FSXA Turbine engine thrust decrease with increase in altitude in FSX

[Roy Holmes]

If you look at any chart of maximum Mach number versus altitude for aircraft, you will find that maximum Mach happens around 30,000 feet altitude when in military (non-afterburning) power. If you flight test any of the stock FSX aircraft, you will find that maximum Mach happens at sea level. With FS maximum Mach decreases all the way up. With real aircraft maximum Mach increases up to around 30,000 feet and then decreases. This can only mean that there is something wrong with the way FS calculates turbine thrust as altitude is increased.

Turbine thrust is a function of several parameters, but if you calculate for zero airspeed it comes down to how static thrust at sea level decreases with increase in altitude. Since a jet turbine is air breathing its thrust depends on the density of the air.

In FSX turbine static thrust decrease with altitude increase is directly proportional to the ratio of ambient air pressure to that at sea level. At 40,000 feet altitude in FSX engine static thrust is about 0.186 that of thrust at sea level. In fact turbine static thrust at 40,000 feet should be about 0.247 that of sea level, which is the ratio of air density at 40,000 feet compared to sea level. This assumes a basic turbojet engine, with a turbofan engine that number would be slightly lower, around .23, but way above .186.

FSX uses ambient air pressure ratio when it should use ambient air density ratio.

So, to get correct performance at high altitudes you need a way to increase the turbine engine thrust, while not making it higher at sea level where FSX calculations are correct.

The Mudpond website has an excellent paper entitled “How the jet engine tables work” and I would suggest having that paper available while reading this article.

The turbine tables are records 1502 through 1507 in the air file or the equivalent engine data tokens in the jet sample files Microsoft released in the ESP SDK. Although the static thrust decrease is directly proportional to atmospheric pressure ratio, the tables use the inverse of that ratio so that values are in increasing order. This is abbreviated to IAP.

The key to correcting thrust at altitude is to get CN1 to be greater at high altitude because thrust comes from CN1 in record 1506. CN1 is driven by CN2 in 1502, so the way to get it greater is to make CN2 greater at high altitude which means playing with 1503 and 1504.

1503 has 3 columns and 9 rows. The top row gives Mach =0, IAP=1 (sea level) and IAP= another number for high altitude. Typically the stock aircraft use 22.57 which equates to 69758 feet altitude! I prefer to use 5.4 which is 40,000 feet. If you actually go over 40,000 feet the number used stays at 5.4, so that is no problem. If I go over 40,000 I’ve probably got afterburners going and that is a whole other story as far as thrust is concerned. The next 8 rows deal with throttle position, CN2 at sea level and CN2 at 40,000 feet (or whatever IAP you use). To get more thrust at altitude the CN2 numbers in the right column should be higher than the numbers in the middle column.
1504 is the same as 1503 except that it is for high Mach. The stock aircraft use 0.9 as high Mach including the FA-18. I prefer to use 1.0 as high Mach. I use the same numbers in 1504 as in 1503 since I’m only concerned with static thrust corrections and 1506 handles the increase in thrust as Mach increases.

Here is the 1504 table from the Jet Sample:
0.90, 1.00 , 22.57 ; Pressure Ratio (Psl/P)
0.00, 62.00, 82.50
0.42, 77.00, 95.40
0.51, 80.50, 98.50
0.61, 84.80, 101.10
0.70, 89.10, 103.10
0.79, 92.20, 105.70
0.93, 96.00, 108.06
1.00, 98.20, 108.06

Here is the 1504 table is use in the F-15E
1 1 5.4
0 65 68
0.6 70 73
0.8 86 88
0.9 94 103
0.93 98 107
0.97 103 120
0.99 105 125
1 107 130

The main difference is the right column where I take CN2 up to 130%. Yes I know that is stupidly high, but it is the only way I know to fix the stupid error in FS. If you use 130 in 1503 and 1504, you have to also use 130 as the maximum value in the other tables in this block. If you are horrified by seeing 130% CN2, scale the readout to a more comfortable number.

