Sunteți pe pagina 1din 2

Case Study

www.nafems.org

CFD for Fun


Over the years, CFD analysis has been widely used in the aerospace and automotive industries, but with a little bit of imagination, such techniques can also be used for fun projects Alex Leong, Ford Motor Company, UK, explains.
he original concept of soapbox racing is simple - a homemade car built from orange crates, tin sheets, buggy wheels and almost everything of junk value but it can only be powered by gravity alone. The name soapbox derby was christened back in 1933, when a newspaper photographer called Myron E. Scott encountered three boys racing homemade carts down a back street in Dayton, Ohio. He was fascinated by the amusing spectacle, and arranged a coasting race for the boys and their friends. In the same year, with the financial backing from his boss, Mr. Scott organized another race. This time, 362 kids showed up with their homemade cars and the event attracted more than 40,000 people watching along the hill. Very soon, the Derby blossomed into a fiercely contested competition across North America. At its peak, the All-American Soapbox Derby attracted 25,000 entrants in 120 races and drew an estimated crowd of 1.5 million people nationwide. In Britain, there was also a strong following in this sport and the National Soapbox Association (NSBA) organized annual championships until 1994. In 2000, the organization committee of the Goodwood Festival of Speed, which is already acknowledged to be the worlds biggest annual historic motor racing event, decided to adopt the soapbox racing as one of its major events. In that year, the Goodwood Gravity Racing Club invited 20 teams from both race and road car manufactures, plus a handful of enthusiastic privateers, to take part in the race. And since then, the Downhill Soapbox Challenge has become one of the most talked-about events in the Festival. Although Soapbox racing started from a very humble beginning, the modern version is much more sophisticated. In the Goodwood downhill soapbox challenge, the design and construction of the gravitypowered vehicles are strictly regulated. Apart from stating that no pedal power or push-starts are allowed, the six-page rulebook also specifies things such as the overall size, shape, driving positions etc. There are two classes of soapbox, either open-topped roadsters or roofed streamliners. The competition comprises of Page 30

two timed runs for each competitor with the result decided by the lowest sum total time of both runs. Although the downhill course is just over 1.1 km long, it will take most of the competitors about 80 seconds to descend and the vehicle will notch up speeds in excess of 95 km/hr out of the final corner. Aerodynamic efficiency, low friction losses and handling finesse are the critical factors of a winning design. Although this may not be the organisers original intention, with only top engineering firms now being invited to take part in the race, the race is not just about winning but the prestige of the company is also at stake. As a result, the competition is fierce and intense. However, this is also a good opportunity to train the new recruits (of those companies) who may not have been involved in any competitive projects before. In Ford, most of the team members are drawn from first and second year engineering graduates who undertake projects such as this to further their knowledge of the company and enhance their engineering skills. This is also reflected in the name adopted by Fords entry: Centennial EGGbox. The acronym of EGG stands for Engineering Graduate Group and 2003 was the centennial year of the company. The Ford team consisted of 14 core members (the author is one of the senior members who provided critical support to the team). Each member was assigned to a group that was responsible for certain aspects of the vehicle, such as the design and construction of the body, wheels, brakes etc. As shown in the picture, Centennial EGGbox is an open-topped design. The streamlined body is built from carbon fibre with a Nomex constructed chassis. The wheels used in this design are 20-inch conventional spoke wheel with wheel covers to reduce aerodynamic drag. However, due to package and cost restraints (a familiar concept within a car company!), it was not possible to use CFD to conduct a complete design iteration exercise, instead CFD was used as a tool to study the airflow around the vehicle and to provide quantitative data for further improvement of the basic design.

Figure 1: The Ford Centennial EGGbox

January 2005

www.nafems.org

Case Study
In the CFD study, the full sized model of the vehicle was simulated at a zero yaw angle (i.e. on level ground). Only half of the computational domain was meshed and a symmetry boundary condition was applied along the center plane of the vehicle. In order to provide a suitable inflow and outflow conditions, the flow domain was extended to an upstream and downstream location approximately 0.5 and 1.5 times the vehicles length, respectively. As an opentopped design, the driver would be sitting in an inclined position with his body partially exposed to the airflow. To take this into account, the CFD model includes the driver and his helmet. However, the parts of the driver body that are shielded by the vehicle were excluded. The ground and the body of the vehicle (plus the driver) are treated as fixed wall boundaries but a moving wall boundary (with a rotating speed corresponding to the forward motion of the vehicle) was applied to the wheels. To simulate the boundary layer flows around the wall regions, two layers of prism cells were created which covered all the wall surfaces. As shown in figure 2 trimmed cell technology was used to create the mesh. Based on a near wall cell size of 8 mm and additional refinement in the wake and ground regions, the final mesh contains approximately one million computation cells. A fixed velocity inlet condition and a zero gradient outlet condition were applied. The inlet flow velocity corresponded to the case in which the vehicle was traveling at 95 km/hr. The flow is assumed to be steady state, isothermal, incompressible and turbulent. The widely used high Reynolds k-e turbulence model with default wall treatment was used. The results in figures 3, 4 and 5 demonstrate how CFD can be used to enhance the design. In these pictures, the pressure distribution (expressed as pressure coefficient) on the car and driver body is shown January 2005 as color-filled contours. The superimposed velocity vector plot shows the flow behavior in the central cross section of the vehicle. In initial simulation, the vehicle had no deflector installed. The CFD predicted that a stagnation region was formed at the front face of the helmet but as the flow sped around the helmet, it suppressed the flow stream coming through the underside of the vehicle. A relatively small recirculation zone was formed immediately behind the drivers helmet. In contrast, figures 3, 4 and 5 shows the pressure coefficient and velocity vectors for the case with a small deflector fitted. As expected, the driver and his helmet were shielded by the presence of the deflector. Although the stagnation zone had gone, the strength of the flow within the recirculation zone was much weaker. When this flow was mixed with the stream coming though the underside of the vehicle, the latter dominated the resulting flow and it created a significant amount of turbulence. In fact, the design with the deflector had increased the drag value by 18%. In hindsight, a common sense approach would choose the design with the deflector, however CFD has predicted that if the combined effects of the whole vehicle are taken into consideration, the design without the deflector has a better overall performance. A final note. Despite all the efforts, the Centennial EGGbox failed to reach the top three at the Goodwood race but after some finetuning; the same vehicle won the downhill race at Prescott, Gloucestershire in the same year. Overall, CFD has been successfully used to enhance the aerodynamic performance of the basic design. Furthermore, the valuable lessons learned from this exercise were passed onto the team who prepared the design for the following year.

Figure 2: Computational mesh

Figure 3: Contours of pressure (rear view)

Figure 4: Contours of pressure (front view)

Contact
Alex Leong, Ford Motor Company aleong@ford.com

Figure 5: Flow pattern around car

Page 31

S-ar putea să vă placă și