X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Received: from mu-out-0910.google.com ([209.85.134.186] verified) by logan.com (CommuniGate Pro SMTP 5.2c2) with ESMTP id 2471131 for flyrotary@lancaironline.net; Tue, 13 Nov 2007 22:40:20 -0500 Received-SPF: pass receiver=logan.com; client-ip=209.85.134.186; envelope-from=wdleonard@gmail.com Received: by mu-out-0910.google.com with SMTP id i10so52215mue for ; Tue, 13 Nov 2007 19:39:43 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; 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; bh=IoRogTQbpHQYCr57fnV0EfPOC4bqutF5UwOq0vDy6sg=; b=UNtG3UrqjVeVWZcccVmigX5rN0G1ytlLnPsXN5GvHW2gPe6bxtl//xBie/3sDeEXP8ErerB01Z1nWSKlvyjq0M4zAPyomggcNgdLSNcgiqiHAvZtDohflAKkqXBBNOjjq/vpkPt5e4gAmTByd4kwGt/EhzD93CliQRabYHliEK4= 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=OIJd8MTXVTssgqg5Xe8tGBWED8PrtaH5cx2klcghwNb+poXIB3vLW/blEwfvH98lZH51zXuxrObfydi554wDE5YylUSmolrtQHqabp5oyfk2tTTjlCU2F5hQidMxuSord/OqhNjbdUHxSV09an/SIduyxUFcAUyFMryw1pRAtTk= Received: by 10.86.73.17 with SMTP id v17mr6471257fga.1195011583348; Tue, 13 Nov 2007 19:39:43 -0800 (PST) Received: by 10.86.98.1 with HTTP; Tue, 13 Nov 2007 19:39:43 -0800 (PST) Message-ID: <1c23473f0711131939w3ac471eas326d3dc50d304fa1@mail.gmail.com> Date: Tue, 13 Nov 2007 22:39:43 -0500 From: "David Leonard" To: "Rotary motors in aircraft" Subject: Re: [FlyRotary] Re: Rebutal to the rebutal {:>) Thick vs Thin was : Diffuser Configuration Comparison In-Reply-To: MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_2127_25966669.1195011583339" References: ------=_Part_2127_25966669.1195011583339 Content-Type: text/plain; charset=WINDOWS-1252 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Thanks Al, Saved me the trouble. That last bit about pressure recovery on exit is a nice touch, and you are right it is important - with either radiator. I know where this discussion is going, and we have there (nowhere) before. And I think we all realize that the difference (if there is one) is too small to be significant in our applications, so everyone just keep building and install what you can. Ed, your jedi mind trick smoke and mirrors will not work on me. The mind trick only works on intelligent life forms.... I think the disconnect there is that you are envisioning a guy with a thick radiator, who changes nothing other than the radiator and minimal duckting changes. But clear your mind, and picture this: Two guys (twins) Two Airplanes identical in every way except the radiator, inlet ducting and exit ducting. Frontal drag is the same, needed heat rejection is the same etc. The ducting systems are designed such that the mass airflow is identical. ie, the thin radiator version has smaller inlet and outlets but a larger plenum or whatever is necessary so that the airflow through the two radiators is identical. GIVEN identical airflow, identical rad volume, inlet air temp and identical coolant temp... the Delta T on average is the SAME. The air, although going slower in the thinner rad, spends exactly the same amount of time in each rad. (same amount of time for heat transfer and same average temp rise of the air and same average temp drop of the coolant) Now, I will admit that heat transfer is more efficient in the higher velocity thick rad. But it is not squared, or even linear. It is less tha= n linear. (ie, as velocity increases in both rads heat transfer becomes less dependent on velocity - they are both turbulent flow and you cannot achieve a lower temp than the ambient air. At some point the faster speeds become detrimental due to heating from friction) I submit that the the speeds of most of our aircraft and relative dimensions of radiators, both radiators are in turbulent flow and increased efficiency of heat transfer is nearly negligible. So in the end the difference between these two systems is that the thin rad has smaller velocities across the rad, smaller pressure drop, improved pressure recovery, higher exit air velocity, same delta T, same heat transfer, same exit air temp, smaller inlets and outlets, and larger volume sucked by ducting. All of which is relatively negligible in our applications. Thanks for the fun discussion Ed. Anxiously awaiting reason for Delta T difference between rads but will probably not buy it. (but still realize there is a very real possibility that I am wrong). Dave Leonard On Nov 13, 2007 8:45 PM, Al Gietzen wrote: > Ed; > > > > I swore after the last one, it was my last on this subject. But, OK; one > more. This is all a bit like the three blind men describing an elephant. > None are necessarily wrong, just different points of view. > > Let's look at it from a sort of systems approach: First I calculate the > heat load. Then I determine the mass flow rate of air I need using a > certain delta T; one that I know is 'reasonable' - or even optimum (radia= tor > size and weight, etc.), based on other analyses which we haven't/won't go > into, but a number between 50 and about 80F is good based on the OAT and > coolant temps we deal with. We already know how to compute the mass flow > rate. > > 1st addressing your statement regarding the inlet sizing. > > > > Assuming that selecting your inlet opening size controls mass flow > is incomplete. It is the total pressure loss for the entire duct (and co= re) > which combined with the available freestream kinetic energy (due to > velocity) that determines mass flow. I can make changes to any of the > intake, diffuser, core, and outlet and make changes in the mass flow - so > its not just the inlet. > > > > Then I compute the required opening area of my ram scoop based on the > velocity of the incoming air; the speed of the airplane at the design poi= nt; > generally a fast climb. Area x velocity =3D mass flow rate. The ram scoo= p is > not necessarily 100% efficient, so add maybe 10% to the area. Now, yes, > that is incomplete; but on first order, unless I do something wrong that > causes the scoop to spill air (like poor diffuser design or too thick a > radiator), then it is complete - that is the mass flow rate. Only change= s I > make that cause air to spill around the scoop is going to change that; so= on > this macro view I can ignore the details of the diffuser of which you are > more expert than I. > > > > Now if I have a well designed diffuser with an area ratio of say about 4+ > I know that the pressure recovery will be greater than the pressure drop > through the rad, so all that air is going through. Right? Well; here is > where we have to start thinking about both the thickness and the density = of > the core because the pressure drop through the core must be less than the > amount of pressure recovery =96 and is why I say that talking thickness > without specifying core density and diffuser ratio is 'incomplete' becaus= e > they depend on one another. > > > > But back to fist order; given the conditions of adequate pressure recover= y > the flow rate is fixed. > > > > 2nd regarding the deltaT "swap" I made: > > I do not agree that you need to compare on the basis of same mass flow or > deltaT - the only factor that really matters is that the needed heat be > removed. There are a large combinations of mass flow and deltaT that wil= l > remove X amount of heat. True, but: > > > > The design point heat load is fixed, the mass flow rate is fixed; > therefore the delta T is fixed (see formula used for computing the the fl= ow > rate in the first place.) > > > > Now, given these conditions; we can look at thick vs thin. We can have a > large frontal area, slow velocity through the core, and have pressure lef= t > over for accelerating the air back toward free stream (along with the hea= t > energy that we pick up which gives velocity by expansion); or we can make > the core thicker up to the limit of the pressure recovery that we have > achieved; and have no remaining pressure at the core exit. But keep in mi= nd > that the have a fixed scoop area for the speed we designed for; and the > frontal area of the rad is the other area in the diffuser ratio; so makin= g > the rad smaller and thicker both cut into the available pressure recovery= . > When we reach that limit where we have used all the available pressure > recovery; we have no pressure left over to accelerate the air back to > something closer to that at which it came in. > > > > Because of the effects of the velocity on heat transfer, as well as > pressure drop, there does happen to be a rad pressure drop (thickness) th= at > results in minimum drag =96 just as there is a corresponding delta T that > results in minimum mass of the core (all other things equal; i.e., > properly designed). > > > > Now the big caveat - It is clear here that to take advantage of the less > pressure drop in the thin rad to reduce drag, we have to have an exit > configuration that efficiently re-accelerates the air. If not, or if we > going to release the exit air into the free stream at a negligibly small > velocity, then it's a different ball game. Then from a drag standpoint > there may be little difference, and that using a radiator thickness > (pressure drop) that exhausts the pressure recovery, may be the way to go > for fitting into a confined space. > > > > So there ya go; cooling system design in a nutshell; minus all the magicJ= . > > > > Al G > > > > > > > > > > > > > > > --=20 David Leonard Turbo Rotary RV-6 N4VY http://N4VY.RotaryRoster.net http://RotaryRoster.net ------=_Part_2127_25966669.1195011583339 Content-Type: text/html; charset=WINDOWS-1252 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Thanks Al,

