PI : DI Ratio Discussion

[arghx7]

^ Ford ran a bunch of experiments with just such a setup. You're basically injecting ethanol as a type of octane booster or racing fuel. If you just want to make power in a simple way (don't care about minimizing E85 consumption) I don't see much advantage to such a complex system.

[Grip Ronin]

Originally Posted by Kodename47

It's definitely something I'd like to understand more.


On a log I've compared recently, I've had very slight FLKC appear at 6.8k after the PI ratio was decreased despite running the same timing. However, the IATs are also 10C more on the most recent log so I'm not willing to put this down to injection ratio alone.


It may also be worth noting that @Shiv@Openflash uses the OEM ratios for the OTS maps. I've also heard that one tuner relies almost solely on DI for boosted applications and only introduces PI when the DI capacity is near it's limitations. It'd be nice to see some results from actual testing, and I recommend that if you do fancy testing that (a) you know what you're doing and (b) that you do it in similar conditions.


I am running a non-standard ratio map on my car, but as I've not had time to change and test for comparison sake I can't even be sure there's much/any benefit to having changed it.


This is as it currently stands:



My line of thinking is that the DI will be used to help prevent knock/cool cylinders in the mid range and at low RPM with high load (20% rather than 50% OEM). When the RPMs get higher and the DI capacity is getting near a "safe" limit then I re-introduce PI at a slowly increasing rate, but hit near stock levels at peak RPM as the ignition timing is increasing.

curious how did you adjust this di ratio if the table only goes to 5200rpm? i cant seem to find out how to do it.

[Kodename47]

Originally Posted by Grip Ronin

curious how did you adjust this di ratio if the table only goes to 5200rpm? i cant seem to find out how to do it.

Change the axes

[Grip Ronin]

Originally Posted by Kodename47

Change the axes

genius, i always forget you can do that on romraider lol. might you know when do the "warm" and "hot" tables take effect?

[Kodename47]

Originally Posted by Grip Ronin

genius, i always forget you can do that on romraider lol. might you know when do the "warm" and "hot" tables take effect?

It's best just to do them all the same. They're the same from the OEM.

[ztan]

I've had a look at port injector duty cycle calculation.

Will post MatLab script later on the MatLab thread.

Calculation:

Cycle time ms/cycle
= 1000 ms/s * 60 s/min * 2 rev/cycle / RPM rev/min
= 120000 / RPM ms/cycle

Injector duty cycle %
= 100 % * Port_IPW ms / Cycle_time ms
= Port_IPW * RPM /1200 %

With stock port injectors and stock PI : DI ratio map I am up to about 23% duty cycle at redline and 1.2g/rev load.

[ztan]

This one is interesting and not much info available. I have made quite a few assumptions here in the absence of hard data on Toyota's DI system. Do your own verification before using these to tune your engine...

Since DI only has one rev to work with for fuelling rather than a whole cycle: at an 8000 RPM redline, there are 7.5ms per rev rather than 15ms/cycle that the port injectors have to work with.

I looked at a number of relationships betweek DI_IPW and load in log lines with Port_IPW=0 and found the most linear relationship with the commanded fuel vs DI_IPW*FuelPress:

Commanded Fuel fuel g/rev
= Load air g/rev / Commanded_AFR air/fuel

From what I understand, most DI injectors have up to a 5ms pulse width time. Flow is also proportional to fuel pressure. Luckily for us, the peak DI IPW is 5ms and peak DI fuel pressure on the GDI flow table is 20MPa. Multiplying the peak DI (5ms) and peak fuel pressure (20MPa) essentially gives us a 100 % value that we can use directly.

Scripts will be posted in the MatLab thread.

Figures:
1. Relationship between DI_IPW*FuelPress vs Commanded fuel
2. DI Duty using stock DI system and PI : DI map

[Kodename47]

Interested on the DI cycle. I run far more DI than the stock setup, my S/C runs about 5.8ms on the ECUtek logs at the higher RPM, compared to ~3.5ms for stock. Is there any way to calculate this at all?


You can log the PI duty on ECUtek although I currently don't....

[P@ul]

Originally Posted by arghx7

^ Ford ran a bunch of experiments with just such a setup. You're basically injecting ethanol as a type of octane booster or racing fuel. If you just want to make power in a simple way (don't care about minimizing E85 consumption) I don't see much advantage to such a complex system.

