TU Racing

(1)

Acknowledgement

I would like to thank the management of University Racing Eindhoven, for giving me the possibility to do my graduation internship at their team. In particular Jeroen Knippenberg and Bram van de Schoot, they both spent a lot of time and effort in helping me with everything I needed. It was a unique experience, where I learned a lot.

I would like to thank the Police Academy in Lelystad for providing their test track on very short term and I would like to thank Torben Beerneart with his help on the test day.

I would also like to thank my supervisors G.J.F. Heijst, W. Zwart and all the other URE team members who helped me during my internship.

(2)

Foreword

This report is written for the graduation internship of Harm Thewissen at University Racing Eindhoven (URE) as a reference for future students at URE. In order to build a race car that has better grip performance, an aerodynamic package had to be designed, built and tested. This will be the first aerodynamic package of an URE car and it focused to the design of an undertray diffuser. The manager and technical manager chassis defined the assignment project on the design, test and improvement of an undertray diffuser.

(3)

Summary

This report explains the development of an undertray diffuser for an electric Formula Student race car. The main goal of this graduation internship is designing an undertray diffuser. An undertray diffuser is just as the front and back wing of race cars an aerodynamic package that generates downforce. The hard part of designing an aerodynamic package for these cars is their top speed. The faster a car drives the more downforce is going to be generated. The Formula student race cars have a top speed around 130km/h. Due to this low top speed (Formula 1 cars reach top speeds of 370km/h), the wings of the car have to work with lower speeds and have to be larger. The undertray diffuser has to generate as much downforce as possible and as less drag as possible.This report will give the students who are going to build the actual undertray diffuser the design and recommendations to build a good working undertray diffuser.

The first chapter of this report gives an explanation of the Formula Student competition at which URE competes every year. By competing in Formula Student competitions students are challenged to design and build a single-seat formula racecar in order to compete with over 400 university teams from all over the world. These competitions exist out of static and dynamic events. In chapter 2 the working principle of the undertray diffuser is explained. The air under the undertray diffuser travels faster than on top of the undertray diffuser. When this happens a lower pressure is generated underneath the undertray diffuser and this lower pressure generates downforce. The set of requirements for the undertray diffuser is given in chapter 3. Willemsen [1] did 2D CFD simulations in order to find the best undertray diffuser geometry. The optimized undertray diffuser of Willemsen [1] is the foundation of this graduation internship and in chapter 4 is his optimized undertray diffuser compared with a 3D undertray diffuser with the same dimensions in Flow Simulation. The 2D ANSYS model of Willemsen [1] is transferred into 3D with Solidworks in chapter 5. This model is optimized to fit on the URE08 and is built out of MDF. Because ANSYS and Flow Simulation give different values for the downforce and drag the undertray diffuser prototype has to be tested. There are four possibilities to test an undertray diffuser: with small-scale wind tunnel tests, with full-scale wind tunnel tests, with tests on the road with the prototypes mounted on the car, and with tests on the road with a test setup. These possibilities are explained in chapter 6 and in chapter 7 are given the actual test. Due to some problems not everything progressed as planned. The tests were not done on a URE car and the drag is not tested. The test data is validated and compared with the simulations in chapter 8. The end of this report contains out of conclusions, recommendations and appendices.

At a speed of 80km/h the most generated downforce is 210N and the generated drag is 56N. At top speed the downforce can reach up to be 4 times as much. The best undertray diffuser for URE to build is the 7⁰ undertray diffuser. This undertray diffuser must be adapted a little bit to make sure the drag will be as small as possible. The edges from the undertray diffuser have to be round and the length of the Gurney Flap has to be revaluated.

(4)

Introduction

URE designs and builds every year a Formula style electric race car to compete in several Formula Student competitions. URE has never built a race car with an aerodynamic package. A race car without an aerodynamic package can have downforce, but the downforce will be close to zero. Some cars can even generate lift, due to the shape of the body. The well-known aero packages are the front and rear wings. A less known aero package is the undertray diffuser. The undertray diffuser is not that well known, because it is not as visible as the front and rear wings. The undertray is the flat bottom underneath the car and the diffuser are the tunnels which have a positive upwards angle.

This graduation report is about designing, building, testing and improving a real undertray diffuser. It is written because, until the 7th car of URE, the URE cars had no aerodynamic packages. In order not to run behind of the competition, URE is going to develop an aero package for their latest car.

Because URE never designed an undertray diffuser, a lot of research is needed. Willemsen [1] performed a 2D CFD analysis about the undertray diffuser, but a 2D design cannot be mounted on a race car. This graduation report is also written as an assignment for “HAN Automotive Technology” by graduate student Harm Thewissen.

The main goal of this graduation internship is designing an undertray diffuser. The undertray diffuser must generate as much downforce as possible, but when the undertray diffuser generates a lot of downforce it will also generate al lot of drag. There is an optimum between the generated downforce and drag, but this optimum is different for every car. The optimum drag/downforce ratio is dependent on the mounting possibilities and whether the traction is more important than the generated drag when looked at the acceleration and skid pad event or at the efficiency event? The first chapter of this graduation report provides information about the Formula student competition and which events have to be done. In the second chapter the working principle of the undertray diffuser is explained. In chapter three is given the set of requirements for the undertray made up out of the rules from FSAE and the requirements of the management of URE. In the next chapter the simulations of Willemsen [1] are discussed, from which his optimized undertray diffuser

Figure 1 : URE07 {1}

(7)

was designed. This chapter also discusses simulations performed with Solidworks to get a first validation of the research of Willemsen [1]. These simulations provide values for downforces and drag. The optimum lengths and angles are used to design the undertray diffuser prototypes. In chapter 5 I present the design from 2D until the final prototypes. There are four different ways of testing the undertray diffuser; small-scale testing in a wind tunnel, full-scale testing in a wind tunnel, road tests with an URE car and road tests with a test setup. These four possibilities are discussed with their pros and cons in chapter 6. In chapter 7 is described how the tests were carried out. The test results are given in chapter 8 with a comparison of the simulations. The conclusion is given with the best design of the undertray diffuser for URE. Finally, some recommendations are given and some additional specific information is presented in the appendixes.

(8)

1. Formula Student competition

Formula Student, also known as FSAE (Formula Society of Automotive Engineers), is the world’s most established educational Student design competition. Students are challenged to design and build a single-seat formula racecar in order to compete with over 400 university teams from all over the world. The challenge set to the students is to produce a prototype racecar for evaluation by a manufacturing firm. They are to compose a complete package consisting of a well-constructed racecar and a sales plan that best matches given criteria. In a real-world engineering experience young engineers are prepared for their future careers all the while learning new skills and showcasing their talent. Through teamwork the multidisciplinary students are challenged to come up with innovative designs, putting theory into practice. Besides developing technical skills, the students also learn about aspects vital to any organization developing management, marketing and people skills. In a series of static and dynamic events the teams demonstrate their understanding and the performance of their racecar, judged by a jury of experts from the motorsport, automotive and supplier industries. The competition is not simply won by the team with the fastest car, but by the team with the best overall package of design, race performance, cost management and sales planning. Worldwide several Formula Student competitions are organized throughout the year. Teams compete at one or more of these competitions every year, which take place on tracks such as Silverstone and Hockenheimring. A Formula Student competition exists of several static and dynamic events in which a total of 1000 points can be earned. Each event holds a number of points and the team which has collected the most points at the end of the competition is declared the overall winner.

