Heat Soaked Restart Issues

[ns158sl]

Thats good to know.

I thought that if there are issues with OL fueling, that should be addressed with the OL tables, not a table that could affect ALL fueling tables.

[GHuggins]

Thats good to know.

I thought that if there are issues with OL fueling, that should be addressed with the OL tables, not a table that could affect ALL fueling tables.

Yes you are correct. If it's ONLY open loop fueling "on startup" having the problem, then only open loop should be addressed, BUT odds are, it's all of the fueling as a whole or at least that's what I often find What I was mainly meaning by the open loop part, was instead PE open loop....

[Cringer]

Is this hot soak restart issue only with alcohol? OR is it all the time? Is it really only "3" seconds to fire? Just making sure I understand it correctly.

I haven't put E85 in for a long time (I assume its also problematic too).. It's E0 - E10 and it's consistent taking taking 2-3 seconds of cranking after sitting for 30+ minutes. I'm sure it does it at the 15 minute mark too, but I'm baselining 30 minutes for my testing. It also happens at the 1.5 hour mark as well.

[ns158sl]

I have to ask, why did you spend the money to have Banish get the data vs buying something like OEM or ID? Future plans for FI?

[Cringer]

My understanding is it is how the actual injector is delayed from actual tip temperature? So if the engine is delivering the same ECT/IAT/MAT/Runtime, it would be calculated the same even if the airflow changed? Especially if the MAF/VE table is calibrated, then the TIP temp would be adjusted for airflow in the PCM?

My understanding is this.

First offset is not a good term. Offset indicates moving the injection event sooner or later (like SOI or EOI). But apparently offset is an adder to the total pulse width.

The ECM knows the air going into the engine (dynamic airflow) and then divides that by the AFR for the fuel. Then it has to calculate how long to hold the injector open to spray that amount of fuel. There are many variables: fuel pressure, manifold pressure, system voltage, etc.

However, the fuel injector is an electro mechanical device. My theory is that there are two ways that heat effects the injector.

1) heat increases the electrical impedance on the solenoid that opens the pintle valve. This means that the applied current to the solenoid will take longer to saturate the coil, thus delaying the initial opening of the valve.

2) metal expands when hot and due to the small orifices in the injector and the pintle valve itself, the effective area shrinks.

So the offset is an adder to counter act the weakened performance of a hot injector.

And it is in this context that I do not prefer to zero out the offset temp table. I'd love to hear everyone else's thoughts on this.

[Cringer]

I have to ask, why did you spend the money to have Banish get the data vs buying something like OEM or ID? Future plans for FI?

It started because I got too good of a deal on a flex fuel sensor. I know adding E on an N/A motor isn't worth much, but I was chasing a community 1/4 mile list times and every ounce of torque was gold. Ran out of LS3 injector. Tried LSA injectors but the spray pattern sucked. Blindsquirrel suggested L96 injectors (slightly bigger than LSA and correct spray pattern) but they were too small. Found a deal too good to pass on these...I misread the ad in my excitement....thought I was getting ID...I'd never even heard of FID fml. Anyway you gotta piss with the cock you got.

So now this is a tuning lesson for me. Unfortunately.

[ns158sl]

I understand the mechanics of heat and an EH/EM component lol. Hence why that table shouldnt be played with Willy Nilly to compensate for something unrelated. I also wouldn't zero it out. Getting it right is going to be the challenge. That's why I said try a known table and "bake" in around it. Best a guy could do I guess?

[07GTS]

the issue at 30mins is only in cranking mode as it dosnt get to the run mode when realistically it should, so injector data, cranking fuel, cranking spark, startup airflow is all that should be in play

[smokeshow]

I have tried that. In fact the tune I uploaded called "hjtrbo 20 pct more FA1.hpt" has the MAF SS values set to 20 g/s and I increased the stepdown value as well.

With all the airflow changes I am looking for changes to the calculated cylinder airmass and a corresponding change to the IPW. The strange thing to me is that the cylinder airmass remains flat while cranking, while VVE, MAF, and Dynamic Air all increase....and IPW decreases.

