Jet Engine Performance

[WarpD]

Ed,

Now you are being slow on purpose. Of course the format was defined by MS...smh. A file format written by Microsoft with the intent of being used in the GUI program that shipped with FS9 pro is still an interpreter. Once again, what it shows in a text file has nothing to do with the variables that MSFS looks up real time, nor how those variables are processed into results.

As your own posts show, it is a 21,3 table... x,y.

First... you toss out yet another insult. You really need to cease such behavior. The format existed before Microsoft's FSEdit (which is fatally flawed and should never be used). As for the rest of what you stated... I'm pretty certain only you know what you're talking about (it makes no sense).... and that doesn't make you correct.

 

[jx_]

I'm definitely not the only one who knows... There are a handful of people out there who understand how the sim works.

 

[Roy Holmes]

Jx_

I am working on a reply to your last post to me. But, I will have ground rules if we are to work together and these will include zero insults OK
Roy

 

[WarpD]

I'm definitely not the only one who knows... There are a handful of people out there who understand how the sim works.

No, really... this makes no sense:

Once again, what it shows in a text file has nothing to do with the variables that MSFS looks up real time, nor how those variables are processed into results

.

 

[jx_]

Of course it doesn't. You're convinced FS only looks up one variable in a two dimensional table query.

 

[WarpD]

Of course it doesn't. You're convinced FS only looks up one variable in a two dimensional table query.

Yeah... show where I stated that. Now you're making things up.

 

[Roy Holmes]

Jx_

FYI, the 1506 table Ed showed in his post #84 comes from
http://msdn.microsoft.com/en-us/library/cc526961.aspx

This is part of the ESP SDK and was the first time MS released details of the contents of the .air file. There is a link to jetsample.asm and you will find the equivalent of 1506 near the end of that file, it is called AIR_70_N1_AND_MACH_ON_THRUST

Roy

 

[jx_]

Yes Roy it is,

But that doesn't change the fact it is an X,Y look up in MSFS or that the asm file is used solely for GUI/compiler purposes.


But now that you pointed it out Roy, this should kill the argument. From the jetsample.asm:

;N1 vs Thrust table (max 21 rows, 11 columns)
;IN: X: Mach
;IN: Y: N1 (corrected)
;OUT: Gross Thurst (corrected) / static thrust

I believe that .asm file says:

INPUTS:
X=Mach
Y=N1 (Corrected)

Thanks for the pointless debate Tom and ED. You're own file format clearly states you are wrong.


So they are using X across, Y down


UINT32 21,3 ;ROWS,COLS <--- Y,X


So Ed was right, Y isn't Mach.... X is Mach, and Y is N1.

Thanks for posting the link Roy.

 

[Roy Holmes]

jx_


Your question was:

Figure out how to reduce your net thrust without changing your fuel flow and you'll see what I mean.

The definition of TFSC in the SDK is: "Jet thrust specific fuel consumption. The ratio of fuel used in pounds per hour, to thrust in pounds. Applies at all speeds."

Note it says thrust, not gross thrust or net thrust. However, since it says it applies at all speeds it must relate to net thrust, not gross thrust. This is correct. TSFC is a function of

net thrust.

To verify this I set TSFC to 1.0 for simplicty, did a test flight and the thrust and fuel flow numbers were the same. So FSX does use Net Thrust/fuel used per hour for TSFC.

Here are my going in assumptions:

A. The only tools available to me are the turbine tables and simconnect.

B. I can not change the equations that use these tables.

D. TSFC is a static variable and I can not change that.

Starting with the turbine tables.

I am going to assume a fixed full throttle setting so I can forget about 1503 and 1504.1502 is no help either because it just dictates CN1 based on CN2 and Mach

1506 scales gross thrust based on CN1 and Mach. A decrease in gross thrust would give a decrease in net thrust, but fuel flow would also decrease because of the fixed TSFC value

Since net thrust is gross thrust minus intake momentum/ram drag. I could look at manipulating values in 1507.

If for instance, I increased the effect of table 1507 ram drag would increase, net thrust would decrease but so would fuel flow, because of the TSFC setting.

I have not looked at doing anything with 1505 because I'm not in a start-up situation and the throttle is fixed.

So that leaves Simconnect.

The A:Var TURB ENG CORRECTED FUEL FLOW is settable via Simconnect. It is calculated by dividing actual fuel flow by air pressure ratio times the square root of air temperature ratio. I'm very much a beginner as far as Simconnect is concerned so right now I do not know how to do that, but I do know who to ask.


