LCA News and Discussions

SaiK · Post by **SaiK** » 30 Sep 2013 20:15

yes, the LCA has them in redundancy layers from physical to transport levels. i think, the discussion was all about the power to signal the wires - the control plane, and what if scenarios... leading to gliding the a/c to safety without the engines. now at what speed the design makes it to break the limits and the drag would let it just fall like a rock is the question. or did i read the point differently?

indranil ji?

PratikDas · Post by **PratikDas** » 30 Sep 2013 20:29

The discussion on the APU, accumulator and redundancy has been great!

nikhil_p · Post by **nikhil_p** » 30 Sep 2013 20:29

The signal processor (I will use this term) will have backup systems in three - four different locations. Each of these are LRU's and will have the flight data coords fed into them before the flight. Each system (electronic) will have access to the main and APU power post which they will shift to batteries (even a 9VX2) will be able to support the electronics and these are not heavy. The 'glass' displays will cut off leaving only a manual backup (probably a horizon indicator, direction indicator and such). leaving enough power for the control systems...

SaiK · Post by **SaiK** » 30 Sep 2013 20:34

to power the actuators (electric/hydraulic) is much more than just enough power for the controls.. so, we need to think a little bit more on the needs.

Lalmohan · Post by **Lalmohan** » 30 Sep 2013 21:56

prem kumar - look up "ACARS"

Lalmohan · Post by **Lalmohan** » 30 Sep 2013 22:02

singha spoke wisely, without any electrical power, the lca will tumble out of the sky
it needs the flight control computers to maintain it in stable flight, no juice in batteries, no gliding

SaiK · Post by **SaiK** » 30 Sep 2013 23:38

redundant APUs is the way to go... comes with cost and weight penalty.

Indranil · Post by **Indranil** » 01 Oct 2013 00:21

nikhil_p wrote:All hydraulic systems, esp. on a/c are built with a certain amount of redundancy. On a car for example, the hydraulic brakes/ power steering/ clutch all have a master/slave setup. The master and slave cylinders are both diagonal split with two intakes and two outlets split by a diapraghm or one way valve. When the pressure on one side of the diagonal is less than the other side the valve actuates as it is probably due to a leak in the transmission system. This will allow at least half the cylinder to still pump fluid into the system. What will happen in this case is that (based on your vehicle setup) you might have delayed response time (allowing you to limp to nearest safe spot) or half of your braking system will still remain effective (allowing you to drive in some cases to the nearest service center at reduced speeds). This even after your brake fluid holder (container) is empty. This holds for the steering rack/ clutch as well.

On aircraft the actuators are designed to work on a op-fail-safe op- fail- safe op- fail - manual system. There are a system of pumps. The main pump will typically work off the engine. In case of an engine failure the pump will have a reserve pumping capacity (IIRC approximately 60%) working off a battery run motor.At this stage the pilot will have a reduced response time, however there is a backup pump (also running off battery power) which he can call upon (software also) in case it is in a fight regime to keep the response almost same. This will allow for a few minutes of full power actuation. After the battery on the secondary (backup) reduces to 20% or below it switches off - conserving it for flaps etc during landing. The main pump still continues to provide some support to the actuators on key surfaces. In most cases the pilot will lock a glide slope and maintain a high angle of attack to increase the glide distance as much as possible. This will reduce the amount of inputs required by the pilot and thus reduce the usage of the emergency backup system.

The weight of the batterys is not high and it is not intended to run a high torque motor. This by itself enables it to be used on most fighter a/c.

The fluid transmission system (hoses/ etc) are also designed with multiple break chambers. In case of one hose failing/ leaking, it can be delinked from the rest of the system while full control is still maintained through backup hoses (the backup hoses are thinner and have higher pressure rating, however a higher fail time - designed to be used only for the emergency use - will be swapped once the a/c is back on the ground.

All this is linked to the FBW system. There will be no instance where the FBW will fail completely unless it breaches multiple points on the hydraulic system. On a highly unstable platform like the YellCeeYay, it will still be able to allow for a glide at certain angles of attack.

