Su27 Wing Thing

[DarkFire]

Anybody else notice that for the duration of the pull, control indicator shows the commanded input riding the edge of the soft limiter? The soft limiter actually move considerably rearward throughout the maneuver.

 

It looks to me like the player was holding a stead stick position, around 60% deflection, and that the soft limit had kicked in around 55%, then ramped up to commanded input, and that this is what broke the plane.

 

So, to the player, it would look like an uncommanded pull.

 

I hadn't noticed but having viewed it again you're right. Interesting. So, if we posit a sharp control input, possibly caused by not having input curves, the soft limiter chases the commanded input up to it's maximum value as opposed to it's value at that particular time, caused by the soft limiter laging slightly behind the true value. Could easily cause the observed behaviour.

[SinusoidDelta]

I still argue it's the pilot's responsibility to react to the alarming onset of Nz. All it takes is some forward stick pressure to survive.

[Ironhand]

I hadn't noticed but having viewed it again you're right. Interesting. So, if we posit a sharp control input, possibly caused by not having input curves, the soft limiter chases the commanded input up to it's maximum value as opposed to it's value at that particular time, caused by the soft limiter laging slightly behind the true value. Could easily cause the observed behaviour.

As far as I can tell with limited experimentation, curve or no curve, the soft limiter and control input behave the same way. If you pull through the limiter, what you get is what you see in that track. At least that's what it looked like to me with some quick experiments early this morning. But I reserve the right to be wrong...

[DarkFire]

As far as I can tell with limited experimentation, curve or no curve, the soft limiter and control input behave the same way. If you pull through the limiter, what you get is what you see in that track. At least that's what it looked like to me with some quick experiments early this morning. But I reserve the right to be wrong...

 

No, you're absolutely right - it does still happen no matter what curvature is applied. Tried it at 0, 10, 15 and 20 curvature values. All nearly identical to the 8th screen shot you posted. I think it might be a little less pronounced with higher curvature values, but that might just be because the soft limiter effectively has a little more time to catch up with the commanded input. So, I guess the question then is whether the observed behaviour is:

 

1. Accurate to the ACS system in the real Su-27
   
2. An artefact of the way in which the game modells the ACS.
   
3. A coding bug.
   

 

I wish I could read Russian or Spanish to see if the actual Su-27 technical manual has anything to say on the subject. In any case I guess the take home is that conducting high-G manoeuvres at anything over ~0.9M is inherently very dangerous.

[rrohde]

I guess the take home is that conducting high-G manoeuvres at anything over ~0.9M is inherently very dangerous.

 

That sums this all up nicely. Thanks! :thumbup:

[lunaticfringe]

It does. And given the manufacturer says so themselves, no one should be surprised by it.

[DarkFire]

It does. And given the manufacturer says so themselves, no one should be surprised by it.

 

Well said. :thumbup:

 

I guess a lot can be explained by the basic design philosophies that produced the Su-27. Sure it has supersonic capabilities and a nice long range, but Russian design philosophy has always believed strongly in WVR maneuverability, hence the corner velocity, the speed produced by cruise throttle setting, the G-load settings programmed in to Betty; it's clearly all designed for fighter combat at fairly high subsonic speeds, probably with a high altitude, high-speed interception mission as a secondary design objective.

[Frostie]

But why do we get instant wing breakage? No deformation, judder etc.

The ultimate structure load is 1.5 times the limit load, an airframe is supposed to be able to withstand up to 3 seconds of ultimate load yet we have instantaneous destruction at 1.3 to 1.5 times limit load.

[DarkFire]

But why do we get instant wing breakage? No deformation, judder etc.

The ultimate structure load is 1.5 times the limit load, an airframe is supposed to be able to withstand up to 3 seconds of ultimate load yet we have instantaneous destruction at 1.3 to 1.5 times limit load.

 

Apparently this actually happened during flight testing. Described in this documentary just after the 53:30 mark, the test pilot was conducting strength testing when he lost the outboard section of his port wing. Unfortunately it doesn't go in to great detail.

 

[Frostie]

Apparently this actually happened during flight testing. Described in this documentary just after the 53:30 mark, the test pilot was conducting strength testing when he lost the outboard section of his port wing. Unfortunately it doesn't go in to great detail.

 

 

I can't look at the video right now but if its the one i'm thinking of he pretty much goes from one extreme to the other, much like bending a ruler one way and then the other, it snaps.

[probad]

But why do we get instant wing breakage? No deformation, judder etc.

you forget the wings visibly deform during regular maneuvering already :)

[Frostie]

you forget the wings visibly deform during regular maneuvering already :)

 

That is normal for all aircraft within standard parameters, it's not permanent.

[karambiatos]

But why do we get instant wing breakage? No deformation, judder etc.

