"EPR increases while N1 remains stable. Read the Swiftair MD-83 accident report for full details: https://reports.aviation-safety.net/...D83_EC-LTV.pdf"

I think Nik was referring to the P3 pressure pipe on the engine not the P2 sensor on the nose cone

PT2 IS on the engine, located in the middle of the inlet cone on the JT8-D (a small tube at the side of the inlet on many other engine models). It was mis-translated from the original French in the report. PT3 is between the LPC and the HPC if my memory is correct. Same thing happened with Air Florida...

For the CFM 56

The PS3 static pressure pick-up is located on the combustion case, at the 9 o’clock position, between two fuel nozzles.

nikplane

I can't keep up with the responses to this question so far

Sticking to this engine model can you precise the reason for your question as well as the location of the sensor you are talking about and of course what possible reason for blocking?

Lomapaseo, unless I'm mistaken Nik was alluding to the fact a blocked P3 pipe could dramatically effect the engine indications and basically that may be a factor, if indeed the engine had anything to do with this event.

On the CFM 56 the P3 pipe connects from the side of the combustion case at the 9 o'clock position, its a formed metal pipe, and gives HPC discharge pressure to the ECU, this enters BOTH channels on the ECU and then is converted to by the MEC/HMU to give movement to the valves in the fuel control unit and hence shaft speed. These P3 tubes are metal and have a drain hole at the lowest point, normally only water would collect and be drained away. ice perhaps but unlikely in normal conditions. The P3 is just one of 5 pressure sensors and does not work in isolation from other signals which are used by the ECU.
If for some unknown reason the P3 line was blocked the ECU may not be able to compute a model to adjust the fuel flow within the range from Surge to Flame out and indications would be erratic, but erratic engine indications are something we are trained for in the sim, and assuming one engine is running "normally" there is always comparison to be drawn to assist diagnosis. When unusual parameters are observed they should be entered in the comments section on the TLB to assist the engineers. One presumes the previously reported A/T malfunctions would have been investigated

Outstanding clarification (even it' not what Nik meant :))

But working with your explanation could anybody suggest why this unique minor failure mode has crept into this discussion thread?. Where is the relevancy in findings to-date?

GUYS!!!
Stop and hold. You're talking about the FADEC CFM56-7 on the NG. This was a 737-500 which has the CFM56-3!!! It's NOT FADEC! There is no FCU, P3 feeds directly into the hydromechanical control.
Lets get back to the scenario at hand.

Even by pprune standards, this is the most ridiculously specific speculation based on the scarce available factual evidence

I can feel the armchair seat restraints tightening at this comment

Er..... EPR?.... that may suit a -100/200 cuz they had JT8s.. The CFM-56-3 on the B737-500 has N1, EGT N2 etc.... no EPR. One of the nice things of using N1 for power is that it avoids an erroneous EPR indication, like Air Florida 90.

Icarus2001

Reuters:-

On Jan 19th 2021 the KNKT reported that the FDR contained 330 parameters of a total of 18 flights. The KNKT have already developed a general picture of the accident, however, need further data from the cockpit voice recorder before publishing any information. Inspections of Boeing 737-300, 737-400 and 737-500 in Indonesia have been ordered.

On Jan 22nd 2021 the KNKT confirmed a preliminary report will be released within 30 days after the accident (until Feb 8th 2021), no other information has yet been released.

On Jan 23rd 2021 the KNKT chairman stated, that the tech log of the aircraft contained an entry about an autothrottle fault a few days before the crash. There has been no other technical anomaly noted. At this time it is entirely unclear whether or not problems with the autothrottle contributed to the crash Crews can easily fly the aircraft manually without the autothrottle system. The CVR will be needed to hear the discussions between the pilots to identify the actual problem they had at hand. (Editorial note: obviously this entire statement came in response to speculations raised by a number of media globally attributing the crash to a malfunction of the autothrottle system).