If all else fails - the pilot will still be able to eject. Man first machine second.

Disclaimer - These systems are possible to use - i am not saying the LCA has them - it may or may not. (hopefully 'may' is the answer).

Thank you. I knew some things and I learnt few more.

While this is true for stable platforms. For unstable platforms like LCA, the FBW could provide upto 80 control signals per second. The reserve juices can only help for so many seconds. Also sustained delayed response can be catastrophic for an unstable platform. If the feedback-actions-reaction loop goes out of sync, you would lose control over the aircraft in a matter of seconds. Having said that I don't know how much delay the FBW can tolerate. It can probably manage instanteneous drops, but definitely not sustained delays. If it was possible, then LCA would be flying with the FBW controlling the control surfaces at half the frequency in first place.

JayS · Post by **JayS** » 01 Oct 2013 18:30

indranilroy wrote: Thank you. I knew some things and I learnt few more.

While this is true for stable platforms. For unstable platforms like LCA, the FBW could provide upto 80 control signals per second. The reserve juices can only help for so many seconds. Also sustained delayed response can be catastrophic for an unstable platform. If the feedback-actions-reaction loop goes out of sync, you would lose control over the aircraft in a matter of seconds. Having said that I don't know how much delay the FBW can tolerate. It can probably manage instanteneous drops, but definitely not sustained delays. If it was possible, then LCA would be flying with the FBW controlling the control surfaces at half the frequency in first place.

Does it actually gives out 80 control signals?? I mean its capable of it, but does it actually corrects the system 80 times a second?? Isn't that too fast for a hydraulic system to react??

Lalmohan · Post by **Lalmohan** » 01 Oct 2013 18:40

the control laws operate on a clock cycle of 80hz, i.e. each control parameter is updated 80 times/second - these will translate into an actual control surface deflection demand, but will be incremental and not absolute. i.e. the control demands are driven incrementally towards the required goal. there is a difference between real time computing and transaction processing - most 'itvity-wallahs' don't understand the former

fanne · Post by **fanne** » 01 Oct 2013 18:49

In one of the LCA article, it was noted (either per second or minute do not remember), the software is capable of generating 500 per second or minute but only 2 get executed. So all the other requirements for Fly by light, or faster computer (for Flight law not jamming etc) is redundant...

NRao · Post by **NRao** » 01 Oct 2013 19:03

Such planes use multiple RTOS OSs. IIRC the main one functions at 80, the rest have their own cycles. Here is one for the F-16 Display. This one is a "memory-protected, partitioned" RTOS.

RKumar · Post by **RKumar** » 04 Oct 2013 23:45

Flight test update

LCA-Tejas has completed 2332 Test Flights Successfully. (28-Sep-2013).
(TD1-233,TD2-305,PV1-242,PV2-222,PV3-371,LSP1-74,LSP2-286,PV5-36,LSP3-157,LSP4-94,LSP5-220,LSP7-60,NP1-4,LSP8-28)

to

LCA-Tejas has completed 2334 Test Flights Successfully. (01-Oct-2013).
(TD1-233,TD2-305,PV1-242,PV2-222,PV3-372,LSP1-74,LSP2-286,PV5-36,LSP3-157,LSP4-94,LSP5-220,LSP7-61,NP1-4,LSP8-28)

ArmenT · Post by **ArmenT** » 06 Oct 2013 12:54

You guys are talking about power supply to the flight control computer, but there are other critical systems in the aircraft that also need electrical power to function. For one, there's the de-icing parts, e.g. the heating elements of the pitot tubes (guess what happens if the pitot tubes ice up) and wing edges. Then, there's the motors driving the pumps for the hydraulic actuators, fuel pumps etc. So there's a lot of other bits in the aircraft that depend on electricity besides the flight computer.

Indranil · Post by **Indranil** » 06 Oct 2013 13:56

nileshjr wrote:
indranilroy wrote: Thank you. I knew some things and I learnt few more.

