AIR file: Tables 1502 to 1507 plus EPR and EGT

[Heretic]

How do you guys tackle those tables?

I want to give a MD-8x better throttle handling, but I have no access to any real world tables (if there are any for the MD-8x or JT8D-200s).
The only thing I could do would be looking at hundreds of cockpit photos/videos and subsequently trying to decypher N1/N2 ratios, mach numbers and else, but that's going to cost a ton of time.

Is there any easier way to do this?
Nicking the data from the 737 or CRJ or 747 isn't going to work (as they are in a totally different engine league) and even other, older freeware MD-8x FDE files don't get the numbers right (at least not from what I see in videos or in pics).


Also, is the EPR modeling in MSFS atrocious or am I just dumb?
Three factors, one of which is basically the upper limit plus a scale and a gain factor.
I never ever got real world correct EPR values to display on the gauges.
E.g. in idle, the EPR gauge in an MD-8x should be in the region around 1.01 while the upper limit of the EPR range should be 2.2 (but gets progressively limited by OAT and airport alt).
In FSX, EPR in idle is 1.4 while full throttle EPR peaks around 2.1 (which is still less than the limit set in the .air).

Same with EGT. You can set upper limits, scale and gain factors, but nothing more. I just can't attain real numbers with that.

 

[Roy Holmes]

To answer the last points first, FS does a lousy job with EGT and EPR and the best thing is to create your own values rather than use the A: vars.

If you want good values for idle and max, I'd suggest buying the flight manual and look under the limitations section. You can get them for around $20 from collectors for example http://www.Flight-Manuals-on-CD.com

I think the next point is that a lot of the ways FS works is based on piston propeller engines rather then jet engines.

For instance the way the throttle works in a real piston is to apply a linear increase in thrust from a slight negative value at flight idle, due to propeller drag, up to its max at full throttle.

For a jet, there is positive thrust at flight idle which slowly increases at first with throttle movement, for instance an inch movement above idle with a jet that has 10,000 lbs static thrust might give an extra 250 lbs thrust. Further throttle movement starts to increase thrust rapidly and above about 80% throttle another inch might give an extra 2,500 lbs thrust. Nothing like linear. For a jet with 10,000 lbs static thrust, the thrust versus throttle position is something like:

Idle 750, 20% 1,000, 40% 1,750, 60% 3,000, 80% 5,000, 90% 7,000, 100% 10,000. Lets call this thrust versus throttle (TVT)

EGT and EPR should behave in a similar manner, but as you say, FS basically scales them in a linear way. So best to make your own values according to thrust.

The non-linear change in thrust with throttle movement is what 1506 should do based on CN1. However CN1 comes from CN2 in 1502, and CN2 versus throttle is in 1503 and 1504. Bit of a mess and hard to understand at first.

The FS turbine model is based on a turbofan where CN1 is different from CN2 because the compressor and turbine are not mechanically connected. I think it is easier to understand the model if you consider a turbojet where the compressor and turbine are connected by a shaft so CN2 is always the same as CN1. Makes it easier to understand how the tables play together when you first study them.

Getting late here. Will continue tomorrow.

Roy

 

[Heretic]

To answer the last points first, FS does a lousy job with EGT and EPR and the best thing is to create your own values rather than use the A: vars.

I was thinking about that. Still need to model or observe the EGT/EPR behaviour though.

If you want good values for idle and max, I'd suggest buying the flight manual and look under the limitations section. You can get them for around $20 from collectors for example http://www.Flight-Manuals-on-CD.com

The web has a flight manual for a MD-8x, but it's not extensive enough.

But the more professional MSFS-addons for this type do come with extensive manuals, but I don't know whether the tables in those were derived from the real deal or whether they are based on flight model data.

For a jet, there is positive thrust at flight idle which slowly increases at first with throttle movement, for instance an inch movement above idle with a jet that has 10,000 lbs static thrust might give an extra 250 lbs thrust. Further throttle movement starts to increase thrust rapidly and above about 80% throttle another inch might give an extra 2,500 lbs thrust. Nothing like linear. For a jet with 10,000 lbs static thrust, the thrust versus throttle position is something like:

Idle 750, 20% 1,000, 40% 1,750, 60% 3,000, 80% 5,000, 90% 7,000, 100% 10,000. Lets call this thrust versus throttle (TVT)

EGT and EPR should behave in a similar manner, but as you say, FS basically scales them in a linear way. So best to make your own values according to thrust.

The non-linear change in thrust with throttle movement is what 1506 should do based on CN1. However CN1 comes from CN2 in 1502, and CN2 versus throttle is in 1503 and 1504. Bit of a mess and hard to understand at first.

