speed-density and TPS

[Orin Eman]

> >>- knowing the mass of the air entering the engine is what it's all
> >>about (IMO) -- so you can determine the correct amount of go-juice
> >>to add to the mix.  I'd say, in that regard, that rpm is a totally
> >>useless datapoint (that's not true, I don't think, but it is in the
> >>context of my simplistic statement above).
> >
> >Okay, since my simplistic explanations are leaving a bit to be desired from others, here is
> >a more thorough explanation.  From Heywood's "Internal Combustion Engine Fundamentals":
> >	"For warm-engine operation, the mass of air per cylinder per cycle m[sub]a is give by
> >
> >	m[sub]a = (vol. eff * disp. vol. * int. pres.)/(R * int. temp.)"

> Joe, I'm sure you know more about this than I do, but here's my thoughts.
> That's the technical definition of the mass of air -- and it makes sense
> but we don't want to "define" it, we want to "measure" it and I don't
> see how you can get there from that formula.  As I understand it, VE changes
> dynamically with rpm and statically with engine design and aging.  But that
> puts us back to what I originally said, doesn't it?  We want to know the mass
> of the air entering the engine -- and you can't get it from the above
> equation since you don't know VE.  You *can* get it by

True.  You have to measure it somehow.

>  1 - measuring VE and setting the controls of an alpha-n or SD system for
> it (or tune by the seat-of-the-pants like I do -- actually, a tach, vac gage,
> and stop watch).

You can start with an estimate, get it set using an exhaust gas analyser
at no load such that it is stoich, then using an O2 sensor, let it
learn from there, working backwards from the fuel required for stoich
to VE.  Now you have a baseline table of VE for the engine.  An O2
sensor is more than enough to track long term variations.

>  2 - directly measuring it by MAF -- but most maf gages measure reversion
> pulses the same as incoming air so they have a deficiency with race cams
> I'm told -- and it makes sense to me ??

Another problem with MAF is it is sensitive to vacuum leaks after the sensor...

To use TPS as you describe, you would need pressure before and after the
throttle as well as TPS, unless you assume atmospheric for one side,
which you can't do for a turbo-charged engine :(
You end up with fuel per cyl = f(pressure difference, TPS, RPM).

The VE equation is fuel per cyl = f(MAP, VE).  Where VE is a function
of MAP and RPM.  The engine doesn't much care about how the MAP
happens to be a certain value, a given MAP stuffs the same amount
of air into a cylinder regardless of whether it's due to WOT on a NA
engine or partial throttle on a turbo-charged engine.
So, simplifying, you get fuel = f(MAP, RPM).

(In either case, you should correct for IAT.)

One thing to consider is that you lose accuracy when you subtract two
similar values (say pressure before and after the throttle) and
pressure sensors may not have the required resolution.

Orin.