[Diy_efi] Re: DIY-WB and methanol

[Garfield Willis]

On Fri, 11 Oct 2002 15:13:45 -0700, "Derek" <derek_obanion at yahoo.com>
wrote:

>Yes I do. I added enough 50/50 mix of methanol/water to allow a =
reduction of 20%
>fuel from the injectors. That works out in terms of volume flow to be =
nearly the
>amount of gasoline.
>
>(6.45/14.7)*.5=3D21.9% (Meaning a 50/50 mix of meth/water has 21.9% of =
the fueling
>capability as pure gasoline).

I see, OK that makes more sense, see below.

>Less inaccuracy because as I understand it, it's the inert components =
that are
>the real problem (IE water). All that water which was in the 50/50 mix =
is now
>gone, replaced by a fuel.

However, your alcohol also has an oxygenate in it (it's a fuel that
carries it's own oxygen), so you will STILL see exhaust component
dilution (over and above the %diluents from petrol) because burning the
meth will produce more exhaust tract water vapor than petrol does. And
the cal curves are based on that petrol diluent percentage. If it was a
simple case of another hydrocarbon fuel that just had a different C:H
ratio, then all you'd have to do is SCALE the AFR readings according to
the ratio of fuel stoich values. But it's NOT that simple when your fuel
also contains a non-zero value of O:C. Not to mention a non-zero N:C
component, as in NOS or nitro.

>I NEVER said you could use 50/50 meth/water as a primary fuel? Did you =
get that
>impression?

Yep, that's what I thot you were saying.

>On a pure methanol motor, widebands work fine, you just shoot for a =
.8-.85
>lambda reading. On a pure gasoline motor, widebands work fine, you shoot=
 for a
>.8-.85 lambda reading. How should mixing the two change anything?

It won't (actual chemical lambda), from a purely chemistry standpoint,
as long as you make it clear what units you're talking in. The problem
in the past has been people talking in petrol AFRs when using alternate
fueling. This just makes an ungodly mess, as you point out below. BUT
(and here's why just a purely chemistry perspective can fool you) don't
forget that the sensor physics enters in here too, along with the basic
chemistry. Any time you have a fuel w/oxygenates or a pure diluent
combined, the more excess water or inert gaseous byproducts (above &
beyond the %diluent coming from petrol) the more your base sensor
calibration curve will be skewed by the diluent effect. Merely scaling
stoich AFRs does NOT compensate for this effect. That's why I took pains
in my last post to demonstrate this by showing that all the pro-level
AFR sensing guys have to take into account the additional O:C and N:C
factors in the alternate fueling cal. This is done to ALLOW for
compensation from the indirect diluents (more H2O vapor and/OR gaseous
N2 from combustion), or direct diluents (non-combustion diluents like
water vapor from atomized water injection).

>I hate the AFR numbers BTW, gets so confusing. Once you start thinking =
in lambda
>you never want to go back (especially if you mix fuels like me). I'm =
getting
>tired of explaning to people why you have to still shoot for about 12:1 =
reading
>on their widebands with methanol, even though it has a stoich of 6.45:1.=
 Because
>that reading is just a translated lambda value. It been corrected for =
gasoline,
>but doesn't actually tell you the ratios anymore once you mix fuels. A =
6.45:1
>mix on a methanol motor will display 14.7:1 on a wideband display set =
for
>gasoline.

That's fine, and that's exactly what you should be doing with alternate
fuels (ESPECIALLY if you have a MIX). Namely, NOT speaking in terms of
petrol AFRs, because there are STILL lots of people who also talk in
terms of true fuel AFRs, so when THEY say they're running straight
alcohol @ 5AFR, those are real *alcohol/air* AFRs, not mangled
'translated/adjusted' petrol AFRs, obviously. The whole
advantage/disadvantage of AFRs is they ARE fuel specific. The Euros
dumped the AFR system long ago, and generally ONLY talk in lambda's, but
you STILL have to remember that all talking in lambda's gives you is
normalization between fuels via their relative stoich values (or
'fueling capability' as you put it earlier), which as a measure of
adequate enrichment, DOES work quite well over a large class of fuels,
so what's sauce for the goose (~0.8-0.85 lambda for petrol, forced to
na) is usually also sauce for the gander (some go even lower Lams for
top fuel). BUT AGAIN, that's the combustion chemistry; you also have to
take into account what assumptions were made when you sensor's cal curve
was given out by NTK, about exhaust gas diluent percentages. The
alcohols are prodigious water vapor producers and exhaust tract
back-pressure produces as a result, and so you have to pay attention to
these effects ON THE SENSOR.

Remember, it's not just combustion chemistry involved, it's the physics
of the sensor that's also involved. In short, I'm suggesting you
re-examine your presumption that these sensors we're using measure pure
Lambda, regardless of fuel chemistry. They don't. Whether exhaust gas
component dilution is direct or indirect (from the 'fuel' being a
fuel+oxygenate), it's still a factor that must be compensated for, in
order to get EITHER a correct AFR *or* Lambda.

HTH,
Gar


_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi