French Judiciary Report Raises Fresh Doubts About MH370 Debris

Zero windage

After French authorities retrieved the MH370 flaperon from Réunion Island, they flew it to the Toulouse facility of the DGA, or Direction générale de l’Armement, France’s weapons development and procurement agency. Here the marine life growing on it was examined and identifed as Lepas anatifera striata, creatures which have evolved to live below the waterline on pieces of debris floating in the open ocean.

Subsequently, flotation tests were conducted at the DGA’s Hydrodynamic Engineering test center in Toulouse. The results are referenced in a document that I have obtained which was prepared for judicial authorities by Météo France, the government meteorological agency, which had been asked to conduct a reverse-drift analysis in an attempt to determine where the flaperon most likely entered the water. This report was not officially released to the public, as it is part of a criminal terrorism case. It is available in French here.

Pierre Daniel, the author of the Météo France study, notes that the degree to which a floating object sticks up into the air is crucial for modeling how it will drift because the more it protrudes, the more it will be affected by winds:
Buoyancy extract

This translates as:

The buoyancy of the piece such as it was discovered is rather important. The studies by the DGA Hydrodynamic Engineering show that under the action of a constant wind, following the initial situation, the piece seems able to drift in two positions: with the trailing edge or the leading edge facing the wind. The drift angle has the value of 18 degrees or 32 degrees toward the left, with the speed of the drift equal to 3.29% or 2.76% of the speed of the wind, respectively.

The presence of barnacles of the genus Lepas on the two sides of the flaperon suggest a different waterline, with the piece being totally submerged. In this case we derive a speed equaly to zero percent of the wind. The object floats solely with the surface current.

This suggests a remarkable state of affairs.

Inspection of the flaperon by Poupin revealed that the entire surface was covered in Lepas, so the piece must have floated totally submerged—“entre deux eaux,” as Le Monde journalist Florence de Changy reported at the time. Yet when DGA hydrodynamicists put the flaperon in the water, it floated quite high in the water, enough so that when they blasted it with air it sailed along at a considerable fraction of the wind speed.

As point of reference, Australia’s CSIRO calculates that that the drifter buoys that it uses to gather ocean-current data pick up a 1.5% contribution from the wind. Here is a picture of one such drifter, kindly supplied to me by Brock McEwen. You can see that more than half of the spherical buoy is out of the water.

DSC_0279

It is physically impossible for Lepas to survive when perched up high in the air. Yet the buoyancy tests were unequivocal. So Daniel pressed on, conducting his analysis along two parallel tracks, one which assumed that the piece floated high, and the other in which it floated submerged. For good measure, he also considered scenarios in which the flaperon floated submerged until it arrived in the vicinity of Réunion, and then floated high in the water for the last two days. (Note that he doesn’t present any mechanism by which a thing could occur; I can’t imagine one.)

After running hundreds of thousands of simulated drift trials under varying assumptions, Daniel concluded that if the piece floated as its Lepas population suggests, that is to say submerged, then it couldn’t have started anywhere near the current seabed search area. (See chart above.) Its most likely point of origin would have been close to the equator, near Indonesia. His findings in this regard closely mirror those of Brock McEwen and the GEOMAR researchers which I discussed in my previous post.

Daniel found that when simulated flaperons were asssumed to have been pushed by the wind, their location on March 8, 2014 lay generally along a lone that stretched from the southwest corner of Australia to a point south of Cape Horn in Africa (see below). This intersects with the 7th arc. However, as Brock has pointed out, such a scenario should also result in aircraft debris being washed ashore on the beaches of Western Australia, and none has been found. And, again, the presence of Lepas all over the flaperon indicates that such a wind contribution could not have been possible.

With windage

Pierre Daniel’s reverse-drift analysis for Météo France, therefore, presents us with yet another block in the growing stack of evidence against the validity of the current ATSB search area in the southern Indian Ocean.

