Quantcast

Liquid cooling

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

trimtab

Well-Known Member
Joined
Apr 30, 2014
Messages
103
Location
rocky mountains, rocky, usa
The bp of water is around 178F at 18k'. The bp of a 50% aqueous glycol solution is about 198F at 18k'. The bp is of concern because the pump will cavitate at the suction side, and flow drops precipitously. Pressure is used to keep the suction side pressure well above the cavitation potential, even at sea level. Beyond a small amount of cavitation, the pump will often unprime. I dealt with this fact all the time in the design of vacuum distillation systems. It's a thing. And even in the 30's, this figured prominently in cooling system designs when the larger engines became common.

You can try and run a system at ambient pressure. You just have to keep the temp low enough so the vapor pressure doesn't get anywhere close to the inlet pressure anywhere in the pump.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,993
Location
Saline Michigan
The P-38 having the intercooler (radiator) INSIDE the wing, can be analogous to putting the radiator of an auto engined ( liquid cooled ) light plane tucked inside the fuselage. Imagine a RV-x with a V-6 or 8 in a tightly wrapped and low drag cowling, with the radiator, no ducting, behind the seats in the tail cone. Eventually the radiant & convective heat will transfer through the fuselage to hit equilibrium, but the total heat you can reject would, I bet, be far less than a few hundred horsepower V-x puts out. Power would be temperature limited.
I don't see what you are getting at here. There are a number of flat four and flat six Subaru powered engines with aft mounted heat exchangers for water and oil, with modest air into the cowling for induction, intercooler, and accessory cooling, and they work fine. Www.sdsefi.com website shows details.

Billski
 

Heliano

Well-Known Member
Joined
Dec 24, 2015
Messages
92
Location
Campinas, SP, Brazil
These postings are most interesting and those participating are clearly knowledgeable on the subject. And it affects me directly because I am in the process of designing the engine installation in my aircraft which is about 70% complete by now. - the engine is a converted Honda Fit engine. RV7charlie is right when he says that what works in a Pietenpol may not work in a Lancair. But here goes a quick reminder about something known since the start of WWII: variable-area cooling inlets and outlets may be the best way to go for high speed aircraft. Too complex? one can simplify it somehow by using spring loaded flaps that open or close with dynamic pressure.
 

rv7charlie

Well-Known Member
Joined
Nov 17, 2014
Messages
845
Location
Jackson
Yeah, us alt engine guys are just beginning to re-learn what the WW2 (and a few years later) guys had down pretty well. Efficient variable inlets are pretty much unachieveium for mere mortals, but variable outlets are relatively simple to do.

One thing rediscovered is that oversized inlets, in proper configuration, don't hurt you from a drag standpoint. According to some of the old NACA/NASA/university/European papers, if the inlet has a relatively fat 'lip' (shape of which some of the documents define), then you can open the outlet (cowl flap) to get good low speed cooling, and when you close down the outlet, the excess air trying to enter the inlet will smoothly flow around and past, as if the inlet was smaller. Lip shape is critical, though.

Hardest thing about variable outlets is the force required to *close* them at speed. A friend with a rotary powered RV6 had to use an RV flap actuator motor to operate his. Cleverest idea I've seen was by a friend who runs a Lyc. Instead of hinging the flap door at the front, he made it in a bent teardrop shape, almost like a ',' and hinged it in the middle. When it opens, it opens like a throttle butterfly (balanced air load on each side of the flap). The teardrop shape smooths the airflow hitting it, and the fat end of the teardrop up inside the cowl acts as a deflector to move more air out of the cowl.

Charlie
 

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
5,409
Cleverest idea I've seen was by a friend who runs a Lyc. Instead of hinging the flap door at the front, he made it in a bent teardrop shape, almost like a ',' and hinged it in the middle. When it opens, it opens like a throttle butterfly (balanced air load on each side of the flap). The teardrop shape smooths the airflow hitting it, and the fat end of the teardrop up inside the cowl acts as a deflector to move more air out of the cowl.
Smart. That makes the typical GA airplane cowl flap look like something off a woodstove from 1900. Way back in about 1985 I built my own swamp cooler (evaporative cooler) for my house. We lived in a hot, dry area, so those things work well there. The duct coming through the wall into the house shot a lot of air across the living room, so I did like your friend did: made a wing and pivoted it across the center of the duct. It would deflect the air up or down (usually up, since cold air sinks and cools the room by scavenging the hot air higher up) and it didn't slow the air or make any extra noise. It deflected the air underneath it upward just as well as it deflected the air above it. Worked way better than I expected. The pivot was just aft of the aerodynamic balance point so little friction was needed to hold its position.
 

