- Email: [email protected]
Aerodynamics of ARTeC's PEC 2011 EMo-C Car
Available online at www.sciencedirect.com
Procedia Engineering 41 (2012) 1775 – 1780
International Symposium on Robotics and Intelligent Sensors 2012 (IRIS 2012)
Aerodynamics of ARTeC’s PEC 2011 EMo-C Car Rizal E. M. Nasira*, Firdaus Mohamada, Ramlan Kasirana, M. Shahriman Adenana, M. Faizal Mohameda, M. Hanif Mata, Amir R. A. Ghania ARTeC, Faculty of Mechanical Engineering, University Teknologi MARA, Shah Alam 40450, Malaysia *Tel: +60-3-55436207, Fax: +60-3-55435160, Email: [email protected]
Abstract A formula-type car design with fendered rear tyres is chosen as the ARTeC’s design for 2011 Perodua Eco-Challenge. This paper aims at investigating aerodynamics of ARTeC’s EMo-C car, measuring drag coefficient in particular and observing airflow around the body. There are two means of measuring the drag, the first is by simulating the air flow via computational fluid dynamics (CFD) suite, and the second is by using wind tunnel experiment. Observation on air flow around the body is also highlighted here with emphasis on the results from CFD simulation. From computational simulation (CFD) and wind tunnel experiment, the Emo-C car’s drag coefficient is 0.42 to 0.48, respectively, the difference between the two is around 12.5 percent. The former has lower value than the latter due to inviscid flow modelling which disregard skin friction, hence the existence of boundary layer on the body panel (viscous flow). Tyres accounts for 25.7 percent of profile drag coefficient. Without tyres, the car experiences down force but opposite effect (lift) is observed when tyres are included. At average race speed of 40 km/h, the aerodynamic drag only accounts for no more than 20 percent of the overall power required from the engine and slightly less than six percent of the estimated total fuel consumption of the car.
© 2012 The Authors. Published by Elsevier Ltd. Selection and/or peer-review under responsibility of the Centre of Humanoid Robots and Bio-Sensor (HuRoBs), Faculty of Mechanical Engineering, Universiti Teknologi MARA. Keywords: Aerodynamics, Automotive, Computational Fluid Dynamics
Nomenclature v A CD CL D U G
airspeed (m/s) car frontal area (m) drag coefficient lift coefficient drag force (N) air density (kg/m3) boundary layer thicknesses(m)
1. Introduction Perodua Eco-Challenge (PEC) 2011 is a competition for the longest distance over a specific race track using a specific amount of fuel provided. In 2011, Perodua Sdn Bhd. only supplies engine and transmission and participants are expected to design and built a single-seated car. The winner of the race is the car with the lowest fuel consumption per litre of fuel. Various car concepts and sketches have been produced by ARTeC to suit to the PEC 2011 regulations [1]. The main issue is that whether it shall be an urban type or a formula type. The latter suits better with the regulations plus the fact that formulatype car has low weight and small frontal area. However, a car of this type may lose some practicality (i.e. ergonomics) of the former. In contrast, the urban-type car requires more weight (roof, ventilation etc.) but has better aerodynamic
1877-7058 © 2012 Published by Elsevier Ltd. doi:10.1016/j.proeng.2012.07.382
1776
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780
characteristics for having tyres enclosed into their bodyshell. Fig.1 shows drag coefficients of various mini cars, sport cars and average for various cars from data extracted from [2], [3] and [4]. Team FKM-UiTM has decided that the car shape shall be close to formula-type car but with rear wheels enclosed into rear bodyshell. The front wheel is left open for simplicity and weight reasons. While chassis design has permitted comfortable room for driver’s cockpit, its resulting frontal area is small enough that it may represent less than 40% of the frontal area of mini or supermini hatchback although many would argue that a formula type car has larger drag coefficient. This concept, as we call it, is known as blended wheel-body concept (Fig. 2) – a sport car which is a hybrid of formula-type car and urban (mini) car. CD of Mini Cars
CD of Sport Cars CD of All Types (Average)
Jaguar XJ-S Por 911 Carrera Cabrio Porsche 928 S Mitsu Starion Turbo Honda Prelude Porsche 911 Carrera VW Scirocco GTX Toyota Celica Supra 2.8i Mazda RX-7 Audi Coupe GT 5E Honda CRX Coupe Renault Fuego GTX Opel Monza GSE Mazda 626 Coupe Nissan 300 ZX Porsche 944 Turbo Porsche 924
