Twinning the Attitude References
On a flight instrument layout safety aspect:
As may be obvious from the soon-to-be-released Stansted KAL 747 accident report and the Afghan-bound B-1B (ex Diego Garcia) accident report (and many others, including SR-111 - releasing 27 Mar):
It is important that Primary and Secondary attitude instruments (Analogue, EFIS, FD, CRT, PFD or otherwise) should be in close juxtaposition (i.e. close alongside each other). It is far too late, once a power source or instrument has failed, to start trying to sort out which is telling the truth (if any). Normally one WILL BE, or maybe you'll have an aural alert or Off Flag or one might just freeze - but if you are into a dynamic unusual attitude BEFORE you start trying to sort out the bogus and balderdash from the bonafide, you have VERY LITTLE CHANCE of making a correct choice and recovering without pulling the wings off.
It can all happen very fast (particularly when distracted) - and that's why a "twinning" of the primary and backup attitude references gives you the best chance for an early heads-up that something is awry. There are many examples around of poor lay-out for the standby AI and the MD-11 is one that springs to mind (centrally low-down and forward of the centre pedestal, hidden behind throttles). When the primary and backup are closely twinned you don't need the peripheral vision that most of us haven't got - in order to quickly pick up a failure.
The B-1B Accident
Cascading failures. With unreliable instruments and no horizon cues in the night sky, the crew ejected, according to an account of the December 2001 crash of a B-1 bomber appearing in the latest issue of Flying Safety. The U.S. Air Force Safety Center publishes the magazine. The article by Maj. Dan Baker recounts a progression of system losses that may bear on commercial aircraft engaged in extended-range operations (ETOPS):
"Shortly after [a nighttime] takeoff from Diego Garcia ... the aircrew experienced multiple systems failures which resulted in the loss of aircraft control and the aircrew ejecting from the aircraft. According to the Accident Investigation Board ... the crew shut down the number one engine due to an oil over-temperature. The associated primary generator fell off-line normally during the engine shutdown. The crew decided to abort the mission and return to Diego Garcia.
"En route, the number two primary generator dropped off line, accompanied by loss of the ... aircraft's computer navigation complex. The pilot switched on the emergency generator, in accordance with the ... emergency procedure for single generator operation. Shortly thereafter, the pilots determined [that] their primary and standby aircraft attitude [i.e., level flight, turning, climbing, etc.] information was unreliable.
"Though the weather at the cruise altitude of FL 200 was clear, there was no lunar illumination and neither pilot could discern the horizon. The OSO [offensive systems operator] and DSO [defensive systems operator] noted increasing uncommanded bank angle displays up to 120º accompanied by rapidly decreasing altitude and increasing airspeed, and advised the pilots. Passing 15,000 feet MSL, the OSO determined the aircraft was out of control and, in accordance with TechOrder guidance, ejected, followed quickly by the DSO ... Convinced the aircraft was out of control and unrecoverable, the pilots ejected. The aircraft was destroyed upon impact with the water and sank. A U.S. Navy vessel in the area rescued all four crewmembers. After lengthy salvage operations, the aircraft was never located, and no parts of the aircraft were recovered."
Wouldn't be surprised if they'd had a bus or wire-bundle fire (which were my thoughts at the time). The cascading nature of the emergency as systems dropped out, culminating in an apparent loss of ALL attitude instruments, obviously left the crew with little choice but to eject. It seems very likely that SR-111 might have progressed along the same lines into its terminal spiral.
The original oil overtemp may have been bogus and just an early manifestation of their burgeoning electrical problems. But how is one to know? Not all electrical problems are going to manifest themselves in smell, smoke or flames, particularly as the problem may easily have been outside the pressure vessel in the case of this B-1. The crew were probably all wearing oxygen masks anyway - and they inhibit the smell of smoke somewhat. As the problems were all initially on the port side of the airplane, I'd guess that the electrical fire was in that left wing. I would imagine that the primary and standby attitude sources were giving conflicting information and that the standby AH was likely correct - however there is only ever a very finite period in which to establish that. Once you have the two main instruments tumbling or toppling they will be obviously "disagreeing" with the standby. However the standby AH will also be (by now) dynamic with a rapidly changing presentation - simply because the aircraft attitude may now itself be changing rapidly). An inevitable pilot confusion about whether any displayed attitude is valid (and the instrument usable) will be soon overcome by the urgency of ejecting before it's too late. That sense of urgency would be heightened somewhat by the OSO's precipitate departure. And he was probably within his rights, as 15,000ft out of control descending has always been the "every man for himself" cut-off point.
It really gets back to the need for "twinning" the primary and standby attitude systems. Early/immediate awareness of a primary or secondary instrument failure will more often than not preclude an unusual attitude developing at all. Looking at the cost of a B-1, I'd say that would be a cheap solution to "early outs" by confused pilots who've allowed the airplane to depart straight and level [and then become confused by the dynamism of unwinding altimeters, spinning HSI's, VSI's locked on the down stop, increasing g and airspeed, audio alarms, urgent intercom calls and a high ambient air-rush noise]. Once it gets to that stage, it doesn't really matter that there's a bright panoply of stars in the night's canopy - everything from there on in is a disoriented blur and avenues of escape are then in the forefront of one's mind. Recovery as a solution becomes quickly discarded when you are sitting on a bang-seat..... and someone has just audibly pointed the way.
You could say that it's an outstanding example of why electrical redundancy matters. No electrical system (or its dependent systems and avionics) can survive once it's been interdicted by a fire or combat damage. What's required is the alternate, stand-alone and switchable integrity of a Virgin Bus. It's incongruous that a military aircraft that needed that level of redundancy (in both war and peace) was lost, happily without loss of life, whilst an airliner with 300 or 400 people would have been a total loss in similar circumstances. In this techno day and age, survivability shouldn't depend upon whether or not there's a bomber's moon in the night sky.