27

On this wiki page (under 'Comparison with heavier-than-air aircraft'), the altitude record of the L-55 Airship states:

The ultimate altitude record for a rigid airship was set in 1917 by the L-55 under the command of Hans-Kurt Flemming when he forced the airship to 7,300 m (24,000 ft) attempting to cross France after the "Silent Raid" on London. The L-55 lost lift during the descent to lower altitudes over Germany and crashed due to loss of lift.

I understand that lift on an airship is created through internal heated air and/or gases, rather than by the movement of air over a wing.

So, how exactly can an airship lose lift? (And how does altitude affect this?)

Cloud
  • 13,280
  • 8
  • 81
  • 150
  • 1
    L-55 was LZ-101 in the numbering scheme used by Luftschiffbau Zeppelin. The run of airships from LZ-100 to LZ-103 were designed for 7300 m altitude - their gas bags were only ⅓ full at sea level! LZ-104 was even designed for 8000 m … – Peter Kämpf Feb 28 '18 at 20:33
  • 1
    One solution to your question title: Pop it with a pin... – dalearn Mar 07 '18 at 13:57

2 Answers2

22

The reason is a temperature difference between the lift gas and the surrounding air, and probably water uptake by the hull when descending through clouds.

A given mass of hydrogen will create a constant lift force, regardless of pressure or altitude, when at equal pressure and temperature with the surrounding air. Therefore, a change in altitude will not change the lift a rigid airship creates. Ideally.

However, air gets colder the higher you climb. The sun heats the atmosphere from below, by way of heating the ground, and space cools it from the top. Therefore, on many days the temperature gradient is bigger than its adiabatic value - that is the way thermals work! L-55 stayed at altitudes where air is -32°C according to the Standard Atmosphere. When descending, the surrounding air got warmer and also warmed up the lift gas, but only slowly. This means that, depending on the rate of descent, the lift gas lagged behind in temperature relative to the air, and this temperature difference reduced its lift capacity.

Note that adiabatic heating will already heat a gas when being compressed. The lapse rate of the atmosphere must be above the dry adiabatic value for this mechanism to work, which it is on many days. Especially behind a cold front. Note that L-55 encountered strong winds - just what you find in and behind a cold front. So it is safe to conclude that L-55 flew in labile air, and when it descended, that motion became unstable, at least close to the ground.

Kapitän Flemming simply descended too quickly. Slowing things down would have warmed the lift gas more, and less lift would have been lost. But delaying the descent has dangers of its own: The gas bags back then were made from goldbeater skin and had a certain amount of seepage. To compensate, Zeppelins started their trips with several tons of ballast water onboard, which was progressively dropped during the many hours of a normal trip. Taking a detour over France delayed the trip, so he was running short on time.

Normally, a loss of lift can be compensated by dynamic lift. With some angle of attack, an airship can create up to 20% of its weight in dynamic lift - as long as all engines are running. L-55 was blown south during the night after attacking Hull and Birmingham, and found itself far more south than assumed when dawn allowed the crew to get a ground fix. When back over Germany, L-55 ran out of fuel and dynamic lift was no longer available to compensate for the lower lift gas temperature. It made a rough landing in the Thuringian countryside near Tiefenort and had to be written off.

