Sid Sharma.CV
Home / Research interests

Fuels, combustor flow & turbine-inlet conditions

How do fuel properties alter mixing, reaction and heat release, and how do these processes jointly govern combustor turbulence, pollutant formation and the non-uniform flow state delivered to the turbine?

Background

Sustainable aviation fuels can cut life-cycle greenhouse-gas emissions using existing aircraft and fuel infrastructure. Drop-in fuels can still burn differently: differences in composition and physical properties change evaporation, fuel–air mixing, reaction rates and heat release [1, 2].

Full-scale tests show particle reductions with low-aromatic fuels [3, 4], but they do not isolate the mixing and oxidation mechanisms behind them. Soot forms intermittently in fuel-rich regions and can oxidize as mixing proceeds [5]. Mean temperature, pressure and mass flow cannot describe these local processes, or the conditions leaving the combustor.

Starting point: my MASc

I measured micro-turbojet performance and emissions [6, 7, 9], and the data supported assessment of collaborators’ engine-performance model coupled to a chemical reactor network [8].

What the reduced model missed
  • NOₓ overpredicted by about 1.8× despite outlet temperature, CO₂ and CO within 4.5%
  • Particle predictions differed from measurement
  • Both consistent with limits in representing local temperature, mixing, reaction history and residence time
The research gap

How fuel, combustor and turbine studies fit together

Each of these three areas answers part of the question. Fewer studies follow one fuel effect from the combustor all the way to the turbine inlet. That link is what I want to work on.

Fuel studies

Fuel → engine output

How fuel choice changes engine performance or exhaust emissions.

[3, 4, 6, 7, 9]

Combustor studies

Local soot physics

Controlled combustors that resolve local soot formation and oxidation.

[5]

Turbine studies

Prescribed inlet → turbine response

Imposed inlet profiles or combustor-like turbulence to assess loss, loading and heat transfer.

[10–13]

Open question

How fuel-dependent combustion, local fuel–air ratio, residence time and combustor flow jointly govern pollutant formation and shape the non-uniform temperature, pressure, composition and flow entering the turbine.

+47%
total-loss increase from combustor turbulence, measured in a cascade experiment [11]
−1.3%
estimated stage-efficiency reduction for an engine-representative turbine [11]
Physical picture

Where fuel effects enter, and where they might persist

InputFuel propertiescomposition, volatility, aromatics
SprayEvaporation & mixinglocal fuel–air ratio
ReactionHeat releasefinite-rate chemistry
OutputsPollutantsNOₓ, CO, soot / nvPM
Combustor-exit statetemperature, pressure, composition, turbulence
DownstreamTurbineloss, heat transfer, cooling, efficiency
Hypothesis

Geometry sets the large-scale flow, and fuel sets the chemistry. How strongly does the fuel reach back into the flow?

The first part is well established. The open part is the size of the fuel effect on the flow itself. At fixed geometry, matched inlet conditions and matched heat input, I want to measure that effect and find out whether it is large or small.

Weak coupling: small effect on the flow

Emissions or temperature change with fuel, but measured pressure profiles and model-supported flow structure do not.

\(\Delta\mathrm{EI}\neq0\)\(\Delta T\neq0\)\(\Delta p(r,\theta)\approx0\)
Strong coupling: large effect on the flow

Reproducible pressure-profile changes, plus model-supported changes in recirculation, dilution mixing or turbulence, beyond experimental and modelling uncertainty.

\(\Delta\mathrm{EI}\neq0\)\(\Delta p(r,\theta)\neq0\)\(\Delta\mathbf{u}\neq0\)

If the differences stay inside the uncertainty, the question stays open. To link a result to one fuel property, such as volatility or aromatic content, I need more than two fuels.

What the size of the effect would change

Each step below is a standard relation. Together they show how one change in the fuel can travel through the whole combustor and into the turbine.

\[\text{fuel properties}\;\longrightarrow\;\dot q(x)\;\longrightarrow\;T_g(x),\,\rho_g(x)\;\longrightarrow\;J\;\longrightarrow\;T_4(r,\theta)\;\longrightarrow\;\mathrm{PF},\;\eta_c,\;\mathrm{EI}\]

Heat release rate \(\dot q\) along the combustor axis \(x\), hot-gas temperature \(T_g\) and density \(\rho_g\), jet-to-crossflow momentum-flux ratio \(J\), exit temperature field \(T_4\), pattern factor, combustion efficiency \(\eta_c\) and emission indices.

