X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Received: from an-out-0708.google.com ([209.85.132.247] verified) by logan.com (CommuniGate Pro SMTP 5.1.11) with ESMTP id 2248242 for flyrotary@lancaironline.net; Tue, 07 Aug 2007 13:12:54 -0400 Received-SPF: pass receiver=logan.com; client-ip=209.85.132.247; envelope-from=rotary.thjakits@gmail.com Received: by an-out-0708.google.com with SMTP id b2so293091ana for ; Tue, 07 Aug 2007 10:12:16 -0700 (PDT) DKIM-Signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=A7GPY8aSWo2jQjdtSE1EJQN5mrAOwy1kTrALPhvuKfz9pJbMYGhZ17SYl+pv479qhOMynxK9NTIXY/EZMsFkWnfjf0eT0JdaU1WsvAAsJH9txs6lYHTsBlxhxmkAK3v7iFU5qIo5MfDhgqppiw6RUW5Sdy/e/ofBagIVvrKyfBw= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=k8IAtR+9yw+Cf61M4By6XINRqGr+oMUBgwyxkFeGJo/UPXB9FupRvict+rvAgvT7EnzMb/CHiPhuC0tpKLc7/6RrPJZ7d0E2lOPKzazpzj/yqg1Yijj9hvzA0zbnREhNqrM4GRx+aTLxirE52ec6cGezSwR6Og9GXVjbuvvIg1E= Received: by 10.100.153.17 with SMTP id a17mr4017967ane.1186506736607; Tue, 07 Aug 2007 10:12:16 -0700 (PDT) Received: by 10.100.96.15 with HTTP; Tue, 7 Aug 2007 10:12:16 -0700 (PDT) Message-ID: <63163d560708071012l3c558385ka7859f835e0ab1fb@mail.gmail.com> Date: Tue, 7 Aug 2007 12:12:16 -0500 From: "Thomas Jakits" To: "Rotary motors in aircraft" Subject: Re: [FlyRotary] Re: RV -7A Cooling Update 8/6/07 In-Reply-To: MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_138336_8366395.1186506736529" References: ------=_Part_138336_8366395.1186506736529 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Going with gut-feelings: Dennis, you mentioned your inlet duct represents a wedge. I believe (other expression for gut-feeling :)), a streamline duct redirect= s the airflow to about 90=BA towards the radiator surface (never mind the fin/cooling tube orientation). So you might want to contact that PL guy on that other list and have him make you a custom 3D drawing for your situation. It should not be too much work to make a duct according to streamline philosophy - cheaper than than a new custom radiator anytime! (Keep your intake opening and distance and have Mr. 3D-Rhino make a coordinate list from which you can cut foam for a duct .....) I did not study deep enough into the cooling theory, but I understand that the incoming air has to be slowed substantially to make the cooling system efficient (less drag). At some point it will break flow and become turbulent/stagnant - the trick seems to be (or so my gut tells me) to get this point as close or right at the face of the radiator - here it can do its magic (take all the heat out of the radiator) and return to orderly flo= w right after the cooler. Hoe does your exit (duct) look, maybe we can squeeze some more delta-P from there..... IF you go the PL route you might as well ask for the "perfect exit" solution, too. Did you post pics in the past? TJ On 8/7/07, Ed Anderson wrote: > > I agree, Tracy. > > Our flow is without doubt never laminar, a boundary layer of laminar flow > has only(mostly?) the molecules next to the metal absorbing significant > heat. Those molecules in the middle of the stream have no(little) > opportunity to pick up heat. A boundary layer with turbulence on the oth= er > hand has molecules shuffling all over the place and every one (most?) get= an > opportunity to contact the hot metal and take away some of the heat. So > turbulent flow is better for conducting away heat. > > Chaotic macro flow (boundary layer folding over itself, eddies, etc) on > the other hand impedes pressure recovery, increases drag and overall > adversely effects cooling. > > My research and experiments leads me to believe that many factors are > relative minor compared to the large scale adverse effect of poor duct > design which leads to early boundary layer separation. However, > significant macro turbulence at the entrance to the core channels might h= ave > a large effect - just speculation on my part. > > Ed > > > > ----- Original Message ----- > *From:* Tracy Crook > *To:* Rotary motors in aircraft > *Sent:* Tuesday, August 07, 2007 11:03 AM > *Subject:* [FlyRotary] Re: RV -7A Cooling Update 8/6/07 > > > No scientific analysis here, just my sum total of gut feel after reading = & > experimenting. > > In my understanding, the airflow through the common rads we use is fully > turbulent. the little louvers in the fins are there to guarantee this. = So > what difference does it make whether the air goes turbulent at the leadin= g > edge of the fins of a dozen or so thousandts later. Again according to m= y > very fallible gut feel, the whole story is whether or not you converted t= he > air velocity to air pressure. Either you did or you didn't. I can't > remember if you have measured pressure at the face of the rad or not but > that will tell the whole story. Angle of tubes, fins, face of rad, etc = is > all relatively insignificant. > > The motorcycle rad stuff someone mentioned is not a good indicator. They > do not depend on high pressure recovery the way we do so the design and > operation of their rads is not necessarily applicapable. > > As always, YMMV and I will gladly amend my gut feel to match reality if > you find it is wrong. > > Tracy > > > On 8/6/07, Dennis Haverlah wrote: > > > > > > > > I've been busy with Family vacation, dealing with the exceptional wet w= eather in > > central Texas and my tennis playing but finally I have some more > > thoughts on radiators and cooling. *My cooling is marginal for Texas in > > the summer.* I want to climb at 120 kts and 26 + inches MP on a 100 de= g > > F day without exceeding 215 on water and oil. > > > > I have the Griffin radiator (core 19 X 13 X 2.5 inches) and stock RX-7 > > '89 oil cooler as shown on pictures I have previously posted. The > > radiators are mounted under the engine at about a 30 deg. angle. My > > latest test flight with OAT of 92 deg F on the ground was encouraging. > > I had temp. probes on the outlet side of the oil and water radiators > > to measure the temp. of the heated air. The temp. probes had an upper > > limit of 160 deg. F. The air exiting the water radiator exceeded the > > 160 Deg. limit soon after take-off. I estimate the air temperature > > rise through the water radiator was at least 80-90 deg. Cooling water > > temp. never exceeded 210 deg. F. > > > > The air exiting the oil radiator was at 135 - 140 deg. F. (A delta T of > > about 40 - 45 deg F.) Oil temperature rose to 213 deg. F. max and at > > cruse 24 in. MP, 160 mph at 5500 feet the oil temp. decreased to 210 de= g. F. > > > > > > I'm close to ideal cooling but I've been surprised how little effect my= air > > flow modifications have have had on overall oil and water cooling. Aft= er > > studying K&W Chapter 12 some more I've decided I mounted my cooling rad= iators > > incorrectly!! As mentioned above, the radiators are below the engine a= t about > > a 30 Deg angle (alpha =3D 60 deg.) to the incoming air stream. The tan= ks are > > orientated fore and aft. This positions the fins across the air stream. > > > > Ch. 12.2 of K & W Fig. 12.6 shows a radiator block at an oblique angle = (*alpha*) > > to the incoming air. The *tubes* are at the angle* alpha* to the flow.= In the > > K & W analysis the tubes are slightly aerodynamic in shape they turn th= e flow > > as it enters the radiator fins. In the radiators I am using the tubes = are > > separated about 1/2 inch. My fins are separated by about 0.080 inch. B= ecause > > I mounted my radiator with the tanks fore and aft, the fins are at the = angle*alpha* to the flow and the fins turn the air. The fins are very sharp= thin metal > > and I believe air flow separation and turbulence is occurring at the le= ading > > edge of each fin. Because the fins are very close together the flow is = restricted > > through the entire radiator surface. *I believe the separated, turbule= nt flow at > > the leading edge of the fins limits the amount of air flowing through t= he > > radiator regardless of how "good" the diffusers are ahead of the radiat= ors.* > > > > If I have to do it over, I will defiantly mount my radiators with the t= anks on the left > > and right side of the incoming air so that the tubes turn the air throu= gh *alpha* - not > > the fins!! > > > > Any comments - Am I out to lunch on this one? > > > > PS. The end of the first paragraph in Ch. 12.2. states "We shall consi= der first the > > simple case of parallel inflow at an angle *alpha* to the *tubes*, as s= hown in Fig. 12.6" > > I have not found a consideration in Chapter 12 of the case of the *fins= * being at > > an angle *alpha*. > > > > Dennis Haverlah > > > > > ------=_Part_138336_8366395.1186506736529 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline
Going with gut-feelings:
 
