E85 Discussion - HPFP lubrication and Flow issues

[phate]

Dano, thanks for the link. I'll have to read the entire thread, later, but I did read the two posts you cited. Do we know if it is a physical malfunction, or an electrical issue when running higher pressure? We know that we do not yet have access to every table in the ECU, so it may be another "safeguard" that hasn't been uncovered.

I mentioned a few threads back that I just changed my oil. I have UOA's done every other oil change, so I'll have this next one changed and analyzed at 3k miles - a full cycle with the E85 (hopefully). Fuel dilution has never been a problem in my car, so any change will be noticeable.

[Lex]

Be careful about believing all that you hear. This person sounds knowledgeable but may not be or may be misinformed. What other vehicles utilized our injectors?

Also when he spoke about lean AFRs, was he talking about stratified direct injection? If so, correct AFR is maintained near the spark - gasoline and air should still burn in correct proprotions for complete combustion and emissions reasons. The entire combustion chamber doesn't have to be filled with it however in modern DI engines.

Spray pattern will change at different pressures. Also, higher drive voltages may be needed. The injectors I believe have a rating 12Mpa. If this is true, there will be a difference between them being able to open once at 3000psi (20Mpa) and whether that will actually work in the vehicle. Ask about voltage levels needed to drive the injectors at those pressures.

EDIT: I am not a doubter - I just like to cross my Ts and dot my Is ... also great work so far!

[djuosnteisn]

Originally Posted by Dano2010 If I read the post correctly, not in practice...ask SSinstaller. IIRC at ~2300PSI rail pressure his injectors will not fire. Its in this thread.

Dano, that's the highest the sensor will read... for all we know he was at 3k psi. No way to tell for sure til someone steps up and does the leg work for a higher reading sensor (which would be easy for a gen pu, cause the scalar tables are exposed in their ATR).

[Dano]

Originally Posted by phate Dano, thanks for the link. I'll have to read the entire thread, later, but I did read the two posts you cited. Do we know if it is a physical malfunction, or an electrical issue when running higher pressure? We know that we do not yet have access to every table in the ECU, so it may be another "safeguard" that hasn't been uncovered. I mentioned a few threads back that I just changed my oil. I have UOA's done every other oil change, so I'll have this next one changed and analyzed at 3k miles - a full cycle with the E85 (hopefully). Fuel dilution has never been a problem in my car, so any change will be noticeable.

I may not have all of the facts correct but IIRC, SS was running the StandBack and thus was trying to mechanically increase fuel pressure beyond what the SB could scale with voltage. He ported the spill valve and feed line to the injectors, then even each injector feed bore. At that point the RV became the bottle neck in reaching higher pressure so he slapped on a 2300-2400PSI RV [PTPs version maybe] and then his injectors quit firing. So then he wanted to negatively scale voltage with the SB to net less than 2300PSI rail which wasn't possible due to the resolution in voltage, so he is going to try the AP which has direct control over CDFP desired pressure.

All that to say, I am guessing mechanical pressure on the injectors being the issue but I am NO expert in this. LOL

*quietly exits thread*

edit: once again Dustin puts things into perspective LOL

[Speed3eak]

Wait... Weren't the injectors having trouble firing at 2100 psi? Or am I missing something?

[djuosnteisn]

Originally Posted by Speed3eak Wait... Weren't the injectors having trouble firing at 2100 psi? Or am I missing something?

AP and SB seem to report different pressures (or appear to assign different pressures to the pressure sensor voltage). AP's only read up to like 2080 or something IIRC, and the SB reads up to 2300 or so...

... but they both flat line noentheless. And IMO, they're likely just interpreting the sensor voltage differently.

[Speed3eak]

You sure?

I've got 1850 requested for all my HPFP desired pressure tables above 1.20 load and 2500 rpm - here's a screenshot:



FWIW my observed fuel pressure is above 1800 most of the time now too.

[phate]

I'm reposting this from the Cobb forums, for those who don't frequent:

What I wrote:

How is the injector duty cycle calculated? 100% is obviously not a constant spray in a direct injection system, because we would be spraying fuel during the exhaust stroke. What is the 100% benchmark - and is it significant? Is it unsafe to go beyond the calculated 100%?

