First a boring brief summary of what is bad for combustion, processes which form carbon and we can usually see black smoke:

1) lots of fuel with little air to circulate with at high temperature (1000K and more, air is not where the fuel is and vice versa)
2) fuel spending too much time in a zone with extreme temperatures without oxidiser (oxygen) or reducer (pyrolysis and carbonisation)
3) bad fuel dispersion so that fuel particles form a very rich mix with little air to burn with. (low droplet speed, large droplets, etc.)
4) ...

Now interesting summary of bad combustion forming white smoke:
1) fuel that is injected too late, has little time to ignite and combustion starts late
2) fuel that ignites too late due to its chemical properties, or much lower speed of sound of the liquid (foamed, emulsion)
3) fuel which needs high temperatures to burn properly but does not instantly give out enough heat to support that (water rich emulsion, improper chemical composition)
4) fuel which can only burn slowly so that the flame and combustion ceases prematurely due to pressure/temperature drop at expansion. (bitumen burns slowly, yet ocean ships have lots of it in fuel, 100RPM engines don't care)
5) fuel which contains inert material (metals, etc.) which form condensate during gas cooling

Pyrolysis is decomposition by heat without the presence of oxygen; forms carbon and combustile gases, gasification is decomposition of substance by heat with the presence of some oxygen, its aim is to make combustile gases only. And that is the aim of prechamber design - to convert liquid fuel into preheated, easily combustile, easily flammable combustile gases or mix with dispersed fuel droplets.

Prechamber with heart insert - function and description

This is the classical prechamber used in Mercedes diesel engines. So, let's come with some pictures and description.. and important part here will be the injector operation and optimal (dream) setting. But let us first have a look at some basic physical law that we have to take into account which basically states that the bounce-off angle, or reflection is the same as the impact angle. This is true for light, for solid elastic objects it is almost correct and for newtonian fluids - we can at least take it as a rough approximation valid at some range of impact angles, drop size and impact velocities. Here is the diagram:

You get the picture.

And now we get to why we should think about it a little.


Now to the real thing. Let us have a look at the nozzle - prechamber heart insert configuration. First the nozzle. First figure depicts fuel spray. There is a deaf zone in the center - at least it should be, so that no droplets, or even streaks are aiming directly at the heart of the prechamber. This would only cause the droplets do disintegrate, vaporize and oversaturate the area in front of the nozzle with fuel vapours and droplets that lost speed and are just waiting for a disaster. The "disaster" is that there is no flame, just well dispersed fuel that has little possibility to combust fluently, it just accumulated and accidentally ignites all at once later. Properly, we would like the fuel to form one long continuous flame ejecting out of the prechamber to the main area between the head and the piston. Well, this can be helped. [* some nozzles eject a tiny stream directly onto the heart, but it is really small amount and velocity, plus the microdrilled hole plugs itself later anyway]

[the second image is how the spray goes around the perimeter of the insert, the third is a magnification]

IF, the fuel spray goes just around the prechamber insert, and IF it only lightly touches, not only the droplet does not lose all kinetic energy, it also is able to continue in the supposed direction forward. The ideal "travel plan" is this: The droplet starts its travel leaving the nozzle at original speed, this rapidly drops down as the droplet loses much of its kinetic energy, also loses some weight due to evaporation, but not as much for the short travel distance when it meets the prechamber heart side and does a touch like a tennis ball on a court. The prechamer heart is already glowing red hot. The droplet evaporates either entirely or from large part and the vapour ignites. In case of droplets flying by, or those who do a touch and continue in forward direction, leaving with a flaming tail behind them (see BOSCH literature). [overly simplified since the spray is dense and one droplet does not travel alone but I think the above illustrates the function well]

In diesel engine we want a long, constant pressure combustion, which could be achieved by one, long lasting flame source - but fuel ignition speed is limited (1-2ms or so), plus one type of fuel can burn at different rate at different pressure, temperature and droplet distribution. The problem is that the combustion itself is what primarily changes the pressure and temperature. So if you get a bad start, you will not correct it later. With the prechamer insert removed, you get a hard nailing sound from the engine. This is caused by spontaneous ignition of large amount of fuel and further resonance - and lots of loud fenomens happen - basically, spark ignition engines work in this way.

The heart shape: when the insert is worn in a way that the middle part is not bulged out, but flat, hard, nailing combustion also mostly occurs - so it can be seen that some of the droplet stream flows around that direction too, plus, a flat center would cause large aerodynamic drag, sucking the sream to the center and behind the insert deflecting fuel droplet stream too much due to aerodynamic drag change. (well, all flow inside prechamber changes with one component aerodynamic coefficient change)

Also, remember, streams flowing at the proper angle around an object tend to bend BEHIND that object, and in this case it would really help the combustion in the right direction. In fact, if this is propably the most important aspect: the prechamber heart insert can be considered an aerodynamic insert that has the right shape to deflect the stream flowing around it so that the fuel stream aims directly at the prechamber exit. [you saw a flow around a wing, the flow lines close behind it]

For those of you who are reading this this far: Example is an OM617 engine which has some hard, nailing sound and lots of black smoke when new DN0SD265 nozzles are used in injectors set up at 150BAR. AT OM601 and OM603 engine it was tried and proved that pressures over 145BAR too cause very unpleasant combustion noise. For the mentioned OM617, lowering injection pressure to 138BAR caused the engine to be as quiet as a sleeping baby. We here used 135BAR or 130BAR in different cases to the same results, depending on the diesel grade the engine could be as quiet that at idle the loudest noise were the valve springs, and that is a lovely sound! (you know, such "yumyumyumyumyumyumyumyum" sound)

The nozzle opening pressure can change the spray angle and distribution a little, but it mainly and radically changes the droplet exit speed. (we consider some constant pressure loss in the injector assembly since the injectors are fed with piston pump). More on that and on pV diagrams later.

Missing letters? diagrams? explanations? grammar? leave a comment.