Advanced search  
Pages: 1 [2] 3 4 ... 35   Go Down

Author Topic: The flight plan, magnetic course, headwinds.  (Read 279897 times)

Gary LaPook

  • T5
  • *****
  • Posts: 1624
Re: The flight plan, magnetic course, headwinds.
« Reply #15 on: December 29, 2011, 03:24:59 PM »

^^^^ Yeah - Pretty much!   Crazier human-error things happened in the aviaition world every once in a while ... Take Tenerife for instance.
Teneriff was much more complex than pouring piss out of a boot or applying deviation and the deviation did not change abruptly after Lae, it was the same 6° East that it had been on the leg into Lae and it gradually changed to 9° East in the vicinity of Howland. And these types of errors are detected and cured every time you get a new fix.

gl
Logged

Erik

  • T3
  • ***
  • Posts: 185
Re: The flight plan, magnetic course, headwinds.
« Reply #16 on: December 29, 2011, 03:49:37 PM »

^^^^ Yeah - Pretty much!   Crazier human-error things happened in the aviaition world every once in a while ... Take Tenerife for instance.
Teneriff was much more complex than pouring piss out of a boot or applying deviation and the deviation did not change abruptly after Lae, it was the same 6° East that it had been on the leg into Lae and it gradually changed to 9° East in the vicinity of Howland. And these types of errors are detected and cured every time you get a new fix.

gl

Taking off without a clearance is not complex.  Neither is watching your altimeter as was the case in the 72' everglades crash.  Landing gear-up happens every day - its not complex either.  Turning on your engine temp de-ice is pretty easy too - which they didn't do in Air Florida 14th bridge during a snowstorm. 

Something bone-headed happened!

Logged

Gary LaPook

  • T5
  • *****
  • Posts: 1624
Re: The flight plan, magnetic course, headwinds.
« Reply #17 on: December 29, 2011, 04:07:22 PM »

^^^^ Yeah - Pretty much!   Crazier human-error things happened in the aviaition world every once in a while ... Take Tenerife for instance.
Teneriff was much more complex than pouring piss out of a boot or applying deviation and the deviation did not change abruptly after Lae, it was the same 6° East that it had been on the leg into Lae and it gradually changed to 9° East in the vicinity of Howland. And these types of errors are detected and cured every time you get a new fix.

gl

Taking off without a clearance is not complex.  Neither is watching your altimeter as was the case in the 72' everglades crash.  Landing gear-up happens every day - its not complex either.  Turning on your engine temp de-ice is pretty easy too - which they didn't do in Air Florida 14th bridge during a snowstorm. 

Something bone-headed happened!
All the examples you gave lasted only moments, not the hour after hour flight of the Earhart plane where any such error would have had many many opportunities to be recognized. 

Teneriff: I don't know how much flying you have done internationally, but even though English is the world-wide language of aviation, controllers who are non-native speakers of it don't it do so well and there have been a number of aviation accidents caused by this and by improper phraseology. The Spanish controller used improper phraseology and instructed KLM to "hold for takeoff." According to ICAO  regulations, the word "takeoff" is only to be used in the phrase "cleared for takeoff." The same is true of Netherland and U.S. and all other countries. The other 747 transmitted at the same time covering up the "hold for" part of the controller's transmission. The KLM captain heard "...takeoff" which in all of his 30,000 hours of flying and at least as many takeoffs had meant "cleared for taekoff." (When you call the tower and say you are ready for takeoff, the controller is only allowed to say one of three things to you: "HOLD SHORT OF THE RUNWAY" or "LINEUP AND WAIT" or "CLEARED FOR TAKEOFF." Notice, there are no words in common used in any of the three phrases.)

Everglades: L1011 on approach to Miami at night from the west over the everglades where it is as pitch black below you as flying over the ocean. They couldn't get a gear down light to come on and in the process of trouble shooting the problem the captain bumped the control yoke  which the auto-pilot interpreted as the captain disconnecting the auto-pilot which is what it is supposed to do when the yoke is moved. The plane descended very gradually into the glades while the cockpit crew were distracted trouble shooting.

