What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Water Cooled vs. Air Cooled- Real Numbers

Easy guys, we're all friends, right?

Here are a few points I have discovered in the three (almost) years I've been flying with my LS1.
? I cruise at the same speeds as the other two RV-10s at our airport
? I burn the same amount of gas as they do
? I change oil as often as they do
? Our most frequent comments are about the high cost of Garmin databases

My biggest question to myself is, why do an engine run up, at all.
There's no mags to check.
All I do is cycle the prop one, or two times, because I always did it in the past.
The real reason I do it is so that other guys don't think I'm being cavalier or reckless....

John
 
I explained the point of this exercise in the initial post. Basically, there was minimal, modern, concrete data quantifying radiator performance and drag for liquid cooled aircraft but plenty of wild conjecture. At the same time there was also widespread belief that air cooled engines would always have less cooling drag than their liquid cooled counterparts due to the higher delta T. My numbers have shown that liquid cooled installations don't suffer higher momentum loss and don't require higher mass flow either. Other flying installations such as Russell Sherwoods Subaru Glasair also confirm the latter case.

I set out to get actual numbers in flight and was keen to compare numbers with some of the great data Dan Horton and others here on VAF had come up with on the Lycoming installations.

Of primary interest was measuring the momentum loss of the cooling air stream- in other words to prove the validity of the Meredith Effect on an actual flying aircraft.

I've already stated I don't have a good way of measuring the extra profile drag of the actual scoop.

Employing exhaust energy to pump the cooling air stream is practically quite difficult on an RV and even a clean sheet tractor design. It has been done on at least one pusher I know of and worked well for ground cooling.

I've seen other attempts at exhaust pumping but the exit geometry was not optimized for low cruise drag, usually with fixed exits and other design compromises. We know for a fact that low cruise drag, while retaining adequate climb cooling margin, requires a variable exit- both on air and liquid cooled installations.

I was not able to complete all the instrumented testing yet due to the unusually long winter here but should soon be able to. So far, I do know that the layout offers much better cooling margins than the previous setup on the ground and in flight. It certainly shows the worth in drag reduction over fixed exit designs.

There is nothing new here other than the data- the WW2 design engineers already knew and proved all this 70 years ago. The fact that most people today don't apply this old knowledge to their designs and continue to use poor layouts with marginal cooling and high drag baffles me.

I won't offer any conjecture here about the numbers at higher speeds. There is no point to speculate. The in-flight data will tell the real story which was the whole point of this rather time consuming exercise.
 
Last edited:
rv6ejguy said:
The fact that most people today don't apply this old knowledge to their designs and continue to use poor layouts with marginal cooling and high drag baffles me.
Click to expand...

Reading the other 7 page long thread you have on this same subject, you stated 300 hours of work on your P51 Rad/scoop. Thats a lot of effort. I do commend your effort, but I see nothing "baffling" about "most people" not investing 300 hours for such minimal gain, if any. (once the drag of the scoop itself is quantified) What are the numbers as far as improvement in airspeed vs your older cooling system? I should think thats all that matters in the end, provided an equally effective cooling system.
 
Last edited:
johngoodman said:
Easy guys, we're all friends, right?

Here are a few points I have discovered in the three (almost) years I've been flying with my LS1.
? I cruise at the same speeds as the other two RV-10s at our airport
? I burn the same amount of gas as they do
? I change oil as often as they do
? Our most frequent comments are about the high cost of Garmin databases

My biggest question to myself is, why do an engine run up, at all.
There's no mags to check.
All I do is cycle the prop one, or two times, because I always did it in the past.
The real reason I do it is so that other guys don't think I'm being cavalier or reckless....

John
Click to expand...

Sounds like you are having good success here John. :)
 
kengps said:
Reading the other 7 page long thread you have on this same subject, you stated 300 hours of work on your P51 Rad/scoop. Thats a lot of effort. I do commend your effort, but I see nothing "baffling" about "most people" not investing 300 hours for such minimal gain, if any. (once the drag of the scoop itself is quantified) What are the numbers as far as improvement in airspeed vs your older cooling system? I should think thats all that matters in the end, provided an equally effective cooling system.
Click to expand...

The 300+ hours spent on my rad mods are mostly due to my poor composite skills. Someone like Dan could have done a way better job in 1/3rd the time I expect.

I haven't quantified speed gain yet because testing was cut short by winter but just as Dan found tangible speed gains on the Lycoming with the throttled exit, I'd expect something useful on mine. Test data from WW2 on various fighter aircraft showed speed increases of up to 32 mph with with rad exit doors closed- hardly something to sneeze at.

The new vs. old on my setup shows a very large increase in cooling margin both on the ground and in the air so that part alone has been worth the effort.
 
OK, thanks for the info. I didn't read past about page 3 of this thread, but did you account for the difference in air volume with different exit door positions? I mean for example that a 25% increase in velocity, at half the exit size is not a 25% thrust advantage. Velocity for velocities sake means nothing if only half the volume is being accelerated 25% more. Gonna need to consider the combination of volume AND velocity to arrive at the optimum exit door setting if you wish to take advantage of the Meredith effect. As for the Meredith effect, ever considered a coolant bypass for a short duration to see what the difference in air velocity is as the air out of the radiator cools, then heats back up on return of the heat energy to it again? Not sure how long you can do this on the Sub before it gets too hot. I would think at least 15 seconds would be enough for a data sample?? Of course the velocity measurement would need temperature compensation for density changes with exit air temp.
 
kengps said:
Velocity for velocities sake means nothing if only half the volume is being accelerated 25% more.
Click to expand...

mass x momentum loss = cooling drag

Reducing mass while increasing velocity is precisely the goal.
 
DanH said:
mass x momentum loss = cooling drag

Reducing mass while increasing velocity is precisely the goal.
Click to expand...

I thought he was trying for thrust with the Meredith Effect. More mass, and more velocity equals more thrust.

Reduce the cooling exit in half, while increasing the velocity less than 100% will provide less thrust, no?

Cooling requires heat transfer to the air mass. So unless the combination of volume x velocity is flowing more mass than a system with more mass at a lower velocity you're gaining nothing. Plus more mass at lower velocity equals more time for heat transfer to that mass.
 
Last edited:
While building race engines for example, my engine builder from my racing days didn't try for more velocity unto itself while spending countless hours on the Flow Bench. More velocity thru less total area thru the ports is pointless. There is a happy medium where the total mass flow is greatest. That point could possibly be with more area at less velocity. Or it could be at higher velocity at less exit area. An extreme example would be an itty-bity cooling exit with very high velocity. It will not flow as much mass as a larger exit area at less velocity. And so I ask the OP if he is considering the exit area in combination with the velocity to arrive at the largest air mass exiting the airplane. Exit area and Velocity are both needed to arrive at a conclusion. Leave one out and your not getting the correct answer.
 
DanH said:
mass x momentum loss = cooling drag

Reducing mass while increasing velocity is precisely the goal.
Click to expand...

I think not. Passing a Given (needed) mass of air thru the cooling system is precisely the goal. Less "mass" by way of higher velocity/less area equals less cooling. More mass is a combination of velocity AND area. It is not velocity alone.
 
Last edited:
I am certainly no expert here. On the contrary, I have learned much from many of you. I have a question though...

The more the air is heated, the more it expands (or wants to all other things considered equal) and therefore the volume of air coming out of the rad ductwork is technically larger than what went in (same mass of course). The air is "contained within the closed cooling system for a brief time, this in turn produces thrust if all other things are considered equal and efficient. Is it negligible?

If the heating effect does create a bit of thrust, this would show up as a loss of drag when optimized. Therefore, ductwork design will have a sweet spot whereby inlet/outlet ratios, aircraft speed, rad size (drag), rad inlet/outlet locations etc. will work together for better ( or worse). Sort of a black art.

definition (mine): "Black Art". A varying mixture of science, theory, intuition, fresh ideas from unlikely sources, testing, failures, more failures, still more failures, and finally some success often caused, at least partially, by accident. :D

Bevan
 
kengps said:
While building race engines for example, my engine builder from my racing days didn't try for more velocity unto itself while spending countless hours on the Flow Bench. More velocity thru less total area thru the ports is pointless. There is a happy medium where the total mass flow is greatest. That point could possibly be with more area at less velocity. Or it could be at higher velocity at less exit area. An extreme example would be an itty-bity cooling exit with very high velocity. It will not flow as much mass as a larger exit area at less velocity. And so I ask the OP if he is considering the exit area in combination with the velocity to arrive at the largest air mass exiting the airplane. Exit area and Velocity are both needed to arrive at a conclusion. Leave one out and your not getting the correct answer.
Click to expand...

High velocity without flow loss in a cylinder head usually has the potential to make more power and exactly what most head designers or modifiers try to achieve, however we are straying off topic here. I never found my flow bench time wasted doing engine development.

The inside of the rad duct is an aerodynamic structure. By careful shaping, we can take in the minimum amount of air (less drag), slow it down at the rad face to increase pressure and velocity (less drag again though the rad matrix) and allow this air to be heated the maximum amount (adding energy and expanding the air), finally being able to expel it close to, or ideally, above the free stream velocity.

The less amount of mass flow passing through the duct and rad, the less drag. Obviously if you close the exit too much, the flow stagnates and you also have insufficient cooling. Since aircraft operate over a wide speed range, the variable area exit door is the answer to control mass flow and exit velocity. Worse scenario is usually full power climb on a hot day. We don't care so much about drag here so we can open the exit door for maximum mass flow through the rad to dissipate the maximum energy from the coolant..

In cruise where we might have double the mass flow available, we can throttle flow with the exit door and increase exit velocity.
 
Bevan said:
I am certainly no expert here. On the contrary, I have learned much from many of you. I have a question though...

The more the air is heated, the more it expands (or wants to all other things considered equal) and therefore the volume of air coming out of the rad ductwork is technically larger than what went in (same mass of course). The air is "contained within the closed cooling system for a brief time, this in turn produces thrust if all other things are considered equal and efficient. Is it negligible?

If the heating effect does create a bit of thrust, this would show up as a loss of drag when optimized. Therefore, ductwork design will have a sweet spot whereby inlet/outlet ratios, aircraft speed, rad size (drag), rad inlet/outlet locations etc. will work together for better ( or worse). Sort of a black art.

definition (mine): "Black Art". A varying mixture of science, theory, intuition, fresh ideas from unlikely sources, testing, failures, more failures, still more failures, and finally some success often caused, at least partially, by accident. :D

Bevan
Click to expand...

Various sources estimate that something like 15 to 25% of total drag on a piston aircraft is due to cooling drag. I think it's worthwhile in most cases to try to minimize it.

It seems like a better approach to instrument the duct and try to learn what is happening by applying science rather than conjecture. This should allow us to arrive at a better rad/duct layout quicker than blind trial and error.

I tried to do this as much as possible given my limited resources, starting with flow bench testing to pick the lowest drag rad design with the highest heat dissipation rates. I progressed next to using some data from Russell Sherwood's successful Glasair SARL winning racer which had a similar ventral rad setup. We both found we had to add inlet guide vanes to stop separation prior to rad entry. This was done with tuft testing and high capacity blowers feeding the actual duct.

Neither of us know how well optimized our layouts really are as we are just at the tip of the iceberg as far as experimentation and measurement goes so far. My goal was never to actually obtain thrust, rather to quantify the the momentum loss/ gain possible at these 100-170 knot speeds and reduce drag over the existing system if possible while increasing cooling margin.
 
Last edited:
 
kengps said:
I think not. Passing a Given (needed) mass of air thru the cooling system is precisely the goal. Less "mass" by way of higher velocity/less area equals less cooling. More mass is a combination of velocity AND area. It is not velocity alone.
Click to expand...

You're making the assumption that the delta-T stays the same - it doesn't have to. If you can dump more heat into a given mass of air, less total mass is required for the same cooling. More efficient rads bringing the approach temps down will make a huge difference, which is what Ross is looking for.
 
