Radiator designs

[wuzak]

Just want to look at the various radiator designs that were used during WW2 and comapre the philosophies of the different countries.

The UK, for example, generally developed radiator designs to suit each particular application - such as the Spitfire's underwing radiators, the Hurricane's belly radiator, the leading edge radiators of the Whirlwind and the Mosquito and various others with chin mounted radiators.

As far as I am aware, the only quick engine change module (QEC) developed in the UK was for the Merlin XX as used on the Beafighter and Lancaster.

The US also left it to the airframe manufacturers. Not sure that any QECs developed in the US went into production (the XB-39's V-3420 was built into a QEC). US designs seemed to be chin or belly mounted.

Neither the US or UK seemed to have pushed the development of the annular radiator, though Napiers did late in the war/post war.

Germany seemed to have mostly underwing or annular radiators. There were exceptions - such as the Me 309 and the Do 335 (rear engine).

Was just reading about the proposed Fw 187C-0, which was to have used the DB 605 engine. Part of the conversion involved fitting radiators under the wings between the nacelles and fuselage. This required removal of the inner wing fuel tank, no doubt costing range.

I wondered why they wouldn't have attempted to install leading edge radiators. These would allow the inner fuel tank to remain, and would not have required too much structural change.

 

[tomo pauk]

Once DB-601/605 is installed, the aft mounted radiator help canceling out the extra weight installed at the front. With radiators installed in the front, too, they would need to make more adjustment to the CoG, and it might involve maybe too much of ballast.

 

[wuzak]

Once DB-601/605 is installed, the aft mounted radiator help canceling out the extra weight installed at the front. With radiators installed in the front, too, they would need to make more adjustment to the CoG, and it might involve maybe too much of ballast.

They lengthened (were going to) the fuselage by 1m for thr Db 605 vs the DB 601.

 

[tomo pauk]

I assume it will be longer behind the wing?

 

[wuzak]

I assume it will be longer behind the wing?

Yes, lengthened in the rear.

 

[swampyankee]

Annular radiators seem to combine the worst features of liquid-cooled engines -- a radiator and its associated plumbing -- and an air-cooled radial, the shape of the nose (I think it's contradictory to say that annular radiators are the best thing in the world and radials are the worst), which is probably why they were commonly used (iirc, there were US aircraft with annular radiators, in the 1920s)

 

[wuzak]

I think the advantage with annular radiators is that the aircraft could be easily changed between radial and LC engine without too much modification. Particularly if the airframe was originally designed around a radial and the streamlining advantages of the in-line could not be achieved.

I think the example of the XB-38 shows my point. The engine nacelles couldn't be as slim as they otherwise could be due to the fact that they had to fit to the same bulkhead as the Wright R-1820. Vega also chose to add leading edge radiators in between the nacelles, which would have required modifications to the production line if it proceeded. If they had done a nacelle with the V-1710 and an annular radiator it could have been just a straight engine module swap.

 

[OldSkeptic]

Trouble with annular radiators and chine ones, was that you were getting a lot of turbulent air onto the radiator, which reduced efficiency.

To get a good Meridith effect, you needed a long entry, shaped to slow the air down, where it passed over the radiator, then is compressed again at the outlet, a simple jet engine in effect.
Any turbulence affects the airflow and reduced the efficiency.

The best positions were in the leading edge, or under the fuselage, where you could get the necessary depth required.
No one ever got (though Hawker claimed they did with their leading edge version of the Tempest, though that is debated) 100% or more of an effect. Rather the thrust offset the drag by X%. The Mustang was the best (of a production version) and was probably in the 90% region. It was in clear air, though the first prototypes were flush with the fuselage, but they had some issues with turbulence, hence the little chin to get it into clear air

The early Meredith versions, as per the Spit and 109 were not as nearly as good, though they helped of course. But under the wing was not a great location, since you got turbulence and/or accelerated air off of the leading edge. It was difficult to get enough depth. And airflow could vary, not just with speed but by angle of attack. Wing twist and bending could alter the inlet angle and so on. Not optimum
Plus they affect the wing airflow.

If you look at the Spiteful's radiators, which they spend a lot of time on, they ran the whole length of the wing basically. And in the end were probably no better than the Mustang's one.

So an annular radiator scores badly. Very turbulent and fast air coming in and little depth.

 

[Aozora]

Annular radiators seem to combine the worst features of liquid-cooled engines -- a radiator and its associated plumbing -- and an air-cooled radial, the shape of the nose (I think it's contradictory to say that annular radiators are the best thing in the world and radials are the worst), which is probably why they were commonly used (iirc, there were US aircraft with annular radiators, in the 1920s)

Depends on the installation; for example, Napier (manufacturer of the Sabre) designed an annular installation and modified two Tempest Vs to test it.

Original installation:

later modified to use a ducted spinner and cooling fan:

From Flight, July 1946:

total drag:

According to Napier via Flight the annular radiator freed airframe designers from the need to sacrifice space in the wings or fuselage, allowing both to be cleaner, while a properly designed annular radiator installation inflicted less drag, provided more efficient engine cooling and was less vulnerable to battle damage or accidents:

View attachment Napier Engine Testing 1946.pdf

 

[tomo pauk]

Aozora covered it.

A Spitfire with leading-edge radiators should've been faster than standard Spit?

 

[wuzak]

The Spitfire's standard underwing radiators could have been improved significantly.

I recall reading that Joe Smith thought that the exit and entry areas were too large, and that the 2 position flap was restrictive.