The static events are judged by a jury of experienced engineers from the automotive industry and even from Formula 1. Before a team can take part in any dynamic event, the teams must first pass all stages of the technical and safety inspection process to ensure their car is in compliance with the strict regulations.

Static Events

There are three separate stages in the static events, which tests the understanding of the car and the business skills of the teams:

• Engineering Design Event: assessment of engineering design and implementation • Business Presentation Event: present business plan to market the car for fictive investors • Cost and Manufacturing Event: complete statement of costs for series production of the car

Dynamic Events

There are five separate stages in the dynamic events, which test high performance driving characteristics such as acceleration, braking and handling – all the while being reliable and dependable:

• Acceleration Event: 75 m straight acceleration from standstill • Skid Pad Event: cornering performance test on a figure-of-eight

(9)

• Autocross Event: performance test on a 1 km track

• Endurance Event: overall performance and durability test on a 22 km course, including a driver change halfway

• Efficiency Event: energy consumption test during the Endurance Event

The autocross event, the endurance event and the efficiency event are all done on a 1 km track made out of pylons. An example of the track is given in figure 4.

In the acceleration, skid pad and autocross events is the generated downforce important. In the endurance event and efficiency event is the drag important.

FSAE rules

The rules are stated in the “FSAE_Rules-2013(Final)” document. The document is studied and the rules pertaining to the undertray diffuser are stated in Appendix A. The most important rules for the undertray diffuser imply that the undertray diffuser may not be longer rearward than 305mm from the rear of the tires, may not be wider than the widest point of the tires and may not touch the ground before and during the race. The use of a power device which removes the air from under the vehicle is also prohibited.

The race cars are built to compete in the Formula Student competitions. So everything has to be done in order for the race car to attend at these competitions. The race cars have to finish all the static and dynamic events and every part on the car including the undertray diffuser must comply with certain rules of the FSAE. When the undertray diffuser does not satisfy the rules of the FSAE, it cannot be mounted, or the entire car can be disqualified from the event.

Figure 4 : Example of a 1km track used on the autocross, endurance and efficiency event {4}

(10)

2. Working principle of the undertray diffuser

The undertray diffuser is an aerodynamic package for cars that generates downforce. It can be compared with a Venturi tube (Figure 5). The cross-sectional area at the start of the Venturi tube is relatively large in relation to the cross-sectional area in the middle. The fluid enters the Venturi Tube with a certain speed, when the cross-sectional area decreases, the speed of the fluid increases. The difference in speed will result in a difference in pressure (Figure 5, dp), the pressure at the constriction is lower. As soon as the cross-sectional surface increases the speed of the fluid decreases. When the sectional surface at the end of the Venturi Tube is the same as the cross-sectional surface at start of the Venturi Tube the two speeds of the fluid will be equal.

Figure 5 : Venturi tube {5}

Figure 6 : Undertray diffuser

The undertray diffuser can be compared with a half Venturi tube, they both work on the same principle (Figure 6). The undertray diffuser essentially is built out of four different parts; the intake, the undertray, the diffuser and the Gurney flap.

The air enters the undertray diffuser at the intake. When the air passes along the intake the cross-sectional area decreases, the speed increases and the air enters the undertray. The undertray is the bottom plate of the car, it will make sure the bottom of the car is completely flat and that the air stream is not slowed down due to the holes between the parts underneath the car. The air speed under the undertray is over the total length of the undertray higher than on the top of the undertray. This higher airspeed generates a lower pressure underneath the undertray and the car than on top of the undertray and the car. This lower pressure generates downforce. With this downforce the tires have more normal force and this higher normal force gives the car more traction. The lower pressure will continue over the total length of the undertray until it reaches the diffuser. Once the air enters the diffuser the cross-sectional area will get larger.

(11)

The increasing cross-sectional surface creates a suction peak and the airspeed underneath the undertray will get even higher. This produces even more downforce. After the suction peak the airspeed decreases due to the increasing cross-sectional surface. This ensures that when the air from the bottom of the undertray and the air from the top of the undertray come together at the rear of the car, the difference in air speed will be as small as possible. When this is done correctly the flow will not become turbulent and the drag of the diffuser will be smaller. This ensures that the undertray diffuser is a more efficient downforce producing device. The gurney flap will increase the pressure at the end on top of the diffuser and will decrease the pressure underneath the diffuser. This will help the boundary layer flow to stay attached longer to the bottom of the diffuser.

There are a few more factors that will make a difference in downforce and/or drag.

 Ground clearance:

When there is a large ground clearance, the proportional cross-sectional area will not reduce much from the start of the intake until the end of the intake. So when the ground clearance will be made very small the air will move faster than with a larger cross-sectional surface. With respect to this fact the ground clearance has to be very small, but this will give two problems. The first problem is that when the ground clearance is too small, the undertray will produce positive lift since there is hardly any airflow between the undertray and the ground. The second problem is that due to irregularities at the surface of the ground, the undertray could hit the ground when the ground clearance is too small. So the ground clearance cannot be too large and not too small. [2, pages 392-397]

 The angle of the diffuser:

If the angle of the diffuser is close to zero the boundary layer flow will not detach, but the air speed will not be reduced enough to make a laminar transition of the air at the end of the car when to two airstreams meet.

If the diffuser has a very large angle the boundary layer flow will detach and the airflow underneath the diffuser will be turbulent.

So the angle of the diffuser is limited with two boundaries: when the angle is too small or too large angle the diffuser will create more drag then necessary.

 Air speed:

With an increasing driving speed the airspeed will also increase. When this happens the proportional difference between the speed of the air on top of the undertray and underneath the undertray will get higher. This means the difference in pressure will be higher.

So the generated downforce and drag are completely dependent on the dimensions of the undertray diffuser and the speed of the air/speed of the car. When the incorrect dimensions are taken the undertray diffuser will generate a lot of drag and will generate a little amount of downforce or even positive lift.

(12)

3. Set of requirements

In the previous two chapters certain dimensions are given for the undertray diffuser. Out of these dimensions a set of requirements can be made that consists of a set of requirements from the FSAE Rules (See also Appendix A) and a set of requirements from the management of URE. Some of the requirements of the management of URE have been discussed by Willemsen [1]. The set of requirements is made, because the undertray diffuser has to meet these requirements.

FSAE requirements

1. The undertray diffuser may not be longer rearward than 305mm from the rear of the tires. 2. The undertray diffuser may not be wider than the widest point of the tires.

3. The undertray diffuser may never touch the ground, neither in the static nor in the dynamic events.

4. The undertray diffuser may not enter the keep-out-zone defined as a circle of 68.6mm around the tires.

5. All the edges of the undertray diffuser that could contact a pedestrian must have a minimum radius of 1.5mm.

6. The undertray diffuser may not complicate egress from the vehicle. 7. The use of a power device is prohibited.

URE requirements

8. The undertray diffuser must have the correct dimensions in order to mount it on the URE08. 9. The Centre of Pressure (COP) from the undertray diffuser must be as close as possible to the

40/60 weight distribution of the URE08.

