glum

I'm currently studying an MSc in avionic system design at Cranfield, with a focus on cockpit / flight deck design.

The first step is to gather top level system requirements, and as well as trawling through the EASA requirements there is an obvious need to gather input from the end user - pilots.

I'd be grateful for any input anyone cares to give, and if it could read along the lines of

"the cockpit shall..."

(ie. allow all emergency controls to be operated by either pilot whilst seated and with harness fastened.)

or "The display system shall..."

(ie. be readable in direct sunlight / be reconfigurable in case of failure etc)

Then so much the better.

Thanks in advance!

Two_Squirrels

You will find that there are relatively few detailed requirements in EASA/FAR regulations. A good source of the sort of thing you might want to use as a starting point is 00-970 Design and Airworthiness Requirements for Service Aircraft, Although military biased obviously, it should give you an idea.

212man

I'd suggest looking at the FAA AC-25 and AC-29s - the guidance material for Certification of Large Aeroplanes and Rotorcraft. A sample from AC-29 extracted below:

Genghis the Engineer

For what class / role of aircraft? The role is paramount in any kind of cockpit design.

G

Nigd3

You could try CS (or FAR)-23/25/27/29 and then filter down to the guidance material that is relevant to the desired class that the aircraft comes under.

In more general terms, here are some starters:
FAA AC25-11
FAA AC23-1311
FAA AC23-1309 (for compliance to the safety aspects)
AC 23-1523 (crew requirements)
GAMA Spec 10 and 12

The above should be able to send you off to other references useful but as Genghis pointed out you really need to define your aircraft class and role, as this will affect your requirements and cert basis.

glum

Thanks for the pointers so far. Our design project is for ... a Tiltrotor (think V22 Osprey, or Bell 609)

Whilst I can read endless reams of certification requirements (and I am) what I'd like from 'real pilots' is your personal requirements - bit of blue sky thinking perhaps?

Stuff like:
"I want a coffee cup holder"
"I want individual heater controls"
"I want furry dice!"

Things I can't capture by reading books...

Thanks chaps!

Hedge36

I'd like a glove compartment perpetually stuffed with $20 bills, please

bfisk

A proper chart light, individual, with easy one-hand control of intensity, red/white color fiter and beam width. Some manufacturers (Beechcraft...) think the paperless cockpit is here -- guess again. It's not yet, and I don't want to use overhead floodlights to do the plog/lmc/company charts not on IFIS etc.

Genghis the Engineer

In a tiltrotor, I think the first and biggest item I'd be worrying about would be rapid emergency evacuation and structural crashworthiness above and beyond any airworthiness standard minima. Second may well be vibration absorption and isolation.

G

captshekh

I haven't seen an aeroplane with proper sun shades that can cover the panels completely. partial covers allow full light intensity at some corners and during head movement the large variation in intensity is discomforting.

At the moment news papers, though not desirable, serve the purpose pretty well.

glum

Genghis:

You're absolutely right, and there are some HUGE risks on this project, but fortunately I'm an avionics chap not structures.

We've taken a long hard look at the Osprey problems and are aiming to incorporate their lessons in our design - Bell have been struggling for years to get certification despite pausing their developement to learn the Osprey lessons.



Thanks for the other inputs - exactly what I was hoping for.

Pilot DAR

Knowing the subject aircraft type, and discipline helps.

In a tilt rotor, I would expect that at certain critical phases of flight, I would want the aircraft systems to take care of themselves, and only notify me when my action will be required to intervene with a decision the systems cannot make.

I do not want distractions at critical phases of flight. I most dislike having to divide my attention both inside and outside at the instant of lift off, to optimize engine power, when it should be able to optimize itself. Example of bad: AS355 helicopter, where just as you reach maximum power as you pick up, you have to have to look at the power indicator to match the engine powers to each other with a switch on the collective. Good: EC120, whose power display for it's single engine, graphically displays the percentage of the most critical engine limitation only.

Thus, I would like a tactile limitation marker, it would work as follows: As you apply power (be it the raising of the collective on a helicopter, or moving the power lever(s) on an airplane), a electric motor driven "bumper" which the pilot would feel at the point where the "system" regonizes the first limitation for that engine, would be motored to the limitation point. If the pilot needs more power, he/she can still pull/push past this tactile bumper and get it, but now you know that you're pulling too much, and it's gonna cost a lot! If there is a "5 minute" rating, it could start to time it at this moment. Maybe the control would have a subtle shaker to tell the pilot they are within a timed limitation. The display should show the relative power being demanded, and the position of the tactile stop at that moment. Interestingly, the Schweizer 300 helicopter has a very basic form of this, which is simply two fixed position rivets on the collective friction slide, you'll feel them if you pull through it to overboost the power, but you still can if you really need to.

