LCA: News & Discussions - October 2016

All threads that are locked or marked for deletion will be moved to this forum. The topics will be cleared from this archive on the 1st and 16th of each month.
Locked
ramana
Forum Moderator
Posts: 59773
Joined: 01 Jan 1970 05:30

LCA: News & Discussions - October 2016

Post by ramana »

Please continue in this thread:

Link to old thread

Thanks, ramana
Indranil
Forum Moderator
Posts: 8428
Joined: 02 Apr 2010 01:21

Re: LCA Tejas: News and Discussions

Post by Indranil »

JayS wrote: Some thoughts on this type of design from top off my head.

LCA does not have a nice place to put the twin tails. IMO some serious work will be needed here. An American company is better placed for this project IMO, if at all we want to outsource it, because they have the most experience with twin tails in the West. The twin tail has buffet issue due to the LE wing vortices, and they know how to deal with it now.

The wing of LCA is too low and the fuselage in too cylindrical. IMO, wing needs to be lifted up, fuselage must be extended outward, flattened and blended with the wing in much better way (like how it is in F22). (I would really love that. The current wing-fuselage blending on LCA is ugly.)

Since LCA is a single engine jet, we don't need to worry about one engine out scenario and the tails could be placed on the engine bay close to each other, avoiding wind structure redesign and also putting in near the wing vortex. We could even consider V tail with included angle >90deg.

Another point is adding tail there will screw up the Area ruling, we know the area curve is very steep in that region and adding tails will only increase the problem. So extending the engine bay slightly outward might be a good idea. It would also give higher lever arm for tails and so smaller tails can be used.

A bubble canopy would be great and along with redesigned air intakes it can be used to remove that kink in the Area ruling curve.

While implementing chin on the nose we need to make sure the chin vortices do not interfere/get coupled with the wing LE vortices at moderate to high AoA. Chin nose suppose to increase directional stability. Could help in reducing tail size further.

IMHO, we should explore tailless delta design by implementing more powerful engine with TVC. That could give us much cleaner configuration. Although it might prove quite challenging in terms of providing directional and rolling stability and TVC will have to be augmented very closely with the flight controls.

Comments are welcome on this.
Here is my "if horses had wings" idea. Start with the LCA Navy Mk2 as the base.
Image

We will start to morph that into something like that in this sketch. But let me give more details.
Image[/quote]

In the sketch, the twin tails are somewhat weirdly mounted. The Navy Mk2 defines a much better way. Take the the wing body fairing behind T.E. of the wing and make it wider. Adopt F-16 style airbrakes. Mount the twin fins just ahead of the airbrakes closer to the engine. This fairing can be made longer-shorter-fatter-wider to smooth the area curve. It would also allow to completely accommodate the chaff dispensers inside the fairing unlike in the MK1.

Have a small housing vertically above the engine for the landing parachute like F-35s proposal. The front of that housing should have an opening for the heat exchanger ram air intake. Alternatively, the intake could remain at the base of one of the fins.

They have already modified the MLG mounts and spaced them out on the Navy Mk2. They now retract straight up into an extra fairing by the side of the inlets. Use this, but adopt a fairing with much better wing body blending. They can be undergo the same transformations as from Rafale A to Rafale B/C/D/M.
Image

Image

Or they can go the F-22 way: Completely flatten out the bottom with flat fuselage bottom extending into flat walls of a rectangular inlet. This flat bottom then rises to meet the wing using the trapezoidal side wall of the inlet, then fairing for the MLG and finally blending smoothly into the wing. Behind the MLG mounts, the fairing tapers out into the fairing containing the air brake section. The flat and wider bottom section will allow the addition of two additional dry hardpoints on either side of the current center hardpoint. These hardpoints cannot be used if the center hardpoint is being used, and vice versa. This could add flexibility when the central hardpoint is not being used to carry fuel. One could look at the GRipen NG, Mirage etc. for this setup.

Image

Inlets should undergo the same modifications as F-18A/B to F-18 E/F. Note that this plays nicely into the design describe above. The splitter plate can be completely done away with.
Image
Image

If the above modifications are made, we won't need to flatten out the fuselage. I would give the fuselage a chin joining the nose to the LE of wing joint.

Can anybody good at 3D modelling draw this out for me. We can refine it out as we go.
vina
BRF Oldie
Posts: 6046
Joined: 11 May 2005 06:56
Location: Doing Nijikaran, Udharikaran and Baazarikaran to Commies and Assorted Leftists

Re: LCA Tejas: News and Discussions

Post by vina »

JayS wrote:ADA has been playing safe and perhaps was restricting itself to this value until recently and now when they have enough confidence, they have increased the static instability margin somehow
:lol: :lol: . What they must have done is taken out some of the ballast from the nose! That is all that needs to be done.Some 300/400 Kg of lead/concrete whatever they were using a the nose/front section of fuselage, when fully removed and the FCS tuned finally will give a 9G performance that will appear in the spec sheets finally. Currently the plane is AoA limited by the FCS.
Cosmo_R
BRF Oldie
Posts: 3407
Joined: 24 Apr 2010 01:24

Re: LCA Tejas: News and Discussions

Post by Cosmo_R »

rohitvats wrote:
Cosmo_R wrote:@Khalsa ^^^I got the number from here

"165 MiG-27Ms licensed built by HAL. Total 120 Upgraded Mig-27ML Active. To be retired between 2018 and 2020.[13]

No. 2 Squadron IAF[13]
No. 9 Squadron IAF[13]
No. 18 Squadron IAF[13]
No. 20 Squadron IAF[13]
No. 22 Squadron IAF[13]
No. 222 Squadron IAF[13]

https://en.wikipedia.org/wiki/Mikoyan_MiG-27
That list is only partially correct.

