Aviation Accident Report: Pennsylvania Central Airlines Flight 19/Analysis of Facts

III.

ANALYSIS OF FACTS — DISCUSSION OF POSSIBLE CAUSES

 

In the preceding section of this report we have set out a summary of the evidence disclosed by the investigation. From the facts previously discussed, it appears that the flight of Trip 19 proceeded normally until 2:31 p.m. when it reached the Herndon fan-marker at 4000 feet, climbing. We also know that the crash occurred approximately 25 miles northwest of the Herndon fan-marker about ten minutes later. In searching for the probable cause of this accident, it is necessary to analyze all of the facts in an effort to discover what event or combination of events occurred during that ten-minute interval which resulted in the crash of the airplane.

In seeking to determine the course of events that led to this accident, in the absence of any direct testimony from any actual observer on board the aircraft and in view of the circumstantial evidence of record, we are obliged to look for the extraordinary, and to examine into the possibility and probability of occurrences that are so rare in practical operating experience as to be highly obscure, if not virtually unknown. Nothing within the ordinary range of experience, and no combination of events of which the occurrence could be regarded at all likely, would have sufficed to produce this accident. The past record of safety of air transportation within the United States, extending to over 130,000,000 miles of transport flying without serious injury to any person in the 17-month period immediately preceding the present accident and to thirteen years of operation without fatal accident by Pennsylvania Central Airlines and its predecessors, is a sufficient evidence that only the highly exceptional or the hitherto unknown would produce so tragic an effect. Since we are here necessarily dealing with occurrences of extreme rarity and consequent unfamiliarity and in some instances are considering phenomena of nature wherein the knowledge of men still remains extremely limited, it is inevitable that much of what follows must appear highly speculative. However, the inherent difficulties presented by the limited evidence available in the present case must not deter us from a full exploration of every possibility and an attempt to reach a conclusion as to what probably occurred to produce the accident.

 

Dispatch of the Flight

No reason has been discovered for believing that Trip 19 should not have been dispatched on August 31, 1940. The airplane had received a "60-hour" inspection at Detroit the day before the flight and had received the required "turn-around" inspection in Washington just before taking off and nothing unusual had been reported as a result of either of these inspections. The pilots who had brought the airplane down from Detroit to Washington that morning had not reported anything unusual in its operation. It was adequately serviced with fuel and oil. The load on board at the time of take-off was 174 pounds below the approved gross weight for the airplane and, according to the record, the load was properly distributed, the center of gravity of the airplane being well within approved limits. At the time of take-off there was a company employee riding in the jump seat in addition to the 21 passengers and crew of three. Under existing regulations this is permissible so long as the approved gross weight of the airplane is not exceeded.

The weather information at the time of take-off did not indicate that unusual weather conditions would be encountered. An overcast condition was reported over a portion of the route but this is not unusual in air carrier operation. The captain and first officer were fully capable, by training and long experience, of flying solely by the use of instruments. Inspection had shown, that the navigation instruments on board were in serviceable condition and monitoring reports showed that the radio ranges along the course to be flown were operating normally. The flight was cleared by the Airway Traffic Control Center in Washington in accordance with applicable regulations. Mild thunderstorms along the course were forecast but such conditions are often encountered in this area during the summer and are not regarded as a reason for cancelling flights. Innumerable trips are made with safety in a perfectly routine manner through thunderstorm areas. Due to the local nature and varied character of thunderstorms, the manner of operating, in such areas is left to the judgment of the captain.

The flights of Pennsylvania Central Airlines were dispatched from Pittsburgh for Washington on the same weather forecast as Trip 19, and although they passed within 10 miles of the scene of the accident within a few minutes of the time it occurred, they encountered no severe weather conditions. In addition, a flight of American Airlines passed 10 miles south of the scene of the accident at 2:10 p.m. and encountered no unusual weather conditions.

 

Local Weather Conditions

From the evidence it appears that during the forenoon of August 31, 1940, there were broken clouds and scattered showers in the vicinity of Short Hill and Lovettsville. Early in the afternoon this broken cloud condition changed to a solid overcast extending at least 4½ miles east of the scene of the accident. At about 12:30 p.m., heavy storm clouds were seen around Short Hill. It was impossible to reach a definite conclusion as to the length of this storm area in a northerly and southerly direction. The pilot of the Luscombe who attempted to fly through this area to Pittsburgh at about 12:30 p.m. estimated that the storm extended about 50 miles south of Lovettsville and an unknown distance to the north. He attempted to skirt the storm to the north and proceeded in that direction to a point 5 miles northwest of Myersville, Maryland, about 20 miles north of Lovettsville, without reaching the northern limits of the storm area.

One person from whom a statement was taken after the hearing stated that at about 2:25 p.m. on August 31, 1940, he was at Bluemont, Virginia, which is about 15 miles southwest of Lovettsville, and that the southern limit of the storm area appeared to be just south of Bluemont. He said that the storm was proceeding slowly in a northeasterly direction up the Shenandoah Valley and that he could see it extending far to the north. Upon the basis of the evidence available to us, it appears that at about the time the accident occurred, the southern limit of the storm area was about 15 miles southwest of Lovettsville and that the northern limit was at least that far and probably much farther to the north.

At about 2 p.m. on August 31, 1940, an extraordinarily heavy rain began to fall along Short Hill. This rain gradually spread during the afternoon in an easterly direction until it covered an area extending approximately 4 miles east and southeast of the scene of the accident. It is estimated that when the airplane arrived in this storm area, the rain had reached a point approximately 1¾ miles southeast of the scene of the accident.^[1] This estimate was arrived at by a consideration of the testimony of Mr. Vincell and a statement made by Mr. Harry E. Everhart. Mr. Everhart, who lives 1½ miles southeast of the scene of the accident, stated that it was raining heavily at his place at the time he heard the "roaring of engines". Mr. Vincell, who lives 2½ miles southeast of the point of impact, said that it was not raining at his place when he heard the crash, but that the heavy rain reached his home shortly thereafter.

It is impossible to reach any definite conclusion as to the altitude of the ceiling immediately to the east and southeast of the storm area at the time the airplane reached that area. The two Pennsylvania Central Airlines pilots who flew near the scene of the accident within a few moments of the time it occurred stated that they entered an overcast near Charles Town, West Virginia, and descended through it almost to Leesburg before they broke out at 3000 feet. Other pilots flying further south and southeast of the scene of the accident came out of the overcast at altitudes as low as 1500 feet. On the other hand, Mr. McGaha stated that he saw the airplane flying through scud at an altitude which he later compared to that of a similar type airplane which was flying at a known altitude of 6000 feet. The testimony of the airline and other pilots flying farther south and southeast of the scene of the accident as to the ceiling encountered is not, however, necessarily inconsistent with Mr. McGaha's testimony, since under the weather conditions prevailing on August 31, 1940, widely varying ceiling conditions could be anticipated within relatively small areas.

From the testimony of witnesses and from the physical evidence of clogged brooks, newly-filled reservoirs, and washed-out roads, it appears that the rainstorm was of really exceptional intensity, both at the scene of the accident, immediately west of Lovettsville, and for some distance to the north and south. It appears, in fact, to have been the heaviest rainfall known in the Lovettsville area for a number of years. The storm's electrical characteristics on the other hand, do not seem to have been at all unusual. Many of the witnesses, however, recalled particularly an extremely violent stroke of lightning and clap of thunder immediately preceding the roar of motors which they associated with the accident; and other lightning flashes were noted in the neighborhood, but were not generally recalled as having been particularly vivid.

All of the witnesses in the vicinity of Lovettsville, except Mrs. Leila Shoemaker, testified that there was little or no wind on the ground during the storm. This testimony is corroborated by the fact that the storm was moving very slowly. Mrs. Shoemaker testified that at the time the accident occurred the wind appeared to be blowing from the west. It is probable, since she lives on the slope of Short Hill, that the air currents observed by her were currents within the storm.