None of these were as simple as applying deviation to a true course. Which, by the way, Noonan and Earhart didn't even need to do since Williams had laid out the courses in magnetic degrees, the magnetic courses, having already applied the deviation.

gl
« Last Edit: December 29, 2011, 05:39:48 PM by Gary LaPook »
Logged

Erik

  • T3
  • ***
  • Posts: 185
Re: The flight plan, magnetic course, headwinds.
« Reply #18 on: December 29, 2011, 05:20:11 PM »

^^^^ Yeah - Pretty much!   Crazier human-error things happened in the aviaition world every once in a while ... Take Tenerife for instance.
Teneriff was much more complex than pouring piss out of a boot or applying deviation and the deviation did not change abruptly after Lae, it was the same 6° East that it had been on the leg into Lae and it gradually changed to 9° East in the vicinity of Howland. And these types of errors are detected and cured every time you get a new fix.

gl

Taking off without a clearance is not complex.  Neither is watching your altimeter as was the case in the 72' everglades crash.  Landing gear-up happens every day - its not complex either.  Turning on your engine temp de-ice is pretty easy too - which they didn't do in Air Florida 14th bridge during a snowstorm. 

Something bone-headed happened!
All the examples you gave lasted only moments, not the hour after hour flight of the Earhart plane where any such error would have had many many opportunities to be recognized. 

Teneriff: I don't know how much flying you have done internationally, but even though English is the world-wide language of aviation, controllers who are non-native speakers of it don't it do so well and there have been a number of aviation accidents caused by this and by improper phraseology. The Spanish controller used improper phraseology and instructed KLM to "hold for takeoff." According to ICAO  regulations, the word "takeoff" is only to be used in the phrase "cleared for takeoff." The same is true of Netherland and U.S. and all other countries. The other 747 transmitted at the same time covering up the "hold for" part of the controller's transmission. The KLM captain heard "...takeoff" which in all of his 30,000 hours of flying and at least as many takeoffs had meant "cleared for taekoff."

Everglades: L1011 on approach to Miami at night from the west over the everglades where it is as pitch black below you as flying over the ocean. They couldn't get a gear down light to come on and in the process of trouble shooting the problem the captain bumped the control yoke  which the auto-pilot interpreted as the captain disconnecting the auto-pilot which is what it is supposed to do when the yoke is moved. The plane descended very gradually into the glades while the cockpit crew were distracted trouble shooting.

None of these were as simple as applying deviation to a true course. Which, by the way, Noonan and Earhart didn't even need to do since Williams had laid out the courses in magnetic degrees, the magnetic courses, having already applied the deviation.

gl

The Spanish controller used improper phraseology and instructed KLM to "hold for takeoff." = Human Error. 

Without picking through all the details, these examples were only designed to act as 'mascots' for human error.  While I admit that the magnetic variation/deviation issue is slim-pickings when it comes to likelihoods of human error for the fate of the electra, it does establish that human error is in fact possible.  Even on very basic tasks such as extending/retracting landing gear, forgetting to turn on de-ice heat, radio calls or whatever.  Mistakes happen. 

Clearly AE/FN made some kinda of human error?  Could one of the contributing factors have been something as simple as a very basic magnetic heading?  Probable - No, Possible - Maybe.

Don't get me wrong.  I am not saying AE/FN didn't have the basic understanding of how to calculate the magnetic headings (pissing in boots or whatever),  it was their bread-and-butter, but rather a simple mistake was made.  Such as subtracting 9 instead of adding 9.  We've all seen stranger things that have happened.
Logged

Gary LaPook

  • T5
  • *****
  • Posts: 1624
Re: The flight plan, magnetic course, headwinds.
« Reply #19 on: December 29, 2011, 05:41:02 PM »




"None of these were as simple as applying deviation to a true course. Which, by the way, Noonan and Earhart didn't even need to do since Williams had laid out the courses in magnetic degrees, the magnetic courses, having already applied the deviation. See attached chart. [ It caused a problem, apparently it was too big of a file.] It was William's strip chart of the Lae to Howland leg. I think I posted it before and I think it caused the same problem then.


[attachment deleted by admin]
I was right, I had uploaded the Williams strip chart before, here is a link to it.



gl
« Last Edit: December 30, 2011, 12:46:02 AM by Gary LaPook »
Logged

JNev

  • T5
  • *****
  • Posts: 778
  • It's a GOOD thing to be in the cornfield...
Re: The flight plan, magnetic course, headwinds.
« Reply #20 on: December 29, 2011, 09:30:24 PM »

Interesting discussion.