On Feb 2nd 2021 the KNKT chairman said in a press conference, that reports distributed by western media about a possible autothrottle malfunction causing asymmetric thrust are wrong. However, the KNKT sent 5 pieces of debris, including the autothrottle unit (but not identifying the other parts), to the USA and UK for further examination stating they want to find out why an autothrottle parameter changed. He re-iterated they don't know why that parameter changed and need confirmation from the parts sent to the USA and UK and the CVR. The maritime search for the CVR is still ongoing.

zero/zero

I think you are right. Why do we need to discuss things like "its a formed metal pipe, and gives HPC discharge pressure to the ECU,......." when the internal workings of a jet engine have nothing to do with the conversation and in any event there is an excellent Engineering Forum on PPrune for those who love the granular construction of turbine blades etc. My point actually is that an engine failure or an auto throttle failure are non events to a trained pilot. So either it was something else, or we have a case of something going wrong with something like pilot training. Not the pilots you see. No pilot sets out to do anything but his or her best. Retired Guy

It's quite normal to answer a question when asked, doesn't mean we are agreeing with the rumoured causes of the event.. on the other hand stating the obvious Can we agree, as the aircraft crashed it was in fact, "something else"!

fdr

It's ok to talk about EPR if you describe what is going on inside an engine. However if you're talking about what a pilots sees then you need to keep to the actual engine cockpit displays. For myself I like to keep to what a pilots sees and only divert to the actual engine operation-cycle to figure out why.

let's move on since I lost track of where we got to in the last week :)

OK, so how about this:
Can anyone say why the Indonesian KNKT has sent bits of the plane to the UK?
The French have more parts onboard a 737 (half an engine at least).
Could it be the UK is still respected. Joy in these Brexit times.

Having fewer national parts on board would presumably be seen as a good thing if an independent perspective was being sought. French involvement in the engines, versus the potential for asymmetric thrust to be a factor? Seems to me a more likely basis for such a decision than respect for any particular state.

Didn’t Smiths Industries make the auto-throttle computers for the classic. If the auto throttle system is being looked at would those be the parts being sent to the U.K. ? That’s total speculation, btw.

Interesting snippet of information released by AVH today. Last communication with aircraft was request for heading 075 to deviate around weather, which was granted. Crew acknowledged clearance. Subsequently a/c was observed turning left on radar instead of right to 075, did not respond to ATC query.

General Electric bought Smiths years ago and renamed them GE Aviation Systems. They're currently headquartered in Cincinnati OH. Not familiar with where the auto-throttle folks ended up, but one former Smiths division that I know of has been resold once and relocated twice since the GE acquisition, and is currently located in Ft Collins, Colorado. They had initially started as part of Dowty-Rotol.

Parts being sent to US
 

https://www.thejakartapost.com/news/...or-checks.html

some possibility that parts like the engine FCU may be examined for mechanical and operation issues (sticky valves ?)

Some times there are issues about the fuel itself and the best place to get a sample is from the undamaged innards of the FCU on an engine. Any issues involving the FCU include before disassembly x-rays of where the valves are set.

I guess we'll just have to wait for an interim report before second guessing why they are sending parts to the US

Translation of a report about remarks made by Chairman of the KNKT Soerjanto Tjahjono on 2021.02.03.
I have seen other translated reports of the Chairman's statement that repeat the quote about the CVR. I was unaware that the CVR memory module had been found and I wonder whether there is an error in quotation or misstatement by the Chairman (CVR vs FDR).

I haven't seen any reference to that before - up to now, the received wisdom seems to have been that it was a problem that stopped the climb at FL110, rather than an ATC instruction.

Regarding status of the CVR. A different report has a subtle variation in the Chairman's remarks and indicates the search for the CVR continues:

Tempo.Co (English Version)
Editor: Laila Afifa
2021.02.04
.
... a maritime search is going on for the cockpit voice recorder’s (CVR) memory unit that Tjahjono said would help explain any human factors behind the crash.

“If we only have the FDR, we do not know why the parameter changed, what was the reason,” he said of the autothrottle. “We need confirmation from the components that we sent to the US and UK and the CVR.”

