Controlled Auto Ignition - Lean Mixture, Why? 1

[pizzaice]

Hi everybody,
I am starting to investigate controlled auto-ignition, also known as HCCI.
All related literature claims that CAI operation can only be achieved with high lambda. I have understood that you need a good homogenious mixture, so ignition is distributed evenly, but why is this not possible with more fuel (also distributed evenly)? Unfortunately I can not find an explanation why this is the case.

Could anybody help me?


Thanks,

Chris

 

[BrianPetersen]

I don't have personal involvement with this (few do) but from talking to another engineer who knows far more about this than I do, there are a couple of problems with HCCI operation at higher engine load.

One problem is that the rate of pressure rise becomes explosive. A tough problem with HCCI is controlling the rate of heat release. They seem to be doing this by using high EGR and "not quite homogeneous" operation, but that only works at lighter engine load. If you have high enough cylinder pressure to get autoignition with minimal EGR at higher engine load, the combustion starts to resemble a massive detonation in the cylinder.

True HCCI with early fuel injection is a nightmare to control. The temperature at end of compression affects the moment of ignition, but the temperature at end of compression is different if the outside temp is -30 C compared to if it is +30 C. At light engine load you can play with the EGR (and whether you cool it or not) to affect this. Near full load, what do you do to control the ignition timing?

I don't know if "true" HCCI will ever see production except maybe in very narrow speed/load regimes on engines that are otherwise direct-injected gasoline or diesel. Production diesel engines seem to be starting to implement lessons learned from this research, but they seem to be using "more-homogeneous but not truly homogeneous" mixture preparation. The new low-emission diesels seem to be using lower compression ratios (i.e. longer ignition delay), and really high injection pressure (to mix air and fuel faster), and EGR systems in which the EGR can be cooled or not, and some of them are using a low-pressure EGR drawn from after the particulate filter. It seems that some of them are using *late* injection timing and relying on the ignition delay to serve as the pre-mixing period.

 

[Lou Scannon]

HCCI amounts to detonation of the entire charge all at once. There's a couple of reasons you don't want this to happen with a stoich mixture (assuming air is the oxidant).
First is, NOx reduction is usually a driver for HCCI combustion solutions. This can only be achieved with an ultra lean mixture (due to the cooling effect of charge dilution)
Second is, detonation is a violent event. Existing commercial engine structures are incapable of withstanding ongoing detonation of a stoichiometric mixture. Today's HCCI combustion recipies are lean enough to be within existing engines' structural and material limits.

 

[SMOKEY44211]

I've been doing some research on HCCI combustion process in the past few years on engines that have been converted that were originally designed for traditional compression ignition. Potentially it offers better BSFC #s and reduction in NOx formation. The near instant release of energy is difficult to contain without damage to the internal engine components. The mechanical limitations prevent being able to fuel the engine to the power level that would consum all the oxygen molecules available for bonding thus the NOx by product. Lowering compression ratio allows more fuel by lowering the post ignition combustion pressure at the expense of lower BSFC due to the lower expansion ratio. There is a slight plus with lower compression ratio in that the combustion occurs closer to TDC reducing the negative work that is produced by earlier ignition and the peak temperature is lower. When running low fuel amounts (lean) the combustion chamber temps. are not sufficient to reach the threshold of large amounts of NOx formation. Introducing inerts such as EGR or water simply displaces some of the oxygen and nitrogen reducing NOx. I think we will begin to see some offerings in engines designed for stationary applications. There is a company that offers conversion kits of model airplane engines to run on diesel fuel using the HCCI combustion process.(Davis Diesel Development) They've been around at least 25 yrs. that I know of. It works very well. When you start scaling up the displacement is when the weak links start to manifest.---------Phil

 

[tbuelna]

pizzaice,

HCCI or CAI combustion is essentially "controlled" detonation. A stoichiometric, homogenous intake charge (ie: an engine at full load) would auto-ignite well before TDC in a fixed CR piston engine (using a CR otherwise suitable for HCCI) causing a very steep pressure rise and high mechanical loads. A very lean mixture is less susceptible to auto-ignition, but unfortunately also produces less power. HCCI combustion chemical kinetics are incredibly complex, and as a result HCCI combustion is very difficult to control reliably.

The typical methods employed to control HCCI combustion are A/F ratio, charge temperature, EGR rate, variable valve timing and variable compression ratio. These are all expensive options for a production engine, but the benefits are reduced part throttle fuel consumption and significantly reduced NOx emissions.

No one has yet gotten an HCCI engine to function reliably outside of a laboratory environment. There are still huge issues keeping it out of the marketplace, such as making sure it works reliably under all operating conditions for the life of the vehicle, and reducing or eliminating its nasty NVH characteristics. Even though there are lots of auto companies, spending millions of R&D dollars to develop it, I don't see it making its way into the marketplace anytime soon. The development challenges are still too great.