Designed Combustor

[ganuganu]

Jul 24, 2015 18:06:12 GMT -5 racket said:

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

Flametube dia is about 65mm and casing dia is 85mm. I am going to use LPG(reduces complexity) as fuel.

[racket]

Hi

65 mm is far too small , even using propane you most likely will have combustion problems, a 65mm FT is only 40% bigger area than your inducer .

Cheers
John

[ganuganu]

Now iam actually planning to increase the flametube dia from 65 to 85mm because in simulation the flame front seems to produce hot spot in the centre, so i want to widen the radius of the flame to prevent hot spot (for even distribution of temperature at outlet of 900K).

Modified dimensions after analysis and yet this to be analysed:
flametube dia from 65 to 85mm
Casing dia from 85 to 105mm
Liner holes same 2mm 10holes
SZ holes 10mm 10 holes
DZ holes 16mm 10holes.

Moved the SZ holes a little forward to stabilize the swirls a liitle more.

[racket]

Hi

Theres still some "strange" hole area division going on .

Your 2mm dia holes in the Primary Zone are going to do nothing but cool the wall as their total area is only 31 sq mms , your secondary air will be used as main combustion air as its 785 sq mms is ~30% of inducer area , your Tertiary holes are far too big at ~2010 sq mms total , for such a small comp inducer of ~2375 sq mms .

Delayed combustion should result despite the increase in FT cross section to 85mm , which I feel is still too small .

A 95mm FT with ~600 sq mms of total hole area for the Primary Zone , ~500 sq mms for the Secondary and ~1,300 sq mms for the Tertiary, should produce what you want .

LOL, but then what would I know , I've never used CFD, only a pencil and paper to do design work ;-)

Cheers
John

[ganuganu]

In primary zone iam using swirlers as combustion air inlet about 0.0353kg/s. As you said the liner hole is just to cool the ft.

[racket]

Hi

That 0.0353kgs/sec only represents ~18% of total airflow .

How do you "regulate" that flow , is there a measured flow area , or is it simply a "theoretical" thats fed into the CFD .

Generally only some of the Primary air ( ~10% of total airflow) is fed through swirlers surrounding the fuel injection , the rest comes through wall/endcap holes.

If you want 18% , then the swirlers would have ~430 sq mms of area if using my design ratios.

Cheers
John

[ganuganu]

The swirlers are calculated and i dont implement any mass flow into each holes in CFD i specify the total mass flow in and the flow will be automatically divided by CFD based on the flow. My geometry looks as shown in the pic.

postimg.org/image/n2e1z5w9f/

[racket]

Hi

Yep , I can see how you're doing it now , very similar to the Dart combustor , may I suggest you do a bit of a Google search on its design , its quite complex with lotsa consideration to the "diffusion" surrounding the combustor for ideal air presentation to the actual entry points across the flametube wall .

Having your swirlers directly exposed to the incoming air , and having the swirlers processing basically all of the primary air will most likely be the cause of your problems .

Cheers
John

[finiteparts]

Ganesh,

The primary zone fuel air ratio is just like every other part that we design for the engine, a compromise. There is a value at the design point that we control and then we need to check how things change over the operating range of the engine. For my senior design we had a very similar arrangement to what you are proposing here.

In my project, we ran two primary zone design cases before we selected a primary mass ratio. The first one allowed 14% of the air into the primary zone and the second was 25%. The 14% air into the primary zone case had trouble at higher throttle settings due to the primary zone going rich (equivalence ratio, phi ~ 2.3), which is well above the suggested limit of an equivalence ratio of 1.5, above which the combustor will suffer from severe smoking issues. I ended up running the primary zone near 25% air for a max rich phi = 1.3 and had no idle stability/flame out problems due to the idle phi ~ 0.3, which was under the suggested lean blow out (LBO) phi~ 0.5.

We found quite a few "suggestions" for LBO, but we finally went with the results from a reactor model that we used, where we modeled the combustor as a Bragg combustor (the primary zone was modeled as a "well stirred reactor" WSR, and the secondary and tertiary zones were modeled as "plug flow reactors" PFR). It gave us a LBO limit phi ~ 0.28.

The primary zone flow was made up of the swirler airflow (14%), a small amount of dome cooling(~ 2%) and the entrained flow from the secondary jets, which was around 9% at full power. Because the amount of swirl created by the swirler and the placement and size of the secondary holes effects the amount of flow entrained back into the primary zone, it is hard to just throw out a number on the percentage of mass flow that should go into the primary zone. But, just to throw it out there, low emission combustors that run very lean headends are running primary zone mass ratios of over 80%, so 18% isn't a very big deal.

Additionally, our design didn't use any cowl diffusion, so our swirler was being fed directly from the discharge air that was coming from an axial cone diffuser at the casing inlet...I don't think a cowl diffuser is a big concern on these low PR engines unless you are trying to maximize your static pressure feed for backflow margin on dome cooling holes or something like that. If the primary diffuser is working properly, the chamber feed should be fairly low in velocity and thus there would be a very small margin gain in static pressure for the added complexity. Now, if the primary diffuser is a tube dumping into the side of a can style casing, it may help to stabilize the static pressure feed, bit there are also many cases were cowl diffusers disturb the flow to a point were it creates a pulsing flow. It should be relatively easy to do a ball park calculation to determine if it is worth the effort.

Ganesh, I think your design is close, but I think your swirler is sitting too far forward. If you transcribe a circle from the centerline of the combustor, the circle should be tangent to the dome and the edge of the secondary jet. It is often called the "magic circle" and it stems from the fact that what you want the secondary injection holes to do is to provide some momentum to the aft end of the recirculation bubble, that the swirler produces, to help energize that recirculation and to provide fresh air into the mix. The swirler and some portion of the secondary jets should add up to the amount of mass that you want in the primary zone. For a preliminary number, you could assume that 50% of the secondary flow gets entrained back into the primary zone. This will change depending on flow conditions, but 50% is a reasonable "guess". I would increase that half angle on that cone, which will pull the swirler back towards the secondary holes.

The design goal of a combustor is to get good mixing with as small a pressure drop across the liner as possible. Usually, it take at least 4-5% pressure drop across the liner to get good mixing, so that is usually a good target.

What kind of static pressure drop are you getting across the liner?

The common practice is to select a desired pressure drop, then calculate the allowable total hole size that would achieve that pressure drop. Once you know the total hole area, you can start to divide that up by region. The usual design goal for the primary zone is to have the equivalence ratio a little over 1 (phi = 1 means stoichiometric) because this provides a very stable flame. The secondary hole size is selected to "complete" the reaction requirements and minimize flame temps so that you don't get high dissociation losses. Dissociation occurs at high temps and effects molecules like N2 or O2. The heat of the combustion is used to break the chemical bonds between these molecules and thus doesn't get used to heat the gas, so it is a loss in the combustion efficiency. Dilution and cooling airflows then get selected last to make sure the liner and the downstream stuff don't melt.

I suggest getting Arthur Lefebvre's "Gas Turbine Combustion" book if you want to know more...it is the best resource out there on the subject!

Good luck!

Chris

[ganuganu]

I used Lefebvre book to design my combustor (even swirler design). I also referred several phd thesis to get an idea about the manual calculations. Mostly what people say is that in a combustor, the L/D ratio of each zones are mentioned based on experience since a proper calculation is not available to size. I have designed my combustor based on aerodynamics method.

The pressure drop across the liner is about 7790pa.