The most important takeaway from this report for me, however, is the stunning discrepancy between how the flaperon floated in the DGA test tank and the “entre deux eaux” neutral buoyancy suggested by its population of Lepas. No doubt some will suggest that the flaperon may have contained leaky cells that slowly filled as it floated across the ocean, then drained after it became beached. However, I find it hard to believe that an organization as sophisticated as the DGA would have overlooked this eventuality when conducting their wind tests. Rather, I read Daniel’s report as evidence that the French authorities have been unable to make sense its own findings. I suspect that this is the reason that they continue to suppress them up to this day.

499 thoughts on “French Judiciary Report Raises Fresh Doubts About MH370 Debris”

  1. @Oleksandr: Please stop playing games with semantics. I was obviously referring to whether the Coriolis forces and wind affect the path of the plane.

    @Gysbreght: No confusion at all.

    @Warren: In HDG HOLD, the navigational heading is held. That would be either the true heading or the magnetic heading, depending on what is selected. It’s that simple.

    I’m done debating topics that were debated and settled two years ago. You three can debate among yourselves.

  2. LNAV is one of the autopilot roll modes that can be selected on the MCP. AFAIK there is only one LNAV mode, and it is controlled by the FMC. There are several VNAV pitch modes, also controlled by the FMC.

  3. @ Victor: please don’t take this the wrong way, but if it’s that simple, then you are simply wrong. In HDG HOLD, the navigational heading is simply held the same, as you say, and as I thought.

    In the simple example of HDG HOLD TRUE 180, the head of the aircraft will always be pointed due south. The thrust component from the engines will also always be to due south. However, the eastward momentum (p = mv) component is still going to be conserved, causing the plane to drift to the east. I got a 2 year-old spreadsheet that quantifies this effect.

    Even nuclear submarines on long underwater cruises have to compensate for Coriolis.

  4. Warren Posted May 5, 2016 at 12:33 PM: “@Gysbreght: there’s a guy on a.net whose job is to take out 777’s fresh from the Boeing assembly line out for test flights before they are handed over to the customer to make sure everything works correctly. ”

    We’ve discussed that post earlier. There are doubts about that guy’s authenticity. As far as I remember, he doesn’t claim to take out 777’s fresh from the Boeing assembly line out for test flights. Perhaps he does that for 737’s, because he doesn’t seem to be quite familiar with the 777’s characteristics. See also comments from RetiredF4 at the earlier discussion.

    In normal mode with autopilot off, the flight control system maintains bank angle without pilot inputs. In the pitch axis the flight control system maintains airspeed (which is different from the Airbus system which maintains flight path angle). Also all the other goodies such as envelope protection, turn coordination, thrust asymmetry compensation, overbank protection, are available.

  5. Victor,

    I noticed all IG members are hyper-sensitive if someone corrects their mistakes. I know what you meant, but other readers may not.

    But I am confused what you tried to say. LNAV is one of the autopilot roll modes that can be selected on the MCP, as Gysbreght wrote. If it is disabled, wind and Coriolis affect flight path. Do you disagree with this?

  6. @Ron said;
    …”presence of a Singaporean airborne radar in the northern part of Malacca Straits.”
    Do you have a reference for this statement?

