Wind Tunnel Testing at the University of Illinois.
Obective
Identify modifications to the wheelchair which reduce the total drag on the chair and athlete
Approach
Design a partial fairing to guide the air smoothly around the athlete
Test Equipment
Wind Tunnel: University of Illinois subsonic, open-return tunnel
Test section: 3 ft. high x 4 ft. wide x 8 ft. long
Max Speed: 160 mph
Turbulence Intensity: < 0.1%
Power source: 125 hp motor driving a 5-bladed fan
Speed Control: Variable Frequency Drive
Instrumentation: Force balance used to measure wheelchair drag, uncertainty of +/-2.5%
Fig. 1 Diagram of the UI 3 x 4 ft. subsonic wind tunnel.
Wheelchair Models
Two wind tunnel models were designed based on measurements provided by Invacare: a small, 1/5th-scale model and a larger, 1/2-scale model. Both models were constructed using standard rapid-prototyping techniques. The purpose of the smaller model was to ensure that the wind tunnel instrumentation was capable of detecting small changes in drag at low speeds, as it was originally designed for airfoil testing at much higher speeds. Three fairings were designed for the small model: a small nose fairing, a large partial fairing, and a full fairing. Since detailed data were not required for this preliminary test, the athlete was modeled out of clay.
Fig. 2 Small, 1/5th-scale wheelchair model used in preliminary testing.
The larger, 1/2-scale wheelchair model was much more detailed than the small model, and contained nearly all of the geometric intricacies present on a full-scale chair. A mannequin of a two-year old was used to represent the athlete; it was 33 inches tall, so represented a full-scale 5'6" athlete. The mannequin was essentially a foam body over a wire skeleton and could be bent into almost any position desired. Once in the appropriate position, tape was used to secure the mannequin in place. Fairings for the 1/2-scale model were constructed from commercially available off-the-shelf materials.
Fig. 3 Larger, 1/2-scale wheelchair model used to determine aerodynamic drag on a wheelchair.
Results
Wind tunnel tests on the 1/5th-scale model showed that the existing wind tunnel instrumentation could measure reasonably small changes in drag. At higher tunnel velocities, these measurements were repeatable and drag coefficient was almost constant over a range of speeds. Reductions of nearly 10% over the unfaired chair were observed using the fairings built for the small model. Given these results, design and construction of the 1/2-scale model began.
Measurements made on the larger model were in good agreement with those made on the smaller model. The drag area, CdA, of the empty chair scaled to 1.05 ft^2 for the full-size chair. When the mannequin was added, CdA increased to 2.96 ft^2. These results indicate that the athlete contributes more than 60% of the overall drag! They are also in good agreement with published data for a person crouching, in which case CdA is about 2-3 ft^2.
The rest of this page shows various configurations that were tested and the measured CdA for each. Values of CdA for configurations that resulted in reduced drag over the baseline case are reported in green font. CdA values for configurations within 2.5% of the baseline case are reported in black font, as these are within the uncertainty of the experimental setup. At the end of the page is a summary showing the percent reduction in CdA for each configuration.
Baseline Case
CdA = 2.96 ft^2
(avg. of several runs)
Fairing behind wheelchair cage
CdA = 2.99 ft^2
Feet tucked under athlete
CdA = 2.84 ft^2
Large partial fairing
CdA = 2.78 ft^2
Small partial fairing
CdA = 2.63 ft^2
Athlete in tucked position
CdA = 2.52 ft^2
Athlete in tucked position with best fairing
CdA = 2.24 ft^2
Fairing around wheelchair cage
CdA = 2.95 ft^2
Summary
The percent differences for the cases shown above, as well as some others not discussed, are summarized in the following chart. Configurations which have bars extending to the left of the left dashed line had lower CdA than the baseline case. Configurations which have bars extending to the right of the right dashed line had higher CdA than the baseline case. Configurations with bars between the right and left vertical dashed lines were within the experimental uncertainty of the setup and are considered to have CdA very similar to the baseline case.
Conclusions from Wind Tunnel Testing
Several "windshield" style partial fairings were tested, all of which reduced CdA and the best of which reduced CdA by almost 12%. The fairing had a similar effect whether the athlete was in a tucked or a more upright position. Even without a fairing, the athlete can reduce CdA by almost 15%, and combined with the best fairing tested, CdA is nearly 25% below the baseline case.
Over a 5000-m race, a difference in CdA of just 10% can result in a savings on the order of 2 seconds, depending on the conditions of the course. This usually represents a substantial fraction of the time separating the top finishers. For example, in the 2008 Beijing Paralympic Games Finals, the first and last place finishers were separated by 3.12 seconds in the women's 5000-m and by 3.77 seconds in the men's 5000-m. This would suggest that the fairing could potentially have a significant effect on the race results. However, it is important to remember that many other factors also play a role, notably drafting, which may mitigate the effect of the fairing in a race situation.