Dennis, you mentioned your inlet duct represents a wedge.
I believe (other expression for gut-feeling :)), a streamline duct red= irects the airflow to about 90=BA towards the radiator surface (never mind = the fin/cooling tube orientation).
So you might want to contact that PL guy on that other list and have h= im make you a custom 3D drawing for your situation.
It should not be too much work to make a duct according to streamline = philosophy - cheaper than than a new custom radiator anytime!
(Keep your intake opening and distance and have Mr. 3D-Rhino make a co= ordinate list from which you can cut foam for a duct .....)
 
I did not study deep enough into the cooling theory, but I understand = that the incoming air has to be slowed substantially to make the cooling sy= stem efficient (less drag). At some point it will break flow and become tur= bulent/stagnant - the trick seems to be (or so my gut tells me) to get this= point as close or right at the face of the radiator - here it can do its m= agic (take all the heat out of the radiator) and return to orderly flow rig= ht after the cooler.
 
Hoe does your exit (duct) look, maybe we can squeeze some more delta-P= from there.....
IF you go the PL route you might as well ask for the "perfect exi= t" solution, too.
 
Did you post pics in the past?
 
TJ

 
On 8/7/07, E= d Anderson <eanderson@c= arolina.rr.com> wrote:
I agree, Tracy. 
 
Our flow is without doubt never laminar, a bounda= ry layer of laminar flow has only(mostly?) the molecules next to the metal = absorbing significant heat. Those molecules in the middle of the strea= m have no(little) opportunity to pick up heat.  A boundary layer with = turbulence on the other hand has molecules shuffling all over the place and= every one (most?) get an opportunity to contact the hot metal and take awa= y some of the heat. So turbulent flow is better for conducting away heat.
 
Chaotic macro flow (boundary layer folding over i= tself, eddies, etc) on the other hand impedes pressure recovery, increases = drag and overall adversely effects cooling.
 
My research and experiments leads me to believe t= hat many factors are relative minor compared to the large scale adverse eff= ect of poor duct design which leads to early boundary layer separation.&nbs= p; However, significant macro turbulence at the entrance to the core c= hannels might have a large effect - just speculation on my part.
 
Ed
 
 
----- Original Message -----
Sent: Tuesday, August 07, 2007 11:03= AM
Subject: [FlyRotary] Re: RV -7A Cool= ing Update 8/6/07

 
No scientific analysis here, just my sum total of gut feel after readi= ng & experimenting.
 
In my understanding, the airflow through the common  rads we use = is fully turbulent.  the little louvers in the fins are there to guara= ntee this.  So what difference does it make whether the air goes turbu= lent at the leading edge of the fins of a dozen or so thousandts later.&nbs= p; Again according to my very fallible gut feel, the whole story is whether= or not you converted the air velocity to air pressure.  Either you di= d or you didn't. I can't remember if you have measured pressure at = the face of the rad or not but that will tell the whole story.   = Angle of tubes, fins, face of rad, etc is all relatively insignificant.=20
 
The motorcycle rad stuff someone mentioned is not a good indicator.&nb= sp; They do not depend on high pressure recovery the way we do so the desig= n and operation of their rads is not necessarily applicapable.
 
