Combustors and Afterburners

Primary, secondary and dilution zones, air splits, liner cooling, pressure loss and pattern factor, combustor types, and real afterburners

The GE9X combustor

For the GE9X, GE raised the overall pressure ratio from the GE90's 40 to 60. Its combustor burns lean, in a Twin Annular Premixing Swirler (TAPS) design, with a 30 % margin under the CAEP/8 limit for nitrogen oxides; its inner and outer liners are ceramic matrix composite, running about 280 K hotter than nickel alloys (the GE9X). The LEAP, on the A320neo and 737 MAX, has the second-generation TAPS II (the LEAP). A combustor is short — a few tens of centimetres — and it must burn every drop of fuel, never blow out, lose only a few per cent of the stagnation pressure, and hand the turbine a gas whose temperature is nearly the same everywhere. This lesson is how it does all four at once.

💡Three zones

Air from the compressor is slowed by a diffuser (Lesson 2-3 showed why) and enters the liner in stages:

The primary zone, fed through the swirler at the dome: a recirculating, nearly stoichiometric flame that anchors the combustion.
The secondary (intermediate) zone: more air through rows of holes completes the burning of CO and soot.
The dilution zone: the rest of the air cools and mixes the gas to an even temperature the turbine can take.

Along the way a share of the air flows along the liner wall to cool it.

\frac{\dot m_{air,\,zone}}{\dot m_{air}} = \frac{f}{\phi_{zone}\,f_{st}}

The share of the combustor's air needed to bring the mixture to an equivalence ratio φ_zone, for an overall fuel–air ratio f.

Derivation

Air splits

All the fuel enters at the dome. If a fraction

x

of the air has joined it by some point, the local fuel–air ratio there is

f/x

, and the local equivalence ratio is

\phi = f/(x\,f_{st})

. Solving for

x

gives the share of air each zone must have received. With the overall

f

fixed by the turbine entry temperature, the designer's real choices are the zone equivalence ratios.

📐Splitting the air

Overall

f = 0.025

with air at 850 K and 40 bar; zone temperatures from Cantera (Lab 1):

Primary zone at $\phi = 1.0$ : 36.4 % of the air, 2679 K.
After the secondary air, $\phi = 0.6$ : 60.7 % of the air, 2151 K.
After dilution, $\phi = 0.364$ : all the air, 1702 K at the turbine.

About a third of the air takes part in the flame; the rest finishes, cools and dilutes it.

💡Liner cooling, pressure loss and pattern factor

Liner cooling: the liner faces gas near 2400–2700 K. Films of air along its inner wall hold metal liners near 1150 K; ceramic matrix composite liners tolerate about 280 K more and need less air.
Pressure loss: the diffuser, the swirler, the holes and the heat addition together cost about 3–5 % of the stagnation pressure — the burner pressure ratio $\pi_b \approx 0.95$ – $0.97$ of Unit 4.
Pattern factor: $(T_{max} - T_{t4,mean})/(T_{t4,mean} - T_{t3})$ , the hottest streak relative to the mean temperature rise. The turbine's first vanes see the hottest streak, so a lower pattern factor allows a higher mean temperature.

📐What a ceramic liner frees up

With 2400 K gas and 850 K coolant, and an illustrative film-cooling law (Lab 2):

A nickel liner held at 1150 K needs a film effectiveness of 0.81 — about 20 % of the combustor air.
A CMC liner at 1430 K needs 0.63 — about 12 %.

The 8 % of air the ceramic frees can go to mixing and dilution: a flatter temperature profile for the turbine, or a leaner, cleaner flame (Lesson 7-4).

Combustor types. Early engines used separate can combustors around the shaft (the Jumo 004 had six). Cannular designs put cans inside a common casing. Nearly all modern engines use a single annular combustor — shorter, lighter, with less surface to cool and a more even exit temperature — and small engines sometimes fold it back on itself (reverse flow) to shorten the shaft.

Afterburners. An afterburner is a combustor with no turbine to protect: fuel is sprayed into the turbine exhaust, held by V-gutter flameholders, and burned to 2000 K and more before a variable nozzle (Lesson 3-3). It pays three penalties: the Rayleigh loss of burning in moving gas (Lesson 2-3), the drag of the flameholders even when unlit (the dry loss), and the liners and screech dampers needed to survive the heat and the acoustic resonances of a large burning duct.

📐A real afterburner

A turbojet at Mach 0.9 with its afterburner raising 1300 K turbine exhaust to 2200 K (Lab 3):

Ideal afterburner: specific thrust from 973 to 1335 N·s/kg.
Entering at Mach 0.2, with a Rayleigh loss and 3 % for the flameholders: 95.4 % of the stagnation pressure kept, thrust ×1.36.
Entering at Mach 0.4: 89.9 % kept, thrust ×1.34, and the TSFC rises from 33.3 to 43.3 mg/(N·s).

The diffuser in front of an afterburner pays for itself — in pressure, thrust and fuel.

There cannot be a greater mistake than that of looking superciliously upon the practical applications of science. The life and soul of science is its practical application.

– William Thomson, Lord Kelvin, lecture on "Electrical Units of Measurement", 3 May 1883 (Wikiquote)

💡Rich, lean and staged

The primary zone's equivalence ratio is the big design choice. Conventional combustors run it near stoichiometric. Rich-quench-lean (RQL) combustors burn it rich, then mix air in so fast that the gas skips through the hottest, NO-forming mixtures. Lean premixed combustors, like the TAPS family, mix fuel and air before burning so the flame never gets hot enough to make much NO — at the price of a harder fight for stability. Lesson 7-4 weighs them.

💡Where you will use this

Lesson 7-4 connects the zone temperatures to emissions. Lesson 8-3 looks at the combustor during throttle movements, when rich or weak blow-out threatens. The burner pressure ratio feeds back into the cycle of Unit 4.

📚Further reading

Arthur H. Lefebvre and Dilip R. Ballal, Gas Turbine Combustion (3rd ed., CRC Press, 2010).
Jack D. Mattingly and Keith M. Boyer, Elements of Propulsion: Gas Turbines and Rockets (2nd ed., AIAA, 2016), on combustors and afterburners.
The GE9X on Wikipedia, on its lean-burning combustor and ceramic liners.

↓But there is a problem

Every zone choice in this lesson trades one pollutant for another. Hot zones make nitrogen oxides; cool zones leave carbon monoxide and unburned fuel. Why should nitrogen — the most inert gas in the air, the gas we breathe in and out unchanged — burn at all? A young Soviet physicist worked it out in 1946.