This is what Travis wrote:

Originally Posted by Travis@COBB This. Basically 100% duty cycle means that the injector is on for 100% of the intake stroke and can't inject any more fuel. In a port injection engine, fuel flow is fixed so to get more fuel into the engine you have to increase the injector's "on" time. The hitch is, as RPM increases this window of time get's shorter and shorter. You have less time to inject the required fuel needed for a given airflow. So if you are approaching 100%+ IDC that means that the injectors cannot inject enough fuel during the intake stroke. So if the injectors are at 100+% that means they are "static" or constantly on for that cycle and you are "out of fuel." To increase fuel flow you either have to get larger injectors OR increse the base fuel pressure. EDIT: I just realized this was in the MAZDASPEED forum! Ha. Anyway, most the principles still apply to the MS platform. Since Fuel Pressure is variable in this scenario, there are some minor differences. FPR and a larger pump will not work here, however. An upgraded pump to handle the higher-comanded fuel flow is a must, however. That said, you will still reach a limit to the fuel system with larger turbos or if you are trying to run something like E85. Until we have a larger capacity injector solution, there will be a very real limit to power output unless you have some for of auxillery fuel injection like auxillery injectors/port injection or meth injection.

Here's the thread: AP's Injector Duty Cycle

[802MS3]

kind of a weak response if you ask me, since no one actually knows the "real" limits of our fuel system yet...

so is he saying that > 100% duty cycle means you are blowing a lot of black soot out the exhaust, aka unburnt fuel?

[phate]

Well, we don't have a point of reference. He is saying that 100% = spraying the entire intake stroke...which I'm a little leery of with overlap. Does anyone know when IVO and EVC occur? I asked the next logical question of which way the injection event moves, but presumably it has to move into the compression stroke.


Edit: I suppose the easiest way to figure this out would be to bump up the fueling until we see no increase in injector pulsewidth at a given RPM.

[Dano]

Originally Posted by rfinkle2 As of right now, the AP maps will not allow you to command over 1798 psi on the gen2. Complete dopiness on my part... The above is not true. rfinkle2 ATR user error!

you can always blame einhorn

Originally Posted by opt_ms3 kind of a weak response if you ask me, since no one actually knows the "real" limits of our fuel system yet... so is he saying that > 100% duty cycle means you are blowing a lot of black soot out the exhaust, aka unburnt fuel?

I would think you are still not getting enough fuel...assuming you are running E and all other things equal. Which is one problem.

[Enki]

Port injection systems are limited (for obvious reasons) to between intake valve open and close... With our engines we can technically spray throughout all four strokes (which of course is a bad idea).

So the real question is, how bad is it to spray during the compression stroke? I understand that spraying late gives less time for fuel to evaporate and burn cleanly, and will increase cylinder washing, etc.

Anyone with some knowledge care to join in?

[Dano]

at some point in the compression stroke the pressure differential between cylinder pressure and injector pressure will cancel each other out to a point and stop or lower injection pressure, no?

I may be using the wrong term but the cylinder pressure will be pushing back against the injector.

We already know if you crank enough on your standback you can spray into the spark event and cause blowout, so IDK at what point the cylinder pressure starts to negatively effect injection.

edit: I suppose a work-a-round would be an IDC triggered WMI setup with some huge nozzles used as an alternative fueling system until larger injectors can be sourced.

I am full of ideas...unfortunately the vast majority are useless

[Enki]

Did some digging on Wikipedia and elsewhere on the nets for crude maths, and found that a typical pressure ratio @ 10:1 compression results in roughly 25:1 pressure ratio; that being said, the absolute pressure estimated with 15 psi boost (not accounting for thermal expansion) would be near 750 PSI or so. This is probably totally wrong but it might be a step in the right direction.

That said, spraying alcohols after intake valve close could likely reduce the pressure inside the cylinder at the beginning of the upstroke due to cooling; I would estimate that no more than 25% stroke would see a reduced increase in pressure though, and likely only at lower RPMs when the cooling effect can be more readily realized.

Again, probably just a bunch of smoke coming out of my ass, but it does make a certain amount of sense.

