Lean burn-thoughts, theorys, and discussion
 

Ok guys I have a 1999 honda civic that has a programable engine managment system(crome). I have been messing around with lean burn and seeing what I can fiddle with to up the mpg. the main thing i am wondering about it is how lean is too lean.. I'm sure there is some point where leaning requires more manifold pressure and thus more fuel to keep accelerating I just wonder if its in the realm of afrs we can acheive?

I have a non-vtec y7 engine and with the manifold pressure around 10-12 inches is around 18-1. Any higher manifold pressures (14"-18") where i normally cruise are leaner than that, guessing 22ish to 1 with no indication of breaking up.. anyone know any gauge type widebands that will read higher than 18 or 20 to one? Also the factory timing in those areas of the tune are around 38-42*. In my tune i'm running anywhere from 48-50 degrees of timing.

I'm considering trying out a set of vx spark plugs(one stage hotter than whats in it now, stock) and leaning farther, but I currently don't have an instaneous mpg gauge so it makes fine tuning difficult to have to drive a few hundered miles to see what changes i've made.. I just started keeping a log on here, but over the past few months of testing i've been getting anywhere from 35mpg driving to work and home(5.5 miles 35-45 mph 5 stop signs) to a few tanks in the mid 60's doing highway trips.

anyone have any ideas or thoughts or concerns about lean burn lets hear what you think!

what I would be worried/concerned about is burning the valves with such lean mixtures and high advance. running even hotter plugs only makes it worse. Not sure there are wide bands that lean on market (other than emission type sensors ).
I would check what your exhaust gas temps are , seeing if its getting much higher than stock and seeing what plugs look like before leaning further or running hotter plugs .

The Insight goes to 25:1, so its O2 sensor must also.

that might be but most aftermarket wideband A/F meters only go to 18/1 . The LM-2 says it can go to 22/1 .

I just tested out 25:1 at < 20% throttle, and around 16% MAP load. on a d16y8. NO BUENO. Ended up melting the electrodes on the plugs. Decided it would be safer to further mod the driver, and not risk melting pistons. I would say 18:1 is about as far as you would want to take the car under light loads.

Try a set of NGK ZFR4F-11 plugs. That's what's spec'd for the '97 Civic HX, which uses lean burn. It's not the same plug as the other Civics of that vintage. Your '99 should be the same generation as the '97.

By the same token, you might try using the same primary wide band oxy sensor as that car. A friend has an '04 HX and he says he found the wide band oxy sensor for about $150ish.

If you are burning components under light loads you are not lean enough. If you watch EGT as you get leaner it will go up to peak (which is where you will start burning valves/pistons/other things) then as you go leaner (50*F at least) temps will drop off. The throttle will have to be opened more to get the same power but you won't have to be sucking air past a throttle plate, which is where a fair bit of efficiency can be gained.

So what your saying is you have to jump to a very lean mode where power feels lower so temps run cooler , so like 17/1 >22+/1 . that might be hard to do for DIY .

These engines that run lean mode , I would guess also have better components (valves, seats ,etc) to handle higher temps .
Do you know about the EGT they run ?
more than 1400F ?

from what I have read and understand, peak cylinder temps occur at 14.7-1, peak egts occur slightly richer than that around 14.0ish but anything after that should start significantly dropping, I know theres quite a few pilots on here, and theres a lot of debate in the aviation world about flying l.o.p. (lean of peak, ie leaner than 14.7 to one) some even suggest running wide open on the throttle and leaning until you acheive the correct cruise rpm.. and most of those engines are dinosaurs with piss poor cylinder head designs air cooled detonation magnets..

Does anyone know the threshold of lean burn for the vx? I currently start richining past 14.7 at around 10" of vaccum and run around a 13.8 at full throttle for power reasons... it makes 106 hp to the crank on a good day, it needs all the help it can get overtaking cars on the highway..lol

About the o2 sensor I know a vx sensor(l1h1 or a ntk sensor) will supposedly read 25-1 but the only controller i've seen that utilize them (ngk... anyone know of any other ntk controllers?) only reads to like 16 or 18-1...

Keep the thoughts coming guys!

Ben

There is a wiki started on this mod as i am looking at doing this mod myself. It hasn't really been filled out well yet but it has a few useful links to threads and external link to some good info
Lean burn - EcoModder

Also when i have been researching i came across a graph which showed the cylinder temp compared to air fuel mixture. I cant find it now of course but i do remember temps going up with a slight increase but then decrease when going leaner.

I wonder does anyone know a good method to measure cylinder temperatures?
Can a temperature sensor just be put into the exhaust stream close to the cylinder?

Just get a diesel and you can run it as lean as you want by taking your foot of the accelerator ;)

It's actually one of the first things I started doing around mid 2008.

