I'm probably coming into this issue a lot later than most of the folks who frequent this site, so I apologize if I'm covering old ground here. I've been interested in this accident for quite some time, but didn't do a lot of web surfing on it as there is almost never any real info available until the accident investigation is complete. Lots of hear-say and speculation, but not much info.
However, I work for Transport Canada, and now I find myself (along with many other coworkers) tasked to implement the nine new recommendations that fall within my area of responsibility. So I am studying the report, trying to better understand what happened.
I'm aware that there were many problems with the IFEN, but from a technical point of view I can't put my finger on any identified problem that could have caused an IFEN wire to arc. What have I missed?
I don't have an axe to grind either way on this issue - I just want to understand what happened. Can anyone point me to some part of the TSB report that sheds some light on why the IFEN wires arced? Or failing that, are there any rumours or speculation that might give a clue?
I do know a few of the TSB investigators, but they work in a different part of the city than I do, so I don't actually see them very often. Next time I run into one of them I'll pick his brain. They have been pretty reluctant to get into details over the past four years, but now that the report is public they should be able to provide a bit more info.
Kevin, I don't think an explanation for that was given in the Final Report. Though I certainly can't answer your question I will say that even the FAA said that the installers (Hollingsead) didn't use good wiring practices in an interview with David Evans of Air Safety Week. Also the installations were hurried & sloppy. The FAA also stated that the system was too powerful for the MD-11 to handle which I assume is why the IFEN was hooked into an essential bus rather than a cabin bus. The wiring to the entertainment system was bundled together with plastic ties with the general aircraft wiring and poorly supported. The mixing of different wiring types was discouraged by the manufacturer of the aircraft. The TSB said that they suspected that another wire was probably involved but were unable to locate it.
Here are some pictures that may interest you.
Here are a couple of interesting links to look at:
These may not give you an answer but they bring up some very alarming points. Barbara
Here is another potentially useful link:
FAA Special Certification Review Team Report on Santa Barbara Aerospace and Swissaiir MD-11 IFE System
As I understand the TSB's analysis, the lead arcing event occured at or near a non-grounded aluminum wire bundle mounting bracket. The bracket had four drill holes though which wire bundels were routed.
I think the theory is that the IFEN wire insulation must have been compromised where it was threaded through the bracket. But this, in itself, would not result in arcing, since there would be no potential to ground. So either the bracket either somehow became grounded, or another wire with a different potential charge was also breached. The latter is considered the more likely scenario.
In the report above, it is stated that inspection of IFEN installations in the remaininig Swissair fleet identified a number of non-conformities (bend radius and spacing primarily). Vic Gerden and others alluded to some deviations from "best practices," but were not specific.
It has been reported that the installation process was conducted in a great rush and there have been questions raised about SRTechnics oversight and QA of the installation work performed by Hollingsead International.
Unfortunately, the bracket itself was not recovered, and it is therefore hard to know exactly what went wrong, although one might infer that the rigorous application of best practices in the installation process might have elminiated the problem.
Engineering drawings were missing or incomplete with respect to the interface between the IFEN and the MD11 electrical system. I'm not sure but I believe the subject wiring may have been among that which had to be "improvised" during the installation process.
The TSB's investigation and report does not contain all the physical evidence needed to establish a link between the fire onboard Swissair 111 and a SPECIFIC documented deficiency in the IFEN's installation. But it does establish that IFEN wiring was involved in the lead arcing event. Given all we know about IFT's and SBA's practices, and the lack of regulatory oversight, there is more than a little circumstantial evidence that these factors contributed to the ultimate chain of events leading up to the crash.
Hope this is useful,
Barbara - Thanks for the links. Some of the issues with the way the IFEN wiring is reported to have been installed could possibly have been at play in this accident. Bending wire with pliers - that could certainly damage the insulation, leading to a bare wire. Inadequate wire support - that could lead to wires chafing against parts of the airframe, which could cause the wire to arc against the airframe. This is the sort of info I was looking for.
The other issues (no power switches in the cockpit, no mention of the system in the smoke checklists, etc) likely would not have been at play here, as the plane was doomed once that initial arc happened and the insulation ignited. These design flaws should certainly have been corrected during the approval process though.
Anyway, I've got to go back in the garage and get back to work. Strangely enough, I'm busy doing the wiring installation in a plane that I am building. I am being very particular with the way everything is installed. No Kapton either.
