Question: Your take on Fly-By-Wire, Power-By-Wire etc. in Light aircraft

[Turnerj]

Hey everyone,

 

I'm currently doing a case study for a uni course on Digital Fly-By-Wire (DFBW). Whilst researching I began wondering what the pilot's community view would be on the technology. Firstly a couple of facts for those who haven't a clue what I'm talking about:

 

1) In a DFBW aircraft, the stick/yoke (I'll use stick) is not connected to the controls directly, but rather acts as an input to a series of computers which take into account a range of parameters including (but not limited to) the airspeed, altitude and vertical acceleration (g's) and decides on an appropriate control surface position that would satisfy the pilot's request.

 

2) The stick now takes on a different role in the aircraft. When the stick is moved sideways, the computer interprets this as a request for a particular roll rate rather than a control surface position like in traditional mechanical/hydraulic control systems. The only exception is when the stick is centralised, this is interpreted as wings level and the computers will again solve complex differential equations to determine suitable control surface positions which will accomplish this. If you've ever watched any fast jet coming into land, you will have seen this illustrated. The elevators (which also act as ailerons) appear to 'twitch' as the plane is coming in to land. This is not because of the pilot, but rather a perfect example of the DFBW computers constantly chasing 'wings level' by adjusting control surfaces. All the while the pilot may be sitting there with the stick centralised.

 

3) As far as vertical (pitch) control goes, the stick is a 'g' setting, with centralised as obviously 1g and max aft position may be the maximum rated g for the airframe, or a little bit more. Depends on the aircraft. It should also be noted that this negates the need for elevator trim in the aircraft. Once a nose attitude is set and the stick centralised, it will maintain that position until the next input from the aircraft.

 

4) Power-By-Wire is also known as Full-Authority-Digital-Engine-Control (FADEC) this is where the throttle takes on the role of a power setting rather than a butterfly valve position. Mixture, Propeller setting and throttle setting are all computed by the computers to find the optimum settings for the requested power from the pilot.

 

The first (and rightly so) concern most pilots will have is system reliability. Typically these DFBW systems are quadruplex systems, meaning that 4 independent computers all create the same data independently. If one computer creates erroneous data (due to any number of reasons) that disagrees with the other 3, then it is 'voted out' by the others and disregarded completely.

 

The benefit of such systems are typically; reduced dry weight, ease of transition between similar types (using DFBW, it should be possible to make all the Tecnams, Jabirus etc 'feel' the same) and a reduction in pilot workload (due to the lack of need for things like trim inflight)

 

Also, usually the computers used for FADEC are separate to DFBW, but I may be wrong about this.

 

My main reason for asking is, with computers getting smaller and more powerful at an exponential rate, it is only a matter of time before this technology is possible in light aircraft. The Liberty XL-2 already has the world's first FADEC system in a light aircraft; the Embraer Phenom 100 (a 6 seat very-light-jet) has a some-what limited DFBW system as well as FADEC. It is coming.

 

I can't wait to see how this technology is currently percieved by the pilot community.

 

Cheers,

 

 

[turboplanner]

You are drawing some very long bows there. Digital can do anything, sizing has to be achieved, so there needs to be a market which will pay - either hundreds of customers who will pay millions for the equipment, or millions of customers who will pay hundreds for the equipment, so there's a low volume issue with recreational aircraft, and that's a fundamental which would have to be overcome.

 

....and it's parked out in the shed for a few months with the mice......

 

 

[Spin]

....and Bazza accidently cuts a control wire and rejoins it with a twist and a bit of duct tape......

 

 

[Guest ozzie]

....and Bazza accidently cuts a control wire and rejoins it with a twist and a bit of duct tape......

According to Qantas you supposed to use staples.

never considered FBW for light aircraft. Although working on ideas for electric conversion i have been considering wireless for the 'throttles'.

 

This would be an interesting project. Not being a 'geek' i'd have to take a guess and say the software is the secret to getting it working right.