While this is true for stable platforms. For unstable platforms like LCA, the FBW could provide upto 80 control signals per second. The reserve juices can only help for so many seconds. Also sustained delayed response can be catastrophic for an unstable platform. If the feedback-actions-reaction loop goes out of sync, you would lose control over the aircraft in a matter of seconds. Having said that I don't know how much delay the FBW can tolerate. It can probably manage instanteneous drops, but definitely not sustained delays. If it was possible, then LCA would be flying with the FBW controlling the control surfaces at half the frequency in first place.
Does it actually gives out 80 control signals?? I mean its capable of it, but does it actually corrects the system 80 times a second?? Isn't that too fast for a hydraulic system to react??

This are extremely small deflections. LCA Elevons are updated at up to 80 hz.

Lalmohan · Post by **Lalmohan** » 06 Oct 2013 14:28

ArmenT wrote:You guys are talking about power supply to the flight control computer, but there are other critical systems in the aircraft that also need electrical power to function. For one, there's the de-icing parts, e.g. the heating elements of the pitot tubes (guess what happens if the pitot tubes ice up) and wing edges. Then, there's the motors driving the pumps for the hydraulic actuators, fuel pumps etc. So there's a lot of other bits in the aircraft that depend on electricity besides the flight computer.

true but most of those have various degrees of fail-safe or fail-tolerant modes allowing for some limited period of controlled flight. on an unstable aircraft, loss of primary control computer is instant game over

NRao · Post by **NRao** » 08 Oct 2013 02:57

Continued from the Raffy thread ...................

On a totally unrelated note, I have no idea why an MK3 (in the first place) and then the idea that somehow a MK3 can contribute towards the AMCA - to me - is silly. (Need to take that up in the LCA thread.)

OK, found ......................... No date on it so far....................

member_20011 · Post by **member_20011** » 08 Oct 2013 03:16

This are extremely small deflections. LCA Elevons are updated at up to 80 hz.

Are they visible to naked eye? I mean if some body starts filming LCA in flight..can they see those movements?

SaiK · Post by **SaiK** » 08 Oct 2013 05:00

another high priority fail-safe ops (at least dual redundancy) is OBOGS.

shiv · Post by **shiv** » 08 Oct 2013 06:05

nileshjr wrote:
Does it actually gives out 80 control signals?? I mean its capable of it, but does it actually corrects the system 80 times a second?? Isn't that too fast for a hydraulic system to react??

The actuators are corrected 4 times a second as per Air Marshal Rajkumar in his book "LCA Story". In any case I doubt if they are hydraulic actuators - I think they are servo motors - but even servo motors have a lag time and I doubt if they could react at 80 Hz.

shiv · Post by **shiv** » 08 Oct 2013 06:14

Pratyush wrote:Can the LCA be fitted with an extendible ram ait turbine inorder to produce power and control the air craft in the event of engine and APU failure.

Some thing like that is fitted in civilian airliners, as fail safe, to deal with the problem as explained in above, by IR.

http://en.m.wikipedia.org/wiki/Ram_air_turbine

Pratyush RAM turbines act like small sails because they are constantly bleeding energy from airspeed to feed the generator. Depending on the altitude and airspeed at which an engine shuts down RAM turbines could aggravate the problem by loss of airspeed. I suppose small turbines can be used for some emergency power, but most aircraft are designed to retain a basic minimum of control when the engine is out and I am sure the LCA too has that feature.

The LCA has passed one engine relight test. (The very test that tragically went wrong in the twin engine Saras) When that engine was not powering the LCA - the plane was gliding. So it can glide.

Indranil · Post by **Indranil** » 08 Oct 2013 06:36

shiv wrote:
nileshjr wrote:
Does it actually gives out 80 control signals?? I mean its capable of it, but does it actually corrects the system 80 times a second?? Isn't that too fast for a hydraulic system to react??
The actuators are corrected 4 times a second as per Air Marshal Rajkumar in his book "LCA Story". In any case I doubt if they are hydraulic actuators - I think they are servo motors - but even servo motors have a lag time and I doubt if they could react at 80 Hz.

