Applied Thermal Engineering 31 (2011) 902e910 Contents lists available at ScienceDirect Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng Towards the control of car underhood thermal conditions Mahmoud Khaled a, b, c, Fabien Harambat b, Hassan Peerhossaini a, * a Thermofluids, Complex Flows and Energy Research Group, Laboratoire de Thermocinétique, CNRS-UMR 6607, Ecole polytechnique, University of Nantes, rue C.Pauc, BP 50609, 44 306 Nantes Cedex 3, France b PSA Peugeot Citroën, Velizy A Center, 2 route de Gisy, 78 943 Vélizy Villacoublay, France c Fluid Mechanics, Heat and Thermodynamics group, School of Engineering, Lebanese International University, Beirut, Lebanon a r t i c l e i n f o a b s t r a c t Article history: Received 19 July 2010 Accepted 10 November 2010 Available online 18 November 2010 The present paper reports an experimental study of the aerothermal phenomena in the vehicle underhood compartment as investigated by measuring temperature, convective heat flux, and radiative heat flux. Measurements are carried out on a passenger vehicle in wind tunnel S4 of Saint-Cyr-France. The underhood space is instrumented by 120 surface and air thermocouples and 20 fluxmeters. Measurements are performed for three thermal functioning conditions while the engine is in operation and the front wheels are positioned on the test facility with power-absorption-controlled rollers. In the thermal analysis, particular attention is given to measuring absorbed convective heat fluxes at component surfaces. It is shown that, in some components, the outside air entering the engine compartment (for cooling certain components) can in fact heat other components. This problem arises from the underhood architecture, specifically the positioning of some components downstream of warmer components in the same airflow. Optimized thermal management suggests placing these components further upstream or isolating them from the hot stream by deflectors. Given style constraints, however, the use of air deflectors is more suitable than underhood architectural changes. Much of the present paper is devoted to heat flux analysis of the specific thermal behaviours in the underhood compartment (especially the absorption of convective heat fluxes) and to a description of a new control approach exploiting air deflectors to optimize underhood aerothermal management. Ó 2010 Elsevier Ltd. All rights reserved. Keywords: Underhood aerothermal management Temperature Heat flux Convection Radiation Physical analysis Air deflectors 1. Introduction Recent trends in the automotive market stress high-performance engines and climate-control systems. At the same time, however, automobile design must respect geometrical restrictions related to style constraints. More and more components must be implemented in the underhood space, and the desire for noise reduction has augmented the use of underhood insulation. The underhood compartment is thus becoming more and more cramped, causing complex airflows and difficult air paths that give raise to complex aerothermal phenomena (especially convection and radiation). These phenomena pose aerothermal management challenges in air-intake design and front-end cooling module layout that exacerbate the complex geometry of the underhood space. Experimental analyses of these phenomena [1e4] are rare; studies have focused mainly on numerical simulations [5e10], which themselves concentrate on the major cases in temperature * Corresponding author. Tel.: þ33 2 40 68 31 24; fax: þ33 2 40 68 31 41. E-mail address: hassan.peerhossaini@univ-nantes.fr (H. Peerhossaini). 1359-4311/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.applthermaleng.2010.11.013 trend analysis. A review of the existing experimental literature in underhood studies shows very little physical analysis of heat transfer, and that mainly serving to comment on calculation/test comparisons [11e15]. On the other hand, the underhood geometry is already so confined and complex that implementing any design improvements is difficult; the remaining solution is to optimize aerothermal management by focusing on its operation mode rather than its architecture. The present paper reports a physical analysis of particular underhood aerothermal behaviors intended as a basis for a new optimization approach [16] of cooling airflow “rearrangements” in the vehicle underhood compartments. It discusses first an experimental study of the aerothermal phenomena in the vehicle underhood compartment by temperature measurement and separate measurements of convective and radiative heat fluxes. Measurements are carried out on a passenger vehicle in wind tunnel S4 of Saint-Cyr-France. The underhood space is instrumented by 120 surface and air thermocouples and 20 fluxmeters. Measurements are performed for three thermal functioning conditions, with the engine in operation and the front wheels positioned on the experimental facility with power-absorption-controlled rollers. In M. Khaled et al. / Applied Thermal Engineering 31 (2011) 902e910 Subscripts and superscripts c convective CF cooling fluid CH cylinder head CJ cylinder jacket COMB combustion COOL cooling EV exhaust valves EX exhaust g global m mechanical max maximal r radiative S surface SEG segment th thermal 0 initial Nomenclature hc n P Q R t T V Wnet convective heat transfer coefficient, W m engine