AF 447 Search to resume

[henra]

Oh well, that's joined up thinking. Now, what surprises might be sitting on some military anti-submarine satellite somewhere.

Indeed, that would be interesting to know.

On the other hand that part of the globe is probably not amongst the premier targets of military intelligence, i.e. the chance of plenty military satellites taking images of the vast ocean in that area is rather slim, if not Null.

The next question would be: If images would have been taken by military satellites (of whatever Country), would this information have been made available to BEA ??

[Lonewolf_50]

The next question would be: If images would have been taken by military satellites (of whatever Country), would this information have been made available to BEA ??

Only if wikileaks passed it along to them.

[Backoffice]

I was ribbing the Cosmo team unfairly, they at least got an image.

The bigger question is why, with the existing technology we have today, nobody seems to put 2 and 2 together. That Cosmo satellite happened per chance to snap the right area 30 hours after the event and, there may be useful evidence there. Did all the satellite operators, civil and military think about taking an image, even if it was sometime after the event, probably not.
A high level U2/TR1 flight over the area the next day may have speeded up identifying where the wreckage could be found, but as far as I know, that didn’t happen either.

We just seem to be using the same search techniques from 40 years ago, which also happens to be the same timeframe for FDR/CVR/ELT development.
Perhaps it’s just unfashionable to find the answers and improve safety.

[bearfoil]

Not unfashionable. Expensive.

[jcjeant]

Hi,

Indeed .. safety first .. when this cost nothing ..
Refer also to the oil offshore activities (I was there ....)

[rfp172]

Bouys

The real question is why all of these "professional or experienced" SAR teams neglected to drop any kind of traceable buoys after the apparent accident for days. The technologies have existed for over 40 years and should have been dropped as a first priority to track the currents and potential disbursement of the anticipated wreckage. The why is the real question - did they not want to locate or just plain stupid??

[bearfoil]

For instance, what would it cost to develop an a/s sensing instrument preferable to the Pitot? A device with no sharp bits to collect ice, no moving parts, and a simple mechanism to measure tagged airflow, from release to capture, and compute an a/s correct to tenths of a knot. Not so very much, especially if applied system wide. At first it would be a back up for the current methods, since its strong suit is accuracy in poor Wx. It would no doubt replace all pitots, though GA would most likely not be interested.

rfp172 A Poll then? I vote "Did not want to find". It would be inelegant to call them "Stupid".


mm43 Sonobuoys?

[JD-EE]

bearfoil asks mm43 "Sonobuoys?"

What would an SAR aircraft be doing with anti-submarine warfare tools? (I did a major portion of the design on the S3A's sonobouy receiver. The sonobouy life time was limited due to its needing to transmit, on channels around 130MHz, the sounds it heard back to the S3A that deployed them. Range was modest.)

[FluidFlow]

Location of AF447?

The following pic shows the most likely crash site for AF447 based on a polynomial regression from the locations of the first 22 bodies recovered, focused using the location of the SAR 'pollutions spot' (slick).
It uses a program called FODDA (FluidFlow’s Oceanic Drift Data Analysis Program) which used the method of least squares to interpolate the body drift from the SAR point and then only has to extrapolate a relatively small distance (30.03 hrs) to determine the most likely crash site. A dataset was selected which would minimize the analysis error. The drift of the bodies was chosen as this is the only data set that will have ‘least different ie most similar’ characteristics, however the actual body drift shows a ‘fanning’ out from this point which should also occur hence supporting the validity of this analysis. The same occurs for the VS and galley which are located further off this ‘body drift line’ than any other body, as would also be expected. The probability of this position being correct is virtually the same probability of the pollution spot being from AF447.

FluidFlow acknowledges the valuable work of the Drift Group which enabled this alternative analysis to be completed and would like to thank the BEA for publishing the Drift report.
I was planning on doing up a report to show the mathematics behind this analysis so would appreciate criticism so I can cover this in my report. I will have limited internet access over the next week but would value your comments. I trust FODDA will provide some ‘food for thought’.
Thanks
Ian

[mm43]

FluidFlow;

Thanks for posting your "least squares" polynomial regression curve deduced position for the crash location. It will be worthwhile to expand on the mathematical basis for the results, and how including the positions and times of as many objects as possible combines to reduce the error rate and increase the probability of the result being correct.

Perhaps, selecting a "notional crash site" at random, then taking the same "least squares" approach could help to validate the probability of the "oil slick" calculation.

mm43

[FluidFlow]

mm43,
Thanks for your feedback. I will do some more work in those areas as time permits. regards
FF

[Machinbird]

And that is about as close as we have come to X marks the spot to date. Nice work FluidFlow!
An impact South of LKP would seem to confirm an early entry into a deep stall and a relatively tight circular path after LKP, IMHO.
Does anyone know what happens when Airbus protections collide? Which takes priority? There aren't that many protections. It should be a relatively compact matrix.

