Higher turbulent mixing due to UnoDuct - air supply system

Transcription

1 Higher turbulent mixing ue to UnoDuct - air supply system Christian Heschl, Anrea Kelz, Fachhochschulstuiengänge Burgenlan GmbH Abstract Previous investigations of free jet propagation are usually limite to axialsymmetrical an qasi two-imensional jets. Multiple jets in the form of nozzle arrays or slot inlets have harly been subject of systematic investigation. In this article the turbulent mixing processes occuring in this case are iscusse with the ai of Computational Flui Dynamics. It is shown that velocity an temperature reuction as well as the turbulent mixing process are epenent on the istance between the air supplies. Introuction The main task of an HVAC system is to provie the require room air quality. To fulfil this requirement suitable air supply systems alreay have to be selecte an imensione in the planning stage. If a mechanical ventilation system is to be use, a isplacement or a mixing ventilation system are available. In the majority of cases the mixing concept is chosen, which means that the air is supplie with high momentum. In orer to satisfy the user requirements the air supplies have to ensure quick reuction of velocity an temperature between the inlet air an ambient air. Therefore a high turbulent mixing process with the ambient air must be guarantee. Experimental an numerical investigations show that the exchange processes are essentially affecte by the turbulence level an the inlet flow velocity [1, 2, 3]. Room epenent influence variables such as room geometry, natural convection an temperature stratification are known [4, 5]. In orer to create best conitions for the user the proucer not only has to optimize the turbulent mixing process but also to provie calculation tools for appropriate utilization of room conitione influences to esigners of ventilation systems. Experience ventilation technicians try to use special flow phenomena such as the Coana effect to improve the flow situation in ventilate rooms. The Coana effect is base on the fact that air is entraine at the jet bounary ue to turbulent mixing processes. If not enough air can be entraine, e.g. on account of ceiling, unerpressure arises an the jet is iverte on the ceiling. This effect can be force by obstructing the lateral entrains of air through long slot inlets. For this reason slot an jet lines are increasingly use. In spite of using the Coana effect it often comes to a col snap in cooling cases. If this happens in the inhabite zone, it can lea to raught an thermal uncomfortableness. This unesire effect is prouce by the Archimees principle an is epenent on the temperature ifference between the ambient air

2 an the jet core. For this reason ventilation technicians strive to reuce the temperature ifference as quickly as possible. In this article the turbulence parameters of jet line inlets an slot inlets are analyse an their effect on velocity an temperature istribution are iscusse. In orer to make any perioic interruption of the jet line inlets possible the investigations are base on the UnoDuct-system of Lufttechnik J. Pichler GmbH. The investigations of the slot inlet system are base on a conventional supply system. Examination metho The classical experiments as well as the numerical experiments, e.g. Computational Flui Dynamics CFD, can be use to get the necessary parameters. Although both methos have ifferent expressiveness they are closely relate an complement each other. The avantage of the numeric experiment to get all information about substantial physical values, however, is gaine at the expense of general valiity. The quality of the results epens on the physical moel an the use numerical methos. Therefore it is necessary to valiate the numerical results with experimental finings. In orer to meet these emans extensive valiations were mae [6, 7]. For the calculations the commercial CFD coe FLUENT 5.5 was use [9]. Analyze air supply systems Two air inlet systems which allow air supply over the whole room with are analyze. The essential flow parameters of the slot inlet an the UnoDuct inlet system are evaluate an interprete. With the UNO-DUCT inlet system the supply air is transporte by a so-calle nozzle uct. This uct consists of a spiral-fol-uct with form-close nozzle elements. The layout of the nozzle elements can be arrange any way you want. On grouns of stability a minimum istance between the nozzle elements in axial irection ( s 89mm ) an tangential irection ( s t 30mm ) must be kept. Every nozzle element consists of seven bores with a iameter of approximately 8.6 mm. In this way you get a moular system which allows a high number of combinations which can be easily aapte to local emans. s Figure 1: Detail nozzle element an air supply system UnoDuct