Show the mathHeat release, mixing, performance and pollutants, link by link

Heat release rate. Volatility decides where the fuel evaporates, and so where it can burn. For a single droplet, the classical \(d^2\)-law gives the evaporation time [18]:

\[d^{2}(t)=d_0^{2}-Kt,\qquad t_{\mathrm{evap}}=\frac{d_0^{2}}{K},\qquad K=\frac{8\,k_g}{\rho_l\,c_{p,g}}\ln(1+B),\qquad B=\frac{c_{p,g}\,(T_\infty-T_s)}{h_{fg}}\]

Initial diameter \(d_0\), gas conductivity \(k_g\) and specific heat \(c_{p,g}\), liquid density \(\rho_l\), and the Spalding transfer number \(B\) with gas temperature \(T_\infty\), droplet surface temperature \(T_s\) and latent heat \(h_{fg}\). A less volatile fuel has a lower \(B\). It evaporates more slowly and releases its heat further downstream. Viscosity and surface tension set \(d_0\) through atomization, and \(t_{\mathrm{evap}}\) grows with \(d_0^2\).

Mixing. The dilution jets enter a crossflow whose density is set by the heat released upstream. For the same jets, at a fixed crossflow mass flux \(G_g=\rho_g U_g\) and pressure \(p\), the momentum-flux ratio becomes:

\[J=\frac{\rho_j U_j^{2}}{\rho_g U_g^{2}}=\frac{\rho_j U_j^{2}\,\rho_g}{G_g^{2}}=\frac{\rho_j U_j^{2}\,p}{G_g^{2}\,R\,T_g}\;\propto\;\frac{1}{T_g},\qquad \frac{y}{d_j}\propto\sqrt{J}\]

Jet \(j\) and hot crossflow \(g\). Jet penetration \(y\) scales with the jet diameter \(d_j\) and with \(\sqrt{J}\) [17]. A hotter crossflow lowers \(J\), so the jets penetrate less and the mixing pattern moves. This is the route by which chemistry could feed back on large-scale mixing.

Engine performance. The exit temperature field goes into the turbine. Rearranging the pattern factor from step 01 of the plan gives the peak temperature at the turbine inlet. Unburned CO and hydrocarbons set the combustion efficiency [19]:

\[T_{\max}=\bar T_4+\mathrm{PF}\,\big(\bar T_4-\bar T_3\big),\qquad 1-\eta_c\approx\frac{\mathrm{EI}_{\mathrm{CO}}\,\mathrm{LHV}_{\mathrm{CO}}+\mathrm{EI}_{\mathrm{UHC}}\,\mathrm{LHV}_{\mathrm{UHC}}}{1000\,\mathrm{LHV}_f}\]

EI in g per kg of fuel, \(\mathrm{LHV}_{\mathrm{CO}}\approx10.1\) MJ/kg. At the same mean exit temperature, worse mixing raises PF and puts a hotter streak on the first turbine vanes. Illustrative values: with \(\bar T_3=800\) K and \(\bar T_4=1600\) K, PF from 0.20 to 0.25 lifts \(T_{\max}\) from 1760 K to 1800 K.

Pollutants. Thermal NO has a very high activation temperature, so small changes in peak temperature matter. CO burns out only if the gas stays hot long enough [18]:

\[\frac{\mathrm d[\mathrm{NO}]}{\mathrm dt}\approx2k_1[\mathrm O][\mathrm N_2],\qquad k_1\propto\exp\!\left(-\frac{38\,370\ \mathrm K}{T}\right),\qquad \mathrm{Da}_{\mathrm{CO}}=\frac{\tau_{\mathrm{res}}}{\tau_{\mathrm{CO}}}\]

Initial-rate form of the Zeldovich mechanism, valid while NO is far below equilibrium. The NO formed grows with this rate times the residence time at high temperature. Illustrative values: from 2000 K to 2050 K, \(k_1\) alone rises by about 1.6×. CO burns out where the Damköhler number \(\mathrm{Da}_{\mathrm{CO}}\gg1\). Dilution air that arrives too early quenches the gas and freezes CO.