[Dano]

so 2K PSI > 750 PSI ? which makes it a non issue?

[Enki]

You'd think so, but taking into consideration heating of the air charge as the piston climbs the chamber drastically increases the pressure. All said and done the effective pressure on the tip could be close to 700 psi (assuming 1750 PSI fuel pressure).

We all know that as pressure drops, flow has to increase which means more time; thus, it snowballs and your injectors turn into flamethrowers spewing raw fire into the cylinder when the exhaust valves open. Not ACTUALLY possible as all the air would likely be burnt up, but it's safe to say that not letting the injectors shut off is probably not the best thing for them.

I do, however, want to see what 120% duty cycle looks like as far as AFRs and power output go; I'm not, however, brave enough to drop my fuel pressure enough to find that answer.

[cld12pk2go]

Originally Posted by Lex Another theory I have is that the buildup is actually from pastic and rubber components along the fuel system that are being corroded by the ethanol fuel. You mentioned it was sticky - was it oily? Did it look like it was rubber based or carbon based? Dark deposits that are carbon based are usually found in places that are very hot (ie combustion chamber)

My best guess is that the E85 is chemically reacting with fuel system components that are rated for E10, not E85.

There is probably some of the general fuel system gunk being dissolved by the ethanol as well since the ethanol is a much more polar solvent compared to gas.

Originally Posted by phate I am not inclined to agree, here. There was a mandate back in the early 90's that vehicles had to handle ethanol fuels. Older vehicles had huge problems with the ethanol enriched fuels - it was doing exactly as you described. The gunk is more carbon-y than anything - it's like the nasty coating that piston crowns get after many miles. I think it is simply breaking down all of the dirt that was in the fuel system. BUT, we shall leave no stone unturned. I called Mazda Motorsports about the possibility of ethanol damaging the lines. The rep wasn't sure, so he's emailing one of the Mazda engineers. Edit: I did find a couple of articles describing a reaction between MTBE and ethanol, but I thought MTBE had been phased out of all gasoline for a few years. I'm going to look into this some more.

See previous answer.

Originally Posted by Speed3eak I don't think it would be a problem - how "stockish" are you? Are you talking about running straight e85? Lots of guys run 3 gallons of e85, with the rest 91 or 93 octane - nobody seems to have had any problems yet. And 3 gallons e85 seems to raise the octane of gas enough to all but eliminate knock from what I've seen with driver311 and clk(his name is longer than that but I can never remember it). I think straight e85 is possible too, and would like to commend phate for having the balls to go for it. Stock turbo I can see it happening, for a big turbo though we'd probably need much larger injectors. I don't know enough about it though - this thread has been great reading and I'm looking forwards to seeing what phate uncovers running 100% E85.

I have been running 5 gal of E85 per tank for the last 3-4 tanks. The last 2 tanks we appear to have gone off the winter blend E70 to actual E85 as my LTFT are now ~+13 instead of +10.

5 gal of E85 and assuming the other gas is E0, works out to ~29% ethanol in the tank by volume. This should require ~13% more fueling to sustain lamba and the fuel mix should require ~48% more heat to vaporize in the combustion chamber on a weight basis (and more than 60% after the extra 13% flow increase is factored in).

All of that extra heat absorption during vaporization should be keeping the temps in the combustion chamber much lower than with straight gas, which should strongly reduce knock potential.

Seems to run well without any MAF curve adjustments, which is very odd. I still am hitting my OL 12:1 AFR targets and my injector PW are ~10% higher than they would be on straight gas. So somehow to ECU is compensating in OL...

[phate]

New fuel line is scheduled for delivery before 1030am, or within one hour from me posting this. I plan to drive the car for ~500 miles before I cut the filter open.

[amoosenamedhank]

I wish you would have known about the inline filter right away. I am curious to know if that had been in there in the first place, would you have saved yourself all of the trouble of trying to clean everything from the CDFP forward. If it does, then the rest of us could be an inline filter and HPRV away from E-85.

[amoosenamedhank]

I live in MN.... you can't stand and piss in a circle with out hitting a gas station with E-85.

Plus with an AP you can always have yourself a 93 octane tune, just in case.

