Highlights From the Forum
December 26 through January 1, 2000
Over the Christmas weekend I indulged myself by playing a computer game that I picked up several months ago but hadn't had time to even look at. It is, supposedly, a very popular fantasy detective game called MYST and I wonder how many forum subscribers are familiar with it.
You travel to a small, remote, uninhabited island strewn with the abandoned structures of the former inhabitants. On one side of the island is a wrecked ship. By carefully inspecting and deciphering the clues on the island you unlock its secrets and are transported to various "ages" in the island's past where you find more clues to ultimately solve the island's great riddle. Sound familiar?
MYST is actually an excellent training device for the scientific method of inquiry. You collect data, form a hypothesis, and test it. If your hypothesis is correct you are rewarded by obtaining access to another level of mystery to work on, but if you don't pay close attention or if you just flail around in unfounded speculation, you get nowhere.
But beyond that, I found it interesting that the game's creators chose an island as the setting for the mystery and it got me thinking about islands as cultural symbols of mystery and escape. Atlantis, Bali Hai, Never Never Land, Robinson Crusoe, etc., etc. A few years ago, just for the fun of it, I cut together video clips from the Niku II expedition with scenes from theatrical films as a humorous spoof to show at a team meeting. What struck me was how easy it was. Expeditions to mysterous islands are a Hollywood staple.
Perhaps, then, what makes the Earhart Project so captivating (at least, for some of us) is that the story we're uncovering, piece by piece, taps into something very basic in human experience and resonates as a tale we've heard many times before - except this time, it's real.
Some time ago you asked me if I thought Amelia's Electra had sufficient range to reach Gardner Island. At the time I had not looked at that possibility and wasn't sure; however, I thought it might be possible. Since that time I have done a much more systematic engineering study of both engine fuel consumption, and aircraft performance for each element of her flight.
Based upon my calculations, there was insufficient fuel to reach Gardner Island. At best, the airplane had a theoretical endurance of about 22 hours corresponding to a range of approximately 2,800 miles. To reach Gardner, she would have needed an endurance of around 24 hours and a range somewhere around 3,000 miles.
My work assumes the following:
What assumptions did you make about total fuel load, route of flight, and winds encountered enroute?
Once you've figured out Myst, you can graduate to Riven, a 5-CD follow-on which has 5 islands and about 5 times as many puzzles to figure out!
Gee, Ric, where have you been? Myst has been out for nearly 6 years! You need to get a regular life, besides this fixation on a dead woman.
I think we've established that I am not well versed in surviving rock stars, current fight promoters, and now, the latest video games. I suspect that it has been equally well established that I don't have anything approaching a regular life.
From Paul Chattey
Myst (One) and Myst (Two) are indeed, I think, popular fantasy detective games. They do offer some rather obvious parallels, like staying alert and looking where you don't expect to find clues---at about knee height in this case, and looking carefully while traveling in both directions through rooms or tunnels for clues. For weeks while I played it, Myst scenes appeared regularly in that state of consciousness between waking and sleep, solutions just slightly out of reach. Sound familiar? Infectious, fun, compelling, frustrating, and a delightful toy. It is also, as you point out, an excellent training device for the scientific method and having gone through both games, I'll go one step further. It's also, with hindsight, a tribute to simplicity. Or maybe things just look simple once we figure them out. I kept notes on and tried to decipher or at least assign meaning to those darn hieroglyphics for the longest time. Eventually I got on with the game and began finding secret doorways and handles, instead of obsessing on "scholarship".
Anybody know if there is a grandson-of-Myst game out there?
LTM, who seldom
gets mysty eyed in the face of facts
Randy says it's called "Riven".
From Bill Leary
> But beyond that,
I found it interesting that the game's creators chose an
One component of mystery (film, literature, etc.) is Suspension of Disbelief*. If the story is presented well enough, we can suspend our disbelief in the story and enjoy it more. An aspect that helps is Limited Scope* or, in this case, Limited Geographical Scope*. The wilder the subject, the easier it is to believe in it if it's far away in a small location with limited access. This "explains" why it's a mystery and why you haven't heard of it in your every day life. For a fairly recent example, this was used to good effect in the beginning of Hunt for Red October.
* "Suspension of Disbelief" is the actual term normally used in the field. The other two, I can't recall the normal terms used, but they're something like that.
Of course, the "willing suspension of disbelief" is a prerequisite for the enjoyment of any theatrical presentation. The interesting thing about the Nikumaroro mystery is that no such suspension is required. It's a classic case of life imitating art.
From Alan Caldwell
Ric, the important thing I got out of your note was that I'm not the only adult playing computer games.:-) Actually I BUY computer games. I have yet to do much in the way of playing. I DO have Myst so you have piqued my interest in it. Do you have "Around the world?"
I HAVE played one game which my son brought over and you are right it DOES require some heavy duty logic and good reasoning to play these things. Good experience for the job at hand.
My legal training and experience has certainly taught me a lot about putting together a good argument. Probably the most important lesson that also applies here is to be so very careful about misstatements and repeating anecdotal bits and pieces as facts. It is not so important that someone does that as it is that others pick them up and they take on lives of their own. A good solid case is built on tiny scraps of supportable evidence until they begin to eliminate other possibilities. Eventually one is left with an almost inescapable conclusion.
It is also good practice to game out possibilities in the hopes that some will obviously be eliminated. The worry is that we have to watch out for false rabbit trails and not get too carried away with irrelevant issues -- except for fun of course.
As the year draws to a close it might be helpful to post a brief summary to sort of get everyone on the same page for the start of the new year. I was just about to suggest content but I can see that is not all that easy. I know it's all on the web site but from postings over the last few months I'm not sure everyone reads it or understands it.
At any rate Happy New year to all,
No, I don't have "Around the World" but I don't even really have time to play MYST (much as I enjoy it). I have all I can handle playing NIKU.
