investigating F404 Engine Issues

[Curly]

TLDR: two parameters of the engines, EGT, and nozzle area may be off,  leading degraded transient performance.  

 

I think there may be 2 discrepancies with the F404 engines leading to marginal performance in the transient response of the engine.  By which I mean the thrust output of the engine  shortly after the throttle is moved. For now I will be discussing the response of the engine from Idle to Mil.  

 

The two issues are, 1, the EGT is too low and 2. The engine’s exit nozzle schedule may be off. 

 

 First, the Exhaust Gas Temperature (EGT) on the engines looks low.  

The charts in the NFM-000, pages III-10-15, provide the operating EGT and RPM as function of temperature. 

 

The chart indicates, At full mil, with an inlet temp of zero celsius, the engine operates at 99% RPM with an EGT of between 830 c and 860 c.   In game the EGT is ~ 815c, which is below the minimum operating condition. Low EGT is indicative of an engine that is under-performing and not putting out enough thrust.

 

 

While the difference seems marginal. Low EGT can lead to a set of interconnected discrepancies. Which results in an underperforming engine. These interdependencies would exhibit themselves most strongly in the areas of transient response. IE The power output of the engine, while and shortly after the throttle is moved. 

 

Now on to the nozzle area issue. Which, maybe a case of the tail wagging the dog. 

 

In the F404, At full mill and above, the engine’s exit nozzle size (A8) is varied with temperature sensed aft of the low pressure turbine. (T5). Below Full mil the area is scheduled with the throttle position.  

 

When the throttle is moved to the full Mill position, the nozzles closes until the EGT limiter kicks on. 

 

https://www.sto.nato.int/publications/AGARD/AGARD-CP-448/AGARD-CP-448.pdf

Page 64. 

 

 

https://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1990/79054/V002T02A035/2399309/v002t02a035-90-gt-357.pdf

 

The result is a rapid rise in thrust.  Below  is engine data taken from a takeoff of  CF-18 with 404-400’s. I’ve overlaid the nozzle position with time and the thrust with time charts.  It shows the relationship between the thrust onset and the nozzle position. Note the rapid rise in thrust once the nozzle fully closes. Max thrust is reached ~ 3 seconds after the nozzle closes.

 

The source is a Canadian paper on the development of computer model of the F404-400. 

 

https://curve.carleton.ca/c6dd200c-1bce-4711-9596-443f1cf85e70

 

In game, the only time I can get the nozzle position to 0 is when the throttle is 50% or lower.  I know our Hornet has a different model engine, a 404-402. However, the 402 has higher thermal limits and therefore the nozzle position should stay closed even longer ~ 4 seconds when the throttle is moved to mil. 

 

If the nozzle position is too large during the transient, performance suffers. 

https://www.sto.nato.int/publications/AGARD/AGARD-CP-448/AGARD-CP-448.pdf

Page 398.

 

Since the resulting loss of performance tends to only show up in the transient response of the engine. It’s easy to miss the effect a misscheduled nozzle has on engine performance.  NASA had the same problem with their dynamic engine model of the F404

 

 

 

https://www.nasa.gov/centers/dryden/pdf/88204main_H-1643.pdf

 

Worse case scenario for the in game Hornet is that the combustor and or compressor isn’t operating properly and the nozzle is scheduled to far open. Or is the nozzle just too far open resulting in lower than optimum EGT’s? 

 

I suspect the latter, that the nozzle schedule is simply off. I can think of two reasons why this may have happened. 

First, there may have been a misinterpretation of  the schedule. Most of the literature about the engine says, the nozzles are fully closed at “Intermediate Rated Power” IPR. Which seems like it would be 50% of the throttle lever range. However GE and NASA define IRP as full mill, 87° Power level angle.

https://www.nasa.gov/centers/dryden/pdf/88068main_H-1375.pdf

https://apps.dtic.mil/dtic/tr/fulltext/u2/a164562.pdf

 

Since the nozzle area only seem to close to 0% at 50% throttle, it may be that it’s simply misscheduled, because the term intermediate rated power is confusing. 

 

The nozzle area may also be off because of how the game translates joystick position into throttle angle.

 

This is related to where full mill is throttle range in game Vs IRL. In game full Mil is approximately 75% of the throttle position.  In the real jet, Ground Idle, is 18°, Flight idle is 32°, full mil power is 87° and max power AB is 130° Power Lever Angle.    So the throttle operates over 98° of PLA in flight and 112° of PLA at ground idle. 

 

At ground idle, if we want full mil the throttle lever should be moved 87° of throttle. This is ~ 77% (87/112) of the range of the real jet. In game full mil occurs at 75% throttle.So how does the game handle the scaling difference. It could be a source of mismatch, you may be taking off at below mil power. With the engine operating on a below optimal engine schedule. 