 

Ozzie

 

 

[Bandit12]

Great concept for the big end of the market and military perhaps, but I can't see it being widely taken up in the light aircraft community. The initial expense of outfitting/purchasing aircraft would probably be a little prohibitive, and I can't see the average LAME able to work on it, which might mean prohibitive costs in maintenance too.

 

Given that the mainstay of general aviation are aircraft often nearly 30 years old, mostly using technology that is closer to 50 years old or more, I'd say resistance to change would be the biggest stumbling block.

 

 

[Old Koreelah]

Very interesting concept, and thanks for the clear explanation, Turnerj.

 

Eventually, if fly-by-wire filters down to our little aeroplanes then we will have truly lost the plot. (We already see autopilots being considered)

 

No more battling with quirky handling, learning individual engine management skills and enjoying the challenge of mastering different and demanding aircraft. Skills lost, interest gone, variation and challenge a vague memory from the good old days.

 

Bland, bland, bland.

 

 

[Guest davidh10]

While it is probably essential for some fighter jets that are actually unstable designs, and thus could not be flown manually, the level of automation already in regular public transport jets is becoming a concern. See the thread "Pilots forgetting how to fly". There's real concern that when the "automated systems" fail or are impaired, the pilot needs to be able to manually fly the aircraft. What you are suggesting negates that option, in addition to changing the actual function of the primary controls.

 

Also look up the "Gimli Glider" incident. Also refer CASA article. Had the pilot not also been a glider pilot and used to using side-slipping techniques to lose altitude, it it likely they would all be dead.

 

In recreational flying, most people fly for the enjoyment of the whole experience, including control of the aircraft and learning how to handle it. Just cannot see much interest, except perhaps for a few experimenters.

 

Most of the computers available in "retail" are not up to the quality standard from a hardware or software perspective that you would need in something like this, where you are going to bet your life on it always being able to fly the aircraft safely. I could see a bigger market in "autopilot" using servos that can disengage or be forced by the pilot., where the pilot can still fly the aircraft if all the automatics go pear shaped. To get the requisite quality will cost more than most recreational aircraft.

 

 

[Guest davidh10]

...Although working on ideas for electric conversion i have been considering wireless for the 'throttles'.This would be an interesting project.

Oh great. every time you get swept by radar, you lose control, due to the interference with the signal. or someone jams it with the microwave!

 

 

[motzartmerv]

Id much prefer to be in direct control of the aeroplane..

 

Although, yesterday I was flying from Goondiwindi to nowra, the forcast said moderate turbulence in area 20.

 

Around coolah the turbs was as bad as i have seen it. Very solid g loadings and some serious vertical gust's, and over 20 knott excursions on the ASI.

 

The Piper sports is not a nice aeroplane to fly in turbulence, its short coupling means that its not overly stable in yaw and pitch in the bumps.

 

It seems to get some aerodynamic feedback through the controls, particularly the ailerons. At one stage the stick whacked hard aginst my leg and I couldnt get it to centeralize, the aeroplane rolled over, past 90 degrees and I found myself looking at the world from an almost inverted position, whith a hell of a lot of G on. I unloaded by pushing forward and got the nose back above the horizen as the contents of the baggae area came forward and joined me in the front. It was only a momentary loss of control, but I was extremely worried about the fact that the ailerons forced the situation through the aerodynamic feedback caused by a severe gust.

 

I will be putting a report about the incident in today.

 

Not nice.

 

 

[Guest ozzie]

Oh great. every time you get swept by radar, you lose control, due to the interference with the signal. or someone jams it with the microwave!

I must remember to remove the MW from the Lazairs galley.