Nope.

See Bernoulli.

http://en.wikipedia.org/wiki/Bernoulli's_principle

Air cooled flat engines use the same principle to induce flow through the cylinders. Look at the expansion in area from your cowl inlets to the plenum space above the cylinders. Carbs use the same effect (in reverse) to generate suction to pull fuel out of the bowl.
 
Last edited:
rv6ejguy said:
By careful shaping, we can take in the minimum amount of air (less drag), slow it down at the rad face to increase pressure and velocity (less drag again though the rad matrix)
Click to expand...

I'm assuming you meant slow it down to "increase pressure" and "Reduce" "velocity". Slowing air down to increase pressure will give you a velocity loss.

I concur on everything you said. We have cowl flaps to reduce mass flow on air-cooled aircraft as well. It works quite simply by reducing exit cross-section of the airflow.
But I thought you were trying to produce thrust via Meredith effect. There will be an optimum point for a given airspeed where thrust produced will be optimised at a certain exit door position (or fixed area). Too big an exit and you lose velocity. Too small an exit and you lose volume. In order to tune that optimum exit size you need to know at what pressure/cross-section maximum mass flow is occurring. My point was that you won't get that answer while only looking at the pressure behind the radiator. The exit area of your variable door must be added to the computation to get total air mass out. You could close the air-door completely and get great pressure. But the air pressure data alone is not telling you what you want to know. See what I'm saying?
The expansion of the air in a jet engine (derived from heat energy) is what gives it thrust when the new, larger volume of air is constricted back down later to create velocity again.
So in your Meredith Radiator you'll need an increasing cross-section at the same time the air is being heated, and expanded. Then a constriction later down the line, to accelerate that larger volume of heated air back up to velocity to create thrust. Otherwise all you'll be doing is creating more pressure inside the radiator as the air gathers pressure from the heat energy. That will accomplish nothing but create pressure. That will slow down the air coming into the Radiator. I think it's not going to be easy designing the proper chamber volumes, and velocity changes to accomplish all that. Plus where are you going to find a radiator that allows expansion of air as it moves thru the core, at the proper expansion rate? I've never seen this in my experience. Probably because radiators were never built to produce thrust. Might need to build something custom. Or perhaps a very expensive high-tech radiator from a formula 1 car where every little bit of efficiency and money is no object? Think of the radiator with a combustion chamber analogy. Combustion chambers allow expansion at the same exact moment air heating/expansion occurs. That'll get you on the correct path to creating thrust.
 
Last edited:
kengps said:
I'm assuming you meant slow it down to "increase pressure" and "Reduce" "velocity". Slowing air down to increase pressure will give you a velocity loss.

I concur on everything you said. We have cowl flaps to reduce mass flow on air-cooled aircraft as well. It works quite simply by reducing exit cross-section of the airflow.
But I thought you were trying to produce thrust via Meredith effect. There will be an optimum point for a given airspeed where thrust produced will be optimised at a certain exit door position (or fixed area). Too big an exit and you lose velocity. Too small an exit and you lose volume. In order to tune that optimum exit size you need to know at what pressure/cross-section maximum mass flow is occurring. My point was that you won't get that answer while only looking at the pressure behind the radiator. The exit area of your variable door must be added to the computation to get total air mass out. You could close the air-door completely and get great pressure. But the air pressure data alone is not telling you what you want to know. See what I'm saying?
The expansion of the air in a jet engine (derived from heat energy) is what gives it thrust when the new, larger volume of air is constricted back down later to create velocity again.
So in your Meredith Radiator you'll need an increasing cross-section at the same time the air is being heated, and expanded. Then a constriction later down the line, to accelerate that larger volume of heated air back up to velocity to create thrust. Otherwise all you'll be doing is creating more pressure inside the radiator as the air gathers pressure from the heat energy. That will accomplish nothing but create pressure. That will slow down the air coming into the Radiator. I think it's not going to be easy designing the proper chamber volumes, and velocity changes to accomplish all that. Plus where are you going to find a radiator that allows expansion of air as it moves thru the core, at the proper expansion rate? I've never seen this in my experience. Probably because radiators were never built to produce thrust. Might need to build something custom. Or perhaps a very expensive high-tech radiator from a formula 1 car where every little bit of efficiency and money is no object? Think of the radiator with a combustion chamber analogy. Combustion chambers allow expansion at the same exact moment air heating/expansion occurs. That'll get you on the correct path to creating thrust.
Click to expand...

No typo or mistake. Air slows down at the rad face due to the increase in cross sectional area of the diffusion duct. As it does so, pressure increases. I measured the pressure recovery at 82% which is pretty good while cooling the engine.

This was an experiment to gather data. If I got thrust, I'd be overjoyed but I didn't expect to see that, especially with 70C coolant temps.

I think you are failing to understand that the exit door throttles mass flow to minimize internal drag in the duct as well as pressure drop across the core. Mass in = mass out. It is all about momentum recovery. In cruise we don't need to pass much air through the rad. If we pass 50 lbs./min through the duct and achieve parity in the exit V compared to free stream or 500 lbs./min., we have offset all internal drag within the duct and rad in either case.

The radiator duct has its maximum cross sectional area at the rad and converges aft of the rad. Exactly the same as a P51. We are adding energy as the air passes through the rad matrix. Heated air MUST expand, the principle is little different from a ramjet, we are just working with about 1/10th the Delta T. The instrumentation clearly shows the theory works in practice. The rad is very efficient at heat transfer compared to an air cooled cylinder head and we are seeing 97% momentum recovery at the exit so far. Not bad for what was essentially an eyeballed shape.
 
