Hacker News Reader

Comment by atharris | Hacker News Reader

Aerospace controls engineer here - while the airframe might not be passively stable (as is common for civilian aircraft), dynamically unstable aircraft have been stabilized with control software since the 70s [0]. If you've flown on an MD-11, you've flown on an 'aerodynamically flawed' aircraft. Most real systems are dynamically unstable without some kind of controller (implying software) in the loop.

[0] https://en.wikipedia.org/wiki/Relaxed_stability

View on HN

Boeing has kind of a poor track record for software lately, from the Max to the SLS. Time will tell if this has resulted in real change.

Ah - I didn't say Boeing's ability to write and test that control software was particularly good (in fact, I think their current track record says exactly the opposite.) I just hate when non-domain experts make judgements about things being 'fundamentally flawed.'

Insufficiently tested and documented? Sure. Bad UI/UX? Most definitely. Irredeemable 'because of aerodynamics' according to some private pilot that flew a 737 once in sim? Absolutely not.

Amen. If we don’t focus on the root cause it will be harder to solve the problem. And understand that we have actually solved it.

(Student Pilot and Mechanical Engineer here)

>> because of aerodynamics

But yes. MCAS was put in place due to concerns over aero. If that was just to avoid the need for extra pilot training then it should have been scrapped since new training will be required now anyway. But since great effort has been made to fix MCAS we can conclude that the root problem is aerodynamic.

Can aerodynamic issues be compensated for with software? Sure. I need to read up on the final hardware/software/instruction solution before passing Judgement.

You're completely ignoring that flying the 737 MAX without MCAS is not an automatic death sentence. Meanwhile a malfunctioning MCAS is actually an automatic death sentence.

The big flaws are in the software, not in the hardware. So stop focusing on that.

The qualifier of 'automatic' on death sentence is fun.

Would the software even be necessary if the airframe was better designed?

Those typically aren't approved for civilian air-transport though, are they?

The MD-11 is a passenger aircraft.

Right, but that's quite atypical, isn't it?

It’s “atypical” because it’s a 30+ year old superseded model and there are more efficient designs available. KLM, a flag carrier, was flying them up until just a few years ago. If it’s certified for carrying passengers, it’s certified. There are no special concessions made to airworthiness regulations for aircraft that sell few in number.

The MD-11 has been certified for air transport since it was introduced, was flying in revenue service until 2014, and as far as I know, that certification has never been revoked.

The MD-11 that was so crashy that airlines eventually sold their airframes to freight companies because cargo can’t refuse to get on one?

I don't think the flight safety record of the MD-11 bears that out[0] - most crashes of significance were either cargo flights (which are much more prone to dynamical issues than passenger flights) or flights in conditions that exceeded design specs (landing in typhoons). It sounds like most airlines sold it because it missed range/fuel burn targets, not because of safety issues.

[0] https://en.wikipedia.org/wiki/McDonnell_Douglas_MD-11#Accide...

My Dad was a frequent business traveller in the 1980s and 90's and I remember him commenting on the MD-11 and saying that he hated them because they were noisy and had lot of vibrations at the back from the center engine.

He said that was the the reason airlines switched to using them as cargo planes.

I realise this is just an anecdote, though.

Reply on news.ycombinator.com