HCCI, fuel selection

[obanion]

In HCCI, the fuel is uniformly mixed with air, usually in excess of 3:1 over stoich, and compressed until it autoignites.

The fuel must match the compression ratio, or more exactly, the in cylinder temperature at a given point during compression. You want the fuel to ignite before TDC, but not too far before.

So it relates to the autoignition temperature of the fuel to be used. Basically my question is this, the ideal autoignition temp of the fuel is the same as the in cylinder temp at what degree BTDC? Should the fuels autoignition temperature coorespond to 40DBTDC, 30, 20, 10?

Based on a few reports I have seen, they a cool flame onset at about 15BTDC. But, does this first start of reaction coincide with the autoignition temperature of the fuel?

 

[crysta1c1ear]

From a naïve non-expert point of view, I would say the angle would depend on the RPM.

Why? Conditions for autoignition are likely to be similar regardless of RPM, the gas doesn't know its in an engine going at a particular RPM, it just reaches a certain temperature and pressure and goes bang. The combustion is then defined as detonation if the speed of the flame front is supersonic. I'd assume the speed of the flame front is therefor pretty constant and independent of the RPM.

The idea of igniting before TDC is so that the flame front can start to travel and that pressure will be building up so that the piston has more downward push on it during the power stroke. Since (I guess) the flame front will move faster with a homogeneous charge, I would expect the (auto)ignition advance timing to be less than in normal engines.

If we take the idea that the flame front will travel at similar speeds regarless of RPM, and the ignition advance would need to be approximately a fixed number of microseconds, it would give bigger advance at high RPM and smaller advance at low RPM.

Personally I wouldn't build a compression ignition engine due to difficulties in predicting exactly when something will autoignite. As technology gets better we might be more and more able to accurately determine the time at which something will auto igninite but I see it as partly down to chance too. I'd take things as close as I possibly could to autoignition and then spark things.

Riding a motorbike, one is much more aware of the weather. There is a river near my parents house, and even on a hot day, there is a noticable temperature drop as you decend and go over the bridge of the river, which has cold water even when there is strong overhead midday sun. I'm not sure that today's technology is up to altering the engine setup to deliver the same autoignition timing with the different air entering the engine as I drive over the river. Correct me if I am wrong! Will these HCCI engines correct for increased air density and decreased temperature as I drive over the river?

Others seems to have similar ideas. But maybe there is a certain pride that prevents commonsense ruling. I've seen plans for HCCI engines to have for example two fuels and the ratio to be mixed slightly differently to adjust the point at which autoignition occurs. I think others plan to compress the fuel and then intoduce a tiny charge of an alternative fuel with a lower iginition temperature, thereby causing 'autoignition' to occur on demand. This is effectively like having a spark plug, but anybody with pride could argue they still had a compression ignition engine.

Anyway, back to the question ...

Should the fuels autoignition temperature coorespond to 40DBTDC, 30, 20, 10?

I'd say it would vary with RPM, more likely to corresponding to a reasonably fixed time before TDC rather than an angle before TDC. And with my assumption of a faster flame speed in a homogeneous mixture, it would also be more important to get get the timing right as a fast flame starting 40° before TDC at low RPM would I suspect be pretty much like knocking or pinging or whatever its called - I get mixed up between the different ideas: autoignition before intended, and detonation which is the supersonic flame front phenomenum, as terms tend to be used interchangably.

 

[obanion]

You missed the essential aspect of HCCI.....no flame front. The whole mixture essentially combusts at the same time. So, the convential rules of ignition timing go out the window.

 

[obanion]

Oh, as for timing control, I'm of the opinion it should be done with ONE fuel, and then use a variable amount of water injection. A very small amount of water will make a significant difference in the heat rise in the cylinder during compression, therby acting as a ignition retard. Having two fuels has the problem of dual ignition. The lower octane fuel autoignites, then the heat rise from the lower octane fuel autoignites the higher octane fuel. Doesn't work as well as a single ignition point, with one fuel. Water just acts to take a little heat out to delay the autoignition when required (higher load, hotter intake temps).

I should also note that if a effective means of intake air temp control existed, that would probably be preferable to water injection for timing control. This makes more sense for a wide spread application of HCCI. In my case if I was to attempt partial HCCI operation as a experiment, water injection would be simpler for me to implement, as I'd already have it there for the purpose of high load, high boost, turbocharged, spark ignited operation.

 

[obanion]

Hmmm...from the presentation here:

http://www.windsorworkshop.ca/downloads/TomRyan.PDF

I get the impression he used a 10.5:1 compression motor and got HCCI operation with regular diesel fuel. I might give it try. Switch to a second set of injectors, injecting diesel at a excess air ratio of 3:1, when I go WOT. If nothing happens, no HCCI. I'm hoping it starts running in a different way. I like this potential means of load control better than ultra lean burn with propane (privious plan), which losses quite a bit of efficiency. HCCI, when working right, is even more efficient than the normal diesel process. I'm planning to exhaustively tune it on a dyno, loading it at certain RPM points and intake temps, and finding the water injection rate that yields the maximum output torque, and therefore maximum efficiency. Will be a hard map to build but I have the capability to do it.

If diesel won't reliably autoignite in a 10.5:1 compression motor, I can always add some kerosene to the mix. Kerosene will lower the autoignition temp. But then I'm back to mixed fuels. The again, isn't pump diesel just a hydrocarbon cocktail anyway?

I'd also be interested in using biodiesel, if it would work. If regular diesel works, then I'll try bio.

 

[SBBlue]

A couple of points here. . . . .

If you are injecting fuel (like in a diesel engine), then you don't have a HCCI engine.

Even with a very high air/fuel ratio, you will still have stoichometric combustion with diesel injection, where you won't with HCCI combustion.