The FS turbine model is based on a turbofan where CN1 is different from CN2 because the compressor and turbine are not mechanically connected. I think it is easier to understand the model if you consider a turbojet where the compressor and turbine are connected by a shaft so CN2 is always the same as CN1. Makes it easier to understand how the tables play together when you first study them.

I think I see where you're going here...basically start out drawing up a curve for N2 vs throttle and then working your way from there from table to table. Sounds doable.

Too bad there's no universal UI-based tool for .air file table editing. AAM does have a graphic display, but you can't add curves in there and working with Notepad, you don't get a graphical display...

 

[Roy Holmes]

Bjoern,
If you want to know how EPR and EGT change with throttle setting, I would recommend EngineSim UI, a NASA applet that models various stock engines or you can build your own design!

One of the stock engines is the F100 and ignoring afterburner it is close to the JT8D. You can set any throttle setting, any speed and any altitude, the outputs include gross thrust ram drag and net thrust as well as EPR and T8 which is EGT. You have to write down the values but it a great tool.

To get back to what I was writing last night. My idea of making CN1 and CN2 the same is just to get a more clear understanding of the FS jet model. Once the model is understood you do have to give them different values to overcome the big failing in the model which was the subject of a thread one or two down. The FS model reduces thrust at altitude according to the ratio of pressure at altitude compared to pressure at sea level, pressure ratio though for convenience they use it as inverse pressure ratio. In fact thrust reduces with altitude according to density ratio. To get a jet in FS to have thrust decrease with altitude according to density ratio you have to not only make CN1 and CN2 different at altitude but make them a lot higher than is realistic. I'll get back to that point later

For now we keep them the same.

In 1503 we need to have our idle CN2 at zero throttle position on the ground,
So in row 2 column 2 lets enter 30. In the last row, usually row 9 enter 100. Engine revs vary linearly with throttle setting, so the first column should be linear between 0 and 1 for seven rows (2 to 9) so the numbers should be 0, 0.1429, 0.2857, etc up to 1. In the second column we need 30 to 100 for seven rows to give a straight line on a graph, that is 30,40,50 etc up to 100.

For now make the third column the same.

For now make 1504 the same as 1503.

Since we have CN1 same as CN2, the entries in 1502 can be similar to those in 1503 except there are four more rows, for some reason.


So far everything has been linear because we have just considered engine speed. In 1506 we have to get a thrust curve against engine speed something like Idle 750, 20% 1,000, 40% 1,750, 60% 3,000, 80% 5,000, 90% 7,000, 100% 10,000 as I described in my last post. 1506 is scalars versus static thrust, and for simplicity we can put the max value in column 2 to 1.0 and the row 3 value in column 2 to something like 0.075.

In the first column the row 3 value should be 30, our idle CN1. To give a better sensitivity near 100 CN1 the values above it should be at say 2.5 intervals then increase. Reading from the bottom we would have 100 97.5 95 92.5 90 85 80 etc.

Based on my suggested thrust curve, which was actually thrust versus throttle position we need to relate these thrust scalars to the CN1 for throttle position.

20% throttle gives about 42% CN1 and 0.1 thrust scalar, 40% throttle gives about 60% CN1 and 0.175 thrust scalar, 60% throttle gives about 75% CN1 and 0.3 thrust scalar, 80% throttle gives about 85% CN1 and 0.5 thrust scalar, 90% throttle gives about 93% CN1 and 0.7 thrust scalar.

Using these as basic values you can make a chart with the actual CN1 numbers for 1506 like I suggested above, then juggle the thrust scalars to make a curve that fits the basic values. There is no need to have different thrust scalars in the second and third columns(0 Mach and 0.9 Mach) because FS calculates gross thrust based on Mach number. Same comment applies to the third column in 1503 and 1504.

I think that the above explanation should show the interaction between tables 1502, 1503, 1504 and 1506.

Before going on lets briefly consider 1505 and 1507 for completeness.

1505 is OK as it is. 1507 can also be left alone. It affects ram drag and the equations that use it also include engine inlet area. Net thrust = gross thrust minus ram drag so you have to have sensible values for inlet area in the [TurbineEngineData] section of the config file if you are seeking accurate net thrust.

Now for the difficult part. What we have so far is a very simple approach to the tables which, unfortunately, will not give accurate results. Even with a turbojet like the Rolls Royce Avon in the Hawker Hunter, I had to split CN1 and CN2 to get adequate high altitude performance. I know the Hunter performance well having flown it for 10 years and I have the manuals for performance numbers. Set me thinking what was wrong with FS and altitude.

Rather than go into the altitude issues here, I'll stop and refer you to my earlier post on the subject.

Hope this was a help and welcome your comments.

FYI, I use FDWB for my work on the turbine tables, have a slightly tailored version.

Regards
Roy

 

[Heretic]

Bjoern,
If you want to know how EPR and EGT change with throttle setting, I would recommend EngineSim UI, a NASA applet that models various stock engines or you can build your own design!

