Lean burn engine (E85?)
 

Hey, I decided to finally register here after a while just reading threads and posts that I found interested and perhaps here I can get and share more information related to what I think would be 'the key' for best fuel efficiency on engine modifications; lean burn.

Lean burn is an old invention, and what I've understood so far it could be possible to run with lambda = 1.40...1.60 if the combustion chamber, ignition and perhaps some amount of egr.. help enough with burning that lean mixture.

E85 is interesting atleast in my thoughts, and I've been running most of my vehicles with E85. Quite hard to cause engine knocking, works with ridiculous compression ratios and is cheaper than gasoline - but not enough to save any money if the fuel consumption is 25..30% more than with gasoline like it usually is.

I have been testing my current daily driver; BMW 523 E39 '99 w/ 5spd manual that I actually bought because it's easy and cheap to get ecu tuning without replacing it with aftermarket ecu (megasquirt or etc) or buying an emulator.
I have been driving around 6000km with only E85, and 2000..2500km of it with lean mixture while cruising at low load, so far no problems.

It's running on lambda=1.15 (+/- 0.05) and ignition timing at low loads is around 10 to 16 degrees more than factory tune.
2800rpm on 5th gear (100km/h gps speed) , low load = 45-48 degrees BTDC. Yes, needs a lot longer gear ratios.. way too much rpm for a 2.5 liter straight six engine.

Fuel consumption is currently about 10% more than with gasoline (95E10), which is quite good at this stage.

An interesting note:
- while cruising around 100km/h (gps speed) for more than 20km, the engine suddenly starts accelerating with same throttle amount, and I'm able to reduce the throttle so the current fuel consumption reads only 7-8 liters / 100km (true fuel consumption about 10% more than it reads) and then after a while it starts slowing down back to normal.

My thoughts are, it starts pre-igniting or it's getting fuel fumes from the tank, because of EVAP.
Spark plugs look good, no signs of knocking etc. I should actually be worried.
Pre-ignition may not be dangerous at part throttle, low load.
But I'm not the right person to tell that.. full throttle, yes - spark plugs gone. :D

There are some faults on the engine (and car) that should reduce fuel consumption more when fixed (hopefully soon), but after that I'm interested on going for more advanced modifications than mostly just ecu tuning.
I have a cylinder head from 2 liter variant, which has smaller valves, ports and combustion chambers.
I've already started modding it; welded combustion chambers and took the valve seats out. Bigger valve seats and valves coming, so it's possible to make the ports better.. intake port requires the most work since I'm trying to change it for 'high tumble flow'.

On factory lean burn engines, are there some clear differences in combustion chamber? Atleast Honda Insight has valve disable function(?) and intake ports made for swirl flow; which is quite odd for a 4 valve head (per cylinder) but understandable when it's disabling the other intake valve making it act more like a 2 valve head.
So in my case it's clear it's best going for tumble flow.

Compression ratio is another thing. Because I'm running the engine on E85, it's possible to benefit of quite extreme compression ratio.
I'm targeting to around 14 to 15:1 ratio, which should still work with the small cams.

Any thoughts/ideas?
If I find the previous threads related to lean burn engines with useful information, I'll post their links here.

Is running E85 actually cheaper than gasoline in your case? In the US, if you're not in a ethanol producing area, E85 is not much cheaper than gasoline. Definitely cheaper than 100 octane or xylene/toluene though :)

As you said, standard spark ignition engines seem to be able to run at around 1.4 lambda with the right adjustments as you said. Using a small amount of hot EGR essentially heats up the air, while the excess air is akin to cooled EGR with better combustion characteristics (you can read papers about mixing hot and cold EGR). Another thing which may or may not be suitable for your car is to run hotter coolant temperatures, as your car has port injection so it'll directly raise the temperature of the fuel disproportionately. I heard horror stories about BMWs of that vintage blowing head gaskets and radiators so I don't know if this is actually a good idea.

If you're raising compression ratio by welding material, I would go a little lower because you probably are interfering with the flow and cooling characteristics, maybe like 13:1.

There are a few ways to run even leaner (like >2 lambda) which you can't really do on your own:
1. Stratified charge direct injection (I believe some Audis used this in Europe for a while, presumably when emissions regulations were more relaxed)
2. Higher energy ignition methods, e.g. laser or prechamber jet or "plasma", these can get you past 2.5 lambda, pretty crazy.

Finally, I think you likely have some significant gains to be found elsewhere, like taller rear tires, underdrive pulley for power steering, thinner fluids, drag reduction, etc.

E85 is around 1,245€ per liter and 95E10 is usually 1,65€ per liter.
So yea, it's right now a bit cheaper. It's fairly well available in my country nowadays, but prices have gone up making E85 in some cases actually more expensive fuel, expect if the fuel consumption is less than 15% higher compared to gasoline.

This is M52B25TU engine, so it can be running even at 102C coolant temperature, but the fan will be then used fairly often.
Normally running 95-97C while cruising 100km/h.
The temp drops quite quickly to around 85-87C when going high load (electronic thermostat).
Oil temperature is about 100C after a while driving.. not sure if I could try 0W20 oil, or just 0W30.. currently 5W30 and there's no really ethanol mixed in it because I drive quite a lot.
If I knew what kind of bearing oil clearances there are, and what the oil pressure is, I would know if it would work without issues on 0W20.

