Single Spool Turbofan Engine Simulation

[Antonio Hadade]

Dear Sirs,

I am trying to simulate a single spool mixed flow turbofan (a configuration I saw on a commercial simulation software called GasTurb 12). I have first built the static model which have given me reasonable results for a given design point. However, the dynamic simulation gives me a result that seems a little unusual. During deceleration from design point, the temperature within the burner seem to INCREASE, when I would expect it to decrease. It seems that the bypass ratio (which is set as variable), is increased and the core air flow decreases too much, ending up with an even higher fuel-to-air relation than it was at higher rotation. I am wondering if any of you have simulated that kind of configuration before and if my model is wrong some how. It has been based on the JT9D fan model, and it is shown at the attached pictures, and explained below:

-----------------------------------------------------------

Model:

1 - Inlet block connected to fan (compressor block)

2 - Fan exit flow goes to splitter, where its divided into fan duct flow and core flow

3 - Core flow from splitter to compressor block

4 - Core flow from Compressor flow into burner (which receives also fuel flow)

5 - Core flow from burner into turbine

6 - Core flow from turbine to mixer and fan duct flow from splitter to mixer

7 - Total flow from mixer into nozzle

8 - Torques from Fan, Compressor and Turbine Goes into Shaft Block

9 - Mechanical rotation goes from iterative solver into Shaft Block (static case) or From Integrator into Shaft Block (dynamic case)

-----------------------------------------------------------

Static Iterative Solver

Independent Variables:

1 - Inlet Air Flow

2 - Fan RLine

3 - Splitter Bypass Ratio

4 - Compressor RLine

5 - Turbine Pressure Ratio

6 - Mechanical Rotation

 Errors:

1 - Fan Flow Error

2 - Compressor Flow Error

3 - Turbine Flow Error

4 - Mixer Flow Error

5 - Nozzle Flow Error

6 - Ndot (Derivade of Mechanical Rotation)

-----------------------------------------------------------

Dynamic Iterative Solver

1 - Inlet Air Flow

2 - Fan RLine

3 - Splitter Bypass Ratio

4 - Compressor RLine

5 - Turbine Pressure Ratio

 Errors:

1 - Fan Flow Error

2 - Compressor Flow Error

3 - Turbine Flow Error

4 - Mixer Flow Error

5 - Nozzle Flow Error

-----------------------------------------------------------

Could anyone give me a help? Thank you very much!!

P.S.: I was wondering if, for single spool, the bypass ratio should be kept constant. If so, some of the errors generated should be ignored, so the bypass could be fixed with the design point value... 

Antonio 

[grc.rhc.dev]

Antonio, 

Thanks for posting. Unfortunately I do not have a mixed flow engine that is releasable at this time. From what I can tell, your block setup looks good. It might be good to take a look at how your scalars (for compressor and turbine maps) and areas (for nozzle and mixer) are calculated. Using the Idesign tool and knowing your design point values should produce good numbers. Note: it might be a good idea to add fuel flow to your steady-state solver as an independent then add the shaft speed minus a desired shaft speed as a dependent. This will drive the system to a shaft speed using the fuel flow and is generally a bit more reliable then using a known fuel flow.

I would not suggest fixing the bypass air ratio (BPR).  It is determined to get a balance in the system, so if you fix it and move off design it wont be correct. If your BPR is getting too large, it may be that the mixer areas aren't right and the pressure in the bypass is just way too low. Decreasing the mixer area to the bypass flow should raise this pressure, though if you use Idesign this number will be calculated automatically. 

You mention this issue only occurs in your dynamic simulation,  does it also occur running steady state at off design points? Also, are you getting any T-MATS warnings? 

-Jeff

[Antonio Hadade]

Jeff,

 

I do am getting warnings even when executing iDes mode (iDes = 0), all of them concerning the mixer. I have downloaded GasTurb free trial for simulating their example of a 1 spool  turbofan with mixed flow and right now I’m trying to reproduce the same results with T-MATS, since it is way more flexible. Below are all characteristics I have been taken in consideration and results I have acquired, if needed.

