DI under high boost.

I just got a dual ECU motec for my Huracan. I was running Motec as a piggyback. The question I have is when running direct injection and port injection. If you split the injection 50/50 for port to DI, the initial port spray would be very lean until the DI spray comes in. Is there any worry of preignition with such a lean initial mix under boost or should I try to phase out DI under high boost in favor of all port injection. I have never tuned DI under lots of boost and I was only running 16 before.

Tim,

This should be accounted for in the fuel film model which is sometimes referred to as wall wetting.

HPA has a module on this within the Motec training course:

https://www.hpacademy.com/dashboard/courses/motec-m1-software-tutorial/

The main table, offset get tuned, engine speed comp, coolant temp comp, and then you have the Fast and Slow portions which are more related to what you're describing. It all works together though.

The in firmware context help for this also has good info if you highlight fuel -> film -> secondary.

Reference engine speed and mixture aim are also critical values so I suggest reading through then making sure those 2 values are appropriate before you get into tuning the fuel film model.

Tim,

What is your Fuel Timing Secondary Cycle Delay set to? This value is used to set the transport delay (in engine cycles, 100% = 1 Engine cycle) from the Port Injectors from the initial pulse until the complete injection mass is seen by the combustion chamber. The Default for this parameter is set to 0, this means that all of the initial injection pulse is presumed to get into the combustion chamber instantly, I would change this to 100 or 200% and see the difference in the lambda trace when the port injectors come in. The Help goes into more detail.

The way that the M1 calculates the balanced between the Primary and Secondary is on the Fuel Volume, not pulse width. The calculation is Secondary Contribution % = Secondary Reference Flow/(Primary Reference Flow + Secondary Reference Flow) * 100.

Direct Injectors must always have fuel running through them when the engine is under load, they are designed to be cooled by the fuel flow and if this isn't present, they rapidly melt and resemble a miniature volcano when they fail. I wouldn't go below 20% Primary contribution.

So is this ever an issue when your port injection is only injecting 50% of the fueling or do the DI injectors inject the rest so early in the compression stroke that its not an issue with the initial charge being not the full amount of fuel?

If the Fuel Volume from the PI is not getting into the combustion chamber in time, then it is an issue as the Fuel Volume from the DI will not be the full required volume, leading to a lean burn/misfire condition in the chamber.

Stephen,

I appreciate your insight and hope you don't mind me leading a few questions for you. I think Tim has brought up something many can benefit from with PI (Port Injection) + DI (Direct Injection) systems on many ECUs. Your comment about never shutting DI down during high load combustion is really critical and I'll try to add a little detail on the PI timing portion because some readers may not have understood the concern on first read.

In addition to selecting optimal or near optimal PI timing which I agree is something we should all be doing in general, due to the nature of PI at high PI injector duty cycles many people use i.e. 80-95%, we can expect some fuel from a single PI event to enter the cylinder each cycle, and some of the fuel from that same injection event will wait outside the cylinder behind the intake valve for the next intake valve opening event, despite attempts at optimized PI injection timing. At high PI duty, the window of opportunity for PI fuel to enter the cylinder on the first cycle is shorter than the injector on time. To be frank, PI injection timing often isn't perfectly optimized either, so the available timeframe per cycle often isn't perfectly utilized either.

Conversely DI injection is brief, occurs in cylinder, and with proper combustion there should not be a significant amount of fuel left over in cylinder from one cycle to the next.

Based on this would it be fair to say the greatest concern comes in large transient conditions including but not limited to a fuel cut starting, or a fuel cut resuming?

Here are two examples which assume PI duty cycle is let's say 85%, or perhaps it's 50% but fuel injection timing isn't perfect, so each pulse worth of fuel is not all making it into the cylinder each cycle.

In the fuel cut start scenario if DI and PI inject, almost all DI fuel is used or exits the exhaust, some PI fuel goes in cylinder and some is on the back of the intake valve. Combustion occurs, DI and PI fuel injection is halted for the fuel cut. The leftover PI fuel from the prior cycle that's been waiting on the back of the intake valve gets ingested as it opens, along with a small amount of fuel that's been wetting port walls, then that inadequate volume burns by itself lean of target since DI injection hasn't occurred to provide the remaining fuel volume required to hit target lambda. On the following cycle there should be so little fuel left over that no significant combustion occurs, but that first "cut" cycle is concerning.

Do you feel spark cut limiting is a viable solution to this concern?

In a fuel cut resume situation, you get all your DI fuel on the first combustion event, but only part of your PI fuel makes it into the cylinder as the injector continues spraying after the intake valve closes, so your first fuel resumed cycle burns with less than the total fuel volume required, lean, and then the next cycle burns normally. This assumes wall wetting, fuel film, and all other fuel settings are perfectly tuned of course.

How do you suggest avoiding this fuel resume concern? Again by using spark cut limiting or some other means?

For those concerned with using spark cut limiting due to their valve train being sensitive, would using higher flow PI injectors to reduce max injection duration, so that in conjunction with proper PI injection timing each fuel injection event could complete in time for "complete" delivery to the cylinder each combustion cycle?

And I mentioned the concern isn't limited to full fuel cuts or resumes, because there are other large transient conditions possible like suddenly going from cruise to WOT, but would you agree that isn't as large a concern since the potential change in air mass per combustion event is only so large?

Thank you as always for your valuable insight Stephen!