I had seen that reducing the throttle a bit after overshooting the target speed by a few kph gave better FE than accelerating to the target speed and staying there.
So I made it into a habit to lightly and slightly lift my right foot to increase fuel economy while the speed remained constant - but overdoing it made the speed drop considerably.

Very tiring, as I had to keep my foot in the exact same position - a bit of a bump was enough to ruin the effect.

Even now with the CC, I hit the "-" button after setting the speed.
It hardly changes the speed, but it improves the (indicated) FE instantly.

I wish I would have taken pictures of the sparkplugs I pulled out of my car. I melted a good 1/16th inch off the trode! I'm interested in this technology as well.

cmj did that setup have a wideband? a vac gauge? what rpms were you turning and what af's and all that... I'm thinking about buying an egt and documenting some hard data to really get a better handle on how af's and timing affect cruise egts.... I have personally ran my car lean everywhere on the map, full throttle included(28* of timing at wot 17-1 afr's) power sucked but no noticeable ill effects what so ever on the plugs. they were white, but thats too be expected no hot spots or indications of melting..

Ben

To get the Exhaust gas temp you can do a couple different methods. One is to just put one probe, close to the exhaust port on the manifold this lets you know what one cylinder is doing. Another is to put one probe just after the header goes down to one pipe, this will get an average of what all cylinders are doing. or you can put one on each cylinder so you can see what they are all doing.

Falcon EGT Gauges from Aircraft Spruce

I removed the factory o2 sensor to put the car into open loop and to use my wideband, but I was tuning beyond it's capabilities. Timing was stock. Map voltage was reduced by 30% last wideband reading was 19:1 before reducing a bit more fuel and maxing out the o2 sensor. Average RPMs were 2200-2600. Average MAP pressure was -19.7inHG and maximum was -11inHG.

Took it for a test run and car started smoking, and stumbling. Wouldn't hold idle. I got it home, checked the plugs and found my problem. This being my daily, I decided to forgo any extreme tuning as such until I can find a wideband that is capable of accurate readings at > 18:1

Those who understand what we actually burn in our engines will find some value here:

Title: Predicting the behavior of a lean-burn hydrogen-enhanced engine concept
Author: Ivanic, Åziga, 1978-
Other Contributors: Massachusetts Institute of Technology. Dept. of Mechanical Engineering.
Advisor: John B. Heywood.
Department: Massachusetts Institute of Technology. Dept. of Mechanical Engineering.
Publisher: Massachusetts Institute of Technology
Issue Date: 2004
Abstract:
--------------------------------------------------------------------------------
(cont.) Lean operation of a spark ignition (SI) internal combustion engine (ICE) offers attractive performance incentives. Lowered combustion temperatures inhibit formation of nitrogen oxides (NOx), while reduced intake manifold throttling minimizes pumping losses leading to higher efficiency. These benefits are offset by the reduced combustion speed of lean mixtures, which can lead to high cycle-to-cycle variation and unacceptable engine behavior characteristics. Hydrogen-enhancement can suppress the undesirable consequences of lean operation by accelerating the combustion process, thereby extending the "lean limit." Hydrogen would be produced on-board the vehicle with a fuel reforming device. Since operating an engine in the lean regime requires a significant amount of air, boosting is required. Hydrogen is also an octane enhancer, enabling operation at higher compression ratios, which results in a further improvement in engine efficiency. The focus of this thesis is on the modeling aspect of the lean boosted engine concept. Modeling provides a useful tool for investigating different lean boosted concepts and comparing them based on their emissions and fuel economy. An existing architectural concept has been tailored for boosted, hydrogen-enhanced, lean-bum SI engine. The simulation consists of a set of Matlab models, part physical and part empirical, that have been developed to simulate performance of a real ICE. The model was calibrated with experimental data for combustion and emissions in regards to changes in air/fuel ratio, load and speed, and different reformate fractions. The outputs of the model are NOx emissions and brake specific fuel consumption (BSFC) maps along with the cumulative NOx emissions and fuel economy for the urban
and highway drive cycles.
--------------------------------------------------------------------------------

Description: Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 2004.
--------------------------------------------------------------------------------
Includes bibliographical references (p. 90-91).
--------------------------------------------------------------------------------
http://dspace.mit.edu/bitstream/hand...pdf?sequence=1
6.314Mb PDF

I don't post to much on here anymore. I found this thread by doing my typical once a month routine of googling "Lean Burn":thumbup:

Anyway keep up the great thread!!!

Here's some post's from the past of my radical lean burn engine that's been a work in progress for the last 11 years.


http://ecomodder.com/forum/showthrea...ign-11500.html

Good Luck with all your lean-burn adventures!!!

