This is not a good look. Looks like regulatory capture.
I’ve always thought it was a bad idea to let Boeing have a monopoly on American air transports, and this is probably one of the consequences.
[Wednesday-morning update]
Some aviation-safety experts have been predicting this sort of “automation surprise” for years.
I think part of the problem is that pilots don’t fly the planes any more; they manage them.
[Bumped]
These civil servants say they knew four years ago but did nothing because they were were more afraid of losing their jobs than endangering the public they were hired to protect. Prosecute the lot of them.
This is looking very ugly for Boeing. I’ve been reserving judgment, but the more I learn the worse it seems. It’s like they did everything wrong here. First, putting the new engines on the 737 seems like a hack when a clean sheet design was probably called for. Second, the MCAS seems like an attempt to cover a change in flight characteristics that probably should have been addressed with training rather than automation. Third, when it came to training on the MCAS system, it appears that they did a terrible job. It looks like these pilots didn’t know this was the issue or how to override it. And fourth, this self-certification seems really fishy.
Back in the earlier thread on the 737 (now on page 5), I pointed out that the MCAS should just move the stab trim to a predetermine point based on AOA and thrust to compensate for the pitch force difference between the 737 MAX and the 737 NG. The system they have acts as if that difference grows over time, and it doesn’t.
Then I suggested an engine change, in part because designing things is fun.
“I might as well discuss the original design mistake of switching to the LEAP-1B engine, with its wider fan diameter that required the new engine location that caused the loss of margin in pitch at high AOA.
They went from the CFM56’s 61 inch fan on the 737 NG series to a 69.4 inch fan on the LEAP-1B, which mean the engine wouldn’t fit under the wing.
The bypass ratio went from 5.1 to 5.5:1 to 9:1 and the fan area went from 2922 sq inches to 3782 sq inches, a 29% increase. But the larger area obviously created a pretty serious problem.
So I’ll throw out a crazy option.
Use the Rolls-Royce Pearl 15 that’s rolling out this year for the new Bombardier Global 5500 and 6500 business jets. It has a 48.5 inch fan (1847 sq inches), but a low 4.8:1 bypass ratio and only 15,125 lbf thrust, compared to the LEAP-1B’s 27,000 to 29,000 lbs.
So use four of them, podded two per nacelle like a B-52. The Pearl 15 is an outgrowth of the BR700 family, and Rolls-Royce is already pitching the BR725 in pods for replacing the B-52 fleet’s engines.
The Pearl would give a fan area of 3694 inches, which is 97% of the LEAP-1B’s fan, and you’ve knocked 21 inches off the fan height, and are back to the vertical size of the JT8D on the original 737’s.
In virtually all respects (seating capacity and range) Boeing has made the 737 fulfill the same role as the original 707, and a four-engine 737 would just complete the transition.
But there’s still a slight shortfall in SFC because of the lower bypass ratio, despite the Pearl 15’s overall 43:1 pressure ratio, and there’s obviously added expense and complexity in maintaining twice as many engines.
So keep both fans and get rid of one of the cores by enlarging an existing Pearl 15 core and adding a duct fed from between its high and low pressure turbines, which goes over to another low pressure turbine on the shaft that runs the second fan. The double-shaft/single core should end up with around a 9:1 bypass ratio, the same as the LEAP, with the same fan area as the LEAP, but with 20 inches less height.
The LEAP-1B costs about $13 million. The CFM56-7 it replaced costs $11 million. The BR700 series is about $7 million, and the Pearl shouldn’t be much different because it didn’t bump the price on the new Bombardier aircraft that use it.”
A safety system in an airliner which does not have redundant sensors. That should be enough to tell you something.
The worst thing is, regardless of whatever minor delay designing a new plane would have made I think Boeing had more than enough time to do it. The oil prices were lower than usual and still are. Instead of designing the 797 they could have designed the new 737.
“Boeing has pointed out that the pilots flying the same plane on the day before the crash experienced similar behavior to Flight 610 and did exactly that: They threw the stabilizer cutoff switches, regained control and continued with the rest of the flight.”
Why was this not passed down to the next crew if it was not repaired? Whoever at LionAir is in charge of maintnance needs to be asked under duress how this was allowed to happen.
