Sams Polo 3.0

[Sydneykid]

On a good note, the new oil cooler setup - different to previous in that it now runs a T-stat sandwich plate and the cooler core is blanked off on the street - is running the oil temps right in the good 90-100 window now. Oil temps now run higher than water temps in normal driving. As a result the catch can is for the first time catching plenty of water. To my mind that means the oil is getting to a temp to evaporate out any water contamination and the crankcase water content is coming out as a gas and being caught/condensed in the can. Better out than in. If I run the catch can system with the PCV connected you dont see as much water in the can as it most likely just gets drawn into the inlet manifold under vacuum as a gas. But if the PCV valve is blocked off and the catch can is only TIP connected, then you see plenty of water caught. Yes there's always going to be more water caught in the cold weather but either way it wasnt even
doing this when the oil temps were so low before.

Yep, absolutely need to ensure that the oil temp gets to 100 deg, regularly.

Cheers
Gary

[sambb]

hmm will have to definitely get a proper tune done for the E20 brew. Due to Covid lockdown boredom I thought I'd play around with ghetto tuning for E20 to/from work (apparently newspapers are essential) to test its viability. I tested my local United as having solid E85 and then did 1/4 E85 to 3/4 of 98RON in the tank and made some 20 litre drums of 5 litres E85: 15 litres 98. That should give 103RON.
After doing a bit of reading the conclusion seemed to be that a good ballpark for this high octane "E20" brew is that you need 5% more fuelling which from what I could gather is equivalent to roughly 0.5 bar more fuel pressure on my adjustable FPR. After driving around for a bit like that I noticed that my short and long term fuel adaptations got better. On petrol the additive (idle trims) stayed solid at -2.0% but the multiplicative trims went from -9% on petrol to -4% with the ethanol. I did some tentative pulls before doing a proper log and the new brews fuelling is bang on the same lambdas as the petrol tunes were. Not as stable throughout a run as the 98 tunes but same same. I did the fuelling just with pressure adjustment because I didnt want to run into duty cycle limits on the injectors. The pump is still keeping up just fine with it slowly going to 0.79-0.80 at 6500rpm.
Before coming to work today I was confident enough to add global 3 degrees of timing and the pull I logged didnt have one dot of timing pull. In contrast, the petrol tune with 3 degrees less global timing still had smatterings of up to 3 degrees timing pull in the log. So it looks like there's even more headroom with the timing on this brew. From now on it'll just be 0.75 degrees added at a time as I'd like to find its ceiling to get an idea of its potential.
The midrange is noticeably stronger but you cant really see it in the data. What you can see is an improvement in the top end. Its gulping 8-9 g/s more air between 6000-6500rpm alone. On such a peak power limited setup as mine (BW K04-001) that really is a plus and is probably happening because of no timing pull further down te rev range and I guess a slightly lower burn temp? Off boost the car feels sharper everywhere with the additional timing too, so there's no negatives. Started aok in 8 degrees ambient too.
So far for just a global bump up of fuel pressure and timing its working pretty damn well! With more timing to come, I'm keen to see where this leads. Now I'm wondering if a dedicated tune for it is even needed (rather than my ghetto tweaks).

questions:

I've seen it written that on E85 you dont want to go richer than 0.79 as it can lead to 'rich knock' in the very top end. Is that even true and if so does it apply to 'E20' brews or should I just stick to mimicking the 98 RON tunes lambdas since its still predominately a petrol brew?

Also are my one stage colder than stock NGK BKR7E going to be too cold on an ethanol brew like this or again is that a full E85 consideration that doesnt apply here? Can plug gaps saty the same as for petrol or is shrinkage a good thing?

[Sydneykid]

"Rich knock", to be frank I think it's a myth, my experience has been that it's simply the knock sensor picking up misfire due to the sparks inability to ignite at higher combustion pressure. Swap to stronger coils, no other changes and the knock is gone. Reduce the plug gap same result.

My rule of thumb with E20/25 is to start at 11.8'ish AFR (0.80 lambda) and tune for power from there.

