How far are you guys letting EOI drop?

[engineermike]

I made 2 pulls at the late EOI and both nosed over up top. I did 3 pulls, then 2 more, with the earlier EOI and they all built power all the way to 7100. It's been surprisingly repeatable.

EOI on a Ford is determined by demand, pressure, and limits. I raised GDI pressure from 2600 to 3200, which pushed the EOI about 20 deg earlier. I also raised the limit up top, which caused it to be 30 deg sooner up top. I suppose some part of the difference could be due to the pressure/atomization but probably not.

Again, this is on the Gen3 Coyote which has no GDI dish. My guess is that anything with any sort of GDI piston trough will tolerate later EOI's than a Gen3 Coyote.

[engineermike]

I feel like I'm missing an opportunity to make money but what the heck.

You asked about repeatability, so here are some to compare. I also have the EOI numbers plotted. Cursor is on 7000 rpm on all snips. Looks like I gained about 60 hp at 7000 rpm by moving EOI 40 deg earlier. There may be more on the table as well. I'm also very impressed by the precision of the VCMscanner power and torque calculations, which are based on change in rpm and speed over time, not the torque PID from the PCM which is inaccurate.

early 1.JPGearly 2.JPGearly 3.JPGlate 1.JPGlate 2.JPGlate 3.JPG

[kevin87turbot]

Wow! That's a huge difference. Thanks for sharing. Did you notice how changing the EOI affected the AF reading on the wideband?

[engineermike]

Wow! That's a huge difference. Thanks for sharing. Did you notice how changing the EOI affected the AF reading on the wideband?

The car was falling over up top and I knew something was wrong but didn't know what. I did lots of research on EOI limits and had convinced myself it was fine at 60 deg BTDCf, but when I ran out of things to check I went back to EOI and discovered this.

The wideband didn't change because Ford controls to a target lambda. If I command .75 it runs .75.

[cadillactech]

The car was falling over up top and I knew something was wrong but didn't know what. I did lots of research on EOI limits and had convinced myself it was fine at 60 deg BTDCf, but when I ran out of things to check I went back to EOI and discovered this.

The wideband didn't change because Ford controls to a target lambda. If I command .75 it runs .75.

It would be interesting to see what a 5 gas analyzer shows us on combustion quality while moving around the EOI. Glad to see the gains you got!

[engineermike]

Cadillactech, you’re actually the second person that’s recommended that to me. I don’t know if the late fuel is wetting the piston, not having enough time for atomization, or a bit of both. Either way, my guess is the particulate matter is higher with late EOI. In the SAE articles they avoided late EOI due to soot formation, but were able to run much later than me. That said, they weren’t really looking for max power and all the tests were at relatively low engine speeds.

[cadillactech]

Cadillactech, you’re actually the second person that’s recommended that to me. I don’t know if the late fuel is wetting the piston, not having enough time for atomization, or a bit of both. Either way, my guess is the particulate matter is higher with late EOI. In the SAE articles they avoided late EOI due to soot formation, but were able to run much later than me. That said, they weren’t really looking for max power and all the tests were at relatively low engine speeds.

I think the main issue with late EOI is getting the fuel correctly mixed and mixed homogeneously. Once the piston is at BDC, the swirl, tumble and air velocity inside the cylinder is much lower than it was as it was dropping down with the intake valve open. Yes some swirl gets going again when the piston rises back up but not as much as when you had a column of air mixing into the cylinder. The GM engineers spent a lot of time with the new LT engines on port design to get everything mixing correctly, lots of CFD time to get it perfect.

But that being said, if you can test this on a dyno with the correct testing tools, you could really dial it in to the degree and make max power without wasting fuel and spitting black smoke.

[TriPinTaZ]

Technically speaking I can run the same calculations on my Z06 with HPTuners. Airmass = calculated torque which can be put into a formula to spit out estimated HP. I tried to get a test in today but its been raining and the ground was to wet to get anything good.

Out of curiosity, did you try the same tests with moving SOI around?

