Commerce Department Report on Hindenburg Disaster (Part Two)
The whole metallic structure of the craft was bonded.
Part III. – The Landing Maneuver
With respect to the meteorological conditions in which the landing was conducted, a summary of the general weather is given as well as the local conditions prevailing at Lakehurst at the time of the accident.
The 7:30 A.M., EST. U.S. Weather Bureau map of the vicinity, including the northeastern tier of states, Shows a disturbance over central New York and northeastern Pennsylvania, with a cold front extending from this center Southwestward to West Virginia. This front separated neutralized polar air to the east of the cold front which had become warmer and more moist and neutralized colder air to the west of the front. The warmer and more moist mass of air covered the Middle Atlantic states, southeastern New York and southern New England.
The cold front advanced eastward during the day from central Pennsylvania at a rate of 12 to 15 m.p.h., passing Lakehurst shortly after 3:30 P.M. There was not quite sufficient surface heating during the early afternoon to set off a thunderstorm at Lakehurst, and it was not until the front passed and some slight lifting of the air mass occurred that a thunderstorm began. The records of the Naval Air Station show that the thunderstorm began at 3:43 P.M. and ended at 4:45 P.M.
Telegraphic reports indicate, the thunderstorms in and to the west of New Jersey were not severe; nor were they of a well defined squall character. Between 12 P.M. and 1:30 P.M. E.S.T., these storms extended in a definite belt over the region of Harrisburg, Pa., northeastward to Bear Mountain, N.Y., and New Hackensack, N.Y. Between 1:30 and 2:40 P.M., none was reported. Between 2:40 and 3:40 P.M., Camden and Fort Monmouth, N.J., only reported thunderstorms. Between 3:30 and 4:30 P.M., Lakehurst, Mitchel Field, N.Y, and Floyd Bennett Field, N.Y.,reported them. Between 4:40 and 5:40 P.M;. none was reported; and between 5:40 and 6:40 P.M., Floyd Bennett only reported one. Summarized, the thunderstorms in eastern New Jersey were of a local character and not severe.
The New York Weather Bureau office bulletin issued at 1:20 P.M., May 6th, follows:
“1800 G.C.T. Moderate wind shift with increasing and lowering clouds possible thundershowers New York and vicinity expected in middle or late afternoon Stop New York scattered cumulus and small cumulo nimbus approaching from west – visibility excellent surface wind south 12 miles – barometer 29.68 falling steadily temperature 66.”
With the passage of the front at Lakehurst, the wind shifted to the northwest with gusts up to 20 knots, and was accompanied by slight increase in barometric pressure, decrease in temperature, heavy showers and several thundershowers. Then there followed a rapid decrease in the velocity of the wind and its direction became variable. The wind at Lakehurst at 6:10 P.M. went into the southeast and remained there for about 45 minutes, shifting again, and then it became mostly southerly. The front, after passing about 3:30 P.M, E.S.T., apparently slowed down to a rate of approximately 7 miles an hour and was in the vicinity of Atlantic City, N.J. at 8:00 P.M., its direction being north northeast-southwest, clearing rapidly after 8:00 P.M. During the afternoon cumulo nimbus and cumulus clouds developed locally and with the approach of the front there appeared a well-defined mild squall line in the west, which moved slowly over Lakehurst and apparently became stationary between it and the shore line until about 5:30 P.M., when it continued eastward. Several heavy showers occurred between 5:00 and 6:00 P.M., with accompanying thunder. Visibility was reduced during these showers. At 5:12 P.M. the thunderstorm then over the field was moving north, and it was believed that by the time the ship arrived at the station the storm would have moved away from the station. The ship at this time was out of sight because of low visibility and the ceiling, in the direction from which it was expected to approach, was not more than 500 to 500 feet.
Conditions at the time of the approach were: Ceiling between 2000 to 3000 feet; clouds .7 stratus; very light rainfall; sky showed signs of clearing to the westward; barometric pressure 29.72; temperature 60° F.; relative humidity 98; surface wind light, variable and shifting and at the precise moment of the
beginning of the landing was southeast 1 knot. It was effected that the surface wind direction would go into the west or perhaps the northwest. Reports from Trenton and Camden, N.J. indicated that the wind was westerly and that at Camden it was about 18 knots just previous to the landing of the ship. Wind at top of the weather tower on the field was west 6 knots. The approach level of the ship was about 200 feet above the ground. The top of the tower is 186 feet above sea level (ground elevation at place of landing was about 90 feet above sea level. The inversion condition was 60° at the lower level, 59° at the second, and 57° at the third level, being temperature readings at various levels from the top to the bottom of the weather tower. As the ship was approaching the landing area, occasional lightning was visible from the distant south and southwest, but none was observed over the field at this time. When the headway of the ship was stopped, a pronounced shift of wind was felt on top of the mooring mast, from southerly to southeast or south-southeast. This wind was colder than the previous wind had been.
Regular reports from the ship were received as scheduled at the Naval Air Station, Lakehurst. At one stage in the latter part of the flight the static was bad but it did not prevent communications between the ship and ground stations. Shortly before arrival at Lakehurst direct communication was maintained by the ship with the Naval Air Station.
