Gasoline Direct Injection...

[DoogenBurns]

I was wondering if anyone has any experience with Toyota's GDI systems that coluld give me some knowledge on the philosophy of the technology. My guess is it's all about getting close to stoichiometric burn without sacrificing torque?

Another thing that's kind of eluding me is the direct injections themselves. I know the already high Compression Ratio of the D4S system is to compensate for the lean port injection, and to actually get flame propagation after spark; but is there room to increase injection pressure, duration and volume to get more torque and power out of the bottom end?

I also assume due to the nature of direct injection, you could probably increase CR at the same time to get a more energetic burn? (just so long as your port charge isn't detonating)

Not exactly experienced with GDI systems so all insight would be extremely helpful.

Much Appreciated,
-Doog

[Dimman]

Check the stickies...

http://www.ft86club.com/forums/showthread.php?t=3172

[serialk11r]

You sound like you think there's a lack of fuel delivery or something...that's not true at all. The direct injectors are capable of supplying all the fuel.

I'm not 100% sure but I believe the few important advantages of direct injection are as follows:
1. Increased charge cooling and thus improved efficiency, by allowing the intake air to absorb more of the latent heat of vaporization (not sure to what degree this happens, as port injection can spray onto the valve and it seems unlikely that the heat on the valve could cause the fuel to vaporize...perhaps someone else could give some input on this).
I imagine that at part load operation, injecting the fuel later in a DI system allows a greater efficiency improvement...again not totally clear on how much of a difference this makes.
2. Allowing "stratified charge", since the fuel is placed directly inside the cylinder at flexible timing you can get more fuel around the spark plug. I imagine there's also the ability to cool hot spots on the piston down.

The reason for D4-S is explained in the papers that arghx7 provided. In short, direct injection has a harder time forming a homogeneous mixture because there's less time for it, so DI engines usually have high tumble ports, which restricts airflow at high engine speeds. To get more power, you need more airflow at high revs, so they fitted higher flowing ports onto the 2GR-FSE, and used port injectors to help low end combustion.

[arghx7]

Originally Posted by serialk11r

You sound like you think there's a lack of fuel delivery or something...that's not true at all. The direct injectors are capable of supplying all the fuel.

I'm not 100% sure but I believe the few important advantages of direct injection are as follows:
1. Increased charge cooling and thus improved efficiency, by allowing the intake air to absorb more of the latent heat of vaporization (not sure to what degree this happens, as port injection can spray onto the valve and it seems unlikely that the heat on the valve could cause the fuel to vaporize...perhaps someone else could give some input on this).

Spraying fuel on the back of the valves wastes much of the cooling effect of the fuel. The fuel cools the valve even more during cold starts, which inhibits vaporization. When the fuel doesn't vaporize well a large portion sticks to the walls or comes out of the exhaust as unburned fuel. The wall wetting and unburned fuel needs to be accounted for inside the ECU. Advancing the intake valve timing during cold operation allows overlap and exhaust heat to warm up the valve. However, port injection allows a lot longer time for mixture formation, so there are tradeoffs upon tradeoffs.

Below is a chart from a very recently published study from the Sloan Automotive Laboratory at MIT. It examines the cooling effects of direct injection under WOT operation.



The study took a Pontiac Solstice engine (direct injected 2.0 liter turbo) and equipped it with port injectors. They figured out an average combustion chamber pressure trace at a borderline knocking condition using direct injection only and then port injection only. Then they heated up the intake air and tried to figure out how much hotter the intake air has to be for direct injection to have the same knock characteristics as port injection. The method was used to calculate "effective charge cooling" of direct injection, especially with ethanol fuels.

I imagine that at part load operation, injecting the fuel later in a DI system allows a greater efficiency improvement...again not totally clear on how much of a difference this makes.

End of injection timing is constrained by mixture formation time. If you spray too late, the mixture doesn't have enough time to spread out inside the cylinder and you could get misfire. Start of injection timing is constrained by fuel impingement on the piston. There are also concerns about fuel washing the oil film off the bore.

2. Allowing "stratified charge", since the fuel is placed directly inside the cylinder at flexible timing you can get more fuel around the spark plug. I imagine there's also the ability to cool hot spots on the piston down.

Stratified charge is generally used during cold start on most mass produced direct injection engines. We know the D-4S system uses simultaneous port + direct injection during cold start for the Lexus 2GR-FSE engine. It's hard to say exactly what little things they decided on for this FA engine--level of stratification for example. There's sort of a "strong" and "weak" stratified charge that can be used.

The reason for D4-S is explained in the papers that arghx7 provided. In short, direct injection has a harder time forming a homogeneous mixture because there's less time for it, so DI engines usually have high tumble ports, which restricts airflow at high engine speeds. To get more power, you need more airflow at high revs, so they fitted higher flowing ports onto the 2GR-FSE, and used port injectors to help low end combustion.

Great way of summarizing it. If you understand that fact, it makes perfect sense that Subaru's new 2.0 FB turbo engine doesn't rev high and doesn't use port + direct injection.

[DoogenBurns]

Originally Posted by Dimman

Check the stickies...

http://www.ft86club.com/forums/showthread.php?t=3172

Wow! I cant believe I missed that. Looks like a good read, thanks!

Arghx7 and Killer, thanks for the explanations, it's very clear this tech works a LOT differently than what was explained to me.

[arghx7]

Originally Posted by DoogenBurns

Wow! I cant believe I missed that. Looks like a good read, thanks!

Arghx7 and Killer, thanks for the explanations, it's very clear this tech works a LOT differently than what was explained to me.

There's a lot of confusion and misinformation circulating about direct injection in its various forms.