What I’m trying to do is increase thrust by the ratio of the difference between density ratio and atmospheric pressure ratio for the high altitude case. At 40,000 feet the aim is to increase thrust by 0.247/0.186 or 33%. That does not mean increasing CN2 by 33% because thrust does not increase linearly with increase in CN2 or CN1. The effects of the other tables have to be taken into consideration and it takes a lot of flight tests to get it about right.

The procedure is to first get sea level max Mach to match the aircraft Manual and all flight testing should be done with afterburner inhibited. The main variables here are Cd0 and the middle column in 1504. If you use 154A to define Cd0 it is a lot more accurate in the 0.9 to 1.1 Mach area than using 1101 and 430. If you do use 154A FS ignores the Cd0 in 1101 and table 430. AirWrench does not use 154A so it will not give you the same performance as you will get in the sim. For the F-15E I was looking to get about 0.9 Mach at sea level and achieved it

Once you have something like real performance at sea level, repeat the tests at 30,000 feet. You should get a higher Mach than at sea level. I was looking to get about 0.98 Mach with the F-15E which I also achieved.

FSX has variables that can be used to create a flight test gauge so you can reach the test points, pause and write it all down. I have used this technique in turbojets like the Vampire, Meteor, Hunter and Jet Provost as well as turbofans like the A-7, F-4 and F-15. Each one required different tuning of 1503 and 1504, but they all ended up performing close to their Manual values. The same technique should work with turbine helicopters and would affect their torque drop-off as altitude is increased. Not tried that yet but it is a future project.

Hope this was helpful and welcome comments
Roy

 

[jetwelder]

very usefull

thanks for providing this....not being an aeronautical engineer..this sort of explanation, helps the fog to lift.....jim

 

[Flusirainer]

Here, this Ploblem is treated comprehensively.
However, in German.


http://www.aerodynamika.com/cgi-bin/yabb2/YaBB.pl?num=1287497583

 

[Geoff_D]

Here, this Ploblem is treated comprehensively.
However, in German.


http://www.aerodynamika.com/cgi-bin/yabb2/YaBB.pl?num=1287497583

NOT PERFECT, but it is just about understandable, with an Online Language Translater !!!

German to English translation

Well then, I lay going views.

First, you need tools to process the Air file, if not yet available.

Multimedia File Viewing and Clickable Links are available for Registered Members only! You need to or, the latter works well in both Flusis

I would also recommend nor the analysis tool Multimedia File Viewing and Clickable Links are available for Registered Members only! You need to or, note here the need FSUIPC version


If everything is on board, then the Air file first time now Aired öffnen.Hier regellbaren the inlet and the outlet regellbaren aktivieren.Das are sections 1521 and 1523

Multimedia File Viewing and Clickable Links are available for Registered Members only! You need to or

The 1 is active in parentheses is 0 disabled.

By Akrivierung these two parameters can now be Metzer the N1/N2 speed depending on altitude and speed changes much easier editieren.Der Flusi namely the rotational speeds depending on the height and speed and the activation may be the much better beeinflussen.Die speeds behave then just as you would edit.

Next we are now working with the Airdatei AAM.Warum?
AAM is doing very well when it is in the Air tables to edit file, WHILE Aired better if you edited pure numerical values ​​or various functions, and indeed with AAM ausschaltet.Das goes well, but has since AAM so its quirks, especially when only numerical values ​​editiert.Mitunter the change is not saved, or is there even an error message.

So work with AAM, we now those tables that are responsible for the speeds as a function of height and speed.
There are still a specialty.
The instruments show the N1 and N2 to speed, WHILE in the Air file CN1 and CN2 are edited, and these sometimes differ significantly from one another.


First, the CN2 speeds as a function of height and Speed.Das the tables 1503 and 1504.Die depending on the speed comes from the fact that one kann.Oben assign two tables Make your own value in the box or Low Mach High Mach

Multimedia File Viewing and Clickable Links are available for Registered Members only! You need to or

edited Specifically, in these two tables, the CN2 speed depending on the thrust lever position and altitude.
The amount is determined by the IAP Low Index.