[Lex]

We already discovered that the ECU injects fuel during the compression stroke when DJ measured the injectors and superimposed it to the crank position signal.

cld12pk2go, the ECU still uses LTFTs up to around 240 g/s airflow. Are you maintaining the correct AFR even beyond that?

[phate]

Mustang filter looks like a total winner. I just put the filter and second line on. No leaks with the system pressurized. I'll work on a solid mount tonight, but I need to head to work, right now. [Car is sitting in garage until the filter is mounted.]


Edit: Lex, I experienced the same thing when I was running much less E85. My car goes into OL at 1.10 load, so no trims are used. With a 50% mixture (actual ethanol content, not 50% e85), I did need to change the MAF curve to compensate for the needed fuel.

[djuosnteisn]

Yeah, time-warp:

Have you measured when the injectors...

[Enki]

Ok, so if you're running e85, dont rev past 6k under boost. Lol

[cld12pk2go]

Originally Posted by Lex We already discovered that the ECU injects fuel during the compression stroke when DJ measured the injectors and superimposed it to the crank position signal. cld12pk2go, the ECU still uses LTFTs up to around 240 g/s airflow. Are you maintaining the correct AFR even beyond that?

Here is a log from about a week ago that shows mid 270 g/s (MAF curve properly calibrated for ~+/-2 LFTF on gas):



AFR flat well beyond 240 g/s...

[superskaterxes]

Originally Posted by Enki You'd think so, but taking into consideration heating of the air charge as the piston climbs the chamber drastically increases the pressure. All said and done the effective pressure on the tip could be close to 700 psi (assuming 1750 PSI fuel pressure). We all know that as pressure drops, flow has to increase which means more time; thus, it snowballs and your injectors turn into flamethrowers spewing raw fire into the cylinder when the exhaust valves open. Not ACTUALLY possible as all the air would likely be burnt up, but it's safe to say that not letting the injectors shut off is probably not the best thing for them. I do, however, want to see what 120% duty cycle looks like as far as AFRs and power output go; I'm not, however, brave enough to drop my fuel pressure enough to find that answer.

heres 2 back to back 3rd gear runs, literally no difference between them.


on this one the car freaks out and stutters and sputters like theres blowout but who knows.





normal run right after.




i dealt with this all day at the track sunday........

[Dano]

Enki,theres your 120% IDC or close anyway


hahahahaha

[Enki]

Yeah and apparently it runs like ass; I read through DJ's oscope tuning thread and didnt see when the injection event actually started. If 100% DC is measured from say 25% into the intake stroke to 25% into the compression stroke, then we are likely fucked as far as that aspect goes

That first log looks fucking crazy man. Talk about washing your cyls.

[djuosnteisn]

Originally Posted by Enki Yeah and apparently it runs like ass; I read through DJ's oscope tuning thread and didnt see when the injection event actually started. If 100% DC is measured from say 25% into the intake stroke to 25% into the compression stroke, then we are likely fucked as far as that aspect goes That first log looks fucking crazy man. Talk about washing your cyls.

It's in there. You can see the injection event in many of the videos. Under normal light load, the injector opens up some time after the intake valve opens. Then as load increases, the start of the injection event first moves to the left (in time), all the way to the point where spray starts right when the intake valve opens.

As load further increases... the injection window spans outward to the right, as necessary, to meet the fueling demands. Even into the compression stroke, and in some logs, when i would let off the gas after a pull, spray would even barely occur after spark (spark would become much more advanced as load falls off).



IMO, a logical 100% IDC would be from intake valve opening, up until the spark event. As anything before (spraying during exhaust) or after (spraying during combustion) would have greatly diminished fueling characteristics. Ideally all the fuel would be injected during the intake event only.