The "Hypothesis and Plan" on the TIGHAR website at The TIGHAR Hypothesis is too long to post on the forum but I do urge that new forum subscribers familarize themselves with it. It will help make sense of the ramblings that happen here on the forum.
With respect to assumptions in my analysis, I selected a volumetric fuel capacity of 1,100 gallons, and a fuel weight density of 5.756 lb/gal at a bulk temperature of 80 degrees F. The weight density number is derived from Pratt & Whitney data.
A useable fuel fraction of 96 percent was chosen. In other words, a 4 percent loss due to evaporation, overboard venting, leakage, tank ullage volumes (due to the 3-point attitude of the airplane), and residual fuel in the feed system that cannot be pumped to the engines. In context with the latter, I was guided by a technical report from Bell Aircraft Corporation wherein a useable fuel fraction of 96.9 percent was measured on a P-39 (internal and external tanks). The increased complexity of the Electra fuel feed system warrants a slightly larger number.
Eight route scenarios were examined. Interestingly, the total mileage from Lae to Howland varies only a little bit depending upon what intermediate points along the route are selected. In any event, this analysis suggests that Amelia flew around 2,585 miles in 20.2 hours (her last radio transmission) at an average ground speed of 128 mph.
My analyses demonstrate that in theory Amelia should have had something like 442 pounds of fuel remaining after 20.2 hours. Similarly, her burn rate at 1,000 feet was about 247 pounds per hour (mixture ratio = 0.07, MAP = 24 in-Hg, rpm = 1,600). This gives a remaining endurance of around 1.8 hours, not enough to reach Gardner.
I believe Gardner Island is about 400+ miles south of Howland. At a ground speed of 128 mph, it would take a little over 3 hours to reach this destination.
Winds aloft were not considered because the analysis was based on total time and total distance. This yields average ground speed.
I agree with you on many points. I agree with you that we have to assume that:
I question three assumptions you have made.
1. Your estimated 4 percent unusable fuel (44 gallons) is based upon one example drawn from a single engined fighter and arbitrarily increased from that figure. I don't know, and you don't know, how much of Earhart's 1,100 gallons was unusable --- but we do know that NR16020 was set up specifically as a long-range, record-setting aircraft. We also know that the fuel system schematic shows a "stripping valve" and a "strip wobble pump" in addition to a "hand fuel pump." This certainly gives the impression that considerable attention was paid to making sure that the aircraft didn't carry around any more unusable fuel than was absolutely necessary. While it is true that some fuel may have been lost thorough evaporation and venting between the time the aircraft was fueled on July 1st and the takeoff more than 24 hours later, it is also true that the tanks may have been topped off prior to departure. (Remember, one of our assumptions is that Earhart and Noonan are competent.) In short, I think your 4 percent unusable fuel assumption is pretty shaky, and those 44 gallons represent about an hour of flying time.
2. I'm a bit puzzled by your assumption that the "Amelia flew around 2,585 miles in 20.2 hours (her last radio transmission) at an average ground speed of 128 mph." It seems quite clear from the 19:12 GMT transmission ("We must be on you but cannot see you...) that the crew believed that they had reached the immediate vicinity of Howland Island and had looked around enough to convince themselves that it wasn't right there. The transmission comes very close to Earhart's scheduled quarter-past the hour transmission time so it seems reasonable to presume that they reached what they thought was Howland's position sometime right around 19:00 GMT. Everyone seems to be in agreement that Noonan was attempting to find Howland by the only means available to him, that is, advancing the sunrise LOP --- and everyone seems to agree that he should have been able to do that with some degree of accuracy in an east/west sense, but not in a north/south sense. In other words, it seems most reasonable to assume that at about 19 hours into the flight the aircraft was on a 157/337 LOP running through (or near) Howland Island but was either to far north or too far south to see the island. This is the time that represents the end of the enroute portion of the flight, not the final transmission heard an hour later. From 19:12 on they're searching for Howland, or any place to land. The 157/337 LOP running through Howland is roughly 2,550 statute miles from Lae (a little less if north of Howland, a little more is south). An airplane covering that distance in 19 hours has made an average groundspeed of 134 mph. For Earhart, cruising at the recommended 150 mph, that would mean an average headwind component of about 16 mph for the entire trip --- not at all remarkable.
3. I don't understand how you conclude that, at the time of the last transmission heard by Itasca at 20.2 hours, the aircraft is 400+ miles from Gardner. If the aircraft was directly overhead Howland at that time it would be 410 statute (356 nautical) miles from Gardner, but that's the one place we know it wasn't. What we do know is that, for the past hour, the airplane had been searching for Howland and, when last heard from at 20:13 was "running on the line." Because the rational direction to search farthest on the line is to the southeast, and because the flight did not reach Howland, it seems most likely that the aircraft was south of Howland and perhaps quite a bit closer to Gardner than you theorize.
I respectfully submit that the endurance calculations need more work before anyone can conclude that the aircraft could not have reached Gardner.
This really interests me as I've worked the figures available backwards and forwards and can't come up with anywhere over 18 hours for an arrival time and 3 hrs remaining fuel.
I've been looking at all the suggested distances for weeks and the estimates of distance from Lae to Howland vary tremendously. The most frequently quoted seems to be "over 2500 miles".
I have checked with software, and various other references and the only way I can get anywhere near that is by using Statute Miles.
I've just been over the figures again (this time with my trusty computer) and I still get the Great Circle Bearing as 079 deg and the Great Circle Distance to be 2263 Nautical Miles. (close to 2600 Statute miles). I even went whole degrees in opposite directions to factor in some extra distance. (the real figure comes up around 2240+)
Early in the flight AE reported "wind 23kts". The only wind a pilot is interested in reporting is either a Headwind (which if she was sitting on a planned 150kts would drop her back to around 125 ground speed) or a tailwind (which would see her speeding along at around 175 kts (200Mph).
