Wichita Falls, Texas: Friday, 5 October 2001

By Jeff Keaton. Edited by Jay Carter, Jr.

Failure of Modified Propeller Leads to Design Improvement

In preparation for flight-testing scheduled September 21, 2001, the CarterCopter Technology Demonstrator (CCTD) suffered significant damage during a qualification proof test on a modified propeller in the "pit." The propeller -- which had been newly constructed as a result of damage that occurred August 5, 2001  abruptly failed, resulting in damage to the tail boom, composite landing gear support, fuselage, and numerous metal parts. The prop drive shaft was bent so badly that the pre-rotator pulley tried to engage the rotor drive with enough torque to break the main U-joint at the rotor head (which had been previously proof tested to 3000 foot-pounds). The prop failed at an engine RPM of 4000 during a full throttle acceleration. In prior tests, the prop had run up to an engine RPM of 5750. Since the rotor was not turning and the prop spinner was not installed, these parts were not damaged.

The propeller spar was fabricated using KEVLAR^® instead of carbon fiber because KEVLAR^® is tougher and less brittle than carbon. The KEVLAR^® was compared against carbon in side-by-side tests for tensile and shear strength with similar performance. There was a potential risk in buckling due to the fact that KEVLAR^® has about 2/3 the stiffness of carbon -- but since most of the loads would be in tension this was not expected to be a problem. However, a full throttle acceleration (maximum torque, but reduced centrifugal force due to the lower RPM) with the prop at flat pitch (spar twisted the maximum amount), caused the spar to buckle. The spar did not fail in tension due to centrifugal force or in shear due to a bond failure.

Required repairs include making a new prop, fixing the composite fuselage and landing gear support, machining 17 aluminum and steel parts, and repairing the starboard tail boom that was nearly severed by one of the prop blades.

Prop failure sequence at 30 frames per second -- 18 September 2001

1/30 sec. Prop spar is starting to fail	2/30 sec.
3/30 sec. 1 blade has departed. 1 blade is cutting tail boom	4/30 sec. Prop shaft is severely bent
5/30 sec.	6/30 sec. Note tumbling blade in background
7/30 sec. Note top antenna movement	8/30 sec. Piece of KEVLAR® spar in foreground
9/30 sec.	10/30 sec. Note upper cowling distortion
11/30 sec.	12/30 sec.
13/30 sec. Rotor moved aft from applied torque	14/30 sec.
15/30 sec.

The failure of the KEVLAR^® spar suggests that although the carbon spar never failed, it was marginal. This test highlighted a potential problem that could have eventually occurred in flight at high altitude and very high speed -- the only flight regime with a combination of low prop RPM, maximum pitch/spar twist, and high engine power/torque.

Test Results and Other Design Enhancements

Some useful test data was obtained prior to the prop failure. For a given constant high horsepower setting, static prop thrust was better at higher RPMs which indicates the prop RPM needs to be increased to more efficiently handle higher engine horsepower. A higher RPM reduces the prop angle of attack and amount of blade stall. This also explains why the aircraft performed better at higher engine/prop RPMs. The ratio of engine-to-propeller RPM will be changed from 2.75:1 to 2.387:1. This should improve static thrust, climb performance, and in-flight prop efficiency at higher power settings -- and increase the centrifugal force for a given torque to help stabilize the spar and avoid a buckling problem (centrifugal force is a squared function of RPM).

Test data also indicates that the revised prop controller algorithm is working well.

A new propeller has since been rebuilt, but the new prop spar was constructed out of carbon with the addition of a carbon shear web to greatly improve the column (buckling) strength. The spar winding process was also improved, which increased the carbon content by 5% and provided better wet-out of the strands. This new prop will be tested on the steel test stand rather than on the aircraft to limit exposure to the aircraft.

Additionally, for added protection both tail booms will be reinforced with KEVLAR^® armor in line with the propeller plane of rotation. All enhancements and repairs are progressing on schedule and flight-testing is scheduled to resume in mid October.

Events of September 11, 2001 Force Pilot Changes

In support of recent military preparations, Rusty Nance will not be available to serve as Chief Test Pilot for the next series of flight-tests. Larry Neal, who has been flying with Rusty as copilot, will move into the left seat and be the pilot. Professional test pilot Tom Currie will fly as Principal Copilot and Flight Test Engineer with Paul Smith and Carl Hawkins as backups.

KEVLAR^® is a Dupont registered trademark.