“Swirl reduces trapping efficiency ( Pinger did a good job explaining the relationship between fluid forces and rate of change of momentum). It is very difficult to balance swirl, trapping efficiency and scavenge efficiency in a 2T.”
With the HCCI combustion, the swirl is no longer a requirement (at least not in the strength it is required for the progressive combustion in the conventional spark ignition gasoline engines, and in the conventional compression ignition Diesel engines).
When the temperature gets above the threshold for auto-ignition, with or without swirl, all the fuel will react with its neighbouring air (and there is plenty of air always) instantaneously (actually in a few crankshaft degrees).
The problem is how to control the beginning of the HCCI combustion (as explains Mazda’s SkyActiv-X expert in the video a couple of pages ago).
If too early, the combustion completes several degrees before the TDC and the engine spends energy to compress at extreme temperature (which means, among others, increased thermal losses) an already burnt gas.
If too late, the efficiency drops (due to the lower expansion ratio) and the hammering torque may even cut the crankshaft (a similar effect to what Ricardo said for the Opposed Piston engines wherein a big phase difference between the two crankshafts over-stresses the late crankshaft and the engine).
For instance, in the following plot:
the curves EGR=0, 10 and 20 (combustion substantially before the TDC) cause a strong increase of the energy required for the compression of the already burnt gas (at the expansion stroke this energy is subtracted, i.e. it offers nothing but friction and significant thermal loss increase; the torque the crankshaft has to apply to compress the already burnt gas is quite strong because the eccentricity of the connecting rod long axis from the rotation axis of the crankshaft is big),
while the curve EGR=35 (combustion substantially after the TDC) is also bad because the expansion ratio is lower and because the impact increase of the pressure while the eccentricity of the connecting rod long axis from the rotation axis of the crankshaft overloads the crankshaft.
You can think of the HCCI combustion as pushing a rock / stone to the top of a hill / mountain (Sisyphus myth):
At some height, somehow (say, by burning the fuel in the cylinder), the weight of the stone (or the pressure in the cylinder) doubles.
Which is the ideal height to get the most energy?
If the weight doubles at the middle of the uphill, the rest pushing to the top of the mountain gets double harder and causes only problems and reduced overall energy.
If the weight doubles at the middle of the downhill, the effort to push the rock from the middle of the uphill to the top of the mountain was at least meaningless.
But if the weight of the rock doubles at the top of the mountain (TDC), it is the best case because the uphill was easy, and the net energy delivered by the proccess maximizes.
In the following image:
the combustion in the HCCI engine (center) is done (completely finished, no flame),
while the combustion in the spark ignition engine (at left) continues strongly (blue flames),
and while the combustion in the right cylinder (Diesel) also continues strongly (red flames (glowing particulates)).
Isn't this (the strict control over the timing of the HCCI combustion) what the PatBam geometry does?