07/31/2003

AFR and Power

When you consider that fuel in excess of just slightly 5-10% maximum rich is
soley there for the cooling mass.  5-10% rich is where peak temperature is
from dissociation effects - tending to remain closer to 5% depending on the
fuel.  

The reason for the use of excess fuel for cooling is that many engines run out
of air before desired power.

Leaning the mixture's below stoic adds large amounts of "cooling" air exactly
the same effect as adding cooling fuel.  Mass is mass.  Naca found gasoline
and water alcohol about as effective as each other for internal cooling up to
about 40%.  Air, being mass, should be just as effective.

Another thing to consider is that with around 10% excess air, the flame and
post flame combustion is the fastest.  It takes more induction time, but less
flame travel time.

Picture the mixture at stoic where the right amount of oxygen is blended with
the right amount of fuel.  There will be a time for oxygen to meet fuel and
have sex.  Sprinkle extra oxygen around, it will take less time for oxygen to
meet fuel and fool around - but the dilution of nitrogen is increasing.  About
10% lean or around 16 to 1.

On an N/A engine, this can be "harmful" to power only at the upper reaches of
power, and on a compressed engine - particularly turbo'd, where excess air is
always available, it is as effective as excess fuel as an internal coolant.

------------------

Dang you Tom.  This is the first recent lean burn data I have seen.

First - On A/C and other engines, there is a low EGT point about 10% excess
air, from maximum power extraction, and then EGT begins to rise.

The post flame combustion improvement runs into a slowing flame speed so that
when the exhaust opening is faster than end of combustion, the temperatures
begin to go up.

You mentioned this as a lean burn engine with appropriate mods.  Does the data
track the decrease to minima and then increase in temperature again at a much
leaner mixture?

The flame speed improvements from a lean burn chamber over a hemi-like chamber
of AC engines should allow a nice decrease in a lean burn time.

Normally, the lower limit of explosiveness is a clue about the maximum amount
of excess air the homogeneous charge will tolerate and still burn.  What
measures other than swirl did they use to extend the lean limit below static
explosiveness?

Was the fuel modified - in particular was it an E-15 gasohol?  Normal gasoline
has as I said about 25% excess, but methanal or ethanol can tolerate up to
100% percent excess air.  Even 10 to 15 percent ethanol could nicely extend
the excess air limit.

Were there any nitrates added to the fuel?  Example, I am fooling around with
cetane improver on my Aerostar and with a half bottle (16 ounces), I seem to
have picked up about 4-5 mpg and not hearing knock on 87 octane with the air
conditioner on all the time.

Recently, there was a thread about R-fluid and how 5% r-fluid allowed the
german engines on gasoline to run with 100% excess air.

And, was there any indication, that for "older" more conventional engines, a
minor chemical change could be used to bring on most of the benefits?

Recently, there was a thread about mental reversals that a middle boost engine
incurs.  Put on a hypothetical hat for a moment.  Excess fuel has been used
historically because air was at a premium.  With late model centrifugal
superchargers and turbochargers, its possible to have excess air even at full
throttle.

Knocking is closely related to charge temperature and as you pointed out, the
charge temperature falls off very rapidly away from near stoic peak EGT.  A
review of Glassman on combustion reveals the purpose of excess fuel is to
increase the mass of combustion and lower the peak temperatures.  Air works as
well as fuel - IF YOU HAVE EXCESS AIR.

Rich of stoic, about 100 hp for each 130 cfm.  Lean of stoic, ignore air and
use .5 lbs of fuel per hp hours unless you have something closer to your
engine.  

If you could adjust for minima of EGT as in AC engines manual tune, could you
not add ethanol or acetone and gain cooling of charge from latent heat of
vaporization and then combined lean to minima?  Both of these not only are
excellent internal coolants but greatly extend the lower limit of combustion,
allowing even leaner mixtures.

Any research of the use of flame accelerant related to the fuel for a lean
burn engine - in particular nitrates (nitromethane is the simplest nitrate - I
think? .

Dang you - you got me thinking again.