Last edited:
You said Slow the air down. Your first post said increased pressure and velocity. Slow down and increased velocity are the opposite.
Agreed "Heated air must expand" But it expands in all directions. Not only the way you want it to expand. This expansion creates pressure. Pressure inside the core will slow down the incoming air into the radiator, negating any gain from the expansion in the downstream direction. The rest I completely agree. Cooling drag is not rocket science. Producing thrust is :)
 
Last edited:
Ken, here is a one page description from McCormick, an accepted reference. I am sure Ross can help you find more extensive works.

http://www.vansairforce.com/community/showpost.php?p=556582&postcount=117

No, I don't expect exit velocity higher than freestream (i.e. thrust) with a piston installation. Our (Ross and I, air or water cooled) goal is to see how close we can get. I think the water cooled installation can get closer simply because the duct system can be so clean internally...less energy robbing stuff in the flow. That is true cooling drag reduction.

On the flip side, the duct system, as a add-on to the RV airframe, is dirty in terms of external drag (unfaired shape, extra frontal area, etc) while the air-cooled RV install can be optimized to reduce frontal area compared to stock.

Obviously, the best solution would be to incorporate water cooling into the airframe design, starting with a clean sheet (umm, screen).

The system Ross is working with on the RV6 is really a test mule for his RV-10 project. What are you building? When can we expect to see your work here?
 
Last edited:
kengps said:
You said Slow the air down. Your first post said increased pressure and velocity. Slow down and increased velocity are the opposite.
Agreed "Heated air must expand" But it expands in all directions. Not only the way you want it to expand. This expansion creates pressure. Pressure inside the core will slow down the incoming air into the radiator, negating any gain from the expansion in the downstream direction. The rest I completely agree. Cooling drag is not rocket science. Producing thrust is :)
Click to expand...

I meant to add the word decrease in front of velocity (corrected now). Sorry for that confusion.

Of course the air expands in all directions and the expansion is contained within the duct. Momentum is always in the aft direction and we use the convergence of the duct to increase velocity at the exit.

We want to recover pressure at the rad face which is why we use a diffusion duct. I am not sure what you are debating here. The pressure, temperature and velocity measurements that Dan and I have done confirm the theory. Others have measured a speed increase indicating a drag decrease. Can we produce any net thrust? Unknown at this point but it seems possible if we can add more energy to the air, possibly with higher coolant temps than I am currently running. We are close to parity already for internal drag. The guys racing at Reno are using very similar layouts. This empirical data seems to contradict your views that there is no gain or point to doing all this.

As Dan said, mine is a test example on a non-optimized airframe and layout, With a clean sheet design, we should be able to do better by submerging the rad to negate the extra frontal area and have a longer and smoother duct shape.
 
Last edited:
No debate you are improving pressure recovery, thus less drag. I'm trying to explain how thrust is produced, and how to benefit from it. I do think the potential is extremely low because of the relatively low energy levels in the system. But in order to produce thrust (which will in net reduce drag) you must have air volume expanding with no pressure increase as energy is applied to it. The purpose of a jet engine is to convert a small volume of highly compressed air, into a larger volume of highly compressed air. This is done by adding energy to expand the air…. as the volume increases upon entering the combustion chamber. But a typical radiator does not allow expansion as energy is applied. It will merely create more pressure in all directions, but no volume gain. When that air (at increased pressure) leaves the radiator to the larger area at the rear face of the radiator, it simply expands again. Energy lost. You will get a net zero return.
Here is the key….A jet engine does not add pressure in the combustion process. It adds "volume" when the energy is applied. Pressure is maintained with heat as the volume is increased. The larger volume is then used in the exit nozzle to create velocity/thrust.
SO……I'm done splainin. My suggestion is this: You want thrust to reduce the drag component from the radiator? Get a radiator that allows increased volume, and not just pressure as the air is heated. Pressure pushes in all directions. It will slow the air going into the radiator, even as it speeds up the air exiting. Net Zero. I am sure the Meredith effect does not occur with a constant chamber volume radiator. There is a difference between a Jet engine, and a space heater.
Honestly I think the gain will be so small that its a huge waste of money. Custom built heat exchangers are very expensive. Keep working on your cooling drag reduction. Sounds like you're getting some good results. Meredith effect is a poor application for money/effort. Good talking to ya.

If I can find it I'll post a photo of a radiator like this. I have seen one. It was much like an airfoil with the cooling tubes getting thinner from front to back. This maintains constant pressure as the heated air expands. That gives you constant pressure, but with more volume to be accelerated in the convergent nozzle. When I stated earlier "I haven't seen a radiator like this" I meant auto radiators of course. Hmmm. Maybe I've wasted all my time here explaining. You using a radiator as I described? Can you tell me where to obtain one if thats what you have?
 
Last edited:
DanH said:
What are you building? When can we expect to see your work here?
Click to expand...

Hey Dan. Not building yet. Research phase. I've previously built an RV4 back in 1984, and a Compair 7 Turbine which I've been using for 12 years in my Aerial Mapping/Remote Sensing business. Just going digital this year from a 400 lbs camera/equipment load to a 35 lbs camera/equipment load. So I no longer need a 700 HP 6 seat airplane. That replaced my C-320 which was the standard in our business prior to Turbines. Just looking at all that is here for data. Internet didn't exist last time I built an RV. Of biggest concern now to me is cooling air exit. The cooling air must not cross the camera field of view. It'll be mounted in a hole in the belly just aft of the trailing spar if I use an RV7. So I've got to either do a lot of ducting, possibly side louvers with an air-cooled engine. Or possibly an exhaust jet exit tube with routing to one side of the belly. Water cooling could be much easier. Perhaps even a radiator behind the trailing spar. Yes, It will indeed be a Clean Screen design. Ross indicated that airframe surface cooling was not enough area, but I question this. I would think the leading edge of a wing has more area that a little radiator. The cooling area would be quite large including conduction thru the aluminum. I know from my High-Powered LED flashlight hobby Aluminum conducts heat very rapidly. I know it would be a lot of work adding a thin water jacket in the leading edge, but I've never been shy about that. I would guess I have 5000 plus hours building time including close to 400 hours putting a 20 x 27 inch camera hole in a C320. Difficult yes, but I enjoy the challenge. Nobody can touch me on cost with the Experimental I'm running. An RV7 could be a game changer in cost over 40-45 gallons an hour with my Turbine.
 