The testimony of the Lovettsville witness would indicate that the storm was very severe and of a type which would produce turbulence in some degree. The testimony of pilots flying in that vicinity during the afternoon, which has previously been referred to, is inconclusive in this respect. One of Pennsylvania Central Airlines pilots reported slight roughness as he came of the overcast just west of Leesburg. Neither experienced any turbulence in the vicinity of Lovettsville. An American Airlines pilot who passed in that vicinity about 2:10 p.m. reported only slight turbulence, but an Army pilot who flew through the Lovettsville area about 12 o'clock noon reported severe turbulence along the course, as did another Army pilot who passed 14 miles south of the scene of the accident at about 2:35 p.m.

The experience of pilots flying in the vicinity cannot be relied upon completely in arriving at a conclusion as to the presence of turbulence. Experience shows that of two airplanes flying at the same altitude at the same time, a very short distance apart, one may encounter very severe turbulence while the other may be operating in comparatively smooth air. Therefore, it is impossible for us to arrive at a definite conclusion that the pilot of Trip 19 was flying through conditions of severe turbulence just prior to the crash. There is, however, a considerable likelihood that he was doing so in view of the general trend of the reports of other pilots, the violence of the storm as observed on the ground, and of the frequency with which turbulent conditions are associated with thunderstorms.

 

Probable Flight Path Prior to Impact

The pilot reported himself over the Herndon fan marker at 2:31 P.M. at an altitude of 4,000 feet and continuing to climb. The intended cruising altitude, as shown by the flight plan, was 6,000 feet. The time from the take-off at Washington to passage over Herndon was therefore ten minutes. The distance from the airport to the Herndon fan marker is 19 miles. The average speed over the ground, assuming the correctness of the pilot's report, was therefore 114 m.p.h.; the average rate of climb, 400 ft. per minute; figures that are reasonably consistent with one another, assuming a light headwind as forecast, and assuming the engines to have been operated during the climb at the power output customary during that part of a flight. If the same airspeed and rate of climb had been continued past Herndon (making a small allowance for the time lost in the take-off before getting into steady climbing flight, and assuming the increase of airspeed with altitude to be offset by an equal increase in the strength of the headwind) an altitude of 6,000 ft. would have been reached just before 2:36 P.M., and at a distance from Washington of approximately 28 miles. Had the airplane then leveled off at an altitude of 6,000 ft. in accordance with the flight plan, and proceeded at its cruising speed of 180 m.p.h. against a 15-mile wind, it would have arrived over the scene of the accident, 44 miles from the Washington airport, at exactly 2:41 P.M. The reported time and altitude at Herndon, and continued adherence to flight plan thereafter, are therefore entirely consistent with the apparent time of the accident. The time of the accident is, of course, not known with such accuracy as to permit depending on calculations of this sort to fix the flight path or airspeed with a great degree of refinement, but at least they appear to exclude the possibility that the aircraft had wandered very far from its course, or that it had proceeded to the far side of Short Hill and then turned back, as one witness believed.

As there has been absolutely nothing, aside from the altitude estimates given by some lay witnesses who in turn were contradicted by others presumed to have viewed the same airplane at approximately the same time, to suggest that the pilot was not proceeding in accordance with his flight plan, and as any change of plan is approaching a region where instrument operation was anticipated would normally be the subject of immediate report by radio, we conclude that the flight was proceeding normally, that an altitude of 6,000 feet had been reached, and that no trouble had been experienced up to the time of reaching the Lovettsville area.

Seemingly the pilot elected to fly through the storm area rather than to make any attempt to circle it. The storm extended for a very considerable distance north of the airplane's path and for several miles to the south, and presumably it appeared to the pilot as he examined the horizon that there would be no gain in passenger comfort or otherwise by any detour of practicable extent, and no hazard in flying straight through.

With the airplane presumed to be on course, at on altitude of 6,000 feet, approximately two miles southeast of the scene of the accident at approximately 2:40 P.M., the analysis of the flight path concerns the course that the airplane may have taken from that point to the final contact with the ground less than one minute thereafter. The most direct evidence on that matter is that of Mr. McGaha and his son, who reported having seen the airplane go into a dive immediately after the lightning flash, and shortly after the machine had passed over their home, Mr. McGaha, in fact, was so sure of his own observation that he had gone out of the house and walked to a considerable distance to look for the wrecked airplane, which he thought might have struck the ground on his own land. Proceeding from that observation, it must be concluded that the deviation from the normal path started at least 1¾ miles short of the point of impact, as it was approximately at that distance that the heavy rainfall began, and rain as heavy as clearly existed at the time and place of the accident would have made it impossible to see anything that airplane might have done after it had entered the storm. Mr. McGaha pointed out to an investigator for the Board the point at which he had been standing when he saw the airplane and the angle above the ground at which he recalled it as having appeared to him when its dive began.^[2] Analysis of that point, however, makes it appear impossible to depend on such a recollection of angles as a basis for analysis of the path. It is not surprising that it should be so, for it is of course externally difficult to recall a line of sight exactly, especially after a considerable lapse of time. Mr. McGaha's recollection was that he had seen the airplane start its dive, immediately after the lightning flash, at a point found to be at an angle of 11 degrees above the horizontal. Since Mr. McGaha's home was at a distance of 2½ miles from the easternmost edge of the rainstorm, the reported angle of his line of sight would have placed the aircraft approximately 2500 feet above the ground at the time of starting the dive. Since Mr. McGaha identified the airplane, by comparison with another machine of similar size subsequently seen at a known altitude,^[3] as having passed over his house and continued at a height of about 6000 feet above sea level, his estimate of the angle at which he had seen the airplane go into a dive (a much more difficult point to fix in memory) must be considered as having been in error.

A similar difficulty arises in connection with Mr. McGaha's recollection that he had watched the airplane until his view of it was cut off by a knoll near his home. The elevation of that knoll fixed another angle, and for the airplane to have been watched until it had gone down behind the knoll would have brought it to a height above the ground (at a distance of 1¾ miles short of the scene of the accident) of only about 1000 feet. Still assuming that Mr. McGaha could only have seen the machine when it was short of the rainstorm (as to the continuous intensity and impenetrability of which, there was general testimony from those who were in it), the airplane would then have had to travel horizontally for a distance of 1¾ miles to reach the scene of final crash while losing only 1000 feet of altitude. That would have indicated a mean angle of path to the horizontal of only about 6 degrees, which does not represent a dive but a comparatively gentle descent. The full acceptance of Mr. McGaha's recollection on this point, combined with the facts that appear to have been definitely established with regard to the extent of the rainstorm, would therefore require the airplane to have started its dive at a point between two and three miles short of the final impact; to have continued the dive at least to within 1000 feet of the ground; to have come back nearly to level flight at that point; to have continued for one or two miles along the path descending at an average angle of not over 5 degrees; and then to have nosed over again, to an angle of at least 20 degrees to the horizontal and probably more, and continued on the path thus established to the point of impact.

Less extreme conclusions concerning the flight path are developed if it be considered as a possibility that the airplane may actually have disappeared from Mr. McGaha's view by diving into the rainstorm beyond the hill rather than by going down behind the crest of the nearby hill, and that he and his son were deceived on that point by the haze and the bad visibility associated with the nearby storm and the general dullness of the day. In that case the airplane might actually have disappeared into the storm at a height considerably greater than 1,000 feet. If it had entered the storm at 4,500 feet above the ground, its path from that point to the point of impact would have been inclined to approximately 26 degrees to the horizontal, or at an angle about equal to that at which the airplane most probably struck the ground. Although the assumption of a steady dive at an angle of 30 degrees or thereabouts is in superficial conflict with Mr. McGaha's impression of the airplane as going down very steeply, the conflict is reduced by taking account of the angle to the horizontal of Mr. McGaha's line of vision on looking at the airplane. With the airplane in a 26-degree dive at a height of 5,000 feet (taking that as the altitude at which the dive might have become well established at a fixed angle), and at a distance of two miles short of the point of final impact, Mr. McGaha would have seen it at an apparent angle of 51 degrees to his line of sight. The apparent length of the airplane, as seen at that angle, would be foreshortened by less than a quarter of the true length, and even an observer quite familiar with the form of aircraft and their appearance during maneuvers might have difficulty in distinguishing the machine seen at such an angle from one diving very near to the vertical.