Mark - Thanks for the additional info. Quite illuminating.
"from a technical point of view I can't put my finger on any identified problem that could have caused an IFEN wire to arc. What have I missed?"
It has been pointed out all along that the soft Tefzel wiring insulation used for the IFEN had no business cohabiting with the much harder BMS-51 (Kapton or aromatic polyimide) - because of chafing concerns. Chafing is a mechanical process that results from the ever-present high frequency vibration of flight. It is a proven wiring menace. However, there are others. If you take a soft insulation such as Tefzel and run through it a continuous high amp loading then that wiring will heat up. If you surround it with many other wires (all heated to some extent) then you have a heat-sink effect. Pass that large wire-bundle through a restricted passage (such as a bulkhead throughway) then you also have a heat-sinked environment (with little convective cooling). How is this relevant? How does this increase its susceptibility to arcing?
Let me go on. Undergauge that wiring conductor in the IFEN wiring and the heat build-up in the wiring is that much greater. What happens to an originally soft wiring insulation when it becomes overheated? Well it becomes that much softer? Does that make it more prone to chafing? Not really, but what it does do is make it more vulnerable to cuts and nicks (such as on swarf surrounding cut-outs) and stress deformation (localized thinning) on the outside of any tight radius bends.
I will ask an expert to post here and see whether he can endorse/add/detract. It is an important aspect of understanding what went on in SR-111 to bring about that (largely now dismissed) initiating event.
Dagger Dirk - I can understand how mixing wires of different insulation types in the same wire bundle could cause problems.
You mention that the IFEN wiring may have been undergauged. Where does that info come from? There is no mention of this kind of problem in the report of the FAA SCR of the IFEN STC. Was this reported undergauging a design problem (i.e. the designer screwed up and specified too small a wire), or an installation error (i.e. the installer didn't follow the wire size called for on the drawings)?
Now here is something I'm wondering about but I admit I am probably wrong about this. I wonder if the draft report gives a little more information about that initiating event but when it was passed around to the manufacturers and they were allowed to submit comments, one of the 'involved' parties objected to it because maybe the TSB didn't have material evidence of what had happened. Or maybe even challenged the evidence that they did have.
I have commented on quite a few draft TSB reports, as they send most of them to Transport Canada for comments. They don't make changes just because someone has objected. You have to make a very strong case for a change before they will do it. Most changes that I have recommended over the years have been editorial ones to improve the clarity, or reduce the possibility of something being misunderstood. I didn't review the draft report for this accident, but some of my coworkers did. I haven't yet heard any of them say the focus of the report changed in any area between the draft and final versions.
I know quite a few of the guys at the working level of TSB (Vic Gerden and below) - they are very careful to not be swayed by pressure from outside organizations. They will listen to well founded technical arguments, but "political" pressure, threats, etc won't sway them. I can't speak to what happens above Vic Gerden's level, but I have no reason to be suspicious.
The guys who worked on this investigation basically dedicated over four years of their life to it. They spent months away from home, at great sacrifice to their families. They worked extremely hard to get to the bottom of the accident, and the issues that it highlighted. They are too proud of the work they do to allow issues to be buried, in my opinion.
Please understand that we (Barbara and I) DO NOT mean to question the integrity nor the dedication of the TSB investigative team. I had the opportunity to spend some time with the investigators in Halifax earlier this week and they are, as you say, a very impressive team.
The engineering disciplines that they routinely apply are reflected in the organization of the final report, in which factual information is (Section 1) is clearly differentiated from analysis (Section 2) and conclusions (Section 3).
Section 1.14.11 (Potential Ignition Sources) of the report is an excellent and relevant illustration of the rigor applied to establishing the factual basis on which findings are drawn (see 2.17 (Fire Development). As one would expect, Section 1 is strictly limited to that which can be directly observed. Section 2 deals with what may be inferred and it is here (and in the area of conslusions built upon the analysis) that we there is room for reasonable people to disagree.
What we call facts, based on direct observation, are those items that we all agree have a negligable (albeit non-zero) probability of being incorrect. In analysis, we are looking for inferences that can be made from the established "facts" that have a very high probability of being correct (on the same order of magnitude as generally accepted statistical inference, typically one to five percent probability of error). We submit that the investigators are conservative in their analysis - the probability of wrong inference is very low.