Surley a sharp avionics guy can design a system where the sytem computer knows what the engine(s) are doing, how much additional power is available, and electrically drive a tactile stop to that point. The pilot will feel it, and can pull through, but has a sense of remaining within limitations, without looking inside at all. The DA 42 has somting like this for pitch limitation, though you cannot pull through it - it's just there!

The system can further assist the pilot by repositioning this tactile stop, in the case of an engine or other failure, which effects total power. In the case of exceeding a "5 minute" limitation, the tactile stop might be motoring back down below the presently set power. As it passes down, it will bump the control down, and the pilot would feel it. Again, maybe a small shaker. This would be accompanied by a display on the critcal power indication display, so the pilot can look at where things are. There are systems with tones for this, but we already have too many tones in the cockpit.

Presuming the tilt rotor has a height velocity curve (I'd sure be courious to see what it looks like!) I would like a computed indication of my phase of flight relative to the HV curve, for the prevailing contitions. Sort of like an angle of attack indicator for a fixed wing aircraft. Green is good, amber use caution, red - do you really want to be doing this? Again, no tones, just a very simple display, in the field of outside view. This display might also offer HIGE/HOGE information, when the airspeed drops below a certain value. (You can/cannot hover out of ground affect right now).

A tilt rotor has some very uncommon failure modes, and related pilot actions. Do you glide or autorotate? If you're gliding, can you turn it into an autorotation at the required moment, or do you smash off the rotors as you land horizontally. The avionics should be propmting these actions, and offering the tolerances available for a successful outcome.

Presuming that the display is offering some kind of terrain awareness, this must be easily disabled when desired. When in operation, the "system" should be constantly displaying as a moving, variable dimension cone, where the aircraft could be in the next minute with the power available (out climbing or out turning terrain or other hazards). The G 1000 does this in a simple way, in the mapping view - I like it!

If you can figure out how to build it, let me know, I'd love to fly it!

Genghis the Engineer

Ah, but you're on the Cranfield MSc, which is probably the best regarded product of its type globally. It didn't get that reputation by making life easy for the students!

Check your PMs - I don't think that there is any substantial tiltrotor experience on here (although I'd certainly pay much attention to Pilot-DAR's comments above, he is a very experienced flight tester of smaller aircraft), but I've pointed you at where the real expert is.

Also, as an appropriate student, you are eligible for student membership of SFTE at about US$15/year - join, then see what's available on their papers database. (Society of Flight Test Engineers)

G

Milt

Don't overlook the need for the "Standard Man" to be able to operate everything and I'm still looking for the standard woman.

Genghis the Engineer

I haven't looked them up for a while, but NASA used to publish some really nice anthropometric tables showing 5th, 50th and 95th percentile caucasian male and oriental female proportions.

I'm not sure where they exist any more, but doing a quick search this should provide you with a reasonable start point.

G

peeush

As far is avionics interface in cockpit goes, I would suggest that an intelligent mix of tones and voice warnings must be considered while deliberating on each 'Phase' of flight likely to be undertaken- i.e hover, forward flight (airplane mode), rearward flight (possible ?), hover turns, glide/autorotation and of course vertical pick up and sit downs. While keeping the certification requirements in focus, the pilot-machine interface will finally depend on the missions envisaged and the cockpit (crew) load while undertaking these acts. Thus, consider each phase of flight separately for all the missions listed out by the user agency. Prioritise the parameters through a table and then select the best feasible option for pilot interface while allowing him to focus his eyes at his priority without jeopardizing flight safety/ life of the expensive components. As an engineer you can only optimise the technology for your machine however, making use of all the sensory organs available with the crew (ALL of them) will provide edge over the contemporary. Please remember that even the flight parameters need to prioritised for every phase of flight and using HUD/HMD may not be bad idea. So the voice warnings, tones, tactile cues must be employed together while providing our pilot with only the information necessary at THAT MOMENT.

Hope it helps !!!!!

glum

Wow, some GREAT posts gents, thank you.

Pilot DAR, you're caused some very creative juices to flow!

Since our aim is to simplify the job of flying as much as possible on such an odd beast, our focus will be on automating many things such as the transition from hover to forward flight and back again (a lot of algorithm writing ahead!)

One of the worst problems for the V22 as I understand it is the susceptability to vortex ring instability caused by descending too quickly - exactly the kind of thing which needs warnings / autonomous control.

Having done some reseach into the training being undertaken with pilots whereby the displays will 'fog out' all non essential information should an emergency event occur, I believe this is definitely a way forward in display design. Again, it will need careful thought and prioritising of dangers (is the engine fire or terrain going to get you first?)


One of the idea's we've gone for is to remove the rudder pedals, since it is a fly by wire aircraft and the 'monkey downstairs' can take care of co-ordinating turns. Any yaw required during hover mode can be actioned with a twisting motion on the joystick. Is this a step too far regarding conversion from other types perhaps?