3 The simultaneous use of the spiral-fol-uct as girer an supply system makes the istribution of supply air in all irections (iffuse) possible. The generate small single jets increase the whole entraine seconary air. For this reason the flow velocity an the temperature ifference are reuce consierably faster. The investigation referre to in this article is limite to in one row assemble nozzle elements. Accoring to the istance s a change of entrainment an reuction of temperature an flow velocity are expecte. In orer to investigate the effect of single jets on jet propagation two simulation moels with ifferent inlet bounary conitions are generate. Simulation moels The simulation results essentially epen on the inlet bounary conitions, i.e. on the flow parameters of the respective air supply. For this reason separate numerical an experimental investigations are mae to etermine the inlet bounary conitions for the UNO-DUCT nozzle system. It can be shown that the simulate time average outflow velocity, the outflow irection an the turbulence intensity agree well with the experimental ata. Besies, as a result of air eflection, a ouble swirl (similar to air passing a pipe ben) was observe. To emonstrate these finings, figure 2 shows the contour plot in the inlet plane of the velocity components. Figure 2: Contour plot of the ouble swirl - Uno-Duct supply system (p stat =20Pa, u uct =4m/s) From these autarkic numerical investigations the necessary inlet bounary conitions were generate. On account of the goo agreement between measure an simulate ata the etermination of the bounary conition for the slot inlet system was also generate via numerical investigations. Therefore the fulfillment of the moel equations is ensure. Figure 3 shows the esign of the use simulation moel. The inlet consiste of eleven nozzle elements with seven bores each an a slot with a with of b = 520mm. The iameters of the bores were = 8. 6mm an the slot height was h = 8. 6mm. The proportions of the simulation omain were efine with L B H = m³.

4 Symmetry - Bounary Conition x Pressure Inlet - Bounary Conition 2 UnoDuct Air Supply System or Slot Supply System Symmetry - Bounary Conition x 2 Pressure Outlet - Bounary Conition Wall - Bounary Conition H/2 x 1 bzw. h x 3 H/2 S S S Pressure Inlet - Bounary Conition b B L Figure 3: Bounary conitions of the simulation moels The chosen inlet geometry efinitions ensure ientical fluxes, i. e. both air supply systems have the same mass, momentum an energy fluxes in the inlet plane. Thus the essential requirements for an objective comparison of the system are met. Discussion of the results The characteristic values are presente non-imensional in orer to make the results easier to access for general interpretation. On account of the three imensional flow the x = 1 0 an x = 0 2 plane (cf. figure 3) are evaluate. As calculation basis a nozzle element spacing of s = 89mm an a slot spacing of B = 979mm was use. The outwar jet entrains the ambient air from top, bottom an sie irection. On account of the chosen moel setup air can not entrain across the symmetry planes. For this reason a quasi plane jet is evelope ownstream. The spee of transition from a multi jet to a single plane jet epens on the spacing between the supply air elements an the turbulent mixing process. Figure 4 an 5 shows the non-imensional time average velocity an temperature in the x = 0 plane. 2 Figure 4: average velocity an temperature air supply system UnoDuct

5 Figure 5: average temperature an temperature air supply system UnoDuct Within a length of x < 100 the velocity graients in Figure 4 can be clearly etecte. They 3 generate aitional, turbulent momentum an heat transfer cross to the main flow irection. With the slot air supply system turbulent transverse motions insie the x 0 plane are also etecte. On account of the big spacing between the inlet elements, the velocity peaks are reuce more slowly. This behavior at the transition between the multi jet an the single jet was expecte. In orer to be able to analyze the global effects of the ifferent turbulent transport processes figure 6 shows the ratio between volume flow at the inlet an the volume flow which is entraine within the jet. 2 = α [1] x 3 /h an x 3 / [1] 10 2 Figure 6: ratio entraine volume flow The parameter α is efine as follows ( x3 ) ( x ) V& UnoDuct α =. (7) V& Schlitzauslass 3 It can be clearly prove that in the area of interest (jet length approx. 100 to 400) up to 15% more ambient air is entraine by the UnoDuct air supply system.

