China Airlines B737 Fire at Okinawa

[Return to base]

I am afraid that the current method of producing Designers/Engineers for industry means that the "old-fashioned" 5 year UK apprenticeship no longer applies and the current design staff have no idea on how to manufacture or assemble items. This is why people like Airbus have to have a third party to ensure that everything will go together in an orderly fashion. I know of designers who never get onto the shop floor so there is no way they will ever gain experience as to what is happening to their design


rtb.

[PBL]

Sunfish gives an excellent intro to FMEA for non-experts. lomapaseo then goes deeper into it

Originally Posted by lomapaseo

I doubt that Boeing assumed a level of "catastrophe" for this error in their FMEA and under what FAR would they perform this type of analysis for such an error....... ?

Good question. Let me say a little more about FMEA.

An FMEA requires you to identify
* everything that can go wrong (all ways in which things can fail, or "failure modes")
* what the possible causal consequences of such a failure can be (and one failure mode might have many different possible consequences)

but it doesn't stop there. You then

* classify the possible consequences into four or five categories. These categories include one typically called "catastrophic" (the worst: say, leading to loss of the airplane) and one called "minor" (the least bad).

and then

* perform a risk analysis: how often are the failure modes likely to appear; given that a failure mode has manifested, how likely is it that a certain effect (from the many that may follow this failure) is exhibited?

This last is not strictly part of an FMEA, but inevitably follows.

There are then requirements that say how often a particular effect is allowed to appear. For "catastrophic" that would be "not expected within the service life of the fleet". An applicant for an airworthiness certificate must show that the failures in their respective categories satisfy these probability requirements. See LLoyd and Tye, Systematic Safety, CAA Publications 1982, for the most concise explanation of categories, classification and risk classification.

So lomapaseo's suggestion requires some refinement. The bolt separating itself from the assembly would likely be a failure mode in the FMEA. That itself would not be assigned a severity category, I don't think, because it is a failure mode, not an effect. The question is then what effects it might engender. And at least one of them turns out (we now know) to be catastrophic. The question would be whether this effect was foreseen in the FMEA.

PBL

[lomapaseo]

PBL

So lomapaseo's suggestion requires some refinement. The bolt separating itself from the assembly would likely be a failure mode in the FMEA. That itself would not be assigned a severity category, I don't think, because it is a failure mode, not an effect. The question is then what effects it might engender. And at least one of them turns out (we now know) to be catastrophic. The question would be whether this effect was foreseen in the FMEA.

yup, you got it

So how can you be expected to assess a catastrophe under the FAR/Jars when the flight is over? I doubt that there is adequate guidance on this.

Of course under continued airworthiness, the FEDs can now react,

[PBL]

A colleague in Japan points to the news item
http://www3.nhk.or.jp/news/2007/08/3...830000144.html
in Japanese, which includes the following.

Air Nippon has found one of its fleet of 13 B737s was missing a washer from a slat bolt. (Apparently the english term "washer" is explicitly used.) The aircraft had been acquired directly from Boeing and entered service in January 2007. The area concerned in the recent inspection AD had not yet been either maintained or inspected.

So a manufacturing fault is suspected.

This from the news item. I have no Japanese myself, so have not read it.

PBL

[barit1]

sunfish

(runs for foxhole)

Wait! I totally agree!
The problem with the FMEA process is that - for a time at least - FMEA's were written by staff people who had relatively little detailed knowledge of the "workings" of the system and hardware. This came to light in the ASA Embraer Brasilia accident (NTSB DCA91MA033) in which a prop failed to fine pitch, and the drag and turbulent airflow led to an uncontrollable roll. (The ERAU website had a copy of the complete report, but I'm unable to locate it right now.)
In this case, the FMEA predicted a failure to coarse (high) pitch, and the FAA cert. staff accepted this analysis. Neither the manufacturer nor FAA personnel seemed to be aware of several decades of service experience with controllable propellers strongly indicating the opposite.
Sometimes we have to relearn lessons the hard way.

Just found this thread; In it Cardinal discusses this very Embraer accident.

[FrequentSLF]

Took today the CI from TPE to HAN. Schedule a/c was a 737-800. At a gate found a A330-300. Number of passengers was in accordance to a 737-800.

[Sunfish]

Barit1, thank you for your advice and the link.

[Volume]

For the record, maintenance protocols for components relate to the severity of failure and the ease of inspection to determine the components condition.
This results in components being classified:

"Hard Time" - set limits on number of hours/cycles/stress history/ etc, which is what you do if there is no way you can gauge the condition of the item by simple inspection - for example disks.

"Condition Monitored" - you inspect the part at set intervals and if its within limits, thats OK. e.g. blades. flap tracks etc. I assume the end stop will be disassembled and inspected in the process of overhauling the LE slat.

"On Condition" - when it breaks, fix it, and not before. e.g seat jacks.

This is the "old style" of maintenance program development. Today we work with the so called "reliability based maintenance", based on the MSG-3 document. (you may google for those key words)
Nevertheless it depends on the skill and mindset of the engineers to decide which failures could occur, before the failure effects can be analysed and an according maintenance task can be developed.
Unfortunately today the beancounters make the wrong conclusion : low number of inspection tasks = low costs = good maintenance program. There is huge pressure on the engineers to come out with a low number of inspection tasks when developing the scheduled maintenance program. One key question is always : is the task economical, does the risk of failure and the severity of the failure effect justify the costs of a task.
If you try to inspect every nut and bolt of an aircraft, you will never be economical. So you have to build on past experience, and you will always have a first for every failure to occur, afterwards (depending on failure effect) it will be taken into account.
You can be absolutely sure, that in 10 years every new developed maintenance program will include the slat track downstop.