Soot forms in fuel-rich pockets and burns out in hot gas with oxygen. Both depend on where the mixing puts the fuel and the air. Mixing and residence time appear in every step, so a change in the flow changes all three pollutants.

A large effect would mean fuel studies cannot treat the combustor flow as fixed. A small effect would mean the flow can be taken from one fuel and reused, which makes modeling much cheaper.

Weak versus strong coupling between fuel effects and combustor flow Geometry & operation liner, swirler, inlet state Fuel properties composition, volatility Large-scale flow recirculation, dilution-jet penetration, mixing, turbulence Chemistry evaporation, reactivity, heat release, composition, pollutants Combustor-exit state T(r,θ), p(r,θ), composition, turbulence → turbine heat release changes ρ, ρu² → strong coupling if detected
Weak coupling: geometry sets the flow, fuel sets the chemistry, and the two meet only at the exit. Strong coupling: heat-release-driven density and momentum changes reach back into mixing and turbulence (dashed).
Research plan

Measure, explain, represent

01
Measure

Fuel-dependent combustor development

  • Limited fuel and operating-condition matrix on an instrumented combustor or engine
  • Pressure, temperature, fuel flow, gaseous and particle emissions, performance
  • Sparse thermocouples and static taps through reaction, dilution, mixing and exit regions
  • Exit temperatures → pattern factor; calibration, repeatability and a full uncertainty budget
\[\mathrm{PF}=\frac{T_{\max}-\bar T_{4}}{\bar T_{4}-\bar T_{3}}\]

Peak-to-mean excess over the combustor temperature rise.

    02
    Explain

    Experimentally constrained RANS

    • Measured boundary conditions and profiles constrain the model
    • Estimate velocity, recirculation, mixing, heat release, composition, residence time and turbulence
    • Withheld spatial measurements test the explanation
    • Computed fields are never treated as measurements
    03
    Represent

    Which exit information matters

    • Test progressively simpler turbine-inlet descriptions: full profiles → means + non-uniformity → means
    • CRNs as a reduced starting point [16]
    • Independent operating conditions check whether simpler descriptions keep the supported fuel effects
    Simplified combustor section: fuel injector and swirler, recirculation zone, primary and dilution jets, wall thermocouples, static-pressure taps and an exit rake at plane 4
    Step 1 in one picture, on a simplified combustor (not to scale): sparse thermocouples and static taps through the reaction, dilution, downstream-mixing and exit regions, subject to geometry, sensor survivability and probe disturbance. The exit rake at plane 4 feeds the pattern factor and the turbine-inlet descriptions of step 3.
    Inlet-description ladder
    Measured T & p profiles + modelled direction, composition, turbulenceMeans + non-uniformityMeans only

    Step 3 stops at the simplest rung that preserves measured exit-profile shapes and fuel-dependent differences within uncertainty.

    Show the mathUncertainty, the RANS equations and the turbine-inlet descriptions
    The math behind each step
    01 Measureoperating point and uncertainty
    \[\begin{gathered}\phi=\frac{\dot m_f/\dot m_a}{\left(\dot m_f/\dot m_a\right)_{\mathrm{st}}},\\[6pt]\mathrm{PF}=\frac{T_{\max}-\bar T_4}{\bar T_4-\bar T_3},\\[6pt]u_c(\bar x)=\sqrt{b^{2}+\frac{s^{2}}{n}},\quad U=k\,u_c\end{gathered}\]

    Global equivalence ratio, pattern factor, and a combined standard uncertainty from the systematic part \(b\) and the scatter \(s\) of \(n\) repeats, expanded with coverage factor \(k\approx2\).

    02 Explainmean flow with modelled turbulence
    \[\begin{aligned}\frac{\partial(\bar\rho\tilde u_j)}{\partial x_j}&=0\\[6pt]\frac{\partial(\bar\rho\tilde u_i\tilde u_j)}{\partial x_j}&=-\frac{\partial\bar p}{\partial x_i}+\frac{\partial}{\partial x_j}\Big(\bar\tau_{ij}-\bar\rho\,\widetilde{u_i''u_j''}\Big)\\[6pt]\frac{\partial(\bar\rho\tilde u_j\tilde Y_k)}{\partial x_j}&=\frac{\partial}{\partial x_j}\Big(\bar\rho D_t\frac{\partial\tilde Y_k}{\partial x_j}\Big)+\bar{\dot\omega}_k\end{aligned}\]

