Tuning for water injection: fuel, ignition, and EGT

[arthurtoo]

I'm running an aquamist HSF-2 kit for my JDM 2006 Subaru STI. injecting only water at 42% IDC onwards. car runs great!

but i realized that there is a very very minor knock during initial injection and it's persistent throughout my logs. should this be of a concern or is there anyway to totally eliminate this?

thanks.

[RICE RACING]

Originally Posted by arthurtoo

I'm running an aquamist HSF-2 kit for my JDM 2006 Subaru STI. injecting only water at 42% IDC onwards. car runs great!

but i realized that there is a very very minor knock during initial injection and it's persistent throughout my logs. should this be of a concern or is there anyway to totally eliminate this?

thanks.

Knock has to be one of the most mis understood terms in the after market performance world.
Typically when an engine is raised in power its vibration recorded through the acceleration sensors will go up (knock sensor), the shape and amplitude of these traces and the zero cross over point can be literally traced within a couple of CA to an in cylinder pressure transducer!, so its a very accurate representation of what is happening inside the combustion chamber.

As the BMEP goes up the 'knock' goes up, this all has to be defined in the ECU you chose to run, some do this better than others, some work, most don't! I would say so far that in all the ECU I have used (not owned or paid by anyone but out of my own pocket!) there is only ONE ECU that you can buy that does this properly and its made by Life Racing in the UK, all the others I would not piss on personally. Why does it work??? they build their own engines that race in LMP1, they are running on knock limit all the time, and they just work as intended, unlike the spec sheet racers with glossy web pages and blogs etc.

Now having got that out of the way and back onto your question.
You need to know what is normal what is not in your own engine, or otherwise the severity of the events, this is easy to define with nothing more than an acceleration sensor and some experience and the correct electronics that have a history of management in the real world at the peak of competition, if your own stuff you run does not meet these then its hard to honestly rely on the information you are being presented with. And note if someone tells you to use audible knock tools then put them in the *fuck head basket* never ever to be consulted with ever again!

Knock has to be quantified and acted upon within one engine cycle, you need to know what it is, if its normal, and where its heading (towards pre igntion) and countered at all costs, the importance of this varies with heat release or potential energy in the chamber IE: the more power it makes the more critical it becomes, some engine types when stressed will fail catastrophically and instantly when moderate levels of knock are exceeded.

Final note, some knock is good! generally the closer any engine runs to destruction the better its performance will be! so its not uncommon with Life Racing electronics to run the engine on the verge of this point all of the time to extract the most power and highest efficiency. Water Injected cars will allow you to run closer to this point without then quickly leading to pre igntion, this is no one really talks about, but is one of the key things that WI allows (not higher ratios of meth than 50% cause that will lead to far inferior performance FYI). So its a long reply but is normal to have light levels of 'knock' and its a healthy by product of an optimized engine. If you levels are right I cant say obviously as that information is not at hand and not validated.

[dlheman]

Originally Posted by RICE RACING

Final note, some knock is good! generally the closer any engine runs to destruction the better its performance will be! so its not uncommon with Life Racing electronics to run the engine on the verge of this point all of the time to extract the most power and highest efficiency. Water Injected cars will allow you to run closer to this point without then quickly leading to pre igntion, this is no one really talks about, but is one of the key things that WI allows (not higher ratios of meth than 50% cause that will lead to far inferior performance FYI)

I was told, when fine tuning REVO software for VW's TFSI engine, the guideline stated to keep adjusting the fuel, timing and boost until the knock retard value reads -3.0 because that's when the engine makes the most power. Maybe it is safe to do that in TFSI engine which comes with forge piston standard.

[dlheman]

Originally Posted by arthurtoo

I'm running an aquamist HSF-2 kit for my JDM 2006 Subaru STI. injecting only water at 42% IDC onwards. car runs great!

but i realized that there is a very very minor knock during initial injection and it's persistent throughout my logs. should this be of a concern or is there anyway to totally eliminate this?

thanks.

On my STI, I find 42% IDC (12 o'clock position) is a bit too late and choose to go for 10 o'clock which injects at around 25% to 30% IDC (I need to look at my HFS3 flow vs FIDC flow datalog). But I am using TMIC so injecting sooner is never a bad idea and I never saw any knock correction.

What is the value of you knock correction during the initial injection anyway? -1.4?

[arthurtoo]

Originally Posted by dlheman

On my STI, I find 42% IDC (12 o'clock position) is a bit too late and choose to go for 10 o'clock which injects at around 25% to 30% IDC (I need to look at my HFS3 flow vs FIDC flow datalog). But I am using TMIC so injecting sooner is never a bad idea and I never saw any knock correction.

What is the value of you knock correction during the initial injection anyway? -1.4?

base on my datalogs injecting at 25%-30% FIDC is around 300-400rpm earlier. what benefits are to be expected for injecting earlier?

yes, the knock correction is -1.4 during the initial injection. few days back i tried advancing more timing where knock is recorded during initial injection, and its at 0 knock now. perfect!