Mailing List flyrotary@lancaironline.net Message #19791
From: George Lendich <lendich@optusnet.com.au>
Subject: Re: [FlyRotary] Re: A lot to learn
Date: Tue, 5 Apr 2005 15:51:23 +1000
To: Rotary motors in aircraft <flyrotary@lancaironline.net>
Ed,
Thanks for that info.
I understand what your saying about necking down keeping the pressure high - very interesting!!
A very well thought out observation on your part - I do believe your right about flow separation causing eddies which blanket the radiator.
I appreciate your comments.
George (down under)
George, you are undoubtedly correct about necking it down causes a restriction to air flow. Also, there is no doubt in my mind that if you have the space to implement a full length duct of the correct profile you will get  better results.  However, for the Streamline duct that takes about 14-16" of space to run a proper pressure recovery duct, I had less than  6".  So what to do?
 
My  reading  indicated that the thing that causes the most severe disruption to good cooling (assume everything else is adequate) is the separation of airflow from the interior duct walls.  This causes eddies to form which block/hinder airflow through the part of the core they are in front of just like a piece of cardboard would.
 
From what I think I understood of the information on diffusers, it is clear that there is a "dome" of higher pressure air that forms in front of the core due to conversion of the dynamic component of the airstream to some pressure increase by the core and/or diffuser.  It actually forms to some extend whether you have a diffuser or not - due to something I think they refer to as external diffusion.  This is simply the conversion of airstream energy (velocity) into pressure by molecules of air meeting the resistance of the core and other molecules which slow down in front of the core.
 
The Streamline diffuser duct describe in K&W chapter 12 makes use of this higher pressure region to keep the airflow along the duct walls from separating (or at least separating very late just before the core which means if an eddy does form, it is much smaller and has less disruptive effect).   But, that is using the 16" to gradually recovery the dynamic pressure with minimum loss.  Which if done correct will give you around 82% pressure recovery (theoretically) according to K&W -  good if it even approaches that figure in any case.
 
Since I did not have that 16" to play with but only 6" (actually even less), I was faced with "what to do?"   I decided that if I made the duct wall curve more drastically than the Streamline duct wall due to the short distance, I was surely looking at airflow separation. But, I decided that if I made the duct walls "pitch" in even further into the higher pressure region that would help - but then (as you stated, George) you start to affect the airflow which causes the higher pressure dome in front of the core in the first place.    What to do?  I decided (rightly or wrongly) that if I increased the velocity of the air by necking it down, then even if the mass airflow were slightly less than ideal,  that the conversion of the higher velocity air to pressure (the old 1/2pV^2 component) would compensate by keeping the pressure high and this in turn keeping my airflow from separating from my radically curving walls until too late to do much disruption of the airflow through the core.
 
There were those who claimed that there was no way that would work.  Well, I reduced my inlet area from 48 sq inch to 28 and it works just fine thank you.  Tracy Crook can vouch that I have flown with the small openings for well over a year and he has never seen steam or smoke coming from my engine - yet {:>).
 
Seriously, my analysis could be all wet, but so long as the ducts cool, I'll stick with it.
 
My previous ducts were simply openings which captured air.  Photo attached.  This shows the inside of two ducts, the one on the left show the original duct shape which simply has an inlet opening into a more or less box.  The one on the right shows my first attempt to see if my theory would work.  The opening of the duct on the right is 9 square inches and it is placed on my left radiator which is my hot radiator (first one in the series of two).  The opening of the duct on left is original 24 square inches.  After successfully flying with this pair.  I was satisfied enough to build two new ducts.  The one on the left radiator is now 18 sq inches inlet and the one on the right 10 sq inches for a total of 28 sq inches.  So a reduction of inlet area from 48 sq inches to 28 sq inches and it cools just fine as far as I am concerned.  In a hotter climate I would probably enlarge the total area slightly.  But, in any case, it proved (to me at least) that preventing/delaying  airflow  separation was the crucial factor.  I am certain my ducts are not as efficient as a full up Streamline duct, but given the space I had to work with I think my approach has worked well.
 
Hope I answered your question, yes, it appears counterintuitive, but I had a rationale for doing it.  Whether the rational was correct or there is some other factor I am not even aware of - I certainly don't know.  I didn't even sleep at the Holiday inn last night.
 
Best Regards
 
Ed
----- Original Message -----
Sent: Monday, April 04, 2005 6:14 PM
Subject: [FlyRotary] Re: A lot to learn

Ed,
I thought necking-down like you have done would cause more of a restriction and would result in less pressure.
I thought that a more direct line of travel from the opening to the core would give less restriction and more pressure.
 
From what I've been told, this particular design seems to want to follow the trumpet ( bell) shaped opening style, whereby 1.5 times the face of the radiator, is required in the length of the duct. I don't think 6" is enough length.
 
What other styles of duct have you tried which are less successful than this design.
 
I must say it does look good but seems to fly in the face of other previous recommendations.
George ( down under)
 

>>  Homepage:  http://www.flyrotary.com/
>>  Archive:   http://lancaironline.net/lists/flyrotary/List.html


>>  Homepage:  http://www.flyrotary.com/
>>  Archive:   http://lancaironline.net/lists/flyrotary/List.html
Subscribe (FEED) Subscribe (DIGEST) Subscribe (INDEX) Unsubscribe Mail to Listmaster