There are several huge problems with HCCI. One is control of ignition. Another is low power density, because of the high air/fuel ratio. A third is incomplete combustion, with high hydrocarbon emissions.

Finally, there is one type of HCCI engine that is widely used. . . .the typical model airplane engine.

 

[obanion]

SBBlue, I said injection, as in PORT injection, like gasoline in a spark engine. This gives me precise control which I will need as fuel flow is load control, but allows the fuel and air to premix before/during compression, so I can have a HOMOGENOUS mix, for HCCI. Sorry about the confusion. I guess saying diesel and inject together, it's easy to assume it's direct injection.

I'm well aware of the ignition control hurdles of HCCI. My thought is I'd want a fuel that would autoignite on cold start and the coldest days I'll be driving it (34F), then use varying concentrations of water injection to maintain a ideal autoignition point over various loads and air temperatures. I fully expect the process of discovering the correct fuel/water/air ratios for maximum efficieny at various temperatures to take a lot of work, but the process exists to find it.

My main question here is if diesel has a low enough autoignition for this use, or should I be looking for something even easier to autoignite (kerosene)? My base simulation of my motor at a typical usage condition shows the autoignition of any diesel in the cylinder to be aroun 15BTDC. That seems almost perfect, and matches the onset of reaction of several optimized HCCI experiment engines I have read about.

I fully understand the problem of energy density. It seems compression ignition of a homogenous mixture is only safe up to about 40-45% of stoich. But because of much higher efficiency, most of the test engines have been able to output 65-75% of the torque of stoich, SI operation.

So HCCI operation could provide as much as 3/4 or the output of stoich SI operation, with only 40% of the fuel consumption rate. For periods where maximum power is desired, I will have the option of turning the diesel off, methanol on, and turn on the spark plugs.

Also, the engine will have a VNT turbo. Which means I should be able to make a good deal of boost during HCCI. So hopefully, I can stay in HCCI opertion almost all of the time, only switching to methanol/SI for a full acceleration.

 

[HydroScope]

HCCI - homogeneous charge compression ignition engine

homogeneous ignition - all fuel and air mixture ignites at once, instantaneous. This would optimally be at top dead centre.

http://www.me.berkeley.edu/cal/HCCI/

DI spark ignited engines where meant to become the next best thing able to reach diesel like efficiency through lean operation by aiming the near stoichiometric mixture of fuel directly at the spark plug. Thus the fuel mixture will ignite in the otherwise highly lean cylinder conditions. This has generally failed most large auto companies as the fuel saturates the spark plug and the fuel does not ignite (other problem such as hotspots on spark sometimes pre-ignition is resulted etc). Their has been some success doing this when the fuel is premixed with air and injected at a lower pressure than that of the HPDI (High pressure direct injected) engines. check out orbital for more info

http://www.orbeng.com.au/orbital/home/home.htm

Diesel engines are already capable of lean engine operation (i.e highly efficient operation) but suffer from the fuel often not completely burning, generally resulting from late injection into the engine cycle, hence smoke i.e particulate emissions are made and the reduction of them is becoming increasingly important to get rid off.

HCCI engines take the best of both worlds, the fuel is introduced at an early stage of the engine cycle so that the fuel has an excellent chance of mixing fully with the air before ignition, unlike a diesel, but does not require a spark plug typically avoiding a hot-spots (hot plug) that can cause pre-ignition, which in-turn prevents high compression ratios which is directly proportional to the efficiency of a ICE. but also avoids the need for a retarding on ignition ( there is no flame front to try and time for. . for an ideal cycle all heat should be added at TDC hence timing is retarded so that most or all heat is added by the time the pistion is at TDC. . . which is why conventional engines fire at 10 deg BTDC etc. . . etc

anyhows . . . Lot’s of people will likely reply with example of auto companies that have successfully achieved HDPI or diesel engines with very good emission controls . . . Please do I love this stuff. . . but the above should give a good discussion for the differences in the technology as some of you seem very confused ;o)

Incidentally a HCCI engine can be designed to run on any fuel and hydrogen is one fuel of consideration with the highest auto ignition of all fuels 585 degrees Celsius.

 

[crysta1c1ear]

obanion
You missed the essential aspect of HCCI.....no flame front. The whole mixture essentially combusts at the same time. So, the convential rules of ignition timing go out the window.

HydroScope
HCCI - homogeneous charge compression ignition engine
homogeneous ignition - all fuel and air mixture ignites at once, instantaneous. This would optimally be at top dead centre.

Crystalclear:
Sure, in an ideal world your cake mix is always perfectly stirred and there is no such thing as lumpy custard.
I'd say compression ignition is something you can put a truth table value on, TRUE or FALSE, but I wouldn't say the same thing about homogeneous charge. The most accurate computer models of HCCI combustion don't assume homogenous charge. They assume a sort of turbulent mixing or the fuel and air. Homogenous charge is really the aim and not the result.

But even with a homogenous charge, you wouldn't get instantaneous ignition throughout; ignition being a random event. Ignition is basically an endothermic reaction and I assume a molecule has to randomly hit another in the right place and with sufficient speed to tear it apart and the loose ends would need the right things to tag on (from a similar random event) to release energy to start a chain reaction - a flame front. With near identical conditions throughout the charge, there would likely be other flame fronts starting elsewhere.

The flame front might be a lot quicker with a (near homogeneous so let's call it) homogenous charge, and some people's timescales might be such that you can say apples are always on the tree or on the floor, but that doesn't make it so, even though a quick glance by me at two trees outside appears to prove that apple falling is an instantaneous event.

==

The point about spark plugs is a good one. For the same reason, even a truly homogeneous charge would autoignite sooner on average in a region around a hot exhaust valve than around a cooler intake valve.