Shiv sir,

Many moons ago me and a fellow poster here had taken this up. All I can say is that my information is correct and comes from one of the highest flying birds. Otherwise, I would not have asserted it for so many years.

nachiket · Post by **nachiket** » 08 Oct 2013 06:37

shiv wrote: Pratyush RAM turbines act like small sails because they are constantly bleeding energy from airspeed to feed the generator. Depending on the altitude and airspeed at which an engine shuts down RAM turbines could aggravate the problem by loss of airspeed. I suppose small turbines can be used for some emergency power, but most aircraft are designed to retain a basic minimum of control when the engine is out and I am sure the LCA too has that feature.

The LCA has passed one engine relight test. (The very test that tragically went wrong in the twin engine Saras) When that engine was not powering the LCA - the plane was gliding. So it can glide.

I thought RATs were standard equipment on fighter jets before they became standard equipment on airliners. So you are right, the LCA must have one too. I can't imagine the LCA gliding for even a few feet without power to its FBW system. An FBW system failure on an unstable aircraft like the LCA would mean that the aircraft will drop out of the sky like a brick.

Of course, the RAT would only be needed if the APU failed along with the engine itself which is highly unlikely. It can happen if the Aircraft runs out of fuel (and the APU relies on the same fuel supply as the engine, which may not be the case. Some APU's run on their own fuel like a mixture of Hydrazine and Water). A regular engine failure will not result in the APU failing. The Engine relight test will also have used the APU to relight the engine.

Some aero gurus here may want to elaborate.

chackojoseph · Post by **chackojoseph** » 08 Oct 2013 06:42

shivagarg wrote:Are they visible to naked eye? I mean if some body starts filming LCA in flight..can they see those movements?

If they have marked those areas with a different colour, it is possible to distinguish them. Even the onboard tracking system can read movements.

Indranil · Post by **Indranil** » 08 Oct 2013 06:55

nachiket wrote:The Engine relight test will also have used the APU to relight the engine.

The APU is not used to relight the engine. What is used is called a jet fuel starter (JFS). In case of LCA, this system is self sufficient and requires only 28V electric supply to a initiate the starting cycle and fuel to continue working till the starting cycle is completed and the main engine is relit.

The maximum altitude at which the LCA JFS (developed by ETBR&DC) can be used is 6km. At this altitude the cutoff speed (the end of the starting cycle) is reached in about 120 seconds, whereas on the ground it takes 60 seconds.

shiv · Post by **shiv** » 08 Oct 2013 06:56

nachiket wrote:
I can't imagine the LCA gliding for even a few feet without power to its FBW system. An FBW system failure on an unstable aircraft like the LCA would mean that the aircraft will drop out of the sky like a brick.

Nachiket it you read accounts of pilots who lost all power on aircraft - they usually know from their training that the aircraft retains a modicum of manual control. I doubt if the LCA would become completely uncontrollable wil total loss of engine power. Batteries and other "power" sources like compressed air bottles probably kick in. There is some info on this from at least two sources that I can recall. One (unfortunately) has been borrowed from me right now - it's about what the MiG 21 does when it loses power.

I would have thought that in an aircraft like the LCA with pitch instability (I think that means it tends to go nose up without FBW) a standard default elevon setting may kick in to correct that combined with some manual rudder control. This is a guess - will ask and post

The folllowing story has an account of oddball thngs that are possible after failure of controls systems
http://www.bharat-rakshak.com/IAF/Histo ... aguar.html