I had another six cylinder engine with 14:1 compression ratio (+20bar cranking pressure) running on E85.. with almost as small cams the BMW has, but without variable valve timing (BMW has vanos for both cams).
But it wasn't for fuel economy, even tho I noticed it was quite fuel efficient compared to it's factory condition.
I couldn't cause any engine knocking, even when I had too much ignition advance (my own mistake) and on the lean side mixture for full throttle / high load.

I do agree with that, I have a lot more where I can improve fuel economy.
Longer gear ratio is on the list (6spd swap + 2.93 vs. 3.23 rear differential ratio) and full undertray (missing two pieces below engine and transmission).

Welcome to the forum!

I'm not aware of combustion chamber differences. I know all lean burn Hondas use a high swirl design, and they can lean out to something like 1.7 lambda and run smoothly.

The Insight also doesn't fully disable one valve - it still opens it fractionally, to prevent fuel from pooling on the backside of the valve. It's really the asymmetrical valve opening that helps. And, above ~2800rpm, both valves become locked together and open fully, to allow the engine to breathe at higher RPM.

I wouldn't worry about knock at moderate to low loads. Even if it does knock (and it isn't likely to), it isn't likely to be dangerous, with cylinder pressures being so low.

What method did you use to determine your ignition timing?

Thanks!

No clue if they have tried tumble flow.. or did they just go and try only high swirl design.
Probably the high swirl design would work the best when fuel economy is on highest priority.

But aiming for tumble flow wouldn't lower maximum power output, since the flow capacity won't be as much restricted.

I'm still looking for a good method, to determine best ignition timing on part throttle, but first I logged mostly injector pulse width, ignition timing, throttle position and engine load (mass air flow sensor), plus rpm of course.
The lower injector pusle width was a sign, that more ignition timing worked better on the same load / speed etc.
But since I'm going with lean burn, I have to keep my eyes on engine load and throttle position. Less load and throttle = better, since that'll mean also lower injector pulse width = better fuel economy.

Higher intake air temperature, and coolant temperature at higher speed would be better.
But I haven't really figured out how I could get higher intake air temperature working at lower speeds, since then they'll be a lot higher.

I'm quite familiar with ethanol as a motor fuel. Running lean might not be so bad it you're willing to avoid much changes to the engine itself, considering ethanol already cools the intake air stream further than gasoline. On the other hand, it may not increase performance to the same extent a more usual and higher AFR would do.

I'll get back to testing after I have replaced my thermostat.
Noticed a few days ago it took quite long time to reach the half way of temperature gauge, and engine temperature sensor tells maximum temperature is only 85C.. should go up to 104C if oil hasn't warmed up before that to around 95C.

That doesn't do any good for fuel economy, and won't allow me to see how much I could raise the intake air temperature by raising coolant temperature (about 96-98C while cruising, maybe higher).

I'll check what the afr is while cruising currently, when I swap the catalytic converter headers to catless. I'm not really sure if I can improve my fuel economy with that.. atleast the noise should reduce a lot, because one of the headers is cracked.

But I have an idea why would the very bad flowing catalytic converter headers be better for fuel economy:
They cause back pressure near exhaust port, and improves 'egr effect' so there isn't need for an external egr.

Hoping of course it either won't effect at all or improves fuel economy.. atleast then I have motivation to fabricate 6-2-1 headers, which is possible when I build a jig with the catalytic converter headers.

No wonder the dedicated-ethanol cars from the '80s and '90s had an auxiliary gasoline tank for cold start, and some also had a thermostatic control to the air intake in order to decrease warmup times. Thermostatic valves set to open at a higher temperature and intake manifolds heated by the engine coolant used to be a common feature back in the day when carburettors were the norm.

Yea, cold starts and cold climate is sort of a problem for ethanol.
Here in Finland temperatures can be something like -25C to -20C on winter (and a lot more colder when going north), but I have always got my E85 cars started.. even without pre-heating (not possible always).

Because knocking is not a problem really with ethanol, atleast on fairly well designed and modern engines, I could try some ideas to warm intake air temperature while cruising.

Maybe I could fabricate another inlet for airbox and a sealed flap between cold and hot air inlets (from exhaust headers).. and control it with a cable untill I figure out if it works or not (for fuel economy).
Like atleast Volvo had on their '90's cars atleast, but it was thermostat controlled (and nowadays most are broken = stuck at half way or for the hot side).

Something like 40C instead of 15 to 18C intake air temperature could help. Maybe even more, I don't know without testing.
On a hot summer day over 40C intake air temperatures would probably be normal, since the engine itself heats intake manifold quite well plus the more than 20C outside temperature not cooling much.

E85 is not available in my country, with E96h (hydrated ethanol) having been mainstream instead. Older dedicated-ethanol and earlier flexfuels resorted to an auxiliary gasoline tank, but since 2011 it's falling out of favor while heater elements powered by the vehicle's own battery and placed either at the fuel rail or right next the injectors became a preferable cold-start aid. Needless to say it only applies to port-injection engines, while direct-injection ones don't resort to any cold-start aid to enable operation with ethanol.


You mean something comparable to the Thermac valve?