There is another aspect which needs to be addressed.
This is the systems/control theory. Granted I do not have specific experience on airplanes but we use the same control systems in power plants, chemical manufacture and oil refineries.
Mathematical/Control systems view : All systems are non-linear in nature (I.e. a small increment in inputs doesn’t produce a proportionate increment in output). However most non-linear systems can be linearized using a Taylor series expansion around an operating point. However, to get to an operating point, you’ll need a human or complicated computing. For example : getting a car from 40 mph to 45 mph is pretty easy, but getting it from off to 5 mph is not. The relationship between the gas pedal and the car speed is linear in the former case but not in the latter.
If you consider the key metrics in an airplane , engine speed, oil pressure, etc etc, they are all step functions or the like during startup or shutdown. Also comparing an airplane to a car is not representative because they have different dynamics at play, impact of failure is different and there are more sub-systems involved.
Human in the loop : Complex systems always benefit from human in the loop. Not all problems can be anticipated and mitigated by control systems - an experienced human is much more aware of conditions on the ground and can anticipate failure modes and check accordingly.
So for complicated control systems, like an airplane or a chemical plant, there will always be human involvement in the startup steps. Some people may move the startup steps to “pre-startup” whereby a human checks an hour or day before startup but it won’t be completely eliminated.
But aren’t all 2 engine jets engineered to be able to take off even if one engine fails? Taking that a stage further, could you cut the time by only starting one engine and then taxiing away? Or would the flight computer prevent initiating take-off with only 1 functioning engine? Obviously would never happen in real life, but if you faced certain death by remaining on the ground, might that be an option?
All 2 engine jets can successfully continue a takeoff if one engine fails late enough in the takeoff that stopping is less safe than flying. In theory given a long enough runway you could make a purely single engine takeoff, although not legally. In practice few runways are that long for a normally loaded twinjet. An empty jet with a light fuel load on a long runway in an zombie apocalypse? That might be a gamble worth taking.
Back in the days of 4-engine jets some of them were fully certified to be ferried (no payload) with just 3 functioning engines. The idea being to fly a hop from a place you couldn’t get that engine fixed to a place where you could. But that was dealing with a 25% reduction in available power, not 50%.
Starting 1 engine near the gate, then taxiing away and starting the other engine while en-route to the runway is a standard approved practice. Management loves it because it un-crowds the ramp, saves fuel, and saves wheel brake wear.
Pilots are more ambivalent about the practice because it increases the odds of damaging something or someone on the ramp with your greater jetblast to get moving. Also because the time and attention spent starting the second engine diverts attention from taxiing safely, watching out for other airplane, trucks, etc. As well, unless there’s a backlog at the runway, much of the taxiing time is already fully occupied with other checks & procedures; shoehorning an engine start in there as well increases the likelihood that something else gets goofed up.
Having said that … If the runway is backed up, or you need to sit on the ground due to traffic or weather at the destination it’s not uncommon to do most of that taxi single engine. And to shut both down if you know you’ll be sitting awhile with no need to move. In each case it’s a matter of balancing the time pressure vs the fuel savings.
At least on the (all non-Airbus) airplanes I’ve flown, the computer won’t notice or alarm for an attempted takeoff with an engine shut down. Nor for the wrong flap setting as long as flaps are set to a possible takeoff setting. Nor for starting the takeoff roll from an unexpected = incorrect point on the runway versus the plan. The last of these is relevant to your (?) posting in the other thread about that Airbus at Lisbon.
Yes, that was me - thanks for your continued patience in laying these things out so clearly for the layperson. I appreciate we are talking about extremes here, well outside operational norms.
Procedures that push the envelope are always a concern though. One of the Admiral Cloudberg write-ups I read recently was about a second-rate pilot for some African airline who had an engine failure mid-flight and didn’t even declare an emergency, instead intending to land at his final destination. Well, he kind of did that - about half a mile short of the airport, when his second engine (of 2) also failed. No doubt cost-saving was part of his thought process in the original decision. He’s not a pilot any more, very sadly nor are any of the 152 others on board nor the six on the ground where the plane crashed.
I forget where I saw it (a YouTube video I think but I cannot find it now) where they noted that losing an engine on a twin engine prop doesn’t cut power by 50% (technically it is 50%) but in effect it is more like 70% or more due to significant drag from the non-running engine (even assuming you feathered the blades).