[PAXboy]

Volume's concern about cost control does not surprise. But, if the new schedules will include the slat track downstop, does that mean that other items will be reduced in the schedule - as they prove themselves to be more reliable?

[Sunfish]

Paxboy:

does that mean that other items will be reduced in the schedule - as they prove themselves to be more reliable?

Sadly, yes, but they shouldn't be.

I refer you to Richard Feynman's Appendix to the report on the Challenger disaster.

http://science.ksc.nasa.gov/shuttle/...Appendix-F.txt

[PAXboy]

Wow! Thanks Sunfish, that was a fascinating (if somewhat tricky) read. But it was well worth seeing the layers being peeled back. What one then finds is, of course, a system designed, built and operated by human beings and we know that humans are very good at fooling themselves. For the most part we get away with it but, Feynman's closing words in that report are superb:

For a successful technology, reality must take precedence over public relations, for nature cannot be fooled.

[barit1]

sunfish - and others interested in the failure of the FMEA process re prop overspeed: Here is the link.

[181.filo]

blaming business!!

What happened to those pilots "know all" blamed CAL crew/mtx as soon as the news came on the CNN.? Why dont we hear from them anymore.
I hope that teaches us the old saying "DONT JUMP IN TO CONCLUSION UNTIL...
Whatch for those loose BOLTS.
happy landings to all my mates.!!!

[Guava Tree]

CAL Maintenance Versus Boeing ?

Hi Filo,
So you do despise anyone who might know it all , or at least know more than you do.
Are you a “Company Man” ?
I must admit that I have been surprised about how Boeing can hire assembly workers with such minimum knowledge and intellect but still train them enough (need to know) to get the job done.But history has proved that, for Boeing, this works.
I talked about Guanxi.I still talk about Guanxi A washer is easy to disappear. Are Nippon engineeers checked for isolated washers, gold and diamonds etcetera when they clock off?
I think we need to wait longer to see if worldwide inspections reveal a significant number of such cases.
Meanwhile with some knowledge about both Boeing staff and CAL maintenance staff, but not “know it all”, I would place my money bet for Boeing and against CAL Maintenance.
How come this CAL aircraft operated for quite a long time since leaving Boeing, before this problem occured ?

[Blacksheep]

Unfortunately today the beancounters make the wrong conclusion : low number of inspection tasks = low costs = good maintenance program.

One of the banes of our life is Accountants poking their noses into everything. They're not terribly good at financial matters and even less useful when trying to control that of which they know nothing. Which is everything apart from accounting. My own experience is that over enthusiastic attempts to minimise costs generally leads to much larger increases in costs downstream.

Example. We were told we had too much inventory. The auditors declared that we had a lot of stock that had never moved and supplies were not to reorder anything with a shelf-life that had expired. Finance Department runs the show and the BOD know nothing of maintenance anyway. Except that it is a cost rather than an earner. They forget that the aircraft are the production machinery that produces the product we sell and they must be kept in full operating condition to ensure a smooth production run. All profit originates from serviceable aircraft.

An aircraft became unserviceable with a no-go defect at an outstation. We had the component but new "O" ring seals are required when fitting a replacement. Yep! the "O" rings were shelf-lifed, it had expired and under the accountants' instructions they had not been re-ordered. Result? Crew ran out of duty hours. Passengers had to be rebooked on other operators or given free overnight hotels. Add knock on effects causing other delayed services and the total bill came to over $1/2 million. The "O" rings cost $20 for a pack of 5.

When counting beans the first step is ensure you have some beans to count.

Bean counters. Don't we all love 'em!

[barit1]

Guava Tree puzzles -

I must admit that I have been surprised about how Boeing can hire assembly workers with such minimum knowledge and intellect but still train them enough (need to know) to get the job done.But history has proved that, for Boeing, this works.

The difference is that in the Boeing shop, an assembler need only address a few tasks and thus he can master them fairly quickly. A maintenance shop needs versatile folks who can work competently on radomes and trim tabs and all in between.

[PAXboy]

Blacksheep a fine story of our times. Let me guess what happened next: The maintenance department told the middle management what had happened. The middle management did not tell the senior management the same story. Consequently, no lesson was learned.

Unless, of course, the maintenance department decided to start ordering key items of stock as before - despite the order to the contrary. This means that the system continued to operate normally, with the exception that some idiot in so called 'management' thinks they did a good job and received words and/or money to reward them.

Please tell me that I am just too world weary and cynical, that the company learnt the lesson and that the accountants were told to keep their nose out of things?

[lomapaseo]

PAXboy

Please tell me that I am just too world weary and cynical, that the company learnt the lesson and that the accountants were told to keep their nose out of things?

You are too world weary and cynical and you also have a good imagination to support your cynicism

[PAXboy]

lomapaseo

You are too world weary and cynical and you also have a good imagination to support your cynicism

Unfortunately, my world weariness is from 29 years of working in commerce and (for a short time) local government, as well as self employment where I got (still get) to watch a very large number of companies at close range. My 'imagination' about what happens (two posts back) is based on bitter experience of 29 years ...

[glad rag]

PAXboy

Sadly, you are on the money with your posts.