regime, rpm engine power, kW heat flux, kW gearbox ratio time, s temperature,  C velocity, m s 1 net work produced in the cylinder, kW Greek symbols 4 heat flux density, W m 2 3 emissivity s StefaneBoltzmann constant, W m s time constant, s h efficiency 2 K 2 K 1 4 a second trial, an optimization approach using static and dynamic deflectors to orient the different airflow paths in the underhood compartment is then used that is based on the physical analysis of the temperature and heat flux measurements. The rest of this article is organized as follows. Section 2 gives a theoretical discussion of the basic heat transfer types in the underhood space and their possible arrangements. Section 3 describes the experimental setup and Section 4 is dedicated to the results and analysis of the temperature and heat flux measurements and to new optimization approaches based on this analysis. 2. Theoretical issues Combustion temperatures for motor fuel are around 2200  C and the temperatures of exhaust gas at the cylinder head outlet are of the order of 1000  C [17,18]. It is therefore necessary to cool the piston, valves, and cylinder walls in order to prevent them from melting. In automotive applications, cooling is most commonly done by water: the engine, particularly the cylinder head and cylinder block, contains cavities (water chambers) in which water circulates, controlled by a centrifugal pump. If a part QCOOL of the thermal energy QCOMB released by the reaction (combustion) is then removed by the water before any further exchange, it is quite similar to what would happen with a transmission of only ðQCOMB QCOOL Þ to the combustion gases. However, for engines of small and middle power, the combustion efficiency defined by: hCOMB ¼ ðQCOMB QCOOL Þ QCOMB Wnet QCOMB Finally, the overall efficiency of an engine is the ratio between the mechanical energy produced by the engine as power and the heat energy provided by the fuel. The overall performance depends on the thermal performance and mechanical efficiency: (3) hg ¼ hth hm Thus, engine function is governed by an overall efficiency that causes the energy restored in engine power to be less than the energy produced by combustion, the overall efficiency being about 0.3e0.4. The difference between primary energy and combustion engine power is then arranged among exhaust energy, energy lost by convection (in the cooling system and under the hood) and energy lost by radiation. Fig. 1 shows a schematic of the heat distribution, noting the orders of magnitudes involved, in the underhood space of a diesel engine [17,18]. The problem of underhood aerothermal management essentially entails controlling these heat exchanges and their impact on component temperatures. 3. Experimental setup and methods This section describes the different instrumentations performed in the underhood compartment, the experimental configurations and phases, and the experimental protocol [19e22]. Fuel combustion (1) is generally significantly high in the internal combustion system (about 70%). During the thermodynamic cycle of the air, net work is produced in the cylinder that corresponds to the energy generated in the pistons. The thermal efficiency of the thermodynamic cycle of an engine is defined as the ratio of the net work produced in the cylinder and the heat energy released by fuel combustion: hth ¼ 903 QCOMB Exhaust Cooling Oil Crank QEX 25% QCOOL 33% QOIL 8% Wnet 34% Exhaust valves Cylinder head Cylinder jacket Segment QEV 4,5% QCUL 12% QCJ 10,5% QSEG 6% (2) On the other hand, the mechanical efficiency hm, defined as the ratio between the energy produced on the piston (effective power) and energy collected on the flywheel of the crank (indicated power), is reduced by internal mechanical friction to 0.7 (0.8 when pieces are in as-new condition). Cooling fluid Radiation QCF 25% QRAD 8% Fig. 1. Heat arrangement in the underhood compartment and order of magnitude for a diesel engine [1,2]. 904 M. Khaled et al. / Applied Thermal Engineering 31 (2011) 902e910 Fig. 2. Schematic of (a) top and (b) side views of some underhood instrumented locations. 3.1. Underhood instrumentation The underhood compartment of the vehicle used here is instrumented by type T and K surface and air thermocouples and fluxmeters of normal gradients (fluxmeters based on temperature gradient measurement in a direction normal to the upper and lower faces of a thin plate of known thermal conductivity). Thermocouples permit temperature measurements at almost 120 positions corresponding to different components, air zones and engine parameters (engine fluid characteristic temperatures). Fluxmeters are attached to the surfaces in pairs (20 fluxmeters in 10 positions) so as to make separate measurements of the convective and radiative heat fluxes. This technique, described in [23e25], entails attaching to a surface of given emissivity two fluxmeters of different emissivities1. In this case, the overall heat fluxes measured by the two fluxmeters are not the same. By considering the surface temperatures and the convective heat transfer coefficient measured by the two fluxmeters to be approximately the same, one can deduce from the overall heat fluxes measured by the two fluxmeters the convective and radiative heat flux exchanged at the surface. Of the components considered, the most important are the exhaust