 

 

[facthunter]

Any loss of "feel" in a sport plane or fun plane makes it much less worthwhile doing it. You need to have as much knowledge of what is happening to it as possible. Any "feel' built into a FBW is artificial. The time may come when you will not be able to do anything directly for fear of not doing it as well as some microprocessor can do it, but hopefully it is a long way off. I've flown this stuff but you wouldn't really call it fun. IF you have 500 passengers behind you maybe the fun doesn't come into the equation, but in our sphere of flying, it is everything. I would also argue that in very small planes, because it enables feel, it is safer, because you fly better. ( Provided you know how to). Nev

 

 

[Bandit12]

A good point Nev. I must admit that I have rarely used an autopilot, and I much prefer an aircraft that has a good "feel" to it.....

 

 

[Guernsey]

A good point Nev. I must admit that I have rarely used an autopilot, and I much prefer an aircraft that has a good "feel" to it.....

I know what you mean Shane, I tried an auto pilot on my wife but it just didn't 'feel' right.

Alan Marriette.

 

 

[Guest davidh10]

I must remember to remove the MW from the Lazairs galley.

Just don't jettison the nice "hostie"

No. Seriously... Wireless communication of control signals can have poor reliability. I only have to look at the indoor/outdoor thermometer I have that works on 433MHz spread spectrum. Every now and then, there's no outside temperature displayed, then it resumes working, minutes later (sometimes 5 to 10 minutes) and the communication should be occurring every minute. Take the spread spectrum garage door opener... effective range varies from day to day and hour to hour. In some environments, I need to stand close, on the right side of my car to unlock it with the wireless key fob. My spread spectrum weather station, is more reliable, but occasionally drops out for a little while.

 

Wireless LAN equipment can be interfered with by microwave ovens... they operate in a "shared use radio band". Probably only a local interference (such as taxiing past a hangar with a microwave that is turned on). That was one of the reasons for the 802.11a wireless LAN standard, using a different band, however it has never really become ubiquitous. Current advances in wireless LANs are using "MIMO" (Multiple In, Multiple Out) using multiple antennas, so that it is likely that if one antenna has poor signal, the other will have good signal.

 

There's a big difference between "wireless" and "fly by wire", the latter of which, my car uses. The accelerator is just a potentiometer that is input to the car computer. The latter technology seems to be pretty reliable, these days, although my wife ended up selling a French branded car that had endless computer based problems, requiring it to be towed on three occasions. No support from the manufacturer :black_eye: In earlier years, we had an Australian car that lost power during cornering and had to replace two computers before finally selling it.

 

 

[flindersflyer]

If FBW were introduced then the entire intake and exhaust management systems would need to be reconfigured, more sensors and relays required etc. . Should an engine misfire for some reason, the computer management system may shut down the engine. This happens quite a bit now days in motor vehicles built after 1990. It would no longer be an easy task to identify the cause, especially by you the owner.

 

The aircraft would have to be hooked up to a scan tool to detect the fault code and then the fault traced back and rectified. Considerable knowledge of the system and components would be required to identify and rectify a fault, often requiring removal of a component and testing with a new component.

 

Becomes very expensive, especially if the tech doing the work has no used parts he can try, and you as the end user will pay for all the components fitted to fault find whether they were the cause or not. I could go on about this but cutting it short, it would turn a relatively cheap excerise of fault finding into a major cost. (EG: 2 yr old Mercedes shuts down. Roadside services can't assist. Vehicle transported to Mercedes dealer. Dealer has vehicle for 3 weeks whilst trying to track cause of fuel system shutting down. Nobody in Aus can figure it out. $100,000 equipment shipped to Aus from Germany. Tests completed. No finding fault. Another tech guy sent out from Germany and off the top of his head at an inspiration, he checks a mercury switch (installed to prevent fuel flow in case of roll over) under the dash. Switch mount has somehow come loose and switch is closed. Fuel system shut down. This was not indicated by the software diagonsis.) Imagine the cost. In this case to Mercedes but caused a lot of headaches for all concerned.

 

My thoughts are stick with easy methods until we're absolutely sure we can trust computerised systems which can identify cause of a problem to the actual failing component.

