scope:

Description of Driving Event:

A mishap occurred with the Propeller Test Rig (PTR) in the 40'x80' Wind Tunnel at an airspeed of about 200 knots. The model was a 25' diameter, three bladed prop rotor, and it was rotating at a speed of about 500 rpm before the mishap.

The mishap was caused by failure of a bearing set in the collective pitch control system during a run in which several data points had been taken and new conditions were being set up. The rotor torque went from about 10,000 ft-lb to ñ2,000 ft-lb (model motors went from acting as motors to generators), and the rotor control system locked up as it was designed to do at ñ2,000 ft-lb of torque to prevent actuator failures from causing damage to the PTR. At this time it was recognized that there was a problem, and the designated procedure was followed which was to open the breaker to the model motors. Opening the breaker (although it had happened automatically) had been successful during an incident that occurred on March 21, 1991, and it was thought that this was a similar event. As soon as the breaker was opened the rotor began to accelerate. Because the rotor blades had gone to a lower blade angle (unknown at the time) the rotor accelerated to a very high speed. Redlines were rapidly indicated on the Bar Chart Monitor, and as soon as this occurred the Wind Tunnel Emergency Stop was initiated. However, the failure had progressed too far, and the rotor continued to speed up until it self-destructed due to overspeed. The rotor reached a speed of close to 1000 rpm; the safe operating limit was 630 rpm, and the structural limit for the blade tension straps was 930 rpm. At overspeed, one blade tore loose and lodged in the top of the test section. The remaining rotor and mast assembly tore loose from the model drive system due to the imbalance. The mast landed on top of the Rig, and the rotor assembly went down the tunnel coming to rest against the safety fence. Some of the debris went past the first fence, but most was collected against the fence attached to vane set 5 which is the last vane set ahead of the Wind Tunnel drive fans. The only damage to the Wind Tunnel drive was a small gouge in one of the blades.

1. Ensure that adequate design and stress analysis have been done and documented for all critical components in models and test hardware. Insure that all components have sufficient load capacity.

2. Perform comprehensive design review and instrument accordingly for health monitoring. An important goal is to do a comprehensive FMECA by knowledgeable systems engineers or designers, prioritize potential failures based on probability, look at consequences of failure in priority order, and then develop instrumentation and emergency procedures on this basis. This is a difficult task, but when there are many potential single-point failures and the consequences of failure is high, it is worth significant effort. Bearings and fasteners are notorious failure points (even if not highly loaded). They must be monitored, maintained, and replaced accordingly.

3. During design reviews, do not overlook parts designed by contractors even though this may be difficult for proprietary reasons. Because some parts have high safety factors does not mean they all do. In fact, hardware that has been used and modified several times is very likely to have a large variation in safety factors.

4. If control systems are designed to lock up to preclude further inputs, Wind Tunnel Emergency Stop should be the only course of action when lock up occurs.

5. Minimize voice directions during emergencies.

6. When open loop control systems are difficult to control, they should be automated (loop closed).

7. When doing research in which control rates are sensitive, human factor considerations become important and must be considered.

8. There should be inspections or verification during assembly and maintenance procedures at critical points. It should not be up to one person to be responsible for the entire assembly or maintenance. Documentation must be comprehensive.