At 1:55 P.M., E.S.T., the station received a message from the commander of the Ship stating that he would depart from Lakehurst as soon as possible after arrival. At 4:42 P.M.., the commander of the station radioed the ship: Conditions still unsettled recommend delay landing until further word from station advise your decision. At 4:52 P.M., the commander of the ship replied: “We will wait till you report that landing conditions are better.” At 5:12 P.M., the commander of the station advised the ship: “Conditions now considered suitable for landing ground crew is ready period thunderstorm over station ceiling 2000 feet visibility five miles to westward surface temperature 60 surface wind west-southwest eight knots gusts to 20 knots surface pressure 29.68.” At 5:22 P.M. station commander radioed ship: “Recommend landing now.” At 6:00 P.M. station transmitted to ship: “0vercast moderate rain diminishing lightning in west ceiling 2000 feet improving visibility surface wind west-southwest four knots gusts under 10 knots surface temperature sixty-one pressure twenty-nine seventy.” At 6:08 P.M. station commander sent last message: “Conditions definitely improved recommend earliest possible landing.” This was acknowledged by the ship.
Prior to the accident all of the ship’s trailing antennas had been reeled in. No high frequency transmissions were being conducted when the trail ropes were dropped from the ship. Both transmitters were turned to the “off” position at that time and remained so thereafter. The radio dynamotors had also been shut off. The last message transmitted over the ship’s radio was shortly after the landing station signal had been sounded, about 15 minutes before the fire. It was sent on the long-wave transmitter to Lakehurst at 6:10 P.M., E.S.T. During the landing, watch was kept on the long-wave receiver. No landing report was transmitted from the ship to Germany while it was over the field at Lakehurst. One of the ship’s radiomen stated that atmospheric disturbances had been encountered during the afternoon of May 6, but that such condition improved toward evening and continued to improve during the last 30 minutes of the flight. No difficulty was experienced during that period in sending or receiving either on the short or long-wave transmitters or receivers. Witness Herbert Dove, ship radio operator stated that he was on watch and actually listening to the radio until the fire started and that he did not notice any interference which could have been caused by improper bonding or shielding and that he did not receive any interference such as might have been transmitted by local station.
There was no oral communication between persons in the ship and on the ground during the maneuver.
The sequence of actions in bringing the ship up to the landing point is in part revealed pictorially by the track of the ship over Lakehurst, drawn on map of the Naval Air Station, with notes on the maneuver by Witness H. W. Bauer (See Appendix III). Among other data the map provides information respecting successive altitudes, speed, operation of engines, released of ballast and valving of gas.
Operation of Engines
About 10 minutes before dropping the bow trail ropes, the engines were running full cruising speed ahead; ship’s speed about 33 meters per second (approximately 73 m.p.h,). The altitude of the ship, according to its altimeters was then about 180 meters (590 feet). About 8 to 9 minutes prior to the release of the ropes all engines were idled ahead; altitude 150 meters (492 feet); ship’s speed falling of to 15 meters per second (approximately 33 m.p.h.) Then, in fairly rapid order the after engines were idled astern and then put full astern to reduce the speed to 12 to 13 meters per second (approximately 27 m.p.h.);after which all engines were idled astern; altitude at this time was 120 meters (393 feet). About 2 minutes prior to dropping of the bow trail ropes all engines were put full astern for a period of about one minute to stop the ship; after which the forward engines were idled ahead and the after engines were idled astern. When the trail ropes had been dropped the forward engines were given a short burst ahead; then idled ahead.
Release of Ballast
Starting at a point about three-quarters of a mile from the landing point 300 kilograms (661 pounds) of water ballast was dropped from ballast bag at Frame 77. Then in rapid order, from the same frame, at about intervals of 1000 feet, ballast was dropped twice again, the second time, 300 kilograms (661 pounds) the third, 500 kilograms (l,l00 pounds). This release of 1,100 kilograms (2,420 pounds) of water ballast took place within a period of 2 to 3 minutes before the trail ropes were dropped.
Valving of Gas
According to Witness H. W. Bauer’s sketch, gas was valved on the wheel for 15 seconds approximately 10 minutes before dropping the bow trail ropes; ship proceeding at full cruising speed. About 6 minutes prior to dropping of ropes, gas in Cells 11 to 16, first five forward cells, was valved for 15 seconds; ship then proceeding at l5 meters per second (approximately 35 m.p.h.). Approximately 4 to 6 minutes before dropping the ropes, gas in Cells ll to 16 was again valved for l5 seconds; speed of ship 12 to 13 meters per second. (Approximately 27 m.p.h.) About 2 minutes prior to dropping of ropes, gas in Cells ll to 16 was valved for 5 seconds.
Crew as Ballast
According to the elevator man who had taken over the elevator helm in the landing approach, the ship was still slightly tail heavy after dropping water and valving gas, consequently six men of the crew were sent forward to the bow in order to equalize the weights. He was unable to account for the tail heaviness of the ship after the ballast had been dropped.
