Iskandar Taib said:
From what I understand, you never get as much thrust out of the Meredith Effect as it takes to overcome the drag of the radiator inlet. You can recover quite a bit, but it is never zero sum or better. In any case, there is only so much thrust that can be generated out of the heat produced by the engine. Make the inlet larger than is optimal, and you get more drag. The inlet size really does matter - the bigger it is, the bigger the drag. You make it just as big as it needs to be to cool the engine, no bigger. If the size didn't matter, then you could make the inlet as big as you wanted. This also applies to radials, by the way.
See:
http://www.supercoolprops.com/ARTICLES/gwhite.htm
http://www.supercoolprops.com/ARTICLES/ducted_cooling.htm
I don't think the Allison was making more power than the two stage Merlin, even on the deck. If they did, people would be dumping the two stage supercharger for racing! Instead, they finesse the inlet, and spray liquid on the radiator core. Incidentally, Supercool is wrong about the Spit IX. It did use the Meredith effect (see Quill's book), though the radiator design wasn't as efficient as on the Mustang, and there were other sources of drag. The Mustang manages to hide most of the bulk of the radiator core inside the fuselage - it takes up a large amount of space below and behind the pilot, while the scoop and inlet represent a comparatively small "bump" in the fuselage profile.
I have to say I'm dubious of the articles above (especially the first one). It claims the meredith effect was tweaked in the Cal Tech wind tunnels. The effect cannot really be evaluated in a static test of that nature. The engine must be running and the cooling system operational for the effect to occur. Certainly the shapes were studied in the wind tunnel, but if you research the P-51 development you will see these guys did most of their work utilizing mathematics rather than physical tests. Physical tests were conducted to confirm or refute the mathematical conclusions, but the process was not one of trial and error as implied.
Also, he seems to take Lee Attwoods story about the P-51 at face value. If you study the P-51, you will see that there was a scism of sorts between Atwood, who was an excutive on the P-51 team, and Edgar Schmued and Ed Horkey, who were the actual designers. Attwood tried every trick in the book to claim the P-51 design features were his ideas, but the truth seems to indicate otherwise. For instance, he did not make claim of having brought the info about the "Meredith Effect" to the team from the British researcher F.W. Merdith done in 1935 until after Edgar Schmueds death in (I think) 1986. But this is refuted by the fact that the heat pump design of the P-51 cooling system had been on the table at NAA for years within Schumed's design team, long before Atwood returned from his sales trip to England.
It is kind of telling that Atwood waited until after Schmued's death to start making his claims as "father of the P-51". Ed Horkey has disputed Atwoods claims from the moment he made them. I tend to believe Schmued was the force behind the P-51's design, not Atwood. Atwood's baby was the B-25 - a plane sorely lacking in any kind of innovation. Schmued's babies, the P-51 and the F-86 were loaded with innovations, and Schmued was the one whose team was designing hypothetical fighters back when they were tasked to design the AT-6 trainer when Atwood got the B-25 project. Atwood had no such reputation of zeal to design a fighter - he was more of... a businessman/salesman.
You can get a glimps into the whole debate by reading the following articles:
http://www.airspacemag.com/asm/mag/supp/jj99/Mustang.html
http://www.airspacemag.com/ASM/Mag/Index/1996/AS/wmtm.html
-------------------------------
I certainly agree there is an optimal size for the scoop. My understanding is that the radiator thrust system overcame 90-100% or more of the cooling system drag, depending on altitude (pressure), ambient temperature, and how hot the particular plane was setup to run. Sometimes they used a Prestone formulation that allowed temps up to 350 degrees F, but these were highly caustic and hard to handle for ground crews, and usually lower temp solutions were utilized.
From the Spitfire thread:
-----------------------------------
First off lets look at the Bf109 scoop/cooling design:
As you can see the radiators are indeed quite small. The "boundary layer diverter" mechanism was a fix for a problem discovered on the E models. The radiator is mounted to the bottom of the scoops and the boundary layer is allowed to flow through a space between the top side of the radiator and the wing. This helps to avoid injestion of the turbulent boundary layer which makes the radiators more efficient than they would be if there were larger but no space was provided for the boundary layer. Because the boundary layer has to make a significant turn upward to follow the diverter, this is only partially effective and at high speeds as the boundary layer gets thicker and has more mass and as pressure builds in the scoop, the boundary layer still lifts off the bottom of the wing and around the scoop entirely resulting in the "gulping" effect. The boundary layer diverter design helps to get a little more efficiency out of the small scoops of the 109 but it hardly "solves" the issue. There is no significant thrust generated for a number of reasons which I'll cover further down.
