Forwarded: Re: Airflow measurement - again...

[kleenair at ix.netcom.com]

Thanks Andrew and Evert for your inputs.  I think I understand the basic 
prinicpals of LDA now.  It doesn't sound like a very practical idea for 
production engines anytime soon.

Someone else on this list posted a message suggesting use of a strain 
gauge for flow measurement.  I've been thinking about that some more from 
time to time and it seems like a good idea.  His idea was basically based 
on mounting a small object in the air flow path via an arm that has a 
strain gauge.  Increased air flow equates to change in the load cell 
reading.  To cancel out vehicle dynamics an identical setup outside the 
flow path is used and the signals are subtracted.  The resulting reading 
needs to be corrected based on air temp since the drag force is a 
function of air density, viscosity, velocity ^ 2.  Knowing density (by 
temp) and viscosity, we can calculate or look-up velocity, and knowing 
velocity, area, and density, we can determine mass flow rate.

Not too many people commented about his message unfortunately, and I 
can't find the original posting.  What are your thoughts?

Best Regards, Mazda



Evert Rosseel wrote:
> 
> > 1)  How would the system determine the direction of travel of the
> > particles?  More importantly, how does it detect direction changes in
> > dynamic flow, or what kind of errors are caused by turbulance and eddies
> > in the air flow, etc.?
> 
> The classical systems measure the velocity along one axis (1D). For
> 2D or 3D more lasers are needed, and different colors are used to
> measure in the different directions (expensive !).
> 
> LDA systems measure the real particle velocity, which - normally - is
> equal to the flow velocity. This means turbulence is measured if
> present. In fact if turbulence has to be measured in flow in engines,
> very often LDA is used.
> 
> direction changes : see end of message
> 
> > 2)  Do you think this is applicable to inside the manifold, or on the air
> > inlet ducting somewhere.  It seems the laser lenses might be hard to keep
> > clean.
> 
> Yes. It has several disadvantages though :
> - it measures at one location
> - only when one and only one particle passes through the interference
> bands
> 
> This means that you need to do measurements over the whole cross
> section of the manifold, and calculate an average value to measure
> the flow rate.
> If the flow is cyclic (as in engines) you also need to average over
> time (1 cycle).
> 
> For these reasons LDA is only appropriate for research, if one is
> interested in the how the flow pattern changes over time for
> instance. This involves a lot of measurements, in several points in
> the flow. Only after computer processing the flow can be visualised.
> 
> Keeping the lenses clean is not a real problem for research, as it needs to be
> done only every few hours.
> 
> > > LDA requires particles in the air-stream to work.  Fortunately the
> > > dust present is enough.  The system works by intersecting two laser
> > > beams at the point you want to measure the velocity.  At the
> > > intersection, the two beams form interference bands.  As particles
> > > travel through the intersection, they reflect pulses of light at each
> > > constructive interference fringe.  A photo-detector senses this
> > > reflected light.  The frequency of the detected pulses is directly
> > > related to the speed of the particle, which is related to the flow
> > > velocity.  Note though that this is not a MASS flow detector
> 
> I also want to add that in the simple system described above, the
> sign of the velocity is not known. If the flow changes direction you
> need to give the  interference pattern a (sufficient) velocity of its own,
> which is done by shifting the frequency of one of the beams.
> 
> In some engines this flow reversion occurs even in the location where
> the mass flow meter (MFM) is installed. OPEL has a 4 cylinder 2 litre
> engine where this happens in a certain rpm and load range. The
> measurements of the MFM need to be compensated
> for this effect (some air passes the MFM 3 times !).
> 
> Also, at low load SI engines will have flow reversal in the inlet
> ports (valve overlap combined with a low inlet manifold pressure: the
> exhaust pressure is higher !).
> 
> Evert
> ****************************************************************
> *    Dr. ir. Evert Rosseel                                             *
> *    Laboratory for Machines                                   *
> *    Department of Mechanical and Thermal Engineering          *
> *    University Gent                                           *
> *    Sint-Pietersnieuwstraat 41, 9000 Gent, Belgium            *
> *    Tel : ++32 9 264.33.06                                    *
> *    Fax : ++32 9 264.35.86                                    *
> *    Email : Evert.Rosseel at rug.ac.be                           *
> *    WWW : http://allserv.rug.ac.be/~erosseel/motoren.htm      *
> ****************************************************************