Basic Energy Sciencnes

Diesel Jet Spray Formation

TSTT Personnel: James Glimm (BNL/SUNY SB), Wonho Oh (BNL), Roman Samulyak (BNL), Myoung-Nyoun Kim (BNL), Andrea Marchese (BNL), Xiaolin Li (SUNY SB)

Combustion Personnel: Constantine Tzanos (ANL)

TSTT members at BNL and SUNY SB have been working with scientists at Argonne to create a simulation code that models spray formation in diesel jet break up. This is an important aspect of the problem as it provides input to spray combustion models and is critical for predictive modeling of diesel engine combustion. The overall goal of such an effort is the design of a nonpolluting, fuel-efficient engine.

Spray formation is a difficult problem for simulation due to

  • the geometric complexity and multiscale nature of the spray,
  • the stiff equation of state (EOS) for diesel fuel, and the need to model phase transitions, shock waves and other strong hydrodynamic transients,
  • the limited resolution of experimental diagnostics, and
  • the sensitive dependence of spray formation on nozzle geometry and other problem parameters.

To model this process, we are using the front tracking code, FronTier which has a high quality interface capability that enables multiscale resolution of complex geometries. For this simulation, we enhanced FronTier with an Equation of State (EOS) module that supports diesel fuel with cavity formation (phase transitions). Using the enhanced version of the code and the flow geometry shown in Figure 1, we performed an initial exploration of input parameters that have a sensitive influence on the jet and its breakup.

Figure 1

Figure 1 The flow geometry for the diesel jet spray formation simulation.

We found the following to be important:

  • The diesel EOS.
  • The boundary conditions inside the nozzle. The flow in the nozzle appears to be turbulent, and the boundary conditions (varying between slip and no slip) should be set according to grid resolution.
  • The rise time of the initial pressure transient (valve opening time).
  • The diesel viscosity.
  • The diesel meniscus at the nozzle outlet.
  • The inlet pressure.

Our simulations are compared to measurements performed on the light source at ANL. In contrast to conventional (optical or photographic data), the light source can penetrate the spray and give data on local densities within the spray. It has thus settled the question (negatively) of an intact liquid core to the jet. The light source data thus gives volume fraction data through out the jet, as a function of time, distance down stream from the jet nozzle and distance from the central axis of the jet. Data from conventional instruments is also available, such as the positions (and velocities) of the tip and tail of the jet. The primary diagnostics are jet tip and tail velocities, and the volume fraction or mass density at various time and space localized cross sections of the jet.

Our present success is semi quantitative. See Figure 2. Due to the importance of the turbulent flow in the very narrow nozzle, further progress will require automatic mesh refinement (AMR), currently being added to FronTier through a merge with the Overture code.

Figure 2

Figure 2 Partial breakup of the diesel jet after injection for a distance of 3.9 mm into the combustion chamber. Blue denotes diesel vapor, red is liquid, and green/yellow is a mixed phase region.