Fuel consumption of the A320

[Hans-Georg Wunder]

Hello all,

I tried to understand the calculation of the fuel consumption
of the A320. In the engine configuration file I found the

TSFC thrust-specific fuel consumption (lbm/hr/lbf)

parameter set to 0.657 . Then I compared the fuel consumption with
thrust (/engines/engine/fuel_flow_pph with /engines/engine/thrust_lb)
and when I set N1=100 %

TSFC = fuel_flow_pph/thrust_lb


But this formula is only valid, when N1 is 100 %
For values lesser then 100 % the formular is not valid.
The fuel consumption is much greater then it should be.

I compared this result with the ICAO ENGINE EXHAUST EMISSONS DATA BANK
where I found some performance values for different power settings (100
%, 85 %, 30 % 7%) of the A320 engine CFM56. There the formula is always
valid.

Is this behaviour a bug or a feature ?

Where is the fuel consumption calculated ?

In the jsbsim-code FGSimTurbine.cpp I only found this peace of code:

if (!Augmentation) {
FuelFlow_pph = Seek(&FuelFlow_pph, thrust * TSFC, 1000.0, 100000);
if (FuelFlow_pph < IdleFF) FuelFlow_pph = IdleFF;

But there the fuel flow is calculated in the way I understand


Thank you for any help

Kind regards

Hans-Georg

[David Culp]

Post by Hans-Georg Wunder
Is this behaviour a bug or a feature ?
Where is the fuel consumption calculated ?

It's in FGTurbine.cpp:

double correctedTSFC = TSFC + TSFC - (N2norm * TSFC);
FuelFlow_pph = Seek(&FuelFlow_pph, thrust * correctedTSFC, 1000.0,
100000);


The published TSFC value is usually for cruise conditions, so I've made it
vary with N2 to better model fuel flows at low N2, where the compressor is
inefficient. I matched the results with known fuel flow values for the
CFM-56, and the results are pretty good.

Do you have a better formula to vary TSFC? If so let's get it into the model.



Dave

[Hans-Georg Wunder]

Post by David Culp
The published TSFC value is usually for cruise conditions, so I've made it
vary with N2 to better model fuel flows at low N2, where the compressor is
inefficient. I matched the results with known fuel flow values for the
CFM-56, and the results are pretty good.
Do you have a better formula to vary TSFC? If so let's get it into the model.

No, I have no better formula, but I found other information:

This site says, that the TSFC is constant for a jet:

http://142.26.194.131/aerodynamics1/Performance/Page8.html

On this site I found tests results of various engines:

http://www.caa.co.uk/srg/environmental/emissions/document.asp?groupid=704

For the CFM56-5B4 I found

100 % is 117,9 kN = 26500 lbf

Power
Fuel flow kg/s TSFC kg/daN *h
100 % 1.166
0,35
85 % 0.961
0,35
30 % 0.326
0,31
7 % 0.107
0.47

This seems to be constant in the range of 30 % to 100 %

At 'Klaus Hünecke: Technic of modern airliners' I found at

TSFC for cruise thrust = 0.611
TSFC for static thrust = 0.35 ( with meets the ICAO values)

May be it is a good idea to meet the published TSFC values
at cruise speed (N1 about 80 or 85 %), not at N1=100 %
To put these conditions in a formula, I have to sleep on it one or two
nights.

Kind regards

Hans-Georg

[David Culp]

Post by Hans-Georg Wunder
May be it is a good idea to meet the published TSFC values
at cruise speed (N1 about 80 or 85 %), not at N1=100 %

That's a good idea. We can assume that most TSFC values will be given for
cruise conditions, about 80 to 85 percent N1. The turbine model uses N2,
only because it has to also make sense for turbojets. This formula looks
better:

correctedTSFC = TSFC * (1.5 - (N2norm * 0.58823));


This will make the TSFC value just right at 85% N2. It also flattens the
curve a bit which will result in less variation over the range of N2.