Since E20/25 needs around 5% more fuel pretty much everywhere, upping the fuel pressure to achieve the AFR target is often enough. Sounds primitive but that's why a lot of people run that level of ethanol, the results are very worthwhile and the cost is not exorbitant.

As you have found the timing adjustments are where the tricks are, around 5 degrees more advance is the often the result. So you might have a bit more to come.

In NGK 7 heat range plugs should be OK, I'd only reduce the plug gap if you find the coils aren't up to it.

You will have to further explore the real benefit, that's mid range torque, your turbo is max flow limited but that's a higher rpm limit, at mid level rpm there should be spare airflow available. The lower combustion temp will let you slip a couple more psi into it from around 2,500 rpm to 4,500 rpm AND still have the 5 degrees or so more ignition advance. That's when you will really feel it.

Cheers
Gary

[sambb]

Yeah I've heard it a couple of times. Like a very dull pop/muffled backfire. Ok I'll shrink the plug gap down from 28 thou to 25 or so and see if that cancels it out. The global timing can be altered in 0.75 degree increments. I went up to 4.5 degrees total added but it was pulling up to 4.5 degrees timing in spots. At 3.75 added it was only pulling a max of 3 degrees and mostly just little pockets of 0.8 and 1.5 degrees. Backed off to 3 degrees total added, its ahppy in a range of conditions, so thats very defibitely the threshold.
But if you like you say the knock sensors are misinterpreting misfires for knock, it may have been false positive timing pull that was causing the timing to come out of it. I'll shrink the plug gap and log misfires against knock and see if there's a correalation. Hopefully if I can get on top of a misfire maybe i'll be able to add more timing.
Yep the midrange is undoubtedly stronger to the extent that I have the non VVT tune in it while I play around. With the VVT tune in it it was actually breaking into wheelspin when I was logging if the surface wasnt super consistent. If I cant get above 3 degrees added, then maybe i'll go up to 1/3 E85: 2/3 98 RON (no idea where that blend sits....E35?) and see if I can improve things.

question: The OEM VW plastic fuel lines and click fittings - are they Teflon based PTFE or basic nylon. Thinking in terms of ethanol resistance.

[Sydneykid]

For checking ethanol suitability I grab a couple of spares and stick them in a bottle of E85 for a couple of days, that gives them a good test.

28/1000 is about 0.7 mm, that's a minimum plug gap that I have run, it's a small spark. The Skylines are at 0.8m on NGK 7's with upgraded coils. Are you running the Audi S3 coils? Do the R8 coils fit?

As I mentioned E25 is relatively common and the easy/cheap fuel pressure upgrade works OK. E30 is about as much as I have seen (until they go full E85), but it may be more difficult to get the AFR's right across the range with just the simple FPR adjustment. Worth a try though.

Cheers
Gary

[sambb]

Right I’ll just see if I can sort out a better outcome on the E20 then.
Yeah the coils are R8 with NGK BKR7E’s.
Maybe a Chevy LS coil upgrade is in the making. A lot of guys with the 1.8t 20v have gone this way.

[sambb]

So on the street in civilian duties I've now got the engine coolant temps where I want them. After going through a couple I've found a good radiator fan switch alternative that holds temps where they need to be. The stock 3 pin dual temp rad switch with high and low speed was just set way too high. The water was kept at 96+ degrees once you hit traffic. Tridon (through Bursons) do a TFS211 which is just a 2 pin single temp fan switch. Its set to come on at 87 degrees and go off at 82 degrees. I think because the rad switch is located half way down the radiator, that translates in the real world to keeping head exit temps between 90 and 85 which is spot on. Its contacts were wired to the low speed radiator fan contacts and the now unused high speed contacts go to a cabin fan override switch. The latter is only active with key on so cant ever be left on accidentally.