    @MH said;
    …”Kate may have seen the Singaporean AWACS aircraft”…
    Singaporean AWACs is a Gulfstram G550 – it doesn’t look like a Boeing 777.
    Also, it was not 9M-TGH, as it had been repainted from a DHL paint livery
    to a Transmile paint livery by 7th March 2014, as proven by one picture
    on the internet (and possibly a 2nd picture which my browser can’t seem to
    access).
    Brief summary of Kates’ points;
    -seen for about 5 to 10 minutes (went below during sighting)
    -whole plane ‘glowing orange’, at least partly seen with a ‘halo effect’
    -no navigation lights
    -apparent trail of smoke behind (tail) of aircraft
    -even lower than 10,000ft
    -no lit windows (but cockpit windows were visible) and;
    ” I was looking”…”for the windows”…”but I couldn’t see them where they
    should have been.”…”I could see where the windows should have been”…
    “but the whole hull just appeared to be one plain surface, as though they’d been painted over”
    -didn’t remember seeing any logo (i.e., airline logo)
    (she may have made a comment about a longer tail or more pointed front –
    I can no longer find such comment unfortunately).
    _
    Given she saw an aircraft with no (emphasis) ‘passenger windows’, that
    narrows the candidate aircraft considerably.
    It therefore would not have been an Indonesian military Surveiller 737.
    I cannot find any pictures of the Chinese purchased Boeing 737s that China
    fitted out and operate for military maritime surveillance, so it is
    unknown if they have passenger windows.
    The US navy P-8 Poseidon (737 airframe) seems not to have been in actual
    navy service by March 2014, but it’s possible pre-operational service
    testing could have been occurring -(can anyone definately rule them out?)
    The U.S. Air Force and the Singaporean military Boeing KC-135 aerial refueling aircraft are most likely candiates. Both countries KC-135’s
    have livery that has lettering that would be too indistinct to read when
    seen at a distance, and have small (country) flags that would be
    difficult to recognize when seen at a distance (other than the flag seen perhaps as a tiny ‘spot’ of colour, easily missed).
    https://en.wikipedia.org/wiki/Boeing_KC-135_Stratotanker
    Interestingly, Kate mentioned that she also saw two other aircraft that
    she thought were flying together (one somewhat below the other) – perhaps
    they were a couple of military jets flying in formation, waiting to be or
    having been ‘refuelled, which would give a reason why a U.S. KC-135 would
    be in the area.
    A Singaporean KC-135R could additionally have been conducting maritime
    surveillance in the area, as they use theirs for a wider range of tasks.
    The ‘orange glow’ would most likely be sodium lighting, which is noted for
    the way it ‘washes out’ other colours to give a slight orange appearance
    to whatever it is ‘lighting up’. (Also noted for its ‘halo’ effect).
    http://www.newyorker.com/magazine/2009/03/23/light-fight
    A military surveillance tasked aircraft may use such lighting as I believe
    it is easier to use infrared night vision devices if the only impinging
    visible light is sodium lighting, which is known for producing only a very
    narrow frequency of light (so only have to filter out that narrow
    frequency of ‘orange’ light).
    Additionally, although a surveillance aircraft would not necessarily have
    any need to have all its hull ‘lit up’, a refueling aircraft would want
    to have as much of its hull visible as possible to maximise the reference frame for the ‘to be refueled’ aircrafts pilot (consistent with not unduly
    using bright lights that make it easy to sight – from a distance! and not
    using any possibly distracting ‘bright’ lights, like navigation lights).
    A final point – Kates opinion when she first saw the ‘orange glow’ plane,
    was that it was a missile because it seemed to be deliberately coming
    straight at her boat – if it was a KC-135R carrying out a near
    surveillance of her boat, it makes sense it would initially be seen to
    fly straight towards her boat, then turn away (possibly she was below
    when it actually turned away at a distance) after her boat had been seen
    from a distance (in infrared?) and possibly photographed.

  7. Victor,

    Re “I’m done debating topics that were debated and settled two years ago.”

    Really? Can you point out where it was discussed 2 years ago?

  8. Jeff,

    “As I recall this is the only autopilot-only endpoint that anyone has managed to come up with in the SIO outside of the current seabed search area.”

    I have not wrapped up a note on FPA+ALT HOLD mode yet, because of the lack of general interest and lack of time. Anyhow, I am sure it will change nothing. Also it suffers from the same drawback as IG’s TRK/HDG HOLD: why would someone switch to FPA +0.1 deg?

  9. @Susie

    “Chillit is proposing a ‘triangulated’ position somewhere near the Zenith Plateau.”

    I hope he can convince of a needed sand search, because i cannot.

    The stuff down there might not even be on the sand, but just beneath it. Metallic, so it is detectable.

