Air Flow Measurement

[tom cloud]

>c'mon, you can't read that?, thats the first message i CAN read and 
>understand!  after having 68hc11 code burned into the back of my brain, 
>this is REAL easy on the eye...
>
>Our experience in the HVAC CONTROLS industry is that a cost effective 
>air flow measuring station (one that low-bidder can sell), using 
>hot-wire, or static and total pressure, becomes REAL inaccurate at the 
>low end, as was mentioned in a previous responce.  mass of air must be 
>assumed, and the duct runs must be of a certain minimum STRAIGHT 
>distance before and after the flow station (no turbulance?). as the 
>price of these stations goes up, honey-comb inserts can be included to 
>mix and straighten the air out as it comes around a bend in the duct. 
>(try that in a 4"dia twisting steel tube). with more cash, you can 
>matrix anamometers or combo pressure pickups in a duct in a grid that 
>samples the pressures at 9,12,16... points. granted, HVAC doesnt have to 
>be as exacting, but a blade-position sensor produces such a non-linear 
>percent open-to-flow relationship, that to use it, you have to FLOW the 
>setup to get a curve to track. (ie 10% open=35%flow). as you add "fixes" 
>to a simple manometer setup to correct for ??? the price goes up and up 
>and you've lost the bid and your broke cause you spent all your wad on 
>the blueprints and lunch for the engineer and your jobs on the 
>line...... 
>sorry, i drifted out of EFI and into rat race there. just another 
>airflow view from another airflow angle.
>
>lucky ECM designers get decent budgets for the controllers or we'd still 
>be cranking that spark-retard knob.  
>

I wrote that seemed t'me one could deduce air flow (and hence mass)
from TPS, BAP and MAP.  Someone said I was reel smart (tanx).  Then
someone else said had to have air temp (oops).

Well, what I didn't say .... just cause it "seems t'me" doesn't mean
it'll fly.  I mean, if it was such a wonderful idea (mine, above, that
is), why doesn't Dee-troit use it?  Is it just not good enough to allow
EPA certification??

I didn't say it ... didn't know .... wanted to hear what others thought,
but, assuming you account for the (I assume sine/cosine) relationship
of the throttle non-linear opening characteristic, you'd still have
the turbulence at different pressures and other probs I can't even
imagine .... would this really work?

Still seems to me (I'm still fishing for a definitive answer) that
TPS, BAP, MAP and TEMP (IAT) would give a very reasonable approximation
(after correction for non-linear TPS vs. opening area) of MAF.  No ???


Tom Cloud <cloud at peaches.ph.utexas.edu>