Why can't an internal combustion engine self-start from zero rpm without being spun-up by a starter? 6

[LMF5000]

Land vehicles that use external combustion engines (like steam locomotives and steam cars) don't have a clutch and don't idle. They simply stop the engine while still geared to the drivetrain. To restart they simply open the steam valve and the engine spins up the load from zero rpm.

Thinking of ancient petrol engines, I can see why some non-zero running speed is required for the engine to start running. The spark plug is fired by a magneto that needs a minimum flywheel rpm to make a usable spark; the carburettor admits fuel via a venturi that needs some airflow to actually pump fuel into the cylinder.

But what's preventing a modern engine from just starting at zero? With a 4-cylinder engine there's a 1-in-4 chance of one of the cylinders being somewhere in its power stroke at any one time. The crank position sensor will tell the ECU which cylinder it is. The fuel injectors and fuel pump can fill that cylinder with a combustible mixture. The battery and ignition coils can produce a spark even though the engine isn't spinning. Igniting that mixture should create enough pressure to get the engine spinning, at which point the other cylinders pick up and the engine accelerates to idle rpm.

Any thoughts?

 

[GregLocock]

They can. I used to work on a farm that had an ex army forklift, basically a light 4wd truck with forks on the front bumper. It had a mediaeval I8 gasoline engine, and a duff battery of course. Starting procedure was to figure out which cylinder was about at TDC and due to fire, remove plug, pour gasoline in, replace plug, flick contact breaker.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[VEBill]

"The crank position sensor will tell the ECU which cylinder it is."

If I'm not mistaken, most such sensors remain Hall effect that provide only a reference pulse once per revolution. i.e. when rotating. But it's not a stretch to imagine that a suitable absolute position encoder could be integrated.

"With a 4-cylinder engine there's a 1-in-4 chance of one of the cylinders being somewhere in its power stroke at any one time."

Assuming your 4-cylinder example is 4-stroke, the engine (not forgetting the valve train) might be 179° away from top of the power stroke. If you're thinking that's the same as only one degree, then you'd be forcing it backwards.

Also, imagine it's at precisely at TDC. Ouch.

.: There might be dead zones on 4-cylinder engines.

 

[BrianPetersen]

Mazda Skyactiv 4 cylinder engines with stop-start do what you describe in start-stop operation. It is only possible with direct injection. It uses the conventional starting motor for cold starts. 4 cylinder engines will generally stop with the pistons partway in the stroke - the compression stroke will push it backwards away from TDC as it comes to a stop.

Bear in mind that the crank position sensor is not an absolute encoder. The ECU has to keep track of which cylinder is in the right position to do this. If it loses power (battery disconnect, key off, etc) it uses the conventional starting motor. If it so happens to stop in a bad position, it uses the starting motor.

 

[BigClive]

Without researching on the internet etc. I seem to remember reading that some cars from the pre-1910 era like some Napiers and Rollers were started like this.

 

[JayMaechtlen]

Certainly some engines have flags on the crank that provide multiple signals per revolution, plus one on the cam.


Jay Maechtlen

http://www.laserpubs.com/techcomm

 

[waross]

A model "T" could do it. My dad used to tell about working after school delivering meat for the local butcher with an old model "T".
When cold the engine had to be cranked to start.
When it was hot, for a short time after it was stopped it could be restarted without cranking.
Simply retard the manually controlled spark and turn on the ignition. It would often start on its own.
The "T" had both a magneto built into the flywheel and it also had buzz coils for starting.
The buzz coils had a vibrator that interrupted the current in the primary and the inductive kick each time the points opened would generate a high voltage in the secondary. Enough to throw a sustained spark over 1/4 inch long.
There was a coil for each cylinder and the distributor was a roller contact that rotated and made contact in turn with a copper strip for each cylinder. The dwell was for almost 90 degrees.
The compression ratio was quite low and there would often be enough fuel-air mixture in one cylinder for the continuous spark to start the engine often but not every time.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[CWB1]

In addition to the methods mentioned above, there were a few historic production engines started via the good ol shotgun blank, Marshall tractors being probably the most famous.