In past discussions, we have assumed, for example, that Vmo/Mmo Protections would be limited by the pitch attitude protections, but it is really all dependent on how the software was coded.
There may have been a programming assumption that protections never collide, and thus a software booby trap could exist.

[GreatBear]

All Over the Map

Very impressive work on the part of the Drift Group, which developed a rectangular search area about 20 to 40nm NW of the LKP with an estimated 95% probability of successful discovery. Months of very high-tech science effort! Very nice regressions recently by FluidFlow based on body drift and the SAR oil slick yielding a "most likely crash site" about 20nm SE of the LKP. MM43's early backtracking work pointing to a site 30nm eastward of the LKP. My own "steady-drift" body recovery drawings from May (#1178). It seems that backtracking the debris drift has put the crash site literally all over the map, depending on the model or the method and the data employed.

The most scientific (Drift Group) results were surveyed by sonar at depth during Phase 3. Perhaps the wreckage of AF447 is indeed in that rectangle, but was missed; there have been many worrisome official comments about crevasses and deep steep valleys.

We do have the ACARS messages. If the upset began or was underway at altitude at the LKP at 0210 and immediately devolved into rapid and uncontrolled descent lasting no longer than five minutes (the A/C failed to report an expected ACARS message), then

The A/C would need to maintain an average speed of 300kts over the ground (in steady descending flight) for five minutes to reach a point on the water 24.9nm from the LKP (to FluidFlow's MostLikelySite, the Drift Group's rectangle, and further eastward to mm43's site). What's the probability of steady flight horizontal when it seems most likely from the ACARS messages that the A/C at 0210 at the LKP was upset (stalled) and already or immediately to be rapidly descending vertically (unrecoverable flat spin, later the cabin pressure alarms)?

I do wish there had been an Airbus pilot to temper the Drift Group's probabilities and explain how the A/C could possibly have been flown those 20 to 40 horizontal nautical miles (see Post #1349) from the LKP in ALT law, given the dire circumstances reported and implied by the ACARS messages delivered from the LKP at 0210.

If there is a Phase 4 search, they should begin at the LKP, where curiously there has been no bottom search yet. We'll know in a few weeks, when BEA makes their promised September announcement.

GB

[henra]

GreatBear,

from the ACARS positon reports the average Ground speed up to 2:10 can be derived at ~470kts.
I have difficulties to see a method in an Airliner at 35kFt to decellerate in such a short timespan that the average speed would drop to 200kts (except immediate loss of a Wing or an attempted loop). with normal and reasonable control input that should be hardly achievable.
(Even the guys who tried to break the high altitude record with a Canadair RJ a couple of years back who deep stalled it all the way down still travelled quite a distance before hitting short of the alternate airport).

However, that wouldn't rule out a position close to LKP though.
Taking into account that the general attitude of the aircraft presumably was out of control, it is even rather unlikely that the flight path in the final 5 minutes was straight.

[thermalsniffer]

Henra

Taking into account that the general attitude of the aircraft presumably was out of control, it is even rather unlikely that the flight path in the final 5 minutes was straight.

Question: How erratic could the flight path be back to LKP? All I have seen to date is MM43's nice bank angle arc analyses, so this seems like a new proposition to me.

[JD-EE]

henra, you touched on a serious point here. The plane had a lot of kinetic energy, mV^2. It also had 35,000' of potential energy. It had a specific velocity vector to that energy. Changing the velocity vector as projected on the ground requires trading mgh for mV^2, external force input, or both sufficiently gentle that the plane does not crumple and sufficiently large that the plane nearly stops in the air. God did not simply swat the airborne example of human hubris from the air. Something else did. We probably need to know how before we can determine what.

So what transitions and energy transformations can take place that would effect these scenarios? How might these energy transitions take place? And for each such transformation "why" might it take place?

We have a set of working presumptions based on the conditions of pieces found. The plane hit the water with "substantial" (how substantial?) vertical velocity and modest forward velocity with a vector roughly along the projection of the axis of the plane's body on the water.

One recurrent question never really answered for that theme is "how do you reconcile the observed fracture patterns with the VS breaking off in flight?" The sideways or even torque type stresses that the VS sees in flight would produce a quite different fracture pattern than the one observed. This is one example of what I mean by "how might these energy transitions take place?"

GreatBear's scenario needs the transitions explained. The data record gives him three tools to work with. Two are related to the communications outage. Was it loss of line of sight from antenna to satellite (body of the plane in the way) or loss of power later made up by restarting engines or the emergency system kicking in. The third tool is his imagination and experience.

In his scenario we have a really big and modestly breakable object containing hundreds of really breakable objects broken in a specific pattern that must come to a near stop along its former track almost instantly and converting it to a much lower velocity off to 90 degrees from its flight path.

mm43 had the same problem. He posted his banked turns graphs which essentially posit controlled turns rather than jet fighter aerobatics. It's a start. It does show a way to account for the forward velocity bleeding off in a slightly more gradual way with the plane ending up off to one side of its former track by around 90 degrees.