10. The undertray diffuser must comply, where possible, with the dimensions of the optimized undertray diffuser of Willemsen [1].

11. The undertray diffuser and the undertray diffuser test stand have to be as cheap as possible. These requirements are used in the following chapters. When the undertray diffuser does not meet the requirements of the FSAE the entire race car can be disqualified from the completion, and when the undertray diffuser does not meet the requirements of URE the possibility exists URE is not going to use the undertray diffuser.

(13)

4. Computational Fluid Dynamics (CFD) simulations

As stated in the previous chapters, the undertray diffuser has to comply with certain dimensions. When the wrong dimensions are used the undertray diffuser will not work as well as it could. The undertray diffuser would generate too little downforce and too large drag. In this chapter the simulations of Willemsen [1] will be briefly discussed, simulations are done with Flow Simulation, and the outcomes of both simulations are compared.

Willemsen [1] performed his 2D CFD simulation with ANSYS. The education automotive engineering at the HAN uses the 3D software Solidworks to draw and calculate parts. A feature ‘Flow simulations’ is implemented in Solidworks, in which 3D CFD simulations can be performed.

4.1 ANSYS 2D CFD simulations

Willemsen [1] made an undertray diffuser without the precise dimensions from the URE racing car. He optimized the undertray diffuser in such a way that it would have the best downforce/drag ratio. The undertray diffuser has no rounded edges and because it is a 2D simulation the undertray diffuser is infinitely wide. The simulations are done without the surrounding body of the car and the simulations are done with a moving road underneath the undertray diffuser. When there is no moving road underneath the vehicle the downforces are not as high as in reality. This has to do with the boundary layer on the ground. This boundary layer without the effect of wind has a speed of close to zero. When the ground in the wind tunnel moves the difference in speeds is also zero. The optimized undertray diffuser is simulated and the contours of static pressure and the contours of velocity are given below. The air travels from the left to the right. Figure 7 shows a contour plot of static pressure. The red areas correspond to high pressure and the blue areas indicate low pressure, with respect to each other. Figure 8 presents a contour plot of the velocity magnitude. The red areas correspond with high velocity, the blue areas with low velocity and the green areas represent moderate velocity. From these two plots can be seen that the undertray diffuser will create downforce.

The drag and the downforce are given in two graphs (Figure 9 and 10). In both graphs the force is plotted against the inlet velocity. These graphs show that both the downforce and the drag are linked with the inlet velocity (driving speed).

Figure 7 : Contours of static pressure near the optimized undertray diffuser (ANSYS) {6}

Figure 8 : Contours of velocity magnitude near the optimized undertray diffuser (ANSYS) {7}

(14)

The two graphs (Figure 9 and 10) and the conclusion about drag and downforce of the BEP (Bachelor End Project) of Willemsen [1] contradict each other. The conclusion gives a drag of 22N/m and a downforce of 97N/m at 50km/h. When the drag and downforce is found in the graphs, a drag of approximately 100N/m and a downforce of approximately 400N/m will be found. After consultation with Willemsen, he explained that the absolute numbers in the graphs are not correct, but the shape of the contours is correct. This means that the values given in his conclusion are correct according to the simulations, but the forces of the conclusion have to be used at 50km/h and then the other values for the downforce and drag at other velocities can be found analytically.

4.2 Solidworks Flow simulations

Flow simulation is used for several reasons: to get a first validation of the research of Willemsen [1], to gain a better understanding of the airflow around the undertray diffuser, and to see if the effect of rounded edges at the start of the diffuser tunnels will improve the drag or downforce of the undertray diffuser.

The same dimensions are chosen as those of the undertray diffuser of Willemsen [1]. There are two different angles simulated, an angle of 18.9⁰ and an angle of 7⁰. The angle of 18.9⁰ is simulated because it is the optimum angle according to Willemsen [1]. The angle of 7⁰ is simulated because it is the largest angle that does not cut through the suspension rods. This is explained more extensively in chapter 5.The simulations are done with an air and groundspeed of 13.89m/s (50km/h).

Figure 11 shows the flow trajectories for the 18.9⁰ undertray diffuser, while figure 12 displays the trajectories for the 7⁰ undertray diffuser. It is easy to see that underneath the diffuser the air is far more turbulent with the 18.9⁰ undertray diffuser than with the 7⁰ undertray diffuser. Due to this turbulence the diffuser will generate drag.

In figure 13 it is seen that when the start of the diffuser has a rounded edge (this edge has a radius of 500mm), the airflow is still turbulent but it is less turbulent than with the straight edge.

Figure 10 : Drag per geometric component for different inlet velocities {8}

Figure 9 : Negative vertical lift per geometric component for different inlet velocities {9}

(15)

Besides these flow trajectory plots it is also possible to generate, just as the ANSYS plots, pressure plots en velocity plots for a certain section.

Figure 14 and figure 15 show the calculated contours of static pressure for the 18.9⁰ undertray diffuser and the 7⁰ undertray diffuser. Figure 14 reveals a large high pressure zone on top of the 18.9⁰ undertray diffuser, this high pressure zone is much smaller with the 7⁰ undertray diffuser in figure 15. The lower pressure zone underneath the diffuser is much smaller for the 7⁰ undertray diffuser. This indicates that the 18.9⁰ undertray diffuser will generate more downforce, but will also generate more drag.

Figure 11 : Flow trajectories for the 18.9⁰ undertray diffuser with an air and ground

speed of 50km/h

Figure 12 : Flow trajectories for the 7⁰ undertray diffuser with an air and ground

speed of 50km/h

Figure 13 : Flow trajectories for the 18.9⁰ undertray diffuser with rounded edges (500mm radius) and

with an air and ground speed of 50km/h

Figure 14 : Contours of static pressure near the 18.9⁰ undertray diffuser (Flow Simulation)

Figure 15 : Contours of static pressure near the 7⁰ undertray diffuser (Flow Simulation)

(16)

Figure 16 and figure 17 provide contour plots of the velocity magnitude for the 18.9⁰ undertray diffuser and the 7⁰ undertray diffuser. The airspeed underneath the undertray of the 18.9⁰ undertray diffuser has a darker color red than the airspeed underneath the undertray from the 7⁰ undertray diffuser. This means the 18.9⁰ undertray diffuser generates a larger downforce. The airspeed underneath the diffuser of the 18.9⁰ undertray diffuser has a dark blue color, implying that the air speed underneath the diffuser is almost 0m/s and slows the air down far too much. In this way there will be turbulence behind the car, because the speed of the air is slower underneath the diffuser than on top of it. With the 7⁰ undertray diffuser the colors above the undertray diffuser and underneath the diffuser are the same. This means the wake flow behind the car will not get turbulent. The only fault with the 7⁰ undertray diffuser is the size of the gurney flap. Because it is too long, the air is slowed down too much and the gurney flap will create excessive drag.

The undertray diffuser with an 18.9⁰ diffuser angle generates 56N/m drag and 209N/m downforce. The undertray diffuser with a 7⁰ diffuser angle generates 22N/m drag and 144N/m downforce. So the 18.9⁰ undertray diffuser generates 2.5 times as much drag, but only 1.5 times as much downforce as the 7⁰undertray diffuser.

When looked at the 18.9⁰ diffuser with sharp edges at the beginning of the diffuser (Figure 11) it generates 56N/m drag and 209N/m downforce. A diffuser with the same angle but with rounded edges (Figure 13) generates 59N/m drag and 220N/m downforce. When this would be an optimization of a previous undertray diffuser it would be useful to adapt these edges. Because the undertray diffuser is a first prototype and it has to be as cheap as possible to built, the rounded edges will not be applied in the prototype.

4.3 Comparison of ANSYS and Flow Simulation

The values of Willemsen’s [1] simulations and the values obtained with the Flow Simulation simulations are not the same. It is very hard to explain what the reason for these differences. Both the programs use the k-epsilon turbulence model. In all the simulations a moving ground underneath the undertray diffuser is used and the ground clearance of 30mm stated in Willemsen’s [1] optimized design is used. ANSYS is 2D and Flow Simulation is 3D, so the simulations of ANSYS are performed with an infinitely wide undertray diffuser with respect to the 1m wide undertray diffuser of Flow Simulation.

Figure 16 : Contours of velocity magnitude near the 18.9⁰ undertray diffuser (Flow Simulation)

Figure 17 : Contours of velocity magnitude near the 7⁰ undertray diffuser (Flow Simulation)

(17)

In Flow Simulation it is possible air travels along the edge from the high pressure zone to the low pressure zone. When the 7⁰ undertray diffuser is made into a 2m wide undertray diffuser the drag is 23N/m and the downforce is 173N/m, and when the undertray diffuser is made into a 4m wide undertray diffuser the drag is still 23N/m and the downforce is 180N/m. When the undertray diffuser is made even wider the downforce will not get higher then 185N/m. This indicates that the air that travels from the high pressure zone to the low pressure zone has a limited effect and when the undertray diffuser is made infinite wide this effect will not be taken into account. Nevertheless a 2m wide undertray diffuser is used to compare the two programs. The domains of the simulations in ANSYS and the simulations in Flow Simulation are: the inlet of the domain is 2380mm in front of the undertray, the outlet domain is 5950mm at the rear of the diffuser, and the height of the domain is 3000mm. The depth of the domain is only given for Flow Simulation because ANSYS is 2D and has no depth, this domain is 3000mm wide for the 1m wide undertray diffuser. Because ANSYS and Flow Simulation are two completely different packages, it cannot be simply stated which simulation is correct. They are used to create a first estimate of the possible forces that work on the undertray diffuser.

The forces are in N/m, the actual diffusers will be around 0.5 meter so the forces have to be halved to obtain the correct forces. The forces are obtained from simulations for a ground speed of 50km/h.

Program Angle Drag (N/m) Downforce (N/m)

ANSYS 18.9⁰ 22 97

Flow simulation 18.9⁰ 56 209

ANSYS 7⁰ 6 75

Flow simulation 7⁰ 22 144

Table 1: Drag and downforce for ANSYS and Flow Simulation with a 7⁰ and 18.9⁰ undertray diffuser at 50km/h

The first simulated angle with Flow Simulation was the same angle as the optimized angle from Willemsen [1]. The second simulated angle is 7⁰. This angle is simulated, because it is the only angle where the suspension does not cut through the diffuser tunnels. The angle is also simulated to see if the differences in downforce and drag between the angles of the diffuser are really as large as simulated.

While the two programs give different forces there cannot be said which program is correct. In order to see which program is correct tests are needed.

(18)

5. Realization of the undertray diffuser

The model of Willemsen [1] is used to make a first concept. This concept is adapted and with the requirements presented in chapter 3 an undertray diffuser is designed that will fit the URE08 and will comply with the rules of the FSAE. In this chapter the 2D design of Willemsen [1], a first concept with the design of Willemsen [1], and an optimized 3D model are discussed. After this the MDF prototype will be shown.

5.1 2D model according to Willemsen [1]

Willemsen [1] optimized his 2D

undertray diffuser by performing 2D CDF analyses. At that stage no correct dimensions of the URE08 were available, because the car was still in the design stage. Thus it could be that his optimized undertray diffuser would not have the right dimensions to fit on the URE08. Figure 18 shows the dimensions of his optimized undertray diffuser.

5.2 3D model according to Willemsen [1]

A simple 3D concept, according to the dimensions of the 2D undertray diffuser from the thesis of Willemsen [1], was made and placed in the 3D model of the URE08 (Figure 19). It was obvious that it would never fit without adaptations. It would cut through the batteries, the monocoque and the side skirts. This concept is made to see how many adaptations are necessary in order to design an undertray diffuser, with the dimensions of Willemsen [1], which would fit underneath the URE08. When the undertray diffuser is placed further to the rear of the car it would not cut through the battery casing anymore, the single diffuser has to be shaped into two diffuser tunnels, and the diffuser has to be clear of the suspension rods. Apart of these adaptations the undertray diffuser has to meet the FSAE rules.

Figure 19 : First concept according to Willemsen’s [1] 2D design

(19)

5.3 Optimized 3D models

The first change was placing the undertray diffuser further to the rear, so that the diffuser would begin at the end of the side skirts. Doing this the COP will be placed further to the rear, but it cannot be solved in any other way. In the first concept the diffuser was as long as possible, according to the rules of the FSAE. Because it was placed further to the rear the diffuser had to be made shorter, or else it would be in conflict with the rules. At this moment the diffuser was one large plate, it had to be made into two separate plates between the wheels and the monocoque. These plates must meet the rules of the FSAE and may not come too close to the wheels. (See chapter 1 FSAE rules) To ensure that the air could not leave at the side of these diffuser plates, side skirts were added. In the first concepts the gurney flaps were only on the top of the diffuser. They are stretched over the side of the side skirts of the tunnels to increase the total stiffness of the diffuser tunnels.

The next step was ensuring that the suspension rods could not touch the diffuser. This can be done with two different diffuser angles. When the diffuser had an angle of 7⁰< α >21⁰ the suspension rods would cut through the diffuser plate in such a way that the holes would get too large. When this would be done the holes cannot be sealed in a proper manner and the airstream would be affected too much. This can be seen in figure 20: this figure show the diffuser tunnels, with the holes made to ensure that the suspension rods would not touch the diffuser in the highest and lowest position of the car. So it was decided to use two different diffuser angles: 7⁰ and

21⁰. The 7⁰ undertray diffuser does not interfere with the suspension rods. Figure 21 shows the assembled 21⁰ undertray diffuser. Figure 22 and 23 display the optimized 3D models. At last three different diffuser tunnels were constructed: two diffuser tunnels are made with the same angle of 21⁰ but with a different length, and one is made with a smaller angle of 7⁰.

Figure 20 Diffuser tunnels at a 15⁰ angle, with the holes for the suspension rods.