If the airplane had gone into a vertical dive at a height of 5,500 feet above the ground, while traveling at normal cruising speed, its speed would have increased to approximately 380 m.p.h. at the time of striking the ground. A similar study for a steady 30-degree dive from the same altitude indicates a speed at contact with the ground in that case of about 330 m.p.h. These figures are very approximate, but suggest the general order of magnitude of the probable speed. Although it would be impossible to determine the speed at contact with any accuracy from the condition of wreckage, the completeness of the destruction at least indicated a speed far above that of ordinary flight. The damage, which technicians who have had past occasion to examine a great many wrecked airplanes that have be contact with the ground in all sorts of attitudes and at all sorts of speeds found almost unprecedented in their experience, would be difficult to with a speed of less than 300 m.p.h.

If the airplane had gone into a vertical dive at 5,500 feet, and had then continued the vertical path very nearly to the ground, with the machine in process of recovering from the dive when it hit, the time from the first deviation from the normal flight path to the impact would have been about 15 seconds. If the descent had been along a steady 30-degree inclination, the corresponding time would have been approximately 30 seconds.

The best estimates of the time interval that elapsed between the time of the lighting flash and the final crash of the airplane, obtained by reconstructing the movements of persons who saw the lightning and heard the crash and engaged in some definite activity in the interval, put it at about 15 to 20 seconds. Though that figure cannot be regarded as a very reliable one, if it be accepted as valid, it would require either that airplane's first deviation from its path preceded the lightning flash with which the crash was so generally connected, or that the dive was very nearly a vertical one through the greater part of its length, and must in that event have started after the airplane was well into the area of the heavy rainstorm.

One difficulty with the assumption of a straight descent at a constant angle is that it fails to provide the negative acceleration of the airplane which would be a possible explanation of the apparent over-speeding of the propellers.^[4] A negative acceleration suggests inverted flight, and consideration has been given to the possibility that the flight path might have had the form of an S, the airplane having been on its back at a midpoint of the descent, and a recovery from that attitude having then been started but not completed before striking the ground. Such a path would be possible in the event of a temporary disabling of the pilots or a temporary interference with the control, a difficulty lasting only a few seconds and followed by resumption of control of the airplane. It would account for the negative acceleration, and consequent over-revving of the propellers. It would account also for the fact that several of the witnesses living near the scene of the accident spoke of the roaring noise that immediately preceded the crash as having seemed to come from the west, over towards the mountain,—for if the airplane had actually taken the shaped path downward, its first deviation from its normal attitude would have occurred when it was approximately over the point of final contact with the ground, and therefore well into the rainstorm area.

Another alternative is that the airplane might have spun from a considerable altitude, or descended on an irregular path after the wing had stalled. The reasons for discarding those hypotheses are explained elsewhere.^[5] In addition to the reasons given there it would appear that there could not have been any large amount of side-slipping or turning on the way down, unless the turns had been extended into complete circles, since the airplane struck the ground substantially on a direct prolongation of the line of flight that it appears to have been following immediately before entering the storm.

In considering the implications of this and other possible flight paths, it must be remembered that the longitudinal stability of the airplane would give it a pronounced tendency to recover from a dive, even without the intervention of the pilots. Tests and calculations on a similar airplane have shown that with the center of gravity in the position that it had at the start of this flight, and with the tab control set for the airplane to trim at cruising speed, it would require a steady push of at least 40 pounds on the control column to keep the nose from rising when a speed of 300 m.p.h had been reached. In this connection there is a possibility which is extremely remote, but may nevertheless be mentioned, in view of the difficulty of finding any combination of circumstances that seems at all probable as an explanation of the maintenance over a period of 15 seconds or more of a flight path of which the abnormality would be expected to have advertised itself to the pilots. If an airplane nosed over very abruptly and very steeply at a time when the occupants did not have their belts fastened, a number of them might have fallen or slid from their seats to the forward part of the cabin. The resultant shift of the center of gravity would make the machine trim at a considerably higher speed than that for which the tab controls were originally set, and so hold it in a dive (though probably a comparatively shallow one) if the pilots, having themselves been disabled by whatever cause produced the sudden change of course, were exerting no force on the controls.

Aside from the complete disabling of the pilots or a shift of load in the airplane, two hypothetical explanations of the prolonged maintenance of a steady dive have been considered. If the pilots were blinded by a lightning flash their immediate concern would have been to avoid stalling of the airplane while their disability continued. In seeking to be perfectly safe on that point they might have over-corrected by pushing the machine into a dive, and with both pilots pushing on the control column together the force required might have been overlooked under conditions of such stress and the dive continued until the ground was reached. The other possibility considered is that the clogging of the airspeed indicator head with rain made the indicator read too low by a gradually increasing amount. The indicator would then have indicated a gradual approach to a stall which the pilot might have tried to off-set by thrusting the control column gradually forward to pick up more and more speed. This explanation does not seem at all likely to be a correct one since other instruments immediately in front of the pilot, notably the altimeter and the artificial horizon, would be supplying an obvious contradiction of the airspeed indicator's reading. Every experienced pilot would recognize the possible fallibility of airspeed indications in heavy rain or freezing weather and would check against his other instruments under such conditions.

 

Mechanical Failure

Investigation of the possibility that a mechanical failure might have been the cause of the accident was made extremely difficult by the condition of the airplane when it was found after the accident. It was in thousands of pieces.

The condition of the engines indicates that they were operating at very high speed at the time of impact but the amount of power being developed is unknown. The switches and the fuel valves were found in the "on" position (the throttles were in the full open position but it cannot be determined in what position they were at the time of the impact). The bends in the propeller blades were identified by experts as power bends and witnesses near the scene of the accident testified that they heard a loud roar of engines immediately prior to the sound of the crash.

An inspection of the engines revealed no indication of a mechanical failure prior to impact. The condition of the master rod bearings and crank pins of both engines showed that the oil film had broken down and the parts showed signs of overheating as did some of the knuckle pins. However, in view of the completely separate lubrication systems and the identical character of the damage to both engines, it appears that the damage was caused not by mechanical failure of the engines but by some factor inherent in the attitude or motion of the airplane after the initial difficulty and previous to the crash.

The evidence indicates that the condition of the master rod bearings, the crank pins, and knuckle pins could result only from a momentary interruption of the oil supply to the bearings or from an overloading of the bearings either of which might have been caused by over-revving of the engines.^[6]

The nature of the damage incurred by the master rod bearings, crankpins, and other parts of the engines indicates that over-revving in excess of 3000 rpm took place on each engine. This is borne out by the testimony of an expert witness during the hearing that the slight indication of heating on the master rod bearings, knuckle pin bushings, piston pin bushings and knuckle pins was caused by overspeeding of the engine rather than by lack of lubrication. This testimony was later corroborated by an independent examination of the parts of one engine by an expert agency.

The expert who testified that over-revving in excess of 3000 rpm took place based his opinion on experience with similar engines which had been over-revved on a test stand. He found that marks made on various rotating parts of the engine, especially the crankpin, varied both in magnitude and direction in proportion to the amount of overspeeding. Data obtained from overspeed tests on master rod bearings of this type of engine indicates that the oil film will remain intact for bearing loads imposed by an engine speed of at least 3000 rpm.

From test experience with other bearings which were damaged in a manner very similar to the damage incurred by these bearings it is estimated that the duration of the over-revving was from five to eight seconds and furthermore it appears that rotation of the engines was stopped very shortly after the over-revving occurred because certain marks on the crankpins which were caused by overspeeding would have been erased if the engines had again been operated at normal speed.

Witnesses near the scene of the accident also presented testimony which indicated over-revving. Several witnesses heard a loud roar of engines for varying periods of time. Mr. Garland Jacobs, who was 300 yards west of the scene of the accident, thought that the roar lasted for about 30 seconds. By retracing Mrs. Thompson's movements the estimated duration of the noise she heard was 19 seconds at 400 yards west of the scene of the accident. Other witnesses made varying statements as to their estimate between the time of the lightning strike and the roar of the engines. Mrs. Jacobs estimated 10 seconds. The noise during descent sounded to her like a "siren" or "scream". Mr. Harry E. Everhart, who lives 1¾ miles southeast of the scene of the accident, estimated 15 seconds. Mrs. Everhart, who lives 2¾ miles east of the scene and who heard an "awful roar," reported 7 seconds. Mrs. Hickman, who lives about 2½ miles southeast of the scene of the accident, estimated 16 seconds. Mr. McGaha, 4¼ miles southeast, also heard a "roar." Mr. Ridgeway, 3¾ miles southeast, and Mrs. Hickman heard a noise that sounded like an old truck.