 

 

Inlet:

 

Altitude: 0

Delta T from ISA: 0

Mach Number: 0.8

Relative Humitidy: 0 %

Inlet Pressure Ratio: 0.99

 

Fan:

 

Inlet Corrected Mass Flow: 25 kg/s

Fan Pressure Ratio: 3

Fan Isentropic Efficiency: 0.9

 

Splitter and Ducts:

 

Design Bypass Ratio: 0.6

Bypass Duct Pressure Ratio: 0.97

Compressor Interduct Pressure Ratio: 0.99

 

Compressor:

 

Compressor Design Pressure Ratio: 5

Compressor  Isentropic Efficiency: 0.84

 

Burner:

 

Burner Exit Temperature: 1500

Burner Design Efficiency: 0.999

Burner Pressure Ratio: 0.96

Fuel Heating Value: 43.124 MJ/kg

 

Turbine:

 

Turbine Design Isentropic Efficiency: 0.88

Turbine Exit Duct Pressure Ratio: 0.99

 

Mixer:

 

Hot Stream Mixer Pressure Ratio: 1

Cold Stream Mixer Pressure Ratio: 1

Mixed Stream Pressure Ratio: 1

Mixer Efficiency: 0.4

Design Mixer Mach Number: 0.3

 

Bleedings:

From Compressor Exhaust to NGV and HPT cooling: 4% and 1%, respectively

 

 

T-MATS model (blocks have been copied from JT9D with the corresponding maps being used) without any bleed or ducts:

 

Inlet Block:

 

Alt = 0;

dTamb = 0;

MN = 0.8;

 

Fan Block:

 

NcDes_M - map Design = 1;

Eff_Des = 0.9;

PR_Des = 3;

Rline_Des – map Design = 2.3706;

No Bleeds;

Rline = 1 defined as surge line;

iDes = 0;

 

Compressor Block:

 

NcDes_M - map Design = 0.993;

Eff_Des = 0.84;

PR_Des = 5;

Rline_Des – map Design = 1.7664;

No Bleeds

Rline = 1 defined as surge line;

iDes = 0;

 

Burner Block:

 

Burner_Efficiency: 0.999;

Fuel Flow: 0.45808*2.2 (same obtained as a result in GasTurb, but in pps for T-MATS)

 

Turbine Block:

 

NcDes_M map design = 99.031;

EffDes_M = 0.88;

PRDes_M = 4.930;

NDes = 20000;

iDes = 0;

 

Mixer Block:

 

s_M_Imp1_M = 1;

s_M_Imp2_M = 1;

s_M_V1_M = 1;

s_M_V2_M = 1;

SWPS_M = 0; -> I assume the primary flow is the core flow

BPRDes_M = 1;

MNDes_M = 0.3;

iDes = 0;

 

 

Nozzle Block:

 

SWType_M = 1;

WDes_M = 35.982*2.2 (the design point mass flow obtained is 35.98 and it has to be in pps for T-MATS)

 

 

Initial conditions for solver:

 

%Minimum Calculation Error

MWS.Solve.C_Lim = 0.001;

 

%Order(->): 1-FanFlowError 2-MixerFlowError  3-NozzleFlowError 4-TurbineFlowError 5-CompressorFlowError 7-Ndot

MWS.Solve.J_Calc_Per = [0.01 0.01 0.01 0.05 0.05 0.01; -0.01 -0.01 -0.01 -0.05 -0.05 -0.01];

 

%Initial Contidions:

%Order(->): 1-Win 2-RlinecCompressor 3-PRt 4-N(RPM) 5-FanRline

MWS.Solve.NR_IC = [25 2 2 26400 2 0.6]';

 

%Derivade size

MWS.Solve.NR_dx = 2;

 

%How many times until next Jacobian calculation

MWS.Solve.NR_Jac_Re = 20;

 

%How many interactions at the most

MWS.Solve.Max_Iter = 200;

 

%Sampling Time or Integration Step

MWS.Solve.T = 0.015;

 

 

The results for the above input data are shown below:

 

 

Parameter	GasTurb	T-MATS
W2 – Inlet Air Flow (kg/s)	35.524	35.52
T2 – Inlet Total Temperature (K)	324.96	325.1
P2 – Inlet Total Pressure (kPa)	152.899	154.6
Bypass Ratio	0.6	1.026
P13 (kPa)	443.4	463.7
T13 (K)	453	457.2
W13 (kg/s)	13.3216	17.99
P3 (kPa)	2270	2319
T3 (K)	757.323	758
W3 (kg/s)	22.2026	17.53
Wfuel (kg/s)	0.45808	0.45808 (assumed)
W4 (kg/s)	21.55	17.99
T4 (K)	1500	1619
P4 (kPa)	2179	2203
W5 (kg/s)	22.66	17.99
T5 (K)	1038	1163
P5 (kPa)	384	407
A1 (Mixer) (m2)	0.11323	0.07582
A2 (Mixer) (m2)	0.02266	0.02635
Aex (Nozzle)	0.06840	0.06153
W8 (kg/s)	35.98	35.98
FN (kN)	16.36	26.33

 

 

I do am getting some warnings and errors during execution of iDes, all of them concerning the Mixer:

 

Warning in GasTurbine_SS_Template/Mixer, Error calculating Rt1. Vector definitions may need to be expanded.