The ship was weighed off to the west of the field and was found a little light. There followed the trimming operations that have been described in the preceding paragraphs. There is evidence to show; that the tail of the ship was heavy during the maneuver. Witness Albert Sammt, second in command of the ship, accounted for this condition by saying that it was due to the consumption of fuel; that it gave him no concern because it was very little. There was diversity of opinion advanced regarding this condition of the ship. Witnesses H. W. Bauer and C.E. Rosendahl considered it to be normal. The latter stated that the ship’s tail heaviness had been logically accounted for. Under the circumstances in which it landed in a light wind with little air flow on the tail surfaces and consequently little aerodynamic lift, 120 pounds midway from the tail of the ship would be felt by the elevator man and be noticed by those in the control car who were watching the inclinometer for that very thing; that the condition did not exist from the time of the dropping of the bow trail ropes during the 4 minutes intervening before the fire broke out.
To other witnesses the ship appeared heavy in the stern, among them witnesses Benjamin May, in charge on top of the mooring mast, and W.A. Buckley, Assistant Mooring Officer. Witness Hugo Eckener indicated, according to his information, that while the ship may have remained in satisfactory trim from the time the trail ropes were dropped until it burned such interval was a short period of time. He did not think that a hydrogen leak would have been so large that in such a relatively short time it could have been noticed. He mentions the testimony of Witness H. W. Bauer, relating to the trimming operations in which a very short time before the accident 6 men had been ordered forward. From this he infers that shortly before the ship reached the landing position it was necessary to trim ship by putting weight forward, and that the elevator man could hardly have noticed anything during this interval because the ship had no more forward speed. He further stated that careful calculation showed that the trimming moment effected by these operations amounted to at least 70,000 to 80,000 meter kilograms (506,391 to 578,933 foot-pounds) of trimming effect; when this effect is compared with the trimming moment that could be obtained aerodynamically at full cruising speed by the use of the elevator controls in the order of 150,000 to 200,000 meter kilograms (1,085,124 to 1,446,820 foot-pounds), then it became clear to him that the ship was very badly out of trim.
Witness Eckener also testified that witnesses in the control car had reported that the out-of-trim condition originated approximately one-half hour before the landing maneuver after going through the rain clouds; that the ship became tail heavy by running through heavy rain because the weight of the rain is greater in its effect on the horizontal fins, which are behind the center of gravity. There is also another apparent effect of rain upon the ship. That is the tail would seem to be heavy to the elevator man while the ship was running through rain, because it automatically has a tendency to nose up since the center of aerodynamic pressure moves aft. This effect, however, disappears very rapidly after passing through rain and in the present instance must have disappeared quickly because the ship as a whole was light. The ship, ten minutes after passing through heavy rain clouds, should have again been in good trim. In the opinion of Witness Eckener, however, it appeared so tail heavy that it became necessary to apply a trimming effect of some 70,000 meter kilograms (506,391 foot-pounds), Furthermore, he indicated that if the ship had been as tail-heavy before it proceeded through the rain clouds, it would not have been operated without the release of ballast. As no testimony was given that ballast had been dropped before the ship moved into the rain clouds, Witness Eckener believed that some unusual condition in the ship might have developed prior to the ship’s landing.
With regard to the amount of rain that the ship had been exposed to during the landing maneuver, there appears to be some difference of opinion. Witness Sammt, stated that there was a little rain as the ship crossed the field at the beginning of the maneuver, not heavy enough to weigh the ship down as much as 500 kilograms (1100 pounds); that was the only rain experienced during the last two hours of the flight because they had avoided the rain carried in the weather front. As the ship took a final bearing on the field it made a wide turn into quiet weather, returning to the field in this condition. According to him, the front had passed and the weather was favorable for landing. The sky was overcast but without disturbances or squalls. Witness Nelson Morris, a passenger, stated that a very light rain fell exactly as the ship came over the field the last time, but until that time there had been no rain, Witness Anton Wittemann, who had commanded the airship GRAF ZEPPELIN, stated that when the Hindenburg approached for its landing maneuver and as it passed through the front, the weather conditions as seen from the ship were entirely favorable; the thunder storm had passed into ordinary rain. The ship entered somewhat heavy rain which became much lighter when closing in on the station. At the approach there were no cumulus clouds; there was a clear-cut stratus layer from which light rain falling. Witness H. W. Bauer, second watch officer of the ship, said that about 20 minutes before the landing approach the ship passed through a heavy rain and through stratus clouds containing rain before making the approach. It did not pass near any 1ightning.
Altitudes at Landing
When the ship was brought to a stop over the landing points its altitude was about 180 feet above the ground. It rose to about 200 feet Then the bow port landing ropes checked its further upward rise. Thereafter, it descended to about 135 to 150 feet when the accident happened.
According to Witness Philipp Lenz, Chief Electrician of the ship, no fuses blew nor did any circuit breakers operate just prior to the fire. The several circuits of the ship were intact, the interior ship lights and the navigation lights were burning as usual.
Two witnesses testified that the top and bottom rudder did not appear to be working in unison when the ship came over the field. From other testimony it appears that the rudders were functioning normally.
Part IV. – The Fire
It was the practice at the Naval Air Station to maintain a log of events in connection with the landing of the Hindenburg. The log of its last landing reveals that the first approach of the ship in landing maneuver was sighted at 6:15 P.M., E.S.T., May 6, approximately over the Officers Quarters on the station. At 6:21 P.M., the bow trail ropes were dropped, on a bearing of 30 degrees from the mooring mast, first the starboard rope, followed immediately by the port rope. Ship was first observed afire at 6:25 P.M.
