"...an air bubble anywhere in the system would auto-ignite..."

Any air bubbles would presumably be infinitesimally small, so the problem would solve itself with an infinitesimally small popping sound.

Maybe.

Why would you expect it to auto-ignite? Fuel vapor catches fire not liquid fuel and the high pressure is very good at keeping liquid as liquid. Air is mostly nitrogen and oxygen which really want to stay a gas and neither are flammable.

I think the actual answer to your question is any air in the fuel system is sent back to the gas tank on the return leg or sucked into the cylinder, whichever comes first.

At a former employer we spent a lot of time rigging ignition and heat sources trying intentionally to combust various fuel systems from within due to their use in the oil fields and other potentially flammable environments, it was one of the more fun series of experiments I've run. Long story short - its commonly more difficult to be within flammability limits than not.

Ever see the Cody's Lab experiment where he filled a chamber with pure propane and then introduced an oxygen stream so as to ignite an oxygen flame in a propane atmosphere? Not as easy as one would expect there either. No worries people; before the 02 is added he fires up a basically continuous spark.

https://www.youtube.com/watch?v=8jmX-TUQkx4 Burning Oxygen In Propane Atmosphere

Your confusing yourself using compression ratio. Inside of the cylinder there is quite a volume change as the air is compressed, 14-22:1 for diesel engines. That means a lot of work is done and there is a substantial temperature rise, 600°F in a diesel engine. In a fuel rail the pressure ratio is quite higher but the compression is minimal, 1.19:1 or so. The work done is minimal so there is minimal temperature rise. Combine the lack of heat with a lack of oxygen and you get no auto-ignition.

The air bubble is surrounded by nearly incompressible fluid (fuel) at 2,500 bar. Wouldn't that liquid compress said air bubble just as effectively as if the incompressible fluid were a metal piston? The relationship of volume and pressure is P2 = P1 * (V1/V2)^y. I calculated V1/V2 (the volume ratio) to be (P2/P1)^(1/y) where P1 is 101,325 Pascals (atmospheric at sea level), P2 is 2,500*100,000 Pascals, and y is 1.4. This calculation yields a volume ratio (aka compression ratio) of 265:1. I assume the temperature of the air bubble would go up as it's compressed just as if it were compressed in any other fashion per T2 = T1*(V1/V2)(y-1)

I think the portion of RVAmeche's answer relating to vapor pressure is the key. As an analogy, a deep sea diver's blood is quickly saturated with nitrogen at depth. Likewise, I suppose an air bubble surrounded by liquid at 2500 bar would quickly dissolve yielding, in this case, an "air/fuel mixture" incapable of igniting.

Heat transfer between the tiny mass of air in the bubble and the surrounding fluid would dissipate the heat of compression *very* quickly.

Instead of a small amount of o2 or normal air make it a large amount! What will happen is the rail pressure will not go as high as with solid fluid, but there will still be a high compression pressure on it. It would be an interesting thing to experiment with.

Brian, Good point! Adding heat loss to the vapor pressure argument makes it highly unlikely if not impossible for anything to ignite.

Quote (RodRico)

Modern common rail systems run as high as 2500 bar (36 kpsi) which is equivalent to a 265:1 compression ratio. At this pressure, I would expect an air bubble anywhere in the system would auto-ignite. Do I have this right? If so, how is air purged from a newly installed injector?

I think it's been adequately pointed out that the geometry (surface-area-to-volume-ratios and such) makes it unlikely that any tiny pockets of air trapped in the injector could ever get hot enough to ignite the fuel they come into contact with.

But what if they magically did manage to light off? we're talking about cubic millimeters of air here. With the tiny amount of fuel you could oxidize with this tiny amount of air, you'd release a tiny amount of heat, but the pressure increase would be pretty small compared to what the high-pressure fuel pump is trying to do inside the rail/injectors. It'd be a real non-event, nothing to write home about. Instead of tiny air bubbles trapped in the injector bodies, you'd get tiny pockets of exhaust gas. Either way, you trigger the injectors a few times, they spray a mix of fuel and air bubbles (or fuel and exhaust gas), and then you've got all the gas out of the rail/injectors.

By the time any significant amount of air makes it's way in there, I'd think that the engine would stop running. That is to say that the volume of most pumps can not bring the system with air in it to a significantly high pressure.

A pressure washer pump, which uses plungers, pistons, whatever, when it sucks up some air the pressure is gone forever it seems. 2600 psi to a trickle in 1 second flat.

A mechanical injection pump, with the air in there, won't pop open mechanical fuel injectors. There's no longer enough pressure.

Similarly, the design of some modern high pressure diesel pumps uses pistons/ plungers. Once air gets there, with it's compressability, if you will, the pressure created inside the pump would be insufficient for it to flow into the high pressure side of the fuel system.

To add to that, if there's check valves it won't go to the high pressure side. The air expands on the intake stroke and then limits the amount of fuel that can be brought into the chamber of the pump.

I've had the unfortunate experience of refuelling my diesel on the side of the highway after running out. I had to go so far as to crack the lines to the injectors to get the air out of there. Even after the lift pump did it's job the high pressure pump could not with the air behind it.

Engineering student. Electrical or mechanical, I can't decide!
Minoring in psychology