Wing snap vs joystick?

[Bucic]

14 minutes ago, AndyJWest said:

TacView data is sampled at too low a rate to give accurate peak G or AoA readings for a snap manoeuvre. 

But in the slow portion the G changed from 1.5 to 3.0 with no change in AoA. It can bee seen on one of the TacView video recordings. Check out the edited post.

[AndyJWest]

Your Tacview file shows you rapidly pulling 6.9 G at 6:00:26, and then rapidly pulling 8.5 G at 6:00:51, pushing to -2.4 G within 1 second. Given that the actual peak Gs may exceed those figures due to sampling errors, I really don't think that whatever happens afterwards can be indicative of anything much beyond DCS modelling cumulative damage. You pulled the wings off with rapid-onset high-G manoeuvres. 

 

Edited December 10, 2024 by AndyJWest

[Bucic]

6 hours ago, AndyJWest said:

Your Tacview file shows you rapidly pulling 6.9 G at 6:00:26, and then rapidly pulling 8.5 G at 6:00:51, pushing to -2.4 G within 1 second. Given that the actual peak Gs may exceed those figures due to sampling errors, I really don't think that whatever happens afterwards can be indicative of anything much beyond DCS modelling cumulative damage. You pulled the wings off with rapid-onset high-G manoeuvres. 

 

Doh, I simply reached for the main menu with charta and haven't found any with "G" in the name

I'll look into it tomorrow. I don't have an agenda - is/isn't the structural limit too low. For me it isn't. It's just the penalty is over the top.

The G onset is fishy though. I have one video with G rising from 1 to 10 in something like 0.2 seconds!

Edited December 10, 2024 by Bucic

[AndyJWest]

You may be being misled by the sampling rate. I can't determine this directly from DCS .acmi files, because they are compressed, but the files from e.g. IL-2 GB are uncompressed, and have a sampling rate of 0.2 seconds. If you take the sample rate into account, along with the fact that Tacview may  interpolate some output data that isn't provided directly, you need to be sceptical about peaks in G etc.

 

 

 

 

 

[NineLine]

I'll review the tracks soon, thanks. 

9 hours ago, AndyJWest said:

TacView data is sampled at too low a rate to give accurate peak G or AoA readings for a snap manoeuvre. 

I have always found tacview very helpful, but you cannot always trust the data 100%. I have seen a few inconsistencies for sure. 

[Bucic]

51 minutes ago, NineLine said:

I'll review the tracks soon, thanks. 

I have always found tacview very helpful, but you cannot always trust the data 100%. I have seen a few inconsistencies for sure. 

Noted! Tacview authors themselves  documented its limitations. And it was the very first time I actually used it so, have faith I have more experience with racing telemetry in motec i2 pro.

[AndyJWest]

Yeah, Tacview is very useful, as long as you don't try to read more into it than it is capable of. And that can vary quite a bit depending on the source - some DCS modules seem to provide more data than others e.g. on things like fuel consumption. It would be nice if there was more consistency with this, though I suppose it is hardly a priority for most people.

Getting back to the F-5E, are we sure that the issue isn't at least partly down to simmers being able to move their joysticks faster than would be feasible in the real aircraft? 

 

[Bucic]

3 minutes ago, AndyJWest said:

Getting back to the F-5E, are we sure that the issue isn't at least partly down to simmers being able to move their joysticks faster than would be feasible in the real aircraft? 

 

I'd say we are pretty sure of the opposite

[NineLine]

41 minutes ago, AndyJWest said:

Getting back to the F-5E, are we sure that the issue isn't at least partly down to simmers being able to move their joysticks faster than would be feasible in the real aircraft? 

I believe that is a big part of it, you have to consider the artificial feel system and what it does, now I do not have FFB yet, but I do ave the AVA base with the heaviest sprints loaded, and I think that has certainly helped to some degree not being so heavy-handed, many of the examples I have seen are people trying to Tom Cruise the stick. I am gonna do a vid of my findings tonight or tomorrow. Thanks all!