Aaaaand there goes the name for my poor man's turbine engine AccuSim, lol.

But thanks, I'll check it out.
Idle N2 vs mach vs altitude (1503, 1504) is giving me a major headache. If only I had more accurate data than those obtained by looking at engine gauges on A.net photos and writing the values down.

One of the stock engines is the F100 and ignoring afterburner it is close to the JT8D. You can set any throttle setting, any speed and any altitude, the outputs include gross thrust ram drag and net thrust as well as EPR and T8 which is EGT. You have to write down the values but it a great tool.

Hey, you can even "design" your engine!

In fact thrust reduces with altitude according to density ratio.

So to maintain thrust, higher RPMs are needed...

To get a jet in FS to have thrust decrease with altitude according to density ratio you have to not only make CN1 and CN2 different at altitude but make them a lot higher than is realistic.

Higher than realistic?
That blows as it implies tons of unnecessarily wasted hours in the actual sim looking at engine gauges.

Since we have CN1 same as CN2, the entries in 1502 can be similar to those in 1503 except there are four more rows, for some reason.

Table 1502 uses N1 (or N2, AAM isn't clear enough on that) on the X axis in 10% increments from 0 to 110%, hence the 11 rows.

I think you could extend 1503 and 1504 to more than 7 rows, but I'm not sure.


So far everything has been linear because we have just considered engine speed. In 1506 we have to get a thrust curve against engine speed something like Idle 750, 20% 1,000, 40% 1,750, 60% 3,000, 80% 5,000, 90% 7,000, 100% 10,000 as I described in my last post. 1506 is scalars versus static thrust, and for simplicity we can put the max value in column 2 to 1.0 and the row 3 value in column 2 to something like 0.075.

In the first column the row 3 value should be 30, our idle CN1. To give a better sensitivity near 100 CN1 the values above it should be at say 2.5 intervals then increase. Reading from the bottom we would have 100 97.5 95 92.5 90 85 80 etc.

Based on my suggested thrust curve, which was actually thrust versus throttle position we need to relate these thrust scalars to the CN1 for throttle position.

20% throttle gives about 42% CN1 and 0.1 thrust scalar, 40% throttle gives about 60% CN1 and 0.175 thrust scalar, 60% throttle gives about 75% CN1 and 0.3 thrust scalar, 80% throttle gives about 85% CN1 and 0.5 thrust scalar, 90% throttle gives about 93% CN1 and 0.7 thrust scalar.

Using these as basic values you can make a chart with the actual CN1 numbers for 1506 like I suggested above, then juggle the thrust scalars to make a curve that fits the basic values. There is no need to have different thrust scalars in the second and third columns(0 Mach and 0.9 Mach) because FS calculates gross thrust based on Mach number. Same comment applies to the third column in 1503 and 1504.

I think that the above explanation should show the interaction between tables 1502, 1503, 1504 and 1506.

Before going on lets briefly consider 1505 and 1507 for completeness.

1505 is OK as it is.

1505 is responsible for engine spool-up or so I've read and can apparently alleviate those ugly peaks during engine startup.

FYI, I use FDWB for my work on the turbine tables, have a slightly tailored version.

Didn't look into the workbook yet, but as I understand it's the (outdated) base for AirWrench.



Anyways, many thanks for the extensive write-up, Roy!


For now, I'll be tinkering with EngineSim.

 

[Heretic]

I need to thank you again for the tip with Engine Sim!

I've modeled a JT8D-219 and blast me, it displays the right numbers throughout the envelope! Now that's an awesome base to start with!

 

[Roy Holmes]

Bjoern,
Glad you like EngineSim!

Just to prove the point about how thrust varies with altitude, in EngineSim, set Mach =0, Throttle = 100%, altitude =0, note net thrust. Then change altitude to 40,000 ft, note net thrust and work out the ratio of thrust at 40,000 feet compared to sea level. Bet it is about 0.25. Density ratio at 40,000 is 0.247, pressure ratio is 0.186. Since FS works on pressure ratio the thrust at 40,000 is 0.75 times what it should be.

Few other points from your previous response.

1. Table size. The jet sample data in the ESP air files section under START OF ENGINE DATA gives max values for rows and columns in what they call tokens
Using the numbers we are used to:
1503 and 1504 can have max 9 rows and 11 columns, but they do not say what the other 8 columns would be about!
1502 can have max 13 rows and 11 columns
1506 can have max 21 rows and 11 columns, same for 1507.

2. 1505 is described as "FuelFlow constant where the constant is a normalized ratio of actual fuel flow parameter to static thrust". The input is Normalized Fuelflow constant, the output is CN2.

I do not think it has anything to do with spool up time since that is covered in the config file and increasing the config value does speed up the response of the engine to throttle movement.