So, assuming a 25kt headwind for the whole flight, and using Statute Miles (2600 from Lae to Howland (Great Circle), we have to convert knots to miles per hour (125kts = 144mph). 2600/144 = 18 hrs Lae --- Howland.
(checking... 2265nm/125kts = 18hrs) it works.
52Gph (worst KJ
figures all the way) x 18 hrs = 936gallons.
All at the worst figures from the Kelly Johnson telegrams.. If AE was able to fly at 10,000 at full throttle for the last 9 hours as recommended she was expected to use 100 gallons less fuel.
What am I not seeing? (This is a serious question - I'd like to know what I missed.. and why it doesn't work).
I would like to suggest that we all throttle back for a minute. The question of the airplane's probable endurance is, of course, central to any hypothesis regarding its fate. Our earlier calculations, based upon a 1,100 gallon fuel load and Kelly Johnson's recommendations, gave us an estimated maximum endurance of around 24 hours and 10 minutes. Others, including Elgen Long and now Birch Matthews, have offered dissenting viewpoints.
We had contemplated commissioning a "blue ribbon panel" of experts to look at the problem but I'm beginning to think that this question mught be best addressed, step by step, right here on the forum. For one thing, if a panel of experts did come up with an answer we would demand to see their work anyway.
Before we start:
Thanks for your response to my present conclusion that Amelia lacked adequate fuel to reach Gardner Island. I appreciate your perspective because it made me think and reconsider what I have done thus far. Following are my thoughts to your items 1 and 2, the 4 percent fuel fraction that could not be used, and the 2,585 mile assumption. As to your comments in item 3, last radio transmission and assumed position, let me defer responding at the moment. You made me rethink this situation.
1) I agree that the useable fuel fraction value I chose is open to debate. However, I stated my estimate and the basis for this number. If anyone has a more realistic number I am more than ready to reconsider. Of one thing I am certain. The useable fuel fraction was not 100 percent. For instance, in all probability some or all of the tanks had small ullage volumes. Typically this is a design consideration to allow for some fuel expansion due to heating. Tank ullage on Amelia's airplane could also occur simply due to the aircraft three-point attitude with respect to the tank filler port position. If this caused some ullage volume, topping off the tanks on the day of departure would not changed this condition. From a more general perspective, I applied the "useable fuel fraction" concept to account for various uncertainties that come into play to one degree or another. This is a common engineering approach in many if not most analyses.
You also mention the modification to the fuel tank feed system made in the spring of 1937. Someone, quite possibly Kelly Johnson, recognized that not all of the fuel could be drained from the all of the tanks. In fact, the residual amount was probably appreciable and something had to be done. The engine-operated fuel pump could not provide sufficient suction to maximize fuel removal from the greatly expanded fuel system. This resulted in the addition of a "stripping valve" and "strip wobble pump." The stripping valve is simply a two-position control device integrating the additional wobble pump into the existing fuel feed system. The strip wobble pump acts as a hand-operated booster pump. It no doubt improved the situation, but would not have achieved the 100 percent goal.
I am probably preaching to you about things with which you are already quite familiar. If so, please forgive me.
2) Your comments about my assumption that "Amelia flew around 2,585 miles" (to Howland) in 20.2 hours is well taken. I tend to agree that this understates her average ground speed based upon her radio transmission of 19:12 GMT. I do not think a distance of 2,250 miles is the appropriate measure, however. Earhart and Noonan did not fly a great circle route towards Howland, not exactly at least. They deviated slightly to the south early in the flight. Depending upon which track one prefers (after the initial position fix report to Lae, some mileage was added to the nominal great circle route.
I considered eight navigation scenarios (one each of which included a 30 nautical mile offset -- one to the north and one to the south of Howland). The mileage spread between the minimum and maximum distance for the eight scenarios was 71 miles. Because they were searching for Howland at the old coordinates, I used a great circle distance of about 2,550 (statute) miles. I took the midpoint of the 71 mile variance and added it to the 2,550 figure. Rounding this to a whole number gives 2,585 miles. If Amelia was in close proximity to Howland at 19:12 GMT, you are correct and her average ground speed would have been around 135 mph. (This is a long winded way of saying you have changed my mind.)
Regarding your item 3, last transmission and possible location of the aircraft at that time, I wish to do some more thinking. Your comments stimulated me to review my conclusion. More later.
Earhart's initial position report, heard at 15:18 Lae time (05:18 GMT) as recorded in the Chater letter, is a problem. Taken at face value, it doesn't make any sense so everyone has to change it in some way. If you accept the postion reported -- 150.7 East, 7.3 south (only about 186 nm from Lae) -- you have to not only assume that they made a huge dogleg in their Great Circle route for some unexplained reason, but you also have to assume that the position reported is several hours old and actually predates the report heard an hour earlier at 14:18 (04:18 GMT). The alternative is to accept that the aircraft made an average groundspeed of under 50 knots.
I find it much easier to accept that the coordinates were either heard wrong or transcribed incorrectly somewhere along the way and that the airplane was, in fact, proceeding along the Great Circle at a "normal" pace.
Like the great "circling" debate, there's no way to know for sure what actually happened and I don't have any trouble with your assumption that 2, 585 is a good compromise to use as the distance traveled up to the point that they started searching for Howland and it sounds like we're in agreement that that point in the flight came at sometime just prior to 19:12 GMT. We're therefore in agreement that the plane's average groundspeed during the enroute portion of the flight seems to have been in the neighborhood of 135 mph. That's progress.
How, I wonder, can we get a handle on what kind of "stripping" efficiency was possible in Earhart's fuel system? I certainly agree that 100 percent is unobtainable but I can't help but feel that your 4 percent is a bit stiff. She had 12 tanks in the airplane. If a gallon of fuel in each tank was unusable that's 12 gallons or about 1 percent of the 1,100 gallon load. Is that within reason for a fuel system with a stripping provision? Or is that hopelessly optimistic? Somebody knows this stuff.