6 With bigger istances the ratio α gets smaller an approaches asymptotically the value 1. This can be explaine by the fact that after a specific jet length a quasi plane jet is evelope by the slot air supply system. u 3 /u 3,in [1] Velocity Magnitue (T-T )/(T in -T ) [1] Temperature UnoDuct Slot Inlet x /h an x 3 / [1] x 3 /h an x 3 / [1] Figure 7: average velocity an temperature on the mean jet centerline Due to the higher entrainment of ambient air an the faster formation of a quasi plane jet, a faster velocity an temperature reuction are allowe. Accoring to figure 7 an about 25 % smaller value for the average velocity an temperature on the jet centerline can be expecte insie a jet length from 100 to 400. Summary Air supply methos have an essential impact on the thermal comfort of the user. In orer to meet the emans of the users various air supply systems are available for the esigners. In the cooling case systems are often use which ensure the avantage of the Coana effect. Apart from slot air supply systems, in a row locate nozzle elements are often use. Due to easy construction these nozzle elements can be irectly integrate in spiral fol ucts. Contrary to slot inlet systems the inlet air is supplie by means of a big number of little bores which are integrate in nozzle elements. After entry of supply air small single jets evelop. The nozzle elements are arrange next to each other, leaving only tiny gaps. These geometrical conitions immeiately lea to high velocity an temperature graients whose effects were analyze. For the investigation numeric methos valiate in other investigations were use. Comparative investigations between a conventional slot air supply system an the escribe UnoDuct system of Lufttechnik J. Pichler GmbH were carrie out. It was assume that the mass, momentum an energy fluxes of both systems agree. In orer to eliminate the influence of room geometry, free jet bounary conitions were suppose. In aition, buoyancy effects were neglecte.

7 Due to the smaller spacing between the inlet elements, the UnoDuct nozzle air supply system allows a consierably faster transition to the quasi plane jet than with the flux-aequate (ientical mass, momentum an energy flux) multiple jet. Besies, consierably higher turbulent mixing processes insie the interesting area (jet length approx. 100 to 400) were observe. It coul be shown that faster transition to the quasi plane jet an more intensive turbulence quantities lea to better entrainment. Uner the escribe bounary conitions the entraine volume flow rate is up to 15% larger than with the UnoDuct nozzle air supply system. The reuction of the time average velocity an temperature at the jet centerline is about 25% higher. Particularly in the cooling case this facilitates the compliance with the requirements of thermal inoor air quality.

8 Bibliography [1] Hanel, B., Beitrag zur Berechnung von Freistrahlen mit erhöhter Anfangsturbulenz, Luft- un Kältetechnik, (1977), [2] Hanel, B., Richter, E., Das Verhalten von Freistrahlen in verschieenen Reynols- Zahlenbereichen, Luft- un Kältetechnik, (1979), [3] Regenscheit, B., Einfluss er Reynolszahl auf ie Geschwinigkeitsabnahme turbulenter Freistrahlen, HLH 27, (1976), [4] Peng, S., Moelling of Turbulent Flow an Heat Transfer for Builing Ventilation, PhD-Thesis, Chalmers University of Technology, Gothenburg, (1998). [5] Zhang, G., Morsing, S., Bjerg, B., Sviit, K., A Stuy on the characteristics of airflow in a full scale room with a slot wall inlet beneath the ceiling, Proceeings of RoomVent 2000, (2000), [6] Heschl, H., Fesharaki, M., Raumluftströmungssimulation, Internationaler Kongress e.nova.2001, Pinkafel, (2001). [7] Hanel, B., Die Berechnung er Mischungszone eines axialsymmetrischen turbulenten Freistrahls, Luft- un Kältetechnik, (1976), [8] Heschl, H., Fesharaki, M., Steinkellner M., Valiation er CFD-Analyse von Raumluftströmungen mittels Messungen, Fluent Anwenerkonferenz, Frankenthal, (2002). [9] FLUENT, FLUENT 5.0 User s Guie, Fluent Inc. Centerra Resource Park, Lebanon, (1998).

9 Nomenclature b B slot with [m] with of the calculation omain [m] bore iameter, uct iameter [m] p stat static pressure [Pa] s t T spacing between two nozzle elements [m] time [s] temperature [K] T ein average inlet temperature [K] T max max. temperature in the observe plane [K] T u i ambient temperature [K] average velocity in the x, y an z-irection (i = 1,2 bzw. 3) [m/s] u i, ein average inlet velocity [m/s] u i,max max. velocity in the observe plane [m/s] x i geometry irection x, y an. z-irection (i = 1,2 bzw. 3) [m] α parameter volumetric flow rate [1]