However, fuel flow is directly proportional to thrust. If you normalize the fuel flow values to that for 100% CN2 and chart them with CN2 as the horizontal axis and fuel flow as the vertical axis, you get a nice chart of fuel flow (and hence thrust) against CN2. Gives another source of data for the shape of the curve in 1506.

3. Corrected N1 and N2 (CN1 and CN2) So far I have not been able to work out what the correction is. I have noted that N1 and CN1 and N2 and CN2 are different according to Mach and altitude but have not collected enough data to see the relationship. You can play with 1503 and 1504 to get N1 stay close to 100% as altitude is increased, so if your RPM display is N1 it looks realistic, but that does not solve the drop off of thrust with altitude issue.

4. FDWB was Jerry Beckwith's original method for people to insert aircraft data and come up with an air file and config file. However, it uses Excel which not everyone has and it is pretty complicated.

AirWrench is a standalone program which is easy to use and anyone can create air and config files with it. AirWrench also uses an iterative method to come up with performance data whereas FDWB is a snap shot. It also has a really neat thrust available versus thrust required presentation for different settings. This is accurate for how FS works it out, but in my opinion is incorrect because FS is incorrect.

Jerry no longer supports FDWB and the available version is quite old. We have discussed the differences between the two concerning which is better for someone who understands aerodynamics and engine performance and both agree that FDWB is better because it makes no assumptions about the type of aircraft concerned. The range of available inputs in FDWB is huge compared to AirWrench which is why I prefer it.

I do use AirWrench as a sanity check because it presents results in a much more understandable way. The available version of FDWB has write protection on most of its worksheets and the cells do not show formulas. My version shows everything and can be edited everywhere, so I can see where results come from and the equations that created them. For me, the ability to adjust the turbine tables and see what happens to thrust etc is really useful. Same applies to AOA tables, drag tables etc. Neither program supports all of the FSXA new tables. However, if you use Jerry's AirUpdate program you can see what is in existing tables, copy them to Excel, create charts, adjust what needs to be adjusted and update every section of the air file. The other tools like AAM show what is in the air file but not what happens when you change values.

Regards
Roy

 

[sergio]

.... AAM does have a graphic display, but you can't add curves in there and working with Notepad, you don't get a graphical display...

Hallo Bjoern,

to add a curve you can export the air file to a text file with Jerry Beckwith's AIRupdate utility, add a colum in the table (to do this I use Excel) and then compile the txt file back to an air file.

Grüße aus der Ostsee.

 

[Heretic]

Glad you like EngineSim!

It's an awesome tool indeed (even more so, once you work out the quirks of the save/load system (every saved file needs an .out as suffix)).

Just to prove the point about how thrust varies with altitude, in EngineSim, set Mach =0, Throttle = 100%, altitude =0, note net thrust. Then change altitude to 40,000 ft, note net thrust and work out the ratio of thrust at 40,000 feet compared to sea level. Bet it is about 0.25.

....and you've won that bet. Haha!

21700 at 0 mach/0 ft/0 °C, ~5700 at 0.84 mach/40000 ft/-56 °C

Density ratio at 40,000 is 0.247, pressure ratio is 0.186. Since FS works on pressure ratio the thrust at 40,000 is 0.75 times what it should be.

Wait, so FS makes thrust lower than it actually is?

1503 and 1504 can have max 9 rows and 11 columns, but they do not say what the other 8 columns would be about!

In-between altitude tables, like for every 5000 ft?

2. 1505 is described as "FuelFlow constant where the constant is a normalized ratio of actual fuel flow parameter to static thrust". The input is Normalized Fuelflow constant, the output is CN2.

You always use the static thrust value at 0 KIAS and sea level to calculate the factor, right?

I've used (approximate) fuel flow gauge readings from pictures to determine the constant.

I do not think it has anything to do with spool up time since that is covered in the config file and increasing the config value does speed up the response of the engine to throttle movement.

Is that the only purpose of fuel_flow_gain?

Also, ThrustSpecificFuelConsumption in the aircraft.cfg is just an override value and isn't actually necessary if you set the tables correctly, right?

However, fuel flow is directly proportional to thrust.

But not linear, is it?

3. Corrected N1 and N2 (CN1 and CN2) So far I have not been able to work out what the correction is. I have noted that N1 and CN1 and N2 and CN2 are different according to Mach and altitude but have not collected enough data to see the relationship. You can play with 1503 and 1504 to get N1 stay close to 100% as altitude is increased, so if your RPM display is N1 it looks realistic, but that does not solve the drop off of thrust with altitude issue.

AAM isn't really helpful on that matter.

"Corrected N1 vs. Corrected N2 and Mach No.
Corrected N1/N2 (vary with temp) relative to Gauge N1/N2"

Corrective factors in a N1 (vertical) vs. N2(horizontal) curve to determine the relation between both?