A big issue with trying to take off on one engine is that below a certain speed it is not possible to keep straight on the runway with just one engine at full thrust. So you’d have to incrementally increase thrust on the one engine only going to full thrust once you were fast enough to maintain control. This would use even more runway than just from having half the thrust available. Like @LSLGuy says, possible in theory but not practical.
Shortly before I stopped flying the BAe146 I crewed a 3 engine ferry flight. The technique was to hold the brakes and set the asymmetric engine, ie the one whose opposite-wing partner was dead, to take-off thrust and mark the thrust lever position on the quadrant with a pen. Then you reduce the thrust on that engine to 50%. The two symmetric engines (e.g. both outboards or both inboards depending on which one was dead) were set to take-off thrust. Then you let it roll and once it’s up to a good speed (60 knots from memory), you increased the third engine to take-off thrust using the pen mark as a quick reference.
It takes more time dicking around with thrust settings than it would’ve taken to just start the 4th engine.
It does make for an odd looking engine instrument display in flight.
Number 4 was shutdown. You’d normally only see that picture in an emergency. Taking off with that set of caution lights was a little bizarre.
Interesting - I would have guessed it was more like put the single engine on full thrust, and the two on the other side on 50% each, then gradually increase those as you built up enough speed for the rudder to be able to correct the asymmetry. But of course, as we discussed in the other thread, taking off in a big plane isn’t usually about just pushing the throttles as far as they will go, there are calculations and stuff!
The challenge with that approach is that there’s really no engine gauge that directly measures percentage of power output. We have gauges we use to set a computed power amount, but the calculations handle the indirection between "I have 20,000# thrust engines and I need 17,000# of thrust so I need to set X% RPM or Y Pressure Ratio or Z degrees exhaust temperature or whatever.
Whether you’re measuring RPM, engine output pressure, or temperature, there’s a very VERY non-linear relationship between those values and the thrust produced. e.g. WAG made up numbers: 30% of RPM is 10% of thrust, 80% of RPM is 50% of thrust, 90% RPM is 80% of max thrust, and 100% of RPM is 100% of max thrust.
And all those number slide around differently depending on weather, age of the engine, and of course are totally different between different engines on different types of airplane.
So as a practical matter there’s no way to set “50% power on both engines on the good side”. But as Richard explains, there is a way to set 100% power symmetrically on the matching good engines.
I recall an accident back in the 1960s maybe early 1970s involving a 3-engine ferry of a cargo DC-8. Bottom line is they got the 3rd running engine up to full power too early, started sliding uncontrollably off the side of the runway then yanked it into the air to avoid that crunch. And duly stalled & crashed from about 100 feet up. Ouch! At least it didn’t hurt for long.
My post was poorly worded and that probably led @Dead_Cat astray (). The engine whose opposite side partner is failed is set to 50% N1, not 50% thrust.
Here is a video recently posted to YouTube that illustrates what I was thinking about when I wrote the OP (this is for a single-engine turboprop). Note it is 54 steps:
That reminds me of something I’ve heard a few times: Back in the 2000s, when Aloha Airlines was still in business, they used old 737-200s for short hops between the islands, and modern 737-700s and -800s for flights between Hawaii and the mainland. Supposedly the reason the kept the old -200s for the interisland flights rather than replacing their entire fleet with the 737NG was because on the very short, high-frequency, short turn-around flights they did between the islands, the high-bypass engines didn’t have enough time to cool down between takeoffs. Supposedly the JT8D engines on the -200s were more tolerant of these very short, frequent flights.
I don’t have 17 odd minutes to watch it all, but from the first bit it looks like he’s doing the full cockpit preparation, which isn’t really what’s required to start the engine. It largely comes down to a) making sure everything is on, and b) making sure everything works. Why isn’t everything just hardwired to be on with the battery? Because you need to be able to turn things off if they are malfunctioning. This isn’t necessary in a car. Why do you need to make sure everything works? Because if it doesn’t, you may be flying illegally and/or it might kill you. Most car systems are far less critical and a driver should be checking they work anyway but many people are lazy and don’t bother.
Yeah, in theory (in winter at least) you should check your deicing system before going for a drive, but virtually no-one does, because if it fails and you can’t see, you can pull over to the side of the road if you end up not being able to see. Whereas if your deicing system fails on a flight, you might kill a bunch of people including yourself.
