Subject: Oil cooler air flow
From: Rotary Engine
Date: 7/17/2007, 4:28 PM
To: AARotary Engine


> > Hey Al,
> >
> > I think there's likely separation as illustrated in the first picture.
> >
> > I think Paul is right about the wedge diffuser, but I don't think you
> > have
> > to move the cooler to the cowl.  Maybe have a look at the 2nd pic...
> >
> > What do you think?
> >
> >
> > Matt-
>
> Al,
> I vote for any form of scoop (ala P-51) that is moved far enough off the
> surface to get out of the boundary layer. Ingesting boundary layer air
is
> never efficient regardless of how cosmetic it is. As an alternative, you
> oversize the heck out of the cooler and get away with passing less air
> over
> it.
> Larry
>
> No doubt that you are right about that.  this configuration began the
way
> it
> did because another Velocity builder had put his standard aircraft oil
> cooler (for a Lyc) in the same location, and said it worked great
without
> a
> scoop ? just the differential pressure above and below the wing was
> enough.
> Go figure.
> Actually in the location that it is, even an inch or less away from the
> surface would make a big difference.  Keep in mind that the scoop is
about
> 23? wide with about 1 ¼? opening.  How do you get BL diversion with that
> configuration?
> ------------
> Matt wrote:
> How can we apply instrumentation in the duct so that we can tell where
the
> flow is becoming turbulent?  Or we can just continue to guess...  Since
> this
> duct is not exactly like any in the books, we can't directly copy any
> solution that's out there..  I believe Al's data is valid (it is what it
> is), but a few more key data points could help make the picture clearer.
> Hopefully we can talk Al into continue playing along.
> -------------
> I'll look at getting a bit more info.
> Meanwhile; I'm looking toward the idea of a vane as what might be a
> relatively easy fix.  I think that my scoop opening is large enough, and
> the
> BL thin enough, that if I can get effective diffusion in the duct it
> should
> work just fine.
>
> Note the attached diagram.
>
> Al
>
> A turning vane is worth a try. They usually don't work all that great as
> they have form and skin friction drag even if they are not made out of
> sheet
> metal
> and are perfectly streamlined.
>
> BTW that is an old  diagram WW I era with an obsolete diffuser shape
> guaranteed
> to separate along the duct walls. Folks don't use that shape.
>
> This plot from Hoerner sums up the situation fairly well.
> Note again "w" is going to be a function of the actual airspeed flowing
> through the
> core. This can be measured with a pito tube pointing at the aft face of
> the
> core.
>
>
----------------------------------------------------------------------------
> This last plot deserves a bit of explanation since there is no factor
> for boundary layer effects. Here they are.
>
> What Hoerner is calling overflowing is now known as external diffusion.
> It happens when the rad is not porous enough and can trigger the main
flow
> to flow around the opening and not in it.
>
> Paul Lamar ...No rotor no motor.
>
>
>   From the pieces you've attached; I'm not getting the big picture, or
the
> message.  Given that I'm working with a case which is in the BL; what I
> glean from this is that a larger radius on the lip of the scoop could
> improve the flow into the scoop. ??
>
> Al
>
> I don't think so. IMHO What happens is the pressure situation inside the
> scoop is
> also affecting the pressure at the entrance of the scoop. This combined
> with
> the boundary layer thickness is enough to divert the high velocity air
> stream
> below the scoop opening.
>
> These CFD's illustrate the problem.
>
> You can use a blower to suck the boundary layer out from in front
> of the scoop. If successful that will allow the high velocity air
> stream to enter the scoop.  Perhaps I'll have time to do another 3D
which
> will illustrate the trick.
>
> You can also just shot gun it and put a two or three times larger
opening
> hopefully scooping everything up including the boundary layer.
>
>
> Paul Lamar ...No rotor no motor.
>
>
>
> Al,
> Food for thought.... don't extend the scoop down. You could actually
make
> it
> shorter in height. Build an extension duct and move the entrance forward
> on
> the wing. The closer you get to the leading edge, the thinner the
boundary
> layer, as a boundary layer gets thicker the farther aft you go on the
> surface. In your case, you would have to have a part of the duct mounted
> on
> the gear door and would loose it with the gear down. The pressure
> differential created on the cooler this way would make the exit area non
> critical. You could louver the exit for cosmetics and the cooler would
> still
> function fine. This duct could start taller and narrow in the front and
> gradually fan out to the width of the cooler while getting shorter,
> creating
> a divergent duct in the process. A wide duct would require some internal
> vertical supports to keep the bottom from ballooning, as it is a lot of
> surface area even at low pressures.
> Larry
>
> Good idea Larry. I'll do a Rhino 3D. Something like a WW II Mosquito
> bomber.
>
> Here they are.
>
> Paul Lamar ...No rotor no motor.

I think there's a wing spar and fuel tank in the way of what you're
talking
about..  Not insurmountable challenges, but the payoff to cost ratio seems
unfavorable..


Matt-

The duct is under the spar and fuel tank. That of course requires
thickening up the strake. I did not show them in an attempt at clarity.
--
Paul Lamar ...No rotor no motor.

Okay..  That makes more sense.  There's obviously a trade off in
efficiency between getting the inlet to a more favorable position (in terms
of local pressure and boundary layer thickness), and having more external
and internal wetted area plus length of intersection drag.  A duct that
extends well forward gets into better pressure and flow, but has more
parasitic drag..

If I remember correctly, Al's data suggested that the boundary layer wasn't
all that thick..  If much hacking and cutting is going to be done, and an
exposed duct is going to be installed anyway, I think the wedge diffuser may
be the most foolproof, least complicated, least picky, and hence the
quickest configuration to implement..  I get the impression from Al that he
just wants something that works well right now.

If it were mine, I'd probably take one or two shots at getting a flow
correcting vane in the inlet, and if that didn't work, I'd go to a wedge
diffuser.


Regards,

Matt-

I think the easiest and quickest thing to try is a few  boundary
layer turbulators in front of the intake.

Next easiest after the wedge diffuser is this boundary layer bypass slot.

IMHO Al needs a lot more pressure taps in the intake to see what the pressure
distribution is across the face of the core.

My offer still stands Al.

"If you send me your shipping address by private mail and the height of the exit
slot I'll send you a 8 tube water manometer, a velocity rake, porous foam  and a bunch
of drip irrigation tubing. All you need is a hot glue gun. In return I expect some
good readings."

To summarize dynamic pressure at 150 MPH it is about 11 inches.
A fair scoop and diffuser can recover 50% of dynamic pressure or
about 6.5". Al is getting .25 inches at 160 MPH.


Paul Lamar ...No rotor no motor.

The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.