As an aside,in many ways the use of "corrected" values for N1 ie CN1, N2, gross thrust etc is strange. Corrected values are used by engineers so as to be able to present a set of

computations on a single chart. I assume that this method lends itself to the use of look up tables and maybe that is why FS uses corrected values. All the stock aircraft output N1 for

display because that is what is used in the real world.

Roy

 

[jx_]

Beautiful Roy!

Now you are thinking with an open mind. I didn't even consider simconnect. Let me know where that leads you.


A quick note on your comments:

Since net thrust is gross thrust minus intake momentum/ram drag. I could look at manipulating values in 1507.

If for instance, I increased the effect of table 1507 ram drag would increase, net thrust would decrease but so would fuel flow, because of the TSFC setting.

Your first sentence is 100% correct. Your second sentence, also correct, is the 1507 bug I speak of. Fuel flow should not be tied to net thrust, it should be tied to gross thrust or mass airflow...if the sim even calculates this, which I don't believe it does (can't think of anyway to confirm). As a result, FS jets don't lose efficiency.

As an aside,in many ways the use of "corrected" values for N1 ie CN1, N2, gross thrust etc is strange. Corrected values are used by engineers so as to be able to present a set of computations on a single chart. I assume that this method lends itself to the use of look up tables and maybe that is why FS uses corrected values. All the stock aircraft output N1 for display because that is what is used in the real world.

Remember how you said you increase your CN1 numbers to get the IAP--->CN2--->CN1 effect? That's basically what CN1 is. That's why they use IAP, because it gives you an indication of altitude.


From Boeing Performance Engineer's Aircraft Design Manual:

It is sometimes useful in engineering work to simplify the presentation 
of data in such a way as to make a single chart show the desired data 
instead of a number of charts. That single chart is sometimes referred to 
as a generalized chart. 

Examples that come to mind are the generalized thrust required chart and 
the generalized fuel flow tables. 

Also, in this chapter you have seen us refer to corrected N1 and corrected 
fuel flow. That simply means generalized N1 and generalized fuel flow – 
data that has been modified in such a way as to make one chart or table 
serve over a range of  variables rather than for just a single value of that 
variable.


To illustrate our meaning, here’s an example. This example will be an 
aerodynamics example rather than a propulsion example simply because 
it will be easier to understand.


Let’s suppose that we wish to have a graph of an airplane’s lift coefficient 
in cruise as a function of its Mach number and weight. That’s easy enough, 
we know the equation:


                   W
CL  = ----------------------
            1481.4 M^2  δ  S


[I]where [B]W [/B]is the weight in pounds, a variable
[B]M [/B]is the Mach number, a variable
[B]S[/B] is the wing reference area in square feet, a constant
[B]δ[/B] is the air pressure ratio, a variable[/I]


You see that the lift coefficient is a function of three variables: the weight, 
the Mach number, and the air pressure ratio δ. Because it’s a function of 
three variables, it would not be possible to make a single chart that would 
be valid for all possible conditions. For example, we could make a single chart 
that is valid for all weights and all Mach numbers, but then we would need a 
chart for each value of δ.


Here’s a way we could “generalize” the data so that a single chart could be 
valid for all altitude, let’s define something called “generalized lift 
coefficient”:

                                                            W
generalized lift coefficient = CL * δ = ----------------------
                                                    1481.4 M^2 S


You see that now the “generalized lift coefficient” CLδ is a function of only 
two things: weight and Mach number. We could therefore make a single chart 
that would be valid for all altitudes!

Thence, the correct usage of IAP to CN2, because generalized thrust is defined
as:

thrust
------
  δ (air pressure ratio)

This will produce Fn/Delta which is basically thrust normalized to sea level, or in other words, what thrust would really be at a specific altitude, if it weren't for pressure loss with height.


All MS did was copied the real world engineering process of designing aircraft...problem with that is, there aren't many real world engineers who have access to real world engineering computations and data in the Flight Sim add-on community. It also makes it far too complicated. It would have been much easier for our purposes if everything was spelled out and there to be tweaked, including thrust by altitude by N1 by Mach and even an TSFC by mach by RPM table.

(I personally try to use CN1 in gauges if at all possible for simplicity, especially if I don't have real world data to cross check.)


Final Note:

CN1 always gives fan speed to thrust generated (like EPR)
N1 gives fan speed mechanically measured RPM

 

[SkippyBing]

In case it's useful I noticed the following effect while testing an aircraft. As I want to reduce the thrust produced by the engine during certain stages of flight, to emulate thrust lost to a boundary layer control system, I've got a SimConnect gauge that directly alters CN1.
For various reasons this was set incorrectly so CN1 went to 99% of CN2 when activated (this made more sense when it was running a turbojet).
Consequently when I made the switch with the engines at idle CN1 went from ~30% to ~60% as did N1, the aircraft started rolling forwards with the brakes applied and fuel flow increased proportionally.
CN1 was the first variable I found I could adjust and affect thrust, altering CN2 had no effect on thrust and I didn't try any of the others.