[engineermike]

Ford sets the soi at 330 (30 atdc I). The sae papers say any earlier sacrifices gdi advantages over pfi because it wets them piston that close to tdc. That said, I tried cheating it and learned there’s some hidden limit in there that decreases with rpm and if you try earlier than 330 you hit that limit at 7300 rpm anyway. You can run >335 at low rpm but you don’t need it there because you have plenty of window.

[cadillactech]

Ford sets the soi at 330 (30 atdc I). The sae papers say any earlier sacrifices gdi advantages over pfi because it wets them piston that close to tdc. That said, I tried cheating it and learned there’s some hidden limit in there that decreases with rpm and if you try earlier than 330 you hit that limit at 7300 rpm anyway. You can run >335 at low rpm but you don’t need it there because you have plenty of window.

I wonder if they changed that when they went to GDI specific piston for the Ford stuff? I am surprised that your engine has normal flat top pistons but I guess it costs money to do that kind of stuff on a large scale and it runs good enough!

[engineermike]

I wonder if they changed that when they went to GDI specific piston for the Ford stuff? I am surprised that your engine has normal flat top pistons but I guess it costs money to do that kind of stuff on a large scale and it runs good enough!

The ecoboost motors have Gdi pistons and have the exact same invisible soi limit.

The Gdi dish is specifically designed for better charge motion during compression-injection. The ecoboost motors use split injection during cold start where they intentionally open the injector twice each cycle. The coyote doesn’t do split injection afaik, so from the factory there would be no injection during compression and thus, no need for a trough. I installed the xdi pump on mine and run 70% more airflow than stock so I needed more windows…but not sacrifice 60+ hp along the way.

[TriPinTaZ]

Ford sets the soi at 330 (30 atdc I). The sae papers say any earlier sacrifices gdi advantages over pfi because it wets them piston that close to tdc. That said, I tried cheating it and learned there’s some hidden limit in there that decreases with rpm and if you try earlier than 330 you hit that limit at 7300 rpm anyway. You can run >335 at low rpm but you don’t need it there because you have plenty of window.

When you move EOI I'm assuming the SOI moves along with it on the Fords? In GM we can only tune the SOI and the EOI is just a product of SOI and how much fuel you're asking of it. For example on an LT4 with a low side upgrade, if I allow the SOI to fall to 380(20 degrees BTDC) the EOI is about 160 on E75 commanding .84 lambda. If I move SOI to 360 then EOI then drops to 140 on E75 commanding .84 lambda. I'm interested in testing if there is any power difference between these two. Then I want to test if there is any difference in power buy backing down to E60 where SOI will be 360 and EOI will be 160. I need to schedule a dyno day but I'm going to try to get some of the local LT4 cars to join me to see if we can repeat the findings on more than one car.

[cadillactech]

The ecoboost motors have Gdi pistons and have the exact same invisible soi limit.

The Gdi dish is specifically designed for better charge motion during compression-injection. The ecoboost motors use split injection during cold start where they intentionally open the injector twice each cycle. The coyote doesn’t do split injection afaik, so from the factory there would be no injection during compression and thus, no need for a trough. I installed the xdi pump on mine and run 70% more airflow than stock so I needed more windows…but not sacrifice 60+ hp along the way.

Ok that makes sense, even on the GM side, since compression injection would have a much different charge motion and more change of fuel puddling since the piston is coming at the injector versus running away from it.

Do you like the XDI pump setup? XDI has a setup for the LT motors now, looks like a belt driven HPFP that can deliver enough fuel for 2000hp plus. Could be fun!

[cadillactech]

When you move EOI I'm assuming the SOI moves along with it on the Fords? In GM we can only tune the SOI and the EOI is just a product of SOI and how much fuel you're asking of it. For example on an LT4 with a low side upgrade, if I allow the SOI to fall to 380(20 degrees BTDC) the EOI is about 160 on E75 commanding .84 lambda. If I move SOI to 360 then EOI then drops to 140 on E75 commanding .84 lambda. I'm interested in testing if there is any power difference between these two. Then I want to test if there is any difference in power buy backing down to E60 where SOI will be 360 and EOI will be 160. I need to schedule a dyno day but I'm going to try to get some of the local LT4 cars to join me to see if we can repeat the findings on more than one car.

Let us know what you find Taz, that would be an interesting experiment!! What kind of dyno do you have available near you?