At the end of 95 minutes, the situation was desperate, as all attempts to lower the undercarriage had failed. The AOC who was at the flying control, in sheer desperation sent word for the HAL specialists (part of JS 120 investigation team) told them that "the aircraft has about 3 minutes of fuel left, suggest anything, even the most ridiculous thing, I will ask the pilot to try it before he ejects". One of the HAL engineers, Mr Jayamohan who had his thinking cap on, quickly came out with a solution, "advise the pilot to put OFF the battery, explaining that this would de-energise the solenoid-operated valve and the trapped fluid would then be available to operate the services. There would be no instruments, lights and radio for a short duration and aircraft may yaw a bit. Battery could be put ON after hearing the thud of the undercarriage coming down". This was precisely what the pilot was asked to do. At this stage, Palit had about 170 kgs of fuel and in terms of duration, just about 3 minutes. He carried out the drill as advised and heard the thud of the undercarriage coming down. Aircraft, which was very light at that fuel state, developed a pronounced yaw (due to small differential in wheels coming down) and Palit put the battery on. The main wheels had locked down, but the nose wheel, which was in the process of coming down, remained half cocked, as the fluid reverted back to controls once the battery was put on. The nose wheel of the Jaguar extends forward, against the airflow and thus takes a while longer. Palit had no fuel to try the procedure once more; he did a quick turn and with his experience and finely honed skill of a test pilot, landed the aircraft safely on the reciprocal runway. He held the nose up as long as he could, used the tail chute and then gently lowered it on to the runway. Aircraft came to a stop with minimum damage to the nose area. One of the engines flamed out during landing run, as the aircraft ran out of fuel.

chackojoseph · Post by **chackojoseph** » 08 Oct 2013 07:10

My understanding is if there is manual input bypassing the computers, the LCA will fall out of the skies. Only option left for pilot is to pull the zero zero lever.

nachiket · Post by **nachiket** » 08 Oct 2013 07:22

Thanks for the correction indranilRoy. It is indeed the JFS which is a separate system on fighter jets, unlike airliners.

Shiv saar, a better comparison would be the MKI crash a few years ago when the FBW system was accidentally switched off by the pilot. Wouldn't that be exactly what would happen if all power to the FBW system was turned off (assuming the RAT does not work for some reason). Keep in mind this is not s simple Engine failure. This is an Engine+JFS+APU+RAT failure. A highly improbable event. This I think would straight up mimic a complete FBW failure as far as basic controls were concerned.

Indranilroy, how would the glide ratio of a fighter like the LCA compare to that of a modern airliner? My guess is it would be considerably worse, meaning that practically the ability to be controllable in the event of a complete loss of power (when there is no chance of an Engine relight) is of limited use. There is no option but to eject.

The longest passenger jet glide is about 20 minutes by an Air Transat A330. They glided about 120km from 33000 ft. They could have possibly done a bit more.

shiv · Post by **shiv** » 08 Oct 2013 07:31

indranilroy wrote: Many moons ago me and a fellow poster here had taken this up. All I can say is that my information is correct and comes from one of the highest flying birds. Otherwise, I would not have asserted it for so many years.

I recall that discussion. I have a technical queries that have not been resolved yet in my mind. What I write below may be wrong.

A 10 ton aircraft flying at say 800 kmph has inertia that would make it difficult to respond to aerodynamic surface deflections lasting as short as 1/80 of a second. Also a gust of wind that lasts for a fraction of a second would be a buffet - not even a bump and may not require a control surface response. A more sustained gust that threatens to change the attitude of the plane would require a response. So it just seemed likely to me that while the measurements are made 80 times a second (or something like that), control inputs are made only after averaging out (say) 15-20 inputs. If the net effect of measurements of 15-20 inputs requires a small elevon deflection - that deflection is actuated even as further measurements continue to see if the last deflection has to be increased or decreased. Mind you this is simply guesswork and not information.

Slightly OT - this sound sample gives an idea of what 80 Hz sounds like and am actuator varying with inputs at 80 Hz would be vibrating at that frequency. You need to wait for the 80 Hz sound
http://www.youtube.com/watch?v=igGroIcga3g

chackojoseph · Post by **chackojoseph** » 08 Oct 2013 07:54

INDIA'S OWN FIGHTER STORY

Just before a stylish flight, pilot Suneet Krishna drives Talk to the light combat aircraft hoping to do India proud, and takes us through the story so far

A fighter plane is not stable enough just to glide. (Even a massive Boeing can actually glide.) Turn off the engine, and a fighter will fall like a rock. But with modern, digital, flight control systems (FCS), inputs from the joystick are converted by computer to signals that keep the plane stable and in control during complex manoeuvres at high angles of attack.

