A computationally efficient combustion trajectory prediction model developed for real-time diesel combustion control

The heterogeneous nature of diesel combustion adds many complexities that make understanding the combustion process difficult. Many researchers have made great efforts in diagnostics, prediction, and control capabilities. In this work, a computationally efficient thermodynamic-based model (15 ms on 2010 dual core processor) has been created that predicts the combustion trajectory (path through the ϕ–T plane) with the goal of bridging the gap between typical off-line engine prediction simulations and on-line real-time engine control strategies. The ϕ–T plane is often used to help illustrate the soot and NOₓ formation behavior during diesel combustion. The experimental engine operating conditions shown illustrate how exhaust gas recirculation influences the combustion trajectory at different timings—that is, showing the typical soot–NOₓ trade-off for diesel engines and the defeat of this trade-off when low-temperature combustion is obtained. The major insight gained is that the low-temperature combustion trajectory looks similar to a conventional one with just subtle differences that keep it from moving into the soot formation region. Additionally, the traditional conceptual explanations for diesel combustion are explored relative to how they are illustrated by the combustion trajectory, especially the transition from premixed to mixing-controlled combustion. Understanding that behavior in this context aids in explaining the different observations for the low-temperature combustion modes. The fact that these observations are made using this simplified modeling approach is promising for future use of this type of thermodynamic-based models in real-time engine control.

Language

  • English

Media Info

  • Media Type: Web
  • Features: References;
  • Pagination: pp 246-258
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 01717387
  • Record Type: Publication
  • Files: TRIS
  • Created Date: May 24 2019 4:53PM