This was published on AVH two days ago, but apparently preoccupied with the autothrottles nobody here picked it up. The information is clearly from ATC tapes, not CVR.

no it’s not Herod. Except yaw damper.

Cessnapete
You’re not missing anything. It’s not the A/T nor the A/P that caused this crash. It’s a 737. It requires no automatic systems to fly perfectly well. It requires - uniquely today- no hydraulic power nor
AC electrical power either. So as some have said it must be something else. Pilot training perhaps? Something catastrophic? But an engine failure, AP or AT fail don’t cut it with correctly trained pilots.
CVR missing is a real problem. But they already know what happened. Just not quite what conversation took place aboard during the event

Have we reached the point where when "nice to have" automatic systems fail there will be an inevitable crash. Amazing.

We've already passed that point, where ,poorly trained pilots are unable to throw away the automatics and manually fly while sorting out malfunctions in a an emergency. Or a trained in mindset that manual flight is an emergency.
Or blind faith in the automatics causing the accident.
I.e. DXB EK B777. Press TOGA, no monitoring that it works, as no hands on A/T during a critical phase of flight. A B777 doesn't Go Around well at idle power! Simple piloting skills lacking, or that disappearing “skill” Airmanship.

Peculiar term
 

"The pilot, he said, then received a traffic guide's instruction to fly at an altitude of 11,000 feet."

TCAS ?

Teddy Robinson:

From the translation supplied in a previous post:

" Chairman of KNKT Soerjanto Tjahjono explained that after takeoff, pilot Captain Afwan had asked the air traffic guide to change direction to avoid bad weather

The pilot, he said, then received a traffic guide's instruction to fly at an altitude of 11,000 feet."

(My emphasis) The "traffic guide" apparently means ATC, not TCAS.

Euclideanplane

Apparently the bahasa Indonesia for ATC is: pemandu lalu lintas udara. lalu lintas means "traffic," and udara "air": pemandu is translated in an online dictionary as "guide", "leader", or "wizard". The name, then, respects the captain's authority whilst praising the skill of ATC.

I mentioned a few weeks ago that the aircraft could have levelled off at 11000’. In fact one of the departures has a level off altitude of 11k. (Different Dep to this one though).
FR does actually show a reduction of V/S from around off 10k.

my post was deleted for whatever reason.

Lets run with the A/T malfunction for a moment.
Previous poster said this has never happened in the climb. However, the A/T clutch etc, could fail at anytime. In the climb this would be no big deal until top of climb when the power comes off. At high altitude and assuming M.72 for the cruise, one TL only retards and the other is stuck. The TL will retard to maintain the selected or managed speed. At high altitude this should not present a problem with the Autopilot as the thrust reduction and subsequent yaw would not be great enough to disconnect it. The pilots should have enough time to figure out what is going on.
However at 10000’ the autopilot would not be able to cope, without any intervention from the pilots the aircraft would quickly roll and enter a spiral dive.

What was the trigger that caused the upset, assuming A/T failure?

On this flight the aircraft was actually cleared to FL290, (from YouTube). There should not have been any level off. The only indication is the few seconds at 10900’ for whatever reason. Question? Did the pilots have a selected altitude of 11000’ in the window.

retired guy

Correct.
An instructor where I work does an exercise in the sim. where you are taken to 30,000 ft and every unguarded switch on the overhead panel is switched off.
The aircraft still flies.
But this incident , and others ( Ethiopian , Lion etc. ) make me think that the days of relying on the manipulative skills of pilots are over.
Airbus philosophy has won.

James7

I don't understand why it's no big deal to have split throttles in climb. Arguably the pilot response time required may vary, but most of the accidents I've seen at various power conditions were because the pilot did not understand or detect what was going on

I think he means that, in the scenario he is describing, the thrust levers would not split until the autothrottle tried to move the levers back at top of climb or interruption of the climb.

Which pretty much eliminates the 'catastrophic' scenario.

lomapaseo

quite right Dave, problem would only arise at level off. aircraft levelling off and possibly distracted by making a turn, easy to miss that one throttle is stuck in the climb position. Even easier if the turn was to the reducing engine. This would also exacerbate the induced roll and yaw caused by the opposite engine on climb power.