Heath did a lot of good math and analysis in that long work-up - fascinating.  Eric's points about errors pique well too, but I tend to think Gary's onto something about the timing, etc.  Too much known ground was covered early enough in the flight for AE to have gone that brain dead for hours on end up front (although I agree, crazier things have happened). 

Not that I fully agree that the other examples of error are so limited to momentary lapses: the everglades crash is a tragic example of many 'moments' for the crew to have cued-in - but they were all so focused on chasing the gear light that none cued-in during the drift-down.  Nobody was 'flying the plane' - tragic, and a lapse of far too many minutes - it can and does happen to the best.  FN and AE would not have been immune - I can agree to that.

In re-reading Friedell's report (Colorado) I was struck by his contemporaneous observation of 'what FN might do' as reported to him by those who knew FN and his habits well and who understood the circumstance (and I will now apologize in advance for the length of this posting):

"The Commander Coast Guard sent word that he had communicated with persons familiar with the methods of navigation of Mr. Noonan, and that Mr. Noonan would take a fix shortly before dawn, correct course for destination, and determine line of position when near the end of estimated run. This procedure would allow a flight of about 3000 miles without a good fix. If short of gas, he probably would follow the line of position to the nearest land. The line of position 337°-157° was given in one of the last reports received from the plane. It was also stated in a report that the plane was short of gas.

Considering the question as to what Mr. Noonan did do, it must be considered which way he would steer on the line. To the northwest of Howland was wide stretches of ocean, to the southeast were spots of land. To a seaman in low visibility the thing to do when in doubt of own position would be to head for the open sea. The land would be the place to get away from. To the Air Navigator with position in doubt and flying a land plane it is apparent that the thing to do would be to steer down the line towards the most probable land. To the Air Navigator, land would be a rescue, just as the sea would be to the seaman. Would and did Mr. Noonan do this or had he other reasons to do otherwise? The answer was of course unknown but logical deduction pointed to the southeast quadrant."


This, like the simplicity Gary's point on the magnetic headings provided by Williams, may take us to the crux of understanding of what may have been in the context of what AE and FN actually had in-hand.

Similarly, as to corrections for ground speed, etc. - it is doubtful to me that much correction for that would have occurred in terms of power settings or chasing optimum speeds.  If one reads the three Kelly Johnson / LAC telegrams carefully, they provide the basis for a beautifully simple set of numbers that can be tabled and used easily for the Lae - Howland flight; in sum the telegrams come to these essential values and segments:

To 8000' initial - climb at 2050 rpm / 28.5" MP = 60.0 gal/hr
Thence 3 hrs each:
8000' / 1900 rpm / 28.0" MP = 60.0 gal/hr
8000' / 1800 rpm / 26.5" MP = 51.0 gal/hr
8000' / 1700 rpm / 25.0" MP = 43.0 gal/hr
Thenceforth:
10,000' / 1600 rpm / 24.0" MP = 38.0 gal/hr


The key of course is what FN and AE actually did do, in each case - hence this interesting string.  But the point is valid - they did have in-hand a reasonable set of guidelines and excellent tools, from the outset.  We can derive some clues about actual happenings from the communications recorded from the flight (but we start to run tantalizingly short at the end, don't we?).

LTM -
- Jeff Neville

Former Member 3074R
 
« Last Edit: December 30, 2011, 08:43:44 PM by Jeff Neville »
Logged

Martin X. Moleski, SJ

  • Administrator
  • *
  • Posts: 2882
Re: The flight plan, magnetic course, headwinds.
« Reply #21 on: December 30, 2011, 12:30:51 AM »

I am having trouble with accessing page 2 of this string to modify my post -

Gary tried to upload another high resolution scan.

I've resized it and am attaching two lower resolutions here (if all goes well).
LTM,

           Marty
           TIGHAR #2359A
 
« Last Edit: December 30, 2011, 12:32:59 AM by Martin X. Moleski, SJ »
Logged

Gary LaPook

  • T5
  • *****
  • Posts: 1624
Re: The flight plan, magnetic course, headwinds.
« Reply #22 on: December 30, 2011, 12:45:11 AM »

I am having trouble with accessing page 2 of this string to modify my post -

Gary tried to upload another high resolution scan.

