Closed loop fuelling and methanol injection

[Jardon]

Hi all,

I have been running my 1s kit for a year now and finally I get it (I think). My car performs well though it has not been on a dyno recently to document anything. It will pull 60-100 mph in 3rd gear in 5.5 secs (even quicker in winter). My reason for posting is just to tell my story really as misunderstanding the available information has led me all over the place in the last year.

My car has Bosch ME7 engine management and as such uses lambda controlled closed loop fuelling at all loads. ME7 also uses allows much finer control of timing when considering knock than previous systems run by Volvo (Bosch Motronic 4.3 and 4.4). I run a lightly modified set up with 1.45 bar peak boost falling to 1.15 bar at my shift point (6000 rpm). This gives peak torque around 4000 rpm - thats where my clutch slips anyway - and boost starts to grow quickly from around 2700 rpm.

My lambda readings are always around 1.0 at part throttle but drop to 0.73 once boost has risen past 1.0 to 1.1 bar. IAT's will reach 60C in winter on a single 4th gear pull without aquamist but never go above 40C in summer with it activated (be it water or methanol). Lambda readings are the same with no aquamist as they are when I inject water or methanol.

I had been concerned that Lambda 0.73 was power sappingly rich (if you do a lot of googling then that's the view of a lot of people). However - some folk say that lambda is application specific and best reliable power may happen with pump gas at say lambda 0.7 or 0.9 depending on the set up (turbo charged) ie: there is no lambda value that will be best for all. Absolute Lambda readings are only part of the story and what happens to fuel in the cylinder (speed of burn, etc) will depend on how homogenous the air/fuel mix is, how much boost and how high in cylinder temperatures are.

In January I was running a 0.5mm jet and 100% methanol when we added 3-6 degrees of advance across the map. The difference in performance was great and the the ecu registered no knock/no retard. Injecting water instead of methanol or a mix instead of methanol makes a big difference - I get flatness with a 0.5mm jet and hesitancy with 0.7mm when running water (This is activating at 0.9 bar). On methanol only I can activate the 0.7mm jet as low as 0.1 bar with no hesitation and performance is even more responsive.

All these variables being chopped and changed but the Lambda readings do not waver - and actually here's why I posted but the above is a background. I have asked lots of people if closed loop fuelling systems like ME7 will trim injector duty to hit the ECU programmed Lambda if I inject Methanol but nobody seems to have a solid yes or no. My numerous Lambda logs and noting effect on performance with various mixtures tell me that yes it does. In my application running a lower Lambda together with some knock suppression and lots of advance is working well.

It is worthy of note that I tried running a leaner map with less ignition advance after the January tweaks and performance was weaker. Simple really but thought I'd share my experience and maybe get some feedback.

[Richard L]

Water only injerction should just as be as effective with you ignition system can cope with it. Since water haa twice the latemt heat of methanol, you only need to inject helf the amount.

Methanol will natually burn when ignited so sparket energy will not be taxed. Imaging burning a wet piece of paper, heat is being taken away as the water evaporates.

Reduce your jet size for water or runM50:W50 for all season performance. The ME7 will find the correct fuel and ignition location for suit your wmi injection setup. But if it goes outside the control area, you will have negative results.

We have seen this in the newer Bosch ECU's. In the old days you hand closed loop running and open loop at WOT, with fueling at timing maps for each.

Now, with the closed loop operation at all times, in some applications there is what we call a Lambda map as well. You can go in and change the fueling maps and the car will still run to a certain lambda value if the lambda tables are not changed as well. There are literally dozens of maps in these ECU's that can be massaged to adjust the fueling/timing/spool/peak boost/boost curve/etc.

Essentially what is occurring is the ECU is pulling fuel to the target values in the lambda map, within certain limits.

As you have said, each car and application is very different. Some like running at different lambda values than others.

[Jardon]

Originally Posted by Richard L

Water only injerction should just as be as effective with you ignition system can cope with it. Since water haa twice the latemt heat of methanol, you only need to inject helf the amount.

Methanol will natually burn when ignited so sparket energy will not be taxed. Imaging burning a wet piece of paper, heat is being taken away as the water evaporates.

Reduce your jet size for water or runM50:W50 for all season performance. The ME7 will find the correct fuel and ignition location for suit your wmi injection setup. But if it goes outside the control area, you will have negative results.

Hi Richard,

that hasn't been my experience - my guess is that water AND low lambda slows the burn too much but methanol contributes as a fuel (a high octane one too) and petrol is trimmed out. With water the petrol volume injected remains unchanged so the water effectively makes the mixture uber-rich. Running Vpower 99 octane and 10-15% methanol injected at 116 octane gives a combined octane of 101-102 by my maths and I will revisit the ignition advance some day - would need to be summer. I like the idea of experimenting with water and ignition advance though.