Citroen 2 CV VW Beetle Daihatsu Charade TS Suziki Alto Fiat Panda Ford Fiesta Mitsu Colt 1200 GL VW Polo Coupe Citroen Visa 17 RD Citroen LNA Opel Corsa TR Honda Civic 1.2 Renault 5 GTL Peugeot 205 GL Fiat Uno ES
0
0.2
0.4
0.6
Average Formula Cars Average Sport cars Average Full-sized cars Average Medium-sized cars Average Mini cars 0
0.1
0.2
CD
0.3
0.4
CD
0.00 0.20 0.40 0.60 0.80 1.00 CD
Fig.1. Drag coefficients of mini cars (left), sport cars (centre) and average for all types (right)
The EMo-C concept’s size and shape has evolved from the large-sized urban-formula hybrid (nearly four metres long) to the size approximately 3.1 metres long with combination of flat, rectangular and angular shape. A car with good aerodynamic characteristics, for example low drag, shall be round, parabolic or curvature in term of profile; however, in a race for fuel efficiency at low speed that is suited for small go-cart track, aerodynamics may not play major roles in overall power required. However, the competition has become very competitive that even low-speed aerodynamics is accounted into the fuel consumption equation. A small advantage in having lower drag than competitors may determine the success of winning the longest range for given fuel. 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
(a)
NOSE
F BODY (UPPER)
F BODY (LOWER)
TAIL
R BODY (UPPER)
FRONT W HEEL
FRONT WHEEL
REAR W HEEL
REAR W HEEL
ENGINE HOOD
F WHEEL ARC
F FENDER
F FENDER (LOW)
R FENDER
R FENDER (LOW)
R WHEEL ARC
REAR W HEEL ARC LOWER EXTENSION
LOWER LERX
chassis point
COCKPIT
SIDE WALL
CANOPY (L)
LERX
REAR SIDE WALL
Series2
3700
3600
3500
3400
3300
3200
3100
3000
2900
2800
2700
2600
2500
2400
2300
2200
2100
2000
1900
1800
1700
1600
1500
1400
1300
900
1200
800
1100
700
-200
1000
600
500
400
0
300
200
100
0 -100
(b)
Fig. 1. ARTeC PEC 2011 EMo-C car concept: (a) side view plots using spreadsheet software and (b) a three-quarter view sketch of the car
This paper aims at investigating aerodynamics of ARTeC’s EMo-C car, measuring drag coefficient in particular and observing airflow around the body. There are two means of measuring the drag, the first is by simulating the air flow via computational fluid dynamics (CFD) suite, and the second is by using wind tunnel experiment. Observation on air flow around the body is also highlighted here with emphasis on the results from CFD simulation. 2. Investigation Methods The car aerodynamics is investigated via computational fluid dynamics (CFD) simulation and wind tunnel experiment (WT). CFD simulation is conducted by using FLUENT software at design speed of 50 km/h (13.9 m/s). A 1/10th scale wind tunnel model is tested at 10 – 30 m/s airspeed to acquire drag coefficients and to visualize airflow around the body.
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780
1777
Similarity between the CFD simulation and wind tunnel experiment is not based on Reynold’s number but only Mach number. 3. Computational Fluid Dynamics (CFD) The efficiency and the financial aspect make CFD a better solution in analyzing the aerodynamics behaviour of vehicle [3]. Flexibility and ease of conducting visualization of air flow around the body, increasing accuracy of computational solution, new turbulence models and the increasing computing power are other aspects which show the advantages of CFD. In this case, GAMBIT software is used as pre-processing tool for modelling and discretization of control volume, and FLUENT is used as a solver and post-processing tool [8]. This software is proven to be a very flexible that can create a very efficient analysis [4]. Car model from CATIA suite is imported in GAMBIT software via .igs format (Fig.3a). The imported model must be converted into solid in CATIA. Since this simulation deals with external flow of vehicle, flow domain must be created. GAMBIT also includes several features that allow users to control the mesh quality, one of which is the application of size functions. For example, size functions can be used to specify the rate at which volume mesh elements change in size in proximity to a specified boundary. The elements were appropriately concentrated near the car, growing in size as the outer domain limits are reached as shown in Fig.3b.