I've resized it and am attaching two lower resolutions here (if all goes well).
I had uploaded the Williams strip chart before, here is a link to it.

gl
Logged

Erik

  • T3
  • ***
  • Posts: 185
Re: The flight plan, magnetic course, headwinds.
« Reply #23 on: December 30, 2011, 08:37:53 AM »

"None of these were as simple as applying deviation to a true course. Which, by the way, Noonan and Earhart didn't even need to do since Williams had laid out the courses in magnetic degrees, the magnetic courses, having already applied the deviation. See attached chart. [ It caused a problem, apparently it was too big of a file.] It was William's strip chart of the Lae to Howland leg. I think I posted it before and I think it caused the same problem then.

Heath did a lot of good math and analysis in that long work-up - fascinating.  Eric's points about errors pique well too, but I tend to think Gary's onto something about the timing, etc.  Too much known ground was covered early enough in the flight for AE to have gone that brain dead for hours on end up front (although I agree, crazier things have happened). 

Not that I fully agree that the other examples of error are so limited to momentary lapses: the everglades crash is a tragic example of many 'moments' for the crew to have cued-in - but they were all so focused on chasing the gear light that none cued-in during the drift-down.  Nobody was 'flying the plane' - tragic, and a lapse of far too many minutes - it can and does happen to the best.  FN and AE would not have been immune - I can agree to that.

I probably wasn't very clear in the context of the above examples regarding commercial aviation accidents.  Maybe they were too complex, but the point still remains that very simple human errors can result unforseen events.  If we need simpler examples, take fuel starvation while in flight, or landing gear up.  These very basic errors happen nearly every day in general aviation.  As Jeff stated, unfortunately AE (or any other pilot for that matter) is not immune.  Fuel starvation is probably the closest example where the duration of the error lasted longer as opposed to 'momentary' errors.

Incidently, when I suggested a magnetic heading related error, I didn't mean to imply that was the ONLY factor.  AE would have also been vulnerable to more 'complex' contributing causes.  Maybe not as great as those found in commercial accidents, but still factors such as fatigue, reduced cockpit lighting conditions, having only one set of eyes on the compass as opposed to two, etc, would have all played a role.

In hindsight, I see where I may have suggested that the entire course was in error.  That wasn't the intention.  If it were, then yes, I would aggree that there would have been to many cross-checks to allow that to happen.  The point trying to be mad was that only a fractional portion of the flight/course could have been a vicitm of a magentic heading related error.

Take this scenario for example:
They are in flight, then lose celestial observations due to overcast.  No big deal.  They revert to dead reckoning using the magnetic compass and directional gyro.  The DG suffers a bit more than its usual gyroscopic pression (maybe from a vacum leak).  AE corrects for it, but turns the DG needle to one hash-mark (or two) shy of it's actual correct heading.  Now they are are off by 10 degrees and don't know it.  They fly this heading for an unknown period of time.  Not knowing how long they have been flying the wrong heading, they dont know how much to compenstate to get back on course.  Fred notices the discrepency.  They are now closer to their destination and decide not to make matters worse by trying to counter-correct the problem.  But, rather to fly a 'parallel' heading an intercept an offset LOP.  Problem now they dont know exactly how far off course they are when reaching the LOP.  To compound the problem, say for example, the magnetic compass wasn't swung correctly or a 'typo' was on the correction card.  Combine that with unknown wind conditions.  Now things are starting to add up in a more complex way - creating a compounding effect.  They hit the LOP with as good of timing they have but even it is off slightly by several seconds.  Unbeknownst to them, the strip chart has Howland charted erroneously 6 miles too far east. 

Things have gotten real ugly real quick! 
Logged

Heath Smith

  • T4
  • ****
  • Posts: 391
Ship in sight, Ontario or Myrtle Bank?
« Reply #24 on: December 30, 2011, 08:47:45 AM »


According to the following page, AE reported seeing a ship in sight at 10:30 GMT.

http://tighar.org/wiki/Nightfall_to_Ship_In_Sight

This telegram seems to be the beginning of the assumption that it was the Myrtle Bank that AE might have seen at 10:30 GMT.