manifold, the cold box (the box protecting the vehicle computer and battery), the alternator, the admission distributor, the air filter, the water outlet plenum, the apron, the charge air 1 A way to do this in practice is to paint one fluxmeter with black paint and the other with aluminum paint. cooler (CAC) inlet and outlet ducts, the cylinder head cover and the right side of the engine. Of the air zones, the most important are those close to the cowl, the apron (The apron is the upper part of the block which separates the passenger space from the underhood compartment), the cold box, the cylinder head cover, the air filter, the CAC inlet and outlet ducts and that downstream of the engine and the charge air cooler. Engine parameters are temperatures: of water at the radiator inlet and outlet, of air at the charge air cooler and compressor inlet and outlet, of gas at the turbine and catalyzer inlet and outlet, and of air at the engine (cylinder) inlet. Fig. 2 shows a schematic of the instrument locations in the underhood space (top and side views) and an example of instrumentation at the cold box. 3.2. Experimental setup and configurations Aerothermal experiments were performed in wind tunnel S4 of Saint-Cyr l’Ecole France, which is a 1:1 wind tunnel with section 5 m wide and 3 m high. To eliminate wall effects, blower S4 has a ventilated test section with longitudinal slots that simulates the flow very close to the actual flow around the vehicle. In addition, the wind tunnel has a roller chassis that can impose rolling resistance on the vehicle. Thus, it is possible to conduct experiments simulating real road conditions. The front wheels are placed on the chassis roller, the car engine runs during the tests, the driver controls different driving modes and the front wheels then entrain the rolls. The roller test facility is equipped with a brake system to adjust and control the power to the wheels and their rotational speed (Fig. 3). M. Khaled et al. / Applied Thermal Engineering 31 (2011) 902e910 905 measurements (Section 4.2), and suggestions for new control approaches (Section 4.3). 4.1. Underhood thermal behaviors During the constant-speed driving phase and for the different thermal functioning points, typical exponential trends are observed in all component temperature variations, air zones and engine parameters. Exponential trends are also obtained for the temporal variation of overall, convective and radiative heat fluxes for all components considered. Fig. 4 shows examples of these exponential trends for thermal point TFP-1; here the different heat flux curves are made dimensionless, by the initial flux for decreasing flux and by the final (infinite) flux for increasing flux. Temperatureetime variations in the constant-speed driving phase consistently have the general form Fig. 3. Rollers test facility equipment. Experiments were carried out for three different thermal functioning points that simulate more or less severe rolling situations from the thermal point of view (Table 1). 3.3. Test phases and experimental protocol For each experiment, data records cover three successive phases, each simulating a real situation with which a vehicle can be confronted: constant-speed driving (at a defined thermal functioning point, one of the three points in Table 1), slowdown and thermal soak. The constant-speed driving phase represents the rolling of a real vehicle at defined engine regime, gear ratio and wind speed. This phase can be TFP-1, TFP-2 or TFP-3. The slowdown phase is simulated in the wind tunnel by passing to neutral after the constant-speed phase and stopping the wind in the tunnel. The thermal soak phase follows the slowdown and simulates the vehicle stopping after a significant heat load. Here thermal inertia maintains temperatures higher even though the engine has stopped. In this case, cooling airflow under the hood is achieved only by fan rotation and/or free convection. At the beginning of this phase, the fan rotates for a brief period (1e5 min) before stopping. For a given configuration, experiment starts by stabilizing the desired thermal functioning point (TFP-1, TFP-2 or TFP-3). The engine regime reaches the preset value and data acquisition starts for the three phases of driving, slowdown and thermal soak. The constant-speed driving phase is maintained until temperatures stabilize. Slowdown comes next, when the driver releases the accelerator: the wind in the tunnel is cut and the car engine is turned off when the air velocity reaches zero. Thermal soak begins when the engine is turned off. During data recording, the beginning of each of the three phases is identified. 4. Results and discussion Here the temperature and heat flux measurements are analyzed in three parts: underhood thermal behavior description (Section 4.1), heat flux analysis by separate convective and radiative flux Table 1 Parameters defining the three experimental thermal functioning points: wheel and wind speeds, engine power, gearbox ratio and engine regime. Vwheel Km h PT-FCT-1 PT-FCT-2 PT-FCT-3 90 110 130 1 Vwind Km h 90 55 130 1 P kW R e n rpm 69 89 98 5 4 5 2600 3800 3780   t exp
T0 Þ$ 1 TðtÞ ¼ T0 þ ðTmax (4) s Here Tmax is the maximum temperature of quasi-stabilization at the end of the constant-speed phase and s is the time constant, i.e. the typical time to attain the stabilization regime. Typical exponential expressions describing the overall, convective, and radiative heat fluxes in the constant-speed driving phase are:
40 Þ$ 1 4 ¼ 40 þ ð4max 4 ¼ 4N þ ð40 4N Þ$exp  4c ¼ 4c;0 þ 4c;max  4c ¼ 4c;N þ 4c;0  4r ¼ 4r;0 þ 4r;max  exp 4r ¼ 4r;N þ 4r;0 t   s
 4c;N $exp (5) 
4r;N $exp (6) exp 4r;0 $ 1  s   4c;0 $ 1  t  t s t s t s   exp     (7) (8) t s  (9) (10) Here 40, 4c,0 and 4r,0 are respectively the overall, convective and radiative heat fluxes measured for each position at the beginning of the constant-speed driving phase. 4max, 4c,max and 4r,max are the maximum values during the phase and 4N , 4c;N and 4r;N are the asymptotic values at which the quasi-stabilization regime is reached. All these heat fluxes can be either positive or negative depending on whether the component receives (positive) or gives up (negative) heat to its environment, and they are essentially functions of the component location and the thermal functioning point. It should be recalled that the overall heat flux corresponds to the sum of the convective and radiative heat fluxes. For the overall heat flux variations, two categories can be distinguished: Category 1 components for which the overall heat flux follows the general form of equation (5). These are components that absorb or lose overall heat fluxes increasing in absolute value with time. In other words, these components heat more slowly than their local thermal environment when they absorb heat or more quickly when they lose heat; 906 M. Khaled et al. / Applied Thermal Engineering 31 (2011) 902e910 Temperature (°C) Overall heat fluxes 90 80 70 60 50 40 Cold box Cold box upstream upstream Cylinder head Cylinder head cover cover Air filter Air filter 30 20 Dimensionless ϕ Temperatures 100 0,80 0,30 -0,70 10 -1,20 500 1000 1500 1000 1500 Convective heat fluxes Radiative heat fluxes 0,30 Cold box Cold box side side Air filter Air filter CAC inlet CAC inlet duct duct -0,70 500 Time (s) 0,80 -0,20 0 Time (s) Dimensionless ϕ r 0 Dimensionless ϕ c Cold box Cold box side side Air filter Air filter CAC inlet CAC inlet duct duct -0,20 0,80 0,30 -0,20 Cold box Cold box side side Cold box Cold box above above Enigne right Enigne right sideside -0,70 -1,20 -1,20 0 500 1000 0 1500 500 Time (s) 1000 1500 Time (s) Fig. 4. Examples of temperatures and heat flux variations at some of the tested components in TFP-1. The same distinction can be noticed in the variation of convective heat flux (equations (7) and (8)) and radiative heat flux (equations (9) and (10)). However, an overall heat flux variation of the first type does not necessarily arise from variations in the convective and radiative heat fluxes of the same category. For example, as seen in Fig. 5 for the cold box component, the radiative heat flux increases in absolute value (category 1), differently from the overall and convective heat fluxes (category 2). a 1000 Part 11 Part Part 22 Part 800 Heat flux (W/m²) Category 2 components for which the overall heat flux follows the general form of equation (6). These are components that absorb or lose overall heat fluxes decreasing in absolute value with time. In other words, these components heat more quickly than their nearby thermal environment when they absorb heat or more slowly when they lose heat. Overall flux flux Overall Convective flux Convective flux Radiative flux flux Radiative 600 400 200 0 -200 0 200 400 800 1000 1200 1400 Time (s) 4.2. Heat flux analysis e components heated by convection b 80,0 Part Part 11 75,0 Temperature (°C) All components absorb heat in the constant-speed driving phase. But the surfaces of all components have at least one portion that receives heat and at least one that emits heat. Consider the first case, i.e. surfaces where the overall heat flux is absorbed (thus positive). Measurements show that, in particular cases, the convective heat flux is also positive. In other words, air that cools the different underhood components induces the opposite effect in some areas and tends to heat them. This is observed, for example, for the air filter or cold box (battery þ computer). Here the focus is on cases in which the components are heated by convection. On the other hand, measurements show that for almost all components, the overall heat flux is driven by the convective heat flux, i.e. the time variation of the overall heat flux follows that of the convective heat flux, whatever the sign and intensity of the radiative flux. Then, two types of variations can be distinguished: increasing convective flux that imposes increased overall heat flux, and decreasing convective flux that induces decreased overall heat flux. 600 Part Part 22 70,0 65,0 60,0 55,0 Surface temperature Surface temperature 50,0 Air Airtemperature temperature 45,0 40,0 0 500 1000 1500 Time (s) Fig. 5. Temporal variations of (a) heat flux and (b) temperature at the cold box and its surrounding air zone in TFP-3. M. Khaled et al. / Applied Thermal Engineering 31 (2011) 902e910 A typical example of the first type is the air filter. Fig. 5 shows variations in the overall heat flux and temperature on its surface and the surrounding air. It can be clearly seen that the overall heat flux exchanged at the air filter surface is absorbed flux (þ1350 W/ m2). Indeed, at first the air zone near the air filter is hotter than its surface (Fig. 5b): the passage of hot air, which has extracted heat from high-temperature components upstream of the air filter, especially the engine, provides a positive (absorbed) convective flux (þ500 W/m2). On the other hand, the radiative heat flux (þ850 W/m2) is absorbed heat flux, since the equivalent thermal environment of the air filter is hotter than its surface. Moreover, it can be noticed that in the first part of the constantspeed driving phase, the temperature of the air zone surrounding the air filter increases faster than that of its surface (Fig. 5b). This explains the increased absorbed convective heat flux (Fig. 5a), the convective heat transfer coefficient hc remaining nearly constant since it is in a forced convection regime. In the second part of the constant-speed phase, the convective heat flux is stabilized because the difference between surface and air temperatures no longer varies. Conversely, the radiative heat flux variation (Fig. 5a) reveals information about the equivalent radiative temperature of the air filter thermal surrounding, Tr. This temperature increases faster than the surface temperature before becoming almost constant in the second part of the constant-speed phase, where TS stabilizes qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðTr ¼ 4 TS4 þ ð4r =3$sÞÞ. A second typical example is the decreasing convective heat flux. Fig. 6 shows data for the cold box. Again, the convective heat flux is positive because the surrounding air is heated near the components upstream of the cold box and its temperature becomes higher than that of the cold box. On the other hand, the convective heat Heat flux (W/m²) a 1400 1200 Part 11 Part 1000 Part 22 Part 800 600 400 Overall flux flux Overall Radiative flux flux Radiative 200 0 0 Convective flux Convective 500 flux 1000 907 flux decreases because here the surface temperature increases more rapidly than that of the surrounding air zone (but remains below that temperature); this decrease now is observed in the overall heat flux with a shift due to the negative (emitted) radiative heat flux. Finally, the examples above show that the exterior air that enters the underhood compartment to cool it by convection can heat up some components; this is exemplified by the cold box, a critical component from a thermal point of view since it contains computers. This problem is a result of the actual underhood architecture, specifically the positioning of these components in the same airflow as and downstream of warmer components. An optimized thermal design may suggest placing these components further upstream or isolating them from hot airflow with deflectors. Note that in the case of decreasing convective heat flux, it is essentially the early part of the transitional curve that is problematic, since the convective heat flux can reach up to five times its asymptotic value. One can therefore imagine a mobile deflector that cuts off the air flux towards the component in the interim period only, letting pass the stabilized heat flux that might be useful for other components nearby or downstream. The deflector might even be connected to a convective flux sensor. The next section gives more details on deflectors based on the heat flux analysis discussed above. Following the analysis of heat flux behaviours above, one may also ask the following questions: is not the absorbed (positive) heat flux the source of emitted (negative) radiative heat flux? If so, does the radiative flux tend to zero if the convective flux is zero? If hot air is not directed towards the cold box, could one not avoid the radiative heat flux emitted by the cold box and absorbed by the surrounding components? In order to answer the above questions, a case in which absorbed convective heat flux is made up for by emitted radiative flux is considered. Consider the overall heat flux variation at the cold box side for the three tested thermal functioning points (Fig. 7). At the end of the constant-speed driving phase, the overall heat flux variations corresponding to the different thermal points are very close. For example, it has an overall heat flux density of 31 W/ m2 in TFP-1, of 99 W/m2 in TFP-2 and 84 W/m2 in TFP-3. One can therefore consider that the thermal situation at the cold box side is independent of thermal functioning point. In fact, however, this is not the case, as one can see by comparing the separate heat fluxes (convective and radiative) for the three thermal operating points at the same position (Fig. 8). It can be seen that the closest overall heat fluxes for the three thermal operating points are the resultant of convective and radiative heat fluxes that are significantly different. 110 90 Part Part 11 100 900 Part Part 22 80 70 60 50 Surface temperature Surface temperature Air temperature Air temperature 40 Overall flux (W/m²) b Temperature (°C) Time (s) TFP-1 TFP-2 TFP-3 700 500 300 100 30 -100 20 0 500 1000 1500 Time (s) Fig. 6. Temporal variations of (a) heat flux and (b) temperature at the air filter and its surrounding air zone in TFP-3. 