So the next thing then was to get the front of the car off and get down to plugging all the massive air gaps that allow all the air that passes through the grilles to go around the radiator - through the headlights, into the guards, straight under it, most straight over it into the engine bay. I foam filled the radiator shroud, covered over any gapping holes and made up aluminium air guides to basically force anything that enters the grilles to have to go through the cooling cores. On the track hopefully now temps will stay in a similar pocket.

On the oil cooling front I've decided to ditch brake ducts fed from the front facias. They just use wishbone scoops alone from now on. With those pipes out of the way I finished the air guides onto the oil cooler so that all air that gets through the fog facia has to go through that core. Before at least 50% was just running between it and the radiator surround and back pressurising the engine bay, so that'll help oil temps a bit more hopefully.


I noticed that when you look into the intercooler that it has turning vanes for the air cast into the end tanks - pretty schmick!


Question: the last two pics show how far proud of the radiator the IC sits. The gap is as big as 50mm on the passenger side. It'll be a bit of fab work to box that in (if thats whats needed) so didnt have time to tackle that just now. Now that the radiator/grilles are all bordered in, any air that goes through or around the IC will have to go through the radiator, just not in the most streamlined way probably.
But the question is, will a gap like that affect the efficiency of the IC, the radiator, or both? Its my guess that FMIC efficiency will be just fine, it'll only be the radiator flow thats compromised. So if next time I'm on track and I find that my water temps are AOK, there's not really any reason to make the effort then is there?

[Sydneykid]

For the last couple of decades vehicle cooling efficiency has been mostly due to the draw effect of the front undertray on the rear of the radiator. Plus the air flow routing through the engine bay. Not so much the push effect from the front. As a result the "old school" thinking of having all of the heat exchangers (intercooler, oil cooler, air con etc) mounted without air gaps between them is no longer necessary. In the Skyline, for example, at over 80kph there is measurable partial vacuum in the 100 mm between the radiator and the intercooler,

I'd just monitor the temperatures, I suspect that they will be OK as long as the draw through hasn't been adversely affected, removing the front undertray is the most common reason. The other is the rear of the bonnet not sealing at the windscreen line, it transfers the high pressure from the front of the windscreen into the engine bay, adversely affecting the draw through. Yes, jacking up the rear of the bonnet on pretty much any car designed since the 90's is a bad idea.

Cheers
Gary

[sambb]

Thanks Gary. Yep I have a good rear bonnet seal and the full engine under tray. It extends back to the subframes crossmember, same as your Polo's probably does. The area over the top of that subframe down the exhaust tunnel with all the steering rack and dump pipe cluttering it up doesnt look to the eye like it would be able to flow the required air out under the car, but I guess it does.
Does the undertray need to run that far back or could it be made more efficient if shortened slightly to say in line with the back edge of the sump or have vents placed in it ahead of that subframe?

[Sydneykid]

Full undertray is good, most of the airflow passes behind the engine and transaxle, down the firewall, then around the footwell curvature. The partial vacuum created behind the car "sucks" the airflow under the floor big time.


Cheers
Gary

[sambb]

Alright cool. All done on that front then. Thx

[sambb]

So in the ongoing E20 tuning of my car I've found a few things.

- The 98 RON petrol tune liked 0.82 lambda which was also described as on the safe side by the tuner. On E20 though it seems to prefer 0.80 sometimes tiping into 0.79's. eg It wouldn't handle 3.75 degrees of timing offset without up to 4.5 degrees of timing pull on lambda's matching the petrol tunes. But richening it up allows it to run that same advance with only a smidge of timing pull near 6000rpm.

- its not worth trying to squeeze that last bit of timing into it anyway even though 1.5-3 degrees of timing pull is considered acceptable if not a target in the dub tuning community. I don't really get why that is. If I look at a 3.0 degree added log and it has 18.5 degrees of timing in there in the upper rev range, but the 3.75 degree log did have that until it was instantly yanked back to only 15 degrees for nearly 800 rpm, then I dont see how the latter is better. Also this tune is what I'll be using on the track for lap after lap of abuse so coming back from that threshold to a log devoid of any timing pull I think is the wiser move.