    Favre-averaged (density-weighted) RANS balances for mass, momentum and species \(k\).

    \[\begin{gathered}-\bar\rho\,\widetilde{u_i''u_j''}=\mu_t\Big(\frac{\partial\tilde u_i}{\partial x_j}+\frac{\partial\tilde u_j}{\partial x_i}-\tfrac23\delta_{ij}\frac{\partial\tilde u_k}{\partial x_k}\Big)-\tfrac23\bar\rho\tilde k\,\delta_{ij},\\[6pt]\dot q=-\sum_k h_k\,\bar{\dot\omega}_k,\\[6pt]\tau_{\mathrm{res}}=\frac{\bar\rho\,V}{\dot m}\end{gathered}\]

    Boussinesq closure of the Reynolds stresses, heat release from the species sources, and residence time of a region of volume \(V\) carrying mass flow \(\dot m\). Withheld spatial measurements decide whether an explanation holds.

    03 Representwhat the turbine actually sees
    \[\begin{gathered}\bar T_4=\frac{\int\rho u\,T\,\mathrm{d}A}{\int\rho u\,\mathrm{d}A},\\[6pt]\sigma_T=\sqrt{\frac{\int\rho u\,\big(T-\bar T_4\big)^{2}\,\mathrm{d}A}{\int\rho u\,\mathrm{d}A}},\\[6pt]Tu=\frac{\sqrt{2k/3}}{|\tilde{\mathbf u}|},\\[6pt]\ell_t=C_\mu^{3/4}\frac{k^{3/2}}{\varepsilon}\end{gathered}\]

    Mass-weighted mean and non-uniformity of the exit temperature, turbulence intensity, and the model-derived turbulence length scale (\(C_\mu=0.09\)).

    \[\mathcal{D}_0=\{T(r,\theta),\,p(r,\theta),\,\boldsymbol\alpha,\,Y_k,\,Tu,\,\ell_t\}\;\longrightarrow\;\mathcal{D}_1=\{\bar T_4,\,\bar p_4,\,\mathrm{PF},\,\sigma_T\}\;\longrightarrow\;\mathcal{D}_2=\{\bar T_4,\,\bar p_4\}\]

    The inlet-description ladder, with flow angles \(\boldsymbol\alpha\). Each description is computed from the one before it. Step 3 keeps the simplest description that reproduces measured exit-profile shapes and the supported fuel-dependent differences within uncertainty.

    Considerations

    Considerations for this direction

    I have not done this work yet. These notes come from a lot of reading and thinking. It is the direction I most want to explore, and I would need guidance to get it right. Items marked Open question are the ones I am least sure about.

    1 · What the turbine would see, and how it could be measured

    QuantityWhy the turbine caresHow it could be measured
    Exit temperature field, pattern factorHot streaks set the vane metal temperatureTraversing thermocouple rake
    Turbulence intensity \(Tu\), integral length scale \(\ell\)They drive the extra loss of combustor turbulence [11, 12]Time-resolved PIV and LDV, which need optical access [20, 21]
    Swirl and flow angle Open questionVane incidence and loadingStereo PIV, or a cooled multi-hole probe where it fits
    Where heat is releasedWhere a fuel effect enters the flowOH* chemiluminescence, OH-PLIF [23]
    Soot inside the flameRadiation to the linerLaser-induced incandescence [22]

    2 · Is the fuel effect large enough to matter?

    These are qualitative links, using the relations in the panel above and our own fuel tests.

    • Jet penetration. A fuel that changes the gas temperature at the dilution jets changes how far the jets reach into the flow. Our FT-SPK test changed the combustor-exit temperature only slightly [9], so this route looks weak on its own.
    • Exit profile. If a fuel shifts where heat is released, the pattern factor can change. Even a small change moves the hottest temperature that the first vanes see, so this route could matter.
    • Radiation. Fuels with fewer aromatics form less soot, and less soot radiates less heat to the liner [19]. We measured much less soot in the exhaust with FT-SPK [9]. But exhaust soot is not the soot inside the flame, so the size of this effect is unknown.
    • Turbulence. Exit turbulence is set mostly by geometry, swirl and the jets, though heat release can raise it [21]. I cannot yet say whether a fuel change would move it.

    My working guess is that the exit temperature profile and radiation are stronger fuel levers than turbulence. That is a hypothesis.

