Point 3 is easy, probably it's that I don't explain well:
At low rpm the spark (in the spark plug) is not strong, because the alternator is working at low rmp. Yes, the car has electronics to keep the spark at the same voltage, no matter the rpm, but they're not perfect.
For example, when you wish to check the spark in a car whose ignition is failing, and you don't have an ohmeter at hand, you disconect the spark plug and you put the end of the plug near the block, to see the spark jumping between the spark plug electrodes.
Spark jumping in a "primitive" test of the ignition coil: the higher the rpm, the spark seems to be more bluish
When I've done that, you can see that the spark that jumps it's "bluer" as rpms increase. Thus, I deduce, the spark has less "power" at low rpm. I've seen that in a model 66 car without electronic ignition, I don't think this is important is newer cars.
Point 1: well, here is the animation where I took the images from (you'll need flash to watch it):
Animation of valve overlap (you need Flash or a Flash plug-in in your browser to watch it)
If you cannot watch the animation, follow me for a moment:
When the cylinder starts the intake stroke, air is still (the piston is just starting to move).
Now, the lower the cylinder goes, more and more air enters the cylinder, right?
When the piston reaches BDC (bottom dead center), that is, the low point of its movement, the air is rushing in at top speed, entering the cylinder through the intake valve.
Piston at the bottom of its movement: the air is entering at high speed, the piston barely has started to move upwards, so air continues to enter
If you close the intake at BDC, you'll shut the intake when the air is at top speed. So, you left the valve open for a little, because even if the piston starts to move upwards after the BDC, the air is not going to "feel" that movement instantly: the "compression wave" has to move from the piston head towards the opening for the air to stop entering.
Under that circumstances, even when the piston is moving upwards, air continues to enter the cylinder for some milliseconds. So, you left the intake open a little beyond BDC.
Now, imagine the opposite: the piston is at the top of its movement, in the exhaust stroke, air is exiting.
The air exits faster when the cylinder reaches TDC (top dead center), that is, the topmost point of its movement, right?
Piston at the top dead center: air is still exiting at high speed, so it pulls some air from the intake (if you open the intake valve some milliseconds before TDC)
So, if you open the intake valve a little BEFORE TDC, the fumes, that are exiting from the cylinder through the exhaust valve at top speed, "pull" some air from the intake valve, helping some extra air to enter the cylinder. This is called "scavenging".
So, a valve camshaft is a compromise: at low rpm you do not want overlap, the air is not rushing in and out at high speeds.
At high rpm you need the overlap to take in account the inertia or impulse of the air entering and exiting.
In F1 cars you have a camshaft optimized for high RPM: this is the reason why idle racing engines misfire so much and why it's so hard to start a car from the pits, at low rpm, stalling the engine: there is too much overlap when the engine is idling.
That's the best explanation I can give.
On a side note, if the engine has fixed cams, you can't control your overlap without changing the cams, but
the effects caused by changing the back pressure in the exhaust pipe are the same.
When you reduce back pressure, it is equivalent to increasing valve overlap, and when you increase back pressure, it is the same as decreasing the amount of valve overlap.
An engine with less overlap will give you more torque and less HP. The opposite is also true: an engine with more overlap will give you more HP and less torque.
That's why some people will say, "
you need a muffler for torque", or, "
you'll have more high-end, but less torque, if you run straight exhaust".
By reducing back pressure in an exhaust system, you increase high-end horsepower at the cost of low-end torque. For example, low overlap cams are called "RV" because they're used in recreational vehicles, that tow large loads and need to have high torque.
On the contrary, high RPM engines with high overlap tend to have more HP and less torque.
As you can imagine, racing engines, with almost no backpressure and no mufflers, create the same effect as valve overlap: they pull unburnt air and fuel from the cylinder and hence, flames on the exhaust.
On top of that:
- you close the throttle, because flames appear when you "coast the vehicle" or you brake, thus creating vacuum on the intake of fuel, sucking fuel from the "idle" ports (idle ports, or whatever the english name, are ports that inject fuel to mantain the engine idling when you don't push the throttle pedal).
- the spark is weak because the engine is not at top rpm (at least in some cars).
- the exhaust valve is open while the intake valve is also open at TDC, between exhaust and intake strokes.
- the droplets might be larger, as EfiOz points out (although I've seen flames in carburated cars).
- the engine manufacturer might have designed the engine map to throw extra fuel under those circumstances to cool the engine.
... so the engine expels unburnt fuel through the exhaust: you get flames.
Yes, it's more or less cool, but you'll deposit carbon on piston heads and increase wear of first piston ring (because of carbon on the cylinder walls).