Here is what I mean...
Attachment 145587

As they should. Cylinder airmass is at maximum with the engine off. Cylinder fills with air. Once you start cranking, airflow begins. The cylinder just isn't empty before. Looks like it needs some fuel back in the OL tables.

I wouldn't bother with anything injector tip-temp related. Outside of a nice set of fine wire thermocouples, it'll just be a guess to calibrate it. Just leave it stock.

[Cringer]

As they should. Cylinder airmass is at maximum with the engine off. Cylinder fills with air. Once you start cranking, airflow begins. The cylinder just isn't empty before. Looks like it needs some fuel back in the OL tables.

I wouldn't bother with anything injector tip-temp related. Outside of a nice set of fine wire thermocouples, it'll just be a guess to calibrate it. Just leave it stock.

I have tried adding more fuel and that doesn't work.

I just don't understand why IPW tapers off when:
1) Cyl airmass, MAP, RPM, voltage are all more or less flat
2) MAF, VVE, Dynamic Air all increase

What is causing IPW to buck the trend?

[Cringer]

Ok so just a follow up here. 07GTS had commented on a different thread that maybe EOIT was to blame. I had come to this same conclusion (or at least an idea to test since traditional/air/fuel/spark adjustments didn't seem to change anything), but was too lazy to try it. But at this point there was nothing left to experiment with!


So long story short, I swapped in the original LS3 injectors and went back to the OL settings from a 2012 Holden PPV flex fuel vehicle. I wanted to take the FID injectors/data out of the equation. The heat soaked restarts improved, but still were problematic.


I tested in two configurations with regard to EOIT and both worked very well:
1) Delay the spray until just before PPV (trying to straddle PPV with half the pulse width on either side) with Make Up Mode disabled.
2) Take advantage of the Make Up Mode RPM table to chop the total pulse width up into smaller chunks across the intake stroke, but still delay the spray until somewhere before PPV.


Basically, my conclusion is that big injectors make hot start ups worse. This is because since it is a much smaller pulse width due to (a) its a bigger injector and able to move more fuel and (b) the commanded EQ Ratio results in less fuel. Cold starts (with a much richer EQ Ratio) just mask the issue since IPW is longer, resulting in fuel being sprayed closer to the sweet spot.


So by moving to a smaller injector (longer overall pulse width) and spraying close to the sweet spot = much better hot starts. My next test will be to put in those larger injectors and see what happens.


As far as the sweet spot, since a big cam is going to have problems moving air at low rpms during cranking, you have to take advantage of PPV. This gives a much better chance of getting the fuel into the cylinder and less effects of transient issues (fuel sticking to the port walls and evaporation). Perhaps the sweet spot is just before or just after PPV...have to experiment more to see what each engine likes.


Here is the latest log.
LS3 injectors - EOIT Adjustments - Heat Soak 32 minutes.hpl

[gtstorey]

As edcmat-l1 says, larger injectors are like filling a shot glass with a fire hose.

[Cringer]

As edcmat-l1 says, larger injectors are like filling a shot glass with a fire hose.

Yeah I have always heard big injectors make idle quality suffer, but I have never heard anyone talk about start up issues.

[smokeshow]

I have tried adding more fuel and that doesn't work.

I just don't understand why IPW tapers off when:
1) Cyl airmass, MAP, RPM, voltage are all more or less flat
2) MAF, VVE, Dynamic Air all increase

What is causing IPW to buck the trend?

Because there is very little connection between the air and the fuel in the early stages of cranking. The FA Mult table is a function of soak time and ECT, and the sequential enleanment is a function of injection events and ECT...none of that relates to the air. Its the final value after crank enleanment that should be relatively close to run fuel - but thats where you transition to run fuel, so it makes sense.

You're right about the large injectors being tougher to manage during cranking. Can't defeat physics...is what it is. Still think it could use some mult put back into the OL IVT table though.

[Cringer]

Still think it could use some mult put back into the OL IVT table though.

Are you seeing anything specific in the log that indicates lean? Or some other reason?