Last edited:
Not enough room. The set-up is quite tall. Camera sits in a Gyro mount, then an IMU is mounted above the camera. Makes the whole thing about 14-15 inches from top to bottom. No need for dual cameras.

Sorry….IMU= Inertial Measurement Unit.

It could be that even moving the air exit over a few inches could do the trick. I intend for the hole to be as close to the right side wall as possible. With the propellor pushing the belly air left slightly, it might be in the clear. The RV7 fuselage gets wider after the firewall, so that helps. I would love a rocket or RV8 but Cooling air is too close to the fuselage edges. Plus the floor structure is difficult. Can be done I'm sure. But moving one rib in the Cargo floor of the RV7 about 2-3 inches left is a piece of cake compared to the changes on an RV8/Rocket.
 
Last edited:
kengps said:
Hey Dan. Not building yet. Research phase. I've previously built an RV4 back in 1984, and a Compair 7 Turbine which I've been using for 12 years in my Aerial Mapping/Remote Sensing business. Just going digital this year from a 400 lbs camera/equipment load to a 35 lbs camera/equipment load. So I no longer need a 700 HP 6 seat airplane. That replaced my C-320 which was the standard in our business prior to Turbines. Just looking at all that is here for data. Internet didn't exist last time I built an RV. Of biggest concern now to me is cooling air exit. The cooling air must not cross the camera field of view. It'll be mounted in a hole in the belly just aft of the trailing spar if I use an RV7. So I've got to either do a lot of ducting, possibly side louvers with an air-cooled engine. Or possibly an exhaust jet exit tube with routing to one side of the belly. Water cooling could be much easier. Perhaps even a radiator behind the trailing spar. Yes, It will indeed be a Clean Screen design. Ross indicated that airframe surface cooling was not enough area, but I question this. I would think the leading edge of a wing has more area that a little radiator. The cooling area would be quite large including conduction thru the aluminum. I know from my High-Powered LED flashlight hobby Aluminum conducts heat very rapidly. I know it would be a lot of work adding a thin water jacket in the leading edge, but I've never been shy about that. I would guess I have 5000 plus hours building time including close to 400 hours putting a 20 x 27 inch camera hole in a C320. Difficult yes, but I enjoy the challenge. Nobody can touch me on cost with the Experimental I'm running. An RV7 could be a game changer in cost over 40-45 gallons an hour with my Turbine.
Click to expand...

Ok.. I will bite....

What forms do you need to fill out to be able to use an experimental for commercial Business work?
 
It's not Commercial. Part 91 all the way. No Passengers, no cargo. My former employer, and other people in the industry I'm aware of always operated under Part 91 using certified aircraft as well. The relevant FAR uses the terms "Incindental to the flight". I don't get paid to fly. I do photography. An airplane is incidental to do my work. Much the same as someone using there experimental for business travel.

BTW- I've had to go to bat on 3 different occasions with the FAA on this issue. It's legal.

But I'm getting way off subject. Don't want another half dozen post deleted as before. PM me for more.
 
Last edited:
kengps said:
It's not Commercial. Part 91 all the way. No Passengers, no cargo. My former employer, and other people in the industry I'm aware of always operated under Part 91 using certified aircraft as well. The relevant FAR uses the terms "Incindental to the flight". I don't get paid to fly. I do photography. An airplane is incidental to do my work. Much the same as someone using there experimental for business travel.

BTW- I've had to go to bat on 3 different occasions with the FAA on this issue. It's legal.
Click to expand...

Thanks... good to know.... Now I can start flying pipeline patrol in my experimental.. all I need to do is photograph while flying, and all is well..
 
n801bh said:
Thanks... good to know.... Now I can start flying pipeline patrol in my experimental.. all I need to do is photograph while flying, and all is well..
Click to expand...

Hard to say. They can be fickle. Especially when one FAA inspector starts to argue with another FAA inspector. Ego's fly, and I end up suffering.
 
kengps said:
No debate you are improving pressure recovery, thus less drag. I'm trying to explain how thrust is produced, and how to benefit from it. I do think the potential is extremely low because of the relatively low energy levels in the system. But in order to produce thrust (which will in net reduce drag) you must have air volume expanding with no pressure increase as energy is applied to it. The purpose of a jet engine is to convert a small volume of highly compressed air, into a larger volume of highly compressed air. This is done by adding energy to expand the air?. as the volume increases upon entering the combustion chamber. But a typical radiator does not allow expansion as energy is applied. It will merely create more pressure in all directions, but no volume gain. When that air (at increased pressure) leaves the radiator to the larger area at the rear face of the radiator, it simply expands again. Energy lost. You will get a net zero return.
Here is the key?.A jet engine does not add pressure in the combustion process. It adds "volume" when the energy is applied. Pressure is maintained with heat as the volume is increased. The larger volume is then used in the exit nozzle to create velocity/thrust.
SO??I'm done splainin. My suggestion is this: You want thrust to reduce the drag component from the radiator? Get a radiator that allows increased volume, and not just pressure as the air is heated. Pressure pushes in all directions. It will slow the air going into the radiator, even as it speeds up the air exiting. Net Zero. I am sure the Meredith effect does not occur with a constant chamber volume radiator. There is a difference between a Jet engine, and a space heater.
Honestly I think the gain will be so small that its a huge waste of money. Custom built heat exchangers are very expensive. Keep working on your cooling drag reduction. Sounds like you're getting some good results. Meredith effect is a poor application for money/effort. Good talking to ya.

If I can find it I'll post a photo of a radiator like this. I have seen one. It was much like an airfoil with the cooling tubes getting thinner from front to back. This maintains constant pressure as the heated air expands. That gives you constant pressure, but with more volume to be accelerated in the convergent nozzle. When I stated earlier "I haven't seen a radiator like this" I meant auto radiators of course. Hmmm. Maybe I've wasted all my time here explaining. You using a radiator as I described? Can you tell me where to obtain one if thats what you have?
Click to expand...