That’s enough of the (crude) aero-engineering lesson.

“When we first flew the Tejas, it was the first time we were testing a fly-by-wire (the FCS) system with no manual backup. There was a lot of apprehension. Even with the F-16 there were major issues, with all their decades of development experience. People were critical, there were worries about what could go wrong. It took a lot of effort to get it off the ground,” he recalls.

Eric Leiderman · Post by **Eric Leiderman** » 08 Oct 2013 08:39

Shivji

There is a reason why so many measurements are taken in a short period of time

Most controllers are PID

ie There is a propotional Integral and a deravative aspect to them

The more inputs in a shorter period of time the more accurately the controller reacts.

I could elobrate but lets simplify it and say the response ramps up if the feedback says the aircraft is not responding the way it should.

So for a loss of power the control surfaces will keep increasing till the deviation between control and feedback signals are same
(theoritically) (practically with a loss of power that will never happen and the softwear will kick in and keep the error signal{error signal=difference between feedback signal and control signal} within a certain limit with minimal usage of control surfaces. )

I wish I could have explained it better

shiv · Post by **shiv** » 08 Oct 2013 09:00

Eric Leiderman wrote:
The more inputs in a shorter period of time the more accurately the controller reacts.

Thanks this is clear enough.

SaiK · Post by **SaiK** » 08 Oct 2013 09:10

why would a feedback necessary say for slat adjustments.. I can map exact positions from 1* gradient to how much ever fraction we want to control.. let us assume say 256 levels of positions are mapped to joystick, then i only need to send that position indicator which the A/D converter does the joystick position to the slat angle angle required. i guess, it is all depends on how it is designed.

no? feedback not required matlab, the joystick position must override whatever current slat position is on a continuous basis? may be i am thinking about a design for RC plane and not some hi-fi stuff that requires proving by some real-time analysis.

ArmenT · Post by **ArmenT** » 08 Oct 2013 10:12

Might as well post this to answer a bunch of questions that newbies might have about this thread:

Stable aircraft: Think of a pencil that is lying on its side on a table. Even if you lift one end of it and drop it, it returns back to its horizontal position by itself (i.e.) stable equilibrium. In a stable aircraft, if you take your hand off the stick, it will return back to horizontal flight by itself.

Unstable aircraft: Consider you are balancing the pencil on its tip on your finger. You can make the pencil stay vertical by moving your hand back and forth, but you have to keep making adjustments continuously to keep it vertical. Now imagine the same problem magnified on a full scale aircraft where you have to make adjustments to various controls, all in an effort to keep the airplane level. This cannot be done fast enough by ordinary human beings, which is why the flight control computer is there to continuously make the adjustments automatically as needed. Of course, due to the unstable configuration, the aircraft can also maneuver quicker.

Now when the pilot pushes the joystick to make a (say) 5 degree bank, the flight computer understands that the pilot needs to make a five degree bank and adjusts the ailerons accordingly to roll the aircraft at the required angle. However, due to the inherent unstable nature of the aircraft, it needs to keep continuously tweaking the position of the ailerons to maintain that 5 degree bank. Similarly, when the pilot moves the joystick back to level out the airplane, the flight control computer keeps tweaking the control surfaces to keep the aircraft balanced in level flight.

Which is where the feedback loop comes in -- the sensors let the flight control computer determine where to make the corrections to hold that angle and how much correction to make. The faster the rate of the feedback samples, the more accurately it can make those minute corrections.

Of course, the software for the flight computer should be able to correctly respond in various scenarios (e.g.) different amounts of thrust, different angles of attack etc. which is why there are so many flight tests to be done to ensure that the software behaves correctly in all scenarios.