From: SYDNEY
Action: STATE
Precedence Datel 07/03/37 Referback KP0302XX (19370703101XKPH) Referforw
Classific Toffl 2200 Referback1 Referforw1
Style Referback2 Referforw2
Group 0 Datez 07/03/37 Referback3 Referforw3
Officeno SY Toffz 1200 Referback4 Referforw4
Text: PLAIN, SYDNEY N.S.W. VIA TUTUILA AND N.R. DATED JULY 3, 1937 RECEIVED 2:42 P.M.; SECRETARY OF STATE, WASHINGTON; JULY 3, 10
P.M.; AMALGAMATED WIRELESS STATE INFORMATION RECEIVED THAT REPORT FROM "NAURU" WAS SENT TO BOLINAS RADIO "AT 6.31, 6.43 AND 6.54
PM SYDNEY TIME TODAY ON 48.31 METERS, FAIRLY STRONG SIGNALS, SPEECH NOT INTELLIGIBLE, NO HUM OF PLANE IN BACKGROUND BUT VOICE
SIMILAR THAT EMITTED FROM PLANE IN FLIGHT LAST NIGHT BETWEEN 4.30 AND 9.30 P.M." MESSAGE FROM PLANE WHEN AT LEAST 60 MILES SOUTH
OF NAURU RECEIVED 8.30 P.M., SYDNEY TIME, JULY SECOND SAYING "A SHIP IN SIGHT AHEAD". SINCE IDIENTIFIED AS STEAMER MYRTLE BANK
WHICH ARRIVED NAURU DAYBREAK TODAY. REPORTED NO CONTACT BETWEEN ITASCA AND NAURU RADIO. CONTINUOUS WATCH BEING MAINTAINED BY
NAURU RADIO AND SUVA RADIO.; DOYLE
________________________
Dztzf 193707031200SYDNEY
Source: STATE Copyno: 0 Record No: 2358

---

The exact position of the Myrtle Bank seems to be up for debate, with two possible positions being estimated (Lexington Search Report, State Department telegram). Apparently TIGHAR has estimated a position of the Myrtle Bank at that time, roughly centered at 2°20′S, 167°10′E. The details are given here.

http://tighar.org/Projects/Earhart/Archives/Research/ResearchPapers/Worldflight/finalflight2.html

---

There are several references to the position of the Ontario given as 2°59.02′S, 165°23.20′E, so I will accept that as being fact.

If we take the earlier calculations that I had done where the ground speed achieved from Lae to check point A and then on to check point B (Nukumanu Island) the ground speed achieved was approximately 124.5 mph. Just as an exercise assuming that she maintained this 124.5 mph speed from Nukumanu Island for some time going forward (say until the 10:30 GMT report) the time elapsed would be about 3.2 hours (10:30 - 07:18) after leaving Nukumanu Island. Calculating the distance, assuming 124.5 mph for 3.2 hours, this is about 398 miles from Nukumanu Island, at about the position 3°7'28.73"S 165°19'45.91"E, along the original flight plan. This would place the Ontario only 11.5 miles to the North East, well within spotting distance.

I am not quite sure why anyone would ascertain that the ship spotted would have been the Myrtle Bank since very little was known about her actual position. Irrespective of the possible scenario above, the Ontario would be a much better candidate for the ship being spotted since it was only 8 miles or so off the original planned flight path whereas the Mrytle Bank would be at least 26 miles away from the original flight path if we assuming the best case approximate position of the Myrtle Bank at 2°20′S, 167°10′E. The only reasoning that I can see for concluding the ship was the Myrtle Bank and not the Ontario was based on some prior speed calculations that were assumed to be fact. From what I can see, the actual evidence of the ground speeds achieved is pretty scant at best.

Using the same speed as above, if Dowdeswell on the Mrytle Bank had his time confused, and he really heard the Electra at 11:00 GMT (local time versus New Zealand time miscalculation?), this would have put the Electra within 30 miles of this coarse estimate of where the Mrytle Bank was estimated to be, nearly at the closest possible distance between the two. Since this report became important only after the fact, I can see how this time calculation could have easily been misinterpreted and / or documented incorrectly. Given that radio log errors and interpretations both at Lae and Howland are demonstrable, I do not think that this is much of a stretch.
Logged

Martin X. Moleski, SJ

  • Administrator
  • *
  • Posts: 2882
Re: Ship in sight, Ontario or Myrtle Bank?
« Reply #25 on: December 30, 2011, 10:01:14 AM »


According to the following page, AE reported seeing a ship in sight at 10:30 GMT.