[=475FG= Dawger]

8 hours ago, AndyJWest said:

Yeah, Tacview is very useful, as long as you don't try to read more into it than it is capable of. And that can vary quite a bit depending on the source - some DCS modules seem to provide more data than others e.g. on things like fuel consumption. It would be nice if there was more consistency with this, though I suppose it is hardly a priority for most people.

Getting back to the F-5E, are we sure that the issue isn't at least partly down to simmers being able to move their joysticks faster than would be feasible in the real aircraft? 

 

This is mostly correct. The F-5 in DCS does not have artificial feel modeled and the rate of deflection of the HSTAB is nearly instantaneous. You can generate 13G without appreciable nose movement. 
 

This can be overcome with some practice but it makes the aircraft extremely unrealistic. 
 

Also, the roll limits are misinterpreted, so that any roll input, however minor takes the wings off at high G level. This is a big issue fighting PvP against Gen 4 opponents and when coming off high closure guns passes. 

[nairb121]

19 hours ago, Bucic said:

The G onset is fishy though. I have one video with G rising from 1 to 10 in something like 0.2 seconds!

I have a suspicion that this ties into the AOA issue documented elsewhere. Since the F-5's L/D range is compressed into a smaller AOA range, less nose movement corresponds to a higher change in lift, causing a realistic pitch rate to create an unrealistic g-onset.

[Bucic]

5 minutes ago, nairb121 said:

I have a suspicion that this ties into the AOA issue documented elsewhere. Since the F-5's L/D range is compressed into a smaller AOA range, less nose movement corresponds to a higher change in lift, causing a realistic pitch rate to create an unrealistic g-onset.

Could you please link the bug report sonthat there's a good reference? I know that bug report, just don't have the link handy.

[nairb121]

6 minutes ago, Bucic said:

Could you please link the bug report sonthat there's a good reference? I know that bug report, just don't have the link handy.

 

Edited December 11, 2024 by nairb121

[NineLine]

19 hours ago, =475FG= Dawger said:

This can be overcome with some practice but it makes the aircraft extremely unrealistic. 

Hmmm, I am not sure I agree with this 100%, while it's true the Artificial Feel system will help this issue, and we do need it, I am not sure practicing and learning to have softer hands makes it 'extremely unrealistic'. The biggest missing part of any flight simulator is not having that butt in the sear or feedback from the stick. I have a feeling when the Artificial Feel system is introduced we will have people complaining about the less responsiveness in the stick, we have seen this when the Bf-109K-4 came out and the controls freezing was modelled. 

The Spitfire is very similar in that you can pull crazy hard on the stick and ruin your day, in this case its true to life and you need very gentle hands in the real thing to not over do it on the stick. 

Now back to Bucic's track, what I saw were 'extremely unrealistic' control movements. No way a pilot is pulling like this in any scenario, I understand it was trying to prove a point, but we already know that the Artificial Feel system isn't there and the limits imposed there are not in place, until then you need to monitor your Gs and how heavy your hands are on the stick. Instantaneous G is much worse than smooth onset and that is what I see in the track. 

One of my many test flights here if anyone is interested. I used Bucic's track and took control, things to keep in mind, its an air start, 100% fuel at the start, 150G centerline tank, full and of course missiles on the wing tips. This cuts down the limits set in the manual as well, as you can see I feel like I put a lot of Gs on the wings before it finally broke. 

 

[=475FG= Dawger]

4 hours ago, NineLine said:

Hmmm, I am not sure I agree with this 100%, while it's true the Artificial Feel system will help this issue, and we do need it, I am not sure practicing and learning to have softer hands makes it 'extremely unrealistic'. The biggest missing part of any flight simulator is not having that butt in the sear or feedback from the stick. I have a feeling when the Artificial Feel system is introduced we will have people complaining about the less responsiveness in the stick, we have seen this when the Bf-109K-4 came out and the controls freezing was modelled. 