On Thu, 31 Jul 2003 10:37:27 -0700 (PDT), you wrote:

> >Tom L wrote:
>> >> I checked some old data, and the change in spark timing isn't huge
>> >> at 16:1 A/F.  When I think "lean burn" my mind jumps to the lean
>> >> stability limit around 19 or 20:1, where the changes in spark timing
>> >> get really large.
> >
> >Greg Parmer replied:
> >-> Installing the EGT gauge just moved a few notches up the priority
> >-> list. 20:1 would be significant. While these heads aren't super
> >-> hi-tech, I've gotta believe the corkscrew design and ceramics will
> >-> be quite a bit better than the old school SBC head.
> >
> >By EGT do you mean exhaust gas temperature?  It peaks at stoich so
> >EGT should not be a concern for running lean.  The typical hot
> >rodder's fear of lean operation should really be fear of "not rich
> >enough" at wide-open throttle.  Running at 16:1 gives about the same
> >exhaust temperature as 13:1.
> >
> >Here's the exhaust temperature from the same data set:
> >
> >2500 RPM wide-open throttle
> >A/F     MBT spark     Exh T (degC)
> >11      22            626
> >12      21            654
> >13      21            683
> >14      21            712
> >15      22            711
> >16      25            680
> >17      27            651
> >18      32            621
> >19      34            600
> >20      38            580
> >
> >Most engines designed before (roughly) the late 1980's won't
> >tolerate lean burn nearly as well as the more recent designs.
> >Certainly an intake port that generates swirl will help.  If you
> >have functional catalytic converters, be careful about running too
> >lean!  If you approach the stability limit and start getting some
> >misfires, the unburned fuel/air mixture will burn in the catalyst
> >and possibly melt it.
> >
> >-> Thanks again for the datapoints.
> >
> >You're quite welcome!
> >
> >

------------------------

Ok - MBT by pony brake - at the mixture specified.  So at 16 to 1, spark was
modified for maximum torque for example.  Thus each egt is for the MBT.  OK

Data is from Heywood and backwards probably on fixed timing sweeps. 

Normally on the lean side you find minima lean, increase air until it comes up
50 - 75 f and run here for best bfsc. AC based I think.

Now the tuning I have seen for lean mixtures has been to find the minima for
EGT, freeze the mixture, then adjust timing for minima.  A couple of
iterations would leave you at combined minimum BSFC.

On your data, was there a timing sweep for each air/fuel?  Something along the
line that peak MBT matched minimum EGT at each specified a/f or very close.

Trying to find out the EGT behavior for the possibility of tuning the engine
without a dyno.  Or other measurable data with either a peak or minima.

The extra oxygen lean causes the fuel to burn more rapidly than at stoic, but
the excess air increases the lag time, requiring, as you point out, a moderate
spark increase.  If the coolant load is appropriately constant, then the
released energy will be split between residual EGT and flywheel.  Normally the
maximum power is when the EGT is the lowest in a sweep with either A/F or
spark fixed.

Turbulence is a good thing.  Lower limits of explosiveness are usually in a
quiescent bombe - so the enhanced swirl tumble etc turbulence would act to
lower it further.  I think that makes sense.

On exhaust overlap, is it increased - with the increase by an advanced intake
overlap?  You well described the effect of the deliberate auto-EGR on
vaporizing the fuel charge by a back blast of EGR.  With leaner mixtures, it
would seem this vaporization would be needed to improve the homogeneity of the
charge.  Also, this extra heat to the charge would not seem harmful as the
excess air greatly cools the charge.  Heywood reports 5% or less residual gas
is meaningless on power thru most of the curves and is correlated with data I
have seen from Rugaloski and Obert among others.

And, if you desire, could you republish the self egr information on fuel
vaporization and the timing needed to get it?  Most new members are not
familiar with it.  Thanks.

On the data for max egt with fixed spark - 5% is about 14.0 mixture and
according to glassman - based on the time preference of hydrogen combustion
over carbon combustion and the generation of highest temperature by
dissociation of co2 to co and free hydrogen.  Rich.  

Since peak power rich NA is at 12.5 ( sortof maybe around there ) and the
power is increased by a max of 5 percent, the entire effectiveness of
enrichment is suspect for full power.  The close vicinity rich for peak EGT
versus the older 5% or so, is this a result of the combustion rate changes in
the fast burn cylinder?

Any recent texts that cover this shift of focus and the various effects?  This
is very nice good stuff - thanks.