Warning in GasTurbine_SS_Template/Mixer, Error calculating Rtout. Vector definitions may need to be expanded.

Warning in GasTurbine_SS_Template/Mixer, Error calculating gammatg. Vector definitions may need to be expanded.

Warning in GasTurbine_SS_Template/Mixer, Error calculating gammasg. Vector definitions may need to be expanded.

Warning in GasTurbine_SS_Template/Mixer, Error calculating iteration gammasg. Vector definitions may need to be expanded.

Warning in GasTurbine_SS_Template/Mixer, Error calculating input static pressures

Warning in GasTurbine_SS_Template/Mixer, Error calculating Psout

Error in GasTurbine_SS_Template/Mixer, unable to calculate PtOut

Warning in GasTurbine_SS_Template/Mixer, unable to caluclate PsOut, PtOut may contain high error

 

I believe I am doing something wrong at the mixer, since I always get all those warnings and errors. Also, the parameters required by GasTurb concerning the Mixer are a bit different from those required by T-MATS. 

Antonio 

[grc.rhc.dev]

Antonio, 

I would probably suggest doing a component by component matching.  You have the Gas Turb model, so you can just feed each block exactly what you see in Gas Turb and locate your problem when the outputs don't match.  

It also seems strange that at your design point the BPR isn't matching,  if you are running with IDes enabled then the on design BPR is defined within the mixer, should probably look into your Ides parameters for the Mixer as well to make sure they are what you intend. Also, the P13 is off by nearly 5%, with IDes enabled you directly define the fan pressure ratio, so it seems like your iDes parameter for the fan might not be what you intended.

Essentially with IDesign you specify your design point, so when running it your maps, scalars, and areas aren't used. The tool will then calculate what the parameters must be for you to run at the specified design point, so when you run off design you can use those numbers (setting IDes to 1). 

Because you have the GasTurb model, you could also just calculate all the parameters manually, enter them in and turn IDes off.

-Jeff

[Antonio Hadade]

Jeff,

Thank you very much! I will do a more careful check. And what about the error and warning messages? Does that mean there is anything wrong with mixer settings? According to the messages, it seems the mixer is not being able to reach an appropriate value, but since the errors did converge to a practically null value if I give the simulation enough time, I assumed that the messages have been printed only during the first iteractions and did not give much importance to them. Should there not be any messages at all?  I think debugging the C code directly would help a lot the evaluation of each block individually (specially because I am not entirely familiar with the mixer's math, since its the first time I simulate such component) and I do intend to do this, at first using an external tool such as Visual Studio (Windows) or X-Code (Mac). If you have any suggestions, please let me know. 

Antonio

[grc.rhc.dev]

Antonio, 

That is definitely possible. The first couple of mixer errors are with the simple table lookups and fail only when your FAR is out of bounds (0 to 0.03), judging from your previous statements on FAR being very high this isn't so surprising. It could be that once your FAR is too high all the internal table lookups are causing the iterative calculation of pressure to fail.  I think the component level test with inputs that are in bounds should help flush out any parameter setting issues.

The paper Chapman, J.W., Lavelle, T.M., Litt, J.S., Guo T.-H., "A Process for the Creation of T-MATS Propulsion System Models From NPSS Data", NASA/TM- 2014-218409,  details an approach to a similar problem. There is no mixer, but it goes through a matching exercise that you may find useful.

-Jeff

[zheng figozfh]

Antonio

            I also met the problem of the mixer when iDesign equals 0. It will make the aera of  mixer inlet become large, A1 nearly 99999 in^2.

[TianDi strong]

I have read the mixer block core code,and I find the error in the code,that's the reason when mixer works in off-design condition,it produces wrong result.I modified the code and totally replace nearly all T-mats blocks written in C by my blocks written in matlab language,creating matlab function block in simulink.I share my mixer code,and you can compare original code to find the issue in mixer C code.Also I have created mixed flow turbofan engine and it works well.If you have any questions when using mixer block,or meeting problem creating mixed flow turbofan engine,email me by 8623...@qq.com(I am chinese and seldom login google email).