0 200 400 600 800 1000 1200 1400 Time (s) Fig. 7. Overall heat flux variation at the cold box side for the three experimental thermal functioning points. M. Khaled et al. / Applied Thermal Engineering 31 (2011) 902e910 a 1400 Convective flux (W/m²) 908 1200 procedures proposed here, which are simple and easily implemented, are based on physical analysis of the aerothermal phenomena obtained from our temperature and convective and radiative heat flux measurements. The deflectors proposed here may be connected to convective heat flux sensors [23e25]. TFP-1 TFP-1 1000 TFP-2 TFP-2 800 TFP-3 TFP-3 600 400 200 0 0 200 400 600 800 1000 1200 1400 Time (s) Radiative flux (W/m²) b 50 TFP-1 TFP-1 TFP-2 TFP-2 TFP-3 TFP-3 -50 -150 -250 -350 -450 0 200 400 600 800 1000 1200 1400 Time (s) Fig. 8. Variations at the cold box side for the three experimental thermal functioning points:(a) convective flux and (b) radiative flux. For example, at the end of the constant-speed driving phase, convective heat fluxes of 378 W/m2 in TFP-1, 332 W/m2 in TFP-2 and 164 W/m2 in TFP-3 and radiative heat fluxes of 346 W/m2 in TFP-1, 233 W/m2 in TFP-2 and 80 W/m2 in TFP-3 are measured. Therefore, it is the compensations between convective and radiative heat fluxes which are imposed as a result of the same overall heat flux trends between the different thermal functioning points. To go further in the analysis, it should be noted that from one operating point to another, the more the positive convective heat flux increases, the more the absolute value of the negative radiative heat flux increases. One can assume by extrapolation that if the convective heat flux was zero, the radiative flux would be too. In this case, it is the absorbed (positive) convective heat flux on the cold box side that increases its temperature with respect to its thermal environment in such a way that the radiative heat emitted is almost equal to the convective heat absorbed. A thermal equilibrium is thus created on the cold box side that makes the overall heat flux remaining almost constant, whatever the thermal operating point. Note that this balance is largely dependent on the cold box position in the underhood compartment and does not necessarily appear for other components that receive heat by convection or absorb heat by radiation (e.g. the air filter or the cylinder head cover). 4.3.1. Passive control by static deflectors The passive version of the control uses deflectors positioned upstream of the components heated by convection. This procedure in fact optimizes the thermal management of components for one of the two typical cases of absorbed convective heat described in Section 4.2, provided that the latter are in the rear part of the vehicle underhood space. The principle of this version of the control procedure is illustrated in Fig. 9. In Fig. 9, the temperature Ta1 of air passing over the hot component 1 of temperature Th1 greater than temperature Tc of the cold component increases to a temperature Ta2 above temperature Tc. Without deflectors, the hot air induces a positive convective heat flux that increases the cold component’s temperature. The first part of the static deflector guides a portion of the hot airflow of temperature Ta2 (greater than Tc) towards the hot component 2 of temperature Th2 greater than Ta2 and Tc. The second part of the deflector, on the other hand, directs a second portion of the hot air stream to the hot component 3 of temperature Th3, also greater than Ta2 and Tc. Therefore, with the static deflector, one can transfer convective heat flux absorbed by a cold component as excesses to the convective heat flux extracted by other components at higher temperatures. It should be noted that the static deflector can have one part (or at one inclination) for a single hot component in its environment, or have more than one part (>2) for many warm components in its environment. 4.3.2. Active control by dynamic deflectors The active version of the control involves placing mobile (dynamic) deflectors upstream of components heated by convection. Looking again at the two typical cases of convective heat absorption from the previous section, the active control can be used in two different ways: one in which the deflector is closed during the transient part of the constant-speed-driving phase and open during the stabilized part, and another in which deflectors are open during the transient part and closed during the stabilized phase. Fig. 10 shows the principle of the first application, “closed-open,” where 1 and 2 designate respectively the transitional and stabilized parts of the constant-speed driving phase. Increasing or decreasing heat flux curves refer to absolute values, not algebraic values. In the absence of deflectors, the first cold component absorbs convective heat flux more in the transitional than in the stabilized period, and the second cold component loses more convective heat in the stabilized period than in the transitional one. To manage the Part 1 of the static deflector Air, T a2 Hot Comp. 1 Th1 4.3. Controlling underhood airflow by deflectors The present section describes a new approach in which static and mobile deflectors are placed in the underhood compartment in order to protect certain components from hot air circulation. These deflectors can also direct warm air passing by low-temperature components to higher-temperature components for cooling. The Air Ta1 Cold Comp. Tc Part 2 of the static deflector Hot Comp. 2 Hot Comp. 3 Th2 Th3 Fig. 9. Schematic of the static deflector principle. M. Khaled et al. / Applied Thermal Engineering 31 (2011) 902e910 Decreasing positive convective flux 1 2 Decreasing negative convective flux 1 2 