34:1 Afr?!

[MAP]

Greetings,

This article in the latest "Design News" amazed me:

Thirty-Year Quest for Lean Burn - 4/28/2008 - Design News

The embedded links make for fascinating reading too.

I had heard Mr. Vizard often speak of 17-20:1 AFR's for good economy, but now 34:1 with a 40% gain in efficiency? Now all we have to do is get rid of catalytic converters, it would seem...

Best,
Mark

[MAP]

Greetings again,

I thought this would work better as a separate post rather than a cumbersome edit to the opening one.

Question: in trying to improve economy, I contacted a few carburetor manufacturers and tuners a few months ago about trying to get 17:1 AFR in cruise. To the very last one, they grumbled that lean surge and tip-in stumble would be the bane of such an attempt.

Do I presume correctly that fuel-injection (the Mass-Flow vendor on this site comes to mind, for instance,) is the ticket out of these difficulties, by artful adjustment of the fuel and ignition maps? Or is that a speed-density (MAP sensor) system trick only? And can any single injector handle the very broad range of duty cycle that would be required, to provide the very lean allotment of fuel at cruise, while being able to support full output at WOT and high rpm? In combination with gearing and chassis refinements, and now reading about the possible benefits of very lean AFR's, I'm wondering whether 460 hp peak from a 383 SBC, with 30mpg cruising at 60mph, is out of the question. (I count myself very fortunate indeed in that there are no emissions requirements where I live.)

Best,
Mark

[Pinhead]

I've always wondered why lean burn for economy motors has been so hard. With fuel heat and hydrogen enrichening, lean burn is not nearly as much of a problem. Of course, hydrogen enrichment would be a PITA for the car manufacturers, but fuel heat would be very simple for them to incorporate. Also, with COP's and the hotter spark that comes with it, I would assume that lean burn would be a given.

[MadBill]

The biggest problem with lean burn in modern engines is emissions control. Until recently, it just wasn't practically possible to reduce Nox to acceptable levels except by running near stoichiometry. European direct injection gasoline engines cannot take advantage of their extreme lean burn capability when calibrated to NA standards and thus suffer significant fuel economy loss.
The latest diesel technologies must deal with tighter Nox standards, and do so with complex strategies like urea injection and I assume this could be applied to a lean burn gasoline engine, but I'm not aware of any that use it.

BTW, in the eighties I was calibrating GM engines to the then more lenient Canadian standards, and often used cruise AFRs in the 16.5:1 range, along with as much as 30° vacuum advance, Certified Highway fuel economy was as much as 17% better than the US versions running 14.7:1.

[DavidVizard-GFN]

Originally Posted by MAP

Greetings again,

I thought this would work better as a separate post rather than a cumbersome edit to the opening one.

Question: in trying to improve economy, I contacted a few carburetor manufacturers and tuners a few months ago about trying to get 17:1 AFR in cruise. To the very last one, they grumbled that lean surge and tip-in stumble would be the bane of such an attempt.

Do I presume correctly that fuel-injection (the Mass-Flow vendor on this site comes to mind, for instance,) is the ticket out of these difficulties, by artful adjustment of the fuel and ignition maps? Or is that a speed-density (MAP sensor) system trick only? And can any single injector handle the very broad range of duty cycle that would be required, to provide the very lean allotment of fuel at cruise, while being able to support full output at WOT and high rpm? In combination with gearing and chassis refinements, and now reading about the possible benefits of very lean AFR's, I'm wondering whether 460 hp peak from a 383 SBC, with 30mpg cruising at 60mph, is out of the question. (I count myself very fortunate indeed in that there are no emissions requirements where I live.)

Best,
Mark

Mark,

You are bringing up some good points here. I had some conversations with carb manufactures on just this issue. It was hard for them to accept that a good ol' SU carb could deal with this very lean cruise situation with no drivability problems when transitioning to a wider throttle opening.
I think I am going to talk to Edelbrock on this one as I think there is potetial to run the cruise mix leaner on their carb without running into a major re-design to get results.
Let me think on this for a while and I will see if Edelbrock have anything to offer here.
DV

[Devious]

Thanks for the above article, it is very interesting.

Originally Posted by MAP

Do I presume correctly that fuel-injection (the Mass-Flow vendor on this site comes to mind, for instance,) is the ticket out of these difficulties, by artful adjustment of the fuel and ignition maps? Or is that a speed-density (MAP sensor) system trick only? And can any single injector handle the very broad range of duty cycle that would be required, to provide the very lean allotment of fuel at cruise, while being able to support full output at WOT and high rpm?