Saved me the trouble.  That last bit about pressure = recovery on exit is a nice touch, and you are right it is important - with = either radiator.

I know where this discussion is going, and we have = there (nowhere) before.

And I think we all realize that the difference (if there is one) is= too small to be significant in our applications, so everyone just keep bui= lding and install what you can.

Ed, your jedi mind trick smoke and m= irrors will not work on me.  The mind trick only works on intelligent = life forms....

I think the disconnect there is that you are envisioning a guy with= a thick radiator, who changes nothing other than the radiator and minimal = duckting changes.  But clear your mind, and picture this:

Two g= uys (twins)
Two Airplanes identical in every way except the radiator, inlet ducting= and exit ducting.  Frontal drag is the same, needed heat rejection is= the same etc.
The ducting systems are designed such that the mass airfl= ow is identical.  ie, the thin radiator version has smaller inlet and = outlets but a larger plenum or whatever is necessary so that the airflow th= rough the two radiators is identical.
GIVEN identical airflow, identical rad volume, inlet air temp and ident= ical coolant temp... the Delta T on average is the SAME.
The air, althou= gh going slower in the thinner rad, spends exactly the same amount of time = in each rad. (same amount of time for heat transfer and same average temp r= ise of the air and same average temp drop of the coolant)

Now, I will admit that heat transfer is more efficient in the highe= r velocity thick rad.  But it is not squared, or even linear.  It= is less than linear. (ie, as velocity increases in both rads heat transfer= becomes less dependent on velocity - they are both turbulent flow and you = cannot achieve a lower temp than the ambient air.  At some point the f= aster speeds become detrimental due to heating from friction)  I submi= t that the the speeds of most of our aircraft and relative dimensions of ra= diators, both radiators are in turbulent flow and increased efficiency of h= eat transfer is nearly negligible.

