Sandia researchers develop conceptual model for Low Temperature Combustion

16 May 2012

Conceptual model for conventional diesel combustion (CDC), left, and LTC combustion, right, for heavy-duty engines. Source: CRF. Click to enlarge.

Researchers at the Combustion Research Facility (CRF) at Sandia National Laboratories are proposing a conceptual model based on findings from their work for low temperature combustion (LTC). The LTC model is in review with the journal Progress in Energy and Combustion Science (PECS), with publication pending, according to Mark Musculus of CRF during his presentation at the DOE Vehicle Technologies Program Annual Merit Review in Washington this week.

LTC strategies—such as Premixed-Charge Compression-Ignition (PCCI), Homogeneous-Charge Compression-Ignition (HCCI) or Reactivity Controlled Compression Ignition (RCCI)—offer improved efficiency and very low emissions.

As such, noted Gurpreet Singh, Team Leader, Advanced Combustion Engine R&D Subprogram of the Vehicle Technologies Program, DOE is seeking to develop the knowledge base for LTC strategies and to carry those research results to products.

LTC strategies face a number of challenges, including combustion phasing; load range; heat release rate; transient control; an increase in HC and CO emissions, and new different species of HC; and fuel characteristics.

The long-term objective for this project has been to develop an improved understanding of in-cylinder process for low temperature combustion. This is a fundamental understanding that will help industry build cleaner , more efficient engines. One big [specific objective] for this year is to distill the observations that we have made from our optical diagnostics and modeling over several years recently into a conceptual model for low temperature combustion. This is something we have been working on for years, now we have taken the time to put it together in a coherent complete model.

We have, of course, a conceptual model for conventional, or high-temperature, combustion. This model, which is a cornerstone of understanding for conventional conditions, was based on multiple laser/imaging diagnostics over many years. It’s been quite valuable. At this point we have many years of optical engine research in low temperature combustion. That begs the question: can you put together a conceptual model like this one but for low temperature combustion. That’s just what we tried to do.

—Mark Musculus

Musculus provided a necessarily brief overview of the salient points of the model that will be published later this year:

• Often with LTC, injection is early into lower density gases. Spray penetration is faster, and with a longer liquid length. This is not unexpected, noted Musculus; what is unexpected, however, is that near the end of injection, the liquid length becomes shorter—i.e., it retreats. When the liquid length gets shorter, that implies that mixing has increased.

  This is really important for what we’re trying to do with low temperature combustion; we’re trying to increase mixing. And so here’s an indirect sign that something might be happening that is increasing mixing.

  —Mark Musculus

  The Sandia researchers used three modeling approaches—1D analytic, KIVA RANS, and Sandia LES models—to understand the observed behavior. The 1D model showed that when the upstream velocity is reduced as the injection rate is ramped down, it draws in more entrainment. Entrained gas slows velocity, driving more entrainment. Further analysis found that the end-of-injection (EOI) ramp-down causes large flow structures to separate rather than collide; the ambient fluid is entrained into gaps.

  This is giving us clues about what we might do to control the injection rate to generate the type of mixing desired.

  —Mark Musculus

• Moving to ignition, they used a laser diagnostic to fluoresce formaldehyde, an indication of first-stage ignition. In addition to the formaldehyde formed, there is a lot of unburned hydrocarbon and CO that persist until second-stage ignition. At that time, there is a spike in OH—OH appears when the fuel and the unburned hydrocarbons and the formaldehyde are consumed.

  Think of formaldehyde as a market of unburned fuel and think of OH as a marker of when that unburned fuel is consumed.

  —Mark Musculus

• Second-stage ignition occurs downstream where the equivalence ratio φ is approximately 1—i.e., near stoichiometric conditions. Soot and PAH are formed in downstream regions where φ>2; in addition, unburned hydrocarbons and potentially CO remain upstream late in the cycle, especially in the region near the injector.

  We can go back to the kinetics model from Lawrence Livermore to help us understand this. What they predict is that when we get to lean mixture—when mixtures are too well mixed—that the delay, or the dwell between first and second stage ignition is increased, and we expect that form and unburned fuel would persiste much longer in the cylinder. And that’s what we see.

  —Mark Musculus

Musculus said he expected the full paper to be published within several months.

Resources

• ACE001: Heavy-Duty Low-Temperature and Diesel Combustion & Heavy-Duty Combustion Modeling (Musculus, DOE 2012 Merit Review). [Presentations will be posted in several weeks at www.vehicles.energy.gov.]