Nowadays Liquid Rocket Engines (LRE) represent a widespread technology in launcher applications. However, the complex phenomena occurring within a rocket combustion chamber, are still far from being fully understood. These systems, due to performance requirements, generally operate at high chamber pressures, exceeding the critical point of typically employed propellants such as cryogenic liquid oxygen (LOx), liquid hydrogen (LH2) or the newly proposed methane (LCH4). The fluid dynamic and reactive behavior of such supercritical fluids is still under investigation. Moreover, in transcritical injection conditions, i.e. when pressure is above and temperature below the critical value, the LOx undergoes a pseudo-transition of state, referred to as pseudo-boiling, as it is heated from the trans- to the supercritical state. Associated to pseudo-boiling is a strongly non-ideal, real gas behavior and an abrupt volumetric expansion leading to drastic changes in flame morphology, which could affect combustion efficiency if not taken into account in the design process. The transcritical condition is systematically encountered in LRE, as propellants are stored in cryogenic state.
In this context, the present research project proposes the developement of a numerical software for the investigation of high pressure turbulent flames in LRE-like conditions, by means of both RANS and LES techniques. Concurrently, the development of a Direct Numerical Simulation (DNS) infrastructure is also proposed in order to assemble a database of turbulent mixing and diffusive flame problems in super- and transcritical conditions, in order to gain fundamental physical insight and assess new mixing and combustion models for high fidelity RANS/LES simulations.
The innovative character of the project first relies on the numerical framework chosen for the proposed architecture: in the low-Mach number limit, in fact, flamelet based tabulation methods for turbulent non-premixed combustion are well-posed [1], and provide a properly filtered/averaged treatment of a real fluid equation of state (EoS) [2]. Moreover the proposed framework features the possibility to accommodate, in a computationally efficient manner, real gas thermodynamics and non-adiabatic combustion in the presence of wall heat transfer: the tabulation technique, indeed, first avoid the numerical stiffness induced by the integration of species mass fractions transport equations at run time; then, allows to use a detailed kinetic mechanism for the description of the combustion and mixing processes without affecting the run time solution. The latter, in particular, is essential for hydrocarbon flames modelization. In the ensemble the proposed architecture results in a time and computational resources efficient platform.
Secondly, in flamelet-based solvers, the Favre averaged mean values which are needed by the CFD solver are obtained through multi-variate presumed probability density functions (p.d.f.) which represent the turbulence-chemistry interactions at a sub grid scale (SGS) level. In particular the joint p.d.f. used to include turbulence effects in the laminar flame structures (flamelet), is usually approximated using Bayes theorem and assuming statistical independence, as the product of single-variate p.d.f. This assumption is widespread among the most common (commercial and in-house developed) flamelet-based solvers, altough the assumption of statistical independence is a rough semplification. In this context the proposed DNS campaign could provide a reference simulations database from which exctracting these statistical informations , nowadays missing in the literature.
Finally, the results obtained with the proposed infrastructure could provide physical insights on trans- and supercritical combustion in LRE and shed light on the wall heat transfer mechanisms inside rocket combustion chambers, which are still far from being fully understood.
Refernces
[1] P. E. Lapenna, R. Lamioni, P. P. Ciottoli, F. Creta, ¿Low-mach number simulations of transcritical flows¿, AIAA
paper (2018) 2018-0346.
[2] P. E. Lapenna, G.Indelicato, R. Lamioni, F. Creta, ¿Modeling the equation of state using a flamelet approach in
LRE-like conditions¿, Acta Astr.