Heliano

Well-Known Member
Joined
Dec 24, 2015
Messages
92
Location
Campinas, SP, Brazil
Exactly as you say, Charlie: flaps can be designed to close or to open with air load simply by changing the flap shape or relocating the pivot axis. As for fat "lips", yes, fat lips prevent flow detachment and the resulting loss of efficiency. The turbofan installations are an example: at high angle of attack, flow detachment is a big problem with those engines (may imply compressor stall, vibration, etc.), and to avoid it inlet lips have a round, blunt shape.
 

Aesquire

Well-Known Member
Joined
Jul 28, 2014
Messages
2,428
Location
Rochester, NY, USA
with aft mounted heat exchangers for water and oil, with modest air into the cowling for induction,
But not installed without ducting in the tail cone, where radiant and convective flow heats the sheet metal cone to dissipate heat. A hypothetical stupid way, that is similar to getting heat through a painted & alodined leading edge sheet metal.

The point is you need airflow over or through a large area to cool your coolant. Fairly direct thermal coupling.

as to coolants... There is a waterless coolant used by car collectors Evans Waterless Coolant, Prevent Engine Overheating it is supposed to prevent corrosion problems in seldom run cars that fools in southern states tend to run with plain tap water. Which has just enough minerals to aid galvanic corrosion of dissimilar metals.

And I say fools, since the addition of Glycol coolant in 40-70% ratio not only serves as an antifreeze, sorta important if flying at altitude or winter, but also greatly increases heat transfer. The invention of Glycol coolant allowed much smalker, lower drag radiators... clear back in the 1920s iirc. Initially they used straight glycol, but that alone was a corrosion nightmare, and leaked through tiny pinholes water didn't. A mixture of water & glycol gave the best compromise, plus lower freeze temperature. The addition of anti corrosion oils both reduces corrosion and lubricates the water pump.

Running a pressurized coolant system allows higher temperatures, higher delta, and thus more efficient cooling.

There was a notable increase in power between the very early Junkers Jumo 211 engines and later models after they went from atmospheric to pressurized systems.
 

Alessandre

Member
Joined
May 28, 2020
Messages
20
I would like to discuss about if this configuration if is possible to apply the meredith effect in this case, maybe using a variable exaust opening, I think for many constructors is much more simple to keep all the engine systens firewall foward than build a special belly for carry the radiator. I think using a functional configuration will be equal or better drag results than an air cooled engine, I'm new in this forum yet, sorry if I'm redundant.

radiator.jpg
 

trimtab

Well-Known Member
Joined
Apr 30, 2014
Messages
103
Location
rocky mountains, rocky, usa
...the addition of Glycol coolant in 40-70% ratio not only serves as an antifreeze, sorta important if flying at altitude or winter, but also greatly increases heat transfer.
The use of an aqueous glycol solution reduces the effective heat transfer coefficient in all flow regimes. The reduced heat transfer performance of using glycols (ethylene and propylene) are partially recovered through the prevention of films that rapidly reduce performance in water only solutions, and also from the increase in volumetric flow for a given pressure drop due to Nusselt number differences.

After it is all said and done, you need more volumetric flow for a given size of radiator for a solution, and that increase in flow is partially offset by slightly lower pumping losses.


 

Royal

Well-Known Member
Joined
Jun 8, 2020
Messages
71
I would like to discuss about if this configuration if is possible to apply the meredith effect in this case, maybe using a variable exaust opening, I think for many constructors is much more simple to keep all the engine systens firewall foward than build a special belly for carry the radiator. I think using a functional configuration will be equal or better drag results than an air cooled engine, I'm new in this forum yet, sorry if I'm redundant.

View attachment 99814
This will help you decide.
 

pictsidhe

Well-Known Member
Joined
Jul 15, 2014
Messages
8,066
Location
North Carolina
I would like to discuss about if this configuration if is possible to apply the meredith effect in this case, maybe using a variable exaust opening, I think for many constructors is much more simple to keep all the engine systens firewall foward than build a special belly for carry the radiator. I think using a functional configuration will be equal or better drag results than an air cooled engine, I'm new in this forum yet, sorry if I'm redundant.

View attachment 99814
A Meredith cooler is the same as a low drag radiator installation. Incoming air is first slowed down in a diffuser, passed through the radiator, then accelerated through an exit nozzle. Flow is almost always varied with a variable exit nozzle. Variable intakes do not work well to control flow. A well designed fixed intake has a very small drag penalty. The trickiest and most important part is getting the diffuser right. They like to be big...
 

Royal

Well-Known Member
Joined
Jun 8, 2020
Messages
71
Most of the planes that did a great job of low drag coolers had the inlets placed on about the thickest part of the wing. Probably getting the fastest air they can and then slowing it down to get the best benefit. They also said keeping the air from making any turns is critical for lower drag. The other option is under the cowling like the oiler cooler in the bf-109. I think putting it under the wing gives you room to expand the air. But we aren't cooling massive v-12s. This article does a great job of explaining it. He also said it worked too well.
 