Description of Landing
The landing made on this occasion has been described as a high landing or flying mooring, a method of landing which is occasionally employed. Some qualified witnesses stated that it was normally conducted in every respect. Among these were Witnesses Rosendahl and A.F. Heinen. Others indicated that the approach seemed hurried; that the ship made what seemed to be a fairly short turn and approached the mooring circle fairly rapidly. Based upon the statements of other witnesses, Witness Eckener expressed the view that the ship must have proceeded in a sharp turn to approach for its landing, Witness Sammt said the turns were normal.
Incidents Before the Fire
Before the fire broke out, the ship was being held by the bow port trail rope which had been coupled to the port yaw line and a strain had been taken on this rope around the niggerhead of the ground winch. The bow starboard trail rope had not been coupled to the ground line, but was being handled by the starboard bow landing party. At no time during the approach did the ship come closer to the mooring mast than 700 feet. The main bow cable of the ship at this time had been let out about 50 feet, but neither it nor any of the cables or ropes in the stern had reached the ground before the fire started. After the trail ropes in the bow had been dropped, the ship no longer had any forward speed. It began to move up and astern and also to swing slowly to starboard. Then a light gust was felt from port.
Fluttering of Outer Cover
Witness R.H. Ward, in charge of the port bow landing party, a couple of seconds before the fire, had his attention attracted by a noticeable fluttering of the outer cover on the top port side between frames 62 and 77, which includes cell No. 5. No smoke or other disturbance accompanied the flutter when he first saw it. It was a wave motion. In his opinion the motion of the surface was not due to the slip stream or resonance effect of the propeller. It was entirely too high from the propeller. It appeared to him to be more like an action of gas inside pushing up, as if gas was escaping. He apparently had seen this action occur in other aircraft. The ship had no perceptible forward motion the time he observed the flutter; its engines were idling in forward rotation. The fabric had not opened up when he first made the observation. The flutter was followed by a ball of flame approximately 10 feet or so in diameter then came an explosion. On a diagram this witness indicated that the first appearance of fire was near the top of the ship and above the point where he saw the flutter. With respect to this testimony, witness Eckener said that a leak in a gas cell, permitting the escape of 40 to 50 cubic meters of gas per second, would be sufficient to cause a flutter in the outer cover which could be observed as reported, but probably would not be enough to draw the attention of those in the control car to a loss of buoancy aft. Witness R.W. Antrim, who was on top of the mooring mast, also stated he saw that the fabric behind the after port engine was very loose and fluttering. It extended rearward and upward from the after port engine to a quarter of the way to the tail.
Strain on Port Trail Rope
The drift of the ship to starboard, according to the mooring officer, Witness Tyler, was finally checked by means of the port trail rope. This rope was hauled up taut on the winch. The starboard trail rope was being handled by the manpower of the starboard bow party. Witness Albert Stoefler, one of the ship’s cook’s, who was looking down from a window in the ship, stated that he “saw how the landing crew came running up, and how they loosened the knot of that rope and fastened it to the lower lines on the ground. Then I saw how the ropes took tension and at the moment I felt a very strong detonation of the ship, vibration of the ship.*** I did not notice any explosion. I only noticed that vibration I was speaking about before.” He thought the ship was striking the mooring mast. Witness H. W . Bauer stated that after the landing rope had been fastened, he went from his position to the port window in the control car and observed the tensioning of the landing ropes. At the time of that observation, there was a strong shock in the control car and his first assumption was that the landing rope had broken. Witness Max Zabel, ship’s third officer, stated that he observed the bow trail ropes being dropped; that the port trail rope became rather tight. He saw the ends of the ropes which were tied together whirl around and tighten. Immediately after this landing rope had become tight, an explosion was heard and the destruction of the ship occurred. He described the vibration that was felt in the control car as an extraordinary one. Witness Dowe, ship radioman, testified that while watching one of the landing ropes being handled by the ground crew there suddenly was some tearing in the ship, a metallic tearing. A passenger reported, “and then as that rope was getting taut, I heard a detonation.***”
Sensations Within the Ship
In describing their nervous reactions at the beginning of the accident, some of the persons within the ship, in addition to such descriptions as are provided in the preceding paragraph, spoke, in effect, as follows: Witness Severin Klein – When the ship was almost standing still, it gave a sudden jolt. Witness Xavier Maier – First he heard detonation; then he noticed the vibration, the shock, ant fell on his back. Witness Heinrich Kubis – First heard or felt an explosion approximately at the time that the ship took a sharp inclination. Witness Lenz – The sound that he heard he thought might have been a landing rope breaking. Witness Claus Hinkelbein – The jerk and the sound of the detonation and the sight of the fire or the reflection of fire were all simultaneous. Witness Kurt Bauer – Noticed a cracking shock which originated in the rear. Witness Witteman – When he heard dull detonation, thud, his first idea was that rope had parted. Witness Water Ziegler – Saw how the port landing rope was hauled tight; shortly thereafter he heard a dull thud or detonation and a heavy shock went through the ship. Witness Kurt Schoener – It was a strong shock he sensed after hearing a rather dull detonation. Witness Sammt – His first intimation that something was out of order was a heavy push, about the same shock as if the ship had been rushed to the side and the landing rope had broken. Neither prior to nor after the bush did he hear a muffled explosion. He did not associate the push with anything that might have occurred in the after part of the ship.