It is in the area of these inferences that we are naturally curious. An interested outside party - say, the manufacturer of a component potentially involved in the origin of the fire - would clearly have strong feelings on how far one might reasonably go with such analysis. As the potentially responsible party, one would certainly want the applicable standard for any conclusion to be extremely rigorous - beyond reasonable doubt. On the other hand, one who was unconcerned with blame or liability - or someone who did not stand to be adversely affected by a particular conclusion, might advocate the application of a somewhat less rigorous standard - preponderance of evidence.
We would argue that greater transparency in the process would render this a moot issue. The point is that everyone involved has a point of view, including the TSB itself, whose "bias" is toward improved future safety (not a bad thing, but there are other perspectives). We simply think that the families of the victims of the crash also represent a point of view that should be represented in the draft review process, particularly inasmuch as our interests may naturally be in opposition to those of manufacturers and operators etc.
It is good to hear from someone who has participated in past draft review processes that: (1) the draft review process did not change the focus of the report; and (2) your experience indicates that suggested changes are subject to rigorous critical review. We would simply like to have the opportunity to confirm this by having the opportunity to see for ourselves.
In fact, safety investigators are undoubtedly the transportation industry's best advocates for the public interest and it is unfortunate that sometimes their recommendations don't carry more clout in the regulatory process.
The TSB's investigation into the crash of Swissair 111 will undoubtedly be the standard against which all such future investigations will be measured. Their work represents great progress in many areas and for this, we congratulate them.
But unfortunately, for us, questions will always remain. More answers simply lead to more questions. The TSB's work has gone a long way to explaining what happened and how, but in the end we ask the more difficult question, "Why?"
The "system" failed our daughter Tara. Was this the unforseeable consequence of the incomprehensibly complex interactions of hundreds of people, all with the best of intentions? Or are there fundamental flaws in the way we promote, operate and regulate air transportation? We believe there are. And in this case, we believe certain parties have acted with anything but the best of intentions. The identification and correction of these problems will require that we look beyond the proximal cause. It will necessarily take us beyond the rigor of engineering and into the great ambiguities of politics, influence, motivation, power and money. On this subject, more to come ...
The opening post of this thread asks exactly the right question, "Why did the IFEN wire arc?" Within days of the crash, the IFEN was the object of scrutiny. Swissair shut it down on other aircraft. The certification process was investigated. The STC and SBA's DAS status were "surrendered". Other IFEN installations were surveyed and the US congress (GAO) is investigating the FAA's designee (DER) system. And in the end we learn that it was IFEN wiring that arced at the source of ignition. And you are right, the TSB's investigation does not identify the specific design flaw or best practice deficiency that led to the arc. Yet the IFEN's represents a small portion of the all the aircraft's wiring. In the context of all we know about the IFEN and its history, it's a real stretch to presume it's a mere coindicence.
So yes, we think there most be more to the story. And we will pursue every possible avenue to learn whatever we can. The draft report and comments are but a few items on a very long list of things we would like to see. For example, we would like to see the documents pertaininig to the "mistaken" suspension and re-instatement of SBA's DAS status during their work on the IFEN. We would like to see the original correspondance between FAA regulators and SBA during the certification process. The list goes on and on.
I hope this clarifies things. I think there's real value in discussion of this sort.
Speaking as Kevin Horton, private citizen, and not as an employee of Transport Canada, I can certainly understand the frustration and concern of those who lost loved ones in this accident. You wait years for the investigation to grind along, then have the other "interested parties" get to review and comment on the draft report, yet you are not allowed to see the draft report, nor the comments that are made on it, nor are you allowed to know how those comments are addressed. If I were in your shoes, I would probably be saying the same things you are.
The parties that review the report are chosen (I believe) because they have unique technical expertise that allows them to possibly identify errors or omissions in the draft report. I can understand how this could be seen to also give them a chance to affect the final wording in a way that minimizes their civil or criminal liability.
Speaking as a private citizen, I would prefer that families, or a representative of the families, was also allowed to comment on the final report. Ideally, all parties should also be told how each comment is dispositioned.
