Page:CAB Aircraft Accident Report, Allegheny Airlines Flight 371.pdf/5

- 5 - card was held firmly by the heat-distorted front end assembly and indicated a heading of approximately 88 degrees. It has been demonstrated that the compass card of the repeater indicator is subject to random displacement when its electrical power has been removed. An attempt was made to determine the heading information being transmitted electrically to the repeater indicator by the MDI at the time of impact; however, impact forces had dislodged the transmitting autosyn of the MDI from its mounting and the rotor shaft, although visibly bent, was free to rotate and it was impossible to establish the angular position of the rotor at the time of impact.

In the examination of the fluxgate transmitter it was found that the lower housing of the unit had incurred impact damage and that while the reacted normally to the erection caging functions it eventually precessed slightly in the pitch axis owing to contamination in the unit introduced as a result of exposure to the elements and contaminants subsequent to the crash. The gyro was caged to preclude gyro tilt and measurements of indicated heading were made on a test stand at 30-degree intervals from 0 to 360 degrees. Errors in the indicated headings ranged from +4 degrees at 240 degrees to -5 degrees at 60 degrees.

During the investigation of this accident it was determined that an inadvertent caging of the fluxgate compass was possible owing to the position in which the caging switch was mounted, coupled with the fact that no guard was provided for this switch. (See Attachment "B")

The purpose of this caging switch is to rapidly erect the gyro in the transmitter unit located in the left wing. This erection system erects the gyro to a vertical position in respect to the aircraft only, and is independent of the rolling-tall type erection mechanism incorporated into the system which erects the gyro to a vertical position relative to the surface of the earth. The caging of this gyro is normally accomplished on the ground prior to flight and with the aircraft in a level position

During the course of the investigation two flight tests were conducted in Martin 202 aircraft with identical fluxgate compass equipment installed. During these tests momentary depression of the fluxgate compass caging switch resulted in the initiation of the caging cycle which required five to six seconds and caused erroneous heading information to the right of up to 90 degrees. These errors were obtained as a result of depressing the caging switch while the aircraft was in a right bank to correspond to the right landing pattern approach of the aircraft at Williamsport.

In addition to the two flight tests performed in the Martin 202 aircraft, similar tests were conducted in a DC-3 aircraft which had similar fluxgate compass equipment installed and tests were also conducted with Similar type equipment installed in a Helmholtz test stand. The data thus collected were analyzed and it was determined that the magnitude and direction of the error in indicated heading is a function of the bank angle and aircraft heading at the time of caging. Completing the cage-uncage cycle while in a banked attitude to the right and while on a southeast or southwest heading produced a positive error in indicated heading. A heading to the left of that intended would be flown if the pilot were to use the compass system to maintain a given heading. Conversely, if the cage-uncage cycle was completed while on a northwest or northeast heading the error would be negative and the actual heading would be to the right of that indicated.

Investigation showed that the secondary or standby means of directional information which was available to the captain was an air-driven directional gyro. This unit appeared to be capable of normal operation prior to the accident.