Does anybody want to hear about Boeing's MAX8 ?
I had promised a few people here in town a lesson in “Bisbonian Aerodynamics”, but the reality is that I am too sick to concentrate on graphics and stuff, so that will be later. I tried to include just enough in this to make it understandable to those who don’t fly for a living:
For many decades, airplane manufacturers have built "stable" aircraft...if the pilot lets go of the controls, the plane will just keep on doing what it was doing before...and if there is a bump or other disturbance, the plane will (by physics, not by control inputs) automatically return to its previous state. This was a big improvement on the original airplanes, that required constant input from the pilots to remain airborne...with predictable results.
Imagine an arrow. Heavy pointy rock at one end, light fluffy feathers at the other end. Center of gravity is pretty close to that rock. Center of aerodynamic pressure (the effects of the air flowing past the arrow) is close to those feathers. You shoot the arrow, and it goes right to where you point it. Little deviations due to wind currents position the feathers a little sideways to the relative wind, and the feathers correct the path back to where you want it. Extremely stable flying machine.
We don't want our planes to be quite that stable, because we would like to be able to turn them. So the center of gravity gets moved aft, to about 1/4 of the way back on the wing. Center of aerodynamic pressure is aft of that, but not a lot. Now, those little bumps and air currents move us around, but the natural tendency of the airplane is to return to it's previous path.
Fighter planes need to be more maneuverable, so for a long time, natural stability was reduced farther all the time, until the F-16, which has "neutral" stability. i.e. none. Four computers keep it flying the way the pilot tells it to go, but the computers are continuously adjusting the flight controls to keep the airplane from going out of control. It's kind of like backing up a trailer, but with a computer.
We don't want our airliners to be that way.
But we got one. Unknown to us.
The 737 has been a fantastically successful airplane. It's the offspring of the 707/KC-135 that I flew in the Air Force, and I have loved flying it. It has been improved over the years, through various models...I rode in 737-100s in Hawaii in the 70's, I flew -200s in the 90's, -300s and -500s through the latter 90's and 00's. (Those two took the well balanced -200, stuck a much bigger engine on it, and that necessitated bigger tail surfaces to balance everything out...but Boeing overshot. They were too big, and the airplane oversteered, especially in the up and down dimension. The -500 was same size as a -200, but with way oversized controls, and was a pretty lousy flying airplane, without the autopilot. The -300 was longer, and not as bad for overcontrolling. Both were naturally stable...just not as stable as I was used to.)
Handling was improved with the -700, and it flies pretty decently. The real problems with it are the autopilot, and I hear that is mostly because we bought the cheapest software for it. The autopilot was programed by some millenial who read about flying in a book, and has never even driven a car, let alone an airplane. Kludgy.
The -800 is a stretched -700. This is where the real problems start. We've taken a small, regional airliner, designed for short hops, and tried to turn it into something that would go to Hawaii. Airlines bought a bunch for that purpose, and then found out that they couldn't make money on that route, because at the fuel loads required for the trip, it was too heavy to get to higher altitudes where it could burn less fuel. A new engine was in order.
The biggest limiting factor on growing the 737 is the short, stubby little landing gear. The first engine upgrade demanded moving the engines partly forward of the wing, and flattening the bottom of the nacelle. The long fuselage of the 800 requires a tail skid, to try and prevent damage to the plane during takeoff and landing. Getting the plane into landing attitudes that were fine in a -700, could cause major damage in an -800 by hitting the rear end on the runway.
In order to get an -800 to go to Hawaii, we needed a bigger, more efficient engine. That combination was called the MAX8. Well, guess what. that bigger engine had to go even farther forward, and be more flattened on the bottom. When we rotate the airplane to take off, the underside of those forward mounted engines create lift, and that forward placed lift moves the center of aerodynamic pressure much closer to the center of gravity, making the plane less stable. (No one told us this). What "unstable" means, in this case, is that a tiny bit of pilot input will, at that point, result in a highly disproportionate aircraft response. To overcome that, Boeing installed a software patch that would, if the nose got too high on takeoff, adjust the tail to forcefully push the nose down. (No one told us this, either). The input to the computer comes from ONE sensor on the side of the plane, that measures the angle of the relative wind. (Not four redundant sensors, like on the F-16) So, if that one sensor has a malfunction, the plane rolls the trim forward for 10 seconds (I NEVER make a ten second long trim change...that's immense). If that still doesn't fix the angle, it does another ten seconds. And another, until the plane is uncontrollable to the pilots. (No one told us any of this.)
So, after the first crash, Boeing's first response was, "Well, all you have to do is turn off the electric trim...then the MCAS can't run the trim forward any more". To which we replied WTF is MCAS? "Oh, it's a system we installed to make the airplane safer. no big deal, really, we didn't figure you needed to know."
So, on takeoff, really the most critical phase of flight, in between talking to the tower, setting takeoff thrust, monitoring engines for proper reaction, steering down the centerline of the runway in a crosswind, calling out the go-no go speed, rotating the nose to the proper angle, slowly enough not to bash the tail into the runway, lifting off (here is where the stall warning started for both fatal flights, due to the failed Angle sensor...the stall warning shakes the control column, and makes a rattling racket). Start climbing away from the ground, and possibly pushing the nose forward to get out of a perceived stall, get the landing gear up, and flaps up, and BAM, ten seconds of nose-down trim (the system only works when the flaps are up...yeah, you guessed it, nobody told us about that) Trim back, at least some of that, which resets the system (nobody told us) and BAM, ten more seconds of nose-down trim. At this point, you are away from the runway, and the GPS knows this, and you are too low, don't have the landing gear down, and are not in position to land. The airplane has various loud warning systems that go off for each of these things, a mans voice yelling TOO LOW, TERRAIN! TOO LOW, GEAR! TOO LOW, FLAPS!, PULL UP! PULL UP!, repeating these things over and over.
So, one thing is telling you to pull up, and the stick shaker is telling you to push forward, and for your entire flying career the most dangerous, immediate action problem is a stall at low altitude...but somehow you are to put that all aside, and calmly turn off the electric trim switches.
I saw an article last night, Boeing is saying they can get a new software patch in place by next week, and everything will be hunky dory. It needs far more than a software patch. Critical flightpath decisions made by a computer need to rely on more than a single input (one Angle of Attack sensor). One can be wrong. It has been, so far, twice. Killing, on average, about 170 people per occurrence. Two would be better, three would be far better, and is the usual aeronutical standard for such questions. If one says one thing, and the other two agree on something else, at least you've got a pretty good chance of figuring out who is right.
I believe that an airliner that is aerodynamically unstable, in a certain flight regime, never should have been certified in the first place. But Boeing/USA/FAA are in an all out commercial battle with Airbus/France, and that's how these things happen.
The bottom line: This patch job solution to keep an unstable aircraft flying would be fine, if it were a fighter. Except that even in a fighter, it would rely on at least three independent inputs from the sensors, to rule out a sensor that was malfunctioning (the F-16 uses 4). To place complete reliance on one sensor on the plane is absolutely insane...and obviously fatal. It should be criminal. The black boxes are now in France, to be investigated; Boeing and the FAA are now bystanders. So who knows...some (partisan based) justice might be forthcoming.
Here is some supplemental reading that is pretty good: https://theaircurrent.com/aviation-safety/the-world-pulls-the-andon-cord...