X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Received: from cdptpa-omtalb.mail.rr.com ([75.180.132.120] verified) by logan.com (CommuniGate Pro SMTP 5.2c1) with ESMTP id 2468344 for flyrotary@lancaironline.net; Mon, 12 Nov 2007 17:14:30 -0500 Received-SPF: pass receiver=logan.com; client-ip=75.180.132.120; envelope-from=eanderson@carolina.rr.com Received: from edward2 ([24.74.103.61]) by cdptpa-omta06.mail.rr.com with SMTP id <20071112221352.ZIEO507.cdptpa-omta06.mail.rr.com@edward2> for ; Mon, 12 Nov 2007 22:13:52 +0000 Message-ID: <002b01c82579$9b8d2c10$2402a8c0@edward2> From: "Ed Anderson" To: "Rotary motors in aircraft" Subject: Generalizations was Re: Diffuser Configuration Comparison Date: Mon, 12 Nov 2007 17:16:00 -0500 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0028_01C8254F.B276BFB0" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.3138 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3138 This is a multi-part message in MIME format. ------=_NextPart_000_0028_01C8254F.B276BFB0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Once again,Al, I agree that generalizations can mislead - if taken too = literally . But, on the other hand I think generalization is very = useful approach to making points and conveying understanding. =20 I think each cooling system starts out as a conceptual generalization. = We generalize about how much power we think we are going to produce = (unless we decide to dyno the engine - certainly a good move, but one = seldom done) and therefore how much heat we need to get rid of. we also = probably start with a rough estimate of core size (perhaps its a core we = already have or readily available-such a GM cores or the space we have = allotted for it) , then we must consider the kinetic energy available in = the air stream for our critical cooling design point (should it be climb = or cruise). =20 I think generalizations are useful in this process. For instance, I = think its fairly safe to say that for our operational environment that a = 1" thick radiator core is probably not the most suitable dimension. Can = one be used, certainly, but its frontal area is going to be large (space = constraints may inhibit) and cooling drag is going to be higher. Now if = you are flying a bi-plane then cooling drag may be such a small part of = your total drag that you wouldn't gain much by reducing it. But, at = your speeds and even my slower speed, it starts to become an appreciable = portion of total drag. Fin density, type fins, core thickness, turbulation or not, core = orientation of the air flow, duct losses, drop across the core, = hydraulic diameter, Reynolds number, etc, etc, all play a role in the = final out come. Then no matter how carefully you "design" you cooling = system, leaks, sharp edges on any turns, distance between back of core = and obstructions, boundary layer ingestion, etc, installation = imperfections can mess up the performance of the "perfect" system. = Again this is all generalizations, but that does not necessarily mean = they are not correct. I have spent weeks, months (slow learner) reading and studying all I = could find and reasonably understand about the airflow side this = problem. It has really pointed out that effective cooling is all about = balancing conflicting aerodynamic and thermodynamic factors. It is = certainly beyond my ability to grasp it all, but there are certain = things that I have become convinced of and more will follow (I hope) as = I study further. =20 1. Mass flow through the core is the most critical element of cooling. = If there is insufficient mass flow then it does not matter how good you = ducting or core is , you will not meet your cooling objective. Your air = mass flow requirement is dependent on your heat rejection needs.=20 2. The maximum duct mass flow possible is a function of free stream = kinetic energy available. This means you cooling design point airspeed = is as much (or more) a crucial factor in your design as any other = factor.=20 3. Many factors determine what you actually mass flow will be, these = include both design, fabrication, installation, environmental and = operational factors. A pretty general statement, but valid just the = same. Its the nailing down of the factors in this area that to me = represents the most beneficial (and the most difficult) factors to = understand in detail. 4. The maximum flow in the ducts (and through the core) is a function = of the free stream kinetic energy and the pressure loss coefficient of = the duct (and core). =20 5. Air Flow separation in the diffuser is the most significant factor = in degrading core effectiveness. Separation reduces cooling by reducing = mass flow, by creating pressure losses, disrupting even velocity = distribution across the core and increasing drag. =20 6. Diffuser's performance depend, in significant part, on the core = characteristics. 7. It is a balancing and optimization problem of opposing aerodynamic = and thermodynamic attributes. 8. If you had enough core and enough air flow - you will cool, but the = penalty in drag and weight may be higher than you would like. 9. Few of us have the knowledge, understanding, tools, time, $$ or = inclination to do it the right the first time , but always time to = re-do-it after the first flight {:>) Besides the generation that appeared to bring this discussion about was = that thicker radiators offer advantages at higher airspeeds. I still = stand by that generalization. note. I did not say that 2 1/2" was too thin or 7" was too thick. But, = I do believe that the Nascar crowd have the resources and inclination to = do the research on radiator size that none of us do have. There speeds = are comparable to ours, so again, I personally feel that a core in the = vicinity of 3" thick sets a bench mark that is probably as valid as = anything we could afford to do.=20 Just because my GM cores happen to be 3 1/2" thick has nothing to do = with it {:>) Appreciate you comments, Al. I will try to hold my generalizations to = an ...A'hem ... acceptable minimum {:>) Best Regards Ed Ed Anderson Rv-6A N494BW Rotary Powered Matthews, NC eanderson@carolina.rr.com http://www.andersonee.com http://members.cox.net/rogersda/rotary/configs.htm#N494BW http://www.dmack.net/mazda/index.html ------=_NextPart_000_0028_01C8254F.B276BFB0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Once again,Al,  I agree that = generalizations can=20 mislead - if taken too literally .  But, on the other hand I=20 think generalization  is very useful approach to =  making=20 points and conveying understanding. 
 
