investigating Hornet's Slow-State Energy Bleed?

[NineLine]

7 hours ago, DummyCatz said:

I hope ED take it more seriously

Both threads are marked as investigating, thanks. 

[DummyCatz]

Apart from the said SEP issues, there're a few more things that require investigation in the provided trackfile:

1. G-limit is not consistent and dropped from 7.6g to less than 7g when bleeding speed from Mach 0.8 to corner speed.

2. The lift limit is at 34° AOA according to the NASA paper. The aircraft is supposed to reach the lift limit at corner speed. However in the trackfile, the aircraft is far from reaching the lift limit (34° AOA), even if it's passing the corner speed.

Edited August 24, 2024 by DummyCatz

[HWasp]

On 8/6/2021 at 11:16 PM, HWasp said:

I could not open your track for some reason, but for me the F-18 can sustain 4,6 G in horizontal flight at 10000 feet easily with a center tank (total fuel between 7000 and 6000)  (and around 3,8 Gs at 20000 feet horizontal ...)

What altitude did you try that?

 

Also comparing the DCS AoA vs G and the video's AoA vs G, it turns out they get less Gs for the AoA compared to DCS:

 

Video: 

290 kts at 13800 feet  ---- AoA 22,4 degrees ----> 4.6 G (or 4,8 its blurred)

 

DCS:

290 kts at 13700 feet ---- AoA 20,4 degrees ------> 5,3 G

 

This can vary with weight a lot of course, but they can't be too heavy in the video since they have 7.3 as max G on the hud

 

 

 

There was on old thread, still with the old FM, where I tried to compare DCS to some HUD footage.

In DCS it took quite a bit less AoA to reach the same G. Just for info, as these are of course connected.

Edited August 25, 2024 by HWasp

[Stackhouse]

 

On 8/24/2024 at 11:44 AM, DummyCatz said:

Apart from the said SEP issues, there're a few more things that require investigation in the provided trackfile:

1. G-limit is not consistent and dropped from 7.6g to less than 7g when bleeding speed from Mach 0.8 to corner speed.

2. The lift limit is at 34° AOA according to the NASA paper. The aircraft is supposed to reach the lift limit at corner speed. However in the trackfile, the aircraft is far from reaching the lift limit (34° AOA), even if it's passing the corner speed.

 

Isn't the HARV in the NASA paper a completely different aircraft than the 18C Lot 20 EPE that we have in the game?

 

Looking at the F-18 HARV Aircraft Specifications this jet is nothing like the one that we have in DCS.

 

For starters is uses two F404-GE-400 engines. Our F/A-18C does not use 400 engines, it is powered by 402 engines. The 402 engines provide increased thrust and increased fuel efficiency over the 400 engines used in earlier 18C lots and the NASA HARV. Namely about 1,700 lbs per engines to 3,400 lbs total thrust.

 

I didn't see in your original post but I may have missed it. What phase of HARV testing did the data that you compared against come from? Because depending on the phase there were either some changes or very significant changes to the aircraft. In phase two alone the thrust vectoring system added an additional 2,200 lbs of weight to the HARV. Not to mention the other 1,919 lbs of equipment that was added to the aircraft (spin parachute, other equipment and wiring). That's a total of 4,119 lbs of additional weight on an aircraft with already significantly less thrust than our 18C Lot 20 EPE.

 

This isn't to entirely invalidate your testing methodology which seems sound. I just don't think the comparison that you are making is related to our aircraft at all nor is it the slam dunk that some people seem to be touting.

Edited September 2, 2024 by Stackhouse

[DummyCatz]

Thanks for your concern. The reasons why I suspect that the EM diagram is related to an EPE aircraft are:

1. The HARV validation flight were not yet started, and the paper is about all the plans of CAMS, in which the EM diagram alone was only used as an example.

2. According to previous speculations, the 7.33g limit is for an f-18 that weighs 33325 lbs, which probably have a fighter escort loadout with (2xa9, 2xa120) and 60% fuel. This coincides the figures from GAO report, that the max STR is as high as ~13 deg/s and the max level speed exceeds Mach 1.3. An f-18 with 400 engine is not able to achieve such performance.

Edited September 2, 2024 by DummyCatz

[DummyCatz]

As said the testing conditions are all speculations. BTW I forgot to mention that in my later testing of STR using the same config of GW=33325 lbs, 2 AIM-9 and 2 AIM-120, Fuel at 62%, the max STR is exactly at 13 deg/s, which matches the EM diagram.

