Etanol.nu

Hi All,

first: thanks for the nice forum and site! I've reading a bit with the help of Systran, but it is rather funny at times .

Well, my "problem": I'm thinking of buying a (used) car with a 1.6l Audi/VW FSI engine. This would be the older type that runs in "stratified", lean or homogenous mode depending on load conditions and comes with odd shaped pistons to get a very lean mixture close to the spark plug [1], [2].

What do you think are the chances on running this on 100% E85?

On the plus side there's enough fuel pressure from the DI system, the compression ration ist pretty high (so it might like the octane of E85). And perhaps (just wishfull thinking ) it might just work without mods since the ECU system needs to cover a large range anyway and *might* not make use of fixed air / fuel ration maps when running under high loads. But no idea really...

On the down side I'm not sure if the tumble flap / wall guided injection will work at all with a different type of fuel when running stratified mode (or it could just run un-strated all the time, no idea how the ECU works at that stage). Are there problems to be expected with lubrication of the injectors or the fuel pump? The high pressure pump is a pretty expensive part (~ 480 € IIRC)...

Well, as you can probably tell I'm by no means an expert on fuel systems or E85 so forgive me if this post is plain nonsense .


[1] http://www.sae.org/automag/techbriefs/07-2002/
[2] http://www.kfztech.de/kfztechnik/motor/ ... tumble.jpg

me, too
but in the German forum http://www.motor-talk.de/forum/alternat ... -b258.html there is a user who have a car with a FSI engine, and he has driven more than 30.000 km with pure E85.

his nick is Dr. Holiday. here one of his statements
translated to English (analogously): Hi E85 fans, just an info: my A3 runs under pure E85 since 30.000km. Until today no problems occured.

maybe you can contact him, and i will send him a message

_________________
best wishes from here
Volvo V70 II on 30% E85 (just for starts) and LPG
Berti
http://www.ethanol-tanken.com
http://www.spritmonitor.de/de/detailansicht[url]/349366.html[/url]

Here in Sweden we have no example known of E85 use in the FSI-engine.
As you say, the stratification could give unexpected results.
The report from Germany seems encouraging though.
In Sweden we have many examples of E85 use in the DI-engine from Mitsubishi, used in Mitsu cars and some volvos (S40,V40 Mk1).
It seems to work pretty well, only issue seems to be the soot build up in different area which could look as problem coming from the ethanol but until now seems to be symptoms getting worse on E85 but wich should come sooner or later on petrol too.
We have seen no report from failure in a fossile-(sot that is)cleaned DI-engine, only from old engines with a lot of soot in them.
Once the petrol use is ended, these engines will not build up any new soot when running on E85.

_________________
http://direktdemokraterna.se
Myndigförklara dig själv.

Isn't this the engine his talking about?
http://etanol.nu/forumrecover/viewtopic.php?t=412
(unfortunately in swedish) Seems to work OK.

_________________
-Magnus-
"Six months in the lab will save you an afternoon in the library"

I think so too.

It works ok and the engine gets much more low rev torque but the car has difficulties to start under +10 degrees and the check engine light lights up with more than 25% ethanol he writes...

/Aryan

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352

Wow, you are quick -- thanks to all for the input!

Had not read of soot related problems in context of E85 usage, so I'll keep that in mind. Could you find a cause for that problem? Possibly some grime getting cleaned out of the system by the alc?

The link to the German forum is very interesting, sounds like his A3 is coping really well. I also read the post concerning the Golf via translator -- so it works but the check engine light comes on at some level of ethanol mixture? Would not be that good if this was due to lean running.

Other issue is that this G5 and the German A3 both use a newer version FSI that is not doing anyting stratified anymore, just "normal" DI. VAG probably wanted to get rid of the expensive stuff for Nox cleaning and kept the "S" in FSI for marketing reasons.

A part from the big question mark about stratified mode, do you think the plug on converter boxes work with FSI engines? If I understood the operating principle correctly, one type of box sits between the original ECU and the injectors and uses the available lambda signal to see if things go to lean. When that happens, injectors are kept open longer to get a richer mix. Sounds interesting because that should work with various Exx mixes -- so far only two fuel stations in the area carry E85 (both already in your database in case you wonder ).

Exactly so, the ethanol cleans the engine very well. We've dismantled sevral engines that had run on E85 after previously driven on petrol for years and they all had lost the majority of carbon built upp. Which is a good thing but on the Mitsubiti GDI that builts upp a lot of soot in the inlet manifold (and needs a cleaning every x tousend km's) it can be too much all of a sudden. Probably slowly increasing the amount of ethanol would solve this and possibly it is not a problem in the VW engine at all.