  10. Thanks Trond

    I think he was just playing around with drift models and then got involved in a mathematical experiment. I’m not sure how convinced he is of its reliability – I was just interested in the method and whether it’s been debunked-in-advance if that makes sense.

  11. Oleksandr wrote “Ken Goodwin, Perhaps you could shed light: are B777 equipped with fuel dumping facility?”

    I believe fuel dumping is a requirement for aircraft that have a much higher takeoff weight than landing weight. See: https://en.wikipedia.org/wiki/Boeing_777 Thus the need to reduce weight prior to landing after an emergency. Better than tossing luggage overboard. I have not checked all aircraft but long range aircraft usually have take off and landing weight differences because of the large amount of fuel that must be carried relative to the overall aircraft weight. The 737, a shorter range aircraft, does not show a landing weight. https://en.wikipedia.org/wiki/Boeing_737

    Have not checked specifics.

  12. Gysbreght wrote @Ge Rijn: The flaperon skin is a honeycomb/composite sandwich, as can be seen where the trailing edge broke off. The sandwich construction eliminates the need for stringers, but not the need for ribs.

    H/C reduces the need for internal ribs/stiffeners. If thick enough, skin to skin, no ribs are required. The T/E panel is relatively small. It does not have ribs except closeout ribs at the ends of the T/E to seal the panel. The H/C has replaced the requirement for ribs to go all the way to the tip of the part.

    The flaperon is a panelized h/c composite assembly with some full depth parts at the T/E and a spar assembly at the leading edge to pick up the loads from the rest of the assembly. The design was developed by balancing the weight, strength and cost requirements for the part. Designs are always a compromise between competing requirements.

  13. @Ge Rijn wrote “The flaperon is essentialy watertight sealed as a whole. It must be for you can not allow water building up somewhere within during service/flight. This offcourse would effect greatly on weight, performance, corrosion etc. I don’t know if it’s got separate sealed compartements. It could be.”

    Assemblies such as the flaperon are designed with drain holes at the bottom of the structure as positioned on the ramp. i.e. full up / retracted and stowed positon. Drain holes inboard / aft. Individual panels within the assembly will be sealed. E.g. A trailing edge panel or a top or bottom panel or spar or web panel will be sealed to prevent water ingestion into the h/c core. Note: The fuselage and dry bays of the wing have drain holes at the bottom of the structure to allow water to drain out after a flight. A lot of the water is condensation from the air due to temperature changes. Many areas of the world with high humidity have a major problem with condensation. Personal experience with a project call “Rain in the plane”.

    So, an assembly like the flaperon will have sealed panels that resist water penetration but the inside of the assembly it will have a few drain holes positioned to drain water after a flight that will allow water to enter overtime and fill the inside compartments. This water will reduce the buoyancy of the hollow structure but once filled with water will not increase/decrease the buoyancy of the sealed panels within the assembly. The water inside the assembly should fill to the waterline of the part floating in the water. Looking inside the part should show a line on the inside surfaces where the waterline might have been. The water will increase the effective mass of the structure as affected by waves and swells thus causing the assembly to act like a much heavier object in the water. E.g. Dead Head log.

  14. @JS wrote “@Ken Goodwin, You implied that a depressurized cabin would be smaller diameter than an 8k pressurized cabin, at altitude. Is that an accurate interpretation? Would this effect impact fuel consumption, even by a very small amount?”

    Yes. The pressurization cycle of an airplane fuselage is one of the key factors/loads governing the design of the fuselage structure. Fatigue of the structure, due to the pressure cycle, is critical to the design. 737 had issues with just this load cycle. Fuel consumption differences would be very very small and probably fall within the noise calculation. E.g. an airplane tested with or without pressurization may give the same answers in a flight test. Calculated values would be different; but actual flight test results would probably show no difference.