ICT_SLB

Glum,
The conversion task for fixed to tilt rotor may be overstated. I used to work with the project pilot for the Canadair CL-84 "Dynavert" tilt wing. He said an competent pilot could convert in a couple of hours. This older program might be a good source overall as well. Similarly I think you'll find that, once you're in a Vortex Ring, it's often too late - the best action is avoidance not any automatic takeover although I'm not sure if there are defined parameters like there are for Ground Proximity or Windshear.

As one who has worked his way through an EICAS development, the one question you must apply to any system alert is "What is the pilot's next action?" It's surprising how many Warnings come down to just a Status message if there's nothing he needs to (or can) do when the failure occurs. There was a very well reasoned CAA publication on alerting & flight phases but I'm afraid I don't have the reference here at home.

Pilot DAR

I sure agree with that! So many times I've wanted to ask an avionics designer, at the very moment of a tone or other warning/annuciation/light; "Yeah, so what is it you actually want me to do at this moment?". We can be overloaded by information, and a person's total capacity can be quite variable depending upon what's going on at the moment. The avionics designer sits in a quite room, considering all the things he wants the system to tell the pilot at any given moment, and how. He/she may not adequately consider that additional avionics which will be added to the aircraft later, which will have all of their own tones and alerts too, along with the busy environment the pilot is in when he's actually flying. Those few tones and lights going off, coupled with the comm radio, poor weather on the horizon, rough air, the pilot thinking about explaining to the boss that he might have picked off a tree branch with a rotor balde and damaged it, and having to get back in time for his kid's soccor.

Recall the basics: "aviate - navigate - communicate". Those are the priorities. There have been many times I have tuned out "communicate", and even "navigate", to assure that aviate was properly accomplished. I can recall during long line training at a controlled airport, telling the tower I would remain in the area assigned, and turn the radio down, so I could concentrate. He agreed, and I concentrated.

It is imprtant that as much as possible, the pilot be able to cancel tones and alerts, so "aviate" can predominate. And, as said, if there is no action the pilot can take anyway, tell him later!

While considering all of this, also consider what the pilot will have on his/her side, when the annuciation system itself fails. If the pilot is expecting a warning from the system, should a certain parameter be exceeded, but the warning sytem itself has failed, how does the pilot know that he must now monitor that function directly?

I'm glad I'm not an avionics engineer, so I can sit here and be critical, without having to have to figure out the solutions!

John Farley

Which is you are trying to do? Sorting out a cockpit and designing flight control systems are two VERY different objectives.

I don’t know who has told you that controlling the transition on a vectored thrust aircraft is difficult, but I can assure you it is not. In visual conditions you do not even need to refer to an instrument in either the accel or decel case. In cloud you need to keep your wings level and your pitch attitude roughly flat with reference to the attitude display to prevent any risk of the wing stalling angle being reached. (This attitude control is what you do by simple visual references when these are available). That just leaves height control.

For HEIGHT control the accel and decel need two different techniques:

Accel - from a VTO leave the power set as needed for a gentle climb or from a steady hover increase the power to give a slight positive climb. Then leave the power as set and push the nozzle lever forward (or use the nacelle pod angle trim switch to reduce the pod angle) until you reach full travel on the control by which time you HAVE to be flying on your wings.

All that is left to discuss is the RATE at which you rotate the thrust vector. That is SO simple it hardly needs describing – move the nozzle lever too fast and you will sink, move it too slowly and you climb unnecessarily high. So during the accel your nozzle lever rate of movement is your height control. Visually this is obviously a piece of cake while in cloud it behoves you to keep an eye on the VSI.



This diagram shows that in the hover if you rotate the thrust vector 15 deg aft of the hover angle you IMMEDIATELY generate a 0.26 g longit accel but thanks to the trig relationship you only loose 0.03% of your vertical component – hence the desirability of the slight initial drift upwards from the VTO or hover. Pure magic and on first acquaintance it appears to mean you get something (0.26 g accel) for nothing.

0.26 g is of course quite a brisk accel and soon leads to airspeed and wing lift. Once this builds a climb will result unless you continue to rotate the vector. This is all a lot easier to do than write about. Many Harrier pilots in two nations were converted to doing this with no simulators and no two-seaters without trouble.

Decel. Transition – in normal wingborne flight go downwind at normal circuit speeds (say 1.3 to 1.5 stall speed – it is not at all critical) deal with gear and flaps as usual and turn base. Roll out a mile or so before you want to hover and SELECT the thrust vector to the hover position. (This is a true ‘selection’ not a flying control type of action).

Once you do that you have chopped off all ‘thrust’ and the aircraft will slow down quickly thanks to the drag. NOW IN YOUR HEAD HOVER the aircraft and ignore the speed which is still present. IE control height with power and maintain the hover attitude. As speed and wing lift washes off increase power to maintain the approach path you require until you find yourself in the hover.

With practice it is easy to select a particular hover point by small adjustments to the pitch attitude or thrust vector angle, but that is exercise two not one.

JF