Everyone who is trying to renavigate the flight or to otherwise simulate its course should consult Randy Jacobson's statement of the variables in his article on the Monte Carlo simulation.  He has a great list in the article of all of the decisions that have to be made to project possible flight paths from Lae to the vicinity of Howland.
LTM,

           Marty
           TIGHAR #2359A
 
Logged

Jeff Scott

  • T2
  • **
  • Posts: 93
Re: The flight plan, magnetic course, headwinds.
« Reply #26 on: December 30, 2011, 11:11:07 PM »

The Monte Carlo simulation is mentioned quite often. The articles describing it talk about constraints on the results but lack detail on the simulation itself. Is it probabilistic or deterministic? What are the equations of motion? What physics is it modeling and how were the math models derived? What methods/software were used to build it? How was the simulation verified and/or validated? Without knowing more about the tool itself, it's hard to evaluate its results.

Based on the description in the article, this sounds more like a prediction of position based on radio messages than an actual 6DOF flight simulation resolving the forces and moments acting on the aircraft defining its motion through space. Is that the case?

Another comment is the two figures in the wiki article are too small to read, so I'm not sure what they indicate.  Are the scales in latitude/longitude?
It's not too late to be great.
 
Logged

Gary LaPook

  • T5
  • *****
  • Posts: 1624
Re: The flight plan, magnetic course, headwinds.
« Reply #27 on: December 31, 2011, 03:00:21 AM »

The Monte Carlo simulation is mentioned quite often. The articles describing it talk about constraints on the results but lack detail on the simulation itself. Is it probabilistic or deterministic? What are the equations of motion? What physics is it modeling and how were the math models derived? What methods/software were used to build it? How was the simulation verified and/or validated? Without knowing more about the tool itself, it's hard to evaluate its results.

Based on the description in the article, this sounds more like a prediction of position based on radio messages than an actual 6DOF flight simulation resolving the forces and moments acting on the aircraft defining its motion through space. Is that the case?

Another comment is the two figures in the wiki article are too small to read, so I'm not sure what they indicate.  Are the scales in latitude/longitude?

GIGO :)

They claim "We used a very conservative scenario, one that assumed that Earhart had no means of determining her position, and that the entire flight was made by dead reckoning." This is not a conservative scenario it is a radical one.

The whole reason that Earhart hauled Noonan all the way around the world was so that he could get fixes to eliminate any errors that would have resulted from dead reckoning alone. This was also the reason, in the original planning, to take two navigators for the leg from Hawaii to Howland so that they could use their skill and equipment to find that Island. Note also, that they had recognized and decided that even the much shorter leg from Hawaii to Howland (only 1900 SM instead of 2556 SM) was too long to complete with only dead reckoning. If Earhart wanted to dead reckon that leg then there was no reason to incur the expenses of having two navigators on board, Earhart could have done the dead reckoning herself as she had done when flying solo across the Atlantic and when flying solo from Hawaii to California. All the planning for the World Flight was based on the knowledge that dead reckoning was not sufficiently accurate for the most difficult leg of the flight, that of locating Howland, coming either from Hawaii or from Lae.  The original plan was for Noonan to leave the flight at Howland and Manning to leave at Darwin with Earhart flying the rest of the flight solo. This is conclusive proof that they knew that finding Howland was THE most critical part of the entire around the world flight

I have attached Earhart's June 30, 1937 radiogram. It requested weather information because " FN MUST HAVE STAR SIGHTS."

In prior posts I have demonstrated how Noonan computed a "point of no return" (PNR) which allowed them to fly until 1407 Z and to within 817 SM of Howland and still be able to return to Lae. Since it is highly likely that Noonan knew of the location of the Rabaul airport after talking to the people at the Lae airport that was 400 SM along the course line to Howland. (And also, very likely, Noonan and Earhart had talked to pilots arriving from Rabaul.) Noonan would have calculated a PNR for a departure from Lae with a return to Rabaul. Since Rabaul was closer to Howland, this PNR would also be closer to Howland. Doing this calculation we find the PNR occurs at 1526 Z, 1901 SM from Lae, only 655 SM short of Howland (only 55 SM short of Tabiteuea in the Gilberts) and only three hours and forty-five minutes before the 1912 Z radio report of "must be on you" from Earhart. So if Noonan had not been able to get fixes they could have turned around and returned safely to Lae or Rabaul from nearly over the Gilberts and try again another day.

