Enhanced Two-stage Ignition Delay Model Based on Molar Fraction of Fuel Components for SI Engine Simulation

Simulation based design and control optimization is widely used to assist the development of highly complex modern downsized turbocharged gasoline direct injection (GDI) engines. In such engines, knock phenomenon is a major constraint that limits performance and fuel economy enhancements. Thus, an accurate knock prediction model is critically important for virtual engine development process. In this paper, an enhanced ignition delay model is proposed for spark ignition (S)I combustion model based on previously developed empirical two-stage ignition delay model using fuel blends [1]. The ignition delay model provides a capability of predicting ignition delay of the end-gas zone for different fuel blends without additional calibration when fuel blending ratio changes. To adapt the ignition delay model to the SI combustion environment, the model is modified to have the sensitivity to the dilution effect by residual gas. Shock tube experimental data from the literature were collected and used to validate the dilution effect model. The ignition delay model developed in this study is implemented into a commercial simulation code (GT-Power®) as a user subroutine for a versatile simulation capability. Experimental data are taken from a 4-cylinder turbocharged GDI engine at various engine operating conditions. The experimental data are processed to extract the auto-ignition timings for the operation conditions in which knock is clearly observed. The ignition delay model integrated with the engine simulation is calibrated at reference cases and validated with various knocking and non-knocking cases.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References; Tables;
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 01726552
  • Record Type: Publication
  • Source Agency: SAE International
  • Report/Paper Numbers: 2018-01-0849
  • Files: TRIS, SAE
  • Created Date: Oct 8 2018 12:50PM