Your approach to assessing fuel consumption during Amelia's flight to Howland is perfectly logical.
As to methodology, I suggest that there are three components to the analysis:
1) Basic to the process is an approximation of Amelia's flight profile. This can be constructed using the radio transmissions and the known starting time of 10:00 local. It has to be rough at first until elemental times calculated from steps two and three below can be added to adjust the first approximation.
2) It is necessary to calculate aircraft performance. (Any standard aeronautical engineering text will describe the necessary steps and equations.) Of particular interest is time-to-climb and descend. I would suggest a starting gross weight at Lae of between 15,000 and 15,400 pounds.
You will have to calculate a zero lift drag coefficient. Use the method presented in Quest for Power (NASA). Assume a range of lift coefficients and calculate the corresponding drag coefficients. From these data you can construct a drag polar diagram. Consider doing this for various flight conditions. You will need this in subsequent aircraft performance calculations.
3) It is necessary to calculate fuel consumption for various power settings at various altitudes from sea level to 12,000 feet. The first calculation to make is with the S3H1 operating at 5,000 feet, full throttle. You can get a handle on a couple of variables because you are dealing with a known condition and known results. Kelly Johnson's data gives you four additional known data points or results.
You start by determining RAM air pressure (suggest 132 mph to start) and the stagnation temperature at the carburetor face; account for a pressure drop across the carburetor, mixture cooling due to fuel evaporation; compression heat addition to the mixture due to supercharging; supercharger recovery pressure; and, heat addition to the mixture in the manifold pipes to the cylinders. Manifold pressures, engine speeds and mixture ratios are known quantities. Iterate assumed values of supercharger pressure recovery and heat added to the mixture in the pipes until the correct answer is obtained for the known conditions mentioned above. Plot engine rpm vs supercharger pressure recovery and vs temperature additions in the pipes for the known conditions. Curve fit the data points. This allows you to select values for these two parameters for other power settings for which you do not have known data.
Refine the end results by using a local atmospheric temperature at sea level and applying a temperature lapse rate to obtain upper air temperatures. Substitute these values in your fuel consumption calculation spread sheet to get a bit closer to what happened on July 2 and 3, 1937.
One of the solutions needed is the rate and total amount of fuel expended during climb operations. Although Kelly specifies a power setting for climb, he does not indicate an estimate fuel consumption rate. Nor does he provide fuel burn rates for engine warm up and takeoff. These have to be calculated once the basic program is set up. Thank goodness for the computer!
What a relief! I was afraid this was going to be complicated.
Give me a little while to chew on this and sound out some of the big words.
Just a note. Now my experience as a pilot is limited, but it is current. Looking at the ACMs for several aircraft here there is a definite trend. Allowance each tank HAS an unuseable fuel quantity. It IS specified. It is NOT taken into account as a part of the fuel load either for flight planning or for Reserves.
I agree with Birch that an allowance would have to be made for some volumetric expansion, and would like to know how it is worked out.
Our visible fuel losses to evaporation here (at 80deg + shade temps and who knows what in the sun, with the aircraft standing in the sun all day) seem negligible (looks like somewhere in the order of 1% per 4 days in the sun if that - in reality you can't see the difference.)
I'd suggest, that very close to 1100gals "by volume" actually did start the trip.
My question? does the weight of the fuel change with temperature, or is the weight constant through the change of volume. (yes I did science -- I just want to see someone else's answer.)
Also, I am prepared to do some actual "tank dipping" on an aircraft sitting in the same tropical conditions as in my parachute postings. It is not being used very often at present and sits a week or more at a time in the tropical sun with vented tanks. I'm sure I can get some pretty accurate evaporation figures. Bearing in mind that the greater the volume of fuel, and the fuller the tanks, the lower the evaporation rate.
Also at a 10,000ft cruising altitude, evaporation would be considerably lower than ground level.
The density of the fuel will change with temperature and so warm fuel will take up more volume than cold fuel.
Because the airplane only sat for one day between the time it was fueled and the time it departed Lae, I can't imagine that evaporation played any significant role.
It has been alleged that, because the airplane was fueled in the morning and sat around all day, expansion would cause overflow through the vents which could amount to a loss of as much as 8 gallons. But to suppose that such a loss, if it occurred, was not replaced prior to take off would seem to imply a level of ignorance, inexperience, or incompetence on the part of the crew that is unjustfied.
In short, I agree with you. I think the airplane went out of Lae with 1,100 U.S. gallons of fuel aboard.
How long Amelia and Noonan could possible lived on Gardner Island? I think was more than a year, could be several years, anyone thinks that's posible? Sorry my English, I am just learning.
Of course, no one can say for sure, but especially if the cache of provisions left by the Norwich City survivors was available and discovered, it seems possible that they may have survived for a long time. If they survived for more than three months, then they were there and alive when a British evaluation party visited the island for three days in October 1937. Is it possible that two people could be on the island and not know that a ship had arrived and that over twenty people were ashore? Absolutely.
I have taken another look at my figures, this time from the perspective that at 19:12 GMT, Amelia was in proximity to Howland Island. (It seems pretty clear from her radio message that she thought she should be very near Howland.) Using my methodology, I calculate that Amelia should have had around 740 pounds of (useable) fuel remaining. Similarly, she was burning fuel at the rate of 247 lb/hr assuming she remained at 1,000 feet.
At an average ground speed of 135 mph, she could have flown a distance of 405 miles or 3 additional hours from that point onward. This says that if she was within 405 miles of Gardner, she could indeed have reached the island. (I am using statute miles unless otherwise noted.) On this basis, I retract my earlier conclusion that it would have been impossible for Amelia to have landed on Gardner. I am glad you made me go back and take a second look. Whether it is probable that Amelia reached Gardner is another question. It seems, however, that it is feasible.
