4 ms·

I've been curious about some questions that I haven't seen discussed anywhere. HN isn't probably the best place to ask them but since I'm not on any forums whe

by jasode 8y ago

I've been curious about some questions that I haven't seen discussed anywhere. HN isn't probably the best place to ask them but since I'm not on any forums where aerospace engineers hang out, I'll just ask here anyway: - Are physics computer simulations (e.g. aerodynamic flow[0]) realistic enough that engineers could have predicted the tendency for the nose to push up into a stall with the forward-mounted bigger engines while still at the CAD drawing stage? (Place some engines on the drawing with X amount of specified thrust, move them forward on the wing, then run a simulation, etc.) Or is the interaction of behavior so complex or so subtle that they had to build the real thing before realizing "Oops, our internal test pilots just noticed the plane is unstable and it looks like we need a software augmentation system!" - Could an unchanged 737-MAX flight simulator from last year reproduce the conditions for the Lion Air and Ethopia crashes? In other words, could a pilot used his/her imagination and said, "What if we make this AOA indicator give a bad reading -- what happens with MCAS?". Similar to how a pilot can choose to simulate a flamed-out engine, could one have chosen a stress-test scenario with a broken AOA sensor? (The infamous deleted video from MentourPilot showed his simulated struggles with turning the stabilizer trim against 400mph aerodynamic loads but it's not clear if a broken AOA can be simulated.) Yes, the 737-MAX was ultimately a failure of management and financial pressure but I'm curious if Boeing even had the technical debugging tools to predict this. [0] examples: https://www.youtube.com/results?search_query=aerodynamics++simulation https://www.youtube.com/results?search_query=aerodynamics++s...