- We've only 2 x upgraded Mig-27 squadrons - 10 and 29
- No 9 (Wolfpack) converted long time back as third Mirage-2000 Squadron
- No 2 squadron flies Su-30MKI
- Of the balance three (18, 22,222), one has been number-plated already as of 2016.
- So, only four squadrons of Mig-27 left.
Please update wiki but remaining squadrons is not the same as replacement need.
vina
BRF Oldie
Posts: 6046
Joined: 11 May 2005 06:56
Location: Doing Nijikaran, Udharikaran and Baazarikaran to Commies and Assorted Leftists

Re: LCA Tejas: News and Discussions

Post by vina »

Somehow I think when we see the MK1A models (in this Aero India hopefully), we will see the Eurofighter like scheme of SPJ and towed array decoy on the wing tips. That is the easiest way to go, rather than try putting those inside the fuselage / mold line . This will also involve the least amount of work and the entire change (the wing has already been stiffened once, thanks to the IAF weapon specs fiasco) can be done in a year or so , well within the MK1A timeline.
rohitvats
BR Mainsite Crew
Posts: 7830
Joined: 08 Sep 2005 18:24
Location: Jatland

Re: LCA Tejas: News and Discussions

Post by rohitvats »

Cosmo_R wrote:
rohitvats wrote:
That list is only partially correct.

- We've only 2 x upgraded Mig-27 squadrons - 10 and 29
- No 9 (Wolfpack) converted long time back as third Mirage-2000 Squadron
- No 2 squadron flies Su-30MKI
- Of the balance three (18, 22,222), one has been number-plated already as of 2016.
- So, only four squadrons of Mig-27 left.
Please update wiki but remaining squadrons is not the same as replacement need.
Sorry, could not get the drift?
jahaju
BRFite -Trainee
Posts: 95
Joined: 26 Apr 2008 18:40

Re: LCA: News & Discussions - October 2016

Post by jahaju »

From earlier post by srai, dated 11th April 2016

Status of LCA manufacturing as on 11th April 2016.

"Let's keep the following tabs and see when a flight takes place and delivery to the IAF occurs:

SP-1 -> first flight (Oct 1, 2014); delivered to the IAF (Jan-17, 2015) [Looks like it takes around 3-months]
SP-2 -> assembly completed; first flight (March 23, 2016) [Let's see how long it takes from first flight to delivery to IAF]
SP-3 -> final equipping and assembly; first flight planned for July, 2016 [Looks like around 3-months to flight once it gets to this stage]
SP-4 -> ready for coupling
SP-5 -> all components on the jig
SP-6 -> some modules on jig
SP-7 -> some modules on jig
SP-8 -> components will be loaded on a jig within a month (by April 24, 2016) [this one will be the one to watch to see how long it takes once it loads on a jig]"

Lets take the discussion to positive side. When is SP-3 slated for handover to IAF?

Since SP-3 is flying, one jig must have become free so as to enable SP-9 assembly to start or started in July 2016.
JayS
Forum Moderator
Posts: 4567
Joined: 11 Aug 2016 06:14

Re: LCA Tejas: News and Discussions

Post by JayS »

vina wrote:
JayS wrote:ADA has been playing safe and perhaps was restricting itself to this value until recently and now when they have enough confidence, they have increased the static instability margin somehow
:lol: :lol: . What they must have done is taken out some of the ballast from the nose! That is all that needs to be done.Some 300/400 Kg of lead/concrete whatever they were using a the nose/front section of fuselage, when fully removed and the FCS tuned finally will give a 9G performance that will appear in the spec sheets finally. Currently the plane is AoA limited by the FCS.
Ohh that's a total no brainer to remove ballast. What I meant was the limit of instability that FCS can handle is increased somehow. The thing is instability limit that can be safely handled is not only restricted by Aircraft dynamics but also the Flight control actuators. There are some variables so I said somehow. Shifting CG is the easiest thing to do on LCA. :D
shiv
BRF Oldie
Posts: 34982
Joined: 01 Jan 1970 05:30
Location: Pindliyon ka Gooda

Re: LCA Tejas: News and Discussions

Post by shiv »

JayS wrote: Ohh that's a total no brainer to remove ballast. What I meant was the limit of instability that FCS can handle is increased somehow. The thing is instability limit that can be safely handled is not only restricted by Aircraft dynamics but also the Flight control actuators. There are some variables so I said somehow. Shifting CG is the easiest thing to do on LCA. :D
I suspect the limit is G forces that pilots can withstand.

Once for timepass I made a video of taking a stable model paper plane and making it so unstable that it crashes - using methods that involved both control surfaces and ballast. Timepass onlee..
https://www.youtube.com/watch?v=b3xdC6VnCHc
JayS
Forum Moderator
Posts: 4567
Joined: 11 Aug 2016 06:14

Re: LCA Tejas: News and Discussions

Post by JayS »

shiv wrote: I suspect the limit is G forces that pilots can withstand.
Shiv, We are talking of aircraft dynamic instabilities with characteristic period of microseconds. Humans can sustain large G values for fraction of a second (even the 9G turn will be sustained by the aircraft for only a few seconds before it settles to sustained turn rate). G is not the limiter here. But how fast your system can respond to the dynamic instability and correct it before it diverges. We have a 80Hz FCS on LCA right?? That's rate 80Hz is limited by the system response - system latency, a large part of which comes from the actuators.
adarshp
BRFite -Trainee
Posts: 44
Joined: 05 Aug 2008 14:19
Location: du weldenwarden