Velocity Inlet
Wall (Bottom)
(a)
Pressure Outlet
(b)
Fig. 3. (a) CAD model and (b) meshed volume using sizing function approach
Unstructured meshing scheme was used with tetrahedral cells. The number of cells is about 200,000. Before the meshed volume is analyzed in FLUENT, boundary conditions must be set-up in GAMBIT. Boundary-type specifications define the physical and operational characteristics of the model at those topological entries that represent model boundaries [7]. Table 1 shows the boundary type for the meshed volume. Table 1: Boundary Conditions Face
Boundary type
In front of the car
Velocity Inlet
Zone name/symbol Vi
Back of the car
Pressure Outlet
Po
Side and above the car
Symmetry
Sym
Below the car
Wall
Bottom
Car
Wall
Car
The meshed volume with boundary conditions is exported to the FLUENT software. In this case, the model is assumed to be inviscid and incompressible. Invisid flow is used instead of turbulent model because it is important at this stage to know the drag coefficient of the car purely based on its shape without skin friction (viscous flow) coming into picture. The operating pressure was set to be 1 atmosphere (101.325 kPa). The material used was air. Due to the fact that the flow is incompressiEOH WKH GHQVLW\ LV FRQVLGHUHG WR EH FRQVWDQW 7KH GHIDXOW YDOXHV IRU GHQVLW\ ȡ NJP3) and viscosity ȝ [-5kg/ms) were used. The velocity inlet was the only boundary condition that required the specification of the physical attributes for the flow. The velocity inlet boundary condition is only applicable to incompressible flow. The velocity profile is assumed to be uniform by default, and when the velocity distribution is set, the static pressure is automatically adjusted. In order to obtain force and moment coefficient, reference area and length must be inserted in reference value monitor.
1778
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780 Table 2: CFD simulation numerical result Velocity = 50km/h, Temperature = 288.16 K, Density , U = 1.225 kg/m3 , Area frontal = 0.0119 m2 Force coeffcients
Without Tyre
With Tyre
Drag Coefficient, CD
0.312
0.42
Lift Coefficient, CL
-0.053
0.339
Two conditions of car have been simulated to obtain the aerodynamic data. The first condition is car without tyre. From the simulation result, model without tyre will give low drag force. The drag coefficient based on frontal area is 0.312 (Table 2). This sport car also produces downforce (negative CL). From the visualization, flow is smoothly attached around the frontal body of the car. However, some turbulence flows also detected around the cockpit and massive vortices is found on top and behind the engine bonnet of the car (rear part of the car) (Fig. 4a.). The second condition is car with tyre. From the result, drag force obtained is higher than the car without tyre. The drag coefficient calculated is 0.42 (Table 2). The hike in drag value shows that tyres and wheels accounts for 25.7 percent of overall profile drag coefficient (inviscid). The simulation with tyre shows lift (positive CL) instead of downforce is generated. Again, the flow attaches smoothly around the frontal body is smooth and unaffected by the front tyres (Fig. 4b.). However, front tyres spoil the air around the side pods making the airflow at the rear part of the body even worse than the car without tyre. Addition of tyre profiles increases frontal are and creating turbulence beginning at the mid-ship section of the car, thus increasing drag. Similar observations are found in ref. [3].
(a)
(b) Fig. 4. air flow path lines of EMo-C car (a) without tyres (b) with tyres
4. Wind Tunnel (WT) Experiment
(a)
(b) Fig. 5. (a) UiTM LST-1 wind tunnel and (b) EMo-C car wind tunnel model
Low speed wind tunnel (Figure 5) is used to determine the drag force and lift force. The cross sectional area test is 50 cm x 50 cm. This wind tunnel had three-component balance with computerized data acquisition system. The maximum speed of this wind tunnel is 50m/s but is it suggested to not exceed than 40m/s when conducting the experiment. A 1/10th scale wind tunnel model is tested at 10 – 30 m/s airspeed to acquire drag coefficients and to visualize airflow around the body. The model has been fabricated by using rapid prototyping machine. Test were conducted at geometrically similar with 1/10th scale wind tunnel model. It can provide dynamically flow situations with prototype [6]. Table 3 shows the result of average
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780
1779
drag coefficient which is at 0.48. The tunnel is unable to measure lift coefficient with model configured (shown in Fig. 5b.) because it uses three-component balance system. Table 3: Wind tunnel experiment conditions and results Air density (kg/m3 )