Just to be clear, this would be the scenario.
assuming right throttle stuck in climb. Aircraft starts to level off, only the left throttle would retard. Pilot makes a turn to the left, he may not notice the right throttle is stuck. Aircraft makes the turn, pilot thinks all is ok as the engine would cause roll and yaw to the left. It would only be after a few seconds or so that things would start to get interesting, then the aircraft would quickly enter an upset. With the aircraft in IMC, pilot not fully up to speed because of the pandemic. The pilot would have to be pretty quick to make the corrections. Recovering from this type of 'powered’ upset, only relying on instruments, requires training. Unless the throttles were closed the aircraft would never recover. - Spin recovery technique.

The Captain was more used to flying the NG by all accounts. A/t system is much more reliable, probably not paying much attention to the throttles.

I was a captain on the NG and classic for many years and from a pilot's point of view I cannot say that there was any obvious difference in the autothrottle operation. I can remember only a handful of flights with the auto throttle u/s, so manual thrust lever operation in all phases of flight was a pretty rare occurrence and required some thought. Once you got your eye in so to speak it was simple. But it is not a mode everybody spends a lot of time practising on the line. Obviously you train a lot of the time in the sim without the auto throttle, but then again that is mainly one engine out and the aircraft does things more sedately. But it is frankly not the sort of failure that you would expect to cause anyone a lot of problems.

James7

Am conflicted by the wording. The scenario would go close to the upset that occurs here. If the APLT trips, the crew are going to be further behind the aircraft when they start putting in any corrections, assuming that they recognize that the problem is an actual roll event, not an instrument error, lots of accidents that the crew has assumed it is an instrument, and as many where the crew have followed flagged info or failed info without flags.

If the trigger event is the split thrust lever condition, with one at climb power, then recovery would still be achievable up to some point in the dive with the thrust lever remaining where it is, Optimal is getting rid of the asymmetry by going to idle, most times a safe bet, or increasing thrust on the other engine. coming back to idle is the fastest way of achieving a balanced known thrust, acceleration is slower at lower RPMs, and thrust output change for a change in RPM is greater at higher RPM. grabbing the levers and bringing them back to idle is not complicated. by omitting to disconnect the ATR. But, in any case, until well into the dive, the yaw and roll authority that remains even with an asymmetry is able to be countered. Not ideal, but possible.

Any time an aircraft is out of control, the action of closing the throttles and neutralizing the controls will result in the aircraft stabilizing in a dive, either right way up, or upside down, but in a dive. There are one or two notable exceptions to that, but historically crew response to a surprise upset is to exacerbate the problem.

Spinning or autorotation, in this case, is not needed but could happen, if the crew had added large amounts of elevator and stalled with the yaw still happening. That would make for a tough recovery in IMC or broken visual horizon conditions. That needs a 3 1/2 to 4 g pull to happen, which could be possible, but it is not needed.

USAir427 may have benefited from hands-off. UAL585 probably not. AF447, certainly would have. Perpignan, nope. CAL Nagoya, Nope/maybe, CAL TPE, yes, Soichi, yes, The B717 stall test pull though, yup.

Years back, checking an examiner in an NG sim, the crew he was instructing was given a microburst scenario, off-axis entry. the aircraft flicked upside down and faceplanted. Repeat did the same thing. They called a coffee break and called in the techs. I flew the same scenario, and if the aircraft was not loaded up, it flew through the microburst happily. Alternatively, pulling hard enough to stall led to autorotation due to the yaw that was coming from the off-axis entry. I flew the QTG for the sim at the same time, and that matched nicely in all respects. The sim wasn't faulty, the crew's awareness of the conditions had holes. Flying that manoeuver essentially hands-off, the microburst was able to be negotiated without a face plant.

Autorotation, off axis entry and flying through a microburst hands off?

You are better than me, I don’t have a clue what you are talking about. Aeroplanes, helicopters or holding patterns?