Re: LCA Tejas: News and Discussions

Post by adarshp »

JayS wrote:
shiv wrote: I suspect the limit is G forces that pilots can withstand.
Shiv, We are talking of aircraft dynamic instabilities with characteristic period of microseconds. Humans can sustain large G values for fraction of a second (even the 9G turn will be sustained by the aircraft for only a few seconds before it settles to sustained turn rate). G is not the limiter here. But how fast your system can respond to the dynamic instability and correct it before it diverges. We have a 80Hz FCS on LCA right?? That's rate 80Hz is limited by the system response - system latency, a large part of which comes from the actuators.
I recall reading somewhere that 80 hz is only for reading system data, i.e. 80 readings per second. These are processed to plot and project aircraft's future state. Actuator response rate is much slower and only 2 control surface corrections are done by the fcs per second. The projections help the fcs decide what interventions will keep the aircraft stable at a slower intervention rate. This becomes more complicated when aircraft manoeuvres are being done. Legitimate maneuvers within flight envelope must be allowed while still ensuring that aircraft does not depart safety limits. I think within flight envelope, pilot instructions are interpreted as a stabled target state in a maneuvering situation and fcs keeps aircraft stable in that manovering state. Pilots dont change maneuvering inputs more than twice a second, if that.
JayS
Forum Moderator
Posts: 4567
Joined: 11 Aug 2016 06:14

Re: LCA: News & Discussions - October 2016

Post by JayS »

IR,

Good Ideas. I particularly like the rectangular intake from F/A18. We could also think of DSI feature instead of splitter plate. (I am not too sure about arrangement of wing+splitter plate for inlet shock at supersonic flight in LCA right now, but I think it can be taken care of).

Why I had been refraining to go full on with the Wing + Fuselage blending a la rafale/F22 (You chose the absolute best two there :-)) is that I am afraid it will engulf the supersonic cambering on the inboard wing section. This might hamper supersonic aerodynamic efficiency and manoeuvrability. I need to think more, but if someone can make a CAD that would be very helpful.
JayS
Forum Moderator
Posts: 4567
Joined: 11 Aug 2016 06:14

Re: LCA Tejas: News and Discussions

Post by JayS »

adarshp wrote:
JayS wrote:
Shiv, We are talking of aircraft dynamic instabilities with characteristic period of microseconds. Humans can sustain large G values for fraction of a second (even the 9G turn will be sustained by the aircraft for only a few seconds before it settles to sustained turn rate). G is not the limiter here. But how fast your system can respond to the dynamic instability and correct it before it diverges. We have a 80Hz FCS on LCA right?? That's rate 80Hz is limited by the system response - system latency, a large part of which comes from the actuators.
I recall reading somewhere that 80 hz is only for reading system data, i.e. 80 readings per second. These are processed to plot and project aircraft's future state. Actuator response rate is much slower and only 2 control surface corrections are done by the fcs per second. The projections help the fcs decide what interventions will keep the aircraft stable at a slower intervention rate. This becomes more complicated when aircraft manoeuvres are being done. Legitimate maneuvers within flight envelope must be allowed while still ensuring that aircraft does not depart safety limits. I think within flight envelope, pilot instructions are interpreted as a stabled target state in a maneuvering situation and fcs keeps aircraft stable in that manovering state. Pilots dont change maneuvering inputs more than twice a second, if that.
You are right at that, my bad. But since I have not read about actual max actuation rate per second for LCA, I wouldn't comment on that. But 2Hz seem low to me (Is this for LCA?). In one of the papers on LCA, one type of dynamic instability is shown to have 200ms time period to double amplitude, that is 32 times in one second. I would think it will need faster than 2Hz response.

This is preciously why the bottleneck in handling the instability is the actuation rate.

Incredibly, X-29) had max static instability margin of some 35% under subsonic conditions. I need to check how were they managing that. But for operations fighters, I have read 8% max so far for EF-2000.

Arguably static stability margin is not the whole sole indicator of extreme instability modes, but a good indicator and increasing it in all probability aggravates the problem.
SaiK
BRF Oldie
Posts: 36424
Joined: 29 Oct 2003 12:31
Location: NowHere

Re: LCA: News & Discussions - October 2016

Post by SaiK »

I think, we have to re-consider the larger nose cross section keeping what kind of radar we want to put inside
adarshp
BRFite -Trainee
Posts: 44
Joined: 05 Aug 2008 14:19
Location: du weldenwarden

Re: LCA Tejas: News and Discussions

Post by adarshp »

JayS wrote:
You are right at that, my bad. But since I have not read about actual max actuation rate per second for LCA, I wouldn't comment on that. But 2Hz seem low to me (Is this for LCA?). In one of the papers on LCA, one type of dynamic instability is shown to have 200ms time period to double amplitude, that is 32 times in one second. I would think it will need faster than 2Hz response.
I am doing this off a phone, so cant find the original source I was remembering. But your question makes me think a little more, and i think what was mentioned was actually along the lines of 2 pilot inputs per second to change aircraft direction, resulting in (now i think) upto 4 actuator updates (2 to change aircraft direction and 2 to stabilize direction change). I think this meant each physical actuator update was done per 15 to 20 readings, combining actual and projection data to arrive at optimal corrections. Will check later if i can find the source.
JayS
Forum Moderator
Posts: 4567
Joined: 11 Aug 2016 06:14

Re: LCA Tejas: News and Discussions

Post by JayS »