[superskaterxes]

Originally Posted by djuosnteisn It's in there. You can see the injection event in many of the videos. Under normal light load, the injector opens up some time after the intake valve opens. Then as load increases, the start of the injection event first moves to the left (in time), all the way to the point where spray starts right when the intake valve opens. As load further increases... the injection window spans outward to the right, as necessary, to meet the fueling demands. Even into the compression stroke, and in some logs, when i would let off the gas after a pull, spray would even barely occur after spark (spark would become much more advanced as load falls off). IMO, a logical 100% IDC would be from intake valve opening, up until the spark event. As anything before (spraying during exhaust) or after (spraying during combustion) would have greatly diminished fueling characteristics. Ideally all the fuel would be injected during the intake event only.

not sure about us but if you read that god damm bosch book i got you it talks about stratified and homogeneous DI fueling which occurs in the intake and compression stroke.

read that shit!

[Enki]

Forgot to factor in advance.

So, who's got a line on bigger injectors?

[wolly6973]

Originally Posted by Enki Forgot to factor in advance. So, who's got a line on bigger injectors?

Maybe here?

Possible bigger injectors?

[djuosnteisn]

I'm sure companies are looking into it, but DI injectors aren't the easiest to build IMO.



And lol at the Bosch book. I'm pretty sure we don't use stratified fueling, and Lex seems to think the same. If we did, you'd see injection event very near spark, and it just never happened in the logs i took. Plus, i'd expect that our piston domes would have been shaped quite a bit different, to help "cup" the mix. but ours are pretty damn plain jane.

[Ziggo]

Originally Posted by superskaterxes not sure about us but if you read that god damm bosch book i got you it talks about stratified and homogeneous DI fueling which occurs in the intake and compression stroke. read that shit!

You gotta have special pistons to run stratified charge, and it is also only for light loads. You would also see very lean AFR (like 20 plus) if we were ever operating in stratified mode.

I would further DJs supposition per the SAE document someone posted awhile back, the injection needs to stop 3ms before the spark to ensure proper atomization and mixing.

This is why I am not a fan of E85 on this platform. Unless you want to get crazy and try out secondary PI fueling, you are just shooting your self in the foot right now trying to use it.

[djuosnteisn]

Originally Posted by rfinkle2 Given the insane thread that I just read where you and Lex were using your scope (way over my head), I read Christian stated that our ecu switches from homogenous to stratified and vice versa. Again, I realize how insanely in depth some of the reverse engineering that was going on in that thread is, and just recall reading the ecu's ability to switch methods.

He may have been talking about DI strategies in general, and not our specific platform, cause i'm fairly certain our ecu never uses stratified fueling strategies.


but i'm always happy to be proven wrong!

[rfinkle2]

Originally Posted by djuosnteisn He may have been talking about DI strategies in general, and not our specific platform, cause i'm fairly certain our ecu never uses stratified fueling strategies. but i'm always happy to be proven wrong!

First of all, I would only be proving that Christian mentioned the ms3 changing from one to another (not that I have knowledge other than what I remember reading) lol, because your brain is moving a million miles an hour faster than mine.

I'll edit this post if / when I can find it...

You (@djuosnteisn , are actually quoting Christian here ... where I believe I read it...)

Have you measured when the injectors...


I would describe this as the logic that switches between or blends from stratified fueling to homogeneous. This is the ECU literally switching fueling strategies back and forth as the engine constantly runs..

[Ziggo]

I have no doubt Christian was talking about the possibilities of DI in general. We don't use stratified charge and he would know this.

I don't think any DI engines on the market currently use stratified charge. When I was studying it 4 years ago the experimental engines showed a tick up in fuel efficiency, but it is murder on the NOx emissions.

[rfinkle2]

I know that you guys hate when I quote the ATR helpfile, but this is from pg. 15/71

- Some DISI ECUs switches logic and blend fueling strategies in different modes. Fueling can go from stratified to homogeneous, and back. Torque targeting can go from boost targeting to load targeting. Closed-Loop (CL) to Open-Loop (OL) transitions may not be smooth on a vehicle where the turbo spools very
quickly.

[Lex]

The VW TFSI uses stratified injection strategies

[djuosnteisn]

RFinkle, alot of that discussion was from a long time ago, when there were still alot of unknowns on the platform. Same goes with some of the verbiage in the ATR helpfile.

Look at the facts and make your own judgement, instead of playing the quote game. Or just email COBB directly, since they are your standard for fact.

/slight sarcasm, only cause i don't want this thread to turn into a quote war and lose focus.