ganuganu
Member
Joined: November 2013
Posts: 30 
|
Post by ganuganu on Jul 17, 2015 1:32:36 GMT -5
  Hello everyone, recently i bought a hx35w turbocharger and i tried to design a combustor by selecting a operating point in compressor map. Mass flow: 0.196kg/s Fuel flow: 0.00205kg/s Combustion air flow: 0.0353kg/s In pressure(to CC): 222.92kpa Operating PR: 2.2 The links are the starccm+ simulated file of my designed combustor with pz,sz and dz holes. This is a straight injection method but i have designed a swirler for air inlet to pz. Swirler method will be simulated soon. In current simulation the average outlet temp is 800K and velocity is 70m/s. Model used: EBU model Turbulence: K-e postimg.org/image/eo8hhyavb/postimg.org/image/7jqo8x3lz/postimg.org/image/jkqcy8w3r/Simulation with swirlers at 45deg angle postimg.org/image/zcbosid4b/The swirls are clearly seen in the vector plot. postimg.org/image/xj8s46rxn/ |
|
|
|
Post by finiteparts on Jul 18, 2015 22:08:29 GMT -5
Hi ganuganu,
The first combustor without any swirl, would probably not work. The reaction is contained deep in the core flow and the dilution jets don't appear to ever get to it, much less reduce the mean gas temperature out of the combustor. This would likely hot streak so bad that it would burn your turbine section.
The second one with the swirl looks like you are moving in the right direction. I would get rid of the row of small hole in the forward cone area. Then I would move the secondary holes forward to help energize the primary recirculation zone. I might even try to reduce the hole sizes to help get the secondary jets to penetrate further into the core flow.
The next row of holes are probably not needed and you should try the simulation without them.
The dilution holes look fine and with the reduced overall hole area (because you deleted some of the other unnecessary holes), the pressure drop across the liner will be higher which helps to push the dilution jets deeper into the core flow. You really want to see the tertiary dilution jets almost touching each other near the centerline.
Finally, don't put a lot of stock in reacting CFD. It is a good indicator of were you would like things to be, but it is a really challenging problem to capture all the physics that is happening in there and usually the CFD is off. That is why engineers often humorously call it "colorful fluid dynamics".
I am not trying to discourage you, because I think you can get a good idea of where to go in your design by using CFD, just don't lock onto the values.
Good luck!
Chris
|
|
ganuganu
Member
Joined: November 2013
Posts: 30
|
Post by ganuganu on Jul 20, 2015 8:35:37 GMT -5
Hi ganuganu, The first combustor without any swirl, would probably not work. The reaction is contained deep in the core flow and the dilution jets don't appear to ever get to it, much less reduce the mean gas temperature out of the combustor. This would likely hot streak so bad that it would burn your turbine section. The second one with the swirl looks like you are moving in the right direction. I would get rid of the row of small hole in the forward cone area. Then I would move the secondary holes forward to help energize the primary recirculation zone. I might even try to reduce the hole sizes to help get the secondary jets to penetrate further into the core flow. The next row of holes are probably not needed and you should try the simulation without them. The dilution holes look fine and with the reduced overall hole area (because you deleted some of the other unnecessary holes), the pressure drop across the liner will be higher which helps to push the dilution jets deeper into the core flow. You really want to see the tertiary dilution jets almost touching each other near the centerline. Finally, don't put a lot of stock in reacting CFD. It is a good indicator of were you would like things to be, but it is a really challenging problem to capture all the physics that is happening in there and usually the CFD is off. That is why engineers often humorously call it "colorful fluid dynamics". I am not trying to discourage you, because I think you can get a good idea of where to go in your design by using CFD, just don't lock onto the values. Good luck! Chris Hello Chris, really thank you for your detailed reply. Today i did another simulation where the mass flow for each holes were not specified instead the mass flow will be automatically divided on its own (inlet diffuser+casing). The mesh count was about 9million done in starccm+ 12 core. The mass flow averaged temperature was about 720K and the exit average velocity was about 54m/s. Tomorrow iam planning to simulate the suggestion given by you(to move the secondary holes little forward and remove the holes in primary zone). Iam doing this simulation just for a reference and confirmation whether the design iam doing is atleast in acceptable range and to improve my design calculation procedure. Sorry for not posting the images today, tomorrow ill do it for sure(pics of both simulations). |
|
ganuganu
Member
Joined: November 2013
Posts: 30
|
Post by ganuganu on Jul 23, 2015 4:15:17 GMT -5
The final Design Considerations and results after the analysis both in ANSYS and Starccm+. They are as follows:
ANSYS (PDF Table) : Average exit temp - 805K
Average exit velocity - 142m/s
Starccm+ (EBU model): Average Exit temp - 760K
Average Exit velocity - 135m/s
Expected Thrust: 50 to 60N.
Even in ANSYS EDM was analysed and the results were same like EBU in Starccm+
Liner Holes: 2mm (Count 10)
Secondary Holes: 10mm (Count 10)
Dilutions holes Set 1: 10.5mm (Count 10)
Dilution holes set 2: 9.5mm (Count 10)
Mass flow Through LH + SH + DH1 +DH2 = 0.158Kg/s
Mass flow through Swirler: 0.0375Kg/s
Total Length of the CC = 205mm
Total Length with Diffuser = 380mm
Swirler Blade Count: 8
Note: Exit nozzle has been designed at the turbine exit.
Fabrication to be started.
|
|
|
|
Post by racket on Jul 23, 2015 4:50:29 GMT -5
Hi
You seem to have an enormous amount of dilution holes , roughly twice the area that I'd expect .
I generally design the flametube total hole area to be the same area as the comp wheel inducer area , with 30% of area to Primary , 20% to secondary , and the remaining 50% to Tertiary .
