Quantcast

Heat pipes and VW cooling

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,586
Location
US
I can see from some recent threads that the cooling of VW engines has been a 'hot' topic.

For the devotees of VWs: Are you aware of folks using heat pipes to assist with cooling?

As many folks may be aware, a "heat pipe" can transfer heat very efficiently from one place to another. They typically are comprised of a hollow tube with a working fluid inside (water, ammonia, freon, etc--chosen to work best at the temps encountered). The liquid boils on the hot end of the pipe, the vapor travels very quickly (typical velocities=100s of M/sec) to the "cool" end, where the vapors condense. The liquid flows back to the hot end via capillary action through a wick on the inside surface of the pipe and the cycle repeats. The heat transfer is very rapid, very efficient (they can move hundreds or thousands of times more heat than a similar sized copper rod), and entirely passive (the pipes are sealed--no pumps or external energy needed. Very reliable). They don't weigh much: just a little bit of fluid is needed and there is typically very little pressure difference between inside and outside, so the walls can be thin.

Many consumer computers now have tiny heat pipes to take the heat away from the CPUs.

Anyway, this IS experimental aviation. If VW engines have increased in output to where it's becoming hard to remove all the heat from the heads (there's only so much real estate for fins, and their efficiency decreases with fin length) I could see a heat pipe with a large finned evaporation end brazed to the fins of the head and carrying heat off to some purpose-built heat exchanger inside the cowl, or even to a finned exterior plate on the outside of the aircraft (under the nose bowl, in the prop blast?). This might be especially useful for turbo-normalized applications: Cooling an 80HP engine at 2000' MSL is a lot simpler than cooling it at 12K' (the cooler air up there doesn't come close to making up for the lower air density as far as cooling effectiveness goes).

Well, there are no new ideas under the sun, so I'm sure this would be being done already if needed/practical. Anyone know?
 
Last edited:

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
5,606
Interesting. First time I've heard of it. Might work on the VW. Even better would be some aftermarket heads that have some serious fins but I think someone's already doing that.

Dan
 

autoreply

Well-Known Member
Joined
Jul 7, 2009
Messages
10,762
Location
Rotterdam, Netherlands
Well, there are no new ideas under the sun, so I'm sure this would be being done already if needed/practical. Anyone know?
Yeah. I doubt it's more efficient than water cooling. In a PC, water cooling has drawbacks (cooling medium degradation of cooler blocks, moving parts), which offset the drawbacks of heat pipes.

Also, heat pipes are only efficient at extremely high power outputs per square inch, but not that well at lower power densities, because the "layer" of cooling medium has to stay "thick" enough.

I do see a future for other things though. Cooling of LEDs (now a major nightmare), EFIS/instruments and cabin heat/cooling (a simple resistance, bolted to a hefty cable and a heat pipe is enough) are ideal candidates. Electric motors too.

Heat pipe - Wikipedia, the free encyclopedia
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,586
Location
US
Yeah. I doubt it's more efficient than water cooling. In a PC, water cooling has drawbacks (cooling medium degradation of cooler blocks, moving parts), which offset the drawbacks of heat pipes.

Also, heat pipes are only efficient at extremely high power outputs per square inch, but not that well at lower power densities, because the "layer" of cooling medium has to stay "thick" enough.

I do see a future for other things though. Cooling of LEDs (now a major nightmare), EFIS/instruments and cabin heat/cooling (a simple resistance, bolted to a hefty cable and a heat pipe is enough) are ideal candidates. Electric motors too.

Heat pipe - Wikipedia, the free encyclopedia
I'm not quite following your point about heat pipes being most efficient at extremely high power outputs per sq inch. In the applications you cited the heat flux is on the order of less than ten watts/sq inch, and I know our little motors put out way more heat than that. On paper a heat pipe would have advantages over liquid water cooling: especially lighter weight and reliability (no pumps to fail, no clamps, rubber hoses, etc).

But, I agree that in practice this might be a "challenging" application for heat pipes. For example, these lengths of sealed copper pipe in a high vibration environment . . . I see some possible trouble there.
 
Last edited:

skeeter_ca

Well-Known Member
Joined
Jun 29, 2005
Messages
1,027
Location
Yucaipa, Ca
Is this the same technology they use for propane operated refrigerators for RV's. If so those don't seem to work that great as it is.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,586
Location
US
Is this the same technology they use for propane operated refrigerators for RV's. If so those don't seem to work that great as it is.
No, I think you are referring to adsorptive refrigeration. That's a different process.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,451
Location
Rocky Mountains
Is this the same technology they use for propane operated refrigerators for RV's. If so those don't seem to work that great as it is.