The Spitfire is very similar in that you can pull crazy hard on the stick and ruin your day, in this case its true to life and you need very gentle hands in the real thing to not over do it on the stick. 

Now back to Bucic's track, what I saw were 'extremely unrealistic' control movements. No way a pilot is pulling like this in any scenario, I understand it was trying to prove a point, but we already know that the Artificial Feel system isn't there and the limits imposed there are not in place, until then you need to monitor your Gs and how heavy your hands are on the stick. Instantaneous G is much worse than smooth onset and that is what I see in the track. 

One of my many test flights here if anyone is interested. I used Bucic's track and took control, things to keep in mind, its an air start, 100% fuel at the start, 150G centerline tank, full and of course missiles on the wing tips. This cuts down the limits set in the manual as well, as you can see I feel like I put a lot of Gs on the wings before it finally broke. 

 

It makes it unrealistic because, in the real world, there is inertia to overcome and it takes time to get the nose moving and for the G to onset. 
 

With nearly instantaneous control deflection, techniques that work in the real world will snap the wing off. 
 

For example, break turns. 
 

The goal of a proper break turn is to alter direction as quickly as possible i. e “square the corner”

To do that, a common technique is to unload, roll (because breaking uphill is a very bad thing), take out the aileron input, pull the stick abruptly back to initiate a hard break and immediately position it forward to avoid the over G. 
 

If you use this technique in the DCS F-5, the instantaneous control deflection will cause instantaneous G onset and the wings will depart.

This is why it is extremely unrealistic. The G onset is so rapid it appears that G is tied to stick position instead of the aircraft acceleration.

It forces one to use unrealistic techniques in order to retain possession of important aircraft components. 

[NineLine]

3 hours ago, =475FG= Dawger said:

It makes it unrealistic because, in the real world, there is inertia to overcome and it takes time to get the nose moving and for the G to onset. 
 

With nearly instantaneous control deflection, techniques that work in the real world will snap the wing off. 
 

For example, break turns. 
 

The goal of a proper break turn is to alter direction as quickly as possible i. e “square the corner”

To do that, a common technique is to unload, roll (because breaking uphill is a very bad thing), take out the aileron input, pull the stick abruptly back to initiate a hard break and immediately position it forward to avoid the over G. 
 

If you use this technique in the DCS F-5, the instantaneous control deflection will cause instantaneous G onset and the wings will depart.

This is why it is extremely unrealistic. The G onset is so rapid it appears that G is tied to stick position instead of the aircraft acceleration.

It forces one to use unrealistic techniques in order to retain possession of important aircraft components. 

All this would be heavily impacted by the Artificial Feel System, the amount you can move the controls now will be impacted. So its not the wings breaking that is really the issue, its this missing system, I am pushing for it to be added. Thanks. 

[=475FG= Dawger]

1 minute ago, NineLine said:

All this would be heavily impacted by the Artificial Feel System, the amount you can move the controls now will be impacted. So its not the wings breaking that is really the issue, its this missing system, I am pushing for it to be added. Thanks. 

Yes, I have been saying this for about two years now.

I actually would have no problem with the current catastrophic wing failure, as unrealistic as it is, if artificial feel was modeled to make it impossible to generate 13G in a few thousandths of a second and the roll limits were interpreted properly so that minor aileron application doesn't snap the wings off.

I posted this request a year ago.  

 

[NineLine]

Yes, and the request for this system long predates your report, with the added focus on the F-5 I am using every bit of sway I have to get it done. 

[Bucic]

On 12/10/2024 at 4:48 PM, AndyJWest said:

Your Tacview file shows you rapidly pulling 6.9 G at 6:00:26, and then rapidly pulling 8.5 G at 6:00:51, pushing to -2.4 G within 1 second. Given that the actual peak Gs may exceed those figures due to sampling errors, I really don't think that whatever happens afterwards can be indicative of anything much beyond DCS modelling cumulative damage. You pulled the wings off with rapid-onset high-G manoeuvres. 