Information received from the N.A.C.A. indicates that a large increase in sound emission from a propeller would result only from overspeeding and that the engines must have been temporarily overspeeding to create the high pitch noise which could be referred to as a "shriek" or "siren" effect. This would explain the "siren" or "scream" which Mrs. Jacobs heard and would distinguish the high tip speed of the propeller from the "roar" of the engines. The testimony of the witnesses therefore corroborates the evidence on over-revving since the noise issuing from an airplane would not ordinarily be described as a "scream."

Over-revving might occur if the airplane were to accelerate faster than the propeller pitch could keep pace with it if in addition there had occurred a momentary interruption of the oil by which the pitch changing mechanism operates. This would create a tendency to rotate the propeller blades toward the low pitch position or at least retard them from moving to the high pitch position which they would normally tend to assume while the airplane was accelerating as in a dive.

Not only would the interruption of oil thus tend to invite overspeeding but the interruption itself could be caused by the same maneuver which created the acceleration. If the airplane went into a sudden dive a momentary interruption of oil could be caused by the negative acceleration which would create a movement of oil away from the oil outlets in the tank, exposing the outlets to air. That such a condition is possible is borne out by the testimony of witnesses during the hearing.

This negative acceleration could be caused by a sudden dive or by having the airplane go over on its back or by a severe down gust. It is not likely that interruption to the oil supply during stable flight conditions could cause over-revving since if the oil interruption had occurred while these conditions obtained, the rate of rotation of the blades to low pitch would be at the slow rate of approximately one degree in eleven seconds.

Under normal conditions the propeller blades would have rotated toward the high pitch stop (about 45 degrees) while the airplane was accelerating and descending at high speed. From the fact that the pitch angle at time of impact appeared to be 24 degrees, which is a cruising angle, it may be inferred that the interruption in oil supply which might have taken place and which would tend to rotate the blades to low pitch, balanced the tendency of the blades to go into high pitch as a result of the high speed of the airplane. Under such conditions the propeller would have the characteristic of a fixed pitch propeller. If the propellers had been windmilling at a pitch setting of 24 degrees and at 3000 RPM, an airplane speed of 320 miles per hour would be indicated. If the propeller had been windmilling at 24 degrees and maximum engine RPM of 2350, it is estimated that the airplane speed would have been about 250 miles per hour.

It appears that the condition of the bearings and crank pins indicates over-reviving immediately prior to impact and not mechanical failure within the engines themselves or their oil systems. An investigation was, nevertheless, made of the mechanical adjustment to the oil filter for the right engine which was required just prior to take-off of the airplane from Washington. As we have previously stated, the oil pressure as shown on the gauge for the right engine had fallen below normal and it was necessary to rotate the blades in the cuno oil filter to relieve it of any sediment which might be blocking the oil flow. It is not believed that the condition which required this mechanical adjustment had any bearing upon the accident or upon the condition of the engines prior to or after the crash. The accumulation of sediment on the oil filter discs to such an extent as to reduce oil pressure is not uncommon and the blades installed between each pair of discs are designed to remove this sediment. After the blades had been turned, the copilot stated that the oil pressure had returned to normal. Moreover, this adjustment was required only with respect to the right engine. Since each engine has an independent oil system, the low pressure on the right engine could not have had any effect upon the left engine. It is not unusual for sludge or sediment to appear in airplane engines making it necessary that filters be installed in the lubricating system to prevent sludge from being directed through the smaller passages throughout the engine which might cause lubrication failure with resultant serious damage or failure to the engine. The oil companies have made an extensive research in cooperation with the manufacturers of engines used in air carrier service, as well as the air carrier operations departments, resulting in only a high grade of oil being approved by engine manufacturers for use in their respective engines. The type and grade of oil used by Pennsylvania-Central Airlines is the same as that used in other of the larger air carrier services. The sludge or sediment content which might be collected in an airplane engine would depend largely upon the usage of the engine, particularly the amount of power applied over extended periods. In other words, if the engine is operated at a high rate of horse power over an extended period of time, more sludge or sediment can be expected than if the engine is operated under lesser horse power conditions. If the cuno type filter which was installed in aircraft NC 21789 was free of sediment to an extent which would permit a normal oil pressure of 60 lbs. at the time of departure, it is not likely that sediment could collect in an amount that would be alarming prior to reaching the next scheduled stop. However, should the filter clog up with sludge or sediment so as to reduce the oil pressure below 40 lbs., the oil would be by-passed around the filter so that adequate lubrication would be provided to the engines and no stoppage of oil would result to the lubricating system. Thus, while the filter is not absolutely necessary to the operation of the engine, it contributes to its economical operation.

It is quite likely that if a material loss in oil pressure was indicated shortly after taking off, Captain Scroggins would have returned to the Washington-Hoover Airport. When this fact is considered with the evidence to the effect that the condition of both right and left engines after the accident was almost identical, the conclusion seems clear that the condition of the oil filter prior to departure from Washington could not have had any bearing upon the cause of the accident.

No reason has been found to suspect an engine failure while in flight. Not only does the direct physical evidence give no suggestion of such a failure, but the circumstances of the accident were such that engine failure partial or complete, could hardly have initiated it. Had Captain Scroggins encountered any malfunctioning of the engines or engine failure, he probably would have followed the usual procedure and returned to Washington. Had one engine failed, the flight could have been maintained over an extended period of time operating on the remaining engine, particularly since the engines were equipped with full-feathering propellers. Had both engines failed, there are numerous farms between Washington and the scene of the accident on which an emergency landing could have been effected. Weather conditions in the Washington area were favorable through the afternoon of August 31 and there is no evidence of a weather condition which would have prevented an emergency landing along the route up to within three miles of the scene of the accident at most. Had any mechanical difficulties arisen prior to the time Trip 19 got into serious difficulties, Captain Scroggins would undoubtedly have informed Pennsylvania-Central Airlines in Washington to that effect by radio. If he had experience mechanical trouble on entering the storm, he would have turned back out of it.

The remains of the control system were examined to determine whether a mechanical jamming of the controls might have caused the accident. While all of the control surfaces were located, it was impossible to discover, due to the condition of the wreckage, whether or not a mechanical jamming had occurred. However, no reason appears for concluding it had occurred.

 

Structural Failure

Consideration was also given to a possible structural failure during flight. In this respect the investigation was made difficult by the condition of the wreckage. However, all major component parts of the airplane were found forward of the point of impact. An examination of all pieces of the wreckage recovered revealed no evidence of any failure, displacement, or distortion of the wings or tail surfaces during flight. All breaks and damage to the structure appeared to have been caused by impact or by the movement of the various parts over the ground following the initial impact. Therefore, it does not seem probable that structural failure during flight contributed, to the accident.

 

Personnel Disability

By reason of the fact that the airplane plunged to the ground with power on, it is possible that for some reason the pilot and copilot were prevented from effectively operating the controls. The evidence shows that both Captain Scroggins and First Officer Moore were in good physical condition and in normal spirits at the time of departure from the Washington-Hoover Airport. There is no record or other indication that either of these pilots was subject to physical disabilities which would incapacitate him without waning. Moreover, it would be necessary to find that both pilots became incapacitated at just about the same time, for if only one pilot should be disabled and fall forward on the control column, the other pilot would not be called upon to exert extraordinary force in order to maintain control of the airplane.

It cannot be definitely said that both the Captain and the First Officer were in their respective seats in the pilot's compartment at the time of the accident. It is the usual practice for air carrier pilots to visit the passenger cabin sometime during the flight, but under ordinary circumstances such visits would not be made until after the airplane had reached its cruising altitude and then only under conditions of favorable weather. Since the accident occurred within about twenty minutes following take-off from Washington-Hoover Airport and the storm condition was being approached just prior to the accident, we find no reason to believe that both the Captain and the First Officer were not in their respective seats at the time of the accident.

