We performed all of our wind tunnel testing in the University of Michigan Aerospace Engineering's 5- by 7-foot wind tunnel. This closed circuit, single return tunnel was designed in the 1950s in combination with a six component force and moment balance to provide a low-turbulence testing facility. While showing it's age, the tunnel and balance are still more than capable of yielding accurate data for our project.

The image shown at right was taken inside the block house containing the 1200hp motor/generator set that drives the wind tunnel at up to 170mph.

Also near the wind tunnel is the Model Workshop, where we performed last minute modifications to our mounting hardware.

(1200hp AC->DC generator)

(Last minute modifications)

Producing accurate wind tunnel results is more difficult when dealing with automotive, as opposed to aircraft, models. A is in ground effect all the time, and neglecting the presence of the ground only a few inches away from the vehicle is not a good approximation. Placing the model on the bottom of the wind tunnel is also not acceptable, as the boundary layer in the test section could be any where from a few inches to a foot thick.

The boundary layer problem can be solved in a variety of ways, including stationary or moving ground planes, tangential blowing, and the creative use of suction. Suction and blowing have been tried before in the 5x7 without success, so part of our origional project goal was to build a moving ground plane (glorified treadmill) capable of matching tunnel speeds up to 150mph. Research revealed some discouraging information, including a 120mph ground plane designed for 1/2 scale models requiring a 300hp electric drive motor. Needless to say, we scrapped the idea of a moving ground plane and selected an existing solution.

The 5x7 tunnel may be fitted with a ground plane, providing needed access to the force balance while raising the model about one foot off of the ground and, hopefully, out of the boundary layer.

The model is supported by an H-frame, which is connected to the force balance and has 4 pins that stick up through holes in the ground plane. We encountered many problems while mounting our 96lb model, including:

• Mounting position 1 requires new mounting plates for H-frame and the new holes drilled in the ground plane
• Calibration reveals that holes are too small, causing the mounting pics to rub and invalidate data, cut larger holes in ground plane
• Model is heavy, and supporting it at the ends of the H-frame causes 1/2 inch aluminum H-frame to bow and rub on sheet metal. Mounting position 2 adapted closer to the center of gravity.
• Warped sheet metal touches H-frame, must be removed and repaired

After 25 hours of troubleshooting and repair, the model was finally mounted, the force balance calibrated, and the wind tunnel ready for testing.

(Types of boundary layer elimination - A. Filippone, Aerodyn.org)

(Drive motor/drum set from ARC tunnel - Tom Hintz)

(2-way adjustable H-frame with custom mounting plates)

(Mounting position 1, front and rear bumpers)

(Mounting position 2, between wheel wells)

With the initial steps complete, we moved on to wind tunnel testing.

We tested two spoilers and one wing and multiple angles of attack, including a base line with no modifications. We chose to test at only three velocities for each configuration, as our time table had been destroyed by mounting troubles, at 80, 100, and 120 mph.

Testing proceeded as planned, with the exception of the computerized data collection software. The software was intended to sample hundreds of times in a few seconds, and return the average value. Unfortunately, the software generated average was obviousely inconsistant with the instantaneous values displayed alongside. Our data analysis attempts to uncover trends in the data and rectify this problem.

Despite fears that our model might take flight in the wind tunnel, the model and mounting hardware survived 120 mph in the tunnel without incident. We had to back off the speed only once, when the highest ride height and most severe angle of attack caused worrisome vibrations.

(Adjusting ride height)

(Model mounted and ready to test)

(Installing the wing)

(Installing a spoiler)