Another perspective is also intriguing. If her average ground speed was approximately 135 mph, it could mean she overshot Howland (135 mi/hr x 19.2 hr = 2,592 miles).
And finally, I come back to the 19:12 GMT message which also indicated "but gas is running low." My methodology may understate her fuel consumption. Certainly there is a +/- tolerance associated with my calculations as with any engineering estimate. I can think of more than one reason fuel consumption may have been greater. Problems with one or both Cambridge Fuel Analyzers or inadequate carburetor heat would be significant reasons. Running leaner mixture ratios would have shown up in the cylinder head temperatures even if the Analyzers were biased or malfunctioning. Even small temperature variations in the charge mixture make a dramatic difference in fuel consumption.
I agree. Many scenarios would result in less than optimum endurance and few would net in better results (like maybe an unexpected encounter with a KC135). Still, historical investigation -- like politics -- is the art of the possible, and with so many independent factors pointing toward something wierd going on at Gardner Island, a realistic assessment of the fuel situation which makes the arrival of NR16020 a possiblity is interesting, to say the least.
We'll go ahead with our own reassessment of the aircraft's endurance. It'll be interesting to see how closely our figures match yours.
>it seems most likely
that the aircraft was south of Howland and perhaps
The possible (and probable) propagation characteristics of the radio frequencies used at that time, tend to add support to your points above.
If the plane was south of Howland (let's assume 50 to 100 miles, just to see how the 'numbers' work out) at 19 hours into the flight, the Itasca could have been hearing Earhart's radio transmissions via ground wave. If, over the next 1.2 hours, Earhart flew 'on the line' south, towards Gardner (figuring an average ground speed of about 130 mph), then this would have placed her 200 to 250 miles from Howland at the time that her last radio transmission was heard. I think 'skip conditions' that time of morning, on 3105 khz., might be just about right for very good signal propagation over a 200 to 300 mile path. This would also have placed Earhart within about 160 to 210 miles of Gardner, which at 130 mph, would have placed her at Gardner in about another 1.23 to 1.6 hours, after her last transmission was heard. (By the way, skip conditions, using frequencies in the 6000 khz range, might not likely work for a path of only 200 to 300 miles, during that time of morning).
Interestingly, considering my 'assumptions' above, and even using the "remaining endurance" time of 1.8 hours that Birch came up with, this would still allow time for Earhart to have reached Gardner, and maybe still have enough fuel left, after landing, to run the starboard engine (I think that was the one?) enough to keep the batteries charged for the radio to work over the next couple of days.
John Rayfield, Jr.
Great postings and exchanges regarding the Electra range!
I have a suggestion - given the uncertainties in the inputs, it might be wise try to define a range of values (rather than a single value) for the flight endurance. Since all the values which make up the analysis are estimates, I think that it is appropriate to estimate the error of these estimates, and report this as well. The end result would be a best estimate +/- an error estimate.
There are at least two (maybe more) ways to do this - either do a sensitivity type of analysis, ie, a reasonable "optimistic" and "pessimistic" analysis in addition to the "best estimate", or use a Monte Carlo analysis with appropriate error distributions on the input variables to generate a cumulative probability curve. In either case, a lot can be gleaned from even the simpler of the two analyses. For example, there is a big difference in an estimated range of xxxx +/- 10NM and xxxx +/- 200NM. Even better is the shape of the cumulative probability curve from a Monte Carlo analysis. Either process can also identify the sensitive inputs -- those which deserve the most care and effort in estimation.
I agree that it's important that, whatever numbers we end up using, we make it clear how confident we are that those numbers are correct. For example, the distance between geographical features probably hasn't changed much since 1937 (plate tectonics aside), so those will be pretty hard numbers. Other numbers will be a bit softer and some will be downright mushy. So -- yes -- that all needs to be stated up front. Let's also keep in mind that the point of this exercise is not to determine what happened (or even probably happened) to Amelia Earhart. All we're trying to do is establish is whether the aircraft "should have" had enough endurance to reach Gardner Island.
I don't know enough about Monte Carlo simulations to know whether that's a good way to approach this question.
> I certainly agree
that 100 percent is
Ric, that raises a question in my mind. If a tank is said to hold 100 gallons could that mean 100 USABLE gallons? Why would we say the plane could hold X number of gallons unless they were all usable? How would anyone plan a flight accurately if the number of USABLE gallons was not known? I would certainly agree it is reasonable to think that not all the fuel could be used from a tank and that some residual fuel would be left at "dry tanks" but I certainly don't know that for a fact. Couldn't someone at Lockheed answer that?
Well, I'm sure somebody can answer that but remember that the answer has to be in a 1937 context and nobody now working at Lockheed/Martin lives in that world.
From Dick Pingrey
Ric, I have read Birch Matthew's and your posting about the unusable fuel and there are several things that bother me. First, why is unusable fuel a percentage of the fuel loaded? If you start with tanks half full or completely full the usable fuel (that fuel which can not be transported to the engines due to the fuel system design) is the same in either case. Second, 44 gal. seems like a lot of unusable fuel for any relatively light two engine airplane. That is 90% of a 50 gal. drum. I would think 10 gal or possibly 20 gal would be much closer. My two tank Cessna has only 4 gal that are unusable, This needs to be checked and compared to similar aircraft or any L-10s that may be available. I would have thought the unusable fuel figure would be in the manufacturer's literature.
Dick Pingrey 908C
Unfortunately, I can't find any mention of usable/unusable fuel in the Lockheed literature we have, and besides, that would be for the standard airplane rather than the "special" with the elaborate long-range tank system.
Well, does it REALLY matter?
What EXACTLY is the 1100gal for?