1.17 - 1.18
Scale (mm)
1: 10
Speed (m/s)
10 - 30
CD
0.48 (average)
5. Analysis on EMo-C’s Drag Coefficient Table 4 shows comparisons between CFD and wind tunnel results. It is expected that CFD yields lower value because the simulations only take account for inviscid flow where as in wind tunnel experiment, the flow is definitely viscous. The drag coefficient yields drag force of 23.8 (CFD) and 27.4 (WT) on actual-sized car which has frontal area of 2.71 m2. Below is an example of how drag is calculated when the car is travelling at 40 km/h (11.1 km/h); D Where
23.8 N ,
, Table 4 : Comparison between wind tunnel experiment and CFD result Sref (frontal)
0.76 m2
Air density
1.225 kg/m3
Airspeed
11.1 m/s
CD
CFD
WT
% diff
0.42
0.48
-12.5%
Drag (N)
23.8
27.4
Power Required (W)
264.2
304.1
Meanwhile, power required to overcome aerodynamic drag from this car is around 260.0 to 305.0 Watt or 0.35 to 0.41 horsepower. This is lower than power required to overcome rolling resistance from tyre-road friction, engine’s thermodynamic efficiency and transmission’s mechanical loss (Table 5). Approximately two horsepower is needed to propel EMo-C car at 40 km/h and aerodynamics only accounts for 19.6 percent of overall power that needs to be delivered by the engine. Given that thermodynamic efficiency of the engine is estimated at 30 percent then aerodynamic drag only accounts for nearly six percent of the car’s fuel consumption. Table 5: Estimated aerodynamic drag power and percentage of fuel consumption Parameter
Value
car speed
11.1 m/s
mass
460 kg
Weight Rolling friction coefficient rolling resistance rolling power mechanical loss (transmission) rolling power required Aerodynamic power required Total power required Percentage of aerodynamic power
including driver
4512.6 N 0.02
standard road tyres and asphalt/concrete road surface
90.3 N 1001.8 W
1.34 hp
20 % 1252.2 W
1.68 hp
304.9 W
0.41 hp
1557.1 W
2.09 hp
19.6 %
of total power required
1780
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780 Thermodynamic efficiency
30 %
Fuel consumption due to aerodynamic drag
5.9 %
of total fuel consumption
6. Concluding Remarks From computational simulation (CFD) and wind tunnel experiment, the Emo-C car’s drag coefficient is 0.42 to 0.48, respectively, the difference between the two is around 12.5 percent. The former has lower value than the latter due to inviscid flow modelling which disregard skin friction, hence the existence of boundary layer on the body panel (viscous flow). Tyres accounts for 25.7 percent of profile drag coefficient. Without tyres, the car experiences down force but opposite effect (lift) is observed when tyres are included. At average race speed of 40 km/h, the aerodynamic drag only accounts for no more than 20 percent of the overall power required from the engine and slightly less than six percent of the estimated total fuel consumption of the car. Emo-C’s drag coefficient is similar to the old mini/hatchback cars such as Citroen 2CV, Volkswagen Beetle and Daihatsu Charade but this is not exactly poor because a formula cars usually has drag coefficients ranging 0.7 until 1.1. Emo-C car was built in April 2011 and entered the competition three months later. Figure 6 shows the completed car. A quick-fix was introduced with intention to reduce the drag further down based on lessons learned from analyses in this paper. A leading edge extension (LEX) on each side of the air cooling/intake pod is incorporated to prevent air spilling from the front end of the pod to spoil the airflow on top of the rear wheel fender. A proper study on the effect of this LEX to the drag of Emo-C car is needed to prove the claim.
Fig. 6: Completed EMo-C car
7. Acknowledgment We would like to express our gratitude to the Faculty of Mechanical Engineering, Universiti Teknologi Mara for the support in terms of funding and human resources. 8. References [1] Perodua Eco-Challenge 2011 Rules and Regulations, Perodua Sdn. Bhd., January 2011. [2] Automotive Analyses Homepage. www.mayffco.com/analyses.htm. [accessed on March 2011]. [3] Kellar, Pearse and Savill. Formula 1 car wheel aerodynamics. Sports Engineering, 2, November 1999, pp. 203-212. [4] Bienz C., Larsson T., Sato T and Ullbrand B."In Front oftThe Grid- CFD at SAUBER PETRONAS F1 Leading the Aerodynamic Development'. presented at the 1st European Automotive CFD Conference (EACC 2003), 25-26 June, Bingen, Germany 2003. [5] J. Katz. Aerodynamics of Race Cars. Annual Review of Fluid Mechanics. Vol 38, January 2006, pp. 27-63. [6] Manan Desai, S.A. Channiwala and H.J.Nagarsheth." Experimental and Computational Aerodynamic Investigations of a Car", WSEAS TRANSACTIONS on FLUID MECHANICS, Issue 4, Volume 3, October 2008. [7] 'DPMDQRYLüDarko, Kozak, DražDQ,YDQGLüäHOMNR.RNDQRYLüDQGMato, " Car Design as A New Conceptual Solution and Cfdanalysis In Purpose of Improving Aerodynamics", Josip Juraj Strossmayer University of Osijek, Mechanical Engineering Faculty in Slavonski Brod, Croatia [8] Scott Wordley and Jeff Saunders," Aerodynamics for Formula SAE: A Numerical, Wind Tunnel and On-Track Study", SAE p Paper Number 2006-010808
Procedia Engineering 41 (2012) 1775 – 1780
International Symposium on Robotics and Intelligent Sensors 2012 (IRIS 2012)
Aerodynamics of ARTeC’s PEC 2011 EMo-C Car Rizal E. M. Nasira*, Firdaus Mohamada, Ramlan Kasirana, M. Shahriman Adenana, M. Faizal Mohameda, M. Hanif Mata, Amir R. A. Ghania ARTeC, Faculty of Mechanical Engineering, University Teknologi MARA, Shah Alam 40450, Malaysia *Tel: +60-3-55436207, Fax: +60-3-55435160, Email: [email protected]
Abstract A formula-type car design with fendered rear tyres is chosen as the ARTeC’s design for 2011 Perodua Eco-Challenge. This paper aims at investigating aerodynamics of ARTeC’s EMo-C car, measuring drag coefficient in particular and observing airflow around the body. There are two means of measuring the drag, the first is by simulating the air flow via computational fluid dynamics (CFD) suite, and the second is by using wind tunnel experiment. Observation on air flow around the body is also highlighted here with emphasis on the results from CFD simulation. From computational simulation (CFD) and wind tunnel experiment, the Emo-C car’s drag coefficient is 0.42 to 0.48, respectively, the difference between the two is around 12.5 percent. The former has lower value than the latter due to inviscid flow modelling which disregard skin friction, hence the existence of boundary layer on the body panel (viscous flow). Tyres accounts for 25.7 percent of profile drag coefficient. Without tyres, the car experiences down force but opposite effect (lift) is observed when tyres are included. At average race speed of 40 km/h, the aerodynamic drag only accounts for no more than 20 percent of the overall power required from the engine and slightly less than six percent of the estimated total fuel consumption of the car.
© 2012 The Authors. Published by Elsevier Ltd. Selection and/or peer-review under responsibility of the Centre of Humanoid Robots and Bio-Sensor (HuRoBs), Faculty of Mechanical Engineering, Universiti Teknologi MARA. Keywords: Aerodynamics, Automotive, Computational Fluid Dynamics
Nomenclature v A CD CL D U G
airspeed (m/s) car frontal area (m) drag coefficient lift coefficient drag force (N) air density (kg/m3) boundary layer thicknesses(m)
1. Introduction Perodua Eco-Challenge (PEC) 2011 is a competition for the longest distance over a specific race track using a specific amount of fuel provided. In 2011, Perodua Sdn Bhd. only supplies engine and transmission and participants are expected to design and built a single-seated car. The winner of the race is the car with the lowest fuel consumption per litre of fuel. Various car concepts and sketches have been produced by ARTeC to suit to the PEC 2011 regulations [1]. The main issue is that whether it shall be an urban type or a formula type. The latter suits better with the regulations plus the fact that formulatype car has low weight and small frontal area. However, a car of this type may lose some practicality (i.e. ergonomics) of the former. In contrast, the urban-type car requires more weight (roof, ventilation etc.) but has better aerodynamic
1877-7058 © 2012 Published by Elsevier Ltd. doi:10.1016/j.proeng.2012.07.382
1776
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780
characteristics for having tyres enclosed into their bodyshell. Fig.1 shows drag coefficients of various mini cars, sport cars and average for various cars from data extracted from [2], [3] and [4]. Team FKM-UiTM has decided that the car shape shall be close to formula-type car but with rear wheels enclosed into rear bodyshell. The front wheel is left open for simplicity and weight reasons. While chassis design has permitted comfortable room for driver’s cockpit, its resulting frontal area is small enough that it may represent less than 40% of the frontal area of mini or supermini hatchback although many would argue that a formula type car has larger drag coefficient. This concept, as we call it, is known as blended wheel-body concept (Fig. 2) – a sport car which is a hybrid of formula-type car and urban (mini) car. CD of Mini Cars
CD of Sport Cars CD of All Types (Average)
Jaguar XJ-S Por 911 Carrera Cabrio Porsche 928 S Mitsu Starion Turbo Honda Prelude Porsche 911 Carrera VW Scirocco GTX Toyota Celica Supra 2.8i Mazda RX-7 Audi Coupe GT 5E Honda CRX Coupe Renault Fuego GTX Opel Monza GSE Mazda 626 Coupe Nissan 300 ZX Porsche 944 Turbo Porsche 924
Citroen 2 CV VW Beetle Daihatsu Charade TS Suziki Alto Fiat Panda Ford Fiesta Mitsu Colt 1200 GL VW Polo Coupe Citroen Visa 17 RD Citroen LNA Opel Corsa TR Honda Civic 1.2 Renault 5 GTL Peugeot 205 GL Fiat Uno ES
0
0.2
0.4
0.6
Average Formula Cars Average Sport cars Average Full-sized cars Average Medium-sized cars Average Mini cars 0
0.1
0.2
CD
0.3
0.4
CD
0.00 0.20 0.40 0.60 0.80 1.00 CD
Fig.1. Drag coefficients of mini cars (left), sport cars (centre) and average for all types (right)
The EMo-C concept’s size and shape has evolved from the large-sized urban-formula hybrid (nearly four metres long) to the size approximately 3.1 metres long with combination of flat, rectangular and angular shape. A car with good aerodynamic characteristics, for example low drag, shall be round, parabolic or curvature in term of profile; however, in a race for fuel efficiency at low speed that is suited for small go-cart track, aerodynamics may not play major roles in overall power required. However, the competition has become very competitive that even low-speed aerodynamics is accounted into the fuel consumption equation. A small advantage in having lower drag than competitors may determine the success of winning the longest range for given fuel. 1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100
(a)
NOSE
F BODY (UPPER)
F BODY (LOWER)
TAIL
R BODY (UPPER)
FRONT W HEEL
FRONT WHEEL
REAR W HEEL
REAR W HEEL
ENGINE HOOD
F WHEEL ARC
F FENDER
F FENDER (LOW)
R FENDER
R FENDER (LOW)
R WHEEL ARC
REAR W HEEL ARC LOWER EXTENSION
LOWER LERX
chassis point
COCKPIT
SIDE WALL
CANOPY (L)
LERX
REAR SIDE WALL
Series2
3700
3600
3500
3400
3300
3200
3100
3000
2900
2800
2700
2600
2500
2400
2300
2200
2100
2000
1900
1800
1700
1600
1500
1400
1300
900
1200
800
1100
700
-200
1000
600
500
400
0
300
200
100
0 -100
(b)
Fig. 1. ARTeC PEC 2011 EMo-C car concept: (a) side view plots using spreadsheet software and (b) a three-quarter view sketch of the car
This paper aims at investigating aerodynamics of ARTeC’s EMo-C car, measuring drag coefficient in particular and observing airflow around the body. There are two means of measuring the drag, the first is by simulating the air flow via computational fluid dynamics (CFD) suite, and the second is by using wind tunnel experiment. Observation on air flow around the body is also highlighted here with emphasis on the results from CFD simulation. 2. Investigation Methods The car aerodynamics is investigated via computational fluid dynamics (CFD) simulation and wind tunnel experiment (WT). CFD simulation is conducted by using FLUENT software at design speed of 50 km/h (13.9 m/s). A 1/10th scale wind tunnel model is tested at 10 – 30 m/s airspeed to acquire drag coefficients and to visualize airflow around the body.
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780
1777
Similarity between the CFD simulation and wind tunnel experiment is not based on Reynold’s number but only Mach number. 3. Computational Fluid Dynamics (CFD) The efficiency and the financial aspect make CFD a better solution in analyzing the aerodynamics behaviour of vehicle [3]. Flexibility and ease of conducting visualization of air flow around the body, increasing accuracy of computational solution, new turbulence models and the increasing computing power are other aspects which show the advantages of CFD. In this case, GAMBIT software is used as pre-processing tool for modelling and discretization of control volume, and FLUENT is used as a solver and post-processing tool [8]. This software is proven to be a very flexible that can create a very efficient analysis [4]. Car model from CATIA suite is imported in GAMBIT software via .igs format (Fig.3a). The imported model must be converted into solid in CATIA. Since this simulation deals with external flow of vehicle, flow domain must be created. GAMBIT also includes several features that allow users to control the mesh quality, one of which is the application of size functions. For example, size functions can be used to specify the rate at which volume mesh elements change in size in proximity to a specified boundary. The elements were appropriately concentrated near the car, growing in size as the outer domain limits are reached as shown in Fig.3b.