The ultimate structure load is 1.5 times the limit load, an airframe is supposed to be able to withstand up to 3 seconds of ultimate load yet we have instantaneous destruction at 1.3 to 1.5 times limit load.

If i remember correctly the su-27 manual doesn't explictly state what 1.5 times limit load implies, whether it's a point where the wings start getting damaged, or a point where they are at risk of breaking apart at any moment.

[ShuRugal]

So, I guess the question then is whether the observed behaviour is:

 

1. Accurate to the ACS system in the real Su-27
   
2. An artefact of the way in which the game modells the ACS.
   
3. A coding bug.
   

.

 

 

I would guess a bit of each. IRL, the pilot would feel the stick hit his soft limiter and would need to deliberately pull through it. He would also feel the soft limit change if he were riding its edge. In the sim, you get no tactile indication that you have hit the stick limiter (unless you have FFB??), and the only way to pull through is to hold the wheel-brake binding.

 

As a result, if you accidentally pull past your desired point, and that point is close to the limiter, you won't know you've done it. Then, when the plane slows down, the flight computer thinks its safe to pull some more pitch, so it increases the soft limit, now you get what appears to be an uncommanded pull.

 

So far, we've only described 1 and 2. But with our brand-new stress-damage model for the plane, which may not yet be correctly dialed to provide accurate results in edge-of-the-envelope situations....

[DarkFire]

I would guess a bit of each. IRL, the pilot would feel the stick hit his soft limiter and would need to deliberately pull through it. He would also feel the soft limit change if he were riding its edge. In the sim, you get no tactile indication that you have hit the stick limiter (unless you have FFB??), and the only way to pull through is to hold the wheel-brake binding.

 

As a result, if you accidentally pull past your desired point, and that point is close to the limiter, you won't know you've done it. Then, when the plane slows down, the flight computer thinks its safe to pull some more pitch, so it increases the soft limit, now you get what appears to be an uncommanded pull.

 

So far, we've only described 1 and 2. But with our brand-new stress-damage model for the plane, which may not yet be correctly dialed to provide accurate results in edge-of-the-envelope situations....

 

I think you're probably right. It's a shame that nobody makes high-end force feedback flight sticks, otherwise I think that would probably be the way to go for hard-core flight sims like DCS. It's also a shame that FFB is more or less broken in DCS at the moment.

 

Hmm. I might start flying with the controller position display always on & keep half an eye on the soft limiter during high speed engagements.

 

I wonder what happens if the ACS is turned off at high mach numbers. Insta-death probably. Might give that a try just to test it...

[Ironhand]

So, if I'm understanding all of this correctly, the G limiter limits you to 8 Gs. What that means is that, if you are somewhere in the neighborhood of 21,400 kg gross weight, you can pull your flightstick back into your lap, hold it there, and the wings will never come off. They won't come off because the limiter keeps you at or below those 8 Gs. In fact, you can watch it doing just that. Your virtual stick rocks forward and back depending on the G loading at the moment as long as you keep your real stick pulled all the way back. And you're safe because that particular G loading keeps you within the aircraft's structural limits at all airspeeds. You'll exceed the operational limits but not the structural limits at airspeeds greater than M 0.85.

 

Problems ensue, however, when you try the same thing at higher gross weights because you will exceed the structural strength of the aircraft before you get to limiter's 8 G setting. I'm pretty sure I've noticed the same forward and back motion of the virtual stick at higher gross weights. I might play around with it this evening. Using the 171,000/(Your Gross Weight)=Max operational G for up to М <0.85 formula, I should be able to see the stick movement before the wings come off with a 26,000 kg gross weight.

Edited October 17, 2016 by Ironhand

[ShuRugal]

I feel like the fact that the OP was doing an inverted pull somehow also factored into this. The pitch channel of the ACS doesn't know if the plane is inverted (unless someone thought to program that in) so all it is going to see is that as that first quarter-loop down progresses, the G-loading changes.

 

Of course, the same theoretical problem could be in the damage model: Does the damage model allow for the fact that you get 1G for free when inverted?

[DarkFire]

So, if I'm understanding all of this correctly, the G limiter limits you to 8 Gs. What that means is that, if you are somewhere in the neighborhood of 21,400 kg gross weight, you can pull your flightstick back into your lap, hold it there, and the wings will never come off. They won't come off because the limiter keeps you at or below those 8 Gs. In fact, you can watch it doing just that. Your virtual stick rocks forward and back depending on the G loading at the moment as long as you keep your real stick pulled all the way back. And you're safe because that particular G loading keeps you within the aircraft's structural limits at all airspeeds. You'll exceed the operational limits but not the structural limits at airspeeds greater than M 0.85.