 

[enginesrus]

It can start like that, but you wouldn't want to power a vehicle with no transmission, like the steam locomotives do.

 

[Jack Gifford]

Didn't I read that Detroit is at least to the prototype stage with no-starter-motor vehicles?
And, yes, Model-T Fords of the brass era would often start without cranking. A late friend demonstrated to me his T starting 17 times before it failed.

 

[mikeofBelAir]

"It can start like that, but you wouldn't want to power a vehicle with no transmission, like the steam locomotives do. "
IC engines produce their power over a rather limited RPM range. Steam engines develop maximum torque at stall making a transmission unnecessary.
If only someone could develop an instant steam generator, steam-powered vehicles might make a lot of sense.

 

[gruntguru]

Steam engines (Rankine Cycle) have lower efficiency than SI piston engines, especially in the small sizes needed for cars.

There are a couple of negatives to uptake of self-start.
1. Electric hybrid transmissions are becoming more common and starting capability is inherent.
2. As pointed out above, any engine with 4 cylinders or less might stop at a "dead spot" eg TDC on a 4 cyl or any of the positions not having a "power stroke" for cylinder counts less than 4. Electronic camless technologies like Koenigsegg's "Freevalve" could extend self-start capability to any cylinder not at TDC, eg 3 cylinder engines would be OK.

je suis charlie

 

[RodRico]

It's hard to imagine how an internal combustion engine could be started without a starter after sitting overnight because rings won't hold compression over a long time period. I imagine it could work if it were stopped just past TDC for a short enough time that a sufficiently compressed air or fuel/air charge remained. A starter is still required after long stops, however, and I don't see the benefit in avoid its use over short stops. As gruntguru points out, the incorporation of even the lowest levels of hybrid operation (P0/P1) using a 48V starter/generator in place of an alternator renders the starter issue moot... a generator *is* a starter, and it's primary purpose is to recapture energy during braking so the starter function is essentially free.

 

[Lou Scannon]

Steam engines (Rankine Cycle) have lower efficiency than SI piston engines, especially in the small sizes needed for cars.

Indeed. Last time I checked, the power density relative to IC engines was pretty horrendous as well.

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[dgallup]

Don't try telling that to the steam fanatics, they are sure the laws of thermodynamics are a conspiracy started by big oil.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[gruntguru]

It's hard to imagine how an internal combustion engine could be started without a starter after sitting overnight because rings won't hold compression over a long time period.

Wouldn't need to be compressed. A cylinder sitting 25% from TDC containing air at atmospheric pressure could contribute significant energy - a quick spray from a DI followed by a spark and things will start moving.

je suis charlie

 

[AwesomeBill]

Because we use fossil fuel and that requires heat to go bang and needs to be squeezed first to warm it up. If you used a charge to go off like the old powder sticks aircraft engines used it will work. in 10 years we will be so electric.

 

[GregLocock]

I can assure you some hydrocarbon fuels go bang without any compression.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[RodRico]

A cylinder sitting 25% from TDC containing air at atmospheric pressure could contribute significant energy - a quick spray from a DI followed by a spark and things will start moving

The key aspect being a direct injection system (vs manifold or port injection) and the ability to control or sense that a piston happened to end up someplace just past TDC on the power stroke. Seems like a lot of assumptions just to eliminate a starter/generator (that's needed anyway to charge the battery)!

 

[LMF5000]

OP here. When I started this question, my main line of thinking was whether we could eliminate the clutch/torque converter and make internal-combustion engines drive the wheels from zero rpm like steam-engined cars did, and like electric cars do currently.

A related thought I had was that with such low-rpm starting capability, a vehicle could operate using a permanently-coupled CVT transmission - no need to add clutches/torque converters or use a full-blown IVT like they do in some tractors (eg. Fendt Vario -

https://www.youtube.com/watch?v=qqrMX3ujgqk

- explained starting at 2:50)