Sunday, April 2, 2006

Look, I've Invented the Ocularly-Administered Sleep Aid!

(This is a post--rated "G" for "Geek"--which those squeamish about flying maybe ought not to read.)

More simulator work. One of the things we worked on today (it's a regular item on this year's recurrent training for all DC-8 crewmembers) is an onboard fire. We don't actually light anything on fire in the simulator, but it's scarcely less torturous for this fact. Quite apart from the difficulty and discomfort involved in trying to work the checklists and fly the airplane with an oxygen mask and goggles on (and this is much more difficult than it might appear at first thought), there is the problem of the checklists themselves. There are separate checklists for smoke or fire in the cargo compartment, for a suspected electrical fire, for air conditioning / heating pack smoke, and for unidentified smoke or fumes evacuation (and these are all separate from fires involving an engine). Each of these checklists takes one to a different remedial place, and the consequences of not getting on board with the correct checklist can be serious, even dire. This is one of the reasons, of course, that we do regular training on these things.

But the actual symptoms of a fire, and the process of identification of the source of that fire, to say nothing of the adrenaline-filled, pressure-cooker situation that a fire will naturally engender, these things are not really adequately addressed in a simulator, in spite of the excellent and diligent efforts of some very dedicated people on the training staff. This exercise is all very instructive, but it only takes us about 30% of the way home. That 30% may easily turn out to be the margin needed for survival, but it leaves an awful lot to our discretion and to chance. To luck. So this is most valuable training, but it's a sobering subject; a fire in flight is, well, serious.

Today's scenario dealt with an unspecified electrical fire, and had us run the initial parts of the checklists in real time, but then the instructors delivered us from the misery of the masks & goggles so that we might learn something from the lengthy procedures that followed. Part of the reason the trainers do this is because the act of troubleshooting an electrical malfunction--that is, one that is not specifically and obviously connected with a particular piece of electrical equipment (which one might simply turn off, maybe)--involves a very long and disruptive checklist that, as it progresses, kills off regular and needed equipment, always when the guy flying (me) doesn't quite expect it. It's instructive, maybe, to run the whole works under the pressures that can be mustered in the simulator, but it's probably more instructive, in this case, to be able to pause and contemplate what these checklists are actually DOING, and to see the consequences of things with enough breathing space for some correlation to occur in one's frantic brain. Habits get formed, experiences stored away, for use on that rainy day.

But we'll lighten the tone a bit. How this checklist unfolds is a direct result of the design of the system in question (and we have similar checklists for problems with the DC-8's complicated fuel and hydraulic systems, among others). As an oft-proclaimed machinery geek, I'm endlessly fascinated by this aspect of things. What switches we have, and what they control, are the result of a brilliant engineering effort by the aircraft manufacturer (in our case, a group of slide-rule-weilding, narrow-tied, white-shirted geeks with pocket protectors and birth control glasses from the Douglas Aircraft Company in Long Beach, California in the early- to middle-50s). Conscious decisions had to be made about what protections were needed, and about the level of human interaction required for the system in question, and these things reflect both the state of technology available or attainable, as well as current trends or accumulated wisdom in the aerospace field at the time of design. A modern aircraft electrical system is a different animal absolutely than what the DC-8 sports, even though both can represent brilliant thinking. A modern electrical system needs virtually no human intervention, and will even troubleshoot itself. Likewise fuel, hydraulic, pneumatic, air conditioning / pressurization systems. (This is, in a nutshell, what happened to the flight engineer as a concept).

(An aside: I rode home yesterday in the cockpit of a Southwest 737, and it's a fabulous little exercise to listen to them run their checklists. The callouts and responses are different from what we say in the DC-8, and those differences reflect the differing design of the 737's systems, systems which exist to solve the same issues as the DC-8 systems address. As I watched, it was like dipping your brain into molasses: you figure out about 10 seconds too late what the question was, what the response meant, and what panel / switch they looked at for verification. This is exactly why each airplane's training is absolutely specific, and why learning one airplane won't really help you with flying another.)

I don't mean to go too far with this (and, shockingly, there's a lot further we might go); books can be, and have been, written about this stuff (and only geeky people buy them). But an introduction might be instructive. Any of us can accept without much difficulty the need for an airplane to have a supply of electrical power. But there is an ocean of questions between that accepted reality and the actual implementation. AC or DC power (or both)? Of what voltage(s)? Generated how? With what redundancy? What backups are needed? None of these questions has an obvious answer, and they represent about 1% of the questions in this vein one might ask about a single system.