Here is a chart on which I assigned to each index, the actual level in ft habe.Der IAP INdex behavior in question is not proportional to height.
Multimedia File Viewing and Clickable Links are available for Registered Members only! You need to or

IAP is an index level NULL NULL

Now to the table.
As you can see, the speeds for the index 1 less than the Index 4
The reason is this, the abnimmt.Um with increasing height, the N2 speed to the speed of CN2 that reducing the speed to compensate, I increase to greater heights, the CN2 speed.
In the N1 and CN1 speeds verhällt the same.
The thrust of the engine but is recognized over the CN1 speed.

So, I refer tables were still far to two.
1502 edited the speed as a function of CN1 CN2 speed.
This table contains two curves that everyone can own Mach value can be assigned.

Multimedia File Viewing and Clickable Links are available for Registered Members only! You need to or

In this example, both are identical.

Then there is the table in 1506
Multimedia File Viewing and Clickable Links are available for Registered Members only! You need to or

With this table you have to edit the shear force as a function of CN1 Drehzahl.zu each curve again, a value can be assigned its own Mach.

I'm assuming that in your flyer in Table 1506, only two editable curves sind.Es there are several ways to do this to ändern.Dazu we can come later.

NOT PERFECT, but it is just about understandable, with an Online Language Translater !!!

 

[Flusirainer]

I allow the images to use with.

 

[Roy Holmes]

Comprehensive description

My German is pretty rusty, so I copied all of your text into Bing Translator and got a "sort of English" version.

I can understand most of the rather poor translation and it seems you did much the same things as I did. I was very interested in the aim of getting N1 and N2 to be at 100%, I had not thought of that approach and will explore it.

Thank you for pointing me at this comprehensive explanation.

However, the bottom line is the same, thrust should be proportional to density ratio, not atmospheric pressure ratio.

Roy

 

[AuJoePilot]

Realistic At Alt Eng Performance vs. Realistic Gauge Readings and lift/drag ratios

Hi Roy, and everybody,

I work with AAM and I came up with the same conclusion as far as getting correct engine performance at altitude according to documentation.

Yes it is certainly true FS calculates something backwards, this is evident because thrust from an engine should be lower at high speed or altitude than it would stationary and at sea level.

I did my process slightly different than you described above, I started with CN2 vs. throttle and IAP ratio, but I didn't set up this graph for thrust purposes, rather, for getting as accurate as could with the N2 reading reguarding whatever info I could find on the particular engine.

I repeated this step for CN2 vs CN1, again, for accurate gauge readings according to research.

Finally, I alter the "thrust @ Mach vs CN1" table, to match Mach performance
, meaning the thrust values are higher with higher mach, which is backwards as far as I know. To avoid speeding out of control, it is important to lower thrust values at a mach value a bit higher than the aircraft's VMO, Let the optimal cruise mach be the thrust peak, this will make the engines fight the air at too high of speeds like they do in the real world, getting the engines performing right is almost as tedious as getting the aerodynamics right, it will take much trial and error to get it right, always get the aerodynamics right 1st, before you go crazy wondering why this or that about the engines, be sure to do the following tests reguarding lift to drag ratio's set in tables such as CL0 vs Mach, CDO vs. Mach, and addional tables with similar parameters, or other parameters settable in the .cfg, although it is recomended keeping as many .cfg scalars at 1.0 when working from scratch with a new FDE

This should be done before the engines are fine tuned.

Working from top to bottom starting with flaps up , gear up , all spoilers retracted.

1. Airplane has the desired cruise attitude profile at altitude and cruise speed.

2. Airplane has a correct all engines out glide ratio at specific weight(s) from the highest possible altitude for this aircraft.

3. Airplane has correct gear drag with ENGINES OUT

4. Airplane maintains desired profiles (with flaps up) throughout all tested altitudes and speeds (all min to max)

5. Airplane Has correct ground effect during landing, and will also liftoff at the desired flaps up speed (if the airplane is big and heavy, this test requires a very long runway) it is also recomended to do the tailskid/ minimum liftoff test per each Config from clean to max takeoff flaps config after you've done the flaps sets in the .cfg, when testing flaps sets, work with 1 flaps position at a time starting with the 1st position, always get the correct lift first, then add the drag, do these tests ascending and decending, calculating takeoff and approach speeds in any particular config can be done by flying level at the specified altitude and monitering the angle of attack's reaction to the changing airspeed.