 

The same is also true while inflight. The throttle operates over 98° from 32° flight idle to 130° Max A/B, With full Mil power at 87°. Full mil in the real jet is at %88 of the throttle, while in game it’s at 75%. Depending on how the scaling is handled, with your joystick at 75% you may not actually be in AB yet. Which could make AB light times seem too long.   

 

Concluding, I think it would be a good idea to check the nozzle scheduling. If that’s correct, Then perhaps check how the joystick to throttle angle scaling is handled and see if there are any discrepancies. Then perhaps look at compressor and combstor engine performance. In order to find out if there is an issue with engine performance in transient response. 

enginerunup.trk

Edited May 14, 2021 by Curly
Grammar

[Hulkbust44]

+1
This definitely needs to be looked at.

Sent from my SM-G955U using Tapatalk

[Swift.]

Interesting wrote up. Remember though that the nozzle position in DcS is likely just ana animation. Engine thrust won't necessarily be at all related to it.

 

Additionally, hornet in game has more thrust than the real thing.

[Figaro9]

Good work curly

 

 

Is this probably the reason why sep is still slightly low at m0.9 in game (m0.9, 33325lbs, 2a9+2a120)?

 

 

10kft: dcs 647ft/s , gao 699ft/s

20kft: dcs 444ft/s, gao 512 ft/s

35kft: dcs 223 ft/s, gao 247 ft/s

 

 

 

Here is the mission if anyone wants to check the above figures. Simply do a 1g acceleration test at the desired altitude. SEP is printed in the logs as m/s and m/s2, which can be converted to ft/s or kts/s to check with gao numbers and envelope…

 

 

acceleration f18.miz

[wilbur81]

8 hours ago, Swiftwin9s said:

Additionally, hornet in game has more thrust than the real thing.

According to what source and testing? Seriously, dude... baseless, data-less, one-liner, throwaway comments like this cause so much "yeah, but..." nonsense on these forums. If you want to post a whole bunch of performance charts for the 404s and then accurate test-results from DCS with detailed and thorough experimentation at all altitudes and temperatures, etc. etc....be our guest. 

Edited May 14, 2021 by wilbur81

[BIGNEWY]

Thanks for the post, the team will take a look. 

[Skysurfer]

2 hours ago, wilbur81 said:

According to what source and testing? Seriously, dude... baseless, data-less, one-liner, throwaway comments like this cause so much "yeah, but..." nonsense on these forums. If you want to post a whole bunch of performance charts for the 404s and then accurate test-results from DCS with detailed and thorough experimentation at all altitudes and temperatures, etc. etc....be our guest. 

 

 

Based on direct SME feedback. The Hornet sustains way too well and reaches too high of a top speed on the deck. No data since the EPE perf. manuals aren't really public.

Edited May 14, 2021 by Skysurfer

[XPACT]

Hornet will steadily but surely become an alien ship if forum doesn't get locked

Jokes on the side if there is real world data to prove it... I just hope if ED actually makes FF MiG-29 that it will be true to what we know about it not being undermodeled or overmodeled.

[wilbur81]

37 minutes ago, Skysurfer said:

 

Based on direct SME feedback. The Hornet sustains way too well and reaches too high of a top speed on the deck. No data since the EPE perf. manuals aren't really public.

 

... by "direct SME feedback," you mean all your friends that flew the LOT 20 Hornet that you've had in-depth conversations with and  then sat down and compared detailed notes about how our DCS Hornet performs at sea-level clean, with/without pylons and stores, at performance-chart fuel capacities, at all temperatures, in full mil and max A/B, confirmed that the DCS Hornet top speeds are not being effected by realistic/unrealistic pylon and weapons drag, simulated fuel weights, etc. etc. right? 

 

My guess is that, between the Lot 20 SME's that ED has spent the last 3+ years consulting with, and the 404 performance data that they DO have access to, that they know what they're doing. "My SME's are better than your SME's..." 

 

I digress. I'm still holding firmly to the fact that ED is going to get us as close to real as possible.

 

A real Hornet driver like Mover who says "Man, this Hornet feels draggy" on his first few DCS flights on YouTube does not equal detailed SME analysis anymore than a real Hornet driver who says, "Man, this Hornet feels too fast" does. And you've admitted that neither you nor anyone else commenting here is looking at real documentation on the matter. 

Edited May 14, 2021 by wilbur81

[wilbur81]

Just now, Skysurfer said:

 

One of them is in contact with Wags and the main advisor for many systems. So take from it what you want. 

That's more like it. 

[bkthunder]

On 5/14/2021 at 6:29 PM, BIGNEWY said:

Thanks for the post, the team will take a look. 

Can the team take a look at the F-5 engine bug which is also related to nozzle scheduling?