4. FDWB was Jerry Beckwith's original method for people to insert aircraft data and come up with an air file and config file. However, it uses Excel which not everyone has and it is pretty complicated.

Yeah, I've had a glimpse at it and was overwhelmed at the boxes. Might give it another try once I have to do the whole deal (aerodynamics AND engines).

AirWrench is a standalone program which is easy to use and anyone can create air and config files with it. AirWrench also uses an iterative method to come up with performance data whereas FDWB is a snap shot. It also has a really neat thrust available versus thrust required presentation for different settings. This is accurate for how FS works it out, but in my opinion is incorrect because FS is incorrect.

Also, it's payware.

I find I'm better off with the manual method since you can actually learn something about aircraft performance.

I'm "just" a standard mechanical engineer, but I want to do my follow-on studies in the aeronautical engineering field, so anything I can learn about aircraft and engine performance is well appreciated.

I do use AirWrench as a sanity check because it presents results in a much more understandable way. The available version of FDWB has write protection on most of its worksheets and the cells do not show formulas. My version shows everything and can be edited everywhere, so I can see where results come from and the equations that created them.

Could you attach a copy if it's not asked too much?

to add a curve you can export the air file to a text file with Jerry Beckwith's AIRupdate utility, add a colum in the table (to do this I use Excel) and then compile the txt file back to an air file.

I'm already doing that.

Still, if AAM had that capability, it would be a few steps less in the workflow. At least AAM can reload the .air file without closing the program, so it's just a matter of "save, dump with AirUpdate, open in Notepad, edit, save, compile, reload in AAM". No biggie, but you don't want to have unsaved edits when you have your .air file open in AAM and Notepad at the same time and press the wrong button...

 

[Roy Holmes]

Bjoern,
Thrust is dependent on Mach number among many other things. To isolate the effect of altitude try the 40,000 check at Mach = zero. Will probably lower the value slightly.

FS thrust is lower than it should be. At any altitude it is lower by the number you get when you divide air pressure ratio by density ratio. So in the 40,000 feet case FS would give you 4275 lbs thrust rather than the 5700 you got in Engine sim. At sea level it is OK because air pressure ratio and air density ratio are both 1.

You could try intermediate altitude columns in 1503 and 1504, be sure when you do that that the header has the correct number of columns declared.
With only two columns FS interpolated linearly between them according to IAP which is also inaccurate because IAP is not linear. Also you should make the IAP number be for about 40,000 rather than the stock one which is for 70,000 feet. I use adacalc to get the IAP values.

In setting up tables I do first try to make static thrust equal the published value for the engine. That value is usually what they get on a test bed and the installed thrust is slightly lower due to intake losses.

Fuel_flow_gain affects the time it takes for the engine to accelerate from idle to max and back again. I tend to use 0.05.

The relationship between thrust and fuel flow is linear. Makes sense twice as much thrust requires twice as much fuel.

I'll check with Jerry Beckwith if he has no problems with me sending the modified FDWB.

Regards

Roy

 

[Heretic]

Thrust is dependent on Mach number among many other things. To isolate the effect of altitude try the 40,000 check at Mach = zero. Will probably lower the value slightly.

I'll do once I'm done with 1506 and 1507.

FS thrust is lower than it should be. At any altitude it is lower by the number you get when you divide air pressure ratio by density ratio. So in the 40,000 feet case FS would give you 4275 lbs thrust rather than the 5700 you got in Engine sim. At sea level it is OK because air pressure ratio and air density ratio are both 1.

So you need to account for that by raising N2 or the N1-N2 constant...

Also you should make the IAP number be for about 40,000 rather than the stock one which is for 70,000 feet.

That was the first thing I've done, since 40k is a better upper limit for the MD-80.

In setting up tables I do first try to make static thrust equal the published value for the engine. That value is usually what they get on a test bed and the installed thrust is slightly lower due to intake losses.

And airfield altitude and temperature, etc...

And yes, I've done that, too.

Fuel_flow_gain affects the time it takes for the engine to accelerate from idle to max and back again. I tend to use 0.05.

Okay.

The relationship between thrust and fuel flow is linear. Makes sense twice as much thrust requires twice as much fuel.

Ah, okay. But what reference do you use to set it?

TSFC at cruise speed or on the ground standing still?

I'll check with Jerry Beckwith if he has no problems with me sending the modified FDWB.

Thanks!

 

[sergio]

Still, if AAM had that capability, it would be a few steps less in the workflow. At least AAM can reload the .air file without closing the program, so it's just a matter of "save, dump with AirUpdate, open in Notepad, edit, save, compile, reload in AAM". No biggie, but you don't want to have unsaved edits when you have your .air file open in AAM and Notepad at the same time and press the wrong button...