I tried running lean with my 85 Chrysler Laser turbo by programming the computer to ignore O2 feedback and run lean under light cruising conditions. I have a LC-1 wideband readout. It would only function as low as 17:1 and anything leaner would cause misfire. I did however get best fuel economy running 16:1 at idle at hot idle and light cruise until +2 psi manifold pressure where it descends to 11.5:1 all the way to 19psi of boost.

I always wonder why this engine won't run lean like the rest of your guys engines.

Going from some info I have picked up along the way and some reading of "Engine Management: Advanced Tuning by: Greg Banish" I would say the AFR you are looking for is around 17.64:1 (lambda 1.2) for close to the same emissions as running at 14.7:1.

For cars that have a issues when going to/from lean burn aim for 15.43:1 (lambda 1.05) according to above book. I have not finished looking over the book but just to let you know there is not much of anything on lean burn in it.

A cylinder head thats designed for power(being the car is boosted) guessing it has excessively large ports with poor low rpm flow and therfore low rpm fuel atomization, and also being boosted its probably low compression, which is another factor which makes firing off a lean mixture difficult..

My other car is a high 11 sec turbo civic, its low compression with a large aftermarket cam, and it won't fire a mixture any leaner than about 16-1.

Ben

I just pulled my 1jz (turbo) back to 15.87:1 from 15.3:1, it seems to show positive results, ice yet to try my new ignition map with it (couple degrees more) and hopefully see more benefit. Hopefully ive yet to reach the cross over point where im no longer gaining ground, still wanna see 40 mpg one day

it was not the lean condition that melted the spark plugs
 

in your car in your system
lean condition DID NOT CAUSE high combustion temperature .

your tampering with MAP voltage DID

by reducing the LOAD signal, the ECM provided ignition timing based on a lesser than REAL / ACTUAL LOAD so ignition timing was ADVANCED , TOO ADVANCED for actual conditions

if the ECM had been provided with a feedback signal from the AFR or front 02 sensor , the ECM would have attempted to compensate for the lean condition caused by reduced load signal .
BUT
the ECM would NOT have attempted correction to IGNITION TIMING advance .
peak combustion pressure was happening TOO soon , the engine was "pinging" and you either did not hear it or chose to ignore it .

...."I removed the factory o2 sensor to put the car into open loop and to use my wideband, but I was tuning beyond it's capabilities. Timing was stock.

>>>> Map voltage was reduced by 30% <<<<

last wideband reading was 19:1 before reducing a bit more fuel and maxing out the o2 sensor. Average RPMs were 2200-2600. Average MAP pressure was -19.7inHG and maximum was -11inHG.

Took it for a test run and car started smoking, and stumbling. Wouldn't hold idle. I got it home, checked the plugs and found my problem. This being my daily, I decided to forgo any extreme tuning as such until I can find a wideband that is capable of accurate readings at > 18:1"....

My experience with lean burn is not exactly the same as you fellas, but I feel compelled to chime in.
I just put a simple voltage divider on my MAF, and used a LM-1 wideband to monitor AFR. When I do this the motor believes it is at a lower than actual load and the ignition timing is advanced to it's maximum available, 40 degrees. I typically run around 17:1 AFR, but because it is a voltage divider, the AFR varies by load...ie. High Load=Leaner, Low Load=Richer. It works pretty well, but the PCM grows wise to my tom-foolery and kicks into limp/open loop after just a few miles. However it is easy to reset it back by using DFCO. This compromise method returns a benefit of approximatly 2-3mpg. I only engage this lean burn under light load cruising, but have found it is still able to pull up hills just fine. I have been considering using the LM-1 to set an offset, so I can keep the motor at 17:1 and switch between the standard O2 sensor and the LM-1 that is offset for cruising, however I would loose my ignition advance because the PCM would not know it was lean, the normal ignition advance is 32-34 degrees while curising.
Just throwing some thoughts out there...

I'm completely dumbfounded.. I can't believe I didn't think of ignition timing and ecu control. Thank you mrwebb!

I'm honestly embarrassed.

Ok I have been busy this weekend, went and got me an EGT gauge and got the bung tig welded in the header and have did *some* testing. Haven't had time to do any changes. Just been driving on current tune. for reference the sensor is in the #4 cylinder about 1.5 inches from the head.
So far what i've noticed-
Idle's around 18-1 afr egt-750
55mph cruise 19-20 afr egt 1150
80mph cruise 19-20 afr egt 1250
flat out wot 14.7 afr egt 1450 (cold weather has caused the wot portion of the map to lean out slightly, intake temp sensor is keeping the low speed lean burn sections of my map at about the same afrs.)

I have also included the current timing map that i'm running.

Keep up the thoughts and ideas guys!