Dynamic deflector Air, Ta2 Air Ta1 Hot Comp. 1 Th1 Hot Comp. 2 Cold Comp. 1 Cold Comp. 2 Tc1 Tc2 Convective fluxmeters Th2 Fig. 10. Schematic of the functioning principle of a “closed-opened” dynamic deflector, in the closed position in the transition phase. thermal situation among the different components of Fig. 10, the mobile deflector closes during the transition period and opens in the steady (stabilized) period. In transition, it directs the hot air stream of temperature Ta2 (which was heated by passing over the hot component 1 of temperature Th1 greater than Ta1) greater than Tc1 and Tc2 towards another hot component 2 of temperature Th2 still greater than Tc1, Tc2 and Ta2. At the beginning of stabilization, the deflector opens to let in the air needed for cooling cold component 2, which is not now critical for cold component 1. Note that this application, “closed-open,” can be used only for a cold component that absorbs convective heat flux (positive) and is located upstream of other components that themselves evacuate convective heat flux (negative) increasing in absolute value with time. The mobile deflector is controlled by two heat flux sensors on the surfaces of both cold components in front of the air stream. The principle of the second application, “open-closed,” is shown in Fig. 11. In the absence of the deflector, the first cold component receives more convective heat flux in the stabilized phase than in the transition regime, and the second cold component loses more convective heat in the transition than in the stabilized regime. To manage the thermal situation of the different components of Fig. 11, the mobile deflector opens during the transition phase and closes during the stabilized phase. In the transition part, the open deflector allows the passage of the air necessary to cool cold component 2 even if the first cold component absorbs convective flux, provided that this latter is very small compared to what is evacuated by the second cold component. At the transition phase, the deflector closes in order to deflect the hot air, which heats the first cold component and is not very efficient in cooling the second cold component, towards hot component 2 of temperature above those of the surrounding air and the two cold components. It should be noted that this “open-closed” application Fig. 11. Schematic of the functioning principle of an “opened e closed” dynamic deflector, in the open position in the transition phase. 909 can be used only for a cold component that absorbs increased convective flux (positive) and is upstream of other components that themselves lose convective heat flux (negative) decreasing in absolute value with time. In conclusion, one can optimize the underhood aerothermal management without changing the architecture or the position of vehicle underhood components. The technical interest of this control procedure resides in the fact that it optimizes the cooling of underhood components that either are insufficiently cooled or are heated by convection due to placement in the underhood compartment. The economic advantages of such optimized underhood aerothermal management (with or without active systems) are indirect: - reduction of thermal problems, thus reducing warranty costs and expenses related to possible emergency fixes; - better interaction with external aerodynamics: better management of the underhood airflow reduces the area of air inlets and thus aerodynamic drag (fuel consumption and carbon emission issues). 5. Concluding remarks The present paper gives a physical analysis of particular underhood aerothermal behaviors, especially convective heat flux absorption, and presents a new optimization procedure [16] based on this physical analysis that entails redistribution of cooling airflow in the vehicle underhood compartment. During the constant-speed driving phase, for the different thermal functioning points investigated, typical exponential trends are observed in the temperature variation of all components, air zones, and engine parameters. Exponential trends are also found for the temporal variation of overall, convective, and radiative heat fluxes for all components investigated. For the overall heat flux variation (as well as for the convective and radiative parts), two categories of components are distinguished: category 1, those components that absorb or lose overall heat fluxes that increase in absolute value with time, and category 2, those components that absorb or lose overall heat fluxes that decrease in absolute value with time. It has been shown by some examples that outdoor air entering the underhood compartment to cool it by convection can in fact heat up some components, such as the cold box, which is a critical component from a thermal point of view since it contains the car’s computers. This problem results from the today’s underhood architecture, specifically the positioning of these components downstream of and in the same air stream as warmer components. Passive thermal management can be achieved by placing these components in areas further upstream or by isolating them from the hot airflow using deflectors. Note that for decreasing convective heat flux, it is essentially the early part of the transitional curve that causes the problem, since the convective heat flux can be as much as five times its asymptotic value. One can also implement active control by using mobile deflectors to cut off the airflow towards the components in the transition period and to let the air pass in the stabilized heat flux phase. This air can be useful for cooling other components nearby or downstream that are at higher temperatures. The deflector can be controlled by convective heat flux sensors. Finally, a new underhood control procedure is presented in which static and mobile deflectors are implemented in the car underhood space in order to protect certain components from hot air circulation. These deflectors can also direct the warm air passing over low-temperature components towards higher-temperature components. The new procedures proposed here are simple and easy to implement in the car underhood compartment and are 910 M. Khaled et al. / Applied Thermal Engineering 31 (2011) 902e910 based essentially on the physical analyses of temperature and separate convective and radiative heat flux measurements. For example, a dynamic “open-closed” deflector lets air pass over (or through) two components in the transition phase but closes in the stabilized phase, when the first component in the airflow absorbs increasing positive convective flux and the second absorbs decreasing (in absolute value) negative convective flux. References [1] E.P. Weidmann, J. Widemann, T. Binner, H. Reister. Underhood temperature analysis in case of natural convection, SAE Paper 2005-01-2045; 2005. [2] E. Fournier, T. Bayne, Underhood temperature measurements of four vehicles. Motor Vehicle Fire Research Institute, by Biokinetics and Associates, Ltd., 2004, Report R04e13. [3] E. Fournier, T. Bayne, Assessment of thermocouple attachment methods for measuring vehicle exhaust temperature. Motor Vehicle Fire Research Institute, by Biokinetics and Associates, Ltd., 2006, Report R06e23b for MVFRI. [4] E. Fournier, T. Bayne. Underhood temperature measurements, SAE Paper 2007-01-1393; 2007. [5] M.R. Jones, D.W. Fletcher. Thermal performance prediction of front-end heatexchange modules, SAE Paper 2001-01-1765; 2001. [6] N. Francois. Using CFD for heat exchanger development and thermal management, Valeo Engine Cooling, European Automotive CFD Conference EACC, Frankfurt, Germany; June 29e302005. [7] G. Xiao. Investigation of conjugate heat transfer for vehicle underbody, SAE Paper 2008-01-1819; 2008. [8] W. Ding, J. Williams, D. Karanth, S.D. Sovani. CFD application in automotive front-end design, SAE Paper 2006-01-0337; 2006. [9] F. Fortunato, F. Damiano, L. P. Matteo Oliva. Underhood cooling simulation for development of new vehicles, SAE Paper 2005-01-2046; 2005. [10] S. Kini, R. Thoms. Multi-domain meshes for automobile underhood applications, SAE Paper 2009-01-1149. [11] O. Bailly, X. Baby, E. Fares, H. de Portzamparc. Piv and numerical correlations on the laguna II underhood flow field, Renault, Congrès de la société des ingénieurs de l'Automobile SIA, Lyon, France, October 26e27; 2005. [12] Z. Yang, J. Bozeman, F.Z. Shen, CFD for flow rate and air re-circulation at vehicle idle conditions. General Motors Corporation, 2004, SAE Paper 2004-01-0053. [13] Jerhamre A, Jönson A. Development and validation of coolant temperature and cooling airflow CFD simulations at Volvo cars, Volvo Car Corporation, SAE Paper 2004-01-0051; 2004. [14] Weidmann EP, Reister H, Binner T. Experimental and numerical investigations of thermal soak, SAE Paper 2008-01-0396; 2008. [15] Hormann T, Lechner B, Puntigam W, Moshammer T, Aimbauer R. Numerical and experimental investigation of flow and temperature fields around automotive cooling systems, SAE Paper 2005-01-2006; 2005. [16] Khaled M, Harambat F, Peerhossaini H. “Véhicule Automobile à Ecoulement d'Air de Refroidissement optimisé”, French patent application filed January 20, Patent no1050365; 2009. [17] Koller E. Machines thermiques, L'Usine Nouvelle e Série Mécanique et Matériaux DUONOD. [18] J.L. Lumley, Engines: An Introduction. Cambridge University Press, 1999. [19] M. Khaled, F. Harambat, H. Peerhossaini, A quantitative method for the assessment of car inclination effects on thermal management of the underhood compartment, Journal of Thermal Science and Engineering Applications 014501 (2009) 1e5 ASME 1. [20] M. Khaled, F. Harambat, H. Peerhossaini, Temperature and heat flux behavior of complex flows in car underhood compartment, Heat Transfer Engineering 31 (13) (2010) 1e11. [21] Khaled M, Harambat F, Peerhossaini H. Effects of car inclination on airflow and aerothermal behavior in the underhood compartment. In: Proceedings of the ASME Fluids Engineering Conference, July 30eAugust 2, San Diego, CA, USA, FEDSM2008-55093; 2008. [22] M. Khaled, F. Harambat, H. Peerhossaini, Underhood thermal management: temperature and flux measurements and physical analysis, Applied Thermal Engineering 30 (2010) 590e598. [23] RC01 radiation/convection (RADCON) Heat flux sensor, HUKSEFLUXÔ THERMAL SENSORS brochure. [24] M. Khaled, B. Garnier, F. Harambat, H. Peerhossaini, A new method for simultaneous measurement of convective and radiative heat flux in car underhood applications, Measurement Science and Technology Journal 21 (2010) 025903 (10 pp). [25] M. Khaled, F. Harambat. Dispositif de Mesure de Flux Thermique et Méthode de Séparation des Flux Radiatif et Convectif, French patent application filed on February 10, Patent no 0950799; 2009.