AFAIK maintenance was done and they replaced a sensor. But the sensor they replaced wasn’t the faulty one.
Regardless of it, having a required safety system which relies on a single sensor is a design mistake no way you cut it. And Boeing making it a “pilot error” is, I think, a bit bonkers. It’s like the car brakes stopped working, you crash, and then they tell you it’s a “pilot error” because you forgot or failed to use the engine to brake the car.
I think it might not b so much regulatory capture as regulatory over reach. Requiring and reviewing 100 pages is a totally different animal than 100,000 pages. Sometimes problems are caused by the very procedures that are supposed to prevent them.
Though I do agree that a single manufacturer of anything is a problem.
I’ve seen a screen shot purporting to be the maintenance log of the Lion Air flight. It appears to have been launched without the fault fix being signed off.
By my count there are 5 other systems than MCAS that run the stab trim motor on the MAX. There’s the manual switch on the wheel, the autopilot, the stability augmentation system, the speed stability system and the Mach trim system.
All of these, when operating, can cause a runaway trim which the pilots must be able to counteract at any time by hitting the two disconnect switches which are very conveniently located and clearly labeled right between the manual trim wheels which if held still will also overpower the electric trim drive. The pilots can easily tell when any system is operating as the manual trim wheels also turn.
There have been several incidents in the USA where the MCAS allegedly falsely triggered but in every case the pilots isolated the stab drive motor.
There’s nothing wrong with the airplane. This is media and politically driven bs and hysteria. There are also commercial interests at work but not just at Boeing. Rationality has been thrown out and scapegoats sought.
Re-engining isn’t going to happen, George. It was airline customers in the US who wanted more fuel efficient engines instead of a later delivered brand new design and Boeing had to upgrade the 737 to compete with the A320NEO or miss out on lots of orders.
Now take a look at some pictures of the new 777-8 and 777-9 with their larger fanned engines, again entirely ahead of the wings and perhaps even lower relative to the vertical center of gravity. They have not been flown yet. It will be interesting to see what happens there when close to stall. Being FBW however, unlike the 737 MAX, any “MCAS” system will simply be hidden in the FBW software.
That’s my understanding as well.
Well, Boeing could have raised the wing by four or five inches or., come up with a way to extend the landing gear, but that apparently would require a new certification. My thought is that they could just add 8″ to the tire diameter by using some wheel shop on the LA strip. Who wouldn’t want to fly in a pimped out 737?
The 320neo has the same difficulty mounting the LEAP engine. They had to move it forward and up, and the nacelle clearance with a forward CG dropped from 22 inches to 18 inches. (320 specs, page 40, clearances.
They’ve probably taken under-slung engine diameter farther than good design practice would warrant.
But back to the flaw in the MCAS system. The goal was to make the plane fly like an NG so the pilots didn’t have to be retrained. So do that by figuring out the increase in down trim required compared to an NG for equivalent flight conditions, and that would fix the problem because the NG is perfectly flyable without any MCAS system at all.
“All of these, when operating, can cause a runaway trim which the pilots must be able to counteract at any time by hitting the two disconnect switches which are very conveniently located and clearly labeled right between the manual trim wheels which if held still will also overpower the electric trim drive. “
Trim runaway, yes I believe that’s a trained emergency procedure for most airliners. However, I think the issue may be that MCAS malfunctions may not show up as a classic trim runaway initially. Now most of the information I’ve gathered comes from aviation blogs and forums, but my understanding is the system could get false high AOA indication, run nose down trim until it thinks it has “corrected” it then stop. Pilot doesn’t see trim wheel running, simply notice that it’s now requiring more back pressure to maintain desired flight attitude. Corrects with manual trim, system stops while trim is running manually, then starts again when he lets go thinking it’s still correcting high AOA. Rinse and repeat, jet’s nose is trimmed down hard and recovery is difficult at low altitude. If this is true, then Boeing was definitely remiss in not including this in the flight manual. It’s one thing to have a trim runaway, it’s another to have a system that can malfunction, counter pilot, inputs, and no one tells the pilot. Honestly, Boeing should have relied on additional training plus conventional stick shaker, or treated the addition of MCAS more like stability augmentation with redundancy and detailed failure mode analysis plus pilot training on the new system.