It is, of course, possible to conceive of a variety of occurrences by which both pilots would have been disabled at the same time. However, the only possibilities that seem to justify discussion in the light of the present record are the effects of a lightning strike, fire, sabotage, or turbulent air. These four subject matters and their relationship to the flight crew will be discussed hereinafter.

 

Fire

Consideration was also given to the possibility that fire might have occurred prior to impact and contributed in some way to the accident because there were evidences of fire on many parts of the airplane, pieces of burnt paper were found behind the scene of the accident, and Mrs. Jacobs testified that she saw a blue looking flame in the air. From the evidence, it is clear that immediately following the impact the fuel tanks broke open and gasoline was sprayed over a wide area. Mrs. Jacobs and Mrs. Thompson testified that at the moment of impact there was a violent flash of fire. Mrs. Thompson stated that immediately after the impact, she saw fire rolling across the alfalfa and corn fields forward of the point at which the airplane struck. No witness, other than Mrs. Jacobs and Garland Jacobs, testified that he saw fire in the air. Carrol McGaha and his son, who testified that they saw the airplane go down, noticed neither fire nor smoke.

It is conceivable that an explosion took place upon impact due to the large amount of gasoline which was being carried and the fact that fire was seen. Such an explosion might have contributed to the disintegration of the fuselage and resulted in a blast of sufficient intensity to carry pieces of the airplane high into the air. The wide dispersion of the gasoline following impact and the intensity of the rain which was falling at that time would limit the duration of the fire. The blast would probably create quite a large volume of hot air which, because of the unstable conditions which exist in the neighborhood of storms, would tend to rise at a very rapid rate carrying particles, such as pieces of paper, with it. Fire and explosion at the time of impact could have been caused by gasoline coming in contact with hot parts of the engine, by breaking of electrical connections, and by metal striking rocks on the ground, creating sparks.

Examination of the wreckage disclosed that partially burned and scorched parts of the airplane were spread over a large area. These parts include the upholstery from the passenger compartment, one tire, various metal parts of the airplane's structure, and the fabric attached to the rudder and left elevator. It is necessary to determine whether these indications of fire resulted from the fire which occurred at the time of impact or a fire which had been burning prior to that time.

The evidence indicates that the marks of fire resulted from the fire at the time of impact. In many instances evidence of fire was found on one part of the airplane, while a part to which it had been attached showed no sign of fire. For example, the right tire was partially burned, but the landing gear parts to which it was attached and the wheel wells into which the landing gear wheels are retracted showed no evidence of fire. The right tire was more severely burned than any other part of the airplane. This tire was found in close proximity to the point of impact and probably collected a considerable amount of gasoline in the wheel depression as the gasoline was thrown from the fuel tanks. A short length of the upper forward part of the passenger compartment had broken in two. One of the pieces was found, with the sound-proofing partially burned and the edges where the break occurred were smoked. The side upon winch the sound-proofing was attached was down. The other piece to which this one had been attached was found nearby with the sound-proofing side up. This piece showed no evidence of fire in the sound proofing nor did the edges where the break occurred show any signs of smoke.

No parts of the engines, nacelles, or pilots' compartment showed any evidence of fire. Some parts of the fuel tanks which had apparently been broken at the time of impact showed evidence of fire. While this evidence consisted for the most part of marks of smoke, a large part of one tank had been melted, probably because it had carried a small quantity of gasoline with it following the impact, which burned on the ground. Ten pouches of mail which had been carried in the mail compartment which is formed by heavy wire mesh partitions in the companionway between the pilots compartment and the passenger cabin were examined and no indications of fire on the pouches or the mail were found. All fire extinguishers on board the airplane were accounted for and none of them had been used.

There is nothing which could be considered as inflammable material either in the passenger or pilot compartment. The upholstery, carpets, seat cushions, etc., which are treated with a fire resistant chemical could be burned only by application of intense heat or after having been soaked with some inflammable fluid. There were three water flares in a closed compartment located about two feet behind the pilots' compartment. A one-gallon can of water to replenish the water supply in the boiler which provides steam heat to the airplane is normally carried in the same compartment about 18 inches below and 15 inches behind the water flares. Part of the flotation gear^[7] is carried between the water flares and the can of water. These water flares are composed of metal cans about six inches in length and two inches in diameter, containing a mixture of carbide which forms an inflammable gas when wet, and a substance which will produce a flame to ignite the gas upon contact with water. These cans are sealed and equipped with a pull ring in the top which, when pulled, opens a small hole. If the flare is then placed in water it will immediately ignite. These flares were, of course, inflammable, but under the conditions above described, it seems impossible that they could have ignited while the airplane was in flight in view of the care with which they are sealed and the unlikelihood of water entering the metal container during flight. The fact that the water carried for the boiler was located below the flares and was separated by a part of the flotation gear would preclude the danger of fire from that cause.

Under all of the circumstances, it seems highly unlikely that the airplane was on fire during the flight. It is possible that Mrs. Jacobs' and Garland Jacobs' impression of fire in the air actually resulted from the fire at the time of impact due to the rapidity with which events occurred. Furthermore, the space interval during which they could have seen the airplane was so short as to prevent an accurate observation of the condition of the airplane as it flashed by at a tremendous speed.

The discovery of a number of pieces of burned paper southeast of the scene of the accident appears at first glance to be inconsistent with this conclusion. One piece, a manila envelope, was found 1¼ miles southeast of the point of impact. One possible explanation of the presence of this burned envelope and other papers was that the airplane was on fire prior to impact and the papers blew or were thrown out. However, in view of the evidence which tends to show that there was no fire during flight, the most plausible explanation of the presence of the burned paper is that at the time of impact the terrific up-draft which followed the blast carried the papers high in the air despite the heavy rain and they floated down to the places at which they were found.

It would not be unusual to find up currents moving at substantial speeds in a storm such as the one in process at time of the accident. Such a current further augmented by a blast of hot air from burning gasoline could conceivably carry papers, such as were found, to a considerable height. Even a comparatively mild breeze from the northwest could then have carried them to the points where they were found. It is noteworthy that material such as the flight calculator and manila envelope were carried the furthest and it is believed that this is due to the fact that such materials would be slower to absorb moisture than lighter papers which were found closer to the scene of the accident.

Tests were conducted by the National Advisory Committee for Aeronautics on a wet manila envelope such as that retrieved from the accident. The result, when considered in conjunction with the evidence of the gasoline explosion at the time of impact and the consequent vertical atmospheric currents, leads to the conclusion that the flight of the envelope from the scene of the crash to a point 1¼ miles distant is not only possible but highly probable.

This theory is supported not only by the evidence, which indicates that there was no fire in the air, but also by the fact that the scorch marks on the manila envelope and other papers found at various points southeast of the accident are similar in character to those on papers found at the scene of the accident.

Therefore, we conclude from the preponderance of the evidence disclosed by the investigation, that the only fire that occurred took place following the impact.

 

Turbulence

The Board has given consideration to the possibility that the accident may have resulted from turbulence so violent as to throw the airplane out of control.

Flying through turbulent air is not, of course, an unfamiliar experience for any transport pilot. It customarily involves nothing more than abrupt accelerations of the airplane, with resulting discomfort for the passengers, and an increased need for alertness on the pilot's part to keep the airplane on its course and to restore it to a normal attitude after any particularly violent disturbance. Instances are known, however, of airplanes having encountered turbulence of such extraordinary intensity as to momentarily throw them completely out of the pilot's control and into wholly abnormal attitudes. Cases have been reported of airplanes having encountered gusts so violent as to stall^[8] the wing completely, from a flight speed considerably above the normal stalling speed. Stalling due to turbulence, without there having been any ice on the wings to increase the susceptibility to such stalling, is extremely rare, but cannot be absolutely eliminated as a possibility.