Is it the TOTAL (air) volume on the designer blueprints? I.E. like the total quantity of gas needed for the INITIAL fill of the DRY brand new tanks out of the factory?
Or is it the actual volume needed to RE fill the tanks after subsequent uses? Aren't aircraft actually tested for volume, until they "run dry", until they suck fumes? Wouldn't that be where the 1100 figure comes from?
Even if this was not actually tested formally at the factory, AE/FN "should" have run their tanks dry, a few on each of their numerous previous flights. And on careful refilling the next time they would have learned the ACTUAL PRACTICAL volume of ea tank...
Or is it impossible to run an aircraft tank dry, even if one has another half dozen full tanks left?
I have 2 tanks on my 1/2 Ton, often run them dry (well, one at a time), and I know they offer exactly 75 and 78 useable liters ea.
Good questions Christian. Yes, you can run a tank "dry", then switch to another tank. If a 100 gallon tank has 95 usable gallons, then when it's "empty" you can only put 95 gallons in it. Why call it a 100 gallon tank?
> Before we start:
Sounds like a good scientific approach and on fuel consumption do we have any better information than from Kelly Johnson? If not I'm not willing to just make up unsupportable guesses for it.
I don't know of anything that beats Johnson. He conducted his tests in NR16020 and Kelly was the best in the business. I don't know how it could get much better than that.
The other thing that makes long-term survival possible is that the island is quite rich in fish, land crabs, birds, and turtles, and has a fair supply of shellfish, all edible and not too difficult to procure. On the other hand, water is very scarce, and cuts become infected rapidly, so both dehydration and death from an infected cut (easy to get on the coral, sharp sticks, broken up airplanes,etc.) are dangers.
It would certainly be possible for people at one end of the island not to know that a bunch of other people had arrived on the opposite end, particularly if one were sick, injured, or otherwise not very mobile. On the other hand, it's a bit harder to imagine live people being missed by the British/Gilbertese team that walked clear around the island only a few months after Earhart's disappearance. Not impossible, but it seems unlikely.
The walk around the island was made by Eric Bevington and a handful of Gilbertese. Eric's diary, and his oral account of the journey as related to me in 1991, make it quite clear that what started out as an exploratory hike turned into an "adventure", the sole purpose of which was to make it back to camp alive. The island was far larger than they thought it was and, incredibly, they had brought along no drinking water.
From Ross Devitt
Ric has obviously spent a little time on a deserted tropical island and found out it is not necessarily "Bali Ha'i". You could be at one end of the island the size of Niku and have a group of people at the other and if you were weak or holed up somewhere you'd never know.
The problem with tropical islands is that they only get rain for a short period each year and unless there is a well or fresh water catchment you have to be fit enough to (and know how to) climb coconut trees for water (milk) the rest of the year.
That is assuming your island supports coconut trees. There are other ways of obtaining fresh water, but.....?
In 1937 there were, reportedly, 111 coconut tree in bearing on the island. The year 1938 was a record drought year in the Phoenix Group.
On the question of possible fuel venting.
If 1100 gallons were put in, and the fuel subsequently warmed, expanded, and spilled through the vent system, then there would not be eight gallons less, or any gallons less. There would actually be more gallons of fuel than had been pumped in (but fewer pounds). If there wasn't more volume then it could not spill out.
The point of this is that range is not based on how many gallons of fuel were in the aircraft, but how many pounds. The original data are unfortunately based on gallons of fuel rather than pounds. To determine the actual fuel load some estimate of the fuel temperature will have to be made.
Yes, and I suggest that we use 80 degrees F as that temperature, as did Birch Matthews.
The aircraft was fueled on the morning of July 1st following the test flight. Presumably that fuel, stored in steel drums, was cool from the night before. During the day the airplane can be expected to have heated up and the fuel expanded, with some unknown amount (poundage) being lost to venting. That night, the airplane cools down again, as does any gas remaining in the drums. On the morning of July 2nd anyone sticking or visually inspecting the tanks should notice that they're down some. In accordance with our assumption that Earhart and Noonan are neither stupid nor negligent, we must also assume that the tanks are topped as part of the preflight preparations. 80 degrees F seems like a conservative temperature for the fuel at the time the flight departs at 10 a.m. local time.
Alan Caldwell asks if we have any better source on the subject of fuel consumption than Kelly Johnson. That's not the question to ask, for Johnson was an undoubted expert. It's what Kelly didn't leave us in the way of information that is important.
Consider or remember that Kelly's data was based upon a standard atmosphere. Amelia did not fly in a standard atmosphere and therefore, these differences have to be taken into account. Fuel consumption will therefore vary slightly over Kelly's findings. Remember also that Amelia flew at one or two different altitudes than planned for by Kelly. Thus you have to view Kelly's recommendations as a general guide as to the flight profile and power settings she should employ. Weather could easily force her to vary from the plan that Kelly envisioned before her first world flight attempt.
In addition, Kelly gave Amelia a climb power setting, but that doesn't tell us what the engine consumed during any and all climbing flights. A small amount of fuel was also consumed in engine start and warm up and taxiing. This is why it is necessary to assess the last flight in elemental detail.
Others puzzle over my terminology and/or use of "useable fuel fraction." If this is not a term used in the aviation world, I apologize. Some background. Interest in just how much of the fuel loaded onboard an airplane could reach the engine(s) certainly came to the foreground in the U.S. during the early part of World War II. With the destruction of the Army Air Corp fighter aircraft at Pearl Harbor on December 7, 1941, aircraft resupply became paramount. The quickest way to replenish fighters was to fly them across the Pacific. This prompted Bell and Curtiss to study the problem of extra long range operations. Bell made the test I referred to earlier. Curtiss did also, although I don't have their data. Other manufacturers probably did similar work.