So in the end the difference between these two systems is that the = thin rad has smaller velocities across the rad, smaller pressure drop, impr= oved pressure recovery, higher exit air velocity, same delta T, same heat t= ransfer, same exit air temp, smaller inlets and outlets, and larger volume = sucked by ducting.  All of which is relatively negligible in our appli= cations.

Thanks for the fun discussion Ed.  Anxiously awaiting reason f= or Delta T difference between rads but will probably not buy it. (but still= realize there is a very real possibility that I am wrong).

Dave Leo= nard

On Nov 13, 2007 8:45 PM, Al Gietzen <= ALVentures@cox.net> wrote:
=

Ed;

 

I swore after the last one, it= was my last on this subject. But, OK; one more. This is all a bit like the three b= lind men describing an elephant.  None are necessarily wrong, just differen= t points of view.

Let's look at it from a sort o= f systems approach: = First I calculate the heat load.  Then I determine the mass flow rate of air I need using a certain delta T; one that I know is 'reasonable' - or even optimum (radiator size and weight, etc.), based on other analyses which we haven't/= won't go into, but a number between 50 and about 80F is good based on the OAT and= coolant temps we deal with.  We already know how to compute the mass flow rate= .

 1st addressing your statem= ent regarding the inlet sizing. 

=  

Assuming that selecting your inl= et opening size controls mass flow is incomplete.  It is the total pressure loss for the entire duct (and core) which combined with the availa= ble freestream kinetic energy (due to velocity) that  determines mass flow.   I can make changes to any of the intake, diffuser, core, and outlet and mak= e changes in the mass flow - so its not just the inlet. 

 

Then I compute the required op= ening area of my ram scoop based on the velocity of the incoming air; the speed of the airplane at the design point; generally a fast climb. Area x velocity =3D m= ass flow rate.  The ram scoop is not necessarily 100% efficient, so add ma= ybe 10% to the area.  Now, yes, that is incomplete; but on first order, unless= I do something wrong that causes the scoop to spill air (like poor diffuser design or too thick a radiator), then it is complete - that is the mass flo= w rate.  Only changes I make that cause air to spill around the scoop is going to change that; so on this macro view I can ignore the details of the= diffuser of which you are more expert than I.

 

Now if I have a well designed = diffuser with an area ratio of say about 4+ I know that the pressure recovery will b= e greater than the pressure drop through the rad, so all that air is going through. Right? Well; here is where we have to start thinking about both th= e thickness and the density of the core because the pressure drop through the core must be less than the amount of pressure recovery =96 and is why I say that talking thickness without specifying core density and diffuser ratio i= s 'incomplete' because they depend on one another. 

 

But back to fist order; given = the conditions of adequate pressure recovery the flow rate is fixed.

 

2nd regarding the deltaT "swap" I made: 

I do not agree that you need to compare on the basis of sam= e mass flow or del= taT - the only factor that really matters is that the needed heat be removed.  There are a large combinations of mass flow and deltaT that will remove X amount of heat. True, but:

 

The design point heat load is = fixed, the mass flow rate is fixed; therefore the delta T is fixed (see formula used f= or computing the the flow rate in the first place.)

 

Now, given these conditions; w= e can look at thick vs thin.  We can have a large frontal area, slow velocity thr= ough the core, and have pressure left over for accelerating the air back toward = free stream (along with the heat energy that we pick up which gives velocity by expansion); or we can make the core thicker up to the limit of the pressure recovery that we have achieved; and have no remaining pressure at the core = exit. But keep in mind that the have a fixed scoop area for the speed we designed for; and the frontal area of the rad is the other area in the diffuser rati= o; so making the rad smaller and thicker both cut into the available pressure recovery. When we reach that limit where we have used all the available pressure recovery; we have no pressure left over to accelerate the air back= to something closer to that at which it came in.

 

Because of the effects of the = velocity on heat transfer, as well as pressure drop, there does happen to be a rad pressure drop (thickness) that results in minimum drag =96 just as there is a corresponding delta T that results in minimum mass of the core (all other things equal; i.e., properly designed).

 

Now the big caveat - It is cle= ar here that to take advantage of the less pressure drop in the thin rad to reduce drag, we have to have an exit configuration that efficiently re-accelerates= the air.  If not, or if we going to release the exit air into the free str= eam at a negligibly small velocity, then it's a different ball game.  Then from a drag standpoint there may be little difference, and that using a radiator thickness (pressure drop) that exhausts the pressure recovery, may= be the way to go for fitting into a confined space.

 

So there ya go; cooling system= design in a nutshell; minus all the magicJ .

 

Al G

 

 

 

 

 

 

 




--
David Leonard

Tu= rbo Rotary RV-6 N4VY
http://N4V= Y.RotaryRoster.net
http://Rotary= Roster.net ------=_Part_2127_25966669.1195011583339--