Rate Shape Design for Gasoline-Like Fuels at High Injection Pressures Using One-Dimensional Hydraulic Models
Journal Article
04-15-02-0007
ISSN: 1946-3952, e-ISSN: 1946-3960
Sector:
Citation:
Tzanetakis, T., Desai, A., Voice, A., and Naber, J., "Rate Shape Design for Gasoline-Like Fuels at High Injection Pressures Using One-Dimensional Hydraulic Models," SAE Int. J. Fuels Lubr. 15(2):137-170, 2022, https://doi.org/10.4271/04-15-02-0007.
Language:
English
Abstract:
Recent research has demonstrated that gasoline compression ignition (GCI) can
improve the soot-oxides of nitrogen (NOx) trade-off of conventional diesel
engines due to the beneficial properties of light distillate fuels. In addition
to air handling and aftertreatment, fuel systems also require further
development to realize the potential efficiency and emissions benefits of GCI.
Injector one-dimensional (1-D) hydraulic modeling is an important design tool
used for this purpose. The current study is a continuation of prior work that
used computed physical fuel properties and hydraulic models to accurately
simulate high-pressure injection behavior relevant to GCI. With respect to fuel
characteristics for the model, physical properties were validated by direct
comparison to measurements at temperatures and pressures reaching 150°C and 2500
bar, respectively. Calibration of the injector model discharge coefficients for
gasoline-like fuel was automated with various multi-objective optimization
approaches coupled to a genetic search algorithm. However, Pareto optimization
showed the best closure with an experimental rate of injection (ROI) and total
injected quantity compared to other current and previous manual methods. The
validated model was then used to determine the injector specifications needed to
approach an idealized, slowly opening rate shape that could enable low-NOx
combustion. Initial parametric studies of key parameters affecting rate shape
showed that changing a combination of nozzle exit, control chamber (or servo)
outlet, and needle orifice diameters could produce the desired single injection
fueling profile. A transient targeting (TT) optimization technique coupled to a
genetic search algorithm was compared to a full-factorial design of experiments
(DoE) and showed that both approaches could reasonably achieve the target rate
shape. However, TT required a significantly reduced computational runtime. In
general, this study provides a robust methodology for accurately simulating
gasoline-like fuels in high-pressure injectors and demonstrates a conceptual
rate shape targeting process for GCI using 1-D hydraulic models. This tool could
potentially be integrated with predictive computational fluid dynamics (CFD)
models to achieve a simulation-led combustion system design process that
includes rate shaping as an additional avenue for optimization.