Cheers
John
|
|
ganuganu
Member
Joined: November 2013
Posts: 30
|
Post by ganuganu on Jul 23, 2015 8:36:18 GMT -5
Hi You seem to have an enormous amount of dilution holes , roughly twice the area that I'd expect . I generally design the flametube total hole area to be the same area as the comp wheel inducer area , with 30% of area to Primary , 20% to secondary , and the remaining 50% to Tertiary . Cheers John Yes you are correct. I need a low temperature of around 500 deg c and to prevent hot spots at the exit, that is why i am using so much quantity of air in DZ. Inducer dia is about 55mm |
|
|
|
Post by racket on Jul 23, 2015 17:20:05 GMT -5
Hi
Do you have a Link to the comp map you used ??
Your mass flow of 0.196 Kgs/sec seems a tad low for a 55mm inducer .
A 55mm dia - 2375 sq mms , would "normally" have a total dilution hole are of ~1200 sq mms .
Why do you need such a low temperature ??
The standard turbine wheel is good for 800 C , at 500 C for a T I T , the engine will barely keep its self spinning around at a 2.2 PR .
At a 2.2 PR,even with a very respectable comp effic of 78%, theres a 94 deg C rise in compression , with a 5% pressure drop across the flametube you'll have a 2.09 PR going into the turb stage .
The small turb stage will probably have an efficiency of ~70% , with a 500 C -773K entry temp and the requirement for a temp drop of ~83 deg C in the turb , you'll be needing a 1.95 PR across the turb stage , leaving only a 1.074 PR in the exhaust gases , or ~1 psi of dynamic pressure with the gases at ambient pressure , that 1 psi is not even going to equate to the necessary gas velocity for the gases to exit the turb exducer , 1 psi is only a couple of hundred ft/sec .
You'll be needing a higher temperature .
Cheers
John
|
|
ganuganu
Member
Joined: November 2013
Posts: 30
|
Post by ganuganu on Jul 23, 2015 21:21:36 GMT -5
This is the compressor map iam currently using.  |
|
|
|
Post by racket on Jul 24, 2015 0:23:46 GMT -5
Hi
Thanks for the Link
Is there any reason why you want to run the turbo at such a relatively low rpm/P2 pressure ??
Cheers
John
|
|
ganuganu
Member
Joined: November 2013
Posts: 30
|
Post by ganuganu on Jul 24, 2015 1:17:34 GMT -5
Hi Thanks for the Link Is there any reason why you want to run the turbo at such a relatively low rpm/P2 pressure ?? Cheers John For this first practical test i dont want to produce more thrust (around 30 to 50N is enough), little is enough because my goal is to validate my experimental results to manual calculation to CFD analysis and generate a comparison report for future reference. There are so many cfd tools available i just want to check how much accurate the CFD results can be. |
|
|
|
Post by racket on Jul 24, 2015 1:53:24 GMT -5
Hi
Ok , thats a good enough reason .
So you'll be needing a gas velocity of ~600 - 750 ft/sec to get your design thrust , thus represents a pressure ratio of ~1.19 exiting the turbine wheel , or ~2.8 psi of combined dynamic/static pressure . I don't think you'll have that at the temperatures you are designing for .
I'd be looking more towards a temp of 600 deg C exiting your combustor if you want that sort of thrust .
Cheers
John
|
|
ganuganu
Member
Joined: November 2013
Posts: 30
|
Post by ganuganu on Jul 24, 2015 2:04:53 GMT -5
Hi Ok , thats a good enough reason . So you'll be needing a gas velocity of ~600 - 750 ft/sec to get your design thrust , thus represents a pressure ratio of ~1.19 exiting the turbine wheel , or ~2.8 psi of combined dynamic/static pressure . I don't think you'll have that at the temperatures you are designing for . I'd be looking more towards a temp of 600 deg C exiting your combustor if you want that sort of thrust . Cheers John You are absolutely right (both on velocity and temp), actually i did the thermodynamic calculations for 600deg C (for which i got thrust of about 55N) but when i did numerical analysis the avg exit temp i got was about 500 deg c. Now iam again working on the distribution of air in dilution zone and secondary zone so that i would be getting the required temp at turbine inlet. Once in your reply you told that at this PR (2.2) my turbine will barely self sustain, does it mean even at 600 deg C i wont be able to sustain my engine? |
|
|
|
Post by racket on Jul 24, 2015 3:45:39 GMT -5
Hi
Small turbos , and yes, your 55mm turbo is a small turbo when it comes to making DIY turbine engines , often have poor performance due to inefficient turbine stages as well as mismatching of compressor and turbine flows , they're OK as a turbo on an IC engine , but not when converted into a turbine engine .
As for getting the required T I T , its really just a matter of how much fuel you feed into the engine and the "backpressure" on the engine , as long as theres a reasonable division of hole area along the flametube , the engine sorts it out "automatically".
If you use the normal ratio of total hole area equal to inducer area , and divide it 30-20-50% , the engine should work .
Cheers
John
|
|
ganuganu
Member
Joined: November 2013
Posts: 30
|
Post by ganuganu on Jul 24, 2015 4:36:19 GMT -5
I accept the point that automobile turbo (turbines) are not sized well to suit gas turbine purpose. I will work on the fuel flow and air distribution and post the sized values here. Thank you very much for the help.
|
|
|
|
Post by racket on Jul 24, 2015 18:06:12 GMT -5
Hi
What is the cross sectional flow area of your flametube ??
I generally use 3 times inducer area as a minimum , so for your 55mm inducer that'd be a 95mm ID for the flametube , anything smaller and air speeds start getting a bit high for burning kero/diesel in the primary zone .
Cheers
John
|
|
