17 May, 1999

We have run the following proof tests on portions of the airframe:

Landing Gear Attachment proof test - the landing gear was installed in the booms, the fuselage mounted to the floor such that it could not move, and the landing gear pressurized with a hydraulic pump. The purpose was to test the aluminum mounting points bonded into the fuselage and booms. The nose gear was loaded with 5000 pounds and completed the test. The main landing gear was to be loaded with 9000 pounds per wheel and the mounting points began to fail at 6000 pounds. Therefore, the test was halted until repairs and reinforcement could be made. Improvements were completed, the test was redone, and passed. The attachments were also improved to prevent galvanic corrosion between the aluminum bosses and the carbon structure.

Wing Bend proof test - The wing was installed in the fuselage and bend tested to 4 Gs. At 2.25 Gs a small area of the top skin initially showed signs of buckling failure. The area was cut open for inspection and it turned out the carbon cloth in that area was too dry and did not bond to the foam sandwich. This area was reinforced and the test was successfully completed to 4 Gs. The wing was retested after the landing gear attachment proof tests to assure there was no damage.

Fuselage pressure test - the fuselage pressure vessel was completely filled with approximately 6000 pounds of water, then pressurized to over 24 PSI (26 PSI at the fuselage bottom). Initially, the fuselage pressurized door leaked due to slippage of the fuselage/door engagement mechanism. By installing cam engagements locking the door and the fuselage together (as was always planned for the final aircraft), this problem was solved and the fuselage completed the test. The maximum flight pressurization will be 10 PSI. The fuselage/wing fairing attached to the pressure vessel came unbonded during the pressure test due to the stretching of the pressure vessel, but this has now been fixed.

Propeller Tests - the version 3.1 prop was tested and failed at 6000 engine RPM. It was determined that the failure was caused by missing layers of unidirectional carbon in one of the propeller skins (a quality control problem not a design problem). The failure caused substantial damage to the test stand but this provided an opportunity to improve the design of the components that needed to be remade. Takeoff RPM in the actual aircraft is expected to be 5500 RPM, but the propeller will be tested rigorously at overspeed conditions on the ground to verify the desired safety margins.

Propeller Tests - the version 3.2 prop was tested and failed at 6500 engine RPM (with the tip travelling at Mach .982). This failure was determined to be due to a combination of blade weave due to the the high tip speed, softness in the pitch linkage, and pitch links weakened to provide clearance to set the pitch low enough to achieve a very high RPM. The test stand was again damaged but not nearly as extensively as in the version 3.1 propeller tests. Propeller 3.3 is now being built with an improved pitch link design which increased stiffness by an order of magnitude and allows the pitch to be reduced without reducing the strength of the system.

Propeller Tests - the version 3.3 prop was tested and failed at 6550 engine RPM. The failure was determined to be because of tip delamination due to failure of the bond between the skins and the foam core. The foam had been substituted for flexibilized epoxy to reduce the weight in the trailing edge of the blades in an effort to reduce the blade weave tendency. As it turned out, the increased pitch link stiffness reduces the blade weave loads to almost zero. The next prop will have solid resin at the tip and an improved counterweight arm design.

Propeller Tests - the version 3.4 prop was tested to 6000 engine RPM and passed. Actual operations will be limited to 5500 RPM (2300 prop RPM). The prop has now been run for several hours at various RPMs with no problems. The test procedure for all prop tests since version 2.1 was to run at engine RPMs of 4500, 5000, 5500, 5750, 6000, 6250, and 6500 (tip mach 1.0). At each RPM the prop was run for four 15-second increments, then four 30-second increments, then four 1-minute increments, the three 2-minute increments with inspections of the prop and test stand after each run before progressing to the next RPM level. A high speed stripchart recorder was used to analyze the pitch link data.

Rotor Head attachment proof tests - the rotor head was installed in the fuselage, and pulled with 12,000 pounds (simulating a 4G loading), passing the test.

Thrust mount tests - the fiberglass structure through which all the propeller thrust loads are transferred to the fuselage was proof tested at 6000 pounds (4 times the maximum expected thrust), and passed.

Propeller thrust tests - to determine the propeller RPM for takeoff, we measured the thrust at full throttle at several different engine RPMs. There was very little difference in thrust between 5250 and 5500 RPM - about 1150 pounds of thrust at 240 horsepower. 5400 RPM was selected as the maximum takeoff RPM. The propeller pitch at flat pitch was adjusted so that the engine RPM cannot exceed 5400 RPM. Earlier tests to 6600 RPM proved the propeller spar capable of 1.5 times the centrifugal load at 5400 RPM. (Static pull tests at 3 times maximum centrifugal force with 45 degrees of twist were also conducted.)

Rotor tests - the rotor was instrumented and tested at several RPMs, looking at the cyclic control loads, the collective loads, the pylon movement both fore/aft and side to side, the lift, flapping, and control system hydraulic pressures. The maximum RPM reached with the propeller also running was 540 RPM, at approximately 88 horsepower. Without the propeller, the rotor went to 580 RPM (Mach .91) at approximately 145 horsepower. Compressibility effects at the rotor tip were causing a steep increase in power required for a small increase in RPM. Therefore, jump takeoffs will be performed at 525 RPM (Mach .825).

Nosewheel Shimmy Tests - the nose gear was attached to a trailer towed behind a truck. The trailer carried two 55-gallon barrels. The test was performed at various weights and speeds up to 55 MPH, with the nose gear at various stages of being raised and lowered. No shimmy was detected in any condition. The nose gear tracked corners well.

Simulated jump takeoff tests - the rotor was prerotated to 525 RPM (3500 engine RPM) with the engine at full throttle, then the clutch disengaged. The engine immediately goes to 5400 RPM as limited by the propeller. The collective is pulled to hold at least 3500 pounds of lift. The cyclic stick is moved in a circle until the lift decreases to 1000 pounds, at which point the collective is returned to zero. The throttle is held full open for one minute. Then the clutch is engaged for another cycle. 250 cycles like this were performed, with thorough inspections every 5 cycles. On 1/20/98, on the 19th cycle, the guard support for the propeller drive shaft pulley failed due to poor welding and went through the propeller, destroying the version 3.4 propeller. The test stand was heavily damaged but the rotor was not damaged. Testing recommenced on 2/20/98 with a new prop and repaired test stand, and on 3/5/98, 250 cycles were completed. All component proof tests have now been completed!

In Aircraft Engine Tests - the engine, drive train, and propeller were tested in the aircraft in preparation for taxi tests. The engine cooling on the ground was much better than expected. Even on a 90 degree day, the engine could sustain 600 pounds of static thrust and short bursts of 1200 pounds thrust without exceeding 210 degrees coolant temperature. As intended, the propeller sucks enough air through the annular exit around the spinner to effectively cool the radiator. Engine compartment temperature actually dropped at full throttle, but immediately after shutdown, the proximity of the exhaust to the fiberglass cowl caused smoking of the cowl because there was no more cooling airflow. This required installation of additional insulation. The propeller spinner began to fail due to hoop loads pulling out the screws holding the spinner halves together. The spinner is being fixed by bonding the two halves together and splitting the spinner in the more conventional manner in the plane of rotation of the propeller.

Return to Press Release