Appearance of Fire
Numerous expert and lay witnesses on the field testified as to where they first observed the fire on the ship. There was great diversity in this testimony for reasons that are very apparent. Among the most important of these reasons were the extreme rapidity with which the fire spread, the different positions of the witnesses with respect to the ship, the size of the ship, more than one-sixth of a mile in length, and an over-all height, equivalent to a twelve story building, and the fact that at the time of the fire it was still daylight. It is estimated that the interval between the first glimpse of flame and the impact of the main body of the ship with the ground was 32 seconds. The great majority of the ground witnesses who testified as to the first appearance of fire were looking at the port side of the ship.
After carefully weighing the oral evidence and transcribing to a master diagram the numerous diagrams on which the ground witnesses indicated their first observations of fire, we conclude that the first open flame, produced by the burning of the ship’s hydrogen, appeared on the top of the ship forward of the entering edge of the vertical fin over Cells 4 and 5. The first open flame that was seen at that place was followed after a very brief interval by a burst of flaming hydrogen between the equator and the top of the ship. The fire spread in all directionsmoving progressively forward at high velocity with a succession of mild explosions. As the stern quarter became enveloped, the ship lost buoyancy and cracked at about one-quarter of the distance from the rear end. The forward part assumed a bow-up attitude, the rear appearing to remain level. At the same time the ship was settling to the ground at a moderate rate of descent. Whereas there was a definite detonation after flame was first observed on the ship, we believe that the phenomenon was initially a rapid burning or combustion — not an explosion. From the observations made, it appears that there was a quantity of free hydrogen present in the after part of the ship when the fire originated.
A brief resume of the observations made within the stern of the ship shows that Witness Helmut Lau, who was standing on the ladder leading up to the lower catwalk from the lower vertical fin and was looking up facing the port side of the fin, heard above him a muffled detonation and saw from the starboard side, down inside the gas cell, a bright reflection on the front bulkhead of Cell No. 4. He saw no fire at first, but a bright reflection through and inside the cell. The cell suddenly disappeared because of the heat. Then Cells 3 and 5 caught fire. This witness said he did not see the center of the origin of the fire, but it must have been further up since he saw the reflection of fire through the cell wall material. It was the same type of explosion that one hears when using a kitchen gas range, when first lighting the flame or turning it off. Witness Lau did not smell any hydrogen at the time he made these observations. Witness Hans Freund was letting out the after mooring cable at Frame 47 and had let out a few meters of it when he heard a muffled detonation. Fire was simultaneous with the explosion. He was surrounded by fire immediately. Witness Rudolph Sauter, who was stationed in the keel of the lower vertical fin, first heard a dull detonation, then saw fire in Cell No. 4, a big fire, which he identified as a hydrogen fire. None of these witnesses in the stern of the ship felt any unusual vibration or heard any breaking of structures prior to the detonation or the sight of fire or reflection of fire. None of the other members of the crew or passengers on board the ship observed fire or reflection of fire until after feeling an unusual vibration or shock or hearing the detonation.”
Part V. – The Combhustible Mixture and its Ignition
Having retraced the course of events and the circumstances surrounding the accident, we come to the question, why did the fire occur? As yet, with the few excentions to be noted, no more has been provided than a hypothetical aproach to the answer. We have weighed the severs1 theories that have been advanced.
The possibility tnat the cause is to be explained by premeditated or willful act has received active attention. Sabotage has been examined under two classifications; the first — external, including the use of incendiary bullet, high powered electric ray, and the dropping of an igniting composition upon the ship from an airplane; the second classification — internal – including the placiing within the ship of a bomb or other infernal device. To date, there is no evidence to indicate that sabotage produced the grim result.
In consideration of accidental causes, two factors must be found together. There mysr be present (a) a combustible mixture of hydrogen and oxygen of the air; and (b) sufficient heat to ignite such mixture. In the analysis of the evidence the mixture and its ignition are treated separately.
Presence of Combustible Mixture of Hydrogen and Air
Accumulation through Diffusion or Osmosis
While it is conceded that the fabric of which the cells were made is slightly permeable to the diffusion of the contained hydrogen, it is not our opinion that this characteristic of the cell walls, under the circumstances prevailing, would account for combustible accumulation of gas and air within the ship; the normal rate of seepage being, as was indicated under description of the cells, about oneliter per square meter per 24 hours.
Failure of Valve Mechanism
According to the testimony, only one valve failure had occurred on the ship. This happened when the ship was new; as a consequence, certain changes had been nade in the construction of the mechanism. In any event, the failure noted occurred to
an automatic or pressure relief valve which would not have been functioning at the time of the accident. However, because the valves were mechanical devices, it was possible that there might be a defect or failure in them, but no testimony appears to show that this possibility was a likely one.