Literature: Heinz Urban, Zeppeline der kaiserlichen Marine 1914 bis 1918

Peter Kämpf
  • 231,832
  • 17
  • 588
  • 929
  • 1
    I just thought that the temperature of the gas will decrease on ascent due to approximately isentropic expansion, and vv on descent. If I am right in guessing that this is (almost) the same for H2 and air, the gas temperature/height profile will lie between the dry adiabatic lapse rate and the actual lapse rate. If these assumptions are correct, and the airship starts down with the gas at the same temperature as the surrounding air, will it not reach the ground at least as warm as the air there, on account of compression, given the actual lapse rate is never > adiabatic? On the other hand... – sdenham Mar 01 '18 at 02:55
  • ...a descent in the morning would be during the time when the lower atmosphere is being heated by the sun. I need to learn more about how the diurnal thermal cycle varies with altitude... – sdenham Mar 01 '18 at 02:57
  • @sdenham: How do thermals work? On many days the lapse rate is above its stable value, so adiabatic warming will not compensate for what is really happening on the way down. October 20, 1917, was such a day. The sun heats the atmosphere from below, by way of heating the ground, and space cools it from the top. Of course that makes the temperature gradient larger than its adiabatic value. – Peter Kämpf Mar 01 '18 at 06:59
  • What time of day day did the descent occur? I think solar heating is the answer, especially as it occurred over the duration of the descent, so that the lapse rate as experienced by the ship was greater than any that existed at any one time. From my recollection of my college meteorology, the lapse rate is not often superadiabatic, and then usually close to the ground, because as soon as it becomes so, convection redistributes the heat - see. IIRC, fair-weather thermals result from surface superadiabacity in a generally ~adiabatic air mass. – sdenham Mar 01 '18 at 12:24
  • @sdenham: Sometimes the air is unstable through most of the troposphere - that is why thunderstorms are possible. If a cold front passes over western Europe, the air is unstable for several days at a time. Only high pressure means stable air with sun-induced instability close to the ground. – Peter Kämpf Mar 01 '18 at 21:39
  • These phenomena are driven by conditional instability from the latent heat of condensation and the difference between dry and wet lapse rates, where the latter is less than the former. This efficiently transfers heat upwards and tends to reduce the lapse rate below the dry rate. If this sort of weather was a necessary factor in this specific accident, perhaps it should be mentioned in the analysis. More generally, as the issue is temperature differences, the compression-heating of the gas should be considered in the analysis, as it is commensurate with the dry adiabatic and exceeds the wet. – sdenham Mar 02 '18 at 00:11
  • @sdenham: I suggest to start gliding and do less theorizing. When you fly gliders you will experience what lapse rate there really is and you will learn quickly that "this sort of weather" is the norm. Also, it's OK if you don't believe my answer, but then please do your own research. – Peter Kämpf Mar 02 '18 at 03:04
  • SE comments are a horrible way to discuss matters... I'm not avoiding your point about the conditions behind a cold front, but if the air was unstable that deep for that long, convection would be active 24 hours a day in that time even without condensation, and we glider pilots would not have to wait for the sun to soar. I think what happens is that whenever instability forms, convection brings it back to the edge of instability (i.e. the dry adiabatic lapse rate, if it is clear) ready for the sun to launch thermals by heating the ground (after compensating for overnight radiation cooling.) – sdenham Mar 02 '18 at 03:13
  • I see that our posts have crossed, and if you don't want to address the physics behind your answer, that is fine by me. BTW, the US Navy's inquiry into the loss of the Macon may have some relevance here... I will follow up when I have finished my research. – sdenham Mar 02 '18 at 03:15
  • @sdenham: For the accident to happen it would have been enough for the over-adiabatic lapse rate to exist for the last 1000 m. I was mentioning thunderstorms only because you seemed so sceptic that over-adiabatic conditions could range through most of the troposphere. And regarding thermals: For the vertical convection to start it needs a temperature difference of several degrees. An unstable atmosphere will just stay that way if not "kicked" by some bubble of heated air rising from the ground. Thermals becoming stronger with altitude are clear proof of an over-adiabatic lapse rate. – Peter Kämpf Mar 02 '18 at 08:35
  • 1
    Assuming that this is a significant variance from adiabatic and is the explanation for the L-55 accident, then it is something that could happen to any airship flying in the lower troposphere in such conditions, and is to some extent a matter of insufficient reserve ballast. As the argument is that the L-55's gas was especially cold on account of the altitude it reached, then the question of compression warming exists for the descent from there to the lower troposphere. – sdenham Mar 02 '18 at 12:20
  • I forgot to say that I think the need for an additional several degrees to kick off convection is evidence of marginal stability, not instability, and perhaps that might also account for varying thermal strength. There is also a sense in which we may be talking past one another: a km-deep adiabatic surface layer on top of a 10 m superadiabatic layer will appear superadiabatic for any pair of measurements where the lower one is at ground level, even if the dT/dA is only superadiabatic in the lowest layer. – sdenham Mar 02 '18 at 13:24
  • @sdenham: Did you never wonder why gliders and balloons don't fly at the same times? Hot air balloon rides are either early in the morning or late in the afternoon, but never with thermals around. And they fly nowhere near the climb and descent rates of Zeppelins. – Peter Kämpf Mar 02 '18 at 16:22
  • 1
    What I am mostly wondering right now is why you would risk your reputation with such a blatantly evasive non-sequitur. Clearly there are no answers to be found here, but maybe the physics SE will be more helpful. – sdenham Mar 03 '18 at 04:52
  • @sdenham: Oh boy, I thought that is obvious. The atmosphere has an unstable temperature gradient. On many days. What is needed to get convection started is either the sun heating through the ground inversion and producing overheated air, or and airship which ascends or descends. The process needs a certain kick to get started, either a thermal or an airship. Maybe you ask a new question; I agree that comments are terrible for explaining things. What you think is "proof" of marginal stability is simply inertia in an unstable system. – Peter Kämpf Mar 03 '18 at 07:18
  • 1
    What should have been obvious is that evidence of the effects of convection is not evidence for that convection requiring deep super-adiabacity. And if inertia is the answer for the delay in convection, and this deep super-adiabacity persists for days on end, why does the convection not continue into the night, or be triggered by night flights? More relevantly, you are persistently avoiding the real issue, which is, even in the presence of this hypothetical kilometer-deep super-adiabatic layer, what effect did compression-warming have on the L55's gas during most of its descent from over 7km? – sdenham Mar 06 '18 at 14:54
  • Maybe the answer lies in the strong wind from the north at altitude, that turned the silent raid into a debacle for the attackers, rather than soaring-type weather in the lower troposphere? – sdenham Mar 06 '18 at 18:56
  • @sdenham: Yes, the strong wind gives a hint, but not what you think. First, back then weather forecasts were rather shaky and Zeppelin navigation, over clouds and at night, was based on expected winds. That L-55 was blown south was less due to the strength but rather due to the wind's unexpected direction. Strong winds equals cold front, and behind a cold front there is high reaching lability - higher lapse rate than adiabatic, in other words. That was L-55's undoing, and it doesn’t matter that you still refuse to believe it. It happened nonetheless. – Peter Kämpf Mar 07 '18 at 01:01
4