When an airplane of the type involved in the Lovettsville accident is deliberately stalled with power on, the stall customarily develops first on one wing-tip and that wing drops sharply, the airplane simultaneously turning toward the low wing. Where the maneuver is deliberately executed by a pilot who is prepared to initiate immediately the required measures of recovery, evidence received in the present investigation indicates that only from 400 to 600 feet of altitude is normally lost in restoring normal flight. The intentional stall, being initiated with the airplane's nose pointed up at a steep angle, does not strictly simulate the conditions of an inadvertent stall due to gusts encountered while in level or nearly level flight. It is impossible to produce that condition for test purposes. Any uncertainty arising on that score is added to uncertainty concerning the effect of intense turbulence during a recovery from a stall, and to uncertainty concerning the effect of bad visibility in making it difficult to estimate the amount of altitude that might have been lost during recovery if the airplane had stalled under the conditions existing at the time and place of the Lovettsville crash. It would seem most unlikely, in view of testimony that the training of Captain Scroggins and all other Pennsylvania Central Airline pilots flying the DC-3 had included many stalls and recoveries therefrom, including some under simulated instrument conditions with no dependence on outside visibility, that the airplane would have lost anything even approaching 5000 feet of altitude in the course of recovery from a stall uncomplicated by other factors than those of mere atmospheric turbulence and inability to see the ground.

It is conceivable that a stall might start a spin, although inadvertent spins with aircraft of this class have probably been even rarer than stalls due to turbulence while flying at normal speed. In view of the limited experience, it is impossible to speak with any certainty of how the aircraft would behave in a spin, although it is again true that recovery from a fully developed spin would normally be expected in much less altitude than is believed to have been available in this case between the height at which the airplane was cruising and the point at which it stuck the ground. It is possible, however, in view of the extreme unfamiliarity of the experience of spinning in a transport airplane and of the special difficulties presented by instrument conditions and rough air, that even a highly expert pilot might lose as much as 5,000 feet before being able to complete the maneuvers for checking the spin and recovering from the subsequent dive; but the need for speculation on the possibility is reduced by the mute denial of any likelihood of the airplane having spun that was provided by the position of the wreckage and by the reading (as previously referred to) of the directional gyro. At the time of impact the airplane was on its normal heading of 310 degrees magnetic, and it would have been most extraordinary if the pilot should have struck the ground during an unintentional spin, or in process of recovery from it, on the same compass heading at which the spin had begun. Tests of the directional gyro by its maker also indicate that if the airplane were spinning with its nose down as much as 40 degrees (as it probably would be in a fully-developed spin) the gyro element would have tumbled in the case and the instrument would have ceased to give any semblance of a true indication of heading. The uniform distribution of the wreckage ahead of the point of impact indicated that no rotation around the longitudinal or vertical axis of the airplane was occurring at the time of impact, which would in itself eliminate the possibility that the airplane was actually in a spin when it struck, although it would still leave the chance that the airplane might have been in a spin, which the pilot had checked, and that it had struck the ground before it had been possible for him to complete the flattening out of the flight path from the subsequent dive before striking the ground.

The direction in which the airplane was headed at the time of impact even argues against the possibility of a stall having taken place after the airplane had for some reason descended to low altitude, since the dropping of a wing in the stall is attended by a considerable turning of the airplane and recovery from the stall is not normally on the same heading on which the maneuver was started. The same reasoning applies, with even more force, to any theory that the airplane might have stalled in conditions of such violent and continuous air turbulence that the pilots were never able to regain control during the 20 seconds or more that would have ensued before the machine would have struck the ground. The presumption of turbulence of that order, and continued over so long a period, seems in any event a virtually incredible one in the light of all that is known of atmospheric structure. It would, in any event, have been impossible for violent turbulence to extend to a very low altitude without having shown itself as a surface wind.

A possibility that may be considered in connection with the discussion of turbulence, though it is not strictly in that category, is the action of a simple down-draft or descending current of air. Thunderstorms are characterized by violent vertical motions of the air, and especially by rising currents, often of very high velocity. Such rising currents have been known to attain a velocity of several thousand feet a minute. In undisturbed air, at 6000 feet, the maximum rate of climb of a fully loaded aircraft of the type involved in this accident is approximately 600 feet per minute at take-off power. Even with one engine dead, the maximum rate of climb would be over 100 feet per minute. If the airplane had flown out of a rising current in front of the storm and directly into rapidly descending air, approximately as it entered the rain, it is conceivable that it might have been impossible, even with the application of full power, for the pilot to avoid a rapid loss of altitude. Aside from the improbability in the light of existing meteorological knowledge of so strong a descending current, and aside from the fact that vertical currents lose their violence near the ground so that the pilot would have in any case been able to check a descent induced in that fashion before the ground had been reached, the time element seems to eliminate a steady vertical current as a possible primary cause of the accident. The earliest point at which the pilot could by any possibility have entered a strong descending current to the point of contact with the ground was only about two miles. It could not possibly have taken over one minute to cover that distance, and to have descended from cruising altitude to ground level within that time would have required a mean rate of descent of over 5000 feet per minute, and a descending current of even more velocity than that. It scene impossible to credit the existence of any such condition.

The Board has also given consideration to the possibility that in the event of severe turbulence, the occupant of the jump seat in the pilots' compartment might have inadvertently disturbed the pilots' control of the airplane. The jump seat installation in NC 21789 consisted of a 12 x 18 inch flat board about ½ inch thick. It is supported by two steel pins about ½ inch in diameter on one side, one of which is spring loaded to hold it in place. These pins are placed in holes in the secondary structure provided for that purpose. The other side is supported, when in a down position, by resting on a channel which supports a part of the mail pit. There were no back or side arms attached. The seat was designed so that it might be raised upward from one side to permit pilots to pass to or from the cockpit. The seat, when in place is located in the companionway directly behind the archway leading to the pilots' cockpit. It was approximately 15 inches from the forward edge of the jump seat to the back of the pilot's seat. The jump seat safety belt is of material similar to that used in the safety belts in the passenger cabin. It was fastened on each end to the airplane structure, on the righthand side by a standard end type eye bolt and on the lefthand side by a special fitting designed by the manufacturer of the aircraft.

The landing gear, flap and engine selector valve controls are located forward of the jump seat on the righthand side. The wobble pump control is located forward of and on the lefthand side of the jump seat and all are easily accessible as a support to one sitting in the jump seat if he finds it necessary to hold on to something. Since the occupant of the jump seat was not a pilot and had not ridden frequently in airplanes, consideration was given to the possibility that he might have grasped the flap control in order to steady himself if the airplane had been passing through turbulent air conditions and the possible effect such action could have had upon the flaps or the controllability of the airplane. The possibility of such action having contributed to the accident is eliminated because of the fact that the flap control mechanism is so designed that the flaps could not be actuated at a speed in excess of 112 miles per hour. If the flaps had been lowered at a speed of 112 miles per hour or less, the airplane would have assumed a settling attitude in a horizontal position and would have had no tendency to assume a diving attitude. There are also a number of other fixtures such as the channels of the archway and the back of the pilots' seats which an occupant of the jump seat could grasp.

A question was also raised as to whether the occupant of the jump seat had his safety belt fastened. If he had not, there is a possibility that he might have been thrown forward so as to interfere in some way with the handling of the airplane controls. The jump seat was found in a badly damaged condition. A part of each side of the safety belt was found and the webbing of both had been broken or cut. Close examination of these parts does not indicate whether either was broken under heavy strain such as might be expected at the time of impact had the belt been fastened and therefore it is impossible to determine whether the belt was fastened at the time of impact.

It is unlikely that the person sitting in the jump seat would have been thrown forward because of up or down drafts such as might be expected in turbulent air conditions. The usual tendency is for one to be thrown straight upwards when the airplane strikes a severe downward gust of air. However, if the seat belt had not been fastened, the occupant of the jump seat might be thrown forward involuntarily into the pilots' compartment if the airplane were suddenly nosed down steeply so as to cause a pitching movement. Therefore, if the airplane suddenly went into a steep dive, the observer might have been thrown forward into the pilots' compartment and might have thereby interfered with any attempts the pilots could make to regain control.

 

Effect of Heavy Rainfall on Performance

The witnesses testified that the rainfall, which was occurring at the time and place of the accident, was the heaviest experienced in a great many years. These statements were substantiated by the fact that small streams in that area were filled to overflowing, bridges were washed out, and highways were flooded. In view of this evidence the Board sought and received a report from the National Advisory Committee for Aeronautics as to the effect of heavy rainfall on airplane performance.