Excellent Video BTW. I have seen it first 2-3 years ago I think.
Indranil
Forum Moderator
Posts: 8428
Joined: 02 Apr 2010 01:21

Re: LCA Tejas: News and Discussions

Post by Indranil »

adarshp wrote:
JayS wrote:
You are right at that, my bad. But since I have not read about actual max actuation rate per second for LCA, I wouldn't comment on that. But 2Hz seem low to me (Is this for LCA?). In one of the papers on LCA, one type of dynamic instability is shown to have 200ms time period to double amplitude, that is 32 times in one second. I would think it will need faster than 2Hz response.
I am doing this off a phone, so cant find the original source I was remembering. But your question makes me think a little more, and i think what was mentioned was actually along the lines of 2 pilot inputs per second to change aircraft direction, resulting in (now i think) upto 4 actuator updates (2 to change aircraft direction and 2 to stabilize direction change). I think this meant each physical actuator update was done per 15 to 20 readings, combining actual and projection data to arrive at optimal corrections. Will check later if i can find the source.
I thought I had cleared this enough number of times. May be not. Here I go again.

There used to be a poster here who really knew his stuff (he has a Ph.D. from one the best universities in control theory). When this discussion first happened on BRF, he said that he had read somewhere that the actuators on LCA worked at 80Hz. People hounded him off because they had looked out of their windows and seen actuators on passenger aircrafts move. They could not fathom movements faster than 2-3 Hz. He argued with the people and said, at 2-3 Hz the aircraft would certainly depart. Unfortunately, people hounded him and he politely left BRF.

Me and him took this conversation offline. I was asking my friends in control theory about the frequencies they were working with. My friends were working with 40 Hz. He emailed Dr. Deodhare directly. Here is the question and answer verbatim.

Question:
Hello Dr.Deodhare
I got your contact information from Air Marshal Rajkumar. I had some questions regarding the flight control algorithm on the LCA.

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?

Thanks in Advance
Answer
At 80 Hz.


There is support for this in another published literature. For example in the paper: "AN APPROACH TO HIGH AoA TESTING OF THE TEJAS LCA",
The stability and control cycle is updated every 12.5 milli secs through highfidelity, rate and acceleration sensors and high rate control actuators.
adarshp
BRFite -Trainee
Posts: 44
Joined: 05 Aug 2008 14:19
Location: du weldenwarden

Re: LCA Tejas: News and Discussions

Post by adarshp »

indranilroy wrote:
adarshp wrote:
I am doing this off a phone, so cant find the original source I was remembering. But your question makes me think a little more, and i think what was mentioned was actually along the lines of 2 pilot inputs per second to change aircraft direction, resulting in (now i think) upto 4 actuator updates (2 to change aircraft direction and 2 to stabilize direction change). I think this meant each physical actuator update was done per 15 to 20 readings, combining actual and projection data to arrive at optimal corrections. Will check later if i can find the source.
I thought I had cleared this enough number of times. May be not. Here I go again.

There used to be a poster here who really knew his stuff (he has a Ph.D. from one the best universities in control theory). When this discussion first happened on BRF, he said that he had read somewhere that the actuators on LCA worked at 80Hz. People hounded him off because they had looked out of their windows and seen actuators on passenger aircrafts move. They could not fathom movements faster than 2-3 Hz. He argued with the people and said, at 2-3 Hz the aircraft would certainly depart. Unfortunately, people hounded him and he politely left BRF.

Me and him took this conversation offline. I was asking my friends in control theory about the frequencies they were working with. My friends were working with 40 Hz. He emailed Dr. Deodhare directly. Here is the question and answer verbatim.

Question:
Hello Dr.Deodhare
I got your contact information from Air Marshal Rajkumar. I had some questions regarding the flight control algorithm on the LCA.

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?

Thanks in Advance
Answer
At 80 Hz.


There is support for this in another published literature. For example in the paper: "AN APPROACH TO HIGH AoA TESTING OF THE TEJAS LCA",
The stability and control cycle is updated every 12.5 milli secs through highfidelity, rate and acceleration sensors and high rate control actuators.
I stand corrected in this case. Apologies for any misinformation in this regards. I did find old discussions on brf now that i looked again on a pc, and i can see my conclusions are based on that discussion and the control law paper linked in there.

Indranil - thank you for your feedback.
Karan M
Forum Moderator
Posts: 20772
Joined: 19 Mar 2010 00:58

Re: LCA: News & Discussions - October 2016

Post by Karan M »

Indranil - follow through - so isn't that a very high refresh rate & how does that correlate to the ADA FCS gents statement about LCA maneuverability & instability - and Vina's hypothesis about 9G capability (my take is weight savings if any will be marginal on Mk1A because ballast will be compensated by AESA).
Indranil
Forum Moderator
Posts: 8428
Joined: 02 Apr 2010 01:21

Re: LCA: News & Discussions - October 2016

Post by Indranil »

Karan M wrote:Indranil - follow through - so isn't that a very high refresh rate & how does that correlate to the ADA FCS gents statement about LCA maneuverability & instability - and Vina's hypothesis about 9G capability (my take is weight savings if any will be marginal on Mk1A because ballast will be compensated by AESA).
What Vina is saying is that if you move the ballast from the nose, the C.G. moves back, hence increasing the static margin. This however does not mean that LCA Mk1A will become 9G capable. Afterall static margin is a measure of the (in)stability of the aircraft, i.e. its tendency to pitch up. There are planes which are statically stable and can manage 9G. Intrinsic to Vina's argument is the assumption that the LCA Mk1 is overdesigned and as the developers realize that it is performing better than expected, they might also find that the current structure itself is 9G capable.