AFIK this feature is allready on Karl-Heinz to-do-list. Some months ago he told me that it would be necessary to change the complete memory allocation and that would need a fundamentall rework of the software. Let see if he find the time to realise it.

Ciao, Sergio

 

[Heretic]

AFIK this feature is allready on Karl-Heinz to-do-list. Some months ago he told me that it would be necessary to change the complete memory allocation and that would need a fundamentall rework of the software. Let see if he find the time to realise it.

Good to hear. I'll wait for it.

 

[Roy Holmes]

Bjoern,
So far I've just explained my theory.

Spent a lot of time past two days, have data for the tables and have flight test results that appear to prove the theory.

I'll try to make this as short as possible, make the key points. Probably best to go over how I developed the tables.

1502. Main aim was to have a specific idle CN1 (31.25%) and make CN1 and CN2 the same throughout.

This is update-1502.txt

Record: 1502 Turbine CN1 vs CN2 and Mach No.
columns: 3 rows: 13
0.000000 0.000000 0.950000
0.000000 0.000000 0.000000
31.250000 31.250000 31.250000
41.125000 41.125000 41.125000
51.000000 51.000000 51.00000
60.875000 60.875000 60.87500
70.750000 70.750000 70.75000
80.625000 80.625000 80.62500
90.500000 90.500000 90.50000
100.3750 100.3750 100.3750
110.2500 110.2500 110.2500
120.1250 120.1250 120.1250
130.0000 130.0000 130.0000

Note the Mach values are 0 and 0.95 because my flight test was planned for a Hunter F6 with no external stores and I expected max Mach to be around 0.96. If I exceeded 0.95, the values used by FS would stay at the 0.95 levels. I went up to 130% as a first guess at how high I might have to go for engine percent.

Having got CN1 and CN2 tied together I then looked at the shape required for 1506, thrust versus CN1. I earlier gave some numbers for thrust versus throttle position and mentioned that throttle position and engine RPM are linear. I have a chart from a book that shows thrust versus throttle position, so I scanned it, worked out the scales per pixel in both axes and came up with a table for thrust versus CN1 given that zero throttle was 31.25% and max throttle was 100% CN1. Because of the way thrust increases very rapidly as 100% is reached, I added points from 90% to 100% to get a good idea of the shape and then extrapolated out to 120% in a straight line because I had no data other than the final slope just before 100%

Here is my update-1506.txt file:

Record: 1506 Turbine Thrust Scalars vs CN1 and Mach No.

columns: 3 rows: 19
0.000000 0.000000 0.950000
0.000000 0.000000 0.000000
31.250 0.07 0.07
41.0714 0.09 0.09
50.8929 0.12 0.12
60.7143 0.18 0.18
70.5357 0.27 0.27
80.3571 0.42 0.42
90.1786 0.63 0.63
92.0 0.70 0.70
94.0 0.75 0.75
96.0 0.82 0.82
98.0 0.89 0.89
100.0 1.00 1.00
104.0 1.20 1.20
106.0 1.30 1.30
110.0 1.50 1.50
115.0 1.75 1.75
120.0 2.00 2.00

The values of interest are 104% where the thrust scaler is 1.2 and 106% where it is 1.3.

The reason this is of interest is that I plotted air density ratio and air pressure ratio from sea level to 40,000 feet. FS computes thrust as far as CN values are concerned according to air pressure ratio whereas thrust actually varies as air density ratio. The difference at 25,000 feet is 1.2 and at 40,000 feet it is 1.3, so you need to increase CN1 at 25,000 feet by a value that gives 1.2 times more thrust and at 40,000 feet by a value that gives 1.3 times more thrust.

1503 and 1504 use inverse air pressure ratio (IAP) and I decided to add an extra column so these are 4 column and 9 row tables with IAP values for sea level, 25,000 feet and 40,000 feet. Because I declared the number of rows and columns in the update file the sim accepted the change.

For the moment I ignored 1503 since it is for zero Mach and concentrated on 1504 which is for max Mach.

For my flight tests I first used a 1504 where the values at sea level, 25,000 feet and 40,000 feet were the same. This was to get a baseline for what FS does when it uses air pressure ratio for altitude thrust changes rather than air density ratio.

I flew it at close to sea level and at 40,000 feet with full throttle. To exclude throttle calibration errors I use gauge code that puts the throttle to full when it is above 96% open. At sea level I got it to 0.95 Mach, which was expected. Thrust was 13245, CN1 and CN2 were 100%, N1 and N2 were 108%.