foobarbecue 8y ago
Absolutely and almost certainly.
dsfyu404ed 8y ago
>Are physics computer simulations (e.g. aerodynamic flow[0]) realistic enough that engineers could have predicted the tendency for the nose to push up into a stall with the forward-mounted bigger engines while still at the CAD drawing stage? (Place some engines on the drawing with X amount of specified thrust, move them forward on the wing, then run a simulation, etc.) Or is the interaction of behavior so complex or so subtle that they had to build the real thing before realizing "Oops, our internal test pilots just noticed the plane is unstable and it looks like we need a software augmentation system!" Yes. They knew full well that the MAX would do that in circumstances where a normal 737 would not. There is nothing wrong with the MAX's performance characteristics. The aircraft itself is perfectly safe to fly. What isn't safe is the half baked MCAS system they used to deal with that edge case so they could avoid having to retrain pilots.
- jasode 8y ago
  >Yes. They knew full well that the MAX would do that in circumstances where a normal 737 would not. To be clear, you're saying Boeing engineers knew this as early as the CAD drawing stage?
  - dsfyu404ed 8y ago
    Probably. A bunch of engineers who do that stuff every day are going to look at the prints and be able to infer the implications on handling. They probably couldn't predict exactly what circumstances would cause the pitch up behavior to manifest itself or whether those circumstances would have even been within the flight envelope but they would have known how moving the engines up and forward would have affected handling. Any automotive engineer who designs that stuff can look at a vehicle suspension and give you a rough estimate of how the vehicle will handle under certain circumstances. People who are well versed in boat hulls can look at a shape and make the same inferences. Yes, commercial airline designers should be able to look at a design and have a rough idea of how it will handle. It's not magic. It's physics. If your day job involved understanding that niche of applications you should be able to look at it and draw some rough conclusions.
- simion314 8y ago
  >There is nothing wrong with the MAX's performance characteristics. The aircraft itself is perfectly safe to fly How do we know? because the FAA approval is not ensuring me that there are not any other problematic systems where Beoing self approved the corner cuts they made. IMO a full re-approval is needed with a trust worthy institution is needed before you can claim that the airplane is "perfectly safe"
  - cmurf 8y ago
    FAR 25, perhaps specifically 25.171-25.181 related to stability but there are separate sections on stall behavior, controllability, maneuverability, performance, etc. Those things must be demonstrated to get an airworthiness certificate under that part. Is this FAR all encompassing of that which makes an airplane safe or not? Because we have an airplane that was so certified, and yet we have two crashes that directly contradict their safety (hence they're grounded). I'm less concerned that these planes are somehow, without computers, not complying with FAR 25. I'm more concerned that Boeing (and FAA) will accept a software only fix, and continue to avoid a type certificate for this derivative model aircraft, and thus avoid any additional training for pilots. Basically, I expect Boeing will take the full hit, i.e. the software was flawed and now it is fixed, so that airlines won't be dinged for any on-going training requirement. Thus far the market isn't significantly dinging Boeing's value. Today's price is the same as it was in September of last year before either accident happened. The market is effectively saying those accidents don't matter that much, cost wise. Unless more information comes out that suggests there was wrong doing by either Boeing or the FAA. That could change the whole calculus.
mimixco 8y ago
I read somewhere that the Max was shipped before simulator upgrades were available to create scenarios involving MCAS. Also, several pilots have said their airlines didn't require new simulator training to fly the Max. Even after the Lion Air crash, few airlines had simulators with MCAS support. This says to me that no, they did not have the debugging tools to predict this because they had no plans to simulate it before going to market. The same was found of Toyota after their uncommanded acceleration lawsuits. In that case, the company was found to have deliberately ignored industry-standard testing procedures for mission-critical software. Could that also be the case here?
- jasode 8y ago
  >Also, several pilots have said their airlines didn't require new simulator training to fly the Max. Right. The media repeatedly reminded us that pilots only had a 2-hour orientation on an iPad for the new 737 MAX. I was wondering if a 737 MAX simulator that could replicate the AOA failures even existed. If Boeing's own internal test pilots had no ability to synthesize a broken AOA with misbehaving MCAS in the flight sim, it means that forcing the airlines to pay more money to have all their pilots more comprehensively trained in a 737 MAX simulator may not have prevented the crashes. The 737 MAX simulator hardware/software would still had to have been modified. Pilots can't train for an edge case scenario that the simulator can't artificially produce. Also in an alternate universe, a "more realistic" sim with AOA failure could have provided feedback loop to the Boeing engineers to adjust the MCAS software to be less aggressive and/or emphasize the need for 2 AOA sensors instead of just 1.
bb88 8y ago
> Yes, the 737-MAX was ultimately a failure of management and financial pressure but I'm curious if Boeing had the technical debugging tools to predict this. If management is depending upon a reality distortion bubble to keep the company out of trouble, then they're not going to listen to anyone who tries to pop that bubble.
cstejerean 8y ago
If they had disclosed the MCAS behavior to pilots and explained how it works in detail then some enterprising pilot might have tried that to see what happens (I’m guessing). But they didn’t tell pilots about this change at all because Boeing didn’t want to overwhelm them with supposedly irrelevant information. That’s the part that boggles my mind, and makes me distrust anything new coming from Boeing.

by jasode 8y ago

foobarbecue 8y ago

Absolutely and almost certainly.

dsfyu404ed 8y ago

>Are physics computer simulations (e.g. aerodynamic flow[0]) realistic enough that engineers could have predicted the tendency for the nose to push up into a stall with the forward-mounted bigger engines while still at the CAD drawing stage? (Place some engines on the drawing with X amount of specified thrust, move them forward on the wing, then run a simulation, etc.) Or is the interaction of behavior so complex or so subtle that they had to build the real thing before realizing "Oops, our internal test pilots just noticed the plane is unstable and it looks like we need a software augmentation system!" Yes. They knew full well that the MAX would do that in circumstances where a normal 737 would not. There is nothing wrong with the MAX's performance characteristics. The aircraft itself is perfectly safe to fly. What isn't safe is the half baked MCAS system they used to deal with that edge case so they could avoid having to retrain pilots.

