Mailing List flyrotary@lancaironline.net Message #35066
From: al p wick <alwick@juno.com>
Subject: Re: [FlyRotary] Ideal cooling
Date: Thu, 4 Jan 2007 05:51:45 -0800
To: <flyrotary@lancaironline.net>
Lot's of true statements. But there's one vital component that's overlooked. If you look at crash history, you may notice that one of the main causes is marginal design. Designing systems on the edge of failure.
We have excellent example from forced landing just a few months ago. The guy that experienced overheating after changing to Evans coolant. If he had robust cooling system, he would not have considered changing coolant mid journey. Marginal cooling contributed to this decision. Forced landing resulted.
I require myself to have compelling reasons to make a design change to any engine system. So if I minimize the hose diameter, minimize the radiator surface area. What do I gain? Likely I save a half lb. So for that trivial weight advantage, I add risk every flight. I lug along a spray bar to compensate. I would have to watch climb rate, make sure I don't fly on hot days. This is the perfect setup for failure.
 
Strongly encourage robust cooling design. Don't let perceived value of weight reduction lead to risky decision.
 

-al wick
Cozy IV powered by Turbo Subaru 3.0R with variable valve lift and cam timing.
Artificial intelligence in cockpit, N9032U 240+ hours from Portland, Oregon
Prop construct, Subaru install, Risk assessment, Glass panel design info:
http://www.maddyhome.com/canardpages/pages/alwick/index.html
 
 
 
On Wed, 3 Jan 2007 13:25:55 -0600 "M Roberts" <montyr2157@alltel.net> writes:
Just so that we are all on the same page here.
 
First the disclaimer:
 
The following assumes that you are flying a subsonic airplane where compressibility effects can be neglected, and that you are not putting out 1000+hp. I think these are fairly safe assumptions. The other thing is it must be steady state with stabilized readings relative to time. The math to look at this case is really very simple. If you can pass algebra, you can do the math for this problem. Just keep the units consistent and use absolute temps for calculating DT (deg Kelvin or Rankine depending on metric or english units)
 
The job of the cooling system is to remove enough waste heat to keep the engine below its maximum temp in all operating ranges. This must be done while keeping the weight and drag to a minimum.
 
The conflict is always between hot day climb and cruise efficiency.
 
The airplane spends most of its time in cruise, therefore cruise efficiency is the most important design point.
 
Hot day climb only happens once in a while. It is the secondary concern. Adding a spray bar or some other band aid to get past this one design point is a perfectly logical and acceptable solution. It is not "cheating". 
 
In steady state cruise there is some heat flux Q from the engine that must be disposed of to keep the engine at its desired operating temp. There is no "optimum" coolant delta T (DT)across the radiator for all conditions to accomplish this. There is only ONE coolant DT for any given steady state operating point that will achieve this and it is found by the equation
 
Q = Mdot x Cp x DT
 
Q = The heat flux that must be removed
 
Mdot = mass flow rate of the water per unit time
 
Cp = The specific heat of the coolant (how much heat each unit of mass of the coolant absorbs per unit of temp increase)
 
DT = T1-T2 The change in temp of the water across the radiator (must be equal to the water DT across the engine block for steady state operation)
 
Now for the optimum part:
 
The absolute best possible performance would be a system that used just enough air Mdot to heat the cooling air to the same temperature as the water radiator exit temp ( water going back to the engine from the radiator). This condition is impossible to achieve in practice. So you try to get as close as you can. How close you get is called the heat exchanger effectiveness. The closer the air and water exit temps are, the better the effectiveness.
 
The air side is also governed by the same equation:
 
Q = Mdot x Cp x DT  
 
This is because Qwater = Qair for steady state operation
 
The Cp for air is much less than for water. The density of water is also much greater than air.  Because of this the water DT will always be much less than the Air DT
 
The optimum condition is to minimize Mdot air (thus drag)  for a given heat exchanger effectiveness to get the required Q. This results in the smallest cooling system and the smallest drag/weight in cruise.
 
The catch is you must oversize the system just enough so that a cowl flap can increase Mdot air enough in hot day climb to adequately cool the engine. In this regime there will be more drag and less Air DT. Who cares! Cruise is where its at.
 
Transient operation can only be analyzed using differential equations, or piecewise analysis and a computer code. Climb is transient in nature, that is the problem with trying to analyze it. The transient nature of climb is why the thermal mass of the engine and coolant help us. It absorbs heat as it comes up to temp, This is all Q that the rad does not have to reject to the air. Taking off with the engine already at red line temp is a different matter entirely. In this case you have no big heat sink to help you. So use a lower rate of climb, a higher speed, open cowl flaps and a spray bar. Or better yet, let the engine cool before you take off then do all those things.
 
For my design I prefer a light low drag installation that is optimized for cruise with some Band-Aids to get me past the once in a while condition that I rarely see.
 
If you are designing a glider tug or a STOL plane for use in North Africa, you will have to adjust your design accordingly.
 
Monty
 
 
 

-al wick
Cozy IV powered by Turbo Subaru 3.0R with variable valve lift and cam timing.
Artificial intelligence in cockpit, N9032U 240+ hours from Portland, Oregon
Prop construct, Subaru install, Risk assessment, Glass panel design info:
http://www.maddyhome.com/canardpages/pages/alwick/index.html
Subscribe (FEED) Subscribe (DIGEST) Subscribe (INDEX) Unsubscribe Mail to Listmaster