If the flight computer crashes or loses power, an unstable aircraft will go into an unrecoverable spin quickly. How soon depends on how unstable the design is. In the case of a heavily unstable aircraft like the F-117, this time was stated to be about 2-3 seconds after the flight computer lost power, per one of its designers (it actually happened to one of their prototypes at the end of their tests, but luckily the test pilot managed to eject safely and they had collected enough data about its flight characteristics by then).

nikhil_p · Post by **nikhil_p** » 08 Oct 2013 10:43

While we are talking about the control systems failure - due to a central computer failure - brought about by the failure of the engine (the main power source) here are a few points to consider.

- The control inputs are at upto 80 Hz - the upto part is important, just as important for men to understand when they go with women for shopping - when a sale says upto 80% off, the actual items on 80% off are typically only 5% of the entire inventory for sale. Most items will be at 25% off.
The point I am trying to make is that the 80Hz input will not be used in most regimes and is typically only during combat type manouvers (e.g.high G Turn with reducing altitude flight). For maintaining level flight typical control inputs are between 20-40 Hz.

- Like a few posters have mentioned more than a certain number of inputs per second actually wouldnt help. An elevon being activated more than 2 times a second might not have any effect on the actual flight as it might overcompensate.

- there is no requirement of a feedback loop from the actuators which is pulsing at the same frequency as the inputs. Typically the feedback loop operates at ratios of approx 1:3 - and is designed to ensure that the components are operating at optimum/ perfectly.

- the control feedback is from the pitot and other equipment(pitch/yaw/airspeed inputs) - not just from the elevons/ actuators.

- like someone mentioned failure of the pitot and these equipment is actually more dangerous.

- heating elements on these control units have multiple redundancy built in.

- Continuously tweaking the position involves also making the 'minimal' number of inputs per second to maintain straight line flight.

JayS · Post by **JayS** » 08 Oct 2013 16:09

shiv wrote:
indranilroy wrote: Many moons ago me and a fellow poster here had taken this up. All I can say is that my information is correct and comes from one of the highest flying birds. Otherwise, I would not have asserted it for so many years.
I recall that discussion. I have a technical queries that have not been resolved yet in my mind. What I write below may be wrong.

A 10 ton aircraft flying at say 800 kmph has inertia that would make it difficult to respond to aerodynamic surface deflections lasting as short as 1/80 of a second. Also a gust of wind that lasts for a fraction of a second would be a buffet - not even a bump and may not require a control surface response. A more sustained gust that threatens to change the attitude of the plane would require a response. So it just seemed likely to me that while the measurements are made 80 times a second (or something like that), control inputs are made only after averaging out (say) 15-20 inputs. If the net effect of measurements of 15-20 inputs requires a small elevon deflection - that deflection is actuated even as further measurements continue to see if the last deflection has to be increased or decreased. Mind you this is simply guesswork and not information.

I too have similar thoughts. For a feedback system to work properly at 80Hz, with control surfaces correction at 80Hz as well, in 1/80th second the complete cycle of process should be finished which would include actually deflecting control surfaces with demanded corrections and then a certain lag time which the control surfaces need to affect the attitude and flight parameters of the aircraft so that feedback signal will be generated. So not only actuation system latency but also aerodynamic latency would come in picture. Both these latencies (lets say electronic systems are far faster than these two) would be limiting factor IMHO. And I agree with following statement by nikhil_p:

there is no requirement of a feedback loop from the actuators which is pulsing at the same frequency as the inputs. Typically the feedback loop operates at ratios of approx 1:3 - and is designed to ensure that the components are operating at optimum/ perfectly.

All this leads me to think that the control surface actuation frequency has to be less than 80Hz. How much less I am not sure. But I am ready to be surprised.

@Indranil:
Well, if you have the info from highly placed source then what could be said about that. But it definitely goes against my intuition. I am trying to find anything which could give some clue but so far no luck.