Another query regarding the presence of such mixture presented itself. Could it have been due to the reduced scavenging of the gas by the ship’s ventiletion system during the last ninutes of the craft’s existence when its speed eventually had been reduced to a full stop, combined vith the last valving operation, about six minutes before the fire? This theory seem improbable because of what was said about the efficiency of the ventilation system and because of the fact that the chimney effect created by the 6 knot vind that tsas bfoning at the ship’s elevation during the last four minutes prior to the fire, should have evacuated practically all of the gas from the shafts. The forward speed of the ship, reported to have been from 15 to 20 knots per hour [sic], when the last valving operation was performed, should have been ample, it was stated, to have cleared the gas rapidly from the ship. A further argument made with regard to the scavenging of gas was that immediately after the last reported valving the ship’s engines were backed down hard, and that this deceleration
should have tended to move the gas in the ship toward the bow and out through the forward gas shafts.
In considering the production of such mixture by the rupture of a cell or cells, there are at least several avenues to explore.
Entry of a Piece of Propeller
One of these might be laid to the failure of a propeller and the throwing of one of its fragments through the adjacent part of the hull into a cell. To this possibility there was devoted an extensive examination by experts of our staff and those of other agencies. The condition of the propeller of engine car No, 2 attracted our attention. Witness F. W. Caldwell, one of this country’s foremost propeller experts, was quite certain that the propeller of the after port engine did not break in flight but was shattered at the time the car struck the ground. He said that there was no indication of the separation of the sheathing from the blades except as the result of shattering on impact. Witness Deutsche, machinist in the after port engine car, indicated that the propeller of his car was still rotating when it struck the ground; that he did not feel any unusual vibration of the engine before the crash.
Fracture of Hull Wire
One other significant possibility must be discussed while the question of cell rupture is being examined. It was suggested that, while in flight, a tension wire might have ripped a hole in a cell and thus permitted a quantity of gas to escape. Coupled to this possibility is the testimony of Witness R.H. Ward, digested briefly in the statment of facts; that he saw
a fluttering in the outer cover above the equator between rings 62 and 77 and believed that this fluttering was caused by gas escaping into the space between the adjoining cell and the outer cover. A shear wire in one of the panels at the place from whicht he gas was escaping could have snapped while the ship was turning during the landing maneuver. Witness Eckener stated that such turns generate high stress in the after part of the ship, especially in the center section close to the stabilizing fins which are braced by shear wires. The gas thus accumulated between the cells and the outer cover must have been a rich mixture. Such a mixture, enclosed in a space between the outer cover and the gas cells, would, if ignited, burn with relatively slow speed until gas in greater volume was released by the burning through of the cell walls. Witness Roseadah1 recalled that in the early years of operation with naval aircraft, shear wires had broken with varying effect, causing no serious damage, however.
Consideretion has been given to the possibility that a major structural failure in the stern of the ship caused the hydrogen to be liberated by rupturing a cell and forcefully breaking an electric lead or metal part, thus producing a spark. The fire broke out when the port trail rope, which held the ship to the ground, became taut. It was reported by some persons that at, or about, the time they observed the fire they heard a cracking sound fron the stern of the ship. An examination of the nreckage disclosed that the rivets, by which the after end of the axial corridor mas connected, through a fitting to the hull, had pulled out; that all of the radial wire in the small frame nearest the stern had broken in tension; that only a few of the small tabs of metal from the periphery of the frame, vhich had been pulled off the bight formed where the radial wires hooked on to the frame, were found on the ground below where the frame struck. The shearing of the rivets and the condition of the wire and the frane might be explained by the force with which the rear end hit the ground; or by the
torsional or other stresses which the tail suffered in its last moments in the air. It has also been pointed out that the ship was stressed for greater loads than the tensional strength of the bow trail rope, and that the rope had not parted. Furthermore, it was observed that the eye through which the trail rope was attached to the ship and the longitudinal member to which the eye was affixed, were intact after the accident. The four members of the crew in the stern of the ship testified that they did not hear or see any such structural failure prior
to the fire.
Ignition of the Mixture
Many of the theoretic aspects of the ignition of the combustible mixture were dealt with at great length by a number of experts. Only a summary of this phase of the investigation is related in this report.
If there had been enough heat generated by the friction of wires or other members of the ship coming forcibly into contact with each other, due to structural failure or breaking, a sufficiently hot spark might have been produced to set off such mixture. There is insufficient evidence to sustain a conclusion based upon this theory.
As has been stated, there are metals which have a catalytic effect upon a mixture of hydrogen and air and would materially lower its ordinary ignition temperature, but it does not appear that any such metal was in that part of the ship where the fire was first observed.
Under the title of chemical possibilities there has also been suggested that a flame might have been produced by spontaneous combustion. The evidence is inadequate to support this theory.
In the examination of thermodynamic possibilities much time at the outset of the investigation was given to the possibility of such mixture being ignited by the sparks from the engine exhausts. It was suggested that sparks or larger particles of carbon thrown out from the diesel engines might have been carried into the openings in the lower part of the hull or have been blown over the exterior of the stern and there have ignited such mixture. While the circulation of the exhaust gases, set up by the direction of rotation of the propellers just before the accident (the after engines idling in reverse and the forward engines idling ahead) was different from that produced while under way, it was maintained by the German experts that this circumstance would not result in sparks or carbon particles reaching the interior of the hull, and, furthermore, that the sparks would not have been able to ignite such mixture on the top of the ship at least 165 feet away from the after exhaust outlets. Witness Ludwig Duerr testified that very extensive experiments respecting this possibility had been conducted by the builders and the results had been reassuring. When the engines are delivering 1100 to 1200 hp. the temperature leaving the piston before it enters the exhaust stack is 500° to 530° centigrade, The temperature of the exhaust is lower. The engines ordinarily develop 800 to 850 hp. At this output the temperature of the exhaust gases is 450° to 480° back of the cylinder. With a mixture of air sucked in, the temperature is reduced to 230° to 250° centigrade. Visible sparks have a temperature over 500° centigrade but lose their heat rapidly as they are impelled through the air.