  15. @Oleksandr

    “1. Forward model. You know there are potentially two origins, one in each tributary. You run simulations starting from a number of points in each of the tributaries, and eventually find two potential origins. The model, however, will not answer which one of the two solutions is the one you are looking for.

    2. Reverse model. Suppose mixing is sufficient to make particles evenly distributed across the river in the place of joining of the tributaries. Then 2/3 of particles will move into the tributary with higher flow speed, and 1/3 of particles will move into the tributary with slower speed. What conclusion will you make? Will you conclude that the point of origin is twice more likely located in the tributary with higher speed?

    The real ocean is a way more complex system of vortices and confluences.”

    Yupp, that’s why we are talking about probabilities…and I have the feeling these ocean experts know what they are doing.

  16. @Warren Platts: I’ll try one last time. In HDG HOLD, the autopilot will bank the plane as required to maintain a constant heading RELATIVE TO THE EARTH. There is NO autopilot mode for a B777 in which an inertial direction is maintained. In HDG HOLD, the corrections to the path that are required because the earth reference is not an inertial frame automatically occur. So, if a particular heading is chosen, such as 186 deg true, rotation of the earth does not cause the plane to drift towards the east relative to ground. In estimating the heading, the ADIRU automatically corrects for earth rotation so that the desired heading may be maintained by the autopilot.

    For those wishing to learn more, you can refer to Section 6.3 of the DSTG report on Bayesian Methods from Nov 2015. This summarizes how the 5 navigational modes (LNAV and {mag or true} and {HDG or TRK}) can be mathematically modeled. Brian Anderson further explained these modes here:

    http://www.duncansteel.com/archives/2115

    Most of us that have reconstructed flight paths (IG, Bobby Ulich, sk999, Richard Cole, etc.) have been modeling the navigational modes for many months exactly as the DSTG has presented them, so I remain surprised that there is still confusion because these questions have been resolved many months ago.

  17. @JS and @Ken Goodwin: If the plane was depressurized and there was no bleed air required for the air packs, that would have a more significant effect on fuel flow, probably on the order of several tenths of a percent.

  18. can anyone suggest or summarise which parts of the plane these 4 debris belong to? all flaperon? does that give some hint as to how the plane failed?

    is it possible that the plane lost one or part of its wing and continue to fly?

  19. @Warren, @Victor, @Oleksandr

    Re. Coriolis Force: because the atmosphere is basically rotating with the Earth, the Coriolis Force would have a minimal effect on an aircraft flying due south or north.

    I would like to point out that I am not a member of the IG, and will therefore not be unduly upset if you challenge the above statement 🙂

  20. Victor,

    Citation from B777 operations manual, section 4.10.16.

    AFDS Roll Modes (Engaged):
    – HDG HOLD
    – HDG SEL
    – LNAV
    – LOC
    – ROLLOUT
    – TO/GA
    – TRK SEL
    – TRK HOLD
    – ATT

    AFDS Pitch Modes (Engaged):
    – TO/GA
    – ALT
    – V/S
    – VNAV PTH
    – VNAV SPD
    – VNAV ALT
    – G/S
    – FLARE
    – FLCH SPD
    – FPA

    Would a trajectory resulting from the selection of, for example, ATT+ALT, be affected by Coriolis and wind or not in your opinion?

  21. ROB,

    “because the atmosphere is basically rotating with the Earth, the Coriolis Force would have a minimal effect on an aircraft flying due south or north.”

    I hope you understand what you wrote, because I don’t. Coriolis force has no dependence on the atmosphere at all.

  22. @ Oleksandr – question Mr. O…..would the Coriolis effect/force be anyway related to a trap/skeet shooter leading the bird (target)….? thanks

  23. @Oleksandr: ATT mode is for bank angles greater than 5 deg. I was referring to “wings level” navigation modes such as TRK HOLD and HDG HOLD.

    Please read the materials I provided.

  24. Victor,

    “ATT mode is for bank angles greater than 5 deg.”