I am no expert on EFI, but I do have a fair bit of experience with a good number of engine management systems. IMHO, mass flow is generally good at improving resolution at low throttle openings, but not for serious performance. However the system from Mass-Flo looks like an easy system for those wishing to convert a carb engine to EFI with minimal injection mapping required. But this type of manifold tends to give away a few of the better charicteristics that EFI can offer over carbs when the intake system is designed specifically around EFI.

I think with normal EFI injectors, and the poor atomization at low fuel pressures, leaner AFR's may require staged injectors (per port) and considerably higher fuel pressures to cover the required ranges. I have had good results on a couple of bike engines using higher fuel pressures and MAP based EFI systems with and without staged injectors with ITB's and PICO injectors. Staged smaller injectors do seem to work better in my limited experience. Also, my experience with marine engines using normal Bosch injectors at pressures of 62-75 psi (and correct injector timing for the engine's needs) has shown a good reduction in fuel required for best power at WOT, as well as being able to run considerably leaner at cruising conditions.

However, I believe port injection is not the final answer. IMHO it will require direct injection at even higher pressures. Look at how well Bombardier has been able to clean up two-strokes and improve mileage compared to four-strokes with their DI technologies.

As a related side note, I own one of the Latham (axial flow) superchargers built by Richard Paul in the late 1980's and early 1990's (now know as axialflow.com). This is a larger, redesigned unit compared to the units made by Norman Latham in the 50's and 60's. It is able to flow 900-930 cfm at design speeds and pressures - not huge, but nice on a small capacity, mild mannered street engine in a light car. This unit works great as a draw through system (using Weber DCOE's or EFI), as it tends to offer very good atomization by beating the air/fuel mix between rotor and stator blades, and tends to even frost up after being run hard at 9-10 psi.

Interestingly, on two different engines, I have been able to effectively run considerably leaner AFR's at part throttle cruise than any other form of forced induction I have run - while using higher than normal compression ratios for the amount of boost pressures seen. As a result of this, I believe that leaner combustion and homoginization go hand in hand. But the results are always open to better interpretation from those with better minds and more experience.

I am now building a small inch V8 (with limited port area and flow) with even higher compression and using this supercharger on E85 to learn more.

I hope this helps,
-DVS

[Devious]

As mentioned in the article above:
CEI ignition - CEI > CEI Lean Burn Ignition
Very interesting.

MAP- in the article you posted, I see the quote "Flow enhanced lean burn, contradicting Exxon's claim". Do you know what the Exxon claim is/was?

[nitro2]

Every engine we have set up in the last 10 years for R&D has had the spark positioned in the area of maximum or near maximum flow. We never shroud the spark from the flow. If the engine needs more ignition energy to light the mixture (lean/rich/alternate fuel) then that's what it gets.

In this scenario a long and powerful spark is important. A short spark ignites what is in the gap then the kernel blows away and hopefully keeps combusting. A long spark (of sufficient power) ignites whatever mixture blows through the plug gap for as long as the spark continues i.e. the kernel becomes a streamer of some length.

Clint Gray
TFX Engine Technology Inc.
Combustion/Intake/Exhaust Pressure Analysis Equipment
www.tfxengine.com

[automotivebreath]

Interestingly enough some natural gas fired turbine engines rely on lean burn to reduce
NOx emissions. Using a precise mixture of gas, H2 and air, the fuel is leaned to a
point just before fire out. At this point combustion temperatures are so low NOx
is virtually non existent.

Spark ignition is not an issue with the constant flame of the turbine engine,
not the case with the SI engine where ignition and mixture homogeneity are major
concerns. If the mixture passing through the plug gap at the point of ignition
does not contain enough fuel to support combustion, the resulting misfire will
result in the drive ability issues (lean surge and tip-in stumble) and high HC emissions.

[automotivebreath]

Some time back I read this paper “Cyclic Variations of Initial Flame Kernel
Growth in Honda VTEC_E Lean-Burn SI Engines”. It describes the same
condition Clint is referring to. Here’s my interpretation of a small portion of
what the paper concludes. All of what is in the paper is focused on the
pent roof design:

Spark duration varied cycle to cycle, long spark duration is beneficial and
preferred over larger spark energy with shorter duration. It is believed the
long duration masks effects of variations in local A/F ratio.

Flame kernel shapes were far from spherical, size and shape differ significantly
from cycle to cycle. The flame develops in the direction of kernel stretch
caused by large scale flow. This is beneficial creating larger flame area as long
as the stretching is not too intense. It becomes a balance, too much flow
causes excessive flame stretch, too little causes poor kernel development;
either can lead to flame quench.

As for plug direction it was found that the preferred orientation is cross flow
meaning if the large scale in chamber flow is east to west the ground
electrode should be north to south.