I  think each cooling system starts = out as a=20 conceptual generalization.  We  generalize about how much = power we=20 think we are going to produce (unless we decide to dyno the engine - = certainly a=20 good move, but one seldom done) and therefore how much heat we need to = get rid=20 of.  we also probably start with a rough estimate of core size = (perhaps its a core we already have or readily available-such a GM cores = or the=20 space we have allotted for it) , then we must consider the kinetic = energy=20 available in the air stream for our critical cooling design point = (should it be=20 climb or cruise). 
 
I think generalizations are useful in this = process. =20 For instance, I think its fairly safe to say that for our operational=20 environment that a 1" thick radiator core is probably not the most = suitable=20 dimension.  Can one be used, certainly, but its frontal area is = going to be=20 large (space constraints may inhibit) and cooling drag is going to be=20 higher.  Now if you are flying a bi-plane then cooling drag may be = such a=20 small part of your total drag that you wouldn't gain much by reducing = it. =20 But, at your speeds and even my slower speed, it starts to become an = appreciable=20 portion of total drag.
 
Fin density, type fins, core thickness, = turbulation or=20 not, core orientation of the air flow, duct losses, drop across the = core,=20 hydraulic diameter, Reynolds number, etc, etc, all play a role in the = final out=20 come.  Then no matter how carefully you "design" you cooling = system, leaks,=20 sharp edges on any turns, distance between back of core and = obstructions,=20 boundary layer ingestion, etc, installation imperfections can mess up = the=20 performance of the "perfect" system.  Again this  is all=20 generalizations, but that does not necessarily mean they are not=20 correct.
 
I have spent weeks, months (slow learner) =  reading=20 and studying all I could find and reasonably understand about the = airflow=20 side this problem. It has really pointed out that effective = cooling is=20 all about balancing conflicting aerodynamic and thermodynamic=20 factors.   It is certainly beyond my ability to grasp it all, = but=20 there are certain things that I have become convinced of and more will = follow (I=20 hope) as I study further. 
 
1.  Mass flow through the core  is the = most=20 critical element of cooling.  If there is insufficient mass flow = then it=20 does not matter how good you ducting or core is , you will not meet your = cooling=20 objective.  Your air mass flow requirement is dependent on your = heat=20 rejection needs.
 
2.  The maximum duct  mass flow = possible is a=20 function of free stream kinetic energy available. This means=20 you cooling design  point airspeed is as much (or more) =  a=20 crucial factor in your design as any other factor. 
 
3.  Many factors determine what you actually mass flow  = will be,=20 these include both design, fabrication, installation, environmental and=20 operational  factors.  A pretty general statement, but valid = just the=20 same.  Its the nailing down of the factors in this area that to me=20 represents the most beneficial (and the most difficult) factors to = understand in=20 detail.
 
4.  The maximum flow in the ducts (and = through the=20 core) is a function of the free stream kinetic energy and the =  pressure=20 loss coefficient of the duct (and core). 
 
5.  Air Flow separation in the diffuser is = the most=20 significant factor in degrading core effectiveness.  Separation = reduces=20 cooling by reducing mass flow,  by creating pressure losses,=20 disrupting even velocity distribution across the core and increasing = drag. =20
 
6.  Diffuser's performance depend, in significant part, on the = core=20 characteristics.
 
7.  It is a balancing and optimization = problem=20 of opposing  aerodynamic and thermodynamic = attributes.
 
8.  If you had enough core and enough air flow - you will = cool, but=20 the penalty in drag and weight may be higher than you would like.
 
9.  Few of us have the knowledge, understanding, tools, time, = $$ or=20 inclination to do it the right the first time , but always time to = re-do-it=20 after the first flight {:>)
 
 
Besides the generation that appeared to bring this discussion about = was=20 that thicker radiators offer advantages at higher airspeeds.  I = still stand=20 by that generalization.
 
note. I did not say that 2 1/2"  was too thin or 7" was too=20 thick.  But, I do believe that the Nascar crowd have the resources = and=20 inclination to do the research on radiator size that none of us do = have. =20 There speeds are comparable to ours, so again, I personally feel that a = core in=20 the vicinity of 3" thick sets a bench mark that is probably as valid as = anything=20 we could afford to do. 
 
 Just because my GM cores happen to be 3 1/2" thick has = nothing to do=20 with it {:>)
 
Appreciate you comments, Al.  I will try to hold my = generalizations to=20 an ...A'hem ... acceptable minimum {:>)
 
Best Regards
 
Ed
 
 
 
Ed Anderson
Rv-6A N494BW Rotary = Powered
Matthews,=20 NC
eanderson@carolina.rr.comhttp://www.andersonee.com
http:/= /members.cox.net/rogersda/rotary/configs.htm#N494BW
http://www.dmack.net/mazda= /index.html
 
------=_NextPart_000_0028_01C8254F.B276BFB0--