jasode 8y ago
>Yes. They knew full well that the MAX would do that in circumstances where a normal 737 would not. To be clear, you're saying Boeing engineers knew this as early as the CAD drawing stage?
- dsfyu404ed 8y ago
  Probably. A bunch of engineers who do that stuff every day are going to look at the prints and be able to infer the implications on handling. They probably couldn't predict exactly what circumstances would cause the pitch up behavior to manifest itself or whether those circumstances would have even been within the flight envelope but they would have known how moving the engines up and forward would have affected handling. Any automotive engineer who designs that stuff can look at a vehicle suspension and give you a rough estimate of how the vehicle will handle under certain circumstances. People who are well versed in boat hulls can look at a shape and make the same inferences. Yes, commercial airline designers should be able to look at a design and have a rough idea of how it will handle. It's not magic. It's physics. If your day job involved understanding that niche of applications you should be able to look at it and draw some rough conclusions.
simion314 8y ago
>There is nothing wrong with the MAX's performance characteristics. The aircraft itself is perfectly safe to fly How do we know? because the FAA approval is not ensuring me that there are not any other problematic systems where Beoing self approved the corner cuts they made. IMO a full re-approval is needed with a trust worthy institution is needed before you can claim that the airplane is "perfectly safe"
- cmurf 8y ago
  FAR 25, perhaps specifically 25.171-25.181 related to stability but there are separate sections on stall behavior, controllability, maneuverability, performance, etc. Those things must be demonstrated to get an airworthiness certificate under that part. Is this FAR all encompassing of that which makes an airplane safe or not? Because we have an airplane that was so certified, and yet we have two crashes that directly contradict their safety (hence they're grounded). I'm less concerned that these planes are somehow, without computers, not complying with FAR 25. I'm more concerned that Boeing (and FAA) will accept a software only fix, and continue to avoid a type certificate for this derivative model aircraft, and thus avoid any additional training for pilots. Basically, I expect Boeing will take the full hit, i.e. the software was flawed and now it is fixed, so that airlines won't be dinged for any on-going training requirement. Thus far the market isn't significantly dinging Boeing's value. Today's price is the same as it was in September of last year before either accident happened. The market is effectively saying those accidents don't matter that much, cost wise. Unless more information comes out that suggests there was wrong doing by either Boeing or the FAA. That could change the whole calculus.

mimixco 8y ago

I read somewhere that the Max was shipped before simulator upgrades were available to create scenarios involving MCAS. Also, several pilots have said their airlines didn't require new simulator training to fly the Max. Even after the Lion Air crash, few airlines had simulators with MCAS support. This says to me that no, they did not have the debugging tools to predict this because they had no plans to simulate it before going to market. The same was found of Toyota after their uncommanded acceleration lawsuits. In that case, the company was found to have deliberately ignored industry-standard testing procedures for mission-critical software. Could that also be the case here?

jasode 8y ago
>Also, several pilots have said their airlines didn't require new simulator training to fly the Max. Right. The media repeatedly reminded us that pilots only had a 2-hour orientation on an iPad for the new 737 MAX. I was wondering if a 737 MAX simulator that could replicate the AOA failures even existed. If Boeing's own internal test pilots had no ability to synthesize a broken AOA with misbehaving MCAS in the flight sim, it means that forcing the airlines to pay more money to have all their pilots more comprehensively trained in a 737 MAX simulator may not have prevented the crashes. The 737 MAX simulator hardware/software would still had to have been modified. Pilots can't train for an edge case scenario that the simulator can't artificially produce. Also in an alternate universe, a "more realistic" sim with AOA failure could have provided feedback loop to the Boeing engineers to adjust the MCAS software to be less aggressive and/or emphasize the need for 2 AOA sensors instead of just 1.

bb88 8y ago

> Yes, the 737-MAX was ultimately a failure of management and financial pressure but I'm curious if Boeing had the technical debugging tools to predict this. If management is depending upon a reality distortion bubble to keep the company out of trouble, then they're not going to listen to anyone who tries to pop that bubble.

cstejerean 8y ago

If they had disclosed the MCAS behavior to pilots and explained how it works in detail then some enterprising pilot might have tried that to see what happens (I’m guessing). But they didn’t tell pilots about this change at all because Boeing didn’t want to overwhelm them with supposedly irrelevant information. That’s the part that boggles my mind, and makes me distrust anything new coming from Boeing.