    Where does it come from? My understanding of B777 operations manual is that this is applicable only when “both flight director switches are OFF” (section 4.10.1). If, however, either flight director is “ON”, the autopilot engages in the selected flight mode according to the manual.

  25. @jeffwise said, “@VictorI, Thanks for refreshing our collective memory. As I recall this is the only autopilot-only endpoint that anyone has managed to come up with in the SIO outside of the current seabed search area.”

    If you allow for scenarios in which there is a loiter between 18:40 and 19:41 and also a slow descent, there is a range of possible end points corresponding to different positions at 19:41 and different starting altitudes. I have not done a comprehensive study over the parameter range, but merely demonstrated that solutions do exist for automated flight paths outside of the search area.

  26. @Oleksandr: Please show me the language which allows ATT mode for zero bank angle. I only see references for bank angles greater than 5 deg.

  27. Victor,

    All the references to 5 deg are applied when both the flight directors are off. If AP is disengaged, and flight directors are OFF, then when the first of them is switched “ON”, the default mode really depends on bank angle.

    But where did you find that a pilot cannot manually select ATT mode if bank angle <5 deg, or that ATT will revert to TRK/HDG HOLD against pilot wish?

  28. FCOM 4.20.6:

    ATT – (engaged) – When the autopilot is first engaged or the flight director is first
    turned on in flight, the AFDS holds a bank angle between 5 and 30 degrees and
    will not roll to wings level. If the bank angle is less than 5 degrees, the AFDS
    returns to wings level (HDG HOLD or TRK HOLD). If the bank angle is greater
    than 30 degrees, the AFDS returns to 30 degrees of bank.

  29. @Oleksandr

    Re the so-called Coriolis Effect:

    I am totally aware of what I said. Coriolis is more accurately described as an effect rather than a force, which why it is referred to as a “fictitious force” However, being called fictitious does not mean it is non-existent.

    It would only affect the flight path if the aircraft were moving south in a vacuum, which is clearly impossible. By it’s very nature, an aircraft can only move through an air mass, in this case an air mass which rotates with the earth as if it were (give and take) fixed to the earth. This movement of the air mass has the effect of nullifying the Coriolis Effect.

  30. @Ken Goodwin

    It’s very good that you have been able to share with us your inside knowledge, particularly in relation to how the flaperon is constructed.

    Slightly off topic: I was reading a pilots’ forum a while back, where they were discussing/comparing the qualities of different makes of aircraft. One of the contributors summed things up by saying that the ideal airliner would have been designed by Lockheed, built by Boeing, and marketed by Mcdonnell Douglas.

  31. Victor,

    “I have not done a comprehensive study over the parameter range, but merely demonstrated that solutions do exist for automated flight paths outside of the search area.”

    The main issue with your solution is the use of magnetic heading. I was probably the first who tried to use magnetic heading instead of the true heading and came to 32S. Then Sk999. I am happy that you guys in IG repeated our work and came to the same conclusions 2 years later. But when did you confirm that magnetic heading can be used for navigation rather than for display purpose?

    I am very skeptical about the use of the magnetic heading for navigation. The primary source of position/velocity is ADIRU. GPS can provide accurate position updates to ADIRU via FMC, but not heading. If ADIRU fails, then GPS provides position data. Why would one use magnetic heading for navigation these days? Magnetic heading can be supplied by SAARU, but that would imply failed ADIRU.

  32. @Oleksandr: What sequence of pilot actions would put the plane in ATT roll mode? I am only aware of entering it while in a bank of greater than 5 deg and engaging A/P. I don’t see any way to select the mode from level flight and I see no way to set the bank angle. Why do you keep insisting that this navigational mode is possible under normal conditions? As I said before, what you describe is CWS, which is not available for a B777.

  33. @Oleksandr

    Why do you think GPS cannot provide accurate heading information? The GPS sensor is more accurate than any sensor on the aircraft with respect to heading. Is it because data fields are not allocated for heading information from GPS?

  34. Gysbreght,

    “When the autopilot is first engaged or the flight director is first turned on in flight.”

