Print

[Irvine John Donald]

One of the main issues, as I see it, with the Navy and Coast Guard search theories is the whole discussion on fuel consumption. The evidence suggests the AE had the proper fuel reserves for the trip and the knowledge to manage her power plants effectively. My question for the forum members is simple, if you're knowledgeable in such matters as "flying".  With the instrumentation of the day, could AE have been burning though her reserves much faster than she thought?  Was the Electra capable of burning 25% more fuel per hour than normal without doing serious damage to the power plant?  If AE determined early on in the flight, by reading fuel gauges, that she was using more fuel than expected, would she the not have aborted the attempt and found nearest airfield to set down on?  It seems to me that because she didn't abort that her fuel was burning at the expected rate and she therefore had the fuel reserves at Howland that she expected.

I ask because it seems like this is one of the key search criteria used. One school of thought being the radio messages suggesting she had one half hour of fuel left at Howland. Therefore spend huge amounts of the overall time searching locally for a ditched aircraft.

[Martin X. Moleski, SJ]

One of the main issues, as I see it, with the Navy and Coast Guard search theories is the whole discussion on fuel consumption. The evidence suggests the AE had the proper fuel reserves for the trip and the knowledge to manage her power plants effectively. My question for the forum members is simple, if you're knowledgeable in such matters as "flying".  With the instrumentation of the day, could AE have been burning though her reserves much faster than she thought?  Was the Electra capable of burning 25% more fuel per hour than normal without doing serious damage to the power plant?  If AE determined early on in the flight, by reading fuel gauges, that she was using more fuel than expected, would she the not have aborted the attempt and found nearest airfield to set down on?  It seems to me that because she didn't abort that her fuel was burning at the expected rate and she therefore had the fuel reserves at Howland that she expected.

I ask because it seems like this is one of the key search criteria used. One school of thought being the radio messages suggesting she had one half hour of fuel left at Howland. Therefore spend huge amounts of the overall time searching locally for a ditched aircraft.

Airplane fuel gauges are notoriously unreliable.  They say the only time you should believe them is when they read empty.

From the old Forum:


Date:         Wed, 17 Nov 2004 12:48:23
From:         Ric Gillespie
Subject:      Re: AE's Fuel

> How did Earhart track her fuel consumption on her flights?  What
> instruments or other advice would they have received (from Johnson,
> Mantz or others)?

We know that the plane was equipped with a Cambridge Exhaust Gas Analyzer and that it was considered crucial to fuel management.  We also know that they completed all of their previous legs without running out of gas or even having to make an unplanned landing enroute to refuel, so it seems safe to say that they were able to monitor their fuel situation adequately and that, by the time they got to Lae, they had lots of experience doing it.

> Do we know how they tracked their fuel on the other legs of the flight?

As far as I know, AE never specifically wrote about it but it seems safe to assume that she (not Noonan) tracked her fuel the same way everybody else did.  You how much gas you started with. You know from the book, verified by experience, how much fuel your engines use at given combinations of manifold pressure and mixture.  You subtract what you presume you have used from what you started with and that's how much you've got left.  If the fuel gauges pretty much agree with your calculations so much the better.  If not, you either have a leak or the gauges are wrong.  It ain't rocket science.

[Irvine John Donald]

Thank you Marty.  Very informative and makes me wonder how some technologies don't seem to advance as much as others. You would think that in today's world we could hope for better. 

[Martin X. Moleski, SJ]

Thank you Marty.  Very informative and makes me wonder how some technologies don't seem to advance as much as others. You would think that in today's world we could hope for better. 

The problems of measurement of any quantities of any kind whatsoever is an art and science all to itself.

The problems of reliably indicating the amount of fuel present in tanks of varying sizes and shapes under wildly varying conditions of temperature, vibration, and turbulence are not trivial.

Besides designing some kind of ideal sensors that are reasonably priced, that can survive in a hostile environment, and that accurately model the amount of fuel in a tank of a particular shape, that system needs to be calibrated and maintained in calibration for a particular aircraft: "Aircraft Fuel Gauge Accuracy."

If we imagine a perfectly uniform container (a cylinder or a cube) on a laboratory table, all you need is a visual ruler (a float, a set of lines etched on the class) to know with a reasonable degree of precision how much liquid is in the container.

Now make that container into an airfoil shape so that it will fit into a wing section.  Or cast it into the form of a tip tank.  Set the contents in motion in that irregularly-shaped tank.  Let the fuel foam and slosh forwards and backwards and undergo plus and minus G forces not only up and down but from side-to-side in various trajectories (a coordinated turn versus a slip, for example).