I don't think anyone thinks you can generate significant thrust from cooling a piston engine with air alone just because the amount we can heat the air is quite low as you say. That is kind of obvious. Your other reasoning is generally very hard to follow. As heat is added to air it will expand, if it's contained at a constant volume, pressure will rise. If we reduce the containment volume, pressure will rise yet more, which is exactly what we want to offset the pressure drop we suffered across the rad.

I think you are just not understanding what the duct shape does here. It has similar cross sectional area vs length as a ramjet. The difference is we only add relatively small amount of heat to the air with a rad rather than burning fuel.

Clearly, the air is significantly heated by passage through the rad- the temp probes prove it. Heated air MUST expand as we've already agreed. I can close the exit door and instantly observe both the exit pressure and velocity increase at any speed- even on ground runup at full throttle. Clearly we are recovering momentum over a fixed exit air modeled for adequate cooling in the climb.

We WANT the air to slow down as it reaches the rad face as this increases delta P, lowers drag and allows more contact time with the fins and tubes to raise the discharge temperature and sink the maximum energy from the coolant to the air with the minimum mass flow.

Airfoil shaped tubes might offer slightly less drag but also less heat transfer due to thermal gradient as air passes though the core. The louvered fins actually create more drag than the tubes. I did extensive flow bench testing and thermal testing with a hot water test rig to choose the most efficient rad design I could, weighing pressure drop vs. temperature drop. I settled on a design with 2.25 inch thickness, 2 rows of tubes .080 X 1.00 with .4375 spacing and 14 louvered fins per inch.

Your gut feelings are not supported by the empirical data. My test article has demonstrated 97% momentum recovery at the exit. This represents a significant reduction in drag over a fixed exit, sized for best climb cooling.
 
Yes, I think we are debating two different issues. I don't dispute anything about your cooling recovery. To make a simple analogy: Your discussing efficiency of a space heater. I'm discussing efficiency of a Jet Engine. And how that might apply to overcoming the drag in your space heater. Let's leave it at that. I won't mention the Meredith Effect anymore. Thats what I thought this thread was about after reading your first post. My mistake.
 
Last edited:
kengps said:
Ross indicated that airframe surface cooling was not enough area, but I question this. I would think the leading edge of a wing has more area that a little radiator. The cooling area would be quite large including conduction thru the aluminum.
Click to expand...

Actually no, A radiator is designed to be the most efficient structure to dissipate heat and has huge surface area in a very small volume. Mine has over 30 ft2 area if you spread it all out. A rad divides the water into multiple very thin sheets with excellent thermal conductivity between the fins and tubes. This is what is required to sink large quantities of heat off the coolant- well known facts in radiator design and thermodynamics.

A wing skin radiator would never be anywhere close to efficient as a conventional radiator in thermal conduction per unit area because of the design and difficulty in getting the water efficiently divided into thin sheets, and achieving high thermal conduction at the union of the two. Piping from the engine to the wings, the complexity of construction of the heat exchanger to the curves wing skin, extra weight of coolant etc. would all add up to tremendous cost and maintenance issues for a gain of perhaps 3% drag reduction over what could be done with a submerged rad using the Meredith Effect.

I don't think all the designers of WW2 liquid cooled aircraft and Reno racers were fools. Do you see any GA aircraft with surface conduction cooling? Maybe the engineers know more than you do. Just a thought.
 
rv6ejguy said:
Actually no, A radiator is designed to be the most efficient structure to dissipate heat and has huge surface area in a very small volume. Mine has over 30 ft2 area if you spread it all out. A rad divides the water into multiple very thin sheets with excellent thermal conductivity between the fins and tubes. This is what is required to sink large quantities of heat off the coolant- well known facts in radiator design and thermodynamics.

A wing skin radiator would never be anywhere close to efficient as a conventional radiator in thermal conduction per unit area because of the design and difficulty in getting the water efficiently divided into thin sheets, and achieving high thermal conduction at the union of the two. Piping from the engine to the wings, the complexity of construction of the heat exchanger to the curves wing skin, extra weight of coolant etc. would all add up to tremendous cost and maintenance issues for a gain of perhaps 3% drag reduction over what could be done with a submerged rad using the Meredith Effect.

I don't think all the designers of WW2 liquid cooled aircraft and Reno racers were fools. Do you see any GA aircraft with surface conduction cooling? Maybe the engineers know more than you do. Just a thought.
Click to expand...

I never said it would be easy. But there is at least 40 sq/ft of surface in an RV D-section. Engineers have learned a thing or two since WW2. It's why you go see an engineer after you produce a theory. They can be quite stuck in their ways, and resistant to change and innovation. I've known quite a few in many different fields of expertise. Of course some are different?...
 
rv6ejguy said:
Do you see any GA aircraft with surface conduction cooling?
Click to expand...

WW2 Engineers all died a long time ago as did the water-cooled airplanes. With the exception of some well cared for Million dollar plus antiques. The engineers moved on to newer things, like turbine engines. Air-cooled engines are still dominant in aircraft piston applications. Do you see any new water-cooled aircraft engines? No. In fact I've seen only one and it went out of business long ago. (and even it was really based on a racing engine big-block) Nobody wanted to to chop the b***s off a King Air and turn it into a liquid cooled Queen Air. Maybe Delta-Hawk will make it. Everything else is just a borrowed Auto engine. Why are there no Honest-to-god water-cooled aircraft engines? Why did the Voyager series never get adapted into the Twin Cessna fleet? If your answer is because there is not a big enough market to justify development cost, you are correct. Same with surface area cooling. Maybe?? It is not an original idea on my part. I have seen over the years serious references to it.
rv6ejguy said:
Maybe the engineers know more than you do. Just a thought.
Click to expand...

Personally that seems like a snotty thing to say. Perhaps you're not used to having your ideas challenged. I'll leave that up to your own resolution. Adios.
 