Not gonna watch the video, but your point has been well-answered. Of those 54 steps about 4 of them are “starting the engine”. The rest are “Prepare the aircraft for flight”.
In your car you should walk around it completely before every drive to see if any tire is flat or has obvious cuts on the sidewall. Also to see if a rock has taken out your radiator or a headlight. Looking underneath for any puddles is also smart. The same applies for all the sit down, strap in, & make everything ready to move steps in a car.
Bottom line: airplanes, even simple ones, are more complicated than cars. In the case of a multi-million dollar TBM, vastly more complicated than a car. As I said up-thread, a better technological comparator would be a 60 foot yacht with all the complications of all the installed systems & toys.
As well flying is vastly more complicated than driving with hazards that are orders of magnitude greater. Which in turn requires professionalism from all pilots, even the weekend hobbyists.
Then when you add the 1950s technological stasis imposed upon much of the industry by the FAA and by limited production runs that make R&D unaffordable, the result is mechanical complexity for the operator.
An analogy I have always used in manual vs automatic car transmissions. One is hard(er) to use but is dirt simple internally. The other is trivial to use but 20x more complicated internally. And much more expensive to build.
I want it to be known I was not posting that to “prove” anything and suggest you and others were wrong. I merely happened across it by chance (because I have been looking up plane videos since I started playing the recently released Microsoft Flight Simulator 2020) and I thought it illustrated what I was looking at when I posted the OP really well.
Honestly, I have been restraining myself from asking a dozen other flight related questions.
Thanks to you and everyone else who has posted great stuff in this thread. I certainly appreciate it.
As to representative, you picked a pretty fancy airplane. If you find a vid of a basic Cessna or Piper or Diamond trainer being prepped for flight that’s a much closer analog to firing up your Camry for a run to the grocery. Still a bunch more steps than a Camry though.
Start as many threads as you like. It’s not like we charge by the word. We also have this omnibus thread for all things about light planes and a smidgen about the big iron.
It’s 9 years old with 1349 posts on an endless variety of airplane topics with the most recent just a couple weeks ago. There’s a treasure trove of trivia and air adventures in that thread.
Don’t forget that once the thread is on your screen you can hit the [End] key to jump to the bottom in just a couple seconds. (That’s on a PC, there is a some Mac option-shift-whatever equivalent).
It seems I did but that was by accident (the plane is modeled in MS FS 2020 which is why I was looking). I had no idea where in the GA field it fit beyond a bit nicer than most till you get to the super-bling jets.
Till about a week or so ago I had never even heard of “TBM” or the company that builds them.
I think it’s safe to say that a successful flight in an aircraft depends far more on what happens before the flight compared to a successful drive in a car. When driving you can just jump in and go. Even if you are in an unfamiliar city, you can set out for a drive for a day without having any significant issues. And if you do have an issue on a drive, like you get lost or something, then you just pull over, stop, and sort it out.
Once you take off in an aircraft, the world starts going by at a pretty rapid speed, and if you haven’t prepared properly you soon find yourself getting “behind the plane”. Being behind the plane (or “hanging on to the tail-plane”) is a bit like going for a drive in a car and realising you were supposed to take the turn you just passed, and then realising that you could’ve fixed it by taking the other turn you just passed, and then realising that another way to fix things would be to take that third turn that is now behind you etc etc. So the pre-flight stuff tends to be more involved because if you do it right then the actual flight is easy.
It doesn’t always have to be that way. If you own your own plane and want to go flying without interacting with the aviation “system”, i.e., you won’t be talking to ATC, no controlled airspace, flying visually in an area you are familiar with, then you can just jump in and go the way you can with a car. But the procedures we follow are “catch-all” procedures that should be suitable for a flight between any two airports in the world.
I found more professionally done startup videos, but the ones I looked at featured newer aircraft than most people fly. One from Embry-Riddle had some sort of battery test I’d never heard of, then went to a Garmin panel. A 172R video showed a fuel pump being turned on. The 172s I flew just had gravity feed. (Of course low-wing airplanes use a fuel pump.)
But yeah, there are a lot of start-up videos out there.
). The engine whose opposite side partner is failed is set to 50% N1, not 50% thrust.