EMU Pro Fuel film stategy / feature update request

[KMP]

I think the EMU PRO fuel film strategy would need some serious updates.

First off all it seems to work a little bit too much in an ‘‘ideal’’ way. I tracks the fuel film but it considers that all the injected fuel will all be burned at some time in the chambers. In the real world it is not really the case. Some engine have better fuel vaporization then others and some fuels also vaporize better then others in the chamber. As of now the strategy doesn’t take into account that some unvaporized fuel will not be burned and sent straight to the exhaust.

What that means is that on an engine or fuel that has poor atomization and / or poor vaporization we have to create a lot of fuel film target increase with load increase to prevent a lean condittion on tip in but this is were the actual film model is loosing track off the ‘‘actual’’ fuel film as a percentage of that fuel did not go on the intake walls but went straight through the chambers, unvaporized , unburned and straight out of the exhaust. What this causes is now we have a ‘‘false’’ large fuel film that we need to decrease when the load goes down and it ends up always creating a lean condition on decel.

What I suggest here would be to add a parameter that states how much % off the added fuel will be lost by going through the engine unburned.

Second problem with film target is that it doesn’t take into account a 2nd or 3rd set of staged injectors. Fuel film is only calculated based on the primary fuel rail that is, most of the time, the one with small injectors close to the valve that does not need a ton of fuel film correction.
With that said if you turn the primaries off to run only on the staged injectors if brings the film target to 0 since the model thinks there is no more fuel injected into the engine when in fact the film target should go up significantly with bigger injectors even farther from the valves.
All that film confusion creates some really weird lean and rich conditions on load changes with staged injectors and even more when the staged are running on a less vaporizing fuel like E85 or even methanol.

I really think the base of the strategy is awesome but needs some adjustments to really work well.

[mkuklis_ECUMASTER]

Hi,
It would be best to provide some actual examples of those situations.
We have to confirm if your conclusions are correct.
A project and log file with timestamps for those situations would be perfect.

@KMP

Some engine have better fuel vaporization then others and some fuels also vaporize better then others in the chamber. As of now the strategy doesn’t take into account that some unvaporized fuel will not be burned and sent straight to the exhaust.

Indeed, some fuel will not be burned in the chamber, but I think you are overestimating how much of it can do that. If fuel doesn’t vaporize in the intake, it vaporizes in the cylinder, where it’s very hot. Assuming the engine is at operating temperature.

Even when you run a rich mixture, it doesn’t mean the fuel is not burned. It is incomplete combustion, which is still a combustion. The engine produces CO instead of CO2 from incomplete combustion.

The only situations in which there might be a significant amount of actually unburnt fuel are cold starts or extremely rich mixtures.

@KMP

What that means is that on an engine or fuel that has poor atomization and / or poor vaporization we have to create a lot of fuel film target increase with load increase to prevent a lean condittion on tip in

That is completely natural behavior of the fuel film.
Low load = low pressure in the intake = high rate of evaporation = low fuel film
High load = high pressure in the intake = low rate of evaporation = high fuel film

What is compounding the increase in fuel film target is the fact that it’s an absolute value.
That means it has to increase with the overall fuel amount as well as with pressure.
That’s why we plan to introduce a relative mode, describing the fuel film target as a percentage of the injected fuel instead of absolute fuel volume. This should simplify the target table, and fewer corrections will be needed.

@KMP

What this causes is now we have a ‘‘false’’ large fuel film that we need to decrease when the load goes down and it ends up always creating a lean condition on decel.

I would have to see that on the log.
I often fixed that by adjusting the film decrease duration.

@KMP

What I suggest here would be to add a parameter that states how much % off the added fuel will be lost by going through the engine unburned.

I would rather make a dedicated multiplier of the correction applied during the film’s lowering. It would be easy to understand and tune. But first, we must see the data to confirm that it’s needed.

@KMP

Second problem with film target is that it doesn’t take into account a 2nd or 3rd set of staged injectors.

This is still basically the first iteration of this strategy.
We know that and will expand it to work with multiple rails.

@KMP

With that said if you turn the primaries off to run only on the staged injectors if brings the film target to 0 since the model thinks there is no more fuel injected into the engine

That is a bug we found about recently.
It will be fixed in the upcoming versions.

@KMP

when in fact the film target should go up significantly with bigger injectors even farther from the valves.

Now I got an idea for a workaround.
You can make a different film target, depending on how much fuel goes to each rail.
There might be some issues with edge cases, but it should be a viable solution.

We have new ideas and improvements for the fuel film model we want to implement. We just need to dedicate some time. We will increase the priority for this task.

Thanks for the feedback.

[KMP]

HI marek.

The decel fuel multiplier would be a game changer yes ! One of the problem I see with making the target decreasing slowly is if you hit / release and hit back superfast the film doesn’t have the time to decrease so the second hit see nearly no film increase and sometimes makes the throttle response worse.

The film target adjustement with stage % is effectively a good idea yes.

I had the chance to do some logs here. The car is my personnal car, it is a turbo 5.3 GM V8 with 1000cc primaries on fuel and 7700cc secondaries on M1 methanol.

This log I cannot really explain and was not described in my original post. I start the car and everything looks fine in the log but it feels like the film correction is non existant. at 3.00 in the log I even add 200% to the film target and despite showing a huge PW increase on tip in it goes super lean an stumbles. at 3:38 I try it removing the 200% target increase and still the same lean tip in stumble. at 3:55 I shut the car down about 3 sec and start it back and all is back to normal no more lean tip in.
This situation was happening randomly when I was shutting the primaries tottally off in the past. After a pull and turning the primaries off and back on it would create that situation about 1/10th of the time. Is I did another pull shutting off the primaries and turing them back on again it would get back to normal 9 times out of 10.
Since I keep the primaries at 1% it never happened again except from time to time on startup where I have to turn off the ecu and turn it on again.

film bug ok after restart.emuprolog

I will do more logs when I have time trying to show the first situation I explained. but a decel fuel multiplier would totally cure that situation.