Ford called this the Bobcat. I recall having a conversation with a coworker that some turboed v8 prototype motor was making 1000 lb ft on a dyno. It was supposed to be an alternative to large diesels, avoiding the costs and packaging of SCR and Urea Injection (IE Scorpion).

http://www.greencarcongress.com/2009...-20090426.html
http://news.pickuptrucks.com/2009/06...el-engine.html

There's some MIT spinout that was working on it, but their webpage is pretty sparse: http://www.ethanolboost.com/

[arghx7]

Originally Posted by Kodename47

Interested on the DI cycle. I run far more DI than the stock setup, my S/C runs about 5.8ms on the ECUtek logs at the higher RPM, compared to ~3.5ms for stock. Is there any way to calculate this at all?

You can log the PI duty on ECUtek although I currently don't....

You have to start thinking in terms of crank angle degrees and injection window. On PFI, the fuel can build up on a closed intake valve or spray into an open intake valve. For open valve injection, we can say you've got the full 720 degrees to inject PFI fuel. That's 100% duty cycle. The start and end of injection don't matter at that point.

For DI it's different. In one sense, pulsewidth doesn't matter as much assuming a fixed rail pressure. You're limited by your injection timing window. For max injector flow you can't spray when the exhaust valve is open, or rather you can but it will go right out of the cylinder and be useless. You've got from the start of injection time until a maximum of when the spark plug fires. So if my SOI is 320 BTDC, and my spark is 20 BTDC, I've got 300 degrees window (oversimplifying a bit). The number of corresponding milliseconds varies with the engine speed.

For DI: log your SOI, log your milliseconds duration, calculate your crank angle duration from that, and then calculate your end of injection. The closer your SOI gets to 360 BTDC (top dead center Intake), the more you are hosing down the piston and/or cylinder wall, getting oil dilution, smoke, and other kinds of problems. The same applies to the EOI: the closer your EOI gets to 0 degrees (top dead center Firing), the more you are hitting the piston and/or diluting the oil.

But even if you don't give a crap about hosing down the cylinder and the piston, you still can't throw fuel out the exhaust valve. So your DI will never get anywhere close to 100% duty, because my exhaust valve is open for 200+ crank angle degrees. From what I've seen typical stock tunes don't run more than 200 degrees injection duration on DI due to concerns about smoke and oil dilution, but as I said that can be stretched.

Originally Posted by P@ul

Ford called this the Bobcat. I recall having a conversation with a coworker that some turboed v8 prototype motor was making 1000 lb ft on a dyno. It was supposed to be an alternative to large diesels, avoiding the costs and packaging of SCR and Urea Injection (IE Scorpion).

whatever the final technical merit ended up being, basically government subsidies for ethanol and fuel economy credits are being cut so there was no economic incentive to go forward with it.

[ztan]

Originally Posted by arghx7

For DI: log your SOI, log your milliseconds duration, calculate your crank angle duration from that, and then calculate your end of injection. The closer your SOI gets to 360 BTDC (top dead center Intake), the more you are hosing down the piston and/or cylinder wall, getting oil dilution, smoke, and other kinds of problems. The same applies to the EOI: the closer your EOI gets to 0 degrees (top dead center Firing), the more you are hitting the piston and/or diluting the oil.

I haven't looked at SOI parameter, but it does make sense that the injection window would be (SOI - spark timing). We'd have to assume that SOI and spark would be referenced off BTDC in the ROM.

Does anyone have the RAM address for SOI?

Kodename47 has had up to 5.8ms on the DI injector - anyone else gone higher?

[P@ul]

Originally Posted by arghx7

whatever the final technical merit ended up being, basically government subsidies for ethanol and fuel economy credits are being cut so there was no economic incentive to go forward with it.

I never heard about the outcome. I think the technology is just a little too early, and more dependent on future diesel regulation and diesel after treatment technology. The D35 GTDI was a happy accident from R&D, so who knows...

....Then again, we have sub $2 gas, OPEC in shambles, ME in flux, and will have a new administration in 2 years, etc.

[Kodename47]

I decided to throw together a little calculation tool for DI angles etc. Let me know what you guys think, it's obviously based on assumptions of the exhaust cam angle and DI SOI so I can't guarantee it's accuracy.

Download Link - Kodename47's Direct Injection Calculator

[thambu19]

Have you guys tried going 100% DI at the lower speeds? It looks like Toyota went 100% DI at the peak scavenging areas closer to 2K rpm to avoid fuel short circuiting during scavenging (HC emissions) caused by PI but in the speeds before that it should be possible to go 100% DI as well and take advantage of the knock relief. I am not sure why they did not do it unless they thought the fuel mixing with 100% DI would be poor causing Soot.