 

 

[turboplanner]

Oh great. every time you get swept by radar, you lose control, due to the interference with the signal. or someone jams it with the microwave!

No problems for Ozzie David, he'd just drop back on to the ground from his five feet fence jumper

 

 

[sain]

I play round with small microcontrollers a lot. These things are essentially a computer on a chip and the power and flexibility they give to designers is pretty impressive. For under $2.00 (two dollars) you can have a single chip with far more computing power and ram than was used on Apollo 11.

 

That puts the ability to do fly-by-wire type stuff into the hands of anybody who can solder a bit and program in c, c++ or assembler. and there are simpler languages available if you don't mind wasting a bit more of your program space.

 

How close is fly-by-wire for light aircraft? Well, the RC plane guys are going nuts with this stuff:

 

http://diydrones.com/

 

and http://www.youtube.com/watch?v=PJF-A3bJGuA

 

ArduPilot is a $24(ish) autopilot control unit built around the arduino platform (essentially an extremely easy to use microntroller + various interface bits). For about $300 you can have all the bits you need to turn a RC plane into a UAV that'll autonomously fly a GPS defined route.

 

I don't think it'll be that long before we start seeing autopilots installed in RA-Aus planes, and I think fly by wire will slowly come with that.

 

That said, I don't think I'd use it much personally. I enjoy the flying too much to hand it over to a machine, and I enjoy the sensory feedback through the controls. I fly for fun.

 

 

[turboplanner]

I don't think it'll be that long before we start seeing autopilots installed in RA-Aus planes, and I think fly by wire will slowly come with that.

The most interesting part will be writing the decision flow charts; you only have to come on here to see the huge disagreements about how to turn, how to stall, how to land etc to know that will drive some nerd mad.

 

And how will it be:

 

You're about to enter the runway, the engine stops and a message comes up "Boot error, restarting"

 

You decide to turn left, pull the stick over and a message comes up "Do you really want to turn left?"

 

You do your shut down checks and a message comes up "Do not shut engine down, important updates downloading"

 

Or you're comfortably cruising and you get a freeze, or worse, a blue screen shut down...........

 

 

[Turnerj]

Some interesting opinions here, great stuff.

 

It seems to me that most people are a lot more open to the thought of computer-controlled engines. Perhaps because it is nearly impossible to purchase a road vehicle these days without a computer controlled engine. So there is less of a fear of the unknown here.

 

Price will not be what restricts this technology, if you had been told 15, 20 years ago that your mobile phone will by 2010 be able to find you on a map, check you emails, submit a flight plan, watch movies, browse the internet, control a remote control helicopter and play all sorts of mind-numbing games. What would you have estimated the cost to be?

 

I think however, as Bandit12 said, Resistance to change will be what keeps this sort of technology away.

 

 

[turboplanner]

however, as Bandit12 said, Resistance to change will be what keeps this sort of technology away.

One of the key problems for Industry is that tertiary institutions turn out academics rather than specialists - a bit like buying a can of beans and finding out they are in a pod and you have to extract the beans first.

 

That costs our industries billions of dollars, and usually about two years before we can get an academic's feet on the ground so he gives us data we can use and sell in the form of finished product.

 

If you are doing a case study you need to get right inside the matter, rather than the throw away line you used; if that was true we wouldn't have accepted mobile phones or ipads, so you need to sift through a little more detail yet - keep at it

 

 

[eightyknots]

Id much prefer to be in direct control of the aeroplane..Although, yesterday I was flying from Goondiwindi to nowra, the forcast said moderate turbulence in area 20.

Around coolah the turbs was as bad as i have seen it. Very solid g loadings and some serious vertical gust's, and over 20 knott excursions on the ASI.

 

The Piper sports is not a nice aeroplane to fly in turbulence, its short coupling means that its not overly stable in yaw and pitch in the bumps.