Do we know that Noonan knew how to calculate a PNR and that his practice was to do this calculation? Yes. On the departure from Hawaii to Howland on March 20th they took aboard an extra 75 gallons of fuel to allow a return to Hawaii after flying for 8 hours on the leg to Howland. This meant that they would have flown 1320 SM and would have come as close 580 SM to howland before turning around. This extra fuel was taken aboard after Noonan calculated the PNR based on having to fight a headwind on the return leg to Hawaii, this is the standard PNR computation.

I could have saved them the trouble of coding the computer (GIGO) to figure where they might have ended up if Earhart only used dead reckoning for the entire 2556 SM leg from Lae to Howland. The generally accepted uncertainty in a position determined solely by dead reckoning is 10% of the distance flown, in this case, 255.6 SM. So instead of doing all the work that they did they could have just drawn a circle around Howland with a radius of 255.6 SM and the circle would contain the location of the plane. This circle appears to contain the results of the Monte Carlo run (but is is hard to tell since they do not provide a chart sufficiently large enough to allow us to read the coordinates of their plot.)

So now the creators of this "Monte Carlo Simulation" take the position that in spite of all the expense and careful preparation and Noonan's prior careful calculations including calculating PNRs that after all that, flying almost all the way around the world, that at the critical moment Earhart and Noonan just decided to ignore all the prior planning and just said "AHHHH, let's just go for it!" And they call this a "conservative scenario."

This "Monte Carlo Simulation" is really just a Pachinko game!

See:
https://tighar.org/smf/index.php/topic,452.msg5586.html#msg5586

https://tighar.org/smf/index.php/topic,452.msg5620.html#msg5620




« Last Edit: January 01, 2012, 02:14:34 AM by Gary LaPook »
Logged

Gary LaPook

  • T5
  • *****
  • Posts: 1624
Re: The flight plan, magnetic course, headwinds.
« Reply #28 on: December 31, 2011, 04:16:35 AM »


Take this scenario for example:
They are in flight, then lose celestial observations due to overcast.  No big deal.  They revert to dead reckoning using the magnetic compass and directional gyro.  The DG suffers a bit more than its usual gyroscopic pression (maybe from a vacum leak).  AE corrects for it, but turns the DG needle to one hash-mark (or two) shy of it's actual correct heading.  Now they are are off by 10 degrees and don't know it.  They fly this heading for an unknown period of time.  Not knowing how long they have been flying the wrong heading, they dont know how much to compenstate to get back on course.  Fred notices the discrepency.  They are now closer to their destination and decide not to make matters worse by trying to counter-correct the problem.  But, rather to fly a 'parallel' heading an intercept an offset LOP.  Problem now they dont know exactly how far off course they are when reaching the LOP.  To compound the problem, say for example, the magnetic compass wasn't swung correctly or a 'typo' was on the correction card.  Combine that with unknown wind conditions.  Now things are starting to add up in a more complex way - creating a compounding effect.  They hit the LOP with as good of timing they have but even it is off slightly by several seconds.  Unbeknownst to them, the strip chart has Howland charted erroneously 6 miles too far east. 

Things have gotten real ugly real quick!
Pilots are trained to check their DG about every 10 minutes so Earhart would have caught the errors you are concerned about after only a short period of time.  There were actually three compasses in the plane and Noonan had one of his own back in the nav station so his job was to check up on Earhart flying the correct heading that he had given her so another reason that any such error was promptly corrected. In addition to the standard "pilot's" compass mounted above the instrument panel, Earhart had a much more accurate and stable compass mounted on the floor in front of the co-pilot's seat, see attached photos. Mounted directly above this second, aperiodic compass, is the correction card to this compass. Aperiodic compasses were generally called "the navigator's compass" because they were mounted at the nav station of our bombers and transport planes during WW2. Noonan had one of these mounted on the floor under the chart table with a window in the chart table to allow him to see this compass. Aperiodic compasses must be read from above, not from the side as is done with the normal "pilot's" compass. Since each of them had one of these very accurate and stable compasses available it makes you scenario very unlikely. Noonan also had an altimeter, airspeed indicator and an outside air temperature gauge at his station so he could compute the true airspeed of the plane and didn't have to rely on Earhart passing this information back to him. This was a very well thought out navigation arrangement, it had worked for 3/4ths of the way around the world.

