Posts by user "Pilot DAR" [Posts: 55 Total up-votes: 178 Page: 3 of 3]¶

Our member WillowRun 6-3 has drawn our attention to the following:

Very significant reporting by WSJ about U.S. - India friction in the investigation.
https://www.wsj.com/business/airline...hare_permalink

Reminder for everyone for this thread, because of the high potential for confusion in this discussion, please write out "Air India" or "artificial intelligence" as you intend, so we know which one you intended...

And, please be factual, rather than speculative in your post, non factual information does not contribute to our discussion...

Thanks...

Not necessarily. I do not speak for how accident investigators do their job, and what information they choose to make public or not, but I can say that: Everything to be publicly reported will be in the public domain. That does not mean it is now, and does not mean that everything will be reported. (Your eyebrows raise....). If the NTSB investigator team does pages of mathematical analysis to make a determination, i could be that they report the determination, though not the pages of math. I can opine this based upon having seen pages of analysis to support a report (for which I was an "official" commenter), where I saw the math, and I know the four pages of comments I provided, but those did not appear in the public report, just the summary.

So, be patient for the final report, and expect to see an appropriate amount of supporting information. If, after the final report has been issued, your qualified and experienced opinion is that the NTSB has been:

As a citizen of the nation producing the report, there is probably a path to follow.

Is nice when it can be done, it cannot always be done. If a citizen group feels that there is an unreasonable risk because one operation is close to another, there are umpteen government levels to talk to in respect of zoning and building permits. I know for a fact that in some cases, private airport operators have been required to purchase adjacent land, to assure it remains a clearway. A citizen group can always approach their local government to pursue that. If that does not come forefront of land use consideration, it wouldn't be the first time....

We mods are entirely happy to keep a thread open, as long as the posts are productive, and the topic is not becoming a repetitive vortex. One really good way to prevent this, is before anyone posts their thought, read/search back through the thread to see if that thought has already been discussed, and perhaps put to rest. New posts, saying not really anything new ('cause it's been posted before) begin thread closing thinking among the moderator team. As any thread, this one will remain open, unless there seems to be a need to close it for quiet time....

Let's recall what the AI-171 preliminary accident report said:

That's caused by a person's hand, not a bit flip affecting a circuit board. Other automated actions (RAT deployment) after that seemed to operate as intended. Sure, bit flips are a thing, but not this thing...

Yes, this thread is a discussion of information published in the reports, preliminary, and final, when it comes. Let's not get wandering off on tangent and speculative ideas about the cause of the accident. If, upon the release of the full report, there are unanswered questions (which happens), we'll discuss then. We are aviation professionals, we do the job within our skills and qualifications. Lets extend that respect to the investigation team also, and consider what they report...

Well, two things about this: In the Twin Otter, for example, operation of the autofeather to feather the first propeller on a failed engine, then disables the second autofeather.. I can't delete a file on my computer without having to click that "I'm sure". I would have to think that the airplane system to shut down the engines could have a check in it, that after manually shutting down one engine then triggers a system which prevents the second engine being shut down unless a pilot does a second and distinct action to enable it. In amongst those actions could be yet another where the "other pilot" could click something out of reach of the other pilot which prevented the actuation of the second shutdown switch.

From training I have done with other pilots, I have discouraged "fast fingers" so I'm nearly always content if there is a "Identify, confirm, (agree with the other pilot) then act" process before critical controls are operated. That couple of seconds to be sure and correct is nearly always worth the short wait. I think in terms of the second of two engine being shut down, it's worth waiting the extra couple of seconds to get it right!

If you need to shut down an engine (for fire or vibration) you need it shutting down in ten seconds, not one second. If the primary (switch, in the case of the 787) fails to shut down the engine, I would expect that there are secondary means, which are effective, just not as quick. Pulling a fire handle would isolate the engine, and shut it down, or selective the fuel off to that engine with fuel system controls? Nothing is perfect about any of this, and anything can fail, but this does seem to be an opportunity to design in a "pause"/"second sober thought" in the next, and much more critical action.

The Twin Otter's protection with autofeather seems to work well enough with that in mind. If a second engine needs to be feathered, the pilot must do it with the propeller lever, the auto system will not do it.

I'm okay leaving the thread open, as long as discussion while we await authoritative information contributes educationally, and does not speculate, nor dredge up topics which we have already put to rest (unless a formal report reintroduces them). I would rather a thread be open, unless it's becoming a hamsterwheel...

It's going around, 'cause it's not going anywhere....

I think the much more persuasive indicator of the switches going to the open state was the engines stopping, and the crew member commenting on the other crew member having selected them open.

Having operated this type of switch for about a half century, I am unconvinced that a properly selected switch could "move towards the cutoff position".