Other than using phase change to move heat they aren't the same thing. There is plenty of info on the net on how they work. When they work they work very well. Like a lot of things if they aren't "tuned" for the particular application they aren't much better than a stick of solid copper.

I've considered using some home built ones for a through the firewall heat exchanger. Using them would mean not having hot oil, water, or CO in the cockpit.
 

NickH

Well-Known Member
Joined
Dec 26, 2010
Messages
101
Location
Portland, OR
They don't weigh much: just a little bit of fluid is needed and there is typically very little pressure difference between inside and outside, so the walls can be thin.
At sea level, maybe. I see this as a major issue at 12K.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,586
Location
US
At sea level, maybe. I see this as a major issue at 12K.
?? Heat pipes can operate at a wide variety of internal pressures, depending on the working fluids chosen. Often they operate at internal pressures >below< STP. The relatively small change in external pressure when going from MSL to 12K' (a pressure change of about 5 PSI) shouldn't make any difference to a copper tube of any "normal" thickness (i.e. mechanically strong enough to put up with the bumps and bruises of life under the cowl). So, I don't think this issue should rule out their use.
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,271
Location
Orange County, California
I'm trying to understand the point of the entire exercise. What makes you think the VW is inadequately cooled in the first place? If you're trying to run any of the VW engines continuously in the 'takeoff power' range, it's highly unlikely that any kind of exotic heat-pipe arrangement is going to be either less expensive or more reliable than simply purchasing a larger VW that puts out the required power within its rated continuous-power band.

Or even buying a Lycoming or Continental, for that matter.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,586
Location
US
I'm trying to understand the point of the entire exercise. What makes you think the VW is inadequately cooled in the first place? If you're trying to run any of the VW engines continuously in the 'takeoff power' range, it's highly unlikely that any kind of exotic heat-pipe arrangement is going to be either less expensive or more reliable than simply purchasing a larger VW that puts out the required power within its rated continuous-power band.

Or even buying a Lycoming or Continental, for that matter.
I didn't say the engine was "inadequately cooled."

Just trying to understand--what would be the limiting factor on continuous operation of a turbo-normalized 2276cc Type 1 at 12,000 MSL? It's not the ability to flow air or fuel.

Anyway, as it was framed in the OP, this might be a solution in search of a problem. If just adding more fins is the answer, then clearly there's no need for anything more exotic. Which might be why no one appears to be doing it. Fair enough--it was just a question.

None of this is a knock on VW power, the VW corporation itself, the city of Wolfsburg, or air-cooled engines in general, and I apologize to anyone who may have believed it was.
 
Last edited:

autoreply

Well-Known Member
Joined
Jul 7, 2009
Messages
10,762
Location
Rotterdam, Netherlands
I'm not quite following your point about heat pipes being most efficient at extremely high power outputs per sq inch. In the applications you cited the heat flux is on the order of less than ten watts/sq inch, and I know our little motors put out way more heat than that.
On a typical chip, power densities are around 1 Watt/mm^2, so that's 625 Watt/sqin. A typical volkswagen (no fins) engine has much more than 200 sqin of area that needs to be cooled. You're right about efficiency, I shouldn't have used that term. What I meant was power density, or weight per unit of cooling power. Heatpipes are great for short distances, very high power on a small area, but don't work that well on a bigger scale.
On paper a heat pipe would have advantages over liquid water cooling: especially lighter weight and reliability (no pumps to fail, no clamps, rubber hoses, etc).
Well, liquid-cooled engines basically work the same. Even without the pump, there will still be a bit of circulation, but you just need much more flow to keep the engine cool enough. If the cooling medium's temperature is high enough, it will also boil and condensate. But I presume the required length of tubing (condensing in-between) will not make that much more efficient. I'd love to be wrong though :)
 

Dana

Super Moderator
Staff member
Joined
Apr 3, 2007
Messages
9,530
Location
CT, USA
Remember that heat pipes simply move heat from one place to another. If you attach a heat pipe to a cylinder head, you still have to put cooling fins on the other end of the pipe to transfer the heat to the air. In the end, it's still an air cooled engine. If there's room to put adequeate fins directly on the head that would be a simpler approach. Where heat pipes (or any liquid cooling system, for that matter) have value is if there's not enough room for air cooling at the source (like a CPU cooler) and you do have room farther away. Either way, the head has to be designed for it; I don't think just brazing it on would be a very efficient solution... and might even cause undesirable uneven cooling.