    3 · Possible platforms Open question

    • A micro-turbojet. This is an option I know well: real hardware and clean back-to-back fuel tests. But it has no optical access, a combustor annulus about 6 mm deep, and vaporizer sticks that may hide spray-related fuel effects.
    • An optical RQL combustor simulator. Windows would allow PIV, LII and chemiluminescence, and an airblast atomizer would bring spray effects back. How results transfer from atmospheric to engine pressure is unclear to me.
    • A combustor simulator upstream of a turbine cascade. This would test the turbine response directly [10, 11]. It needs an existing facility, or we could build one. That seems like fun work.

    Which platform, or which combination, is the right one is the first thing I would want to work out with a supervisor.

    Open questions

    • how to measure turbulence and flow angle in a hot, sooting exit flow, with reliable seeding and clean windows
    • how to scale from an atmospheric rig to engine pressure, where soot and radiation are stronger
    • how to link a measured exit state to vane loss and heat transfer [12]
    • whether RANS is good enough to explain the measurements, or LES is needed
    • which fuels, beyond two, would separate one fuel property from the others.
    Expected outcomes

    Linking pollutant formation and the non-uniform turbine-inlet flow through their shared fuel-dependent physics could give a baseline for passing experimentally constrained flow information between combustor and turbine models.

    If mixing leaves no detectable fuel-dependent spatial differences, shared temperature or pressure profiles may represent the exit state without being uniform. If it does, the work identifies which spatial quantities to retain, since they may affect turbine loss and heat transfer [11–13].

    Scope

    The baseline could support later high-fidelity simulation [14, 15], turbine-response calculations, and larger-engine comparisons with collaborator models and data; none is required for the PhD.

    Low-volume tests support screening and model development. Differences in pressure ratio, geometry, cooling, efficiency and residence time rule out assuming scale equivalence.

