) of CFD and within 5.2C (9.4F) ofexperiment. Ongoing work is in progress to formul
36、ate semi-empirical correlations for between-the-panes convective heattransfer that can be used in simple models to design and opti-mizate BIPV/T systems.EXPERIMENTAL SETUPA total of 12 Venetian blind slats (49 mm (1.9 in.) wideand 43.9 mm (1.7 in.) spaced apart) are placed mid-waybetween a double gl
37、azing unit on the interior side and a single-layer glazing on the exterior. The channel is 92 mm (3.6 in.)wide. The window is 498 mm (19.6 in.) high and 956 mm (37.6in.) long (depth into paper). The single-layer glazing is clearand 5 mm (0.20 in.) thick. The double glazing unit consists of3 mm (0.12
38、 in.) thick glazing panes, spaced 12 mm (0.47 in.)apart, with a low-emissivity coating of 0.1 applied on theindoor pane facing the exterior. A variable-frequency fan,operating between a frequency range of 0 to 60 Hz, forces airto move upward through the ventilated channel.A total of nine scenarios w
39、ere tested corresponding tothree blind slats angles (), 0, 45, and 75, and three fan speedsettings, 10, 20, and 30 Hz on Sep. 22, 2008 around noon(Table 1). During each scenario, radiation-shielded T-typethermocouples (accuracy 0.5C (0.9F) were used tomeasure the blind surface temperature (Tb) and t
40、he outdoorand indoor glazing surface temperatures (Tg,oand Tg,i) at threelocations inside the ventilated channel: close to the top frame(11th slat from the bottom), center-window (7th slat from thebottom), and close to the bottom frame (3rd slat from thebottom). Tbwas measured, for each slat, on var
41、ious top andbottom, as well as shaded and unshaded spots. Also, duringeach scenario, the outdoor and indoor ambient air tempera-tures (Textand Tint), wind speed, total irradiation on a south-facing vertical surface (I) were measured using an integratedweather station. Also, the air streamwise veloci
42、ty and temper-ature profiles were measured at the inlet (just below the lowestslat and just above the PV panel) using a translating hot wireanemometer (temperature accuracy of 0.3C (0.5F) andvelocity measurement repeatability of 0.03 m/s 1% (0.10 ft/s 3.2808%) of reading). The incoming air mean air
43、speed( ) and temperature (Tin) were calculated from the anemom-eter readings through: (1)(2)PIV is a non-intrusive method of calculating the velocityvector map of a fluid. It is based on the spatial and temporalresolution of motion of small particles released inside the flow.These so-called seeding
44、particles glow when a laser beam ofspecific wavelength shines upon them and so their locationscan be captured in digital recording media. By comparing thelocation of a group of particles in one interrogation cell at thebeginning and end of a short time interval (measured in mi-VVVxdY=00w xdY=00w-=Ti
45、nVT xdY=00wVxdY=00w-=Table 1. Experimental Scenario DesignScenario Fan Speed, Hz , , m/s (ft/s) Tin, C (F) Text, C (F) Tint, C (F) I, W/m2 (Btu/hft2)1 10 0 0.13 (0.43) 40.5 (105) 13.0 (55.4) 22.8 (73.0) 680 (216)2 20 0 0.31 (1.0) 36.8 (98.2) 13.0 (55.4) 22.8 (73.0) 680 (216)3 30 0 0.56 (1.8) 32.7 (9
46、0.9) 13.0 (55.4) 22.8 (73.0) 680 (216)4 10 45 0.12 (0.39) 34.4 (93.9) 10.8 (51.4) 22.0 (71.6) 650 (206)5 20 45 0.35 (1.1) 32.7 (90.9) 10.8 (51.4) 22.0 (71.6) 650 (206)6 30 45 0.62 (2.0) 29.1 (84.4) 10.8 (51.4) 22.0 (71.6) 650 (206)7 10 75 0.13 (0.43) 39.3 (103) 13.1 (55.6) 22.2 (72.0) 692 (219)8 20
47、75 0.46 (1.5) 35.3 (95.5) 13.1 (55.6) 22.2 (72.0) 692 (219)9 30 75 0.64 (2.1) 30.8 (87.4) 13.1 (55.6) 22.2 (72.0) 692 (219)V 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal u
48、se only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission. ASHRAE Transactions 385croseconds) the velocity vector of the fluid can be calculatedin that interrogation cell during that time interval through
49、 aseries of PIV correlation algorithms.A complete PIV system, including the laser and thecomputer processor, was installed inside the test room(Figure 2). The laser is of type Nd:YAG with a laser-sheetthickness of 1.5 mm (0.59 in.). A 10-bit camera (1600 1186pixels) and a 60 mm f /2.8D lens were used to capture the PIVpictures. While the laser and camera are normally placedperpendicular to each other, they had to be placed parallel toeach other because limited space was available on