 

It seems to get some aerodynamic feedback through the controls, particularly the ailerons. At one stage the stick whacked hard aginst my leg and I couldnt get it to centeralize, the aeroplane rolled over, past 90 degrees and I found myself looking at the world from an almost inverted position, whith a hell of a lot of G on. I unloaded by pushing forward and got the nose back above the horizen as the contents of the baggae area came forward and joined me in the front. It was only a momentary loss of control, but I was extremely worried about the fact that the ailerons forced the situation through the aerodynamic feedback caused by a severe gust.

 

I will be putting a report about the incident in today.

 

Not nice.

That sounds awful, MozartMerv. That must have resulted in a huge adrenaline rush.

 

The most interesting part will be writing the decision flow charts; you only have to come on here to see the huge disagreements about how to turn, how to stall, how to land etc to know that will drive some nerd mad.And how will it be:

 

You're about to enter the runway, the engine stops and a message comes up "Boot error, restarting"

 

You decide to turn left, pull the stick over and a message comes up "Do you really want to turn left?"

 

You do your shut down checks and a message comes up "Do not shut engine down, important updates downloading"

 

Or you're comfortably cruising and you get a freeze, or worse, a blue screen shut down...........

Here are a few more error messages that could worry aviators:

 

Hardware failure in port or attached device. (I hope the starboard one still works)

 

The request has timed out. (how do I reset that digital clock?)

 

Unknown Error (well then, how can we solve this kind of problem?

 

The structure size is incorrect. (if only I had a larger wingspan)

 

The route is not available. (I can’t see any witches hats)

 

An operation is pending (where’s the doctor?)

 

Error attempting to create the destination file (well then, where are we going then?)

 

Invalid Drive (perhaps I need a re-drive for my aeroplane?)

 

You must be an Administrator to use this application. To use this application, you can log on as an administrator, or contact your technical support group for assistance. (who was the administrator again?)

 

 

[Guest davidh10]

... "Do not shut engine down, important updates downloading"Or you're comfortably cruising and you get a freeze, or worse, a blue screen shut down...........

or the classic..."An unknown error occurred!"

 

...It seems to me that most people are a lot more open to the thought of computer-controlled engines. Perhaps because it is nearly impossible to purchase a road vehicle these days without a computer controlled engine. So there is less of a fear of the unknown here.

No actually it is because:-

 

• The technology has matured and is not as problematic as it used to be and;
   
   
  
• Most engine management systems have a "limp home mode" that allows you to drive home at 25kph with no performance and;
   
   
  
• If the engine management module actually dies completely and stops the engine, you can just roll to the side of the road and call a tow truck.
   
   
  

 

 

Aircraft engines run at about 80% of full power most of the time or you descend. Having it go, for example, from 4,200RPM (level cruise) to 3,000 RPM (final approach power) or lower as a "limp home mode", would be an air emergency, causing a "mayday" call and a forced landing. A forced landing is almost certainly going to be somewhere other than an aerodrome, and statistically that means there is a high probability of damaging the aircraft significantly and sustaining minor injuries. There is of course a statistical probability of serious injury or death, albeit lower, if you weren't over tiger country at the time.

 

Risk assessment is a two dimensional matrix. We will often accept a risk that may have serious consequences, if its likelihood of occurring is infinitesimal, on the other end of the scale, a risk that is likely to be realised often, will not have many takers unless the consequences are trivial.

 

Lets face it the risks are not assessable by the end user in the aircraft fly-by-wire case.

 

The automobile case that may be vaguely similar, is where the Japanese have been testing computerised collision avoidance systems that now not only remove power and apply the brake, but also take over the steering and attempt to drive around the obstacle. Now that would scare the pants off me. Imagine the case where it steers you into a head-on collision because a pedestrian jumped in front of you! I don't think we are going to see this in our cars any time soon.