But tell me what kind of f-18 with 400 engine with the above config can reach a max level speed of M1.3 and above as indicated in the EM diagram?

Edited September 2, 2024 by DummyCatz

[DummyCatz]

14 hours ago, Stackhouse said:

I could understand why you might conclude that the aircraft used was an 18C rather than the HARV but I'm not understanding why you are concluding that the tests utilize the 402 engines. Especially when the HARV itself uses 400 engines.

Perhaps there is a better source but in my view it would only make sense that if the purpose of these tests was to then validate with the HARV that they would compare to an 18 with the same engines. They even went so far in the digital simulator testing to compare against an aircraft with the same aerodynamics and thrust as the 18 itself. Would they suddenly abandon this in the real world testing?

 

Additionally - are the E-M diagrams coming from simulator runs in this testing because it reads almost like there have been no real-world flights and instead they were all simulator based? This is further supported by the 'Current Status' page where they state that so far the testing was all 16 and 18 simulator runs and the next phase is for HARV Validation Testing (I'm assuming this is from where you concluded that it was not a HARV generating the data).

 

To continue my arguments, I'd like to point out the following:

1. As quoted from the paper, 'the chart following this illustrates the change in turn rate with loss in bleed rate and will serve to illustrate why CAMS is important'. There was no mentioning of such a diagram was generated from simulation or flight test data.

2. The inputs to the CAMS control law are sensor information, bleed rate (or Ps) level, and override commands from the cockpit. For an existing simulator, the only software job to do is to modify the control law. There's no need to generate an EM diagram in order for the control law to function. Bleed rate (or Ps) level is calculated in real time by using V-dot (or derivative of speed, dV/dt, as I did) and displayed on HUD. It is not taken from a diagram.

Remember this is all about the modification of the existing control law. The underlying aerodynamic and thrust data can be whatever it is. That’s why you need to jump out of the whole HARV thing and start looking at the EM diagram itself, the g-limit, the max STR, the max level speed, all of them give hints.

Edited September 3, 2024 by DummyCatz

[Figaro9]

Agree.

We can check and compare the em chart to the gao envelope at several  points…

1. the eidetics chart is for a 33’325lbs f18 based on its 7.33g-limit
2. the design-limit of m1.5 shown in the eidetics chart is identical to the gao envelope for the f18 402 (33325lbs, 2a9,2a120)
3. the max sustainable 1g speed of ~ m1.35 corresponds with the gao envelope
4. The 1g sep of +10kts/s  @ ~m0.4 shown in the gao corresponds with the 200ft/s shown in eidetics at ~m0.4
5. That 1g sep at m0.95 of ~ 500ft/s in eidetics corresponds with the ~ 10 kts/s in gao
6. This eidetics chart can’t be for a 400 engine f-18 cause that engine type  at 7.33g/33325lbs has max speed of <m1.27
7. That sep in eidetics is too high for an f-18 with 400 engines. The a-model m0.9 sep is ~600ft/s at 10’000ft (not 15kft), while the epe has a max sep of 700ft/s at 10kft in fe-load out.

Just speculating once more of corse…. 

[DummyCatz]

Thanks @Figaro9 for the analysis. So we have a pretty high confidence that this EM diagram is related to our EPE aircraft.

 

BTW our DCS hornet can reach M1.45 (overperforms) at 15000ft with the same loadout (33325lbs, 2a9,2a120) and 15°C mission temperature (standard atmosphere). I think the 1g envelope is worth investigating too.

Ref page 84 of https://www.gao.gov/assets/nsiad-96-98.pdf

 

Edited September 18, 2024 by DummyCatz

[BIGNEWY]

On 9/2/2024 at 1:39 PM, Stackhouse said:

 

Isn't the HARV in the NASA paper a completely different aircraft than the 18C Lot 20 EPE that we have in the game?

 

Looking at the F-18 HARV Aircraft Specifications this jet is nothing like the one that we have in DCS.

 

For starters is uses two F404-GE-400 engines. Our F/A-18C does not use 400 engines, it is powered by 402 engines. The 402 engines provide increased thrust and increased fuel efficiency over the 400 engines used in earlier 18C lots and the NASA HARV. Namely about 1,700 lbs per engines to 3,400 lbs total thrust.

 

I didn't see in your original post but I may have missed it. What phase of HARV testing did the data that you compared against come from? Because depending on the phase there were either some changes or very significant changes to the aircraft. In phase two alone the thrust vectoring system added an additional 2,200 lbs of weight to the HARV. Not to mention the other 1,919 lbs of equipment that was added to the aircraft (spin parachute, other equipment and wiring). That's a total of 4,119 lbs of additional weight on an aircraft with already significantly less thrust than our 18C Lot 20 EPE.

