Air Midwest flight 5481 crashed in Charlotte, NC in 2003. I’m watching a documentary on it now. Two problems combined to cause the crash. The aircraft had undergone maintenance two nights before the crash, and the elevators were mis-rigged. Instead of 14º of downward travel, there was only 7º. The other problem was that the aircraft was 580 pounds over gross with an aft CG. (A baggage handler asked the pilots about the loading; and after the crash, told an investigator the aircraft ‘looked heavy’ and was a bit tail-down.) The pilots performed weight-and-balance calculations, and found they were within the CG envelope. But they used the then-standard average weights of 175 pounds for passengers and 20 pounds per bag. People and their baggage are heavier now. When the undercarriage was raised, the CG was moved that much more aft and the nose pitched up. With limited elevator travel, there was no chance of recovery. Since the crash, FAA-standard passenger weight for U.S. males is 195 pounds. I don’t recall what they said the weight of the bags were on the accident aircraft.
One thing I always do when I preflight an airplane is to make sure that the elevators have a full range of motion. I realise this is impractical to do on a big airplane like a Beech 1900 (the accident aircraft) or a jet. What procedures are in place to ensure that control surfaces have a free range of motion?
Is there any way flight 5481 could have been saved? It was already in an aft-CG situation, at low altitude and at low speed. Might putting the gear down have helped at all? What about using the trim tab? If they trimmed up, could the little strip of metal hanging downward do any good? What does the nose do when you lower flaps on a low-wing, T-tail airplane? (I used a Piper PA-28 for my BFR, and I discovered that the nose acts opposite to a Skyhawk’s when flaps are lowered.)
Given that things were happening fast and time and altitude were in short supply, was there anything that could have been done by a ‘theoretically perfect pilot’ to salvage the situation?
Sitting here on my couch, I’d say that the appropriate action would be to bank steeply to recover the nose toward the horizon. Assuming you have enough thrust to accelerate to a manageable airspeed, you may now have the control authority to keep the nose down with the wings level. If not, use bank to keep the nose just above the horizon in a steep turn. This will at least buy a little time until you can figure out just what in the hell is going on. That being said, getting forced into a surprise low-speed, low-altitude steep turn in an overgross turboprop is not a place I hope to ever find myself. Yikes.
A steep bank might get the nose pointed a bit better, but it would also dump a good portion of the lift overboard.
Not what you want when you are already overloaded and tail-heavy.
I’d call that a doomed plane - once out of ground effect it had nowhere to go.
Which might just have been its last, best chance - retard throttle to lose altitude until ground effect kicks in.
Unfortunately, the Beech 1900 has a short wingspan. I don’t know where that plane went down, but if there wasn’t a large body of water or ultra flat ground under it, ground effect wasn’t available.
(Ground Effect: an increase in lift which exists while the plane is at or below an altitude equal to its wingspan)
You can’t really. All you can check is that the control wheel and column have full range. On some aircraft you can see the ailerons and you may have roll spoiler indicators if applicable. E.g, in the BAe146 you can see the ailerons move and the roll spoilers have indicators. The elevator is more complicated, I don’t know how the B1900’s is set up but when you get to certain size the control column doesn’t directly move the elevator at all (this is true for ailerons as well but on much larger aircraft.) Even if you still have full manual control rather than hydraulics you may still have a design where the control column moves a tiny servo tab on back of the elevator and the elevator just floats up or down in response to the servo tab’s movement. In this case when you do a full and free check, all you are checking is that the movement of the servo tab is full and free.
The photo below demonstrates how the elevators are free to move on their own. The one closest to camera is fully up and the other is down.
The B1900 is pretty small though so that probably doesn’t apply. The systems description at smartcockpit.com doesn’t mention anything at all so presumably it is just a simple direct control of the elevators. In this case you can do a full and free check of the control column but you still can’t check that it corresponds to appropriate movement of the control surface.
Stranger has the right idea for immediate actions. Get some bank in to keep the nose down and hope that gives you enough time to get some airspeed and have a think about things and start trying some of the other options you’ve mentioned. I haven’t flown a B1900 and don’t know what trim changes the flaps and gear have. You’d think putting the gear down would drop the nose a bit.
**Richard Pearse **has, as he so often does, covered it all well.
I have no experience with B1900s, so can’t add anything directly to the aircraft specifics of this mishap.
Some thought-bites:
Airplanes big enough to have directly powered flight controls have some cockpit indicator to show actual control surface displacement. Driving all 3 axes to both stops while checking the indications is part of the typical pre-taxi check. As RP says, the intermediate-sized airplanes where the flight controls deflect tabs which in turn deflect the control surfaces tend not to have position indicators.
Severe out of trim conditions are rapidly fatal. Very few crews successfully recover from them. A DC8 freighter crashed on takeoff from (IIRC) Kansas City back in the mid-80s that way. A 747 freighter crashed departing Bagram a couple years ago. There have been many others over the years.
Conversely, a Delta crew saved an L1011 departing San Diego in the early 90s. They were well out of trim & stuck that way. The situation was salvageable, just. And the guys flying happened to be especially good/cool.
The FAA’s updated standard weights mentioned by the OP have changed the situation with weight & balance computation from patently hogwash to merely mostly fake. IMO there ought to be another 30 lbs. allocated per passenger for big jets. And using actual weights by a certified scale should be mandated on all RJ & turboprop flights.
On big jets the difference between actual reality and the FAA’s standard (i.e. unrealistically optimistic) numbers amounts to a rounding error, and is very unlikely to result in a make or break situation in flight. Conversely on the 19 seater’s it’s a big deal. Those are marginal aircraft anyhow and they’re being operated a lot closer to the edge than the paperwork indicates.
Oh shit!
I seem to recall a similar incident during WWII with a B17, there was a lock on the control column that prevented the stick from moving forward. It was not removed during the preflight. The plane took off and they could not bring the nose down.
Do they have anything like a huge truck scale that they could install somewhere between the gate and the runway?
I’m not thinking about the massive mechanical old-school beam ones, more along the lines of the modern truck scales with multiple load cells and strain gauges. Someone in the tower keys in the airplane type and out comes a perfect CG calculation before takeoff.
Sounds so obvious that there must be a huge obvious flaw that I’m not seeing (e.g. too costly to implement or too costly in down time to weigh each plane)
That happens more often than you might think. A couple of months ago I saw a Cessna 172 for sale on eBay that crashed because the pilot couldn’t find the control lock (which blocks the master and magneto switches) and used a bolt instead. He forgot to take it out before he took off. (The plane has been repaired.)
On some aircraft, the flight controls don’t move directly with the pilot input. On the DC-9 there were little tabs on the ailerons, elevators, and rudder that moved in the airflow and pushed those surfaces the opposite way. The only control surface, besides the spoilers, that had any power assist was the elevator and that was only in one direction. A hydraulic actuator would slam the elevators full down in case the aircraft was in a stall and needed to drop the nose fast.
Installing that system at every runway at every airport would be silly expensive. Not gonna happen.
I hear the 787 has something like this. No other’s I’m aware of do. My knowledge is not definitive however.
The main challenge is the regulations require all calcs to be done down to the *single pound *level of accuracy. And gear-mounted load cells can’t do that. Another example of measure with a micrometer, mark with a crayon & cut with an axe. One of the FAA’s favorite doctrines.
