1 squirt/simultaneous driveability?

[pmbrunelle]

Previously, I had a port-injected V6 using 2 squirts/alternating. At low revs, following a cold start (lets say 0°C weather), the transient response was not great, but acceptable if I was smooth on the throttle.

If I change to 1 squirt/simultaneous, how is the driveability? I suppose that with 1/simultaneous now there may be a lag in fuel delivery of up to two crankshaft revolutions, rather than one revolution as with 2/alternating.

Not that I want 1 squirt/simultaneous, but I want staged injection, so my regular port injectors would all need to be ganged on one output...

Or is it time to bite the bullet and change to sequential injection?

[rickb794]

You say staged injection, but I think you mean timed.

Staged injection is an secondary set of injectors to supplement a small primary set.
They are used when high power demand exceeds the primary injectors capacity.

Before tearing into it I suggest raising the VE numbers directly above the idle area and in the first column to the right and above idle.
When you crack the throttle there is an instantaneous MAP spike that hits an area of the table that is not used in normal driving, raising the numbers in that area will help eliminate stumbling without affecting normal driving.

You did not post an MSQ so there is a lack of clues to what is happening.

[pmbrunelle]

Previous MS2/Extra setup:
mainboard injector staging: 2 squirts / alternating
INJ 1 - gasoline port injectors #1, #3, #5
INJ 2 - gasoline port injectors: #2, #4, #6

Not great, but OK cold transient performance. I've had the MS2 since 2009, so I don't think I'll be able to make it any better...
I use a table called "Warmup AFR Target Multiplier". It multiplies the target AFR by the amount in the table; this lets me use CL-EGO even in warmup.

AFR Table 1.gif

Warmup AFR Target Multiplier.gif

13th 12 October 2015.msq

Now the engine is out of the car for a rebuild (with turbo), it is time to redo the wiring harness...
I want to add water/methanol injection as a "second stage" to take over some of the fuel delivery at high loads (table-based)so here are my options:



Future choice 1: MS2/Extra setup:
mainboard injector staging: 2 squirts / alternating
INJ 1 - gasoline port injectors: #1, #2, #3, #4, #5, #6
INJ 2 - water injection solenoids: bank A, bank B

By going to staged injection, I need to gang all my gasoline port injectors on one output. So in terms of the gasoline port injectors, it's equivalent to saying 1 squirt / simultaneous. The "refresh rate" of the fueling will be halved, compared to what I had before. This is where I'm wondering about the effect on driveability.



Future choice 2: MS3X
mainboard injector staging: 1 squirt / simultaneous
INJ 1 - water injection solenoids: bank A
INJ 2 - water injection solenoids: bank B

INJ A - gasoline port injector: #1
INJ B - gasoline port injector: #2
INJ C - gasoline port injector: #3
INJ D - gasoline port injector: #4
INJ E - gasoline port injector: #5
INJ F - gasoline port injector: #6

Obviously, this will work... I'm just not sure if it's really worth spending the money on this.

[pmbrunelle]

Post to add my *.ini file. Without it, you won't be able to view my *.msq file.

megasquirt-ii.ms2extra.ini

[racingmini_mtl]

You don't add water injection using staged injection. You are not providing more fueling with water. On the MS2, you can only use a programmable on/off output while on the MS3 you can use the built-in water injection feature.

Jean

[pmbrunelle]

The methanol in the water/methanol mixture (-35°C windshield washer fluid) combines with oxygen, therefore, I need to pull fuel from the gasoline injectors. The staging feature allows me to do this.

At full boost, I would expect 90% of the oxygen to be consumed by the gasoline, 10% to be consumed by the methanol.

Obviously, this scheme wouldn't work with pure water injection system, but I need the methanol to prevent freezing. I will be obligated to always use the same -35°C stuff so my tune doesn't changed.

The biggest reason I plan to use the staging feature, is that the water/methanol will be port-injected. Therefore, I don't want to use a fixed PWM frequency.

With a fixed PWM frequency, depending on engine RPM, sometimes the intake valve will be in-phase with the water/meth, sometimes out-of-phase. This may cause an alternation between water-rich / water-lean cycles.

So by using an injector output, and having one pulse per two crankshaft revolutions, the water/meth delivery will be repeatable from cycle-to-cycle.

Dead-time for a water/meth solenoid is longer than for a standard fuel injector, but I think that a 62% duty cycle is possible at 6500 RPM (expected redline).

[racingmini_mtl]

Ok. I'm not convinced but that's not based on experience so that doesn't mean much.

If you can make this work, you may want to try 3 squirts/sim. That would give you a more consistent behaviour between all cylinders and a more up-to-date fueling for each squirt. Of course, you also have shorter pulses and more dead times so that may not be ideal for your injectors.

Jean

[pmbrunelle]

My theory regarding the necessity of maintaining the water injection in-phase with the engine cycle is just a guess.

However, this is a $12k engine build, so I'm not willing to try fixed-frequency PWM, based on a chance that it might be OK.

Someone else with more cash to burn than I do can test the theory of fixed-frequency PWM with port injection. Most people would inject before the throttle body, so the discrete PWM pulses of liquid would average out before reaching the cylinders, and this would not be a question.

Also, I'm planning on larger-than-usual amounts of water/meth:
Gasoline flow (per cylinder): 202 g/min
Windshield washer fluid flow (per cylinder): 144 g/min

Engine will be non-intercooled, so its survival will depend on the large flow of water/meth... so I better do it right!

3/simultaneous should be OK for the gasoline injectors (normal high-impedance injectors with 820 ms dead-time), but it is not feasible for the water/meth solenoids to operate that fast... they will spend the majority of their time opening and closing.

Patrick

[jamies]

However, this is a $12k engine build, so I'm not willing to.......

Also, I'm planning on larger-than-usual amounts of water/meth:


Engine will be non-intercooled, so its survival will depend on the large flow of water/meth... so I better do it right!

Water / methanol is a "bandaid" solution , i suggest you dont rely on it to keep the pistons from melting.

[rickb794]

High boost without an intercooler is a risky proposition.

An intercooler is almost an additional power adder due to the reduction of spark retard needed.

Look at what the drag guys are doing with Iced intercoolers.

[pmbrunelle]

Water / methanol is a "bandaid" solution , i suggest you dont rely on it to keep the pistons from melting.

Not really more of a bandaid than an intercooler which is relied upon to prevent the pistons from melting...

If I would choose tried-and-true methods, such as air-to-air intercoolers, this would be a cookie-cutter project, and not unique to my car. It would be a boring project, and not worth doing.

Anyway, we have veered off-topic, my question was to know if 1 batch-fired injection per cycle was enough for good driveability.

Jean doesn't seem so sure, and suggested I try 3 squirts, myself I am not so sure... looks like I might be going for the sequential injection.

[Paul_VR6]

I have tried one squirt per cycle and I am generally not a fan. On small injectors 2/3 simult is better.

On the cars I have done on wmi I have never had to pull fuel when wmi is engaged. Generally the additional cooling (that the iat doesnt see) from wmi changes the fueling enough that its a non issue. I have seen this across a range of setups from pretty mild to borderline stupid (30+psi boost on pump gas).

[hybrid]

Mine is definitely richer when water/meth engages (by about 0.5AFR).
I run 50/50 mix.

My IAT is after the meth, so it sees the cooler air.
I'm putting a pressure sensor in the meth line to detect when there's pressure. I will use that to engage table switching to more aggressive fuel and spark tables as a failsafe method.
I'm also using MAT retard as another failsafe.