If true, I would agree that’s a problem for Boeing. However, I don’t think that’s the case. My information comes from a former 737 IP, but admittedly (due to other reasons), left that role before the role out of the MAX and now flies a 777. First, I don’t know how you miss the trim, as it is loud. But, your scenario describes putting in manual trim. If by this, you mean grabbing the trim wheel (rather then clicking the yoke toggle), then that deactivates the automated system. My friend is puzzled as to why that wouldn’t be a trained reaction, as he taught it for the 737NG.
Please note, I’m only noting the specific scenario suggests it may not be true, but neither I nor my friend knows if MCAS might work differently. And if it does, in the manner you described, that’s a big problem.
From the linked article “The limit of the system’s authority to move the tail applies each time MCAS is triggered. And it can be triggered multiple times, as it was on the Lion Air flight.
One current FAA safety engineer said that every time the pilots on the Lion Air flight reset the switches on their control columns to pull the nose back up, MCAS would have kicked in again and “allowed new increments of 2.5 degrees.”
Seems to say that yes, pilot use of trim only disables the system temporarily. Of course, I’d trust it more if it were a more technical publication. Either way, I can’t understand why both AOA probes weren’t used for input.
Part of my understanding, perhaps mistaken, is that we could have a series of cycles where MCAS is adjusting the horizontal stabilizer for more and more down trim (leading edge up), while the pilot corrects by using the elevator trim tab, adjusting the elevator for more positive pitch (trailing edge up on the elevator). So the tail would end up looking like a V and eventually the the pilot runs out of elevator authority because the elevator alone can’t counteract full down stabilizer trim.
Concur on the audio cue. Check out Mentour Pilot’s youtube channel; he has several videos taken from inside a simulator, and the manual trim wheels make enough noise to wake the dead every time he touches the trim tab on the yoke.
George, they extended the nosegear on the Max by 8″ which was all that was needed.
I doubt we’ll ever find out what the larger engines and nacelles do to the A320NEO stall characteristics and in any case that will be hidden in the FBW software. I wonder what the differences training for the NEO vs the Classic is?
Ah, here’s one: http://www.avimeter.com/training/a320-neo-leap-1a-differences-course-theory-practical
New updates from an experienced Boeing 727, 737, 757 pilot.
0:00 to 1:20 – He says the cockpit voice recorders on the Lion Air crash show they were going over the checklist till the end. On the previous flight of that aircraft there was also an MCAS upset and another pilot in the jump seat aided with recovery.
4:45 to 5:15 – He says the French have just revealed that the Ethiopia crash data indicates a problem with the MCAS system.
6:25 to 7:40 – The jack screw is not the problem. The aircraft are brand new. It’s the input to the jack screw that’s the problem.
7:40 to 12:08 – Q: If recovering is as simple as throwing the STAB TRIM cutout switches, which have been in the same place on all 737’s, why aren’t the pilots flipping the switch? A: The MCAS problem is much more insidious than runaway trim, which is both obvious and a memory item. The symptom is entirely different. It’s an intermittent problem that probably starts out with a whole series of warning lights that leads you down the wrong path.
12:08 on – History of the MCAS, the engine change, Airbus competition, and then lots of engineering gold. I didn’t realize that the MCAS system is also going to override and run the throttles full forward.
Lots of updates towards the end. New training in the simulator and design changes, including added redundancy, and the system has to be scaled back to the original design intent so that the system adds just some trim and doesn’t have the control authority to cause a disaster.
There’s another aspect of “automation surprise” that I’ve been tracking for decades: structural analysis software is constantly being “improved,” and the natural temptation is to lower design margins as the analysis becomes more “accurate.” A colleague of mine in the ICBM world was a structures engineer, and maintained that structural design software would one day allow a raft of failures. The more widely used the software, the worse it would be.
I was skeptical of his view until climate “science” came along, and proposed the destruction of the world economy based on computer predictions. I’m qualified to know the limitations on computational fluid dynamics (and that climate models are BS), but no one needs a technical education to see that climate models don’t predict climate. So I’m a little leery of reliance on structure models now.