 

Let's take another automotive case. Anti-lock brakes are on all new cars. I almost had an accident caused by anti-lock brakes, where there was a typical scenario of someone jamming on their brakes on a freeway and everyone behind has a progressively shorter distance to stop. I had applied the brakes to just short of locked and was slowing nicely until the front tyres went over the metal fish plate across a bridge road joint. The tyres of course locked, but would have gripped on the other side of the fish plate 9about 30cm) if it weren't for the anti-lock brakes that released braking power and lost be two metres of braking distance. I stopped about 30cm from the car in front. Without anti-lock brakes, I would have been two metres from the car in front. Anti-lock brakes have been mandated by the "nanny state" because there's so many drivers who don't know how to brake effectively under differing road conditions. My car does not have them fitted and I know that some people remove the fuze to disable them.

 

 

[Yenn]

From your description of how it works I cabn see that fly by wire would be contra indicated for recreational pilots. If we got used to a central stick levelling the wings, we would lose the ability ot fly if the computer collapsed.

 

When computers are as reliable as the taps in your bathroom and equally as cheap they may be an option. I have watched an auto pilot in action in an ultra light and it impressed me with its accuracy, but I could see the downside of it leading a pilot into trouble.

 

 

[Litespeed]

If FBW were introduced then the entire intake and exhaust management systems would need to be reconfigured, more sensors and relays required etc. . Should an engine misfire for some reason, the computer management system may shut down the engine. This happens quite a bit now days in motor vehicles built after 1990. It would no longer be an easy task to identify the cause, especially by you the owner.The aircraft would have to be hooked up to a scan tool to detect the fault code and then the fault traced back and rectified. Considerable knowledge of the system and components would be required to identify and rectify a fault, often requiring removal of a component and testing with a new component.

Becomes very expensive, especially if the tech doing the work has no used parts he can try, and you as the end user will pay for all the components fitted to fault find whether they were the cause or not. I could go on about this but cutting it short, it would turn a relatively cheap excerise of fault finding into a major cost. (EG: 2 yr old Mercedes shuts down. Roadside services can't assist. Vehicle transported to Mercedes dealer. Dealer has vehicle for 3 weeks whilst trying to track cause of fuel system shutting down. Nobody in Aus can figure it out. $100,000 equipment shipped to Aus from Germany. Tests completed. No finding fault. Another tech guy sent out from Germany and off the top of his head at an inspiration, he checks a mercury switch (installed to prevent fuel flow in case of roll over) under the dash. Switch mount has somehow come loose and switch is closed. Fuel system shut down. This was not indicated by the software diagonsis.) Imagine the cost. In this case to Mercedes but caused a lot of headaches for all concerned.

 

My thoughts are stick with easy methods until we're absolutely sure we can trust computerised systems which can identify cause of a problem to the actual failing component.

Sounds exactly like my brothers experience, new merc c class.

 

Car just failed to keep running after only a month on road. 6 weeks of angst and entire fuel system, injectors, pumps etc replaced. No one had a clue what was wrong.

 

It seemed ridiculous to me- they spent a huge amount of money and time and still no fix.

 

No one seemed to think of the obvious- check the bloody computer. Turns out a reboot of system- a very easy and sensible thing to ensure system is working was never tried.

 

Dumb asses, 15 mins work and car performed perfectly. Well over ten grand was spent for a 15 min fix.

 

Quality of systems is something merc have really dropped in recent years- I doubt my brother will ever trust another merc and certainly not their service techs.

 

Meanwhile my 12 year old V8 BMW has never stopped, never had a computer problem. Yes It has a full computer system and fly by wire throttle, but stability control etc can all be turned off with one button- the god button.

 

Some systems are well designed and rarely have problems and some are the windows of cars.

 

I would be real worried if windows had control of whether I die or live.

 

 

[Litespeed]

Back to actual aircraft-

 

I am happy to have systems that assist in navigation etc but only with a complete back up system of actual maps etc.

 

I have seen the uav stuff for radio control, very good stuff- but a failure is not a big risk in that field. In a real aircraft, it could be fatal.