The DC-8 represents one proven way to answer these thousand questions. It has an engine-driven generator on each of its four engines, any one of which, if the engine is running, produces enough alternating current to supply the entire airplane's needs (though with some restrictions if you were actually down to a single generator). But it's not simply a matter of an electrical line running from a generator to the equivalent of a power strip into which things are plugged. Or rather, it IS something like this, except there are 10 or 15 such large "power strips" (called electrical buses), which are protected by current limiters and off of each of which spring a bunch of powered items, each item protected along the way by circuit breaker(s). These buses are arranged in a hierarchical manner so as to be redundantly powered by a couple different sources; thereby certain, key things are very difficult to de-power. And to this mix is added a rudimentary brain that seeks, by opening and closing relays, to keep these buses powered with little or no human intervention, so that a loss of a given source of power will tend to be imperceptible except by the clunk of switching relays or to the engineer who now sees a warning light glowing on his panel.

If this seems abstruse reading from one's computer screen, it is decidedly less so when the two crewmembers fighting an electrical fire are throwing switches on the engineer's panel in an attempt to deprive a burning item of electrical power. X piece of equipment is powered by the Y electrical bus, which in turn gets its power from Z generator. But shutting off Z generator depowers virtually nothing (because of rudimentary brain above), and even shutting off power to the Y electrical bus will not necessarily deprive us of X. All according to some grand inscrutable plan. This is really cool, actually. What seems random or haphazard on the surface will, if dug into, reveal a pointed engineering effort behind it. The loss of power to the galley, or certain lights, for example, is unimportant. The loss of the autopilot is rather more serious. Losing the flight instrumentation is really fucking serious. The system is designed accordingly. Cool. The simulator lets us get to know these systems intimately.

Lastly (I've come this far, I might as well wallow in my unabashed geek out), I can't help noticing the similarities between these systems questions & solutions in my DC-8 and what one sees with the spacecraft that were being designed around the same time as these early jet aircraft. Without overstating the point, there are a lot of similarities between the problems faced by aircraft and spacecraft engineers. Much moreso than, for example, between aircraft engineers and early computer engineers. Both need to control a craft that maneuvers freely in all three dimensions; both need flight instrumentation for this control; both must provide an atmosphere under pressure for their occupants; both need instruments and indicators for these systems. And, of course, they had the same technology available for most of these issues. (The similarities were enough that military pilots were chosen for the early astronauts.) I've thought for a long time about doing a book of large format, high-resolution photographs of the early spacecraft cockpit environments, with a detailed discussion about what controls were there and why, and how those controls were actually manipulated for spaceflight. And maybe looking at the actual checklists, which would be instructive in this vein.

Like I said, only geeks buy these kinds of books. My wife, bless her heart, is sound asleep long before I finish explaining this book idea. Each time. I can even make her sleep standing up with this subject. Ambien in pulp form.


Anonymous said...

Glad to see the flying related post. I enjoy them. Its good to hear about the systems and workings of an old aircraft. The only "old" aircraft I ever flew was the T-37s...some of who rolled off the line in 57. Since then its been the automated, multiple-redundant, computer controlled type. Which, of course, have their own issues when the computer or electrics act strangely...kicking out a switch when it should not be or not responding to a command. Our best fix- pull breakers or go down to a black jet...which can be a pain when you are pressed for time...and you have to go all the way...even shutting down the APU.

Thanks for the writing...there are interested readers out there.

Anonymous said...

I know the feeling of being a living sleep aid. I too can put my spouse to sleep with my long and interesting explanations of subjects with which I am especially familiar. At times I even get special requests and perform on command - all the benefits of an over-the-counter sleep aid, with none of the nagging drowsiness the next morning!

And, just for your information - I enjoyed your geeky posting and read it without nodding off even once - at 2am no less.

I find it interesting that in flying so much time is spent rehearsing disaster-management scenarios. I wonder if there are other areas that are similarly inclined to be well-prepared. In most situations the disasters are not as well defined and not as certainly disasterous, and I think that in many cases people are left to wing it when the time comes. Also, in flying you have enough riding on your ability to handle disasters that it is worth the cost and time to not only spend it in training, but to develop the simulators necessary to train completely.

-- Jeffy

wunelle said...

I often read to Susan at night to put her to sleep, but I argue that I should be able to choose the material since she's generally out like a light in about 20 seconds! Lately, she'll ask about my latest post for this purpose...

Anonymous said...

Why we sim...why we spend 90% of our training on the less than 1% chance we have a problem

wunelle said...

Wow. An amazing pic!

BrianAlt said...


No, actually that was pretty interesting.