 

This isn't to entirely invalidate your testing methodology which seems sound. I just don't think the comparison that you are making is related to our aircraft at all nor is it the slam dunk that some people seem to be touting.

 

Hi Stackhouse, 

The references used is based on the F-18 HARV that used the 404-400 and not the 402 which were entering service around this time. The DCS F/A-18C uses the 402 EP engines. As such, the basis for much of this discussion is not sound. We are happy though to review applicable data if/when it becomes available. 

thank you 

[DummyCatz]

Dear all,

Please actually start reading the NASA paper and refer to the discussion above. I won’t be stating again how the EM diagram was used as an example and unrelated to HARV.

Thanks.

[NineLine]

13 hours ago, DummyCatz said:

Dear all,

Please actually start reading the NASA paper and refer to the discussion above. I won’t be stating again how the EM diagram was used as an example and unrelated to HARV.

Thanks.

The thread is marked as investigating, let the team investigate. We were addressing another user about the document mentioned. We are well aware of it and BN responded. Thanks.

Edit: The team is investigating this, I will lock it for now and when/if we need more info we will share it. Thanks. 

[TheFreshPrince]

A good demonstration of the capabilites of the Hornet is this airshow video imo:

Unfortunately you cannot see the HUD clearly, but one can get an impression of the speeds from the wind noises. I have tried to somewhat reenact maneuvers like the high alpha pulls and then rolling near stall speeds, but I keep falling out of the sky. 6:30-7:30 is a good example. The DCS Hornet loses all of its speed when pulling hard but is not controllable at slow speeds.

Edited September 21, 2024 by TheFreshPrince

[Figaro9]

To further speculate an expand on the above:

8. another very interesting public nasa document shows a bstr of 12.7°/s for an f-18a (32’366lbs, 60% fuel) @ 15kft / M0.8. That bstr is lower compared to the eidetics one, although she is almost 1000lbs lighter…

9. Also, the max sep of that f-18a (32’366lbs) at m 0.94  is 590ft/s compared to the 600ft/s+ in the eidetics (33’335lbs) chart. Again, that is lower although that aircraft weighs less…

https://ntrs.nasa.gov/citations/19910008802

[DummyCatz]

Thanks again @Figaro9

For the specific test that I conducted at M0.755 / 15000 ft, the eidetics EM chart (supposedly 33325lbs, 60% fuel, 402 engine, fighter escort loadout) indicates an SEP of -800 ft/s, while the SEP of a lighter and less powered F-18A (32366lbs, 60% fuel, 400 engine) is clearly beyond -1250 ft/s. (In DCS it's -417 ft/s)

Edited September 22, 2024 by DummyCatz

[DummyCatz]

On 9/21/2024 at 3:32 PM, TheFreshPrince said:

Unfortunately you cannot see the HUD clearly, but one can get an impression of the speeds from the wind noises. I have tried to somewhat reenact maneuvers like the high alpha pulls and then rolling near stall speeds, but I keep falling out of the sky. 6:30-7:30 is a good example. The DCS Hornet loses all of its speed when pulling hard but is not controllable at slow speeds.

 

Hi, between 6:30-7:30 I saw some brief pulls but were quickly returned to normal AOA ranges before the roll begins, so maybe you were pulling too much? There was indeed an FCS issue reported by me about the high AOA rolling capability that might related to the issue you described (https://forum.dcs.world/topic/337280-missing-important-features-in-fcc-ofp-v107-lack-of-high-aoa-roll-and-yaw-performance-as-if-with-v1051), but is not quite the topic of bleed rate of this thread.

 

[Kefa]

The FLXTRM simulation mentions the G override, see pg 12:  "With a G override capability, a slightly greater sustained turn rate is possible with full power."   Note that it refers to the increase in sustained turn rate from the paddle as "slightly", and if one compares the PS=0 line at 7.5G to 8.2G (at 5,000ft), that is roughly only a .5dps increase.    After 8.2G the line is on the way down.  I would expect it similar to that for sea level.  

 

Edited September 23, 2024 by Kefa

[Figaro9]

20 hours ago, Kefa said:

The FLXTRM simulation mentions the G override, see pg 12:  "With a G override capability, a slightly greater sustained turn rate is possible with full power."   Note that it refers to the increase in sustained turn rate from the paddle as "slightly", and if one compares the PS=0 line at 7.5G to 8.2G (at 5,000ft), that is roughly only a .5dps increase.    After 8.2G the line is on the way down.  I would expect it similar to that for sea level.  