 

[wuzak]

If you look at the Spiteful's radiators, which they spend a lot of time on, they ran the whole length of the wing basically. And in the end were probably no better than the Mustang's one.

Hardly. They are completely within the undercarriage's track.

Supermarine Spiteful Mk.14, RB518 by San Diego Air & Space Museum Archives, on Flickr

 

[OldSkeptic]

Yes. They made major improvements with the Spit III. If you see photos of it the radiator is further forward and bigger, despite that it had less drag because is was more efficient.
Now the Spit III was about 10-20mph faster than a 'standard' Spit on the same engine power. Some of that from detail tidying up, but the better rad had to contribute. My guess is about 50% of that improvement.
Ditto my estimate of the Mustang, at least, maybe a bit more of its higher speed came from the lower drag of the rad.

The Mustang's was the next generation Meredith radiator, with had an infinitely (sort of) varying exit flap. The Spit's (and 109) were more basic and had poorer positions as well so were not as effective, though they did help of course.

Remember what you want to do. You want smooth air coming in, you want to slow it down, so heat transfer is better (you want to pass the max volume of air as possible over the hot areas as slowly as possible), then you want to compress the air as it exists, so it then it speeds up and gives thrust. Basic equations P1V1/T1 = P2V2/T2. Which means the hot air, compressed now, will have a higher speed and thus more thrust. Do it right and that thrust offsets the drag coming from the cooler air entering into the inlet.

The extreme form of this is in (good) supersonic engines (not like the disasters like the B1). The 2 best were the SR-71 and the Concorde, where the majority of the thrust came from that Meridith like effect (not quite the same, but similar principles). In fact the Concorde, at mach 2, used less fuel than a comparable modern aircraft at mach 0.85 (common cruise speed these days). The downside was that it would use more than a comparable sub-sonic aircraft taxiing and take off than the other would use on an entire short trip. The regulations that crippled Concorde, coming from the US no less cut that 'efficient time' down.

 

[OldSkeptic]

Hardly. They are completely within the undercarriage's track.

Supermarine Spiteful Mk.14, RB518 by San Diego Air Space Museum Archives, on Flickr

Yes 'full length' was a bit if a hyperbole, but they were were much more than a std Spit. Particularly depth, which is needed. But here is the mistake they made, they moved them back too far in the wing. Now Supermarine's own analysis (forgot the date late 43 early mid 44?) showed that a Mustang like radiator would increase the Spits speed by about 20mph. MAP rejected it because of the production issues (and Messers R&R kept coming to the party with more power). Their own, with the development of the Mk III, experience showed that you needed an under wing radiator right out front to get non-accelerated/non-turbulent air.

The end affect was to have a 'new' plane, very little faster than a 'standard' Spit with the same power .. and a poorer mach limit .. and handled poorer.

So why, they stayed with the under wing radiators I don't know.

 

[tomo pauk]

The Spitfire III have had fully covered undercarriage. That will add 10 mph alone?

 

[davparlr]

The Germans, Focke Wulf 190D and Ta 152, seem to use the annular type radiators to good advantage. The Fw-190D-9 with 2100 hp was capable of 385 mph at SL. The universally accepted aerodynamically efficient P-51B, with 1860 hp was capable of 386 mph at SL. While the Fw-190D-9 was not as efficient as the P-51, it didn't seem to suffer too much from its radiator design. The annular radiator design effectively buries the radiator within the area constraints of the fuselage. In addition, the exposed frontal area of an annular radiator is much less than that of a radial engine. It is notable that the FW190D-9 was similarly efficient to the Spitfire XIV, which made 389 mph with 2100 horsepower. Actually, I am sure all of these numbers are within the error rate of manufacturing and maintenance. The Tempest V, with 2235 hp was capable of 393 mph at SL. These, of course, represent the aerodynamic efficiency of the total airframe and not just the radiator, but it is interesting to note the performance comparison of aircraft with annular (Fw190D-9), nose mounted (Tempest V), wing mounted (Spitfire XIV) and rear fuselage mounted (P-51B) radiators.

 

[Shortround6]

Probably the best results it would be possible to find would be for the Tempest since it seems to have used at least 3 different radiators on the same airframe using essentially the same engine, reducing other factors to a minimum.

 

[l'Omnivore Sobriquet]

Were the leading edge radiators not getting just as much 'troubled air' than the annular ones closer to the spinners' center ?

(Further from the propeller axis, would they have gotten a 'cleaner air' ? or not ? Thanks the more scare presence of the prop's paddle through space (in a given time frame) in the outer region, as the (similar) relative aera section covered in a given time was done so more through radial speed... and little matter presence than near the axis, where one found lower actual speed but more 'obstructive' or 'effective' matter presence of the paddle...)

(the relative aeras travelled by the padde were the same but by the axis the overall scale of the scheme was smaller...)

Yet that same paddle in the outer ring was much more air blasting than at the near-axis region.

Some questions being easier to ask than to answer.

Rationnaly, could anyone explain why a leading edge radiator situated behind the outter ring of the propeller, blasting at its full might, might be getting a 'cleaner air' than the annular radiotor situated by the axis region of the same propeller, producing much weaker 'air trouble' but, hm, 'taking more place' ??????

 

[wuzak]

Probably the best results it would be possible to find would be for the Tempest since it seems to have used at least 3 different radiators on the same airframe using essentially the same engine, reducing other factors to a minimum.

Right.

435mph with Sabre II and chin radiator.
466mph with Sabre IV and leading edge radiators.
~465mph with Sabre VII and annular radiator.

 

[tomo pauk]

What was the power used for those speeds?