traindoc From United States of America, joined Mar 2008, 329 posts, RR: 0 Posted (3 years 3 weeks 6 days 6 hours ago) and read 3351 times:
The problems with the Trent 900 and 1000 engines and the 787 raise the question: "Are these things getting too complex?". I am not being critical of or attacking Airbus, Beoing or Rolls. However, you need "almost" perfection with a commercial airplane. You cannot pull off the road to fix the problem. As planes and engines become more complex, there is a greater possibility of failure.
I am a physician, so I know something about complexity and systems failure. But I don't (usually) have to treat them at 35,000 feet while going 500+ MPH. How about some input from those of you with expertise in these areas?
Well, yes and no. Undeniably, modern aircraft are more complex in their systems than aircraft of a generation or two ago, but all of that complexity exists for a good reason. Whether that reason is safety, reliability, efficiency or productivity, all of the systems perform a job that is either required or desirable on a modern aircraft, otherwise airlines would not be willing to pay for those systems!
Just as cars are much more complex these days, the erosion of the ability to self fix a car is well offset by the massive increase in safety, reliability, efficiency and productivity that a modern vehicle possesses (even if some may argue they have given up their "souls" in the process!)
eta unknown From Comoros, joined Jun 2001, 2036 posts, RR: 0 Reply 3, posted (3 years 3 weeks 6 days 3 hours ago) and read 3043 times:
Well on the subject of cars, don't forget VW had to tweak the engines on the latest Golf- the previous model had a new generation engine that caused so many reliability problems VW's reputation was getting tarnished so they simplified it (or dumbed it down) for the current model.
comorin From United States of America, joined May 2005, 4861 posts, RR: 16 Reply 4, posted (3 years 3 weeks 6 days 2 hours ago) and read 3004 times:
In a simple system, x=tau(y) where Tau, the transformation function is generally monotonic and easily deducible over a finite range of y values. x is behavior of aircraft, and y is an independent variable ( or scalar) that describes aircraft state and control inputs.
In a complex system, tau is non-linear, may have multiple values and not esily deducible without a complex solution that requires months of training in a simulator across all values of y.
So modern aircraft are too complex if the learning curve to understand the tau function is costlier than the benefit received.
While cockpit automation has helped reduce crew and workload, has it made flying any safer? What does the data show?
Seems to me that the Airbii pilots here seem OK with the degree of automation on aircraft.
Fly2HMO From , joined Dec 1969, posts, RR: Reply 5, posted (3 years 3 weeks 6 days 2 hours ago) and read 2984 times:
I'm gonna go on a limb here and say airplanes today are overall less complex than airplanes of yore. Take a SuperConnie for example, each engine had 18 cylinders (aircraft total 72), which also equates to 144 spark plugs in the plane (two per cylinder) That's a TON of things that can break right there.
Then take jet engines, which even in their infancy, were vastly more simple. In the grand scheme of things, jet engines only have one moving part (dual and triple spools notwithstanding). Also, planes nowadays have much more redundancy than ever before. Not to mention you can start a plane nowadays literrally with just pushing a button. I don't necessarily aircraft systems nowadays to be more complex, maybe different, but certainly automation and redundancy enhances safety greatly.
Quoting eta unknown (Reply 3): Well on the subject of cars, don't forget VW had to tweak the engines on the latest Golf- the previous model had a new generation engine that caused so many reliability problems VW's reputation was getting tarnished so they simplified it (or dumbed it down) for the current model.
Not to derail the thread, but that's utter hogwash. VW's have been getting more and more reliable with every new generation. Maybe in the 70s I would have believed that statement, nowadays, no way. All the late VWs I've owned (and friends own) have been extremely reliable, much more so than any Asian or US car I've owned or driven. I've worked on the MKV Golfs a LOT and the MKVI (current generation) are basically identical. The new FSI engines have been very reliable, even the first versions. May want to post a source for those ill-informed claims.
You do, and in the past the way to ensure that was to overengineer and overbuild everything. It worked pretty well, but wasn't the most efficient. The industry needs to push the envelope but I think that a lot of the bean counters and managers don't have a good handle on just how difficult pushing the envelope can be, which leads to some problems.
Why do Aerospace Engineering students have to turn things in on time?
spacecadet From United States of America, joined Sep 2001, 3365 posts, RR: 13 Reply 7, posted (3 years 3 weeks 6 days 1 hour ago) and read 2866 times:
Quoting comorin (Reply 4): While cockpit automation has helped reduce crew and workload, has it made flying any safer? What does the data show?
The data shows a steady decline in accident rates attributed to all causes.
I'm assuming the OP's post was not just about cockpit complexity but really overall engineering complexity. But the thing is, with the extra complexity of the design also comes higher quality engineering and better build quality due to things like computer aided design and testing and improved production methods. You don't see planes just break up in mid-air anymore, for one example, which was not really unheard-of in the old days.
And all the extra complexity in the way airplane systems are designed is actually to lessen the pilots' workload. Pilots no longer have to scan dozens of dials in the cockpit and hand-fly their aircraft. One of the earliest "complex" systems fitted to airliners was auto-pilot, and they have only improved over the years to the point that now they're integrated with the FMC and can control the plane from takeoff to landing, freeing the pilots up to do other tasks. But the auto-pilot and FMC are themselves complex systems that early airliners didn't have. The airliners became more complex, but that reduced the chances of pilot error.
You could say the same for weather radar, GPWS, and TCAS. All complex computerized systems, but all there for extra safety.
Most of the added complexity in any aircraft design over the years is in response to a problem that needed to be solved. And the solutions have for the most part worked, even if it means airliners are not as simple as they used to be.
I'm tired of being a wanna-be league bowler. I wanna be a league bowler!
comorin From United States of America, joined May 2005, 4861 posts, RR: 16 Reply 8, posted (3 years 3 weeks 5 days 17 hours ago) and read 2646 times:
Spacecadet, thank you for the reply.
One factor at play here is also the trend towards growing machine autonomy - an aircraft trying to fly itself. I think Airbii were the first to do it with their rules-based 'laws'. It's the difference between riding a passive object like a bike, versus learning how to ride a fuzzy-logic Horse.
As for the comparison with the old Connies and their recip engines, I would say that they were more complicated rather than more complex.
Viscount724 From Switzerland, joined Oct 2006, 23184 posts, RR: 23 Reply 11, posted (3 years 3 weeks 5 days 6 hours ago) and read 2213 times:
Quoting Fly2HMO (Reply 5): I'm gonna go on a limb here and say airplanes today are overall less complex than airplanes of yore. Take a SuperConnie for example, each engine had 18 cylinders (aircraft total 72), which also equates to 144 spark plugs in the plane (two per cylinder) That's a TON of things that can break right there.
Or the 28-cylinder (2 spark plugs per cylinder) Pratt & Whitney R-4360 used on the Boeing 377 Stratocruiser and various military types. Total 224 spark plugs for the 4 engines on the B377.