Ben

Frank316 looks like some great data there.
Seems like your car is already running a bit lean since it is up at 20:1 afr (air fuel ratio). I am interested to see what results you get when you start doing some changes.

For other peoples reference i assume the acronims mean the following (since i had to look some of them up)
AFR: air fuel ratio
WOT: wide open throttle
EGT: exhaust gas temperature

You can't create REAL lean burn in a combustion chamber/piston crown design that wasn't intended for it. If you read the simple Wikipedia entry on it, you'll see the mention of stratified charges. At the really lean lean A/F ratios used in lean burn engines, you can't support a flame, if those mixtures were homogeneous through the cylinder. Lean burn engines, through various methods, concentrate fuel around the spark plug, generally. You end up with an area that is say 15:1, w/ the remainder of the chamber all air. You end up with a TOTAL ratio of 24:1, for example.

If it were just as simple as the right combination of timing, throttle opening, and fuel mapping, etc, all cars would've been set up for it!

REAL lean burn? anything leaner than a factory "optimized" 14.7-1 is lean burn... The reason lean burns not all the rage.. EPA, and $.. Lean burn does create more NOx emissions, which can be curbed but with a MUCH more expensive catalytic converter. the 3 way cats in use today NEED 14.7-1 so they have the right mix of chemicals to get rid of polluntants and not get plugged up.

Hondas solution to a homogonous mixture was to use swirl ports on the vtec-e motors, all new heads since 96 use similar swirl ports because the mixing of the fuel and air to estabilsh the "lean burn" also promote better emissions at richer air fuel ratios.

Anyways been doing some more testing, although colder winter weather, snow tires, winter gas, all have a pretty detrimental effect on mpg, car still does a pretty regular 32-36mpg better than epa average of 28 and thats about 80% city/back county driving.. and I don't really hypermile, i still drive like an idiot.. i know if i slowed down i'd get better fuel milage but ??

I have found out that there is a limit to how lean you can go.. not so much as it missing, or maybe it is and its just to hard to tell, but it doesn't seem to do as well if i lean it WAY out(past 20-1) and have to apply more throttle, than if i leave it slightly richer.

The other thing is that at around a 20-1 af at a 55mph cruise, since the temp is below 30 outside, simply turning on the heater causes the coolant temp to plummet, I really think the thermostat isn't even opening on the highway the motors running cool enough to not have too. The egts at 20-1 are lucky to be over 1050, at 14.7 its over 1300. I also have the start up tables leaned out too, but it still has to be slightly richer when the motors dead cold to get it to warm up. I need to go a grill block, but like i said its been below 30 and i'm not that hardcore..

For those still reading i'm gonna put some of my warm up air fuels and what not just for reference..
Start up dead cold from 25* or so it idles around a 14.7-1
Slowly leans out to about 18-1 at idle when warm
Crusing at part throttle low load is 18-1
Cruising at 70-75 it dips down around 14.7-1
Full throttle drops to 13.6-14 for peak power

Block radiator in winter
 

>at around a 20-1 af at a 55mph cruise, since the temp is below 30 outside, simply turning on the heater causes the coolant temp to plummet

Block off part of the front your radiator with cardboard or sheetmetal.

...GOOGLE on "HCCI" or Homogenious-Charge Compression-Ignition.

...diesel-like operation at low-speed cruising with reversion to spark-ignition for controlled burn periods during accelerations.

...here's an old GM test article:

http://green.autoblog.com/2007/08/26...s-hcci-engine/

HCCI engines are pretty cool in theory but would require a LOT of engineering to get it right, not to mention a heavy ass block to withstand the constant abuse of "detonation" the main principal there operating under.. Lean burns much more of a practical application.

I'm currently working on building a new engine for the beater. It all used parts that people were planning on throwing away or I had laying around. It'll be around 10-1 compresion with flat top pistons, a head that has much larger quench pads and much tigter piston to head clearence than a normal engine would have, it'll also be slightly less displacement (1.5L vs the 1.6L) thats currently in the car. It will have a vtec head that has a "smaller" camshaft profile than whats in the car now (when vtec is not engaged), and hopefully when in vtec it will make a little more power.

Ben

A goodly chunk of the increase in efficiency of LOP is that you are opening your throttle farther for less vacuum.

Frank316, do you have an instant mpg readout to see what your mpg is at a 20:1 AFR at 50-55mph?

No I do not, been thinking about building myself a mpguino, but haven't got around to it yet. And I relize running leaner lets you open the throttle more and decrease pumping losses, but it gets worse gas mileage past 20-1 than it does below it, not sure whats up with that but don't have the instrumentation to see everything thats going on. The cars converted to run obd1 so I can't use a plug n play scan-gauge or any of the like, the obd2 port doesn't work.

Ben

Any update Ben ?