Velocity Inlet
Wall (Bottom)
(a)
Pressure Outlet
(b)
Fig. 3. (a) CAD model and (b) meshed volume using sizing function approach
Unstructured meshing scheme was used with tetrahedral cells. The number of cells is about 200,000. Before the meshed volume is analyzed in FLUENT, boundary conditions must be set-up in GAMBIT. Boundary-type specifications define the physical and operational characteristics of the model at those topological entries that represent model boundaries [7]. Table 1 shows the boundary type for the meshed volume. Table 1: Boundary Conditions Face
Boundary type
In front of the car
Velocity Inlet
Zone name/symbol Vi
Back of the car
Pressure Outlet
Po
Side and above the car
Symmetry
Sym
Below the car
Wall
Bottom
Car
Wall
Car
The meshed volume with boundary conditions is exported to the FLUENT software. In this case, the model is assumed to be inviscid and incompressible. Invisid flow is used instead of turbulent model because it is important at this stage to know the drag coefficient of the car purely based on its shape without skin friction (viscous flow) coming into picture. The operating pressure was set to be 1 atmosphere (101.325 kPa). The material used was air. Due to the fact that the flow is incompressiEOH WKH GHQVLW\ LV FRQVLGHUHG WR EH FRQVWDQW 7KH GHIDXOW YDOXHV IRU GHQVLW\ ȡ NJP3) and viscosity ȝ [-5kg/ms) were used. The velocity inlet was the only boundary condition that required the specification of the physical attributes for the flow. The velocity inlet boundary condition is only applicable to incompressible flow. The velocity profile is assumed to be uniform by default, and when the velocity distribution is set, the static pressure is automatically adjusted. In order to obtain force and moment coefficient, reference area and length must be inserted in reference value monitor.
1778
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780 Table 2: CFD simulation numerical result Velocity = 50km/h, Temperature = 288.16 K, Density , U = 1.225 kg/m3 , Area frontal = 0.0119 m2 Force coeffcients
Without Tyre
With Tyre
Drag Coefficient, CD
0.312
0.42
Lift Coefficient, CL
-0.053
0.339
Two conditions of car have been simulated to obtain the aerodynamic data. The first condition is car without tyre. From the simulation result, model without tyre will give low drag force. The drag coefficient based on frontal area is 0.312 (Table 2). This sport car also produces downforce (negative CL). From the visualization, flow is smoothly attached around the frontal body of the car. However, some turbulence flows also detected around the cockpit and massive vortices is found on top and behind the engine bonnet of the car (rear part of the car) (Fig. 4a.). The second condition is car with tyre. From the result, drag force obtained is higher than the car without tyre. The drag coefficient calculated is 0.42 (Table 2). The hike in drag value shows that tyres and wheels accounts for 25.7 percent of overall profile drag coefficient (inviscid). The simulation with tyre shows lift (positive CL) instead of downforce is generated. Again, the flow attaches smoothly around the frontal body is smooth and unaffected by the front tyres (Fig. 4b.). However, front tyres spoil the air around the side pods making the airflow at the rear part of the body even worse than the car without tyre. Addition of tyre profiles increases frontal are and creating turbulence beginning at the mid-ship section of the car, thus increasing drag. Similar observations are found in ref. [3].
(a)
(b) Fig. 4. air flow path lines of EMo-C car (a) without tyres (b) with tyres
4. Wind Tunnel (WT) Experiment
(a)
(b) Fig. 5. (a) UiTM LST-1 wind tunnel and (b) EMo-C car wind tunnel model
Low speed wind tunnel (Figure 5) is used to determine the drag force and lift force. The cross sectional area test is 50 cm x 50 cm. This wind tunnel had three-component balance with computerized data acquisition system. The maximum speed of this wind tunnel is 50m/s but is it suggested to not exceed than 40m/s when conducting the experiment. A 1/10th scale wind tunnel model is tested at 10 – 30 m/s airspeed to acquire drag coefficients and to visualize airflow around the body. The model has been fabricated by using rapid prototyping machine. Test were conducted at geometrically similar with 1/10th scale wind tunnel model. It can provide dynamically flow situations with prototype [6]. Table 3 shows the result of average
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780
1779
drag coefficient which is at 0.48. The tunnel is unable to measure lift coefficient with model configured (shown in Fig. 5b.) because it uses three-component balance system. Table 3: Wind tunnel experiment conditions and results Air density (kg/m3 )
1.17 - 1.18
Scale (mm)
1: 10
Speed (m/s)
10 - 30
CD
0.48 (average)
