possible that a static system error caused the barometric altimeter to read higher than the actual altitude of the aircraft and produced a decrease in the indicated rate of descent on the vertical speed indicator. in these circumstances, the pilot would reduce power and possibly lower the nose of the aircraft in order to regain the desired rate of descent. This in turn would result in the aircraft's being lower than indicated on the altimeter, and descending at a rate greater than that displayed on the vertical speed indicator.
The existence of an error such as that described above is consistent with certain indications on the flight data recorder. For example, there are several increases in the rate of descent recorded by the flight recorder during the final approach indicating that the captain may have been attempting to compensate for the lower-than-actual rate of descent. Since these descent rates were all in excess of 1,000 feet/minute, the absence of any required callout would support the premise that the vertical speed indicator was reflecting a rate of descent lower than the actual descent rate. Moreover, during the last 10 minutes of flight, there were two instances in which the flight recorder reflected descents which resulted in overshoots followed by gradual returns to the desired altitude. We recognize that these overshoots may have resulted from either the pilot's technique in the manual operation of his flight controls or the use of the aircraft autopilot. It is possible, however, that these overshoots could be symptomatic of a lagging of the aircraft instruments due to an error within the static systems. It is also conceivable that there could be an error in one or all static systems such that it would manifest itself while the aircraft was descending but not after leveling off. The first of these descents was 175 feet and resulted in an overshoot of 50 feet. The second descent, which occurred at 6 minutes 7 seconds before impact, was 575 feet with a resultant overshoot at 150 feet. Both of these overshoots were corrected by a gradual climb back to the desired altitude. The ratio of the amount of overshoot to the total descent is 0.286 and 0.261, respectively, or 26 to 29 feet for each 100 feet of descent. The final descent of 1,200 feet with an apparent overshoot of 318 feet results in an error ratio of 0.265 or 27 feet per 100 feet of descent, which closely parallels the error ratios of the two earlier overshoots.
With respect to physical evidence pertaining directly to the barometric altimeters, it appears that both were correctly set at 29.67, thereby eliminating any indicated error from that source. The displacement of the outer and inner drums, 600 feet and 3,000 feet for the captain's and copilot's barometric altimeter, respectively, was the result of impact forces overcoming the light spring tension holding them in place. Since the drum assemblies in both altimeters were essentially identical, the variation in displacement is attributed to the difference in impact forces encountered.
- 26 -