At 40,000 feet, thrust was 2516, CN1 and CN2 were 100%, N1 and N2 were 94%. Mach was only 0.93, less than at sea level which is wrong. The ratio of thrust at 40K to sea level was .18 or very close to pressure ratio. The reduced thrust obviously comes from N1 and N2 being only 94%. Big drop in thrust from 100% N1

I then repeated the test with a 1504 having modified values for the 25,000 feet column and the 40,000 column.

update-1504.txt.

Record: 1504 Turbine HiMach CN2 vs Throttle and Pressure Ratio (Modified)
columns: 4 rows: 9
0.950000 1.000000 2.691 5.383
0.000000 31.250000 31.250000 31.250000
0.142857 41.0714 41.0714 41.0714
0.285714 50.8929 50.8929 50.8929
0.428571 60.71430 60.71430 60.71430
0.571429 70.53570 70.53570 70.53570
0.714286 80.3571 80.3571 80.3571
0.857143 90.1786 90.1786 90.1786
1.000 100.00 104.00 106.0

The only change I made at this point is in the last row where I used 104 CN2 at 25,000 feet (IAP 2.691) and 106 CN2 at 40,000 feet (IAP 5.383) to get the 1.2 and 1.3 increase in thrust needed at those altitudes.

I do need to make minor changes for throttle settings less than full, but that is fine tuning.

This flight test gave the same results at sea level. No surprises there since there were no changes that affected sea level flight.

At 40,000 feet, thrust was 3377, CN1 and CN2 were 105%, N1 and N2 were 99.1%. Mach was 0.96 which was correct. The ratio of thrust at 40K to sea level was .25 or very close to density ratio. The increased thrust obviously comes from N1 and N2 being 99% not 94% as with the unmodified 1504.

Summary. FS for jet aircraft gives reduced performance as altitude is increased because the turbine tables use pressure ratio rather than density ratio in the thrust calculations. To fix this problem the turbine tables need to be changed as described above.

Roy

 

[Heretic]

Great write-up and straight to the point, Roy!

I need to dive into my numbers again and make corrections. Thanks to your post, it's now clear what to do exactly and where to do it.


P.S:
Does the Hunter really perform the same at sea level and altitude?
I always thought performance at SL was less than at alt, or does this just apply to the supersonic range?

 

[Roy Holmes]

Bjoern,
Thanks, glad to help.

Unless the aircraft has an IAS limit that prevents it from achieving max performance at sea level, the limit on max Mach at sea level comes from the steep rise in Mach drag above the critical Mach number. The Hunter with no wing stores has a 620 IAS limit and no Mach limits. So sea level performance is drag limited. At higher altitudes drag is less because IAS is less. Highest Mach is usually achieved around 30,000 ft where there is enough thrust to get another 0.01 Mach or so. Above this the lower IAS and greater AOA starts to bring induced drag more into the equation, and of course thrust is less.

However, the thing that got me started on this topic was that every stock FS aircraft got its highest Mach at sea level and at 30,000 ft Mach was always lower, which was flat wrong.

The performance increases at altitude come from increased TAS and reduced fuel flow, mainly in the form of greatly increased range. At 25,000 feet you get twice as far per unit of fuel as at sea level and at 40,000 feet it is close to 3 times as far. Of course, you use a lot of fuel climbing to 40,000 feet!

I do not know the max IAS for the MD8x, but my guess is it is 350 or thereabouts. When I tested the stock aircraft I ignored the overspeed warning.

Had a funny incident after loading the 1506 update. Was getting 35000 lbs thrust and doing 1.2 Mach at sea level. Even throttled back I was still accelerating and thrust was still over 20,000 lbs. Checked the table with AirEd and noticed the data was swapping columns about half way through.

Found an extremely small extra decimal point in 1506 where it did not belong.

That was just after I had only copied 19 rows instead of the 21 stated in the header, which caused the Hunter to leap 500 feet into the air. Garbage in garbage out, got to be so careful. For that reason I only do one update at a time so I know what causes problems.

Looking at N1 it was interesting that it went to 108% at sea level. I did not plot it again other than at 40,000 where it was 99%. It should be easy to write gauge code that compensates for altitude change and apply a correction to give an indicated 100% all the way up.

I've started looking at the turboprop and helo turbine tables air file tables since they use density ratio for torque and their 1503 and 1504 tables are for CN1, not CN2 as used by jets. Could be something interesting there.

Supersonic performance is another story. In general (F22 being the best-known exception) they are all just subsonic in military power. What I did with F-4, F-15 and F-104 was model their performance with afterburner inhibited until it was close to mil power figures, then use the afterburner on thrust 1524 table to get the max power performance correct.

Regards
Roy

 

[Heretic]

I do not know the max IAS for the MD8x, but my guess is it is 350 or thereabouts.

It's 340 KIAS or mach .84.

I've started looking at the turboprop and helo turbine tables air file tables since they use density ratio for torque and their 1503 and 1504 tables are for CN1, not CN2 as used by jets. Could be something interesting there.