Had this been the cause of the ignition, it is believed that it would have come into play before the elapse of the four-minute interval between the dropping of the trail ropes and the accident.
That the heat of the exhaust gases caused the havoc is also improbable. If ignition had happened at the exhaust it would have been necessary that the temperature of the band of air between the outlets and the place of the first flame would have had to be about 507° centigrade. According to Witness Duerr, the temperature at the exhaust outlets was much lower than 507° C. With the Hindenburg and the Graf Zeppelin, no difficulties had been experienced from this quarter.
Under the classification of electrical sources of ignition several were considered. A combustible mixture of air and hydrogen could have been ignited by the overheating of wires carrying current within the ship, e. g., by a short circuit. Barring the possibility previously alluded to, of a substantial failure in the stern structure of the ship, which might have produced a sudden breaking of such wires in the aft end of the ship, it is thought to have been only remotely possible that the mixture was fired by a defect or failure of the ship’s electrical wiring.
According to Witness Lenz, who was stationed in the electrical power plant at the time of the accident and had most of the ship’s electric indicators, fuses, and circuit breakers, under observation, the various circuits were functioning normally just prior to the conflagration. No fuse blew or circuit-breakers operated at that time. It was also observed that the cable carrying the current to the stern light was very sturdy and was installed so as to provide plenty of slack to compensate for expansion and contraction of the frame of the ship.
.A theory introduced by Witness Heinen was that the cause of the fire was due to the ignition of such mixture in one of the gas fulness or pressure electric meter actuating units fixed to the axial corridor in the vicinity of cells No. 4 and 5. He believed that a small pocket of gas accumulated in the folds or ridges of the cells surrounding the corridor and found its way into the inner recesses of the meter and was there ignited by an electric spark; that the fire thus created traveled up along the radial wires to the space between the cells and the outer cover igniting the free hydrogen collected along the longitudinals at the top of the ship on the inner surface of the outer cover; that the relatively slow burning of such free hydrogen would account for the peculiar manifestations of illumination described by certain witnesses; that the fire in the second sequence then destroyed gas cell No. 4, as seen by Witness Lau.
With regard to the presence of gas in one of the meters it was estimated that in one hour the seepage in the axial corridor would have amounted to one-fortieth of one per cent of the volume of the corridor; that even in the motionless condition of the ship, the corridor would have been well ventilated due to the chimney effect created by a wind of six knots blowing over the gas shafts; that the ventilation in the corridor would have prevented pockets of hydrogen from forming because the air current through the corridor was not laminated but was made up of whirls and eddies. However, if it could be shown that a rent occurred in a cell below the axial corridor, then it is possible that some free hydrogen might have found its way into one of the meters. In regard to the ignition of such mixture within a gas pressure or fullness, meter the following is quoted from a report of the Bureau of Standards, relating to Exhibit 74, one of the meters taken from the ship:
“It is evidently intended for measuring and giving a remote indication of small gas pressures by electrical means. The gas pressure acts on a diaphragm in oppostion to a helical spring. A plunger attached to the diaphragm carries a coil of wire which has a resistance of 100 ohms. Two rollers, connected in parallel, make contact with the sides of the coil. Two flexible connections run to the ends of the coil. The change in the relative resistances of the two parts of the circuit between the contact rollers and the ends can cause suitable electrical inoicating instruments in the control cabin to indicate the position of the coil and diaphragm and hence the pressure.
“All electrical parts are enclosed in a cylindrical metal box. The only openings into this box are (1) the hole, 10 mm in diameter at the top through which the operating rod passes with a clearance of not over 0.5 mm and (2) the opening at the bottom which is completely filled by the 3-conductor cable (covered with metallic braid) which connects to the rest of the circuit. The conical housing surrounding the metal box is well ventilated.
“The device seems to be excellently designed and constructed from the standpoint of safety, and there appears no way by which it could with any reasonable probability have caused a fire.
“An overheating of the device by short-circuit seems impossible. A short-circoit external to the device would impose on it only the full voltage (24 volts) of the circuit and produce a rate of heat dissipation of less than 6 watts. A short-circuit inside the device would not draw more than the 1 milliampere fixed by the external instruments. A simultaneous short-circuit both inside and out would blow a fuse, if one was present, before a dangerous temperature was reached. Good practice requires such fuses on all such circuits, and one was probably used.
“The normal operation of the device should produce no sparks. Deterioration of the contact rollers or of the coil, or a breaking of a wire inside the metal box might produce a spark inside. It seems impossible that hydrogen should be present inside as it could get there only by diffusion down the narrow clearance between the operating rod and its guide tube, 50 mm long. A spark could be produced outside the box only by the breaking of the 3-conductor cable.
“This cable is strengthened by the metallic braid and runs in a protected location along the structural member. It could not be determined whether or not the cable was definitely anchored to the member, nor whether the metallic braid was originally clamped to the metal box, because of damage in the fire.”