How many of these multiple forces and conditions will the model used by your system take into account?  With you equip it with accelerometers and gyros (more parts that can fail!) to smooth out errors caused by these variables?  How often will you update the display if it is digital?  By what means will you transmit the information if the system is analog?  What kinds of errors are introduced into the system by the methods of transmission of information that you choose?

What kind of safety margins do you want to build into your system?  Do you want to deliberately err on the side of caution, so that when the gauge reads "E" there is still a little bit of usable fuel left?  Or do you want the prop to stop turning at the same time the needle hits bottom? 

Gasoline expands when it is hot and contracts when it is cold.  Will you add a temperature correction to your system so that it accurately reflects fuel by weight instead of by volume? 

Suppose you decide to use a float as your basic gauge.  When the tank is full, the float is highly accurate.  No float can keep floating when the gas gets low.  At some point, the float hits bottom while there is, perhaps, some usable fuel still in the tank but not enough to keep the float up.  How much gas is that?  Should you put the float into its own cylinder that goes deeper than the bottom of the tank?  Should you try to isolate it from the wave action in the tank?  By what means do you measure and report the height of the float?  Do you average a set of readings periodically or simply let the gauges shiver with the changing conditions?

I am not a pilot.

I fly radio-controlled airplanes.

I read a lot about aviation.

I'm not sure that I've indicated all of the variables correctly.

I imagine that today's gauges are, on average, better than those of 1937.  Persuading pilots to use all of the information at their disposal to make sure that they don't run out of gas until after landing seems to be one of the best ways to improve aviation safety, but there seems to be room for improvement.

A friend of mine--an A&P and a CFI--took off on a short hop in a Taylorcraft to reposition it for delivery to a customer.  He ran out of fuel very shortly after takeoff, tried to turn back to his runway, and destroyed the plane among some trees.  To this day, he swears that there should have been five gallons in the aircraft and he is convinced that someone stole the gas.  Nevertheless, he should have stuck a stick in the tank before he took off--the most primitive and effective method of making sure that there is gas available at the start of a flight (as a general rule--I guess that even sticks can be deceived). 

I am an armchair pilot.  I serve in the battalion of Captain Hindsight.  I survey the results of others' mistakes.  I know what they should have done differently.  I'm not putting my money where my mouth is.  You may take the things I say with as much salt as your doctor allows in your diet.   

[h.a.c. van asten]

LJ.Dnld . The fist leg of 874-900 mls asked 10% more fuel burnt whereas a 44 mintes delay on ETA-Nukumanu was incurred . Recomputation (1996) showed that @ 1912 GMT the reserves were sufficient for 1h05m , the 100 oct special avgas included . The hourly mean gas consumption was (1,100 US gals.) / 20h17m = 54 gls. These figures when fed into the entire , 2729 mls  , flight´s performance formula give good compliance with 20 mph averaged headwinds and groundspeed 134-136 mph.  154 mph /54 gph ( 1 - 20 mph /154 mph) = 2.48 mpg , 2 mls taxi-and-run distance included. [154 mph = Ind.Air Speed , 54 gph = US gls/hr , 20 mph = av.equiv. headwinds]. The 1996 inquiry needs updating due to new vistas but however , figures given will not essentially decline. Initial fuel loaded was 1,050 gls 80 oct + 50 gls 100 oct (remaining after Lae take off).

[Chris Owens]

Off topic, but as an engineering curiosity and a follow up to Marty's post, here's one way to measure fuel level in a tank.  Sounds ridiculously overcomplicated on the face, but NASA did it in some spacecraft applications, where you have the added problem of no gravity to hold all the liquid fuel at the bottom of the tank.

Completely sealed system. Liquid fuel is contained in a rubber bladder which is inside a larger rigid tank. Pressurize the space outside the bladder with a slightly radioactive gas. Have a sensor that measures radiation. As you use fuel, the bladder shrinks, the radioactive gas expands, and therefore the space right in front of the sensor will have fewer atoms of the gas in it, hence a lower level of radiation.  Run the particle count  through a computer and presto: you have yourself a nice gas gauge.

I woulda thought you could have done it by measuring the pressure in the tank and correcting for temperature, but then again I'm not a rocket scientist.

[Martin X. Moleski, SJ]

Off topic, but as an engineering curiosity and a follow up to Marty's post, here's one way to measure fuel level in a tank.  Sounds ridiculously overcomplicated on the face, but NASA did it in some spacecraft applications, where you have the added problem of no gravity to hold all the liquid fuel at the bottom of the tank.