No that is true, but it _could_ also mean that the ECU is only complaining about needing to adapt more than usual but still function normally, this happens in some cars too. To know for sure you need to measure lambda values.



Wasn't it because it was hard/impossible to compile with tougher exhaust regulaions where NOx needed to be lower? It is quite "unfair" that diesels are allowed to have 10x more NOx in their exhausts, no wonder they are more fuel efficient, nevertheless an ethanol (cools better) engine with a lot of EGR should be able to lean burn with low NOx emmisions.



Good question! Don't these FSI engines use a wideband lambda sensor instead? I 'd think that these piggyback systems can work well as long as the do not attempt to keep the engine around lambda 1 all the time. Vielleicht mahl bei http://alcoholix.de/ nachfragen und hier wieder melden!



Good and the price is correct too?

/Aryan

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352

Jup, that sounds reasonable, thanks.

Ah, ok. Would also be interesting to read out the error codes or do some logging with VAGCOM. I wouldn't like to do that with a test drive car from the dealer though since that would require pumping in and out some fuel . An additional labmda meter sounds like a good idea as well!

Oh yes, Euro5 could be another reason. And I think our car reviewers are a bit to blame -- they generally drive like maniacs and the little fuel economy possible just did not show up in tests. Diesel is some sort of national religion here, the local automakers are just starting to support ethanol for the local market and I'm sure we'll only get the expensive high tech stuff. It's quite ridiculous since VAG has had the FFV tech for ages in Brazil and I'm sure an E85 Fox would sell really well. But you probably are aware of all that .

The EGR part sounds interesting. In stratified mode these engines seem to pump in quite a bit of EG [1] and if NOx is kept low that would mean less regenerating of the cat (there's a NOx sensor behind the cat). Apparently the engines were specced to 98 Oct fuel only because at the time of release that was the only way of assuring low sulfur fuel and using higher sulfur stuff meant not reaching the fuel consumption target due to more frequent cat regeneration.


The tech docs I found so far [1], [2] confirm the use of wide band lamda sensors. Despite the "clever" talk my knowledge is theoretical at best -- the wide bands are used because a conventional sensor would not be much use with the "on purpose" lean running going on. They are not precise enough at lambda >1, right? So my question about using a piggy back system with lambda input was pretty stupid, especially if one wants to retain the lean burning. In fact these things should not work well, with the ECU aiming for lambda >1 and the piggy back prolonging injector time to get it back to 1. Just like my wife arguing about directions with the sat nav .

I've checked the alcoholix page and they explicitly say it *does not* work with FSIs. But I've googled around a bit and found [3]. They claim to support it. I'll ask them about engine type I'm considering.

I hope this is correct, but if I get it right, this type of engine runs with lambda >1 under certain load conditions, with other loads it switches to lambda=1 and at WOT it probably goes a bit rich. So it must be taking some sensors in account (throttle position, MAP) and then look up or compute a target lamba and injection time. I wonder if it is in closed loop all the time or if they use fixed values. While it's in closed loop it just seems to be a question to keep it in a window where it can sort itself out, at least as long as the injectors can supply enough fuel. All a bit confusing for a newbie...


Haven't been there in a while but the prices sound about right. I'll update when I never/better data.

One last thing (soooory for the long post) -- actually the engine in the G5 and the A3 could still be set up for lean operation. I was wrong about the time of the change to homogenous operation.

Could someone ask the poster with the Golf the engine code [4] (should be on a sticker in the manual and in the boot)? They changed to non stratified during the production [5]. I'll ask the A3 owner.

Cheers and if you made it this far big thumbs up for your patience !
David

[1] http://home.wirehub.nl/~eygauto/Touran- ... rkette.pdf
[2] http://www.ngk.de/Neue_NTK_Lambdasonden ... 38.98.html
[3] http://www.rmg-rapsol.de/html/fsi__gdi__direkt.html
[4] http://www.doppel-wobber.de/wbb2/golf_info.php?id=140
[5] http://www.a3-freunde.de/forum/t74935/8 ... 6-FSI.html

Basically you are right.
The only problem is that with an DI-engine, the timing for injection can be critical if we are talking late injections, just before start of combustion.
If the injectro pulses are prolonged by a piggyback system by say 25%, the majority of the fuel amount can come to arrive at a too late time in relation to the ignition spark.
This could lead to a too late combustion (which might be not programmed to take into account by advancing the injection+ spark to an earlier crankshaft degree to the next cycle) and/or even worse, a missed combustion if the fuel cloud "misses" the spark.
All this leads to the conclusion that the most safe way of converting to ethanol fuel would be to increase the fuel pressure to keep the injection timings as unchanged as possible.

