Pressure Coupled Representative Interactive Linear Eddy Modeling for Internal Combustion Engine simulations

Doctoral thesis, 2025

The ongoing battle over the environmental ramifications of internal combustion engines is continuously intensifying. As the automotive industry gradually shifts towards adopting light- and medium-duty electric vehicles, the transition for heavy-duty transport solutions such as trucks and ships remains challenging. This is due to the inherent limitations of battery technology in these specific applications, such as low energy density, low range, insufficient power requirement, and inadequate charging infrastructure. This implies preserving  Internal Combustion Engines (ICE) as a valid technology for high-power applications. However, the environmental impact of operating ICEs by burning fossil fuels, such as diesel, is significant. Therefore, enhancing the combustion process within the ICE is essential. The reduction of pollutants emitted by the combustion engine is realized via operating the engine in nonstandard conditions such as Homogeneous Charge Compression Ignition  (HCCI) or Premixed Charge Compression Ignition (PCCI), which require in-depth Computational Fluid Dynamics (CFD) simulations investigations. These simulations require specialized combustion modeling techniques capable of accurately representing finite-rate chemistry and mixed-mode combustion.

The Linear Eddy Model (LEM) was utilized in this work to predict the turbulent combustion process. LEM is distinctive in its ability to simulate turbulent combustion on a 1D line in physical space, thereby capturing all scales down to the Kolmogorov length. The LEM advanced three processes: i) Turbulence in 1D physical space via stochastic rearrangement, i.e., Triplet maps. ii) Molecular diffusion and heat conduction via advancing 1D zero-Mach number equations of species mass fractions and energy in physical space. iii) Chemical advancement in each LEM cell. This approach allows for a detailed simulation of unsteady turbulent combustion processes occurring in ICEs on the LEM.

This research led to the development of a novel stand-alone LEM model for engine combustion simulations called Spherical Stand Alone LEM (SSALEM). It is based on coupling the spherical formulation of LEM to precalculated CFD quantities using a pressure constraint. The pressure coupling enabled the direct capturing of heat effects such as the latent heat of evaporation and wall heat losses on the LEM with no modeling, as these effects are an intrinsic part of the enforced CFD pressure trace. SSALEM simulated an engine with a pressure coupling constraint based on a simple slider-crank model, where initial investigations were realized. Later, LEM was coupled to a CFD simulation using the same pressure constraint in the Representative Interactive LEM (RILEM) configuration. However, it was observed that advancing one line was not sufficient to adequately resolve the turbulent scalar statistics. For that, several LEM lines were advanced in parallel with different turbulence rearrangements coupled to one CFD solver, i.e., Multiple RILEM. MRILEM was utilized to simulate the combustion process for a heavy-duty truck engine for part- and full-load scenarios. In this investigation, the progress variable was defined based on O2, and a novel PDF for the progress variable, namely a piece-wise step function, was utilized. MRILEM demonstrated a strong agreement with experimental data for the pressure trace and heat release both for part- and full-load cases. Afterward, a duct fuel injection was simulated with MRILEM, where the duct was implemented physically on the line(s). In this study, two different turbulence regions were implemented on the line, to simulate the high mixing rate inside the duct. MRILEM showcased a good correlation with experimental findings and results of other models when comparing results of heat release, lift-off lengths, and ignition delay. In addition, soot was also quantified on the CFD based on mapped LEM mass fractions. Finally, MRILEM simulated another heavy-duty engine case with a low compression ratio, where MRILEM was initialized with the solution of unsteady homogeneous reactors. This investigation analyzed the effect of varying key parameters for the combustion progress: Progress variable definition O2 and formation enthalpy h298 and progress variable PDF (Step function and presumed beta. This study displays and analyses the configurations that yielded the best matches with experiments. In addition, this work also introduced a tabulated RILEM method and analyzed the effect of advancing a TRILEM compared to conventional MRILEM on the combustion process and the simulation time.