Looking at it from the accident investigators' points of view, I can understand why they want to keep the families away from the draft report. The draft report would then certainly become public knowledge, and every change in wording between the draft and final report would trigger great speculation and accusations about why there was a change in wording that was in some parties favour. If the draft report was going to receive that kind of scrutiny, the investigators would take even longer to produce it, as it would have to be as perfect as the final report now is. And they would have to "waste" valuable time defending each change in wording, rather than using their time to investigate other accidents.
I'm frankly not sure where the proper balance lies between these two opposing view points. I can honestly say that I am comfortable with the integrity of the current process, but I know that is small comfort to those who lost family members for no good reason.
quote:And this is really the issue, that the perspectives and interests of investigators and victims' families are, to a certain degree, opposed. To the extent that investigators limit the nature and scope of their inquiries to that which promotes future improvement, this divergence is inevitable. Of course we support actions that will avert future tragedies, but grief is not forward-looking. In fact it is quite the opposite.
To be specific, given what we know about the entertainment system, including what we learned from the TSB's final report, we believe there is more than enough evidence to warrant further investigation (by the appropriate authorities in the US). But such things are political. And unfortunately, so is the regulatory process through which safety improvements are mandated and implemented. We are therefore competing for media attention to our respective issues.
When Vic Gerden was asked whether there would have been a crash in the absense of the IFEN, he avoided the question. Had he done otherwise, the headlines would have been quite different. Our objectives would have been better served, but the TSB's would have suffered.
So it's not a question of the integrity of the TSB investigators. Their actions are consistent with their clearly stated objectives. At one level, the best we can do is to agree to disagree. But on another level, we are simply attacking the same problem on different fronts. Investigators are forced to working within a system that we believe to be fundamentally flawed. But it is not their mission to change the system, but rather to do the best they can within it. In fact, they are dependent on cooperative relationships to implement their recommendations. On the other hand, we are unconstrained by these factors and will do everything we can to hold "the system" accountable for its failings. (Please note that I refer here to the US based complex of legislation, regulation and industry. I could not presume to comment on TC.)
So of course we want the draft report, responses and dispositions. We want everything. The list is long.
So we have our squabbles, but should not forget that we agree far more than we disagree. Ultimately, our interests converge. For now, there may be the occasional verbal bar-room brawl. But at the end of the day (or night as the case may be), we'll buy a round for everyone.
But here are the key facts from below, as I see them and understand science.
Chafing is a mechanical process that results from the ever-present high frequency vibration of flight.
It is a proven wiring menace.It is the most worried-out thing in wire harness installation design, but hardest to quantiy a precise vibe spectrum of non-ideal airplane and strange airplane structural zones. Measurements & inspections over the life of a specific type design are the only ways to quantify that type design's wiring's susceptibility to chafing over its life.
soft Tefzel wiring insulation used for the IFEN had no business cohabiting with the much harder BMS-51
And interaction of dissimilar materials is yet another factor that is even harder to predict because each different material will have its own resonant vibe frequencies, inducing different chafing than if all wire materials were the same.
Under-gauge that wiring conductor in the IFEN wiring and the heat build-up in the wiring is that much greater.
Another fact. And under-gauging of wire, when coupled with allowing higher amperages will lead to trouble on its own, much less when you are talking dissimilar materials.
Pass that large wire-bundle through a restricted passage (such as a bulkhead throughway) then you also have a heat-sinked environment (with little convective cooling).
I know those feed-thrus on that side of the fwd bulkhead into the flight deck very well. It is very carefully designed, as it should be since so many critical wires to critical breakers pass thru this area. It is telling indeed that this area is where the fire started.
A related story from my experience: During the very early days of what is becoming known as Electronic Flight Bag, we did and in-flight evaluation of two Onboard Maintenance Terminal systems on two MD-11s in the FedEx fleet. I had to design the add-on wiring installation, and I KNEW I did not want to run these non-critical OMT wires thru that same feed thru behind the CB panels. However, since the OMT "lived" on the flight deck (as opposed to the IFEN), it did have to have wires go up there, and get power from that same panel, and the wires went to the main avionics rack.
Because I did not want to disturb ANY of those production cables in the production feed thru, I purposely brought my add-on OMT wires forward of the cockpit bulkhead, along the fuselage, and then UP to the flight deck behind the FO's seat and behind the 3rd observer's station (old Flt Engr station). Physical isolation, a simple & effective design-for-safety tool.
Rainman original MD-11 Flight Controls and LSAS Design Engineer
|Powered by Social Strata|