I also wonder if the wick could move enough fluid to handle the heat generated by an engine... it might require a large heavy heat pipe.

Some PPG engines have pumpless thermosiphon cooling, with a radiator mounted above the engine. Natural convection causes the circulation, and it seems to work quite well.

-Dana

Lottery: a tax on the mathematically challenged.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,586
Location
US
In the end, it's still an air cooled engine. If there's room to put adequate fins directly on the head that would be a simpler approach. Where heat pipes (or any liquid cooling system, for that matter) have value is if there's not enough room for air cooling at the source (like a CPU cooler) and you do have room farther away. Either way, the head has to be designed for it; I don't think just brazing it on would be a very efficient solution... and might even cause undesirable uneven cooling.
Agreed. Of course, even the term "liquid cooled engine" is a bit of a misnomer, since ultimately all our engines are air cooled (with a small nod to charge-cooled engines that derive some of their cooling from evaporation of the incoming fuel). Liquid cooling of engines is just a way to move the heat to a place where it can more effectively be moved to the air. I think that heat pipes sized to move heat from the head to a more convenient place (the aircraft's skin, a purpose-built heat exchanger, etc) wouldn't need to be very large (and they do scale up well--industrial-sized ones are several inches in diameter). Overall, the biggest challenges would be, 1) as you mentioned, designing or adapting the head to facilitate the efficient removal of heat and 2) dealing with vibration. In every application I know of, the heat pipes are rigid, and that's a big problem if we're going to attach them to an engine.

And, along these same lines, a nit: When are we going to stop using the term "radiator"?:) "Heat exchanger" or "convector"?

autoreply wrote:
On a typical chip, power densities are around 1 Watt/mm^2, so that's 625 Watt/sqin. A typical volkswagen (no fins) engine has much more than 200 sqin of area that needs to be cooled. You're right about efficiency, I shouldn't have used that term. What I meant was power density, or weight per unit of cooling power.
Thanks. I understand now.
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,271
Location
Orange County, California
The problems with VW cooling have mostly been application and useage related. Pilots don't understand that the engine cannot be run WOT all the time. Like most auto engines, it simply wasn't designed for that operation. Yes, head cooling is generally the factor that limits this on the VW series.

However, trying to 'fix' the limitation begs the question of the application. If you really need 65hp continuous, then trying to make a smaller-displacement (1600, 1835cc Type 1) VW produce that through adding fins or exotic cooling schemes is just a bad idea when there are larger-displacement VW models perfectly happy to churn out 65hp all day long without modification. If you need much over about 90-100hp continuous, then a VW is simply a bad choice of engine, regardless of the Type or displacement.

People try to 'go cheap' with VWs for some reason, but they're still an engine and they do only what they're designed to do. If you need 180hp, you don't try and figure out a way to get it out of an O-200; you go get an O-360. Same principles apply to VWs.

As far as the continuous power limitations of a particular engine under particular conditions, the people who could best answer that question would be the particular manufacturer involved. I assume that, in this case, that would be Great Plains, Sonex (Aerovee), or Revmaster. They'll know the limitations of their product better than anyone. I'd try them.
 
Last edited:

DaveK

Well-Known Member
Joined
Apr 21, 2007
Messages
365
Location
Northern California
Thought about this awhile back but stopped looking into it for a couple reasons, some already mentioned. The big one was you still need to have fins (aka a mass / radiator) to transfer the heat to the air, which you have to put somewhere.
Another was the heat pipe works great up to some critical heat load at which case it overloads and the thermal conductance drops to that of the pipe material only. So if you didn't get it right and some part overloaded you suddenly have much less heat being pumped out of the head which wouldn't correct itself until you let the whole system cool down again.
To make it worse the heat pipes would tend to block air to the head so you wouldn't be able to rely on the normal heat rejection off the fins. But saying all that I could see it as a way to cool limited areas of a head where there is little airflow or insufficient fin area. But bigger fins (as already pointed out) would likely be easier to do.
I think to get a robust system together that could handle the vibration and other operational issues would turn out to be more work then water cooling. Though it was fun to think through the problem.
 

Latest posts

Top