    References

    Bibliography

    1. [1]S. Blakey, L. Rye, and C. W. Wilson, “Aviation gas turbine alternative fuels: A review,” Proceedings of the Combustion Institute, vol. 33, no. 2, pp. 2863–2885, 2011. doi:10.1016/j.proci.2010.09.011
    2. [2]E. Corporan, T. Edwards, L. Shafer, et al., “Chemical, thermal stability, seal swell, and emissions studies of alternative jet fuels,” Energy & Fuels, vol. 25, no. 3, pp. 955–966, 2011. doi:10.1021/ef101520v
    3. [3]R. H. Moore, K. L. Thornhill, B. Weinzierl, et al., “Biofuel blending reduces particle emissions from aircraft engines at cruise conditions,” Nature, vol. 543, pp. 411–415, 2017. doi:10.1038/nature21420
    4. [4]L. Durdina, B. T. Brem, M. Elser, D. Schönenberger, F. Siegerist, and J. G. Anet, “Reduction of nonvolatile particulate matter emissions of a commercial turbofan engine at the ground level from the use of a sustainable aviation fuel blend,” Environmental Science & Technology, vol. 55, no. 21, pp. 14576–14585, 2021. doi:10.1021/acs.est.1c04744
    5. [5]M. Stöhr, K. P. Geigle, R. Hadef, I. Boxx, C. D. Carter, M. Grader, and P. Gerlinger, “Time-resolved study of transient soot formation in an aero-engine model combustor at elevated pressure,” Proceedings of the Combustion Institute, vol. 37, no. 4, pp. 5421–5428, 2019. doi:10.1016/j.proci.2018.05.122
    6. [6]S. Sharma, D. Lieder, S. Saini, A. Kumar, P. Kirchen, J. Goeing, and S. Rogak, “A micro-turbojet test rig to analyze the impact of jet fuels on emissions and performance,” in Proc. CSME International Congress / 33rd Annual Conf. CFD Society of Canada, Vancouver, BC, May 24–27, 2026, Paper 361. My work
    7. [7]S. Sharma, D. Lieder, N. Sapkota, H. Lintusaari, A. Kumar, P. Kirchen, J. Goeing, and S. Rogak, “Experimental and model characterization of jet fuel emissions in a micro-turbojet engine,” presented at the Combustion Institute–Canadian Section Spring Technical Meeting, Concordia University, Montreal, May 11–14, 2026 (non-archival). My work
    8. [8]D. Lieder, S. Sharma, P. Kirchen, S. Rogak, and J. Goeing, “Development of a validated chemical reactor network model for non-CO2 emissions in RQL aircraft combustors,” Journal of Engineering for Gas Turbines and Power, 2026, GTP-26-1516, accepted June 19, 2026. My work
    9. [9]D. Lieder, S. Sharma, H. Lintusaari, et al., “Measurement and prediction of particle emissions from a micro turbojet using conventional and alternative jet fuel,” in Proc. Global Power and Propulsion Society, Montreal, Sep. 9–11, 2026, GPPS-TC-2026-102. doi:10.33737/gpps26-tc-102 My work
    10. [10]M. Folk and R. J. Miller, “Design and test of a combustor simulator,” in AIAA Propulsion and Energy 2019 Forum, 2019. doi:10.2514/6.2019-4256
    11. [11]M. B. Folk, R. J. Miller, and J. D. Coull, “The impact of combustor turbulence on turbine loss mechanisms,” Journal of Turbomachinery, vol. 142, no. 9, 091009, 2020. doi:10.1115/1.4047615
    12. [12]K. Gakhar, M. Folk, E. M. Greitzer, and C. S. Tan, “A framework for estimating effects of combustor turbulence on turbine profile loss generation,” Journal of Turbomachinery, vol. 145, no. 3, 031018, 2023. doi:10.1115/1.4056407
    13. [13]M. D. Barringer, K. A. Thole, and M. D. Polanka, “Effects of combustor exit profiles on vane aerodynamic loading and heat transfer in a high pressure turbine,” Journal of Turbomachinery, vol. 131, no. 2, 021008, 2009. doi:10.1115/1.2950051
    14. [14]C. Koupper, L. Gicquel, F. Duchaine, and G. Bonneau, “Advanced combustor exit plane temperature diagnostics based on large eddy simulations: Going further than the radial temperature distribution factor,” Flow, Turbulence and Combustion, vol. 95, no. 1, pp. 79–96, 2015. doi:10.1007/s10494-015-9607-3
    15. [15]P. Koob, F. Ferraro, H. Nicolai, R. Eggels, M. Staufer, and C. Hasse, “Large eddy simulation of soot formation in a real aero-engine combustor using tabulated chemistry and a quadrature-based method of moments,” Journal of Engineering for Gas Turbines and Power, vol. 146, no. 1, 011015, 2024. doi:10.1115/1.4063376
    16. [16]H. Khodayari, F. Ommi, and Z. Saboohi, “A review on the applications of the chemical reactor network approach on the prediction of pollutant emissions,” Aircraft Engineering and Aerospace Technology, vol. 92, no. 4, pp. 551–570, 2020. doi:10.1108/AEAT-08-2019-0178
    17. [17]J. D. Holdeman, “Mixing of multiple jets with a confined subsonic crossflow,” Progress in Energy and Combustion Science, vol. 19, no. 1, pp. 31–70, 1993. doi:10.1016/0360-1285(93)90021-6
    18. [18]S. R. Turns, An Introduction to Combustion: Concepts and Applications, 3rd ed. New York: McGraw-Hill, 2012.
    19. [19]A. H. Lefebvre and D. R. Ballal, Gas Turbine Combustion: Alternative Fuels and Emissions, 3rd ed. Boca Raton, FL: CRC Press, 2010.
    20. [20]C. Willert, M. Schroll, J. Heinze and T. Soworka, “High-speed PIV at the exit of a lean-burn combustion chamber operated at elevated pressure,” in Proc. 13th Int. Symp. Particle Image Velocimetry (ISPIV 2019), Munich, 2019. elib.dlr.de/128097
    21. [21]R. Orbay, K.-J. Nogenmyr, J. Klingmann and X.-S. Bai, “Swirling turbulent flows in a combustion chamber with and without heat release,” Fuel, vol. 104, pp. 133–146, 2013.
    22. [22]K. P. Geigle, M. Köhler, W. O’Loughlin and W. Meier, “Investigation of soot formation in pressurized swirl flames by laser measurements of temperature, flame structures and soot concentrations,” Proceedings of the Combustion Institute, vol. 35, pp. 3373–3380, 2015.
    23. [23]M. Lauer, “Determination of the heat release distribution in turbulent flames by chemiluminescence imaging,” PhD thesis, Technische Universität München, 2011.