While Peter Kämpf has addressed the L-55's case, the question asks about loss of lift accidents in general, and the loss of the USS Macon (ZRS-5) provides an additional perspective.

As mentioned elsewhere, rigid airships flew with the lifting gas at the ambient pressure: their lifting cells within the envelope were only partially inflated on the ground, and as the craft gained altitude, they expanded. If they reached their full extent, any additional gain of altitude beyond this ‘pressure height’ would lead to gas being released through safety valves, in order to avoid an overpressure that could rupture a cell.

The demise of the Macon started with the failure of the incompletely-repaired upper fin, which punctured one or more of the rear gas cells. The officer-in-charge’s response was to drop large quantities of ballast and fuel, causing the ship to zoom well above its pressure height, which was less than 3000 ft., leading to an additional loss of gas. This climb was exacerbated by the change in pitch resulting from the loss of gas from the rear cells, which was not fully compensated for by the elevator man, and which produced additional dynamic lift.

As pointed out by Peter Kämpf, venting gas from exceeding the pressure height is not generally sufficient to leave the ship deficient in buoyancy: given that the amount of gas remaining is sufficient to support it at this altitude, it is sufficient to support it at any lower altitude, so long as the gas is no colder than the surrounding air. This is a consequence of the ideal gas law (and the fact that hydrogen, helium and air are all very nearly ideal gases at atmospheric pressures and temperatures): a mole of one gas will displace a mole of another if they are at the same pressure and temperature, regardless of what that temperature and pressure is, and so, by Archimedes’ principle, will create a similarly-independent buoyancy equal to the weight of one mole of the displaced gas.

In the Macon’s case, however, losing additional gas would not have helped in dealing with the leakage from the punctured cells, and some forty minutes later, it settled on the water. It is the opinion of historian Richard K. Smith that the excursion above the pressure height was decisive, and without the additional loss of lifting capacity that it caused, the Macon may well have remained airborne. He believes that mishandling of the ship led to dynamic lift contributing to the zoom, in which case the above analysis is not necessarily sufficient, as in the presence of dynamic lift (or upwards momentum), we cannot assume the ship was buoyant above the pressure height.

This is essentially the mirror image of Peter Kämpf's argument: if the airship did not contain sufficient gas to be statically buoyant at the apex of its trajectory, then it did not contain sufficient gas to be so at any lower altitude, a situation that could only be remedied through dynamic lift or by jettisoning weight - something the crew was working on until almost the last minute. Once it became apparent that a crash was likely, the commander had to confront the choice between slowing down or endangering everyone on board, with the former robbing the craft of dynamic lift.

In the case of the L-55, launching with its cells only one-third filled, the pressure height would have been about or perhaps a little above its record-breaking altitude, where the density is about one third of that at sea level. The pressure height of an airship is not fixed in construction, but by the degree to which it is filled before launch.

sdenham
  • 1,348
  • 8
  • 27
  • 1
    Flying close to the maximum height was avoided in order to not have the pressure-limiting valves vent the hydrogen gas. Bursting cells were much less of a concern - that needed failed overpressure valves first. And the temperature lag effect works both ways: In case of the Macon, the lift gas was warmer than the surrounding air and accelerated the climb, once that got started. This effect is even sometimes called the aerostatic phygoid. – Peter Kämpf Mar 02 '18 at 16:25
  • 1
    @PeterKämpf I do not think anyone is suggesting there was a danger of cell bursting, given the presence of automatic relief valves. It subsequently occurred to me, however, that in the absence of such valves, perhaps the biggest danger of over-expanding cells might be to the structure of the airship rather than to the cells themselves - especially in a lightly-built height-climber (though the Macon was more strongly built, following the loss of the Shenandoah.) – sdenham Mar 06 '18 at 15:02