    Isn’t it consistent with what I wrote above? First engaged… First turned on… If after that pilot manually disengages TRK/HDG HOLD and engages ATT?

  35. DennisW,

    Heading describes the direction the vessel is pointed in. GPS can only provide an accurate track.

    To provide an accurate HDG via GPS, you need 2 GPS, spaced far apart and derive the vessel’s pointing direction from the two GPS positions.

  36. @Rob: We are getting off into the weeds, but if the plane was on an inertial path and the air was still relative to the earth, there would be a relative lateral velocity between plane and the earth which would tend to counteract the Coriolis effect and “drag” the plane along with the air. But since the Coriolis effect requires a continuous force, there would have to always be some relative lateral velocity between the plane and the air. The amount of this relative velocity would depend on the lateral drag coefficient of the airframe. If the drag coefficient were low, then the plane would follow an inertial path, and the Coriolis effect would have a big effect on the path. If the drag coefficient were very high, the Coriolis effects on the path would be small.

    Luckily, there is a not a “wings level” roll mode of a B777 that produces a path that is influenced by Coriolis effects.

    Also, if a plane was on an inertial path, the altitude would also change due to the curvature of the earth. Again, we don’t have to worry about this because the vertical modes use altimeters to control altitude, vertical speed, or flight path angle (FPA).

  37. Victor,

    “Why do you keep insisting that this navigational mode is possible under normal conditions?”

    Because it is listed in the manual as normal mode. It is clearly stated that if flight director is ON, such a mode can be selected. Why not? Why do you keep insisting it is impossible mode? And why do you think bank angle >5 deg is required? ATT is probably the simplest flight mode.

  38. @MuOne

    That is absolutely not true. GPS is a data rich sensor, and two of the most frequently used outputs are the vectors [x,y,z,t] and [Vx,Vy,Vz,F] where the first vector is position and the time offset of the local GPS clock, and second vector is the velocity and the frequency offset of the local oscillator. In fact, the velocity data derived from the Doppler residuals has less noise than the position data derived from code phase measurements. It is common practice even in cheap GPS receivers to use Doppler smoothing to make the position outputs less noisy.

    Multiple GPS antenna platforms enjoyed a very brief trial period as sensors for pitch, roll, and yaw, but that application fell out of favor a couple of decades ago.

  39. @Oleksandr: First, I am not aware of any previous work to mine that showed that a slowly descending flight using FPA vertical mode, autothrust, and autopilot would end near 29S. The point is it is a completely automated mode–pitch, roll, and thrust. Yes, many of us prior to this showed that magnetic headings would produce curved flights, but only if the speed was manually changed during the flight.

    As for your insistence in believing that there is an ATT mode for level flight, and magnetic heading hold does not exist, I can’t persuade somebody that doesn’t want to be persuaded, despite the mountain of documentation to the contrary. I give up. I can’t even find a button or knob selector to enter ATT mode from level flight. So please explain to us the sequence of buttons to enter this mode.

  40. Dennis,

    Re: “Why do you think GPS cannot provide accurate heading information?”

    GPS updates are much less frequent than ADIRU updates. That is why velocity derived from GPS may be significantly less accurate. The other aspect is reliability. ADIRU is self-sufficient internal source of data. GPS relies on external source. You can imagine what will happen if GPS network is down.

    Re “The GPS sensor is more accurate than any sensor on the aircraft with respect to heading.”

    No. IRS are a way more accurate. If I recall correctly, laser gyros have drift of order 10^-7 deg/s. State-of-the-art piezoelectric gyros – 10^-8 deg/s. But position is comparatively inaccurate, mainly due to errors in accelerometers and numerical integration.

  41. DennisW,

    I agree, GPS is very accurate in sensing TRK, i.e. Vx, Vy, Vz.