 

Problems ensue, however, when you try the same thing at higher gross weights because you will exceed the structural strength of the aircraft before you get to limiter's 8 G setting. I'm pretty sure I've noticed the same forward and back motion of the virtual stick at higher gross weights. I might play around with it this evening. Using the 171,000/(Your Gross Weight)=Max operational G for up to М <0.85 formula, I should be able to see the stick movement before the wings come off with a 26,000 kg gross weight.

 

Yes. At an all-up weight of ≤ 21,400 Kg no matter what you do, no matter what crazy stuff you command the aircraft to do, if the ACS is operational it'll make sure that you'll never exceed the structural limits. At any all-up weight > 21,400Kg you have to take the max G in to account via the calculation as stated in the manual.

 

Of course switching the ACS to manual using the Button Of Death is an entirely different matter :music_whistling:

 

ShuRugal: I don't think the damage model does take aircraft attitude in to account. Maximum structural G load is a hard limit and I think that exceeding it by any means results in airframe destruction.

[Pocket Sized]

I feel like the fact that the OP was doing an inverted pull somehow also factored into this. The pitch channel of the ACS doesn't know if the plane is inverted (unless someone thought to program that in) so all it is going to see is that as that first quarter-loop down progresses, the G-loading changes.

 

Of course, the same theoretical problem could be in the damage model: Does the damage model allow for the fact that you get 1G for free when inverted?

 

This was discussed in depth a few pages ago.

 

TL;DR it doesn't matter. The G you feel in the cockpit is the same the wings feel regardless of attitude. Turn rate for a given G will be different but that doesn't really matter here.

 

The plane was grossly overstressed given the weight and Mach number.

 

On top of that, it was still experiencing some roll/yaw coupling because the plane didn't have time to settle before the pull, which led to asymmetric loading and further stress on the airframe.

[GGTharos]

Yes. At an all-up weight of ≤ 21,400 Kg no matter what you do, no matter what crazy stuff you command the aircraft to do, if the ACS is operational it'll make sure that you'll never exceed the structural limits. At any all-up weight > 21,400Kg you have to take the max G in to account via the calculation as stated in the manual.

 

Of course switching the ACS to manual using the Button Of Death is an entirely different matter :music_whistling:

 

There's an AoA limiter. I don't recall there being a G limiter, just posted G limitations and the Nadia programming.

 

 

ShuRugal: I don't think the damage model does take aircraft attitude in to account. Maximum structural G load is a hard limit and I think that exceeding it by any means results in airframe destruction.

 

Attitude is irrelevant. The only thing that is relevant are G's 'felt' by the airframe.

[Pocket Sized]

And for my last post in this entirely redundant thread:

 

No, the Russians aren't too stupid to design a plane with care free handling. Neither are Americans. It's just not worth the effort. You can train the pilots and/or add a warning system which is much easier and more useful. (Something both sides did)

 

There was an F-15 which, long story short, literally tore its wings off. A combination of factors including an inoperative overload warning system lead the pilot to accidentally pull way too many G's.

 

Not to mention more primitive aircraft like aerobatic prop planes, airliners (albeit you'd need a few hundred pounds of force), and a vast majority of general aviation planes can all be overloaded to the point of catastrophic failure.

[Ironhand]

There's an AoA limiter. I don't recall there being a G limiter, just posted G limitations and the Nadia programming...

You're right. That's what its supposed to be. My error. It was just that it was holding me to roughly 8 Gs that made me think of it that way.

[DarkFire]

There's an AoA limiter. I don't recall there being a G limiter, just posted G limitations and the Nadia programming.

 

You're right of course, though G will be proportional to AOA up to the critical limit.

 

No, the Russians aren't too stupid to design a plane with care free handling.

 

No they're not. In fact both the Su-27 (and -33) and the MiG-29 were all deliberately designed not to have care free handling.

[jackmckay]

As a refresher to this conversation:

 

You have been shown, repeatedly, the same data direct from the manufacturer showing that you well exceeded the capability of the Su-27 to withstand overstress at the airspeed (M 1.26) and weight (in excess of 21,400 kg).

- Direct from manufacturer? You want to say that you contacted Sukhoi Design Bureau and they gave YOU structural data? Or you(ED) just read the Su-27 manual?

The track you provided, as outlined by Ironhand, showed explicitly that you still had the stick well aft of 65% at the point of the wings coming off- the exact point that just prior to him I showed that in *any* relationship between inertial coupling- gross, slight, or none, the airframe will disintegrate- in specific accordance to the relationship of AoA and G that was previously outlined- that 10 units of AoA at your stated speed and weight well exceeds the G limit as outlined by Sukhoi (6.5).