[engineermike]

When you move EOI I'm assuming the SOI moves along with it on the Fords? In GM we can only tune the SOI and the EOI is just a product of SOI and how much fuel you're asking of it. For example on an LT4 with a low side upgrade, if I allow the SOI to fall to 380(20 degrees BTDC) the EOI is about 160 on E75 commanding .84 lambda. If I move SOI to 360 then EOI then drops to 140 on E75 commanding .84 lambda. I'm interested in testing if there is any power difference between these two. Then I want to test if there is any difference in power buy backing down to E60 where SOI will be 360 and EOI will be 160. I need to schedule a dyno day but I'm going to try to get some of the local LT4 cars to join me to see if we can repeat the findings on more than one car.

Sounds like Ford and GM work the same. I command SOI and that’s what it does (as long as it falls after the invisible limit). It calculates EOI based on the fuel quantity needed, fuel pressure, and limits. I manipulated EOI by first raising GDI pressure from 2600 to 3200, which made eoi about 20 deg sooner. Then I lowered the eoi clip for another 10 deg. The next problem is that my blend is being driven down to 80% since I reduced my window substantially.

[engineermike]

Do you like the XDI pump setup? XDI has a setup for the LT motors now, looks like a belt driven HPFP that can deliver enough fuel for 2000hp plus. Could be fun!

I like it, but it’s expensive and don’t expect much help from xdi on tuning. Also, the coyote upgrade I believe is actually the LT5 pump adapted to the coyote. It runs off the cam like stock, not belt driven.

[engineermike]

Update….I clipped the eoi another 5 deg and gained more power.

The difference is almost unbelievable. Since I started, I’ve gained about 45 hp peak-vs-peak and up to 140 hp at the shift point.

[cadillactech]

Update….I clipped the eoi another 5 deg and gained more power.

The difference is almost unbelievable. Since I started, I’ve gained about 45 hp peak-vs-peak and up to 140 hp at the shift point.

Just shows that a lot of power is left on the table without adjusting these to their optimum angles.

[TriPinTaZ]

I know this is a ford engine in a GM thread, but I think these findings would likely translate to the GM platform. I think the unspoken rule for the GM world is you want to stay sooner than an EOI of 160 and optimally stop spraying by 180 EOI (BDC).

[engineermike]

I've been thinking about this a lot to try to understand the mechanism behind the power changes with EOI based on the data I've gathered. Here are my ideas and thoughts at this time:

- Geometric distance to piston, leading to wetting: I don't think this is the issue because it would be a function of only EOI not rpm. My data indicates the EOI effect on power is strongly correlated to rpm.

- EOI angle vs BDC or IVC: I don't think this is it because a low EOI well into the compression stroke and well after IVC doesn't seem to affect power much at lower rpm.

- Rapid pressure rise during compression: My thinking here was that fuel injected during compression would vaporize rapidly in a trapped space, causing rapid pressure rise and pressure forces working against the piston motion. I don't think this is it because the total mass of air and fuel vapor in the cylinder during compression are the same regardless of when you inject the fuel in the case of a PD blown engine.

- Vaporization time: I think this is the most likely culprit. What I've noted is that the difference in torque at an EOI of 85 vs 140 at 5000 rpm made very little difference in power output. However, the difference between 100 and 115 at 7000 rpm was significant. Having a stratefied charge could work with late EOI, but the Coyote piston geometry clearly doesn't have the slightest chance of achieving a stratefied charge, so full vaporization must be achieved before spark. Another data point that contributes to this confusion is that in my late EOI pulls, once the EOI clip is encountered then it stays constant with rpm. During that time, power is steadily and dramatically dropping as rpm rises. This also discounts geometric distance and relationship to IVC and BDC.

I would expect the DI GM engines to be able to tolerate a later EOI than a Coyote due to piston geometry perhaps contributing to stratefied charge. It's looking more like about 120 is going to be the magic number for me to stay above, which is unfortunate because it will limit me to about 75% GDI blend due to injector/pump limitations. I would guess GM could run lower, but you guys could do some testing like mine. The power and torque math functions have tuned out to be highly repeatable.