You are right of corse and sure enough, that‘s how the ed f-18 works in game.

Try that: jump in a edf18 (33325lbs, 2 aim9, 2 a120, standard atmosphere) at sea level and accelerate  in full burner to max sustainable speed (which is ~ M 1.13).

Then start a max g-pull in full ab (do not paddle, so 7.3g) and watch your air speed. Keep pulling until airspeed stays constant (which will be ~ M 1.01).

Str at M 1.01 will therefore be ~11.8°/s , 7.3g.

Now keep turning and pull the paddle. Watch the speed decrease again. You will see, str now is 16.3°/s @ M0.95 (9.4g).

You can do the test of corse while accelerating too. Accelerate to M0.86 and pull 9.4g. Your wings will ripp off soon… lol. It will work tho if you start your pull at M0.93.

Our hornet sustains 9.4g (+) between ~M0.83 and ~M0.95. The str at M0.83 is of corse much higher than at M0.95….

 

Edit: By the way, i get a str of 19.3°/s at M0.62 (7.3g) and 19.4°/s at M0.83 (9.4g).

 

Edited September 24, 2024 by Figaro9

[DummyCatz]

There have been 1g SEP (acceleration) issues that can of course be investigated together:

 

[Gierasimov]

Since long time ago we were told by Wags and through official ED social media, that DCS F/A-18C product IS NOT simulating everything in 1:1 precision on purpose. Including the performance values and flight envelope abilities. This was stated already multiple times. 

How valid is the argument about performance displayed by virtual Hornet as compared to real life Hornet? Even if we take real documents and show EDSA where DCS product is lacking, it is still on purpose, therefore we should not expect a fix. Correct ? 

I wonder though, what will come out of the FM coding change that is being made right now.

[Muchocracker]

14 hours ago, Gierasimov said:

Since long time ago we were told by Wags and through official ED social media, that DCS F/A-18C product IS NOT simulating everything in 1:1 precision on purpose. Including the performance values and flight envelope abilities. This was stated already multiple times. 

How valid is the argument about performance displayed by virtual Hornet as compared to real life Hornet? Even if we take real documents and show EDSA where DCS product is lacking, it is still on purpose, therefore we should not expect a fix. Correct ? 

I wonder though, what will come out of the FM coding change that is being made right now.

He said that about the AoA abilities years ago but here we are with the most recent FM update basically doubling the max AoA it'll yank to 60 degree's. 

[Gierasimov]

3 hours ago, Muchocracker said:

He said that about the AoA abilities years ago but here we are with the most recent FM update basically doubling the max AoA it'll yank to 60 degree's. 

It used to be at least 47 units of Alpha since day one, if it can go to 60 units now, its evidently better, even though not doubled. Still my point is, DCS Hornet is not real life 1:1 copy 

[Schmidtfire]

It obviously will never be a 1:1 copy in a desktop simulator.

However, I think we all expect the flight model to be as close as possible and not altered/dumbed down on purpose - IF there is public info available.

[Gierasimov]

20 minutes ago, Schmidtfire said:

It obviously will never be a 1:1 copy in a desktop simulator.

However, I think we all expect the flight model to be as close as possible and not altered/dumbed down on purpose - IF there is public info available.

IF and WHEN public data is available for sure it fuels community expectations, however those are not moderated by the fact that Hornet is just a bolt-on module to the core game (code) which might influence module code in such a way that certain 1:1 mapping of available data is simply impossible.

[DummyCatz]

On 10/10/2024 at 5:25 PM, Gierasimov said:

IF and WHEN public data is available for sure it fuels community expectations, however those are not moderated by the fact that Hornet is just a bolt-on module to the core game (code) which might influence module code in such a way that certain 1:1 mapping of available data is simply impossible.

It's a good question on what kind of realism do we seek.

This thread is only about whether the flight performance agrees with publicly available data, which are scarce at best and limited to a few public NASA papers and GAO reports. It surely doesn't suggest that we should seek for a real life 1:1 copy, or even we know that the remaining large portion of non-public part is accurate at all.

These scarce public data points aren't classified specs or obscure details, but rather fundamental performance (SEP) characteristics. I believe replicating these accurately should be the bare minimum for a module that prides itself on FM realism and sold for 80 bucks.

Still, the ongoing extensive FM review and rework suggests that ED are open to enhancing accuracy when presented with compelling evidence, even if perfect replication isn't the goal.