 

[Roy Holmes]

That is interesting. I can see how increasing CN1 should lead to a thrust increase, but I am surprised that increasing CN2 would not also increase CN1 and therby increase thrust, but then I am ignorent about whether Simconnect in some way bypasses/overrides the normal relationship between these variables.

For what it is worth, if I want to check out engine params on the ground I choose a twin engined airplane and shut one of them down. That way the brakes usually hold even at full throttle. Much easier than trying to move one throttle with the mouse.

Any more data on those lines would be of great interest.

Roy

 

[SkippyBing]

Adjusting individual parameters via SimConnect can have some odd effects at times, as if FSX makes its calculation based on the variable value before you altered it. On the flip side I managed to emulate a two speed supercharger by lowering the carb air temperature on a piston engine so I think it may have to do with where in the calculation chain the variable is.
If I find the time I may look at writing a gauge that lets you alter individual engine parameters to note the effects. It'll at least make a change from trying to tune the throttle response to match the ODM...

 

[jx_]

Are you able to scale the fuel flow like Roy was trying to figure out? Would make things more interesting.

 

[JB3DG]

Just one interesting aside i found during some simconnect tests, I set corrected N2 RPM to 100% with the engine off and had sufficient thrust to leave the ground. This is with corrected fuel flow at 0 as well. Any ideas as to why?

 

[Roy Holmes]

Skippy,
Figure 9 of Zskowski's paper and the associated text show that the piston engine simulation is conventional, it follows each step of the Otto cycle and is easy to follow. I should imagine that an early Simconnect input gets carried through the program.

Figure 10 and text shows the jet engine simulation which is not conventiional. Commanded N1 is an early calculation in the flow chart, so perhaps it does get carried through. N2 starts as a function of fuel flow (which appears to have been calculated from N1 earlier). Then N2 is recalculated as a function of N1 and Mach giving gross thrust.

This is where I get lost, because the way the tables are set up, N2 is commanded N2 and N1 is a result of that command. Makes sense since N2 does drive N1 in a twin spool turbofan. Thrust is calculated from N1 according to the tables, yet it appears to come from N2 in the flow chart. Again, the tables use normalised corrected CN1 and CN2 not actual N1 or N2.

I need to learn more about the way normalised chrts are used by engineers, but there is a clue in the paper "the SimEngine incorporates extensive use of normalized look up tables that can be scaled by the required thrust output of the engine". So perhaps they calculate the normalised required thrust first based on commanded N1, altitude and Mach, then do a second routine to get the normalised value to represent an actual value for the flight conditions and throttle setting. More stuff to learn.

I have a lot of data now on the relationship between TURB ENG CORRECTED FF and TURB ENG FUEL FLOW PPH and it is not easy to understand. Since the former is a function of altitude and Total Air Temperature, I did my flight test runs at different altitudes, Mach and throttle settings.

My initial impression is that the CORRECTED value comes from the flow chart calculation that follows Commanded N1 and the uncorrected value comes from the TSFC config file entry and the calculated thrust. If that is the case, adjusting the CORRECTED value may just affect the calculation of the required thrust, but the actual fuel flow will stay in a constant relationship with thrust according to the TSFC constant.

However it would be very interesting to see the effects that Simconnect make by setting the TURB ENG variables and for sure it would lead to a better understanding of SimEngine.

Roy

 

[Roy Holmes]

Jonathan,
Would you do that test again and look at TURB ENG CORRECTED FF as well? I suspect it should be zero, but you never know. I think what happened is that your CN2 input resulted in a "required thrust" value like I mentioned in my previous post and so you had thrust without combustion or CN1. Very interesting input!

Thanks
Roy

 

[JB3DG]

Done. Both affect thrust although the fuel flow does more so than RPM

EDIT: sorry i misread. To clarify my test, the corrected FF was 0 which makes the other FF values 0 as well.

 

[Roy Holmes]

You said you had thrust, but was there a readout of TURB ENG JET THRUST? If so how did it compare to the static thrust value in the config file
Roy

 

[JB3DG]

Nope. I could only tell i had thrust because the jet was accelerating rapidly. For some reason whenever i have tried to get a readout of TURB ENG JET THRUST i always get 0s. Even when im not messing with simconnect.