Last edited:
kengps said:
WW2 Engineers all died a long time ago as did the water-cooled airplanes. With the exception of some well cared for Million dollar plus antiques. The engineers moved on to newer things, like turbine engines. Air-cooled engines are still dominant in aircraft piston applications. Do you see any new water-cooled aircraft engines? No. In fact I've seen only one and it went out of business long ago. (and even it was really based on a racing engine big-block) Nobody wanted to to chop the balls off a King Air and turn it into a liquid cooled Queen Air. Maybe Delta-Hawk will make it. Everything else is just a borrowed Auto engine. Why are there no Honest-to-god water-cooled aircraft engines? Why did the Voyager series never get adapted into the Twin Cessna market? Maybe the engineers know more than you do. Just a thought.
Click to expand...

The Continental liquid cooled engine adoption had nothing to do with engineers. It was the airframe modification changes and cost of development with the low number of aircraft to be sold to recoup the design and development costs. If one engine model went into one aircraft model (and brand) then it might have been economically feasible.

We did thermodynamic analyses on the cooling circuit that Ross is doing here and could not find thrust, but it very nearly neutralized the cooling drag. We concluded that if the total aircraft "system" might end up with an "effective" thrust due to reduction of total airframe losses, but did not even do the amount of work that Ross has here in terms of actual testing.
 
kengps said:
I never said it would be easy. But there is at least 40 sq/ft of surface in an RV D-section. Engineers have learned a thing or two since WW2. It's why you go see an engineer after you produce a theory. They can be quite stuck in their ways, and resistant to change and innovation. I've known quite a few in many different fields of expertise. Of course some are different?...
Click to expand...

Clearly if you ran the calcs, you'd see that 40 ft2 of D section won't cool a 200hp engine. Given the design and construction difficulties, nobody will ever do it anyway these days. It's just a dream that many people have had before.

Yeah, I've heard what you say here so many times- engineers can't think outside the box- usually by people who disregard the math and practicality of the situation. So I'll wait for you to build it and show me how well your idea works.
 
kengps said:
WW2 Engineers all died a long time ago as did the water-cooled airplanes. With the exception of some well cared for Million dollar plus antiques. The engineers moved on to newer things, like turbine engines. Air-cooled engines are still dominant in aircraft piston applications. Do you see any new water-cooled aircraft engines? No. In fact I've seen only one and it went out of business long ago. (and even it was really based on a racing engine big-block) Nobody wanted to to chop the b***s off a King Air and turn it into a liquid cooled Queen Air. Maybe Delta-Hawk will make it. Everything else is just a borrowed Auto engine. Why are there no Honest-to-god water-cooled aircraft engines? Why did the Voyager series never get adapted into the Twin Cessna fleet? If your answer is because there is not a big enough market to justify development cost, you are correct. Same with surface area cooling. Maybe?? It is not an original idea on my part. I have seen over the years serious references to it.


Personally that seems like a snotty thing to say. Perhaps you're not used to having your ideas challenged. I'll leave that up to your own resolution. Adios.
Click to expand...

The Rotax 912 is air and liquid cooled, over 40,000 built and it's the best selling piston aircraft engine in the world at the moment. There are other liquid cooled engines being produced: http://www.d-motor.eu/nl/home-1.htm http://www.adeptairmotive.com/site/default.asp plus the Deltahawk. The latter two undergoing certification right now.

I don't mind being challenged if someone has some reasonable numbers or some better ideas but if it's just based on feelings and conjecture- well that's not very meaningful for discussion. On top of this, most other people in the know on this subject and aerospace engineers would disagree with you too.
 
I wouldn't pretend to know all that Dan & Ross know, but...

Names escape me at the moment, but the wing skin as radiator idea has actually been tried in the past, & it doesn't work. Even ignoring the complexity, etc, it doesn't work because of...wait for it...laminar flow.

The air next to the wing isn't moving relative to the skin. It actually insulates the wing skin from the moving air, killing heat exchanger efficiency.

Heat exchangers (radiators, oil coolers) are designed to cause turbulent flow of both fluids within the heat exchanger, which strips away the laminar layers & improves heat transfer. (And is one of the reasons air needs to be slowed for its trip through the exchanger; with turbulent flow, high speed = even higher drag.)

Hopefully Dan or Ross will correct anything I might have skewed in this post.

Charlie
 
rv7charlie said:
Names escape me at the moment, but the wing skin as radiator idea has actually been tried in the past.
Charlie
Click to expand...

Curtiss with the R3C-2 Schneider Cup seaplane racer. Supermarine with their S-4, S-5 and S-6 Schneider Cup racers. The Italians with their Macchi Schneider Cup racers. They all had cooling issues because they couldn't get enough surface area for adequate cooling. All of those airplanes are about the same size or slightly larger than our RV's.
 
Perhaps "Adequate" is the word. I would imagine those old racers had pretty poor heat transfer for the HP. We've come a long ways since the 30's. It may require an innovation in materials. Specifically ceramic engine technology that greatly reduces the need for external cooling. At some point surface cooling could become "adequate".

rv6ejguy said:
Yeah, I've heard what you say here so many times- engineers can't think outside the box- usually by people who disregard the math and practicality of the situation.
Click to expand...

Didn't say all engineers. Most. I've had an innovative idea or two that needed an engineers knowledge to make it work correctly. "Most" told me it couldn't be done. Then I found one who did it. Engineers may know a lot. but "Most" are stuck in their box.
 
kengps said:
Perhaps "Adequate" is the word. I would imagine those old racers had pretty poor heat transfer for the HP. We've come a long ways since the 30's. It may require an innovation in materials. Specifically ceramic engine technology that greatly reduces the need for external cooling. At some point surface cooling could become "adequate".



Didn't say all engineers. Most. I've had an innovative idea or two that needed an engineers knowledge to make it work correctly. "Most" told me it couldn't be done. Then I found one who did it. Engineers may know a lot. but "Most" are stuck in their box.
Click to expand...



Now that I can agree with you on..... If everything was so cut and dry and predictable, then experimental /homebuilt planes would not be on the cutting edge of the future of GA... IMHO.
 
So now we are in to bashing "most" engineers.