In the light of all the available evidence on this point we believe that the possibility of igniting such mixture by the means just described was very slight.
An attempt was made to discover if the ignition of such mixture could have been laid to spark emission due to resonance effect upon metal parts of the ship’s interior caused by received radio waves of high frequency.
There was on the field at Lakehurst a localizer beam radio transmitter of low power, maintained by an airline company, the on-course portion of which was so situated as to pass through the space occupied by the ship at the time it took fire. This transmitter was at that time about 1800 feet from the ship. Its power output was 15 watts; its frequency 278 kilocycles. The maximum field strength authorized for this type of station is 1500 microvolts per meter at one mile, which represents fifteen ten-thousandths of one volt, per meter measured at one mile on the on-course portion of the range which, incidentally, is the area of weakest radiated power. The strength of this field is so low that it has been compared to the power of a fly. So far as could be determined, this localizer was the only transmitter that was operating at Lakehurst at the time in question. It is not believed that other high-fresuency stations, at some distance from the field, could have had inductive effect upon the airship.
Witness Dieckmann, of the German Commission, stated that he and his colleagues had been particularly interested in the possibility of ignition through high-frequency radio induction, especially after hearing the testimony of Witness Freund who was engaged in paying out a length of the stern cable at ring 47 when the accident took place; that this part of the cable might have received impulses and thus electrical energy would have been conveyed into the inside of the ship. However, it appears that if such result was to occur due to inductive effect, a transmitter relatively close to the ship and of considerable power would have had to be operating at the time of the event. These conditions were not present.
Resonance effect due to high-frequency generation within the ship was impossible because all the ship’s transmitters had been shut down before the appearance of fire. Furthermore, the ship was carefully shielded against resonance effects generated from within. Witness J. B. Whitehead put no stock in this possibility because of the small amount of energy that could have been involved. Furthermore, once inside the ship in the form of oscillations in the structure no damage could have been done, because the structure itself was so large and so complex that there was no possibility of a small amount of energy setting the whole ship in oscillation and that oscillation in separate parts, which perhaps contained high resistance, would be short-circuited by other parts of the ship. In view of the facts and the expert testimony given on this possibility, it may be said that in such inductance there was only the remotest chance that it was responsible for the elusive spark.
Under this designation of electrical possibilities there is now to be considered a group distinguishable from current electricity and known as electrostatics. In this group, there is first mentioned a possibility due to the nature of the materials employed.
In the older type of cell fabric, containing a rubberized element, it was apparently possible to create a static spark by tearing the fabric. The cell fabric used in the Hindenburg as far as we could learn, did not include material possessing this characteristic. Since virtually all of the cells were consumed by the fire, no test could be made of the cell fabric.
The cause of the accident was the ignition of a mixture of free hydrogen and air. Based upon the evidence, a leak at or in the vicinity of cell 4 and 5 caused a combustible mixture of hydrogen and air to form in the upper stern part of the ship in considerable quantity; the first appearance of an open flame was on the top of the ship and a relatively short distance forward of the upper vertical fin. The theory that a brush discharge ignited such mixture appears most probable.
South Trimble, Jr., Solicitor
R. W. Schroeder, Asst. Director, Bureau of Air Commerce.
Denis Mulligan, Chief, Regulation and Enforcement Division, Bureau of Air Commerce.
Officers and Crew on board the Airship Hindenburg on its departure from Frankfurt-am-Main Germany, on May 3, 1937 were as follows:
* Captain Ernst Lehmann
Captain Max Pruss, Commanding
* Willy Speck, Chief Radio Operator
* Franz Eichelmann
Rudolf Sauter, Chief Engineer
* Wilhelm Dimmler
* Ludwig Felber
* Ernst Huchel
* Alfred Bernhard
Philip Lenz, Chief Electrician
*Ludwig Knorr, Chief Rigger
*Frau Imhoff, Stewardess
Dr. Ruediger, Ship’s doctor
Xaver Maier, Chief Cook
Al fred Groezinger
Werner Franz, Mess Boy
Captain Anton Wittemann
*Indicates those who died in accident.
Passengers on board the Airship Hindenburg on its departure from Frankfurt-am-Main, Germany, on May 3, 1937, were as follows:
Adelt, Gertrude Berlin, Germany
* Angers, Ernst Rudolf, Dresden, Germany
Belin, Peter,Washington, D. C.
* Brink, Birger
Clemens, Carl Otto, Bonn, Germany
* Doehner, Hermann, Mexico City, Mexico
* Doehner, Irene
* Dolan, Curtis, France
* Douglas, Edward, New York
* Erimann, Fritz
Ernst, Elsa, Hamburg, Germany
* Ernst, Otto C.
* Feibusch, Moritz, Lincoln, Nebraska
Grant, George, London, England
Heidenstamm, Rudolf von
Herschfeld, George, Bremen, Germany
* Knoecher, Erich, Zeulenroda, Germany
Leuchtenberg, Wm., New York
Osbun, Clifford, Chicago, U.S.A.
* Pannes, Emma, New York
* Pannes, John
* Reichold, Otto, Vienna, Austria
Vinholt, Hans, Copenhagen, Denmark
* Indicates those who died in accident.”