Calculations were based on an estimated rain density of 50 grams per cubic meter. This rain density is equivalent to a rainfall of 1.4 inches per minute if the falling velocity is taken as 12 meters per second. According to a recent report prepared by the United States Weather Bureau this is maximum rain density likely to be experienced anywhere in the eastern part of the United States and represents extreme conditions of actual rainfall in a cloud burst.

According to the National Advisory Committee for Aeronautics report, the increase in weight due to the accumulation of rain, as well as the impact of rain, assuming a mean rain drop velocity of 20 feet per second, is negligible.

Using a rainfall of the same intensity, calculations were also made to determine the effect of the drag which might result from rain impinging on the frontal area of the airplane. The results show that this effect, while not negligible, is not likely to force an airplane down. The drag effect varies with the size of the rain drops, but, assuming that the rain consisted entirely of large drops, the power absorbed by the rain while substantial, would be less than the power reserve available. Even if the reserve power were not drawn upon, the path angle and rate of descent would have to be maintained for several minutes and over a distance of several miles to force an airplane down from 5500 feet above the ground.

The roughening effect of rain due to fixed or splashing rain drops on the airfoil is unknown, but the percentage increase in drag from that cause is believed to be small in airplanes such as the DC-3 with overlapped skin construction, exposed rivet heads, and other departures from an absolutely smooth wing surface.

While the conclusions of the National Advisory Committee for Aeronautics indicate that heavy rainfall will not disturb the performance or behavior of an airplane such as a DC-3 to any marked degree, the Committee believes that heavy rainfall would have a substantial effect on the performance of the airspeed indicator. Beyond a certain critical combination of airspeed and rain density, the airspeed head will flood and the water will accumulate in the pressure line. If such be the case the airspeed would no longer serve as a guide to the true flight condition.

However, in the case of NC 21789, a hand pressure pump was incorporated in the airspeed indicator system so that any accumulation of water could be manually discharged. It was impossible to determine whether this pump had been used, but had the airspeed indicator been affected due to water in the pressure line, it should not have caused serious complications for one of Captain Scroggins' experience because he had undoubtedly encountered a similar condition resulting from the pitot head freezing under icing conditions. Moreover, any inaccuracy of the airspeed indicator would be revealed by the attitude of the airplane, as shown by the artifical horizon, the altimeter, the power output of the engines, and the feel of the controls.

Based upon all the available evidence, it appears that the effect of the rainfall could not alone have caused the accident.

 

Sabotage

Examination of the wreckage did not reveal any evidence of the aircraft or its controls having been tampered with prior to the accident. An alarm clock was found in the wreckage in a damaged condition. This clock was turned over to the Federal Bureau of Investigation for inspection to determine whether it could have been associated with a detonating mechanism. The inspection of this clock revealed that the alarm had been set for 9:15. There was no connection on the clock which could have been used for wiring connections necessary to have used it as a detonating mechanism.

Since the door between the pilots' compartment and the cabin was not locked, there is a possibility that a passenger might have entered the pilots' compartment and interfered in some way with the control of the airplane. However, there is no evidence whatsoever in support of such a presumption and the jump seat upon which the observer was seated in the aisle running between the pilots' compartment and the passenger cabin would have made it especially difficult for a passenger to enter the pilots' compartment.

Robert Williams, an airplane cleaner for Pennsylvania-Central Airlines, employed at Washington-Hoover Airport testified that he saw a man enter the passenger cabin of the airplane two or three minutes before the other passengers and that he did not see this man leave the airplane before the passengers boarded. He could not state positively that the Stewardess or the rest of the crew were on board at the time this man entered. Since it was about time for the passengers to board the airplane he assumed that the man whom he had observed go on board was a passenger and he paid no further attention to his movements. No evidence was found which would indicate that this person entered the airplane for the purpose of committing sabotage.

Two employees of the Federal Bureau of Investigation, one a special agent, were on board at the time the airplane crashed. An investigator from the Federal Bureau of Investigation testified that the special agent was making a routine trip and that the special agents of the Federal Bureau of Investigation who came to the scene of the accident did so only for the purpose of identifying the bodies of their colleagues.

No evidence was found during the course of inspection or investigation to justify a conclusion that sabotage caused or contributed to the accident.

 

Lightning

The testimony of witnesses who were in the Lovettsville area at the time of the accident clearly indicates that a flash of lightning occurred in the vicinity of the aircraft while it still was proceeding in a normal flight attitude and at a time less than thirty seconds before the plane's impact with the ground.

A previously mentioned, Mr. McGaha and his son both saw lightning in line with the plane's flight and they testified that immediately thereafter they saw the plane dive toward the ground. Mrs. Everhart likewise saw a blinding flash of lightning on the path of the flight which caused her to lose sight of the plane. The young boy, George Pendley, in the house of Miss Virgie Mentzer testified he felt shocks at the moment of the lightning flash and roar of thunder preceding the roaring noise of the plane's motors.

The testimony of a great many witnesses agreed on a sequence of a flash of lightning, or at least the thunder which accompanied it, immediately followed by the roaring sound of the plane's motors apparently ending upon impact with the ground.

Several experts who testified and many data made available to the Board indicate that conditions in the Lovettsville area at the time of the accident were of a nature which could produce very strong discharges of lightning between the cloud which was over Short Hill and the ground. Descriptions of the unusual darkness and other characteristics of the cloud indicate that it could have generated large charges of electricity. Furthermore, it had been raining for some time in the vicinity and the dampness of the earth would have increased its conductivity so that electrical charges from a considerable area of the earth might collect at one point for a strong lightning discharge. Although there is no record of a great number of lightning flashes over any considerable period of time, there was at least one flash and possibly two flashes a very short time before the crash of the airplane and one of the results was the well-established splintering of the wooden butt of a rifle which was in Mr. Baker's barn some distance from the scene of the accident.

A flash of lightning close to an airplane in flight, and less than thirty seconds later, the crash of the same airplane into the ground, are two events of so much importance taken together that they cannot be dismissed as a mere coincidence and their relationship must be analyzed to the fullest possible extent.

The experts who testified and the data collected reveal the extent of the knowledge thus far accumulated concerning the causes, character, and the various results of lightning discharges, together with their effects on aircraft in flight and their crews. As relating to the accident under investigation, our analysis appropriately may be divided into four general effects of lightning;–thermal, electrical, optical, and mechanical.

The thermal or heat effect of lightning is the one typically found on the many aircraft which have been struck in flight. Generally, lightning enters and leaves airplanes at two different and often widely separated structural extremities, such as the nose, the tips of the wings, the units of the tail assembly, propeller blades, radio antenna masts, pitot tubes, etc. The point where the lightning enters or leaves the plane usually can be discovered by a hole or indentation "as small as a pin prick" or "as large as two silver dollars." Occasionally, the external fabric of an airplane may be burnt at the point of entrance or exit of the lightning. Experience has revealed that the speed of the airplane through the air quickly extinguishes these fires leaving indications of burns in straight lines rather than in curves.

The thermal effect of lightning is recognized in the electrical and radio systems of airplanes through burnt fuses, melted or destroyed wires, and other damage of a like nature, and there may be similar effects upon other parts of the airplane's structure, controls, or equipment.

Past experience has failed to reveal any case where lightning has caused the fuel of an airplane to catch fire while in flight, nor is there any record of any other serious form of fire in flight from this cause.

In this case, although all the parts of the plane were not found, those parts which usually are struck by lightning were examined, together with the remains of the electrical and radio systems. All the technicians and experts who examined the wreckage agreed that there were none of the recognized indications that lighting had struck the airplane before its crash near Lovettsville. There is no reason to believe that any usual form of lightning struck the plane or that there were any thermal effects of lightning upon the plane.

The electrical effect is not known to have caused any injury to persons on all-metal airplanes of structure similar to the one involved in the present accident, which have been struck by lightning while in flight, nor any serious damage to the airplane itself. Mr. L. P. Harrison, United States Weather Bureau,^[9] summarized reports on a great many cases of lightning strikes on aircraft including one instance of a pilot being incapacitated because of electrical shock which was conducted to him through the mechanism of his radio headgear, but that incident involved an airplane of a very different type and it was agreed by all the experts who testified that the protection provided by the all-metal fuselage and wings and the characteristics of the electrical, radio, and other equipment would make any significant electrical shock to persons virtually impossible on the airplane under consideration. While the electrical effect of lightning frequently produces a magnetic field which may temporarily or permanently influence metal parts, compasses, or other instruments, there is no record of serious results therefrom on aircraft in flight.