I'm interested in those findings as I still want to do a FDE for my Do-328.

 

[Roy Holmes]

More flight test results

Hi Bjoern,

Done a few more flight tests with the Hunter F6 to fine tune performance, now I'm getting accurate thrust figures at high altitude as well as sea level. The net thrust is within 1 percent of the values the FDWB gives me at max Mach for both sea level and 40,000 feet, so I'm happy that the turbine tables are pretty good.

I did find that max Mach was slightly higher (about 0.02 Mach) than the specification at both altitudes and I reduced it by slight increases in table 154a in the 0.9 to 1.0 mach range. If you use 154a, (FSXA only) FS ignores Table 430 CD vs Mach. 154a gives the ability to reduce the interval between Mach values in the table, so you can focus closely on the area between 0.80 and 1.0 where drag rises quickly and in a non-linear way. In 430 you are stuck with Mach values at 0.2 intervals out to 3.2 Mach, so most of the table is wasted since there is little change in CD vs Mach above Mach 1. With 430 FS interpolates linearly between 0.8 and 1.0 which is an extremely non-linear area for drag rise and is pretty crude for fine tuning of performance. FDWB does not address 154a but you can edit the air file through AirUpdate and there are ways of fudging the FDWB Drag results through its 430 tables and 1101 data overrides.

Next thing I looked at was how to cosmetically cover up the fact that N1 is too high at sea level and a bit too low at 40,000 feet as a result of correcting the thrust vs altitude issue with FS. What I mean by cosmetics is that I would like to have a readout for "Cosmetic N1" that basically stays at 100% with full throttle at all times. Or since max RPM is 8000 in the Hunter, I would like my RPM gauge to be close to 8000 at all speeds and altitudes with full throttle. Just makes the simulation more realistic if the gauges say what they did in the real aircraft. So it is a cosmetic cover up.

Since the thrust versus altitude issue in FS is related to pressure ratio I thought that pressure ratio might give me a way to achieve the cover up. I plotted N1 versus CN1 and altitude with full throttle. At sea level N1 was about 1.09 times CN1, at around 30,000 feet they were the same and at 40,000 feet N1 was lower by about 0.93 times CN1

You can get ambient pressure ratio from (A:Ambient Pressure,inHg) divided by 29.92.

I called that (LRESSRATIO,number) I wanted a negative factor to reduce my "Cosmetic N1" below 30,000 feet, where it needed to be lower than N1, and a positive factor to increase "Cosmetic N1" above 30,000 feet. I subtracted LRESSRATIO from 0.2968 which is the pressure ratio at 30,000 feet, scaled it by 0.1, added 1 and multiplied the result by (A:TURB ENG N1:1,percent) and saved the result as (L:RPM,number). This resulted in a "cosmetic N1" within 1 percent of 100% whenever the throttle was at 100%compared to 9% too high at sea level and 2% too low at 40,000 feet.

To make the code easier to debug, I split it into two lines, calculating the ambient pressure ratio first.
<Value>(A:AMBIENT PRESSURE,inHg) 29.92 / (&gt;L:PRESSRATIO,number) /Value>

Then I completed the equation:
<Value>0.2968 (L:PRESSRATIO,number) - 0.1 * 1 + (A:turb eng n1:1,percent) * (&gt;L:RPM,number)</Value>

For a "cosmetic N1" percent readout that's all that is needed, for a "Cosmetic RPM" readout simply multiply by max RPM divided by 100.

This worked pretty well with the tables I posted earlier and I believe the principle would hold good for other tables with different CN1 and CN2 values.

Regards
Roy

 

[Heretic]

(FSXA only)

I "only" have SP2, so this isn't an option for me.


Also, thanks for the hint about the N1/RPM gauge. I'm not sure if I have to incorporte it, but I keep it in mind (or the post bookmarked).


Btw: What Hunter model do you fly? Alphasim's/Virtavia's or Dave Garwood's?

 

[Roy Holmes]

You only need 154a if you are trying to get performance established above crticial Mach number. It the MD8x has 0.84 max Mach you do not need it, but you do need the altitude performance increase.

I made the Hunter model.
I have models of everything I have flown (12), everything I would have liked to have flown (6) and three helicopters that are fun to fly in the sim. I have one payware package which I hardly use because, although the external models are excellent, their cockpits and gauges are awful and only one has decent flying characteristics. Within the airplane types I have variants, for example one of my F-4Ms has no external stores so I can fly it through the full envelope and the other has 3 external tanks so I can do long flights. My Hunters are the F6 which is clean, an FR10 that has a different cockpit layout and four drop tanks, and a two seater T7. They all have VC cockpit only.

Tried to get them all in one line as AI, couple missing and of course, no helos except the one I took the screen shot from.

Regards
Roy