The same consideration was made with regard to getting fighter aircraft to England as fast as possible. I don't think (but admittedly don't know) that people studied the useable fuel concept prior to World War II.
As to tank volumetric capacity, this is often calculated by the tank designer based upon his drawings. I have been through every Lockheed technical report existing in the Lockheed archives that was generated for the Model 10 and Model 12. None addressed verification of tank volume. These tanks, including the special ones made for Earhart, were shaped to conform to the envelope in which they were installed. They may or may not have incorporated an ullage volume. I don't know; however, it not an uncommon practice.
One cannot simply attribute trapped residual fuel to the tanks. There were a lot of lines, fittings, valves and controls associated with the Earhart fuel system. I assure you that World War II manufacturers took into account trapped fuel and oil in their systems. You will find this information on many detail weight statements.
Trapped fuel on the P-39 amounted to 3.1 percent. Somewhere I have a similar figure for the P-51 (can't lay my hands on it at the moment). Memory tells me it was just slightly greater than the P-39.
Hope this helps answer some questions as well as explain some of my thought processes and considerations.
This is from a flight in 1928. Probably before Amelia thought of her attempt.----From First Across the Pacific.
Kingsford-Smith and Ulm believed their critics wrongly identified their Pacific effort with other attempts that they saw as ill-advised, ill-prepared and therefore ill-fated. For instance, they put the Fokker through a number of testflights with increasing overloads and had the Douglas company strenghten the parts most afflicted. Also, they severely tested their ability at blind flying and radio navigation, both still quite a novelty but essential for their success. To prepare themselves for the long flights, they frequently went out driving for fifteen hours, then flying for four hours, running for two, back to flying and so on for a total of some forty hours. Apart from their fixed radio transmitters, they had a watertight, battery-powered emergency radio, the aerial of which could be lifted with a kite or some gas balloons they carried.
In short, their preparations covered every eventuality. And so with complete confidence in their aircraft, its engines and their own ability Kingsford-Smith, Ulm, Lyon and Warner awaited the right moment. It came on May 31, 1928, with favourable weather predictions for the route to Hawaii and the advantage of a full moon during the night. The take-off from Oakland, close to 9 AM, went well and Kingsford-Smith describes how, flying over the Golden Gate which lay glistening in the morning haze, he felt relieved that after all their preparations they were finally under way.
The flight itself was quite uneventful. In fact, to the two pilots who could hardly move and were sitting in the horrendous roar of the engines, it was probably boring more than anything else. Kingsford-Smith, a hopeless addict, continually longed for a cigarette. But the sunset was spectacular and for the first hours of night, the moon lighted their path, and stars filled the sky. At ten, Lyon threw out a flare, to check for drift. At close to midnight, to avoid some rainclouds they climbed to 4,800 feet, but otherwise the weather was as perfect as predicted.
Shortly before two, they noticed some lights far below. Kingsford-Smith used his spotlight, fitted with a Morse key, to make contact with the ship and let them know the "Southern Cross" was doing fine. Within an hour, they passed another ship, from which Warner received a radio bearing. They were the only two ships they encountered on their entire Pacific crossing.
When dawn finally came, Kingsford-Smith decided to descend below the clouds, as it had become rather chilly in the cockpit. At 8 AM, Lyon estimated Hawaii was just over 200 nautical miles away. A number of times, the land Kingsford-Smith or Ulm spotted turned out to be yet another cloud bank, until finally, close to 11 AM, Mauna Kea came into view. At 12:17 PM, the "Southern Cross" touched down at Honolulu's Wheeler Field, after a flight of more than 1,700 nautical miles.
The full text is at: First Across the Pacific.
It is not a direct quote, but I have read the published journals of Ulm and Smithy and it is pretty close especially the descriptions of looking for Hawaii.
Notice that looking for a tiny island (Hawaii) from a 200 mile estimate they faced at least as difficult a time as Earhart and Noonan would have faced looking for another tiny island (Howland).
Interesting given the similar radio equipment, and that Kingsford Smith's worked -- and that he knew morse code...
Compared to Howland or Niku, the Hawaiian Islands are like continents.
What about actual figures from previous legs of the flight? There have to be some records of consumption / fuel purchased during the flight. Didn't she fuel up at the military airfield after her flight from Calif to Hawaii? The military surely would have that recorded somewhere. Probably fueled up at Pan Am in San Juan. Seems like the fuel bills would have been forwarded to some central accounting point who was footing the bill for the entire venture. Putnam? Duke Univ? someplace.
LTM (Who likes to
keep track of her mileage)
There's actually quite a bit of fuel information available regarding the Oakland/Hawaii flight and I think we discussed that at some length on the forum a couple of months ago. There is at least one fuel receipt (Darwin) from the second world flight attempt and other bits and pieces of information, but nothing as technical or comprehensive as Johnson's telegrams.
Folks, much of what I will say here is based upon my modeling simulations (shared with TIGHAR) and recent material uncovered for the 8th edition. What is beginning to fall into place is that AE and FN might well be on the LOP passing through Howland, but are very far away, well to the SSE of either Howland or Baker, possibly 80 miles away! My monte carlo simulations of the flight indicate a southerly excursion due to the changing wind direction (from the time of 1030 GMT onwards) from that predicted, so that a slight northerly cross-wind component drove them south of the intended flight path. By itself, its kinda interesting, but can also be misleading.
Now, Bob Brandenberg has done a marvelous job of radio signal strength analysis, which indicates AE was about 80 miles away at her closest approach to Howland (at one of her radio transmissions). If all of this is true, this makes a lot of sense: why AE and FN didn't find Howland or Baker or see the smoke from the Itasca, it also fits the lack of radio reception after her last message (both frequencies were basically out of range during those times), and it helps AE get to Gardner quicker than expected, as she seems to be well south of Howland, helping out the fuel situation. I'm beginning to become convinced that this might be the right scenario. I wonder how Noonan could be so far off in a N/S direction: the only explanation is that he wasn't able to take a celestial fix at all during the latter half of the flight. That seems sooo strange, but stranger things have happened.