The electrical effect of lightning has been known to produce a sudden and tremendously increased noise or vibration in the receiving unit of telephones or radios of such intensity as to produce acoustical shock, especially under circumstances where earphones are held closely to both ears. In such an event, characteristic marks are usually left upon the diaphram of the receiving unit. Although the earphones of the plane were not recovered for examination, the testimony of experts indicated that an electrical charge sufficient to produce acoustical shock upon the pilot probably would not reach the earphones due to the design and installation of the radio system in the airplane involved in the accident.

There is no reason to believe that any electrical effect of lightning other than accoustical shock may have been even a contributing cause of the accident.

The optical effect of lightning, or of any sudden and bright light, is well known. Mrs. Everhart testified, that the lightning flash which occurred in the vicinity of the airplane while she was watching from a distance of about 1½ miles was so brilliant that she was blinded for a brief time during which she lost sight of the airplane. Experts testified that the nature of lightning illumination was especially injurious to human eyesight and a considerable number of pilots have reported temporary impairment to their eyesight from lightning flashes near their aircraft while in flight for varying lengths of time depending upon such conditions as the brilliance of the flash, whether the pilot was looking directly at the lightning, the degree of darkness in the cockpit and outside the plane. Some persons have reported that the blinding effect of a lightning flash in their vicinity has lasted for a good many minutes or even hours with a recurring after image and other forms of interference with normal eyesight which have been noticeable for several days after.

It is evident that the cloud toward which the airplane was flying was very dark and that the airplane was quite close to it at the time of the lightning flash. If the pilots had been looking forward through the windshield at the time the flash occurred in front of the plane, the result could have been so severe as to produce virtual blindness throughout the dive of the airplane to the point of impact.

It is obvious that the sudden blinding of a pilot might seriously interfere with his efficient control of an airplane in flight. While the blinding of the pilots in the present case might have been an added and grave complication in attempting to regain control once the dive had started, it is not believed likely that the optical effect of lightning of itself represents a basic cause for the airplane to change its normal flight attitude and dive toward the ground.

The mechanical or pressure wave effect of lightning was described at some length by one of the expert witnesses, Dr. Karl B. McEachron, an electrical engineer of the General Electric Company, who has been in charge of lightning research for the past several years and is generally recognized as an authority of great eminence on the subject.^[10] In one type of lightning stroke, there is a relatively long time, slow discharge of small value current which produces a burning or thermal effect. However, there is a very different type of lightning stroke which lasts an extremely short space of time but involves a very high value of current and results in a mechanical effect. The first type of lightning may set a tree on fire, whereas the latter type may splinter a tree and toss heavy pieces of wood a considerable distance away. Brief, powerful discharges set up pressure waves with characteristically steep fronts which are capable of a damaging effect at a short distance away from the path of the lightning itself. It should be noted that a single lightning discharge may involve not only mechanical but also thermal and electrical effects. Thus a severe lightning stroke involving a high value of electrical current may disintegrate a tree while also setting it on fire.

If the particular discharge of lightning with which we are concerned in this investigation were of a nature to produce a strong pressure wave, the pilots might have suffered from acoustical shock or concussion of a severity depending upon the proximity of the lightning to the cockpit and whether or not the cockpit windows were open. Experience has revealed that such pressure waves have produced severe concussion upon human beings which might result in unconsciousness. The possibility of acoustical shock or concussion as one of the basic elements in this case cannot be dismissed.

Pilots quite frequently have reported that when a flash of lightning has passed near their planes they have felt a slight "bump." This would be a result of the pressure wave, but all available data indicate that there never have been any serious effects upon the stability and flight path of an airplane from this cause. We do not believe that lightning itself produced such severe turbulence as to have caused or contributed to this accident.

The mechanical effect of lightning has been known to smash in the windows of buildings even though the discharge itself may have passed some little distance away. Although the cockpit windows were designed, installed, and approved to withstand any reasonable impact or pressure load, it appears entirely possible that a lightning flash near the nose of the plane might result in their being smashed in. The plane must have entered a torrential downpour of rain at the time of, or immediately following, the flash of lightning in its vicinity, and if the cockpit windows had been smashed in, the pilots might have been subjected not only to the violent impact of flying pieces of glass but also to a withering stream of water striking them with all the force of the airplane's speed through the air of 150 miles an hour and upwards. The smashing of the cockpit windows as a result of an unusually powerful lightning discharge near the nose of the plane with a consequent serious interference with the pilot's control of the airplane remains a possibility.

No airplane has been reported to have suffered structural damage resulting from the mechanical effect of lightning while in flight. Furthermore, Dr. McEachron testified that the available data indicate that the destructive force of lightning decreases in proportion to the altitude above the ground and, as already mentioned, it appears that the airplane in the present case was flying at an altitude of 6,000 feet at the time of the lightning flash. However, as Dr. McEachron said during his testimony, "We are looking here for the unusual thing, not the ordinary thing," and the rare chance of serious damage to the airplane or its controls caused by lightning should not be overlooked. If the mechanical effect of lightning may have been a factor causing the accident, the smashing impact with some part of the plane's structure typically would have left none of the usual indications of a lightning strike of a thermal nature. A violent pressure wave conceivably could damage the tail of an airplane to such an extent as to cause the loss of all normal flight control, but the testimony of eye witnesses as to the relative position of the airplane and the flash of lightning would lead to a conclusion that the lightning discharge occurred in front of the plane rather than to one side or in back.

According to Dr. McEachron's testimony, the greatest damage by lightning occurs at or near the earth terminal of a brief, high current value discharge. It is natural to consider the destructive force of such a discharge at or near its cloud terminal. There is some reason to believe that the airplane in this case must have been just about to enter the cloud and its rainstorm at the exact moment of the flash of lightning. It is useless to speculate as to whether or not this airplane may have been in the cloud terminal of the lightning discharge or what may have been the effect of its being there since Dr. McEachron pointed out the limited knowledge on the subject and the obvious difficulty of research.

Upon the evidence of record, we cannot conclude that the airplane was struck by the usual type of lightning which produces a thermal effect; that the airplane or its crew were injured by any electrical effects of lightning other than acoustical shock; or that lightning itself produced any turbulence which changed the flight attitude of the plane. We, therefore, conclude that none of these phenomena of lightning were related to any cause of the accident.

We do think it possible that lightning may have temporarily blinded the pilots or that the pressure wave resulting from the lightning may have subjected the pilots to acoustical shock or concussion; may have smashed the cockpit windows, or may have caused other damage to the structure and controls of the airplane through mechanical effect.

1. See Appendix C.
2. Chapter II, Section on Course and Position of Aircraft immediately prior to the Accident.
3. See Chapter II.
4. This point is discussed in detail in the section of the present chapter dealing with Mechanical Failure.
5. Section on Turbulence.
6. Over-revving is a condition which occurs when the engine speed exceeds the revolutions at which its governor is set. The maximum speed at which the governor is permitted to be set is the rotative speed which the engines are permitted to attain at takeoff, in this case 2350 rpm.
7. The flotation gear carried by Pennsylvania-Central Airlines is to safeguard passengers in the event of a forced landing in crossing over Lake Erie and Lake Michigan. It consists of two rubber rafts which can be instantly inflated by releasing a valve in a carbon dioxide tank that is carried with each raft for that purpose. Life preservers to accommodate all passengers and crew are also carried on the airplane.
8. Stall. A wing is said to be stalled when the angle at which it is moving through the air becomes so large (usually as a result of attempts to reduce the speed too far, or climb too steeply) that the air no longer flows smoothly along the upper surfaces of the wing, but eddies irregularly, resulting in sudden changes of lift and great instability.
9. Mr. Harrison is employed by the United States Weather Bureau as an Assistant Meteorologist and has participated in the work of the Sub-Committee on Lightning Hazards to Aircraft of the N.A.C.A.
10. Dr. McEachron is Director of High Voltage Research Laboratory, General Electric Company; Member of Sub-Committee on Lightning Hazards to Aircraft of the National Advisory Committee for Aeronautics.