Presume first that the plane is not decelerated at more than about 3g. Much more than that would potentially cause things to break making everything in the analysis more difficult. Start simple work in complications later only if needed.

This may be a little over the top, mind you. The damage to the recovered corpses, perhaps people not strapped in, suggests they were not subjected to being thrown around the cabin so much as thrown to the floor of the cabin while in sitting positions. (At least that is my understanding.) Violent maneuvers that satisfy this condition seem questionable. Perhaps 1/2g to 1g is a better limit.

This isn't decrying GreatBear's analysis. It's simply requesting that energy and the energy state transitions be considered, too.

[henra]

Question: How erratic could the flight path be back to LKP? All I have seen to date is MM43's nice bank angle arc analyses, so this seems like a new proposition to me

Very good question!

theoretically it could be very erratic. Depends really on the mode how and why it came down. If we had a radar track of the final 5 minutes we might already have a much better idea what happened (besides obviously being able to locate the wreckage much easier).
However, assuming control of the plane got lost at altitude, keeping a defined flight path was probably rather low on the agenda of the pilots.

JD-EE:
This energy loss is really what is most mysterious to me. As you said 210t with almost 500kts at 35000ft is a HUGE amount of emergy. This obviously has been dissipated within only 5min to a very low energy state.
(maybe 100 -150kts at 0ft). Doing this without breaking up the aircraft isn't easy.

[HazelNuts39]

some numerical constraints

Originally Posted by henra

210t with almost 500kts at 35000ft is a HUGE amount of emergy. This obviously has been dissipated within only 5min to a very low energy state.

The kinetic energy difference between 475 and 200 kTAS is equivalent to 8230 ft height. The airplane has lost the energy equivalent of 35000 + 8230 ft height in 5 minutes, i.e. at an average rate of 8650 ft/min. In my opinion that is quite high, but isn't necessarily associated with disintegration in-flight. On the other hand, it doesn't leave much room for continuation of cruise flight, in other words it suggests LOC very quickly after 2:10.

To end up close to or south of LKP the airplane must have changed direction. That requires a force. 3 g requires a force of 3 times the airplane's weight. At FL350, M.8 the airplane hits the buffet boundary at 1.65 g, corresponding to a bank angle of 53 degrees, rate of turn 3 degrees per second, radius of turn 2.5 nm.

regards,
HN39

[Machinbird]

grity

I think, a line based on a polynomial regression, is a little bit to simple for a 3D problem

Grity, I think we realize that as a first order linear solution to a complex (chaotic) problem, there will be a lot of inaccuracy and uncertainty remaining, however given the choice of 6 days uncertainty and 30 hours uncertainty, reason says the one with the significantly shorter period of uncertainty will have much greater accuracy. The key question is, is the SAR detected slick relevant?
If the slick was not relevant, I would still expect another potentially relevant slick to show on the same SAR image taken 30 hours post accident. With an airframe that impacted with as much energy as AF447, I would expect tanks to rupture and the fuel to promptly be on the surface after impact, with a further trail of fuel coming up from the wreckage with a relatively short half life (perhaps 10 minutes)
FluidFlow made a number of simplifications from the drift group's methods, so a methodology that re-integrates those methods can't help but be yet more accurate.
The choice of 22 bodies with almost common recovery times as a data subset for analysis is a good idea. Uncertainties in this group come from initial depth of insertion in the water (3D flow vectors) , time until bodies reach the surface and begin moving with the surface flow vectors, and value of the resultant aerodynamic/hydrodynamic drag coefficients as bodies begin to float progressively higher in the water.
Hopefully consideration of these various factors will make only small adjustments in the X marks the spot prediction. If so, that location will be ready for the next search phase to check out.

[henra]

HazelNuts39:

The loss of kinetic energy between 475 and 200 kTAS is equivalent to 8230 ft height. The airplane has lost the energy equivalent of 35000 + 8230 ft height in 5 minutes, i.e. at an average rate of 8650 ft/min.

If the forward velocity had reduced to 50 kts, the equivalent altitude loss would be ~9850 ft and overall equivalent potential energy loss ~45000 ft.
Equivalent speed would be 1010 kts.
regarding linear deceleration that would equal 1,73 m/s^2 linear deceleration. That requires significant drag.
More than what an aircraft in mormal configuration in straight and level flight would achieve.

So, regarding energy state the final horizontal velocity doesn't make that much of a difference, i.e. we have a good idea of the medium energy dissipation over that time span.

/Edit:
As a comparison, to give an idea of the level of deceleration: The average acceleration during takeoff and landing is roughly comparable: 150kts in 40-50s ~1,75 m/s^2 (45s). To achieve that by aerodynamic drag alone wouldn't be easy