You may now Fine Tune your engines!

1. Do the same tests as above with all the engines at Idle at empty and Max gross weight, it's important to make sure you don't give too much thrust at idle, make sure the plane slows down almost as much as it did during the glide ratio test with all engines off but a bit less as the engines idleing should provide a very little amount of thrust.

2. Ground idle thrust tests, at mtow and empty, be sure you don't go into a high speed taxi while reaching for a checklist, get that right.

3. Takeoff roll, check.

4. Climb Cruise profile, check.

5. Descent Profile, check.


I hope this organized process helps save some people some time, it's a long process to have to learn all by oneself.

Now I DO want ppl to give any constructive critisism to this process if they have any, here is my question reguarding gauge readings.

Since fs doesn't provide a realistic EPR variable, or any form of CN3 variable (the third stage compresser in 3 spool engines such as Rolls Royce RB211 or Trent series engines) I am in the process of programming these gauges with their own custom variables which include Mach, Altitude, Temperature, and Percent of Max RPM.

Heres the big question, how do real pilots, aircraft/jet engineers feel about the way FS models engine readouts? By that I mean, what we can't program in an airfile, or cfg file, things like how will CN1 or any other engine parameter react if the temperature changes? How do you feel about the level of accuracy FS simulates these things? since we cant change the way FS affects the scenario.

Please anyone with real Pilot or Engineering knowledge of how jet engines really work Please Respond.

 

[Roy Holmes]

That was a long post! You do make a lot of good points

Best place to start understanding how FS models jet engines is the Mudpond article I mentioned early in the thread:
http://www.mudpond.org/jet_flow_chart.pdf

Couple of points you made:

thrust values are higher with higher mach, which is backwards as far as I know.

Net thrust=Gross Thrust minus Ram drag.

Ignoring altitude effects and Table 1506 for the moment,
Gross thrust = static thrust*(1+(M^2)/5)^3.5 where M is Mach number. At M=0.5 the gross thrust is higher than static by a factor of 1.19, at M=0.9 the factor is 1.69, at M=1.3 the factor is 2.77. So gross thrust increases as Mach increases.

Ram drag also increases with Mach but by a smaller factor. It varies directly as the area of the intake (which explains why the Harrier was subsonic). For more reasonable sized intakes net thrust is around 1.3 times static thrust at M=0.9. So thrust does increase with Mach and quite dramatically above M=1.0. FS models this correctly and there is no real need to have different values in the second and third columns of Table 1506

FS reduces thrust at altitude by the atmospheric pressure ratio whereas it should reduce by the atmospheric density ratio. That was my main point in this thread. In a later thread about the turbine tables I explained a way to get around that problem.

FS does accurately model the effect of temperature changes on thrust.

I have to disagree with you that fine-tuning aerodynamic factors comes before fine-tuning the engine. You need to have the engine thrust correct first.

To consider drag for example. At maximum speed drag = thrust, so you can not adjust drag coefficients to get the maximum speed correct unless you know the thrust is correct. Another good Cd0 test is time to height test values, compared to aircraft performance, but only when the thrust is correct.

Once you have the drag correct, you can consider lift and angle of attack. Key here is to have table 404 as close as possible to what the modeled aircraft airfoil section would have. You can get the key values from A:STALL ALPHA, degrees, A:ZERO LIFT ALPHA,degrees and A:LINEAR CL ALPHA, per radian and compare what your model has compared to the airfoil section.

Once the thrust, drag and lift are correct, the rest of the performance follows and you only need a few test profiles to confirm them. Then you can fine tune the stability and other secondary effects.

It is simple to calculate what the thrust should be for any altitude and Mach, trick is getting FS to give that thrust at altitude. Get that right and the rest is straightforward.

In general, (other than the altitude issue) FS does quite a good job of modeling jet engine performance.

Please anyone with real Pilot or Engineering knowledge of how jet engines really work Please Respond.