(20)

5.4 Undertray diffuser prototypes

Now that we have seen the images of the virtual undertray diffuser in this chapter, here are finally some pictures of the undertray diffuser prototypes placed under the car (Figure 25). The undertray diffuser is built out of 6mm MDF according to the 3D models shown in section 5.3. The plates are glued together and strengthened with 20mm aluminum L profiles. Figure 24 shows the 21⁰ undertray diffuser prototype photographed at several view angles. Figure 25 contains two photographs of the almost finished URE08 with the 7⁰ undertray diffuser prototype placed underneath the URE08. MDF is used because it is, together with the aluminum L profiles, a strong and cheap solution to build the prototypes.

Figure 22 : Optimized 3D model (α=21⁰)

Figure 23 : Optimized 3D model (α=7⁰)

Figure 25 : 21⁰ undertray diffuser prototype shown from different angles

Figure 24 : 7⁰ undertray diffuser prototype placed underneath the URE08

(21)

6. Test possibilities

In chapter 4 is shown that tests are needed to investigate how much downforce and drag the undertray diffusers actually generate, and in chapter 5 is shown how the undertray diffuser prototypes are designed and built. These tests can be done in four different ways: small-scale wind tunnel tests, full-scale wind tunnel tests, tests on the road with the prototypes mounted on the car, and tests on the road with a test setup. These four possibilities were all examined and are explained in this chapter.

6.1 Small-scale wind tunnel tests

Usually the first aerodynamic test that will be done is a scale test of the vehicle in a wind tunnel. This is the fastest and cheapest way to do the tests. Only the aerodynamic properties of an undertray diffuser are linked to a phenomenon called the ‘ground effect’. The ground effect is a phenomenon where the air under the undertray accelerates. Normally this would not be a problem, but when a model is scaled down and tested in a wind tunnel the velocity of the air has to be scaled up. This has to be done in order to keep the Reynolds number the same. The Reynolds number gives the inertial forces and the viscous forces. This ratio has to be kept the same in the model in order to have a dynamic similarity.

The Reynolds number is defined as:

The dynamic viscosity (μ) and the density of the air (ρ) cannot be changed, so the only variables in the equation are the mean velocity ( ) and the characteristic linear dimension (L). The characteristic linear dimension will be scaled down, so the mean velocity has to be scaled up with the same factor. This also would not be any problem, because the existing wind tunnels can handle these airspeeds. But due to the ground effect, there has to be a moving a moving ground underneath the vehicle or the entire test would not be reliable. When the car is scaled down, the airspeed and the speed of the moving ground have to be scaled up. So with a top speed of roughly 140km/h (The top speed of the URE08) and a scale of 1:10 the airspeed and the speed of the ground have to be between 700 and 1400km/h. To design and build a moving ground with these speeds is not possible within the given time. Because of this, the possibility for scale testing was rejected.

6.2 Full-scale wind tunnel tests

Another possibility is performing wind tunnel tests at full-scale. Full-scale wind tunnel testing is an expensive way of testing, but the great advantage is that every parameter can be controlled and can be kept the same for every test. Just like with the scale model test, a moving ground has to be implemented in the test. Designing a moving ground that can move 130km/h is hard but not impossible, so this is a good possibility.

As said, wind tunnel tests are very expensive and URE is a racing team based on sponsorships. So the next hurdle was to find a company that was willing to let us use their wind tunnel based on a sponsorship.

(22)

First a list with requirements for the wind tunnels was made.

 The wind tunnel had to be large enough to fit the URE08, without the influence from the walls or ceiling on the airstream. The dimensions of the URE08 are 2.5mX1.75mX1.25m (LxBxH). For the dimensions of the wind tunnel a minimum of 5mX2.75mX2.5m is chosen. Larger is better, but when only a wind tunnel was available with these dimensions it would be sufficient to do the tests.

 The wind tunnel must be able to generate airspeeds up to 140km/h.

 It is preferred that the wind tunnel has a built-in moving ground. When a wind tunnel is accessible for URE without a moving ground, the moving ground will be built.

An inventory was made of available wind tunnels that meet these requirements. There are seven companies in Holland and Germany with wind tunnels that meet the requirements. This list is given in Appendix B. Each company was mailed and phoned, but they all declined any sponsorship. Only TNO agreed to get further in detail. Eventually also TNO replied negatively.

6.3 Road tests with an URE car

This possibility of testing is maybe not the most accurate way to test the undertray diffuser, but all the factors that can influence the test will have an influence on the undertray diffuser when it is used in races.

Because the prototypes have to be mounted and demounted, the prototypes have to be fixed onto the monocoque in such a way that it can easily be taken of the URE08. In order to do this, three brackets have to be made. These are one rear bracket (Figure 26) and two front brackets (Figure 27). The two front brackets will be placed on the outside of the seatbelt brackets and mounted on the undertray. The rear bracket will replace the quick jack bar and mounted between the diffuser tunnels.

The downforce will be measured with strain gauges. This will measure only the downforce of the undertray diffuser and not the combined downforce of the car and the undertray diffuser. This is the needed data, because the effect of the undertray diffuser is examined and not the downforce of the total car. There will be placed 2 strain gauges on the front brackets and 4 strain gauges on the rear bracket. Each red mark in figure 26 and 27 proposes a strain gauge. On the opposite sides are placed another 3 strain gauges.

Figure 27 : Front bracket Figure 26 : Rear bracket

(23)

This is done to make sure that only the normal force will be measured and not the possible flexion of the brackets. These forces are stored on the MoTeC data acquisition system. The MoTeC data acquisition system has a minimal input resolution and the amplifiers have a maximum amplification. When the first brackets were designed the resolution was not taken into account. The strain gauge calculations are elaborated further in Appendix C.

The drag will be measured with a roll out test. First this test has to be done without the undertray diffuser and subsequently with the three different undertray diffusers. This has to be done, because the drag of the whole car with diffuser will be measured. In order to know the contributed drag of the undertray diffuser, the difference between the two has to be measured.

The corner stability will be measured with an enlarged skid pad test. The circles of the skid pad events at Formula Student competitions have a radius of 10.5m. The circles for the tests will have a radius of 15m so that the speed will be approximately 40km/h, the speed at which the most corners in a Formula Student event will be taken. The test will be done first without the undertray diffuser and then with the three different undertray diffuser prototypes. The lateral acceleration will be measured with the sensors of the car and the test with the highest lateral acceleration has the most traction in corners.

Every undertray diffuser prototype will have the same test sequence and every test will be done two times. The straight test track will be driven in both directions to see what the effect of the wind direction is and when possible to eliminate the wind. The driver will pull up to 30 km/h, then drives 5sec at this speed, and then pulls up to the following speed:

30 km/h, 40 km/h, 50 km/h, 60 km/h, 80 km/h, 100 km/h and top speed.

This speed will be held constant with the cruise control of the URE08. The speed can be a little bit different for each test, but the speed will be stable to get an accurate measurement. Small variations in the speed are not so problematic, because these variations will be too small to measure.

After this the drag test will be done in order to measure the drag, while the skid pad test will be done in order to examine the effect of the undertray diffuser on corner stability.