 

 

All in all Tacview isn't what I expected. I know it ain't sub-microsecond but... Not to say it's a bad piece of software but I was hoping it's going to be a "be all" in cases exactly like the one at hand here.

That said, what the hell is this supposed to be? It's over a 4 second time period. The G rises disproportionatelly to AoA. G is directly proportional to lift force, lift force is directly proportional to AoA, with only airspeed as a variable. So what is happening here?

 

[=475FG= Dawger]

4 minutes ago, Bucic said:

All in all Tacview isn't what I expected. I know it ain't sub-microsecond but... Not to say it's a bad piece of software but I was hoping it's going to be a "be all" in cases exactly like the one at hand here.

That said, what the hell is this supposed to be? It's over a 4 second time period. The G rises disproportionatelly to AoA. G is directly proportional to lift force, lift force is directly proportional to AoA, with only airspeed as a variable. So what is happening here?

 

An educated guess would be slat deployment.
 

One of the “techniques “ to prevent snapping the wings off is to set the flap system to off because when the slats auto-schedule at about 320 knots, the G spike can induce wing separation. 
 

 

[Bucic]

@=475FG= Dawger

Yeah, I should have watched my own track again before posting that comment

[=475FG= Dawger]

22 hours ago, NineLine said:

Yes, and the request for this system long predates your report, with the added focus on the F-5 I am using every bit of sway I have to get it done. 

I appreciate that you are working to make it happen.

I am sure you understand my frustration due to issues introduced by ED left to linger for over three years now in what once was my daily MP ride.

I hope whoever decided to introduce catastrophic wing failure to the F-5 without fully exploring its impact has since been re-educated.

Its too bad this wasn't recognized years ago and the F-5 reverted.

At this point, I am pushing the issue out of principle. The past is gone and cannot be re-created.

I am unlikely to fly the F-5 again, in any case.

I will say that I do not trust ED to not do something like this in the future so I won't buy an ED produced module unless that trust is somehow restored.

How that would happen is beyond me.

 

[RustBelt]

On 12/11/2024 at 6:10 AM, =475FG= Dawger said:

 

This can be overcome with some practice but it makes the aircraft extremely unrealistic. 
 

Unless you’re flying with a floor mounted stick with an extension, and progressive springs or FFB, it’s inherently unrealistic whatever you do. Desktop joysticks are a compromise of convenience. They, not the sim, is the problem. As someone with a seat mounted FFB stick, faked in “damping” would be extremely obnoxious. Because I did learn the feel to fly the plane as it’s built. With my elbow and shoulder, not my wrist. 
 

if you’re using the wrong tool for the job, don’t expect great results. 

[=475FG= Dawger]

23 minutes ago, RustBelt said:

Unless you’re flying with a floor mounted stick with an extension, and progressive springs or FFB, it’s inherently unrealistic whatever you do. Desktop joysticks are a compromise of convenience. They, not the sim, is the problem. As someone with a seat mounted FFB stick, faked in “damping” would be extremely obnoxious. Because I did learn the feel to fly the plane as it’s built. With my elbow and shoulder, not my wrist. 
 

if you’re using the wrong tool for the job, don’t expect great results. 

Floor mounted stick, 20 cm extension. 

15,000 hours in real airplanes.

The DCS F-5 is the issue. 

The F1 and F-4 are the examples to compare the F-5 to.

But thanks for adding to the huge pile of "GitGud".

I use it to fertilize my garden.

[RustBelt]

So what you’re saying is, you want DCS to airbus you because it’s easier than learning what not to do. 
 

And yes, I’m going to go out on a limb and guess some of that 15,000hrs was in an airbus. Because only FBW drivers can’t understand how you have to fly a sim different to a real plane.