Because mechanical components are not the only things that fail, even if they’re the most common failure causes. Sensors, wiring, computer components and even programming logic can be failure points. Complexity inherently creates many different points of potential failure.
But car engines have multiple cylinders. Won’t one of them always be at or near that position?
Yes, and the computer will know which one it is. That’s a main difference between a cold start and a hot restart–the engine is in a known state.
Sorta.
There are 720 degrees = two revolutions in a complete single-cylinder cycle of intake, compression, power, and exhaust. Most engines are designed with a uniform distribution of cylinders around that circle. Not all, but most. Assuming an even distribution (commonly termed an “even fire” engine), a 4-cylinder engine will have TDCs that are 180 degrees apart, while an 8-cylinder engine (typically in V8 shape) will have one every 90 degrees.
In an 8-cylinder engine w 90 degree spacing, there will always be one cylinder between TDC and halfway down the power stroke. And just as one is passing / past that 90 degree point, the next is passing TDC.
My bet is that igniting a cylinder at almost 90 degrees past TDC is going to be pretty weak. But as long as it ignites at all it’ll give some push. And then you immediately fire the next one in sequence that’s just past TDC and you get a full power deflagration and you’re running. You might need a bit of boost from the electric starter motor, but only in the worst case geometry or engine not yet warm.
In a 4 cylinder that stopped at completely random positions I’m thinking you’d need a boost from the starter motor maybe half the time. But that boost will be very, very brief.
Certainly there must be a time limit on the “pause,” though, since pressure & temperature are decreasing.
I do not drive like an American: my braking profile tends to be a lot shallower than most people, slowing down gradually instead of stopping hard (which sometimes gets me through a light that the average person would have had to stop for). An Escape that I was driving a lot seemed to be built for the typical American driving style, because my stopping habits often failed to shut the engine off.
And it kinda makes you wonder how good this auto-shutoff/restart is. Does the engine sometimes shut off when it should have kept running for a brief stop? It cannot realistically take the most appropriate action all the time. It seems like an iffy feature.
Ah, right, I was thinking 360º for a cycle. I forgot that internal combustion engines are fermions. For 4 or 6 cylinder engines, that would often put all cylinders far from TDC.
That seems very low - a quick search suggests savings of 2-15%, depending on the car and how often you’re in stop-and-go traffic or gridlock. Is the vast majority of your driving on a highway or on roads where stopping is very infrequent?