Why bigger injectors?

[Angrey]

Would be interesting to dive into what's at play here. I'd imagine that the BTDC timing correlates into cylinder pressures pretty directly so the longer you wait, the higher the pressure in the combustion chamber will be and therefore the pressure delta across the injection would be less.

What exactly do you think is changing the timing or the end of the injection based upon the blend, the simple duration/pulse of the injection? Obviously this is where your theory about higher pressure would lend a hand, you could run more GDI blend without extending the injection length if the pressure is higher (thus reducing the shift of the end of it).

Sponsored

 

[engineermike]

Would be interesting to dive into what's at play here. I'd imagine that the BTDC timing correlates into cylinder pressures pretty directly so the longer you wait, the higher the pressure in the combustion chamber will be and therefore the pressure delta across the injection would be less.

The GDI does have to overcome the cylinder pressure when injecting on the compression stroke, but if you do the math it's not a problem since the cyl pressure is generally under 250 psi before spark and GDI pressure is in the thousands. I believe Porsche actually floats the EOI as a function of GDI pressure and time-until-spark, which supports my theory that the problem is evaporation time. My power losses I documented were generally only above 6500 rpm, which further supports the theory that basically there just isn't enough time for the fuel to transition from tiny droplets to well-distributed vapor before spark.

What exactly do you think is changing the timing or the end of the injection based upon the blend, the simple duration/pulse of the injection? Obviously this is where your theory about higher pressure would lend a hand, you could run more GDI blend without extending the injection length if the pressure is higher (thus reducing the shift of the end of it).

Basically what happens is that the PCM determines how much fuel mass is needed per cylinder per cycle based on air load, stoichiometry, and commanded lambda. It multiplies this by the warm blend x safety factor. This value is divided by the injector flow rate at actual pressure, and the result is the time the direct injector must stay open. EOI in degrees can then be calculated based on SOI, rpm and the time needed to inject that fuel mass. SOI - EOI is then compared to the max injection angle x safety factor and EOI is compared to the EOI clip and the greater of these 3 values is commanded with the remainder going to the port injectors.

 

[Angrey]

The GDI does have to overcome the cylinder pressure when injecting on the compression stroke, but if you do the math it's not a problem since the cyl pressure is generally under 250 psi before spark and GDI pressure is in the thousands. I believe Porsche actually floats the EOI as a function of GDI pressure and time-until-spark, which supports my theory that the problem is evaporation time. My power losses I documented were generally only above 6500 rpm, which further supports the theory that basically there just isn't enough time for the fuel to transition from tiny droplets to well-distributed vapor before spark.



Basically what happens is that the PCM determines how much fuel mass is needed per cylinder per cycle based on air load, stoichiometry, and commanded lambda. It multiplies this by the warm blend x safety factor. This value is divided by the injector flow rate at actual pressure, and the result is the time the direct injector must stay open. EOI in degrees can then be calculated based on SOI, rpm and the time needed to inject that fuel mass. SOI - EOI is then compared to the max injection angle x safety factor and EOI is compared to the EOI clip and the greater of these 3 values is commanded with the remainder going to the port injectors.

Interesting, so in theory, you could end up with a situation where you're trying to command more GDI and it's actually resulting in less direct injection because it runs the guideline calculations and whatever it can't fit in it dumps the rest of the duty off on the port injectors?

 

[engineermike]

Interesting, so in theory, you could end up with a situation where you're trying to command more GDI and it's actually resulting in less direct injection because it runs the guideline calculations and whatever it can't fit in it dumps the rest of the duty off on the port injectors?

Yes but the neat thing is that it does this automatically. The warm blend table is only needed when you want to command something less than what it’s capable of. So for instance at idle when idle stability and soot are a concern but knock and VE isn’t, then you might command 10% or even 0. But at wot when you want max VE and knock suppression there is no reason to command less than 90% blend as long as eoi clip and max injection angle are set up properly.

The only time it won’t supplement with port injection is if you set the gdi blend to manual and run out of gdi window. It’ll just reduce fuel flow and go lean when it can’t add more gdi blend to meet the fuel demand.

 

[robvas]

Has anyone experimented with only running DI to increase airflow from not having to use the port injectors like with the LT engines?

 

[engineermike]

Has anyone experimented with only running DI to increase airflow from not having to use the port injectors like with the LT engines?

I have but it causes a couple of problems. Cold starting is bad due to the lack of a gdi trough in the piston and thus lack of stratified charge. Also there’s an issue with sudden turning on of port injection.causing a momentary lean condition. And finally the stock gdi system can only support about 450-500 hp without port. Ford does command 90% gdi stock though.

Sponsored