On the other hand, we have seen good results from the Mitsubisch GDI-engines anyway. unconverted, these engiunes seems to adapt very well to a larger fuel need.
If the same capability is in place in the FSI:s, the injections are probably already programmed to take place in a well working timing window.

The best way is really to test with higher and higher concentrations of E85 in steps of 10-20% units until the engine protests or emit CEL.
If 100% E85 can be reached without any problem and a warm engine, we can make the conclusion that E85 or similar oxynegates are already taken into consideration during development of the engine.
And there is a rumour that this is the case with the FSI:s...

_________________
http://direktdemokraterna.se
Myndigförklara dig själv.

Hm, yes -- thanks a lot for pointing that out! The late injection happening in this type of engine is something I overlooked. It's probably enough of a gamble to play around with fuels that have different combustion speeds. Not a good idea to also mess with injection time.

I'll have to study a bit more to understand how the fuel volume is actually controlled here. There could even be a simple solution by software (Lemmiwinks?) [1] or by fooling the mechanism that regulates the pressure in the fuel rail. The LPG people are having more or less the same problem with FSI engines and there finally are solutions coming out next year. Perhaps I can just steal from that .



Sounds good! BTW, speaking of Mitsus, is there anything at [2] that could be useful for you? The performance crowd in the US has discovered E85 and I think they made quite a progress playing around with PC software and the ECUs of their cars. No Idea if the EVOs use totally different electronics though,



Yup, that's the plan. Problem is that I haven't got a car to test with. I'd really like to learn more about E85 compatibility before buying the box (an Audi A2 1.6 to be more precise). The German guy with the Audi A3 was so kind to reply to my questions: his car is fitted with the more conventional BLF engine which is DI but lambda = 1.

Would really be easier just to get a nice old Golf II for the winter and stick a washer under the FPR .

[1] http://etanol.nu/forumrecover/viewtopic.php?t=412
[2] http://www.openecu.org/index.php?title=Downloads

This Lemmiwinks http://www.vwfixx.com/forums/lofiversio ... 22492.html I'd never heard off, it is very very intresting stuff for VW/AUDI owners!

One should easily be able to convert the car for E85 with good cold start capabilities or even full FFV!

For E85 I's start with changing:



I'd put it to 200% you need a lot of extra fuel on cold days, during summers or when the engine is warm I suspect this fnction is not used?



150-200% there too? Maybe itis better to add a (NTC) resistor instead to the coolent temperature sensor though, making the ECU "think" that the engine is colder and thus prolonging the starup enrichment time.



Id start with 130% here
112% should theoretically be enough to raise the fuel injection time to 140% if you put +24,8% in step 10. Still for FFV use and a good fast lambda sensor the higher the better?



I'd shift down 8 degrees and see how that works



+24.8% would be probably enough to run the car ok on E85. Theoretically you need +40% but the short time fuel map and some more cold start enrichment should be able to fill the gap.

Mmm very intresting, unfortunately I do not have a VAG to try with. A Audi A2 or a VW Lupo could make a very fuel efficient Ethanol car, escpeccially with raised compression.

/Aryan

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352

Certainly!
Hey, any Mitsu GDI-fan here?





Oh!!!!
Why is only Vag-com available for VW???
The crack of vag-com can't be overestimated, and this download, what can I say? -GREAT STUFF!!

_________________
http://direktdemokraterna.se
Myndigförklara dig själv.

Great stuff indeed, although I haven't got a VAG car I played around a little and put in E85 values:



Dave70 thank you very much!

To answer your questions about this Golf 1,6 FSI :

It has a BAG type engine and the longterm fuel map goes to its max value with E65.

In order to get overcome coldstarting problems he put in 4.7Kohms NTC resistor in series with the original engine temperature sensor (the one on the drivers side o the engine) with a switch over the NTC so he can turn the cold start enrichment on and off.

/Aryan

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352

Well, thanks again to all of you for the input! A great help.

The engine code BAG is very good news since this should be one of the lean burners, similar to the A2 engine. Using an NTC with a switch is pretty clever!

Just a *big fat* disclaimer about lemmiwinks: I've done a bit more reading and it seems they have a focus on VAG engines with a certain type of Bosch ECU [1], notably some of the 1.8T engines and stuff used in faster Audis. So this is probably not nearly as compatible as VAGCOM for example. For fun I'll try it on our Lupo but I'm 99.99% sure it's not going to work because the car has a Simos ECU.

[1] http://s4wiki.com/wiki/Lemmiwinks

Aha it already sounded to good to be true, still very intresting for drive by wire VAG cars!

/Aryan

_________________
Komplettera gärna data om dit piggybacksystem här: http://www.editgrid.com/user/aryan/E85_piggyback_system och här viewtopic.php?f=10&t=6352