    Not so for HDG. Gedanken experiment. Put the GPS in line with the plane’s axis on a flat surface. Read the Vs, now turn it to be in line with the wings. The Vs read the same, because the track is the same. However the “heading” of the GPS unit is now 90deg away from its track.

    You’d need a built in compass, or the abovementiond second independent GPS unit for a second position fixed to the vessel, to derive HDG, i.e. the orientation of the vessel.

  42. @VictorI said: The possibility of an automated flight that ends near 29S was discussed on this blog. It requires a loiter between 18:40 and 19:41 and a descent of -0.1 deg Flight Path Angle (FPA) and a constant magnetic heading.

    Just a thought … If the aircraft started off at (say) 35k ft around 19.41 and descended at -0.1 degree FPA, what altitude would it have descended to when it ran out of fuel?

  43. @Victor

    Intuitively, a plane flying through Earth’s atmosphere is not flying an inertial path, whether it’s travelling from north to south, or from east to west.

  44. VictorI is correct – in heading mode, Coriolis forces do not affect the direction of travel. However, understanding why that is so turned out to be an interesting problem, so here is how I explain it.

    First, ignore the atmosphere, and consider the surface of the Earth to be frictionless. Also, consider the Earth from the point of view of an observer in inertial (non-rotating) space. An aircraft moving in any direction while on the surface will circle the Earth along a great circle path (NOT the same great circle as in LNAV mode, since that is executed in a rotating frame.) The only force is that of gravity, which is in a direction perpendicular to the direction of travel. Let’s put in actual numbers. The equator rotates at 900 knots. Take an aircraft that, when viewed in a rotating frame, starts at the equator traveling due South at 450 knots. In inertial space, the aircraft is actually traveling on a path that is inclined to the equator by arctan(450/900) = 26.6 degrees. In a rotating frame, the aircraft does not follow a straight path South, but rather follows a path that seems to be deflected left (to the East) by that fictitious Coriolis force. Further, the heading is not constant, but varies continuously along the path. To emphasize, the aircraft would not feel any force beyond that of gravity.

    In HDG HOLD mode, the plane actually needs to bank a bit to the right in order to maintain a constant heading.

    I hope that helps.

    P.S. regarding the use of magnetic heading for travel, yes, it is the normal mode when traveling a fixed heading route, even for commercial jet aircraft. Airport runways are desginated by magnetic heading. Most general aviation aircraft do not have IRUs and rely on magnetic instruments for heading information. Since all planes fly in the same
    airspace and use the same air traffic controllers, the lowest common denominator rules.

  45. @Oleksandr and MuOne

    I’ll just say your understanding of GPS is wrong.

    The GPS receiver will output the same heading regardless of how it is oriented. It is based on the motion of the antenna phase center relative to the satellites. Heading can be derived from the velocity vectors with a trivial amount of math.

    As far as velocity update rate is concerned modern receivers are capable of updates, both position and velocity every 20ms (although 50ms is a common spec requirement). Typical velocity accuracy is better than 0.05 meter per second (non-averaged single fix). GPS does not suffer from gyro drift, and is almost always used to calibrate inertial sensors during flight (even ring laser gyros). In fact, many conferences have been and are dedicated to the blending of GPS and inertial data.

  46. @MuOne,

    Because if the plane was pointing at 180, but actually heading 190 because of high side winds?

    So the GPS says the heading is 180 when it’s really 190?

    Makes sense.

  47. @sk999: Yes, some banking would be required for a constant heading, but the amount of banking is reduced by the lateral air drag on the plane as the air moves with the earth and this imparts a lateral force on the plane.

  48. @Dennis – but the plane isn’t necessarily pointing in the same direction it’s traveling.

    Isn’t that a dealbreaker?

  49. @sk999: You would also need some down elevator action since if you are flying tangent to the surface and holding inertial direction, the ground would curve away from you, increasing your altitude.

  50. @Middleton: For that particular example that I provided, the (pressure) altitude started at 38,000 ft at 19:41 and was at about 15,600 ft at fuel exhaustion.

Comments are closed.