5. Analysis on EMo-C’s Drag Coefficient Table 4 shows comparisons between CFD and wind tunnel results. It is expected that CFD yields lower value because the simulations only take account for inviscid flow where as in wind tunnel experiment, the flow is definitely viscous. The drag coefficient yields drag force of 23.8 (CFD) and 27.4 (WT) on actual-sized car which has frontal area of 2.71 m2. Below is an example of how drag is calculated when the car is travelling at 40 km/h (11.1 km/h); D Where
23.8 N ,
, Table 4 : Comparison between wind tunnel experiment and CFD result Sref (frontal)
0.76 m2
Air density
1.225 kg/m3
Airspeed
11.1 m/s
CD
CFD
WT
% diff
0.42
0.48
-12.5%
Drag (N)
23.8
27.4
Power Required (W)
264.2
304.1
Meanwhile, power required to overcome aerodynamic drag from this car is around 260.0 to 305.0 Watt or 0.35 to 0.41 horsepower. This is lower than power required to overcome rolling resistance from tyre-road friction, engine’s thermodynamic efficiency and transmission’s mechanical loss (Table 5). Approximately two horsepower is needed to propel EMo-C car at 40 km/h and aerodynamics only accounts for 19.6 percent of overall power that needs to be delivered by the engine. Given that thermodynamic efficiency of the engine is estimated at 30 percent then aerodynamic drag only accounts for nearly six percent of the car’s fuel consumption. Table 5: Estimated aerodynamic drag power and percentage of fuel consumption Parameter
Value
car speed
11.1 m/s
mass
460 kg
Weight Rolling friction coefficient rolling resistance rolling power mechanical loss (transmission) rolling power required Aerodynamic power required Total power required Percentage of aerodynamic power
including driver
4512.6 N 0.02
standard road tyres and asphalt/concrete road surface
90.3 N 1001.8 W
1.34 hp
20 % 1252.2 W
1.68 hp
304.9 W
0.41 hp
1557.1 W
2.09 hp
19.6 %
of total power required
1780
Rizal E. M. Nasir et al. / Procedia Engineering 41 (2012) 1775 – 1780 Thermodynamic efficiency
30 %
Fuel consumption due to aerodynamic drag
5.9 %
of total fuel consumption
6. Concluding Remarks From computational simulation (CFD) and wind tunnel experiment, the Emo-C car’s drag coefficient is 0.42 to 0.48, respectively, the difference between the two is around 12.5 percent. The former has lower value than the latter due to inviscid flow modelling which disregard skin friction, hence the existence of boundary layer on the body panel (viscous flow). Tyres accounts for 25.7 percent of profile drag coefficient. Without tyres, the car experiences down force but opposite effect (lift) is observed when tyres are included. At average race speed of 40 km/h, the aerodynamic drag only accounts for no more than 20 percent of the overall power required from the engine and slightly less than six percent of the estimated total fuel consumption of the car. Emo-C’s drag coefficient is similar to the old mini/hatchback cars such as Citroen 2CV, Volkswagen Beetle and Daihatsu Charade but this is not exactly poor because a formula cars usually has drag coefficients ranging 0.7 until 1.1. Emo-C car was built in April 2011 and entered the competition three months later. Figure 6 shows the completed car. A quick-fix was introduced with intention to reduce the drag further down based on lessons learned from analyses in this paper. A leading edge extension (LEX) on each side of the air cooling/intake pod is incorporated to prevent air spilling from the front end of the pod to spoil the airflow on top of the rear wheel fender. A proper study on the effect of this LEX to the drag of Emo-C car is needed to prove the claim.
Fig. 6: Completed EMo-C car
7. Acknowledgment We would like to express our gratitude to the Faculty of Mechanical Engineering, Universiti Teknologi Mara for the support in terms of funding and human resources. 8. References [1] Perodua Eco-Challenge 2011 Rules and Regulations, Perodua Sdn. Bhd., January 2011. [2] Automotive Analyses Homepage. www.mayffco.com/analyses.htm. [accessed on March 2011]. [3] Kellar, Pearse and Savill. Formula 1 car wheel aerodynamics. Sports Engineering, 2, November 1999, pp. 203-212. [4] Bienz C., Larsson T., Sato T and Ullbrand B."In Front oftThe Grid- CFD at SAUBER PETRONAS F1 Leading the Aerodynamic Development'. presented at the 1st European Automotive CFD Conference (EACC 2003), 25-26 June, Bingen, Germany 2003. [5] J. Katz. Aerodynamics of Race Cars. Annual Review of Fluid Mechanics. Vol 38, January 2006, pp. 27-63. [6] Manan Desai, S.A. Channiwala and H.J.Nagarsheth." Experimental and Computational Aerodynamic Investigations of a Car", WSEAS TRANSACTIONS on FLUID MECHANICS, Issue 4, Volume 3, October 2008. [7] 'DPMDQRYLüDarko, Kozak, DražDQ,YDQGLüäHOMNR.RNDQRYLüDQGMato, " Car Design as A New Conceptual Solution and Cfdanalysis In Purpose of Improving Aerodynamics", Josip Juraj Strossmayer University of Osijek, Mechanical Engineering Faculty in Slavonski Brod, Croatia [8] Scott Wordley and Jeff Saunders," Aerodynamics for Formula SAE: A Numerical, Wind Tunnel and On-Track Study", SAE p Paper Number 2006-010808