This possibility is examined and calculated, because it would be used in the last week before the end of this graduation internship. One week before the actual tests, the battery package of the URE08 had a short circuit and this would not be fixed before the test day. So another solution had to be found. First, the possibility was considered to run the tests on an older URE car. After a careful examination it was clear that the undertray diffuser would never fit on one of the older URE cars. So the possibility of testing on another URE car was not a possibility anymore until the URE08 would drive again.

6.4 Road tests with a test setup

Due to this unfortunate event a whole new test stand had to be developed and built. A frame was designed and built in order to test the undertray diffuser at the side of the URE van (Figures 28 and 29). A frame built with Bosch profiles (40mmx40mm) is placed in the van. Attached to this frame is a second frame that is shown in figure 28 and 29. In order to have as little as possible air disturbances from the van, the frame was made 2.5m long and there is 1.5m between the van and the side of the undertray diffuser.

(24)

Because of the length of the frame and the generated downforce the van would tilt, the frame and the undertray diffuser would not be horizontal to the road and the measurements would not be correct. Not to speak of the fact that the undertray diffuser could hit the ground and could break.

Therefore a side wheel is attached to the end of the frame, to make sure the frame would stay at the same height at different speeds. This side wheel shall have an influence on the airstream next to the undertray diffuser. This disturbance is small and is the same with every test, so it is taken as negligible. Because of the van the wind from the side can be totally different than in reality, but due to the low height of the undertray diffuser and the side skirts over the total length of the undertray diffuser the influence will not have a very large effect. Every test shall be done in both driving directions in order to make the effect from the wind as small as possible. Just as with the road test with an URE car, every test will be done in the same way. Only now there will be no skid pad tests and the drag is going to be measured in the same way as the downforce. Every test will be done two times, the straight test track will be driven in both directions to examine what the effect of the wind direction, and when possible to eliminate the wind effect. The driver will pull up to 30 km/h, then drives 5sec at this speed, and then pulls up to the following speed:

30 km/h, 40 km/h, 50 km/h, 60 km/h and 80 km/h. Higher speeds will not be possible on the length of the test track so 80km/h will be the max.

Both the downforce and the drag are going to be tested with 1KN load cells (the two watertight sealed blue cells in figure 30 and the green square in figure 31). The red U-frame (Figures 30 and 31) is attached at the front and back to the undertray diffuser (the small black rectangle at the bottom of figure 31).

Figure 29 : Front of the road test setup

Figure 28 : Back of the road test setup

(25)

When the undertray would be attached to the URE08 it could not bend, but because this is not the case the undertray is strengthened with two extra profiles attached between the red U-frames. The only connection between the red U-frame (undertray) and the frame of the bus are the two load cells to make sure that the downforce and the drag will only pass through the load cells. The two short vertical profiles next to the load cells are placed to make sure that the undertray diffuser will not rotate. (The loadcells have a strain of max 0.1mm) The drag will be measured with two extra U-frames in order for the load cells to be turned 90⁰. The ground clearance can be adjusted by moving the black horizontal profile up or down (Figure 31).

(26)

7. Testing

In the previous chapter we discussed the different possibilities to test the undertray diffuser. Finally the fourth possibility; testing on the road with a test setup was chosen. In this chapter we will discuss the actual road tests.

After contacting the Police Academy in Lelystad they offered to sponsor the entire test day at their facility. The test track is given in figure 34. Because the fourth option was chosen to test the undertray diffuser prototypes, only the 800m long straight track was used.

Figure 32 : Test track of Police Academy in Lelystad {11}

There are many factors that may have an influence on the behaviour of the undertray diffuser. For example: the surface of the road, the ground clearance, braking, accelerating and cornering, the driving speed, the wind speed, rounded edges of the diffuser, and the thickness of the plates of the undertray diffuser. Although the undertray diffuser will in reality also be mounted underneath the URE car, in the test setup it is freely constructed on the frame.

Not all these factors will be taken into account in this study. The surface of the road and the wind speed cannot be changed. The thickness of the undertray diffuser plates can make a difference, but it will be too expensive to make and test multiple undertray diffusers with different thicknesses. The ground clearance will not be the same during braking and accelerating over the total length of the undertray. This will give different downforces, but this will not be investigated in these tests. The angle of the diffuser can make a large difference in both the downforce and the drag, but due to the suspension of the car only two different angles are tested.

Because of the late developments, the entire test stand had to be altered and the test stand could not be tested before the test day. There were a few errors that had to be fixed at the test location and the test started later than planned. Additionally it was a very rainy day and the tests had to be stopped or had to be repeated several times. When the second test began, it started to rain. The test was done, but it soon became obvious that due to the heavy rain the test data were useless. So the tests were stopped during rainy periods and repeated afterwards. This caused several delays and not every test could be done. Due to these factors the drag could not be tested and only the downforce was measured. The centre of pressure will be calculated from the downforce from the front and back load cells.

(27)

Figure 33 : Mounted 7⁰ undertray diffuser prototype

Figure 32 and figure 33 show the mounted undertray diffuser with angles of 7⁰ and 21⁰, respectively.

After the first test it turned out that the obtained data was very unsteady and at low speeds the generated downforce was defined unreliable. Because time was limited I decided to perform the tests in a different way. The driver had to accelerate until 80km/h and hold this speed for 5sec. By doing this the difference in downforce between the different undertray diffuser prototypes would be larger and the data could be used better. The disadvantage of these tests is that information for the lower speeds was lacking, but the max downforce and drag would give enough information. The measured data is given in the next chapter and is compared with the simulations.

Figure 34 : Mounted 21⁰ undertray diffuser prototype

(28)

8. Validation of the tests

Chapter 7 shows the progress from the test day. The test day has given data in the form of voltages from the front and back load cells. A maximum of 5V was possible so 1V equals 200N. Out of these data the downforce for the three different undertray diffuser prototypes and the centre of pressure (COP) are generated. After this is examined which simulation gave the best values and the drag is taken out of these simulations.

8.1 Centre of pressure

The undertray diffuser is placed approximately 400mm to the rear in comparison with the 40/60 weight distribution of the car. The COP of the undertray diffuser will therefore also be placed 400mm to the rear. When is watched at the shift of the COP of the entire car, due to the shift of the COP of the undertray diffuser, the COP of the entire car will shift 47mm to the rear.

Because the force is measured with two different load cells, it is fairly easy to calculate the centre of pressure (COP). The graphs in Appendix D (Figure D1-D3) show the measured downforces, for the front and back load cells, as a function of the speed for the 7⁰ and the two 21⁰ undertray diffusers. Note that figure D2 and D3 show results for the same diffuser angles, but for two different diffuser lengths.