Dave

[Hans-Georg Wunder]

Post by David Culp

Post by Hans-Georg Wunder
May be it is a good idea to meet the published TSFC values
at cruise speed (N1 about 80 or 85 %), not at N1=100 %

That's a good idea. We can assume that most TSFC values will be given for
cruise conditions, about 80 to 85 percent N1. The turbine model uses N2,
only because it has to also make sense for turbojets. This formula looks
correctedTSFC = TSFC * (1.5 - (N2norm * 0.58823));
This will make the TSFC value just right at 85% N2. It also flattens the
curve a bit which will result in less variation over the range of N2.
Dave

I think this is a good first approach. But it covers not the static
thrusts condition, when the aircraft is at the airport. There must
be a dependency with the airspeed. I search for hours on the net to find
some hints regarding this item. Nothing.

Never mind. To have a correct fuel consumption for cruise is
more important than the consumption at the airport. We just have to
hurry up with take off :)

Hans-Georg

[Jon S Berndt]

On Tue, 09 Nov 2004 22:41:16 +0100

Post by Hans-Georg Wunder

Post by Hans-Georg Wunder
May be it is a good idea to meet the published TSFC values
at cruise speed (N1 about 80 or 85 %), not at N1=100 %

Never mind. To have a correct fuel consumption for cruise is
more important than the consumption at the airport. We just have to
hurry up with take off :)
Hans-Georg

Dave:

is this something that you have/will implement in JSBSim for the
turbine model?

Jon

[David Culp]

Post by Jon S Berndt
is this something that you have/will implement in JSBSim for the
turbine model?

I'm not sure yet. Now that my internet connection is back up I'm trying to
find out what exactly is the problem that Hans-Georg is having. I need
specifics.

An overview of how FGTurbine calculates thrust and fuel flow:

Idle thrust is looked up in the idle thrust table, which uses speed and
altitude as inputs.

Full thrust (military thrust) is looked up in the milthrust table, also using
speed and altitude as inputs. Idle thrust is subtracted from this here. It
is added back in later.

N1 and N2 targets are based on the throttle position and the idle N1 and idle
N2 values from the configuration file. A function called Seek() acts as a
lag filter to spin the engine to the target N1 and N2 realisticaly.

A value called N2norm is found. This is just the N2 range (from idle N2 to
maximum N2) normalized.

A "thrust coefficient" is found, ...

Oops, that's not my code. It looks like we're getting thrust squared?

I'll fix it.


Dave

[Hans-Georg Wunder]

Post by David Culp

Post by Jon S Berndt
is this something that you have/will implement in JSBSim for the
turbine model?

I'm not sure yet. Now that my internet connection is back up I'm trying to
find out what exactly is the problem that Hans-Georg is having. I need
specifics.

I have problems with the fuel consumption, not with thrust:

Davids last approach was:

correctedTSFC = TSFC * (1.5 - (N2norm * 0.58823));

The correctedTSFC will vary from

N2norm = 0 -> correctedTSFC=1.5 TFSC
to
N2norm = 0.85 -> correctedTSFC= TFSC
(This value should be the value at cruise speed,
which you will find in the docs)
to
N2norm = 1 -> correctedTSFC=0.91 TFSC

For the CFM56-5 TFSC at

http://www.caa.co.uk/srg/environmental/emissions/document.asp?groupid=704
and
at 'Klaus Hünecke: Technic of modern airliners' I found

TSFC for cruise thrust = 0.611

but also

TSFC for static thrust = 0.35 ( with meets the ICAO values)

The TSFC-value of 0.35 seems to be a value in a test bed, but
should also be the value, when the aircraft does not move (airport).
And this value we will not get with the approach of David.
But I have no better idea :)

Hans-Georg

[David Culp]

Post by Hans-Georg Wunder
I think this is a good first approach. But it covers not the static
thrusts condition, when the aircraft is at the airport. There must
be a dependency with the airspeed.

Engine performance does vary with speed, but this is done in the "idle thrust"
table in the engine configuration file.

Post by Hans-Georg Wunder
I search for hours on the net to find
some hints regarding this item. Nothing.
Never mind. To have a correct fuel consumption for cruise is
more important than the consumption at the airport. We just have to
hurry up with take off :)

Do you mean that your fuel flow at idle is not what you're expecting?


Dave