- before I worked out that it wanted more fuel, it was pulling some timing around the torque/boost peak. It did this less so when I was running with the intake VVT activated. I thought that the VVT would enable better cylinder filling in this area and then as it was squeezed and banged that it would make for higher cylinder pressures than if there was no VVT. But the opposite has seemed to be the case. VVT 'on' between 2400 and 4250 seems less prone to torque peak knock - is that because the overlap it creates actually reduces the peak cylinder pressure?

- When I started out with this I'm pretty sure I was getting spark blowout. Short of LS2 coil converting it all I could do was shrink the plug gap from 28 thou to 24 thou. This worked instantly and the spark blowout that could be heard/felt is gone. But it could still be there. When it was happening it wasnt visible in the misfire detection logs which are notoriously inaccurate for anything other than idle misfires or literally dead coil pack diagnosis.
So the question is how can I look for spark blowout that might be too subtle to hear/feel? I assume spark blowout means no combustion takes place. The O2 sensor would see this as a sudden lean blip wouldnt it ie its only looking for oxygen content yeah, so a misfire will be seen as a whole lot of fresh air coming at it from 1 of the 4 cylinders right? Can spark blowout be seen in this way?
I'm reading that iridium/platinum plugs might be the go for ethanol fuels as they have a stable spark position making them more appropriate for blowout issues. On the flip side I always thought that they were higher resistance compared to copper and its all about peak current flow right which is why I've used BKR7E's and just changed them at every oil change... thoughts?

- oh and for the first time I've topped out the MAF on a few occasions. 194 g/s through a 2.5 inch stock MAF housing. Yes that was a cold dry night after a long cool cruise, but still its never gotten anywhere near that on a petrol tune with any K03 frame turbo before. Usually its hitting 185 g/s though, but thats kind of a moot point because this turbo is all about midrange anyway. So E20 blends work!

[Sydneykid]

Lots of questions Sam, hope I don't miss any.

I have seem lambda down to .75 work well, but that was on E25, so .79 on E20 isn't outrageous.

As per previous posts, I have never used a gap below 0.7 mm (28/1000), so my suggestion would be a coil upgrade if you need to further reduce the plug gap. You are going to need it eventually anyway.

In my experience more ethanol ALWAYS reduces knock, due to the cooling effect. There is a point where adding more no longer reduces the knock, which is where I stop.

You are correct, more cam overlap reduces the combustion pressure and hence the knock. But as usual you don't get something for nothing, there is a torque loss, but you do get a bit more exhaust turbine efficiency. So as usual it's a trade off, if it's in an rpm range where you need instant response then more cam overlap isn't a bad thing.

Spark blow out results in a sudden rich spike of unburnt fuel, ethanol does have an oxygen content so it's not as noticeably rich as with pump petrol.

I have always used copper spark plugs for ethanol, at the end of the day copper is better conductor of electricity than iridium and platinum, which are really there to prolong the intervals between plug changes. I have never had an engine run better on iridium or platinum spark plugs.

Topping out the MAF, now there is a good thing, more air = more power.

You know my thoughts on ethanol, best thing I ever did was to change when leaded Avgas became unobtainable from Bankstown airport and unleaded race fuel was $7 a litre (~$500 a race weekend in fuel alone). We used to mix our own E85 about 15 years ago, and then started buying it around 2008 when the United servos stocked it.


Cheers
Gary

[sambb]

Alright I might look at getting it out to E25 and do a bit more testing. I'll be buried in LS2 coil conversions now too.

Hmm surprised that a mis causes a lean condition at the O2 sensor. My thinking was that combustion consumes oxygen. If there's no combustion then oxygen levels in the virgin air remain high. I thought the sensor is not looking for unburnt fuel vapours, that it can't see them. No worries though, i'll take your word for it and now I know what to look for now so it'll be interseting to see if anything jumps out.
appreciate the help Gary

[GRP]

Where is Sam?

The only interesting ongoing thread on this forum and we have nearly 4 weeks of no posts.

Hope that all is well in Gladys land.