When the front and back forces would be the same the COP would lay, due to the positioning of the load cells, 50mm in front of the start of the diffuser. The load cells are located 840mm from each other. (Figure 35)

(29)

The COP for the different undertray diffuser is given in Table 2. Back load cell

force (N)

Front load cell force (N)

COP removed from the start of the diffuser to the rear (mm)

7deg diffuser 100 85 13

21deg short diffuser 115 95 23

21deg long diffuser 140 70 160

Table 2 : Centre of Pressure for the given undertray diffusers

8.2 Downforce

The forces of the front and the back load cells are added and plotted in the graph below (Figure 36). The data of the load cells fluctuates due to the vibrations and the gear shifting of the van. Due to this every run is done two times and the best graph is chosen. When there is more time available on a test day it is better to do all the tests more than two times so that false measurements can be detected better. The best graph is not certainly the graph with the highest downforce, but the graph with the best transition of the entire graph. In this case the downforces at 80km/h were close to each other so there had to be no guessing which maximum downforce was the correct one. Every run contains accelerating from 0 to 80km/h, maintaining the speed at 80km/h for 5sec and slowing back down to 0 km/h. The slowing back down to 0km/h is not given in the graph, because only the maximum downforce is wanted. The 5 sec driving at 80km/h is also not given in the graph, but the data is used to make sure the top of the three downforce lines is correct. The tests are not done with a maximum of 50km/h as the simulations, but are done with higher speeds because the difference in downforce is larger this way and the errors of the load cells are smaller.

Figure 36: Downforces for the three tested undertray diffusers

0 50 100 150 200 250 0 10 20 30 40 50 60 70 Do w n fo rc e ( N ) Speed (km/h)

Downforce undertray diffusers

7deg diffuser 21deg short diffuser 21deg long diffuser

(30)

The larger angle of the diffuser gives more downforce, but the length of the diffuser does not generate more downforce. It gives more frontal surface so it will generate more drag, but in order to be sure the drag tests have to be done.

8.3 Drag

Because the drag is not tested, but the simulations of Flow Simulation are fairly accurate, the drag of the undertray diffuser will be taken from these simulations. The drag from Flow Simulation for the undertray diffuser with an angle of 21⁰ will be too high, but the drag of the 18.9⁰ undertray diffuser will be close to the drag of the 21⁰ undertray diffuser prototype. This is a rough estimation, but it shows that the drag of the 21⁰ undertray diffuser is a lot higher that the drag of the 7⁰ undertray diffuser. To get the real data for the drag, the tests have to be done over and now with the drag tests.

Program Angle Drag (N)

Flow simulation 18.9⁰ 78

Flow simulation 7⁰ 32

Table 3 : The drag for the undertray diffusers at 80km/h taken out of the Flow Simulation simulations

8.4 Comparison with ANSYS and Flow simulation

In order to compare the measured downforces at the test day the Flow Simulation simulations are done another time at 80km/h, the downforces for ANSYS at 80km/h are analytically taken from the graphs. These downforces are together with the measured downforces from the test day given in Table 3.

Downforce (N) Diffuser angle

ANSYS 121 18.9⁰

Flow Simulation 302.5 18.9⁰

Real undertray diffuser 210 21⁰

ANSYS 94 7⁰

Flow Simulation 193 7⁰

Table 4 : Downforce at 80km/h out of ANSYS, Flow Simulation and the undertray diffuser tests Even after the tests it is very hard to compare the two programs and the real test. ANSYS gives with both the angles far too low downforces. Flow simulation on the other hand gives with the angle of 7⁰ a very good downforce, but gets a too high downforce number with the angle of 18.9⁰. With a simulation of an even larger angle (21⁰) as with the tests the downforce would even be larger. Comparing the undertray diffuser prototypes, the 2D ANSYS undertray diffuser, and the 3D Flow Simulation undertray diffuser will always give different values. In the ANSYS simulation a 2D undertray diffuser is used, in the Flow Simulation simulations a 3D undertray diffuser is used, and at the road tests an undertray diffuser with diffuser tunnels and less wide intakes is used.

(31)

The undertray diffuser prototypes a mounted with a frame which generates turbulence above and behind the diffuser while the ANSYS and Flow simulation undertray diffuser fly in the air. All these factors make it hard to compare the three to each other, but it gives an indication whether the simulations give too high or too low forces.

8.5 Downforce drag ratio

The downforce drag ratio of the undertray diffuser prototypes are calculated with the drag of the Flow Simulation simulations and the downforce of the undertray diffuser prototypes. This is only an indication of the r=downforce drag ratio. In order to get the correct downforce drag ratio the drag tests with the undertray diffuser prototypes have to be done.

The downforce drag ratio for the 21⁰ undertray diffuser is => The downforce drag ratio for the 7⁰ undertray diffuser is =>

(32)

9. Conclusion

The main goal of this graduation internship was designing an undertray diffuser.

The eventual designs of the undertray diffuser are three different diffuser tunnels. Two diffuser tunnels of 21⁰, with a different length and one diffuser tunnel of 7⁰.

These undertray diffusers are tested and the downforces are compared with the simulations. Downforce (N) Diffuser angle

ANSYS 121 18.9⁰

Flow Simulation 303 18.9⁰

ANSYS 94 7⁰

Flow Simulation 193 7⁰

Table 5 : Downforce at 80km/h out of the different simulation programs and the undertray diffuser tests The drag tests are not done on the test day so an estimation has to be done to get the drag forces. Because Flow Simulation gives better downforce data, the drag is taken from the Flow Simulation simulations. When the drag data has to be precise the drag tests have to be done.

Program Angle Drag (N)

Flow simulation 18.9⁰ 78

Flow simulation 7⁰ 32

Table 6 : The drag for the undertray diffusers at 80km/h taken out of the Flow Simulation simulations Simulations can be used to get a first impression of the range in which the forces will be, but it cannot be used without real tests.

The short diffuser tunnels have a COP close to the start of the diffuser tunnels. When the diffuser tunnels get larger the COP of the undertray diffuser moves more rearwards.

Back load cell force (N)

Front load cell force (N)

COP removed from start diffuser (mm)

7deg diffuser 100 85 13

21deg short diffuser 115 95 23

21deg long diffuser 140 70 160

(33)

The best undertray diffuser design for the URE08 is the 7⁰ undertray diffuser, because:

 This 7⁰ undertray has a better downforce drag ratio of 5.8 against the downforce drag ratio of 2.7 from the 21⁰ undertray diffuser

 The 7⁰ undertray diffuser generates only 25N less downforce than the 21⁰ undertray diffuser, which is only 10% less downforce

 The 7⁰ undertray diffuser does not cut through the suspension rods. This will make it a lot easier to do maintenance on the car

 The COP of the 7⁰ undertray diffuser is better, because it is closer to the 40/60 weight distribution off the car

(34)

Recommendations

1. The tests have to be done once again and this time also the drag has to be measured. This can be done by placing the load cells in a horizontal position.

2. The test frame has to be improved. When the test is done again the diagonals within the frame, now these are tension straps, have to be replaced with Bosch profiles.

3. The original tests of the undertray diffuser under the URE08 have to be done to get the most reliable data. The downforce can be measured in two different ways. The travel of the dampers can be measured and out of this the downforce can be calculated. Or the downforce can be measured with strain gauges on the undertray diffuser brackets. (Appendix C)

4. All the sharp edges have to be made into roundings. This will make sure the airflow will be better underneath the undertray and the drag will be kept to a minimum.

5. The eventual undertray has to be embedded into the monocoque. Because the undertray will always have a certain thickness, there will be a transition from the monocoque to the undertray. This can only be solved by making it one part with the monocoque or by making it attachable but the undertray starts at the sides of the monocoque. When the second option is chosen the transition from monocoque to undertray has to be as smooth as possible. 6. The length of the Gurney Flap has to be revaluated. There was not enough time in this

graduation internship to do this, but in chapter 4.2 is explained that the Gurney Flap is too long.