The 8th edition is just about done by the co-authors, isn't that right, Ric? and will definitely be worth the price of admission. We're sitting on a veritable gold mine of information, so when will it be ready for publication? I cannot wait!
We're probably looking at March for the 8th edition to be ready to ship.
I'm not aware of any Air Force investigations in the 60s. But in Dec 1960, ONI (Office of Naval Intelligence) conducted an investigation concerning Thomas Devine's claim that he knew where AE's grave was located along with numerous photos. Agent Patton's investigation disclosed "no evidence that AE landed at Saipan". Patton, however, cites two Saipanese sources claiming they overheard the Japanese military in 1937 "talking about the crash of AE's plane at Jaluit Atoll".
You probably have his full report dated 23 Dec 60: ONI-2345-7(b). I know the current mission of TIGHAR is to prove (or disprove) with credible evidence, either physical or by witness testimony,the hypothesis that AE flew on to Niku after failing to locate Howland and crashed on or near the Island. I must agree with forum member Dennis McGee. No matter what the millions of hours it will take to estimate within whatever reasonable parameters the Electra's range, if the artifacts and bones can be conclusively linked to AE she made it!
If none of the TIGHAR'S investigation yields any viable, concrete evidence,then move on to other theories.
One for instance: re your evidence:
What is the final expert evaluation by the Cat's Paw people of the Niku heel? Did AE normally fly with these shoes? Were similiar shoes found in her effects after the crash? Your photo says it was taken ten days before the crash. Did she have more than one pair? Other reserchers say she wore a size 7? What inquiries were made with living relatives? Any effort to locate shoe repair shops ? Any documentation from checkbooks, relatives, friends that she had a replacement heel added? Who took the photo? Many potential inquires left. This evidence may be your best single and irrefutable link to Niku and AE.
This is probably old stuff to you but as you warned, it gets addicting!!
No, I don't think we have a copy of the ONI investigation. We really haven't spent a lot of energy collecting background material on the Saipan theory. Never seemed worth the effort.
I'll try to answer your shoe questions.
>What is the final
expert evaluation by the Cat's Paw people of the
Remember, there are two heels, only one of which is a Cat's Paw replacement heel. Not much information can be had from the second heel because it has no markings, but the Cat's Paw has been tracked to a mold that was in use in the mid-1930s. The nail holes fit the remains of a hard rubber sole (fragmented) bearing stitching holes that indicate that it was a woman's blucher oxford. A small brass shoelace eyelet seems to confrim that the shoe was a woman's.
>Did AE normally fly with these shoes?
There are many photos showing AE wearing such shoes on the world flight.
>Were similiar shoes found in her effects after the crash?
I assume you mean the Luke Field crash. No. She took her personal luggage with her and it was not inventoried by the Air Corps.
>Your photo says
it was taken ten days before the crash. Did she have
Photos taken during the world flight also show a pair of two tone saddle shoes with light colored soles.
>Other reserchers say she wore a size 7?
So did I when I was a kid. Our assessment of Earhart's shoe size is based upon photogrammetric measurement.
>Any effort to locate shoe repair shops ?
How many shoe repair shops have been around for 62 years? Is it realistic to think that we're going to find some old guy who remembers putting new heels on AE's shoes?
from checkbooks, relatives, friends that she had a
No. I'm not aware of any such records that could be checked.
>Who took the photo?
That particular photo is in the Purdue collection. The photographer is not credited, but there are many other photos taken before and during the world flight that show AE wearing shoes that look just like those.
Randy's modeling work is very interesting. Sounds like you have a proponent of Monte Carlo methods right there in the TIGHAR's den.
As Randy notes, you have to pretty much accept the fact that their celestial nav broke down to some degree for this to be possible. If that were the case, it is possible that Noonan might have shifted their targeted arrival point away from Howland to something roughly halfway between Howland and Baker. Given 15-20 mile visibility (which might have seemed reasonable), that would create a backstop of some 70-80 miles (approx) in length across their path where either of the islands should have been visible. In other words, if Noonan felt the nav accuracy was deteriorating (but visibility near sea level was good), he might have opted to increase his chances of making landfall by going slightly off course to widen his target area.
This would place them on a course roughly 18 miles south of Howland. If they were flying by DR, the crosswinds that Randy postulates might already have had a head start on drifting them southward. Just a thought.
On another subject, my understanding is that Noonan had 2 chronometers, a bubble sextant and a marine sextant as backup on board. I have several marine sextants, but know very little about aircraft types.
1. Anyone familiar with brand and the performance characteristics of Noonan's bubble sextant? How accuracte were they (generally)? How high an altitude could be reliably measured? Did it have a chronometric averager or could you just snap a quick sight? Even if it had a chronometric averager, could you override it if you had limited sight time? Were they reliable (not prone to missalignment or other problem during normal handling that would render them useless)? Would it be unusual to carry just one?
2. Has anyone out there ever actually tried using a marine sextant from an aircraft? What was the result?
Happy New Year,
The Monte Carlo simulation that Randy was involved in was done by a company called Wagner Associates in California as part of their effort to sell the Navy on their program.
I can't answer your questions about Noonan's bubble octant. We don't know for sure what type of instrument he carried. We do know that just prior to the first World Flight attempt, a Pioneer bubble octant was borrowed from the Navy by Harry Manning (at Noonan's insistence) and that the instrument was signed over to Noonan when Harry left the team following the Luke Field debacle. It seems likely that Noonan used that same instrument on the second attempt but we don't know that for sure.
As for the number of chronometers, it would seem prudent to carry a spare but Chater's references to Noonan's "chronometer" are all singular.
|Back to Highlights Archive list.|