As always, YMMV and I will gladly amend my gut feel to match reality i= f you find it is wrong.
 
 Tracy

 
On 8/6/07, D= ennis Haverlah <cloudust= er@austin.rr.com > wrote:=20 


I've been busy w=
ith Family vacation, dealing with the exceptional wet weather in=20
central Texas and my tennis playing but finally I have some more=20
thoughts on radiators and cooling. My cooling is marginal for Texas in=
=20
the summer.  I want to climb at 120 kts and 26 + inches MP on a 100 deg=
=20
F day without exceeding 215 on water and oil.=20

I have the Griffin radiator (core 19 X 13 X 2.5 inches) and stock RX-7=20
'89 oil cooler as shown on pictures I have previously posted.  The=20
radiators are mounted under the engine at about a 30 deg. angle.  My=20
latest test flight with OAT of 92 deg F on the ground was encouraging. =20
I had temp. probes on the outlet side of the oil and water radiators=20
to measure the temp. of the heated air.  The temp. probes had an upper=20
limit of 160 deg. F.  The air exiting the water radiator exceeded the=20
160  Deg. limit soon after take-off.  I estimate the air temperature=20
rise through the water radiator was at least 80-90 deg. Cooling water=20
temp. never exceeded 210 deg. F.=20

The air exiting the oil radiator was at 135 - 140 deg. F. (A delta T of
about 40 - 45 deg F.)  Oil temperature rose to 213 deg. F. max and at=20
cruse 24 in. MP, 160 mph at 5500 feet the oil temp. decreased to 210 deg. F=
.


I'm close to ideal cooling but I've been surprised how little effec=
t my air=20
flow modifications have have had on overall oil and water cooling.  After=
=20
studying K&W Chapter 12 some more I've decided I mounted my cooling=
 radiators=20
incorrectly!!  As mentioned above, the radiators are below the engine at ab=
out=20
a 30 Deg angle (alpha =3D 60 deg.) to the incoming air stream.  The tanks a=
re=20
orientated fore and aft. This positions the fins across the air stream. =20

Ch. 12.2 of K & W Fig. 12.6 shows a radiator block at an oblique angle =
(alpha)=20
to the incoming air.  The tubes are at the angle alpha to the=
 flow.  In the=20
K & W analysis the tubes are slightly aerodynamic in shape they turn th=
e flow
as it enters the radiator fins.  In the radiators I am using the tubes are=
=20
separated about 1/2 inch.  My fins are separated by about 0.080 inch. Becau=
se
I mounted my radiator with the tanks fore and aft, the fins are at the angl=
e
alpha to the flow and the fins turn the air. The fins are very sharp=
 thin metal
and I believe air flow separation and turbulence is occurring at the leadin=
g=20
edge of each fin. Because the fins are very close together the flow is rest=
ricted
through the entire radiator surface.  I believe the separated, turbulent=
 flow at
the leading edge of the fins limits the amount of air flowing through the=
=20
radiator regardless of how "good" the diffusers are ahead of the =
radiators.
 
If I have to do it over, I will defiantly mount my radiators with the tanks=
 on the left
and right side of the incoming air so that the tubes turn the air through <=
u>alpha - not
the fins!!

Any comments - Am I out to lunch on this one?

PS. The end of the first paragraph in Ch. 12.2. states  "We shall cons=
ider first the
simple case of parallel inflow at an angle alpha to the tubes=
, as shown in Fig. 12.6"
I have not found a consideration in Chapter 12 of the case of the fins being at=20
an angle alpha.=20

Dennis Haverlah


------=_Part_138336_8366395.1186506736529--