Anyways,
1. The ballast in the nose of Mk1 is not 300 kgs. It is much lesser but can't be disclosed in public.
2. AFAIK MMR is about 140 kgs, 2032 is about 90 kg. 2052 with LCA size antenna will be around 145-150 kg. You know this space better than me, so please correct me if I am wrong.
3. HAL's claim of making Mk1A 700-800 kgs lighter is a big stretch. The weight savings will be of the order of 300 kgs. It will be higher when they can get cocured wings.
malushahi
BRFite
Posts: 351
Joined: 16 Jul 2008 03:08
Location: South of Berkshires

Re: LCA: News & Discussions - October 2016

Post by malushahi »

f-18 claw (longitudinal). 90s tech.

http://i.imgur.com/MDWu1Oh.jpg

Image


tejas should be at least as good.

present day piezo-actuator response can approach several hundreds (even thousands) of hz.
Indranil
Forum Moderator
Posts: 8428
Joined: 02 Apr 2010 01:21

Re: LCA: News & Discussions - October 2016

Post by Indranil »

JayS wrote:IR,

Good Ideas. I particularly like the rectangular intake from F/A18. We could also think of DSI feature instead of splitter plate. (I am not too sure about arrangement of wing+splitter plate for inlet shock at supersonic flight in LCA right now, but I think it can be taken care of).

Why I had been refraining to go full on with the Wing + Fuselage blending a la rafale/F22 (You chose the absolute best two there :-)) is that I am afraid it will engulf the supersonic cambering on the inboard wing section. This might hamper supersonic aerodynamic efficiency and manoeuvrability. I need to think more, but if someone can make a CAD that would be very helpful.

I will try to find out the static margin of the LCA. But as you know, the C.G. moves by 3% without the active fuel proportioners. So 5% static margin is on the very low side.

I chose the Rafale because its lower camber of the wing is a similar to that of the LCA's.

Image
vina
BRF Oldie
Posts: 6046
Joined: 11 May 2005 06:56
Location: Doing Nijikaran, Udharikaran and Baazarikaran to Commies and Assorted Leftists

Re: LCA: News & Discussions - October 2016

Post by vina »

Indranlroy wrote:What Vina is saying is that if you move the ballast from the nose, the C.G. moves back, hence increasing the static margin. This however does not mean that LCA Mk1A will become 9G capable. Afterall static margin is a measure of the (in)stability of the aircraft, i.e. its tendency to pitch up. There are planes which are statically stable and can manage 9G. Intrinsic to Vina's argument is the assumption that the LCA Mk1 is overdesigned and as the developers realize that it is performing better than expected, they might also find that the current structure itself is 9G capable.
1. When you remove ballast from the nose, the CG moves back and the airframe becomes MORE unstable , decreasing the static margin. ie. you are making it it MORE unstable. Also, removing weight from the extremities has the max impact in CG (which depends on weight * distance from CG , which is then weighted by the total weight). A classic weigh and cg estimate sheet will have exactly that, weight, and distance from center (in x,y and z) and you estimate the CG from the individual weights.

2. The airframe as it exists is full 9G capable. It is not a question of weight, and drag (the engine has enough excess installed thrust), but rather maturity of control laws. To pull the full 9G, in a sustained turn, the bank angle has to be 1.459 radians or 83.62 deg . That is the load factor. To generate the 9G lift, the AoA has to increase for the wings to produce 9 times lift. That will need an AoA of around 26 deg to 28 deg, depending on the CL curve . My submission is that the LCA is not YET fully certified for the designed angle of attack. Once that certified , the full 9G will appear on the display boards. Watch out for news of AoA achieved to be upwards of 26 deg. As of now, it is certified to 24 deg or so. The plane is hence currently limited to 8G or whatever due to the FCS preventing from pulling the required AoA.
rahulm
BRFite
Posts: 1247
Joined: 19 Jun 2000 11:31

Re: LCA: News & Discussions - October 2016

Post by rahulm »

Just a nit pick. Due to the CLAW and control,system characteristics, Tejas reaches within striking distance of its certified AoA and G rapidly and then slows down. There is an internal proposal to set the CLAW to just past the certified AoA to correct this.
vina
BRF Oldie
Posts: 6046
Joined: 11 May 2005 06:56
Location: Doing Nijikaran, Udharikaran and Baazarikaran to Commies and Assorted Leftists

Re: LCA: News & Discussions - October 2016

Post by vina »

rahulm wrote:Just a nit pick. Due to the CLAW and control,system characteristics, Tejas reaches within striking distance of its certified AoA and G rapidly and then slows down. There is an internal proposal to set the CLAW to just past the certified AoA to correct this.
Yes, that will improve turn performance as well as it will reach the required alpha faster.

BUT WHAAT! You want to play around with control laws? Will you do that with the Gripen /F16/Rafale on your own ? Or if you go back to the OEM and tell them, wont they laugh at you and point you to the door to get out ?

