Reactivity-controlled compression ignition (RCCI) is a promising combustion
strategy for improving engine efficiency while reducing emissions. Within the
framework of the dtec.bw MORE project, this study investigates an innovative
homogeneous RCCI (hRCCI) combustion concept using port fuel injection of ethanol
combined with early direct injection of octanol in a modified single-cylinder
gasoline research engine. Although three-dimensional (3D) computational fluid
dynamics (CFD) has been increasingly applied to low-temperature combustion (LTC)
concepts, the use of partially scale-resolving approaches remains limited
because of their higher computational cost and the resulting challenges for
multi-cycle simulations in comparison to conventional Reynolds-averaged
Navier–Stokes (RANS) simulations.
To address this challenge, a CFD methodology in AVL FIRE M is developed that
combines partially averaged Navier–Stokes (PANS) turbulence modeling with port
and direct fuel injection, fuel-specific spray and evaporation behavior,
wall-film formation and evaporation, and chemical kinetics. A novel
reduced-species definition strategy is introduced to decrease the computational
effort associated with chemistry and enable the application of the PANS approach
over multiple engine cycles. After preconditioning cycles, nine consecutive
cycles are simulated and analyzed to characterize combustion variability.
The simulations reproduce the experimentally observed cylinder pressure and heat
release rate trends with good agreement. The predicted cycle-to-cycle variations
remain within the experimental minimum–maximum range and are tightly distributed
around the experimentally observed mean cylinder pressure evolution. The
predicted combustion duration agrees well with the experiments up to
approximately 0–85% mass fraction burned. Beyond 85% burn, significant
deviations occur, which are attributed to the increasing contribution of
diffusion-controlled combustion and effects that are not fully captured by the
present modeling framework.
The results demonstrate that the proposed PANS-based methodology, together with
the developed computational reduction strategy, enables computationally
tractable multi-cycle 3D simulations while retaining the capability to resolve
important flow–mixture–chemistry interactions. This study provides new insights
into the combustion characteristics of dual-fuel hRCCI operation, while also
demonstrating the influence of fuel injection parameters and identifying current
modeling limitations. The proposed framework therefore provides a basis for
further model development and combustion system optimization toward higher
efficiency and reduced emissions.