I have both.

Roy

 

[AuJoePilot]

Thank you Roy

So then is is normal for the curves in the thrust table to go higher with airspeed? I am confused because of a document I read regaurding the engines of the concorde, it was said at 60,000 ft at Mach 2.0 the engines were only producing about 10,000lbs thrust each, and they are rated at 38,000. I probally misunderstood.

The problem I encounter with adjusting the engines 1st is that the flight profile never seems to match, when using "realistic" values for all the parameters set in the air file, something is always off, what I've found that works for me is tuning the whole thing together slowly, when I start from the ground up, I always find myself having to start over at the top on down after it just looked great all the way up. I've not seen one airplane that used real data in their graphs for lift/drag/thrust that fly's like anything close to the real plane, this became very apparent to me when I looked at the FDE's of some very good payware aircraft and found they did their graphs in a similar manner.
You mentioned Ram Drag several times, Good point I never thought of, thank you sir, is there a ram drag parameter in the airfile? Iv'e not heard of it. Thats what I was simulating by dramatically decreasing the thrust between optimal cruise Mach and +0.02 - +0.06 > VMO/"Overspeed"
In the case you can decide to do the opposite, you can simulate a RamJet at very high speeds because the increase in speed keeps feeding the thrust which in turn keeps increasing the speed, this happened to my 1st FDE project when I did this accidentally, Boeing 2707 SST, would go at 2.73, but if encountered overspeed at 3.0 would go uncontrolably to Mach 4.7 and up at 84,000ft and could circle the globe on a half tank needing 500 miles to slow down very gradually using the spoilers until the RamJet effect wore off, if you didn't get the ram effect, the plane flew like they said it would as far as takeoff/landing, speed, Fuel consumption (boeing SST project documents).

So? How would you say that the temperature should affect EPR reading? Same as the others? Up if Hot?

Any clue how my homemade graphs for EPR / N3 vs. ? should go? "EPR" and "N3" aren't truly supported by the airfile, so those are going to be custom gauges settable by developers or advanced users. They Will have to undergo a similar. If you want to School me on how real gauges read, calculate and are chained to each other in any way, please, I'm honored talking to a real engineer/pilot and gauge readings vs enviorment/flight profile have everything to do with something bigger I'm trying to create, it all starts with the engine gauges.

 

[AuJoePilot]

Actually

Roy,
come to think of it,
if you want to swap airfiles so I can better understand what you were explaining above.
I view the airfiles in AAM Aircraft Airfile Manager, so I see a graph when I view most parameters.
We may have solved this problem in the same or similar way and are just misunderstanding each other. If I'm correct you are saying that by raising CN2 value vs the throttle position vs iap ratio on the higher iap ratio end? And you are getting out of this the correct thrust, but too high of a reading from N2/N1?
If this is the case then the way I described earlier is a way of fixing the thrust vs alt problem without sacraficing the chance of getting accurate gauge readings according to source. TBL 1506.

 

[Roy Holmes]

If you have not done this already, I suggest reading through:
http://www.fsdeveloper.com/forum/showthread.php?t=47164 There is a reference there on how to get EngineSim and you can use it to see data on how EPR etc varies. It also includes examples of the turbine tables I modified to get thrust/altitude about correct.

The Concorde thrust figures at 60,000 ft compared to 38000 static thrust on the ground would be reduced by the density ratio 0.095 (3610) but increased because it is at Mach 2 by 2.77 to give 10,000.

The ram drag I mentioned is calculated by FS using table 1507. The result is subtracted from gross thrust to give net or available thrust.

If you change the turbine tables the way I outlined, you do get values of N1 that are higher than a real aircraft would see, but you do get the correct thrust at altitude. I simply apply a scale equation so what I see on the N1 gauge is realistic. This is what I have in my F-15E as an example.

<Value>(A:Turb eng1 N1, percent) abs 9 - (A:Indicated Altitude, feet) 5000 / - (&gt;L:N1,number) (L:N1,number) 0 &lt; if{ 0 (&gt;L:N1,number)</Value>

Keep at it, you are on the right track.

Roy