(35)

Resources

[1] D.H.J Willemsen - CFD-based aerodynamic optimisation of a 2D race car diffuser – TU/e report R-1812-S (2012)

[2] Wolf-Heinrich Hucho. (1998). Aerodynamics of Road Vehicles (Fourth Edition). Warrendale, Pa: Society of Automotive Engineers, Inc.

[3] Simon McBeath. (1998). Competition car downforce. California: Haynes North America, Inc. [4] V. Samuntran. (2000). Vehicle Aerodynamics. Warrendale, Pa: Society of Automotive Engineers, Inc. [2] http://www.f1technical.net/articles/10 February 2013 [3] http://www.formula1-dictionary.net/ground_effect.html February 2013 [4] http://eprints.soton.ac.uk/48823/1/Internal_Report_AFM_07_06.pdf February 2013 [5] http://www.symscape.com/blog/secrets_of_diffusers February 2013 [6] http://eprints.soton.ac.uk/42969/1/GetPDFServlet.pdf February 2013 [7] http://dave.ucsc.edu/physics195/thesis_2010/guarro_thesis.pdf February 2013 [8]http://www.aero.calpolysae.org/technical/Vortex%20Behavior%20in%20F1%20Underbody%20Co nditions.pdf February 2013 [9]http://www.aero.calpolysae.org/technical/Edge%20Vortices%20for%20Underbody%20Diffusers.p df February 2013 [10] http://www.sciencedirect.com/science/article/pii/0167610586900875# February 2013 [11] http://books.mcgraw-hill.com/EST10/site/spotlight/automobiles/articles/Race-CarAerodynamics.pdf February 2013

Figures

{1} Database URE {2} http://www.formula1-dictionary.net/diffuser.html {3} Presentation of URE {4} Presentation of URE {5}http://www.globalspec.com/learnmore/sensors_transducers_detectors/flow_sensing/gas_flow_s witches_volumetric

{6} D.H.J Willemsen - CFD-based aerodynamic optimisation of a 2D race car diffuser – TU/e report R-1812-S (2012)

{11}https://www.politieacademie.nl/onderwijs/Documents/Plattegrond%20CHM%20Lelystad%2001 -12.pdf

(36)

Figures

Figure 1 : URE07 {1} ... 6

Figure 2 : Ferrari Enzo with undertray diffuser {2}... 6

Figure 3 : Skid pad event {3} ... 8

Figure 4 : Example of a 1km track used on the autocross, endurance and efficiency event {4} ... 9

Figure 5 : Venturi tube {5} ... 10

Figure 6 : Undertray diffuser ... 10

Figure 7 : Contours of static pressure near the optimized undertray diffuser (ANSYS) {6} ... 13

Figure 8 : Contours of velocity magnitude near the optimized undertray diffuser (ANSYS) {7} ... 13

Figure 10 : Negative vertical lift per geometric component for different inlet velocities {9} ... 14

Figure 9 : Drag per geometric component for different inlet velocities {8}... 14

Figure 11 : Flow trajectories for the 18.9⁰ undertray diffuser with an air and ground speed of 50km/h ... 15

Figure 12 : Flow trajectories for the 7⁰ undertray diffuser with an air and ground speed of 50km/h ... 15

Figure 13 : Flow trajectories for the 18.9⁰ undertray diffuser with rounded edges (500mm radius) and with an air and ground speed of 50km/h ... 15

Figure 14 : Contours of static pressure near the 18.9⁰ undertray diffuser (Flow Simulation)... 15

Figure 15 : Contours of static pressure near the 7⁰ undertray diffuser (Flow Simulation)... 15

Figure 16 : Contours of velocity magnitude near the 18.9⁰ undertray diffuser (Flow Simulation) ... 16

Figure 17 : Contours of velocity magnitude near the 7⁰ undertray diffuser (Flow Simulation) ... 16

Figure 18 : Optimized undertray diffuser according to Willemsen [1] {10} ... 18

Figure 19 : First concept according to Willemsen’s [1] 2D design ... 18

Figure 20 Diffuser tunnels at a 15⁰ angle, with the holes for the suspension rods. ... 19

Figure 21 : Assembled 21⁰ undertray diffuser ... 19

Figure 22 : Optimized 3D model (α=21⁰) ... 20

Figure 23 : Optimized 3D model (α=7⁰) ... 20

Figure 25 : 7⁰ undertray diffuser prototype placed underneath the URE08 ... 20

Figure 24 : 21⁰ undertray diffuser prototype shown from different angles ... 20

Figure 26 : Rear bracket ... 22

Figure 27 : Front bracket ... 22

Figure 29 : Back of the road test setup ... 24

Figure 28 : Front of the road test setup ... 24

Figure 30: Mounted watertight sealed 1KN load cells ... 24

Figure 31 : Schematic principal of the watertight sealed 1KN load cells ... 25

Figure 32 : Test track of Police Academy in Lelystad {11} ... 26

Figure 33 : Mounted 7⁰ undertray diffuser prototype ... 27

Figure 34 : Mounted 21⁰ undertray diffuser prototype ... 27

Figure 35 : COP removed from the start of the diffuser to the rear ... 28

(37)

Tables

Table 1: Drag and downforce for ANSYS and Flow Simulation with a 7⁰ and 18.9⁰ undertray diffuser at 50km/h ... 17

Table 2 : Centre of Pressure for the given undertray diffusers ... 29

Table 3 : The drag for the undertray diffusers at 80km/h taken out of the Flow Simulation simulations ... 30

Table 4 : Downforce at 80km/h out of ANSYS, Flow Simulation and the undertray diffuser tests ... 30

Table 5 : Downforce at 80km/h out of the different simulation programs and the undertray diffuser tests ... 32

Table 6 : The drag for the undertray diffusers at 80km/h taken out of the Flow Simulation simulations ... 32

TU Racing

Acknowledgement

Foreword

Summary

Table of Contents

Introduction

1. Formula Student competition

Static Events

Dynamic Events

FSAE rules

2. Working principle of the undertray diffuser

3. Set of requirements

FSAE requirements

URE requirements

4. Computational Fluid Dynamics (CFD) simulations

4.1 ANSYS 2D CFD simulations

4.2 Solidworks Flow simulations

4.3 Comparison of ANSYS and Flow Simulation

5. Realization of the undertray diffuser

5.1 2D model according to Willemsen [1]

5.2 3D model according to Willemsen [1]

5.3 Optimized 3D models

5.4 Undertray diffuser prototypes

6. Test possibilities

6.1 Small-scale wind tunnel tests

6.2 Full-scale wind tunnel tests

6.3 Road tests with an URE car

6.4 Road tests with a test setup

7. Testing

8. Validation of the tests

8.1 Centre of pressure

8.2 Downforce

Downforce undertray diffusers

8.3 Drag

8.4 Comparison with ANSYS and Flow simulation

8.5 Downforce drag ratio

9. Conclusion

Recommendations

Resources

Figures

Figures

Tables