Lalmohan · Post by **Lalmohan** » 08 Oct 2013 17:23

^^^ computation is incremental not absolute

saik - the joystick and throttle indicate what the pilot wishes to do with the aircraft as a whole. the FCS translates that into the optimum deployment of control surfaces such that the pilots objectives are met given the current flight environment. its not like the pilot thinks about slat angle and aileron position

SaiK · Post by **SaiK** » 08 Oct 2013 18:08

I agree, and that should be the case with the FCS design. And in that design, we have inputs from Slat, altitude sensor (precision matters), airflow sensors, ring laser gyros, and what not that goes into the dynamics of the flight). It can be easily said, LCA can be made into UCAV just for the FCS alone.

Instead of pilot sending the signal, it can come from secured sat comms.

Indranil · Post by **Indranil** » 08 Oct 2013 19:37

@Hakimji,
You are wrong on what is required to be done for a FBW aircraft flying through momentary buffet. But you know enough people to find out.

@ArmenT
You are right except that on a catastrophic failure of the flight computer, the plane won't enter a spin. What would actually happen is that the pilot would try to stop the plane from pitching up. But the controls of a plane are different from that of a car (I am sure you know this, but I am providing a over simplified explanation for the sake of others). While turning a car, holding the wheel at a certain deflection, maintains the rate of turn. Whereas on a plane, holding the "wheel" at a certain angle increases the rate of turn. Therefore to turn, a pilot rotates the "wheel" till the desired rate of turn is achieved, turns it back to "straight" for the rest of the turn, and then turns it in the opposite direction to complete the turn. Now, going back to the scenario we were discussing, the plane pitches up, the pilot provides correction through the stick, but has to let go, as otherwise he would push the plane to negative AoA. But immediately as he lets go the plane starts pitching up again. So he has to provide a correction again. So his corrections are more like pulses rather than a holding the stick at a particular place. The problem is that these required pulses are way too fast and short for human beings and there is a lag between the control commands and the reaction to those commands. So within a couple of seconds, his commands will go out of sync. The plane would start oscillating around the pitch axis. Instead of mitigating the problems, the pilots inputs would be aggravating the problem. In the best case, the amplitude of the oscillations would increase till in a matter of seconds, the plane would reach AoA, where it stalls. The same thing happens if your flight computer is slow in providing control commands. In case that you are close to the ground, you will crash like the F-22 incident. If you are in the air, the characteristics is highly non-linear and unpredictable. If you are flying fast, there is a good chance that you would roll over backwards like a leaf. If the pilot has not ejected, he won't be conscious to eject later, and even if he manages to remain conscious, most probably his ejection will be fatal. Beyond this point the plane is just a rock in ballistic trajectory affected by aerodynamic forces.

@Nachiket ji
It is actually not as bad as you think. The wetted aspect ratio is not that bad for a fighter (flying clean) vis-a-vis a airliner. For example this is a F-16 landing with main engine out.

And the F-16 which had a bird strike immediately after take off. He banks and has time to try one relight.

Having said that I would expect modern airliners like the B787 and A350 to glide longer.

@Nilesh,

None of us are at a disagreement here that the sensors provide measurements much faster than what the CPU can process. And that none of the sensor readings are wasted. This is what Lalmullah has been saying for quite some time regarding using the sensor readings to incrementally arrive at a control signal. In Nikhil's case, 3 sensor readings would lead to one control signal. Our disagreement is only at the rate of elevon commands provided by the FCS. This was the simple question asked to the high flying bird (verbatim)

I just wanted to find out how many times a second the commands to elevons are updated on the LCA? in other words how many times a second does the elevons move on the LCA?

Here was the simpler answer (verbatim).

At 80 Hz.

I have nothing more to say.

@Saik ji,
UAV will obviously be FBW, but there is no requirement of making a UAV unstable unless you want it for A2A roles. The problem will UAVs is not flying it but relaying commands. India currently is quite limited in this regard. That is why Rustom-2 will have a 24 hour endurance and only a range of 350 kms.

negi · Post by **negi** » 08 Oct 2013 20:05

Those videos confirm one thing though i.e. even after an engine flame-out the FCS functionality and obviously all the underlying sub systems which control the lifting surfaces remain functional.