Regarding "precession" of the directional gyro, pilots are not taught all the details of what they see in the cockpit. There is "real precession" and "apparent precession" but pilots don't know this, only flight navigators are taught to deal with these distinctions. Real precession is cause by friction in the bearings of he gyroscope that causes the DG to change from its setting and this is all that pilots are taught. Real precession is normally very small (unless the DG has been damaged or the bearings are very worn.) What pilots are actually seeing is "apparent precession" which is caused by the rotation of the Earth on its axis. The Earth rotates 15.04° per hour (in inertial space) and even though the gyro is maintaining it's direction in space, since the earth is turning under it, the DG appears to precess in the opposite direction. A plane flying over the North Pole would see the DG precess at this rate of 15.04° per hour. However the rate of apparent precession varies with the sine of the latitude so pilots in the U.S. see their gyros precess about 10° per hour (the sine of 45° latitude is 0.7 times 15.04° per hour equals 10.5° per hour.) The reason I have gone through this exposition is because the sine of zero degrees is zero which multiplied by 15.04 equals zero thus making the apparent precession at the equator, latitude zero, also zero. Since almost all of the precession seen in the DG is actually apparent precession, Earhart's DG would not have precessed much, if at all, since she was flying along the equator.

The first three attached photos show Earhart's cockpit, the aperiodic compass and its correction card. The fourth photo is of an exemplar aperiodic compass, not Earhart's, which has finer graduations on the azimuth scale than Earhart's. We don't have a photo of Noonan's aperiodic compass but it may also have had the finer graduations.

gl

« Last Edit: August 03, 2012, 08:06:40 PM by Gary LaPook »
Logged

Martin X. Moleski, SJ

  • Administrator
  • *
  • Posts: 2882
Re: The flight plan, magnetic course, headwinds.
« Reply #29 on: December 31, 2011, 08:57:43 AM »

The Monte Carlo simulation is mentioned quite often. The articles describing it talk about constraints on the results but lack detail on the simulation itself. Is it probabilistic or deterministic? What are the equations of motion? What physics is it modeling and how were the math models derived? What methods/software were used to build it? How was the simulation verified and/or validated? Without knowing more about the tool itself, it's hard to evaluate its results.

I recommended the article as a starting point for those trying to understand the initial conditions and constraints of a simulation.

I did not participate in the design of the simulation and can't answer those questions.

Randy Jacobson, who, I hope you will concede without further argument, did amazing amounts of research for TIGHAR, no longer participates in this Forum.  He has not said why he does not participate.

Quote
Based on the description in the article, this sounds more like a prediction of position based on radio messages than an actual 6DOF flight simulation resolving the forces and moments acting on the aircraft defining its motion through space. Is that the case?

That's how I read the article.  Why design something that requires a supercomputer and measurements we don't have.  Calculating the movement of a body based on fundamental forces with six degrees of freedom over twenty hours of flying is more work than you can or need to do.  There are an indefinite number of paths her aircraft might have taken to reach the various "checkpoints" available for a simulation--with a very region for the "checkpoint."

Quote
Another comment is the two figures in the wiki article are too small to read, so I'm not sure what they indicate.  Are the scales in latitude/longitude?

Yes.  The second image is easier to interpret than the first.  There is an "H" for "Howland" and a "B" for "Baker."
LTM,

           Marty
           TIGHAR #2359A
 
Logged
Pages: 1 [2] 3 4 ... 35   Go Up
 

Copyright 2018 by TIGHAR, a non-profit foundation. No portion of the TIGHAR Website may be reproduced by xerographic, photographic, digital or any other means for any purpose. No portion of the TIGHAR Website may be stored in a retrieval system, copied, transmitted or transferred in any form or by any means, whether electronic, mechanical, digital, photographic, magnetic or otherwise, for any purpose without the express, written permission of TIGHAR. All rights reserved.

Contact us at: info@tighar.org • Phone: 610-467-1937 • Membership formwebmaster@tighar.org

Powered by MySQL SMF 2.0.15 | SMF © 2017, Simple Machines Powered by PHP