- G safe operational limit not ultimate structural limit. In fact I posted several tracks in which I exceeded stick pull up to 100% at very same conditions and survived the very same maneuver in which my wings broke off in MP. Repeatedly.

You were accurately refuted as to your claim that ED fails to calculate G correctly in relation with AoA at the speed given.

- Yes, ED fails to calculate wing snap-off G limit point, point of wing separation. Until ED shows CFD/FEM analysis data there's no at least some guarantee that damage model is accurate at all. Actually its completely ****ed up. Period.

You were shown, repeatedly, the charts concerning your weight and load being well in excess of what the aircraft was designed for at the speed in question.

- Su-27 aircraft is designed as all weather high maneuverability fighter with just predicted right load-out for exact purpose and just right predicted flight envelopes. The safe margin excess is given by factor of safety that you or the ED does not know. They can just speculate but I can tell you that Russian air-frames are designed for low maintenance front line operations traditionally and that means higher safety factor than US airplanes.

 

You were refuted as to the intent and translation of the charts you mentioned being in relation to AoA limits regarding stall, rather than structural limitations. And the specific relationship between Mach, altitude, AoA, and G was outlined. Twice.

- Wing load is direct proportional to frontal projection of wing itself, that gives correlation between speed, AoA and vertical force vector on wetted, stream exposed wing area. High speeds implicit high loads on high AoA and high (ultimate) loads are matter of discussion here NOT stall behavior.

You present US FAR regulations, rather than nation of regulations regarding the aircraft in question- regulations that Sukhoi, in fact, does not have to meet, because they're not under the auspices of the FAA for the construction of tactical aircraft.

- Yes, just for example case but again Russian military safety standards are higher than Western. Guaranteed. Russian FOS is always higher than US FOS as a matter of fact it goes up 2,5 (2vs5max at static loads!) as I was tough at mechanical engineering college from certified professors from former Yugoslavia that had both insight from USSR and US technology. Prof for that is maintenance time needed for Russian versus US airplanes.

You present material deformation data that, as previously stated, has absolutely nothing to do with the construction of the Su-27.

- That stress-strain chart is for example and it shows steel elastic properties as focus was on ELASTIC properties. In fact, you have to know that airplanes are not made entirely of aluminum. Some very important high load cycles parts are made of either steel or titanium or any other suitable material properly strengthened for right design given function. Air-frames are composite structures.

 

The only thing consistent in this conversation at this point is your continued denial of the facts at hand. Fault for this failure does not lay with ED. Not the laws of physics. Not Sukhoi. And not anyone here presenting the correct evidence to refute your wholly disjointed and entirely mediocre claims to the contrary. You broke the aircraft. Anything you say from this point forward in

opposition to this amounts to the grasping of straws.

- I stand before my claims because I am mechanical engineer Sir. That is my profession. You on the other hand are, as I figure out from your words, programmer, and you read lot of manuals. How many manuals for new design airplanes have you written, or how many mechanical components have you designed, tested and verified? And if you don't believe me, what is your legit right, why don't you first ask some mechanical engineer about airplane structure design basics.

And as far as anyone with any sort of knowledge of the subject should be concerned, this matter is closed.

So this is my conclusion on this topic regarding Su-27 airframe design aspects in DCS.

 

- DCS Su-27 IS NOT CERTIFIED by TsAGI institute who designed winbox of Su27S;

 

- DCS HAS NO VALIDATED independent structural tests committed on ANY flight (damage) model using certified and industry approved FEM/CFD software at least (for ex. ANSYS/FEMAP/COMSOL).

 

- Flight Manuals are written BY designers (aeronautical/mechanical engineers) FOR users (pilots) to maintain high availability and to fulfill airframe predicted lifetime - that means flight performance margins are NOT kill switches. Life of pilot is too precious. Period.

 

Tight relation to real world damage model with ED's damage model shows this picture. Here we can see perfectly functional F-15 returning RTB from SATAC mission after killing one of my teammate. Praise the western engineers.

 

 

PS: My stick has wide (5cm) play without extension. Its worn out. That could cause that excess in MP. Not sure.

 

Birdstirke OUT.

Edited October 20, 2016 by jackmckay

[GGTharos]

- Yes, ED fails to calculate wing snap-off G limit point, point of wing separation. Until ED shows CFD/FEM analysis data there's no at least some guarantee that damage model is accurate at all. Actually its completely ****ed up. Period.

 

Until you show some data to support any claims at all, the industry 1.5x factor is a fine assumption.

 

Everything else is sheer drivel.

 

PS: My stick has wide (5cm) play without extension. Its worn out. That could cause that excess in MP. Not sure.

 

Maybe that should have been your first point of investigation instead of the salt-truck rodeo.

 

You screwed up, suck it up and deal with it.