From post 166 of this thread:

Perhaps the switch could be defective, but that is (a) easily detectable by the pilot) and (b) still not likely to result in it moving on its own, rather just not locking well in the selected position. This is a situation where readers need to apply their understanding before accepting an unsubstantiated report. If we receive an authoritative report about how this switch "moved toward cutoff" twice on its own, I will read with great interest!

Let's remember that totally independent of the external pull the toggle to unlock mechanism, internally, the switch still has the original over center mechanism which will spring the switch and contact to one extreme, or the other. Irrespective of the secondary locking feature, all quality toggle switches will spring to the intended position to prevent the switch contacts from resting just or just short of contact, and possibly arching internally.

Yes, if defective, the pull part of the toggle can rotate, and then the motion of the toggle will be abnormal. This would be entirely detectable in the moment by a pilot familiar with the operation of the switch. I see one of three situations here: The switch would operate properly, and the report is not accurate, the switch was operable, but the locking part of the toggle was not moving correctly (so the switch was defective), ans someone was satisfied that once positioned to run, it would remain there safely (suitably qualified mechanic, I hope, or the switch was entirely defective, so the flight could not depart until the switch was replaced. All three of these conditions are very easy for the pilots to understand. One does not require maintenance activity.

All of that said, I see this as peripheral at best to the Air India 171 crash. The preliminary report tells of both fuel cut off switches being found in the run position, and states that they were both moved from run to cutoff after takeoff within a second or so of each other, and then back to run. Nothing authoritative I have read so far from the Air India 171 crash suggests that either one of the fuel cutoff switches were defective. Indeed, the events of the accident suggest that when operated, they functioned exactly as intended! Their being found in the run position removes doubt that they (the locking feature) were operating properly. I think that the report of today, if credible, is unrelated to the 171 crash in causal information.

I'm not clear on what an "incorrect direction" would be in this context. Does that imply that when the switch was in the RUN position pushing it toward run caused it to easily move toward CUTOFF?

This requires more explanation to make sense.

Hmmm... Were it to have been a circuit breaker, how many times would a reset attempt be acceptable practice? But ultimately, it latched correctly in RUN. But then someone evaluated the switch locking, and reported as quoted above? This is all very slushy.

Though this discussion being about the RUN/CUTOFF switch(s) of a 787 being accused of improper operation, does this really relate to the Air India 171 crash? The reported information has already stated that both switches were selected from RUN to CUTOFF within a second, and then back to RUN, with a pilot remark about that action. It is a statistical infinity that on 171 both switches would fail into the cutoff position by themselves within a second, and there not be a pilot remark about that in the CVR, when there was a remark "Why did you do that?" associated with the switches.

I have moved a lot of switches and other controls in my flying career. In each case, I made that action with the intent that doing so would cause the desired action of a system. Moving the control itself was a means to an end. So, in moving the control, I determined that the ultimate objective was achieved. If there was a "feel" or locking system associated with the motion of the switch/control, then I would assess that too - was it working as expected? Like a flap or landing gear selector, I don't just push it toward the other direction, and hope that it pops into the desired position, I actually move the control the whole way, and confirm by feel/sight/locking device, that it is where I intend it to be, and will stay. Let's not forget that very basic expectation of piloting! Yes, there is the phrase "throw the switch", but that is for Igor. We pilots will move the switch the whole way, and assure that it got where it was going and is going to stay!

Both of these topics relate to the switch(es) in the 787, but I think that the similarity ends there.

I read a report of a problem. I'm not sure that I actually see a report that there truly was a problem.

Preceding a "however" This was reported:

If a "however" is written after that statement, it is either meaningless and to be ignored, or it has the effect of negating the statement. There is no "partly airworthy" for these switches. It can't be both statements. Thus, my doubt about the credibility/applicability of this whole situation in the context of the 171 crash, subject to these two statements being reconciled with each other.

We've had some good discussion about engine switches, and I think we've covered the topic very thoroughly. If any doubt remains about how these switches function, or could malfunction, I suggest that the curious person (a) go back and read all of the posts about switches in this thread, and the previous thread about the 171 crash. If that doesn't cover it, go and find one of the switches, and understand it in hand. They are very common in aircraft application, and certainly other non aviation applications. The posters here asserting a position of knowledge about the switches are probably very familiar with them, having used them in airplanes for years. If you are not familiar with the nature of operation of these switches, continued posting here is not going to help, all the information is here now.

We'll come back to this topic when there is new authoritative information to discuss. I'll leave the thread open for a bit for any "closing this phase" thoughts, so you're not feeling truncated, but let's end the hamsterwheel for now, or a moderator will close the thread for a while. I have started a general discussion on cockpit controls and switches in tech log. Design & operation discussion there, 171 accident discussion here - when there is something new and authoritative to discuss.