Attachments

rv7charlie

Well-Known Member
Joined
Nov 17, 2014
Messages
845
Location
Jackson
Alessandre,

People who are much smarter than me (and have actual engineering degrees) tell me that the Meredith effect is almost like Bigfoot. Lots of people claim to have seen it, but no one has ever brought back one for peer review.

I'm not an engineer, but I think that you can be confident that your in-cowl drawing will never achieve a true Meredith effect. There's just too much chaotic flow in the cowl, both before and after passing through the heat exchanger.

While I don't doubt that Ross (in the Kitplanes article) got exit air moving faster than inlet air, it does not necessarily follow that he achieved *net thrust*. If you read the article carefully, he does qualify his findings by saying that he didn't measure the *drag* of his attached cooling system. He made a *lot* of other changes in/around his cowl while converting to the belly scoop coolers, and his previous cooling drag profile had been quite high, so almost anything would have improved his cooling drag, increasing a/c speed. That's not the same thing as true Meredith effect, where the cooling path literally functions as a ram jet, with waste heat being the 'fuel'.

My friend who is an engineer and plays with aero stuff all the time tells me that while the P51 *might* have achieved it, it requires quite high input pressures (achieved through the 300-400 MPH speeds) to have any chance of making it truly work.

Just my thoughts; I could be wrong... :)
Charlie
 

jedi

Well-Known Member
Joined
Aug 8, 2009
Messages
2,156
Location
Sahuarita Arizona, Renton Washington, USA
Alessandre,

......

“While I don't doubt that Ross (in the Kitplanes article) got exit air moving faster than inlet air, it does not necessarily follow that he achieved *net thrust*.”

.......:

Charlie
Exhaust velocity needs to be referenced to free airstream velocity to determine thrust. Reference to inlet velocity does not help unless inlet pressure and static pressure are also known and inlet total pressure is compared to exhaust total pressure.
 

rv7charlie

Well-Known Member
Joined
Nov 17, 2014
Messages
845
Location
Jackson
Even if the exit is higher velocity than free stream, it still doesn't mean *the system* is producing net thrust. A venturi tube
Search Results | Aircraft Spruce
used to create vacuum for gyros on old a/c produces flow faster than free stream, but it won't make any thrust. It adds drag.

It's likely that any system with a large inlet and a clean interior transition to a small outlet will have an exit velocity higher than free stream. Doesn't mean that it's making thrust.
 

Jay Kempf

Curmudgeon in Training (CIT)
Lifetime Supporter
Joined
Apr 13, 2009
Messages
4,032
Location
Warren, VT USA
The Meredith effect was never a net thrust. It was a recovery of cooling drag to a high efficiency factor. So say your radiator was 2 sq feet, you reduce the effect of hanging it in the breeze by like 90%. There is no way to get that sort of efficiency through a cooling duct with a giant right angle turn and short passages. The paths need to stay laminar and diverge and converge properly. Can be done. Not inside a cowl most likely.

The other myth of the P51 is that the cooling circuit had low drag all the time. It was only designed for the interceptor mission mode. Go up high, pull back throttle and go a long way due to low drag or high L/D at high altitude carrying droppable fuel tanks and munitions for the rendezvous. And to be able to return to protect the bombers. Very much a single point design. And that is how you have to use something like heat pressure recovery. Single point on the performance envelope.
 

rv7charlie

Well-Known Member
Joined
Nov 17, 2014
Messages
845
Location
Jackson
Well,
Meredith effect - Wikipedia.

No doubt the straight through system a la P51 is the most efficient, but according to all the stuff I've read (including that Wiki link), the 'Meredith effect' means that near-mythological net thrust, due to the added heat via the heat exchanger(s).

Charlie
 

Jay Kempf

Curmudgeon in Training (CIT)
Lifetime Supporter
Joined
Apr 13, 2009
Messages
4,032
Location
Warren, VT USA
Nope, that's mythological reading of secondary sources. The original paper was a "drag reduction" effort. Yes the pressure recovery in the aft section of the converging duct was accelerating the flow to near ambient but not a net thrust. And physics does not allow a net thrust if you do the energy balance. Too much pressure drop through the radiator to allow it.

Wikipiedia is shite.
 

pictsidhe

Well-Known Member
Joined
Jul 15, 2014
Messages
8,066
Location
North Carolina
The Meredith effect is one of those things that is theoretically possible, but hard to achieve in practice. It's general layout is the way to minimise cooling drag, even if you don't make any net thrust.
The P51 was designed as an interceptor. Early ones sucked at altitude thanks to the Allison. Schmued tried his best to reduce drag and was quite successful despite many efforts only being partly as good as envisaged.
Once the P51 got the Merlin, it became the great escort it is famous for.

Many WWII aircraft were designed with thicker than optimal radiators for the simple reason that they presented a smaller target...
 
Top