Completely sealed system. Liquid fuel is contained in a rubber bladder which is inside a larger rigid tank. Pressurize the space outside the bladder with a slightly radioactive gas. Have a sensor that measures radiation. As you use fuel, the bladder shrinks, the radioactive gas expands, and therefore the space right in front of the sensor will have fewer atoms of the gas in it, hence a lower level of radiation.  Run the particle count  through a computer and presto: you have yourself a nice gas gauge.

I woulda thought you could have done it by measuring the pressure in the tank and correcting for temperature, but then again I'm not a rocket scientist.

They may have started with the bladder first as a solution for getting the fuel to flow in a zero-G setting, then realized that they could create a radioactive dipstick as icing on the cake.

Control-line pilots have used pressurized fuel cells inside of tanks for many years.  I don't know whether their technology precedes NASA's.

"Combat and some speed models use rubber tubing ('bladder' tank) , baby pacifiers, or fountain pen ink bladders, inflated with fuel in a veterinary syringe, to hold the fuel under fairly high pressure. The fuel line is pinched off to prevent fuel loss until the engine is started. The high pressure of fuel delivery permits the use of a larger intake on the engine, allowing more air flow than would otherwise be possible, and thus more power. The fuel delivery is very steady with this method until the very last bit of fuel runs out, but is logistically difficult because any leak results in streams of fuel spraying out" ("Control Line," Wikipedia).

[h.a.c. van asten]

Chr.Owns. Seems easier to hire an anaerobic garden gnome who continuously by GSM reports until what body part he stands in the fuel , fly as cautiously as possible : no waves permitted , never fuel to full.

[Monty Fowler]

 

*goes back to pondering the many mysteriesl of Amelia*

[Thom Boughton]

Chr.Owns. Seems easier to hire an anaerobic garden gnome who continuously by GSM reports until what body part he stands in the fuel , fly as cautiously as possible : no waves permitted , never fuel to full.

Ummmmm....what?

[h.a.c. van asten]

One of the main issues, as I see it, with the Navy and Coast Guard search theories is the whole discussion on fuel consumption. The evidence suggests the AE had the proper fuel reserves for the trip and the knowledge to manage her power plants effectively. My question for the forum members is simple, if you're knowledgeable in such matters as "flying".  With the instrumentation of the day, could AE have been burning though her reserves much faster than she thought?  Was the Electra capable of burning 25% more fuel per hour than normal without doing serious damage to the power plant?  If AE determined early on in the flight, by reading fuel gauges, that she was using more fuel than expected, would she the not have aborted the attempt and found nearest airfield to set down on?  It seems to me that because she didn't abort that her fuel was burning at the expected rate and she therefore had the fuel reserves at Howland that she expected.

I ask because it seems like this is one of the key search criteria used. One school of thought being the radio messages suggesting she had one half hour of fuel left at Howland. Therefore spend huge amounts of the overall time searching locally for a ditched aircraft.

For zero wind 950 US gls was computed, with 10.5 % (regression factor 0.905 airspeed vs groundspeed consumption) addition for equivalent winds the 87 oc amounted to 1,050 gls . Add 50 gls 100 oc in wing tank and fuel load @ take off was 1,100 US . On the Lae-Nukumanu track wind (25 mph east) was out of forecast speed and compass (15 mph s-se) so as to ask 10% addional fuel burn (regress.factor 0.82 instead of 0.905). @ 1912 GMT  reserves were : 45 US , 23 of 87 oc , 22 of 100 oc for respective 33 min & 32 min flight time @ 41 gph. Thence, the 1912 Z fuel reserve was 1/2 hour for the 87 oc to finish the flight as planned , plus 1/2 hr for the 100 oc left after take off from Lae. The total reserve was for 1 hour , reasonable for the conditions & expectations given . The to Howland closest land points are Winslow Reef (210 mls) and McKean (350 mls) : neither the one , nor the other was within endurance range, whereas the A/c was not alighted @ Baker (35 mls).

[h.a.c. van asten]

I researched the fuel supply and management (quantitatively by aircraft performance theory) and published in the magazine of Royal Netherlands Air Force Museum , Soesterberg , 2 issues of 1996 . E.Long arrived at the same 1912 GMT reserves , but did not in his book deliver computations on the subject . The book appeared after 1996 but it is hardly possible that he knew my mentioned articles . H.

[Chris Johnson]

Have you got a link to your article or could you scan it and upload it to the forum for us to view it?

[h.a.c. van asten]

Articles need some updates , and I have only one of two available here since I am not at my home address now , will try nevertheless . If ready I will post signal . H.

[Chris Johnson]

Articles need some updates , and I have only one of two available here since I am not at my home address now , will try nevertheless . If ready I will post signal . H.

That would be good for the forum!