Buy More Rafales! Buy More Gripen! , Buy More F16 ,The Village Idiots in the media and otherwise will do the Rah Rah dance with you! :roll: :roll:
shiv
BRF Oldie
Posts: 34982
Joined: 01 Jan 1970 05:30
Location: Pindliyon ka Gooda

Re: LCA Tejas: News and Discussions

Post by shiv »

JayS wrote:
shiv wrote: I suspect the limit is G forces that pilots can withstand.
Shiv, We are talking of aircraft dynamic instabilities with characteristic period of microseconds. Humans can sustain large G values for fraction of a second (even the 9G turn will be sustained by the aircraft for only a few seconds before it settles to sustained turn rate). G is not the limiter here. But how fast your system can respond to the dynamic instability and correct it before it diverges. We have a 80Hz FCS on LCA right?? That's rate 80Hz is limited by the system response - system latency, a large part of which comes from the actuators.
Air Marshal Rajkumar, in his book on Tejas stated that the actuators work at 2 Hz/twice per second
Image
JayS
Forum Moderator
Posts: 4567
Joined: 11 Aug 2016 06:14

Re: LCA: News & Discussions - October 2016

Post by JayS »

As IR said, Static instability has nothing to do with 9G capability. Its just how much willing the airctaft is to go there, thats what static stability tells you. MiG29 originally was statically stable aircraft, did not have FCS, but still very agile and 9G capable.

And yes, LCA aero-structures is "designed" for 9G. Limitation seems to be the aerodynamics as VIna points out. The envelop opening will happen someday.

5% is what the only available figure so far about LCA that I have seen. I have considered 5%+3% in past. Let me know what you find.
shiv
BRF Oldie
Posts: 34982
Joined: 01 Jan 1970 05:30
Location: Pindliyon ka Gooda

Re: LCA: News & Discussions - October 2016

Post by shiv »

I would be happy for someone to explain this to me and tell me I am wrong because..

I can see absolutely no reason for any aircraft actuators to move 80 times per second. Given that each actuator activates a control surface and those control surfaces cause a change in the direction of movement of a 10,000 kg object flying at 600 plus kmph, getting 80 or even 40 changes in control surface position every second is unlikely. To me it seems more likely that inputs are analysed at 80Hz and an average of those inputs should be sent to the actuators to move the control surface 2-4 times a second which is more likely to have some effect on the movement of the aircraft.

I cant imagine that the actuator is asked to move an elevon up and then down again 12.5 milliseconds later, only for the decsion to be reversed again in 12.5 ms. For a surface controlling the air movement over a 10,000 kg mass moving at 600 kmph that sounds meaningless. To me.
shiv
BRF Oldie
Posts: 34982
Joined: 01 Jan 1970 05:30
Location: Pindliyon ka Gooda

Re: LCA: News & Discussions - October 2016

Post by shiv »

JayS wrote:As IR said, Static instability has nothing to do with 9G capability.
Please. I need to understand this if possible in lay terms and this is related to my own games with model aircraft for many years.

A statically unstable aircraft will not fly, but if it is forced to fly it will perform the most unbelievable and uncontrolled "manoeuvres" before crashing. A stable aircraft cannot be made to perform those mad manoeuvres. So a statically unstable aircraft CAN be controlled by computer by forcing it to fly stably by keeping control surfaces oriented in a particular degree and direction. This would actually add to drag because the statically unstable plane, left to itself without those pesky control surfaces forcing it to fly stably, "wants" to perform mad manoeuvres and crash.

But when hi-G "mad" manoeuvres are required - the fly-by-wire system allows those movements to occur by reducing control over the control surfaces allowing the plane to depart from its stable flight regime to the unstable manoeuvre in a controlled fashion, while preventing complete loss of control.

To that extent it seems to me that Hi-G manoeuvres and static instability are inextricably linked. Instability is not needed for Hi-G manoeuvres, but it helps
shiv
BRF Oldie
Posts: 34982
Joined: 01 Jan 1970 05:30
Location: Pindliyon ka Gooda

Re: LCA: News & Discussions - October 2016

Post by shiv »

Some further thoughts/questions/doubts. The calculations inovlved are way way beyond my knowledge or capability

When you have a 10,000 kg plane flying at 600 kmph - there is a huge mass of air flowing (at 600 kmph) past a control surface - say an elevator.

The question is as follows: If the plane must roll the aileron must move. A miniscule deflection of the aileron will initiate a slow roll. But for a rapid roll - say 180 deg/second the aileron must be deflected much more than a "miniscule" deflection and that deflection has to happen fast. A fast, large elevator deflection against a 600 kmph stream of air on a large wing surface will require a large force applied quickly.

What would be the sort of force required to move such a elevator 5 degrees in 0.1 seconds. What force would be required to achieve the same movement in 0.02 seconds (20 ms)? How long after that movement would the aircraft actually start rolling and roll 10 degrees? How much difference would it make to the roll rate if the elevator moved in 0.1 seconds rather than 0.02 seconds?
LokeshC
BRFite
Posts: 697
Joined: 15 Aug 2016 04:36

Re: LCA: News & Discussions - October 2016

Post by LokeshC »

Drag Equation with changing effective area. Let me get off my cell and will post.
vina
BRF Oldie
Posts: 6046
Joined: 11 May 2005 06:56
Location: Doing Nijikaran, Udharikaran and Baazarikaran to Commies and Assorted Leftists

Re: LCA: News & Discussions - October 2016

Post by vina »

shiv wrote: Please. I need to understand this if possible in lay terms and this is related to my own games with model aircraft for many years.

A statically unstable aircraft will not fly, but if it is forced to fly it will perform the most unbelievable and uncontrolled "manoeuvres" before crashing. A stable aircraft cannot be made to perform those mad manoeuvres. So a statically unstable aircraft CAN be controlled by computer by forcing it to fly stably by keeping control surfaces oriented in a particular degree and direction. This would actually add to drag because the statically unstable plane, left to itself without those pesky control surfaces forcing it to fly stably, "wants" to perform mad manoeuvres and crash.

But when hi-G "mad" manoeuvres are required - the fly-by-wire system allows those movements to occur by reducing control over the control surfaces allowing the plane to depart from its stable flight regime to the unstable manoeuvre in a controlled fashion, while preventing complete loss of control.

To that extent it seems to me that Hi-G manoeuvres and static instability are inextricably linked. Instability is not needed for Hi-G manoeuvres, but it helps
Ok. Point by point

1. Static Stability - Means that when a plane is flying on a straight line path, it gets disturbed /perturbed in any way, it comes back to it's original path. However,note , for this discussion we will restrict stability to pitch alone - ie. if you push the nose up/down or Musharraf up/down, the plane regains level flight, though in reality , the motions are coupled and you don't have a pure motion in axis alone

2. Statically Stable Aircraft - (refer point 1).. For this, CG is before the centre of lift . Ie, aircraft is nose heavy and will glide with nose pointing a bit down and Musharraf a little up.

3. How does it work . (Statically stable). The wings and tail work in opposition. Wing try to lift ,while tail pushes down. So total lift = wing lift LESS tail lift. So for a given weight of aircraft, the Wt = Lift --> Wt = Wing Lift LESS tail lift

4. Statically unstable --> It puts Nose in the air and Musharraf to ground in a twink of an eye. So to keep nose pointed level, you need to LIFT the Musharraf . So wings and tail work together and not in opposition. So total lift = wing lift PLUS tail lift. So for a given weight of aircraft, the Wt = Lift --> Wt = Wing Lift PLUS tail lift

5. Between 3 and 4 , you can see that 4 is far more efficient as a configuration. 4 gives you greater range , greater payload, because you need only a smaller wing .. So for a given wt and power, 4 carries stuff further and faster.

6. Manoeuvrability --> To raise the nose, 4 will also raise it faster , because gravity does the job (and will do it very quickly), while in 3, you have to work against gravity to raise the nose. Raising nose and putting Musharraf down is foundation of maneuvrability


Remember I told you that there are no "pure motions" . Consider a roll. If you have to roll clockwise, the the left wing has to develop MORE lift than the right wing (ideally for max rate roll, left wing develops max lift upwards, while right wing develops max negative lift pointed downwards). However, as a result, the drag on both the wings will be different (drag is induced along with lift) , so if the lift in the right wing was not changed and you roll by just increasing lift in the left wing, the drag experienced by the left wing will be more and the aircraft will yaw to the left. Now this has to be corrected by a rudder input to keep the nose from yawing.

So in an aircraft with an FCS, all this can be programmed in and the pilot wont have to input rudder manually, like he would have been taught to do while flying a basic trainer. Also , if you are flying a mig 29, you have to keep an eye on the AoA indicator. In an FCS controlled plane, you don't have to, because the FCS wont allow you to exceed alpha without you knowing it and keeps you safe and prevents you from going to unsafe areas. That also gives you superior confidence to throw the plane around (as Dr Deodhare said in the interview). In a Mig 29, you probably wont do so and keep it to 95% of max limits, lest you stray and self destruct in unsafe flight regimes.
LokeshC
BRFite
Posts: 697
Joined: 15 Aug 2016 04:36

Re: LCA: News & Discussions - October 2016

Post by LokeshC »

shiv wrote:Some further thoughts/questions/doubts. The calculations inovlved are way way beyond my knowledge or capability

When you have a 10,000 kg plane flying at 600 kmph - there is a huge mass of air flowing (at 600 kmph) past a control surface - say an elevator.

The question is as follows: If the plane must roll the aileron must move. A miniscule deflection of the aileron will initiate a slow roll. But for a rapid roll - say 180 deg/second the aileron must be deflected much more than a "miniscule" deflection and that deflection has to happen fast. A fast, large elevator deflection against a 600 kmph stream of air on a large wing surface will require a large force applied quickly.

What would be the sort of force required to move such a elevator 5 degrees in 0.1 seconds. What force would be required to achieve the same movement in 0.02 seconds (20 ms)? How long after that movement would the aircraft actually start rolling and roll 10 degrees? How much difference would it make to the roll rate if the elevator moved in 0.1 seconds rather than 0.02 seconds?
No aero eggspurt, only basic 1/10 JEE level engineer here. So plizz to excuse any mistakes. An aileron increases the effective area and changes the "equillibrium" so to speak. So if the aileron has area surface area A. And it is held up 10 degrees in 600kmph. The "reference area" (the area that really matters) in the drag equation is a rectangle of size A*sin(10). Then you can plug the rest into the drag equation here https://en.wikipedia.org/wiki/Drag_equation. Other than reference area, rest all of them are pretty well known constants.

BTW what changes between completing the movement from 0.1 to 0.02 what changes is not force, its power. Work Done = Energy = Force*distance. Power = Energy/Time. i.e. You need higher power to get that thing moving faster.
Last edited by LokeshC on 14 Oct 2016 11:45, edited 1 time in total.
JayS
Forum Moderator
Posts: 4567
Joined: 11 Aug 2016 06:14

Re: LCA Tejas: News and Discussions

Post by JayS »

shiv wrote:
JayS wrote:
Shiv, We are talking of aircraft dynamic instabilities with characteristic period of microseconds. Humans can sustain large G values for fraction of a second (even the 9G turn will be sustained by the aircraft for only a few seconds before it settles to sustained turn rate). G is not the limiter here. But how fast your system can respond to the dynamic instability and correct it before it diverges. We have a 80Hz FCS on LCA right?? That's rate 80Hz is limited by the system response - system latency, a large part of which comes from the actuators.
Air Marshal Rajkumar, in his book on Tejas stated that the actuators work at 2 Hz/twice per second
Image
I actually have always doubted the "80Hz" actuation rate. That why I was quick to accept what Adarsh said above about 2Hz. Now we have two conflicting but equally credible sources. I will wait until I get more confirmation on this one way or the other. May be looking at US fighters or asking the question in AI-2017.

