Air flow measurement - again...

[renko at iastate.edu]

At 10:25 AM 12/19/96 -0500, you wrote:
>  Just had a curious idea.  Im going to spit out every pro/con I can think
>of just to get the idea out of the table for discussion.
>
>  Instead of using a MAF w/ its flow reversion problems, why not use a
>strain gauge?  Put a small object, like a sphere, in the airstream connected
>to an aerodynamic shaft to a strain gauge oriented to measure force in the
>direction of flow.  Should be able to read drag on the object quite
>precisely and quickly.  There could be non-linearities due to the shaft and
>object shape, but a look-up-table/linerization-circuit could correct these.
>After linerization, there should be no problems w/ flow reversion (it would
>measure reverse flow) or non-symetric operation.  It should measure
>something very similar to a MAF (and could probably be made to replace one.)
>It may also have difficulties at low flow rates like the MAF due to very
>small drag force, but...  If the object is something nice like a sphere, one
>could even find the drag vs diameter equation in any good text.  There would


Drag=Coeff_drag*.5*air_density*air_velocity^2*projected_area

For turbulent flow you need to find COeff_drag off a graph

>be a tradeoff of object size vs low speed sensitivity, but I think that it
>would be MUCH less restrictive than a flap type meter.
>
>  Ive had a bit of practical experience w/ strain guages (in a small wind
>tunnel even - to measure lift and drag) and could cook up something to stick
>in a throttle body, but I dont have any way to test it yet.  
>
>  Any thoughts/problems/opinions out here?
>
>SMD
>--
>Stephen Dubovsky
>dubovsky at vt.edu
>
>95 Yamaha FZR600
>83 Porsche 911SC
>84 Jeep Cherokee
>
>


I would suspect that a look up table would be a must.  The drag/velocity
relationship on a smooth sphere is linear so long as the flow around the
sphere is laminar.  This is at very low speeds.  At any appreciable speed
the flow would become turbulent and pressure drag becomes a factor.  This
makes the curve predictable but odd.  See eqn in body of msg.


John Renko
Helser 2631 Elwood
Ames, IA. 50012-0002
(515) 296-5502
renko at iastate.edu