A comprehensive numerical framework has been established to simulate reacting §ows under conditions typically encountered in rocket combustion chambers. The model implemented into the commercial CFD Code ANSYS CFX includes appropriate real gas relations based on the volume-corrected PengRobinson (PR) equation of state (EOS) for the §ow ¦eld and a real gas extension of the laminar §amelet combustion model. The results indicate that the real gas relations have a considerably larger impact on the §ow ¦eld than on the detailed §ame structure. Generally, a realistic §ame shape could be achieved for the real gas approach compared to experimental data from the Mascotte test rig V03 operated at ONERA when the di¨erential di¨usion processes were only considered within the §ame zone.

To account for heat losses near cooled walls an extension of the flamelet model is proposed based on an enthalpy defect parameter. A definition of the enthalpy defect and its transport equation is introduced. The inclusion of the enthalpy defect into the flamelet generation and the integration in terms of a probability density function for this parameter is discussed. The near wall extension is implemented into the OpenFOAM architecture and compared to ANSYS Fluent finite rate data for a testcase of a reacting laminar cooling film over a cooled flat plate. The near wall extension seems to improve the predicted heat flux compared to the original flamelet model, but profound validation was not possible due to a lack of suitable experimental or DNS data so far.

The demand for increased performance and lower weight of gas turbines gives rise to higher fuel-to-air ratios and a more compact design of the combustion chamber, thereby increasing the potential of fuel escaping unburnt from the combustor. Chemical reactions are likely to occur when the coolant air, used to protect the turbine blades, interacts with the unreacted fuel. Within this work, Reynolds-averaged Navier–Stokes (RANS) simulations of reacting cooling films exposed to high temperature fuel-rich exhaust gases are performed using the commercial computational fluid dynamics (CFD) code ansys fluent and validated against experimental results obtained at the Air Force Research Laboratory in Ohio. The results underline that the choice of the turbulence model has a significant impact on the evolution of the flow field and the mixing effectiveness. The flamelet as well as the equilibrium combustion model is able to predict an adequate distance of the reaction zone normal to the wall. Its thickness, however, is still much smaller and its onset too far upstream as compared to the experimental results. According to the present analysis, the flamelet combustion model applied along with k–ω shear stress transport (SST) or k–ε turbulence model turned out to be an appropriate choice in order to model near wall reacting flows with reasonable prospect of success.

The demand for increased performance and lower weight of gas turbines gives rise to higher fuel-to-air ratios and a more compact design of the combustion chamber, thereby increasing the potential of fuel escaping unburnt from the combustor. Chemical reactions are likely to occur when the coolant air, used to protect the turbine blades, interacts with the unreacted fuel. Within this work, RANS simulations of reacting cooling films exposed to high temperature fuel-rich exhaust gases are performed using the commercial CFD code ANSYS Fluent and validated against experimental results obtained at the Air Force Research Laboratory in Ohio. The results underline that the choice of the turbulence model has a significant impact on the evolution of the flow field and the mixing effectiveness. The flamelet as well as the equilibrium combustion model are able to predict an adequate distance of the reaction zone normal to the wall. Its thickness, however, is still much smaller and its onset too far upstream as compared to the experimental results. According to the present analysis the k-ω SST turbulence model applied along with the flamelet combustion model turned out to be an appropriate choice in order to model near wall reacting flows with reasonable prospect of success.

No abstract