However what I said still remains equally valid, whether its 80Hz or 2Hz.

But I have doubts its as low as 2Hz, I have already said why. May be 2Hz for pilot input but more than that in total by the FCS in order to control the system, just as Adarsh said in his post later. It is more believable to me at this juncture.

A question to all - What kind of actuators are there in Tejas?? I have always assumed its Hydraulic actuators.

PS: F35 uses Electro-Hydrostatic Actuators EHA. Interestingly it seems the cooling is the limiting factor in sizing the actuators.
Last edited by JayS on 14 Oct 2016 11:56, edited 1 time in total.
LokeshC
BRFite
Posts: 697
Joined: 15 Aug 2016 04:36

Re: LCA: News & Discussions - October 2016

Post by LokeshC »

Ummm. While I am no aero-engineer. I have good experience in imaging and signal processing.

Re: The claw image posted above. That 80 hertz is converting from Analog domain to digital domain and is calibrated to remove instabilities. Basically making sure the "region of convergence" (http://fourier.eng.hmc.edu/e102/lecture ... node3.html) of a discrete time domain system completely covers the operating characteristic of the aircraft. The 80Hz is most likely NOT the actuator frequency. It reflects how sensitive the internal modeling dynamics are to signal discretization.

As you can see the initial filters are in s domain (laplace/continous/analog) after the a/d they are z domain (discrete time).
vina
BRF Oldie
Posts: 6046
Joined: 11 May 2005 06:56
Location: Doing Nijikaran, Udharikaran and Baazarikaran to Commies and Assorted Leftists

Re: LCA: News & Discussions - October 2016

Post by vina »

LokeshC wrote:Ummm. While I am no aero-engineer. I have good experience in imaging and signal processing.

Re: The claw image posted above. That 80 hertz is converting from Analog domain to digital domain and is calibrated to remove instabilities. Basically making sure the "region of convergence" (http://fourier.eng.hmc.edu/e102/lecture ... node3.html) of a discrete time domain system completely covers the operating characteristic of the aircraft. The 80Hz is most likely NOT the actuator frequency. It reflects how sensitive the internal modeling dynamics are to signal discretization.

As you can see the initial filters are in s domain (laplace/continous/analog) after the a/d they are z domain (discrete time).
We are switched! Me no elec engg, though I know the math and do this signal processing stuff in a totally different context.

Ah, then perfect guy to ask the question. I wrote about this when this was discussed earlier here as well. This 80 Hz CANNOT be the actuator frequency. This would be the sampling frequency as per the Nyquist rate (from what I remember of my hostel mates rave about sampling etc..for signal processing).

So the actual signal freq it has to filter and process will be 40 hz (max), with most probable around 25 hz . The actuator response will be around 2 hz. You cannot have a large flap or tail or elevon or something fluttering around at 80hz. Just the mechanics of that will be mind boggling, given the inertia of that kind of thing.

Also look up my post in response to Shiv on stable /unstable etc.
LokeshC
BRFite
Posts: 697
Joined: 15 Aug 2016 04:36

Re: LCA: News & Discussions - October 2016

Post by LokeshC »

Yes, the 80hz a/d means the system dynamics that are being modeled at that point would need to be band-limited to 40Hz. In other words whatever those dynamics maybe (changing airflow, heat transfer etc), change no more than 40 times a second in order to achieve stable, effective control of the system.
LokeshC
BRFite
Posts: 697
Joined: 15 Aug 2016 04:36

Re: LCA: News & Discussions - October 2016

Post by LokeshC »

vina garu, agree with your response to Shiv garus question.

Statically stable aircraft will have "stabilizers", 'usually' the smaller back wings to keep them "flying correctly", which comes at a cost of increased drag and limited maneuverability because "stabilizer" is supposed to prevent rapid movements.

Therefore the idea is to remove the big stabilizers and artificially stabilize the aircraft using computerized control systems that fake a stabilizer. It is a virtual stabilizer can be switched off when needed or can change its virtual size and shape when needed. The cost of stabilizing such an airframe (in terms of fuel and maintenance) might be more than a physical stabilizer but is highly useful when super high maneuverability is needed.

The A/D system in there is just to convert everything into a discrete input to a computer. Which then runs the control filter (the notch filter is a second order IIR filter, while the lags are low pass filters that look like delayed step functions with exponential decay: Look at #4 in "common z transforms" section in here: https://en.wikipedia.org/wiki/Z-transform )

very limited knowledge about aircrafts so, pliz to excuse onlee.
prahaar
BRF Oldie
Posts: 2831
Joined: 15 Oct 2005 04:14

Re: LCA: News & Discussions - October 2016

Post by prahaar »

LokeshC, sampling frequency does not necessitate operation at half the rate. I do not understand the operations described in the F-18 picture so cannot comment. But for example, audio sampled at 48KHz (A/D) contains all the audio frequencies. For example, if it is just stepper motor actuator information derived from A/D converter, it need not have the 40Hz upper limit.
LokeshC
BRFite
Posts: 697
Joined: 15 Aug 2016 04:36

Re: LCA: News & Discussions - October 2016

Post by LokeshC »

BTW: I think I know what the claw is doing at a 100000 feet level:

Its a PI controller: https://en.wikipedia.org/wiki/PID_contr ... controller

With a lag compensator: https://en.wikipedia.org/wiki/Lead%E2%8 ... ompensator
Locked