Now lets look at the P-51 radiator-thrust design:
First off, as is quite apparent, the P-51 radiator is HUGE compared to the two scoop radiators of the Bf109. Furthmore, the radiator has three to four times frontal area of both Bf109 radiators combine, which makes it inherantly more efficient for transfering heat.
Next, lets consider the boundary layer diversion method. On the P-51 the scoop is spaced more than an inch and a half away from the bottom surface of the wing (this varied a bit through different models). This means the boundary layer misses the scoop inlet entirely and encounters no obstruction that could rip it away from the scoop inlet until it is well past the inlet. The problem is completely solved.
Now lets look at how the thrust system works. First high speed cold air enters the scoop and proceeds down a widening passage which acts as an expansion chamber. The expansion chamber futher cools the air, slows its velocity, and increases the pressure (I know this is counter-intuative but its true). Then the (relatively) slow moving air passes through the radiator grilling, which is designed in the form of little ">" shapped fins stretched over the tubeing to form a sort of one-way valve. The heated air then passes into a narrowing passage which acts as a compression chamber.
When air passes through the radiator it is heated unevenly. Air molecules which make contact with the radiator fin elements are super-heated. Those that pass mid-way between the elements are much less heated. These molecules exchange heat in the compression chamber. One hot molecule and one cold molecule take up less volume than two warm molecules (assuming the total heat energy level is the same). So the air in the compression chamber is being compressed by its momentum into the narrowing passage and at the same time it is expanding as the heat in the molecules is transfered from the hottest molecules to the cooler molecules.
Finally, the hot air is vented through the thrust nozzel, which is designed and regulated for pressure. This provides thrust. At medium-high to high speeds, the jet of air comming out the thrust nozzel is supersonic, which provides usable thrust beyond speeds where a prop is no longer able to provide much thrust. Not only that, but the stream of hot expanding air is directed right into the wake of the fuselage. This wake is where parasitic drag normally "sucks" the plane back, and is the biggest part of an airplane's drag. Just like a tracer bullet, the P-51's exhaust fills the vacuum wake and reduces drag.
The Bf109 cooling system lacks both an expansion chamber and a compression chamber. The cold boundary layer air is re-introduced into the radiator exhaust in the space behind the radiator, virtually eliminating the chances of producing much thrust from expanding air. The cooling flaps at the back of the scoops are not designed to sustain high pressures behind the radiator, nor to control the outflow to generate a supersonic thrust stream, they are there simply to regulate the radiators to prevent excess cooling, primarily in dives. And finally, the radiator exhaust does not flow into the fusealge wake to help cancel out the parasitic drag.
The 109 has no meaningful "Meredith effect" thrust!
Note: The "Meredith effect" explanation is incomplete, trying to attribute all of the advantage of the radiator design to the thrust generated. A good portion of the advantage was the projection of super-heated air into the vacuum wake of the fuselage, nullifying parasitic drag.
----------------------------
The Spitfire lacked the boundary layer diverter portion of the Bf109, so it may have generated more thrust than the Bf109, at the cost of injesting the boundary layer. To overcome the cooling inefficiency caused by boundary layer injestion, the British simply made the scoops and radiators larger and larger. But without the expansion chamber, compression chamber, and pressure nozzel, not much thrust would be developed. And because the radiators don't feed into the vacuum wake behind the fuselage, the reduction in parasitic drag is minimal.
Finally, I'd also point out that the Mossie ustilized the Meredith effect, and it did so rather effeiciently - but only at two speeds (one being maximum)where the cooling flow was just right.
Also the Japanese Zero utilized the effect, and the this was stolen and transfered to the later model Corsairs. But the effect on these planes was much less (1/3rd?) than on the P-51. I'm not really sure how it was done.
=S=
Lunatic