I have a lot more respect for people that rise to the top of what they do based on their own accomplishments than those who try to show how good they are by putting other people down. It's a sign of maturity. Most folks don't last very long around here without respecting others.
 
rv7charlie said:
I wouldn't pretend to know all that Dan & Ross know, but...

Names escape me at the moment, but the wing skin as radiator idea has actually been tried in the past, & it doesn't work. Even ignoring the complexity, etc, it doesn't work because of...wait for it...laminar flow.

The air next to the wing isn't moving relative to the skin. It actually insulates the wing skin from the moving air, killing heat exchanger efficiency.

Heat exchangers (radiators, oil coolers) are designed to cause turbulent flow of both fluids within the heat exchanger, which strips away the laminar layers & improves heat transfer. (And is one of the reasons air needs to be slowed for its trip through the exchanger; with turbulent flow, high speed = even higher drag.)

Hopefully Dan or Ross will correct anything I might have skewed in this post.

Charlie
Click to expand...

Surface conduction cooling was used on quite a number of record breaking aircraft in the '30s but almost the entire wing/ fuselage and floats were covered in heat exchangers under the skin. For a race aircraft, this could make sense where manufacturing complexity on basically one off builds and high maintenance is of little concern. For everyday use aircraft it makes little sense. It's really heavy too, limiting useful load. Engineers have already run the pros and cons lots of times before and they choose rads based on the physics and math.

You are absolutely right about turbulent flow being needed for best heat exchange between liquid and air which is why the rad is so good and skin so poor. Again, this has been known for many decades in thermodynamics. 80 years of time passage has not changed the fundamentals of thermodynamics.
 
Last edited:
Michael White said:
Curtiss with the R3C-2 Schneider Cup seaplane racer. Supermarine with their S-4, S-5 and S-6 Schneider Cup racers. The Italians with their Macchi Schneider Cup racers. They all had cooling issues because they couldn't get enough surface area for adequate cooling. All of those airplanes are about the same size or slightly larger than our RV's.
Click to expand...

These airplanes had about 10-14 times the power output of an RV though...

Surface conduction works but is not very practical for production aircraft for the other reasons I mentioned above.
 
kengps said:
Perhaps "Adequate" is the word. I would imagine those old racers had pretty poor heat transfer for the HP. We've come a long ways since the 30's. It may require an innovation in materials. Specifically ceramic engine technology that greatly reduces the need for external cooling. At some point surface cooling could become "adequate".



Didn't say all engineers. Most. I've had an innovative idea or two that needed an engineers knowledge to make it work correctly. "Most" told me it couldn't be done. Then I found one who did it. Engineers may know a lot. but "Most" are stuck in their box.
Click to expand...

Ceramic engines are a long time in the future for the masses. In the meantime, with our conventional aluminum engines, most people are going to be using radiators to cool them.

I've seen people come along with innovative concepts and ideas who are not engineers and made them work. You don't have to be an engineer to be smart and creative and trial and error may eventually get you to a good solution however, the science of engineering is the foundation for our technological success today and I wouldn't underestimate the importance of that. I figure I can learn something from anyone who has either practical or theoretical experience in any field. Of course, who hasn't met myopic, closed minded engineers? They populate this discipline just like any other discipline. That's just the human race.
 
Last edited:
n801bh said:
Now that I can agree with you on..... If everything was so cut and dry and predictable, then experimental /homebuilt planes would not be on the cutting edge of the future of GA... IMHO.
Click to expand...

Many things in the mechanical world are predictable to an engineer who really looks at the problem and runs the numbers. This is why engineering is an important science but I have seen some closed minded and arrogant engineers who discounted ideas because they are outside the norm without even running the numbers. That's not really engineering then.

In your case and mine, we've both had these types tell us something wouldn't work but in fact it has worked fine in practice. The engineers in question do indeed look dumb when faced with the realities of something performing well in real world use for hundreds of hours. So, we can be proud of our accomplishments and know that not all engineers are right all the time, any more than we are right all the time.

Your Ford/ 801 and other auto conversions with many hundreds of flight hours proves that you don't have to be an engineer to make automotive conversions successful. Experience in a field is also very important. An engineer with all theoretical experience and no practical experience can be equally inept at designing something practical...
 
Ironflight said:
So now we are in to bashing "most" engineers.

I have a lot more respect for people that rise to the top of what they do based on their own accomplishments than those who try to show how good they are by putting other people down. It's a sign of maturity. Most folks don't last very long around here without respecting others.
Click to expand...

Paul and others... Let me qualify my feeling...

First off, (most) is an overreach and I should have corrected that when I quoted the other poster. As you know the "alternative engine" forum can and will elicit some strong feelings on both sides of the issue. .. So, if anyone is a engineer reading this and can't take my opinion on what I am going to type, then may I suggest they hit the delete button now... <G>

As a commercial building contractor for over 30 years, I can say this with confidence. ( Some ) engineers design structures that are so far over the top in strength and rigidity that common sense and safety are thrown out the window in favor of creating a grossly expensive project. ( Some ) of these engineers have never even nailed two boards together, but are experts in fasteners.

What has compounded this problem immensely is the fact our local building dept will not approve ANY plans unless it has an engineering stamp on it. Wanna build a 4X6' dog house for Fido,, ya need a stamped set of plans.. Don't get me wrong.. some complex structures need a darn good engineer to get it right, but.. There are schedules and tables for load bearing capabilities for dimensional lumber that it time tested and easy to calculate so average people can design and build a dog house, or a simple structure. My beef is when the government dictates a requirement that is excessive and burdonsome..

It is a slippery slope and we could be next. So, what if a builder of a RV-10 wants to install a V-8 automotive engine in it and has to fabricate a mount, brackets, cowling,etc,etc,etc..... I suggest that if builders who are thinking and building "outside of the box" are required to hire and pay a licensed engineer would give up and stuff a Lyc/Conti in there and not explore the true spirit of EXPERIMENTAL aviation....

Peace to all the engineers out there...

Rant off...

Ben Haas.
 
You must log in or register to reply here.
Back
Top