Jump to content
Sauerkraut-LW

P51 Radiator Design

Recommended Posts

I understand that the P51's radiator was designed to take advantage of the Meredith effect, but there is one thing about the design that has always puzzled me.

 

Why is it designed like a "scoop"? Notice that the top of the radiator inlet is NOT the bottom of the fuselage, like other aircraft with that radiator placement. There is a gap between the top of the radiator and the bottom of the aircraft's body.

 

P-51_small2.jpg

 

P51_Av_4407_DA_scoop_p132_W.png

 

This is not very noticeable in earlier versions of the aircraft but it becomes much more pronounced in the definitive D version of the plane. (If you still can't see or picture what I'm talking about, it is VERY noticeable on the F-82 twin mustang).

 

Every intuition would say that this would essentially create an air trap that would cause extra drag. I have no idea how that could possibly contribute to the design in any way.

 

Does anyone have insight into this? I've never actually seen anyone discuss this before and I've always been intrigued by it.

 

EDIT: Replaced the "design" in the second line with "placement".

Edited by itsthatguy

Share this post


Link to post
Share on other sites

I understand that the P51's radiator was designed to take advantage of the Meredith effect, but there is one thing about the design that has always puzzled me.

 

Why is it designed like a "scoop"? Notice that the top of the radiator inlet is NOT the bottom of the fuselage, like other aircraft with that radiator design. There is a gap between the top of the radiator and the bottom of the aircraft's body.

 

P-51_small2.jpg

 

P51_Av_4407_DA_scoop_p132_W.png

 

This is not very noticeable in earlier versions of the aircraft but it becomes much more pronounced in the definitive D version of the plane. (If you still can't see or picture what I'm talking about, it is VERY noticeable on the F-82 twin mustang).

 

Every intuition would say that this would essentially create an air trap that would cause extra drag. I have no idea how that could possibly contribute to the design in any way.

 

Does anyone have insight into this? I've never actually seen anyone discuss this before and I've always been intrigued by it.

I think it was meant to avoid the turbulent air attached to the fuselage. Like in the modern jets engines inlets.

  • Upvote 1

Share this post


Link to post
Share on other sites

IMHO it could also be like this to minimize exchange time after damage during emergency belly landing.

Share this post


Link to post
Share on other sites

Marsh is the winner.

 

That design is actually less draggy than the one say, on the Mig 3.

Share this post


Link to post
Share on other sites

As zun already mentioned, it takes the inlet out of the boundary layer of the air flowing around the fuselage. Basically the lower velocity and turbulence of the boundary layer is not optimal for cooling while increasing drag. This way they could keep the radiator smaller

Share this post


Link to post
Share on other sites

Here is some good info (at least applicable for earlier P51s)

https://www.youtube.com/watch?v=8_5kSxWO5rg

 

There were 2 excellent articles on the web about Mustang's radiator design, but website is not available anymore?! :(

 

As I remember the P51 has:

- a turbulent-less inlet (it isn't directly lowering drag, but avoids turbulence inside)

- a diffuser

- a radiator itself (early version had a circular segmented core)

- a variable nozzle

 

Technically, the P51' radiator is a ramjet powered by engine waste heat. It had enough thrust to (almost) nullify cooling drag. To work well, the plane had to be moving faster than 450km/h, then it was effectively recovering 300-400HP.

 

Here you will find some nice pictures including parts of the circular radiator:

http://americanaeroservices.com/a-36-apache/

Edited by Ehret
  • Thanks 1

Share this post


Link to post
Share on other sites

 

 

then it was effectively recovering 300-400HP

 

That much? Seems extrodinary on the face of it..............from a non-engineer's point of view anyway. 

Share this post


Link to post
Share on other sites

If you look at the performance of the razor back B and C models, it may be easier to believe that the radiator design played it's part in the speeds that the 51 could attain.

 

Same power output as Merlins in the much lighter Spitfire, yet the B and C models could hit 445+ mph at altitude.  The draggier D models lost about 20mph from the razor back types because of drag from the bubble canopy.

Share this post


Link to post
Share on other sites

The draggier D models lost about 20mph from the razor back types because of drag from the bubble canopy.

No it didn't

http://www.wwiiaircraftperformance.org/mustang/p-51-tactical-chart.jpg

http://www.wwiiaircraftperformance.org/mustang/P-51D_15342_Level.jpg

http://www.wwiiaircraftperformance.org/mustang/eglin-p51b-level.jpg

Edited by RoflSeal

Share this post


Link to post
Share on other sites

If you look at the performance of the razor back B and C models, it may be easier to believe that the radiator design played it's part in the speeds that the 51 could attain.

 

Same power output as Merlins in the much lighter Spitfire, yet the B and C models could hit 445+ mph at altitude.  The draggier D models lost about 20mph from the razor back types because of drag from the bubble canopy.

 

Not only speeds but unusually fast cruise and fuel efficiency. I hope we will get other version as a collector plane, because, frankly, the D is worst of the bunch. The handling and damage resilience were compromised, too, and it's clear P51Ds were needed for range mostly.

Share this post


Link to post
Share on other sites

But you want turbulent air inside a radiator as it increases mixing and hence heat transfer.

 

Laminar air flow sucks for heat transfer as it stays in the layers having an insulating effect.

 

Defo don't want the rad itself causing turbulence elsewhere though.

Edited by AeroAce

Share this post


Link to post
Share on other sites

Not only speeds but unusually fast cruise and fuel efficiency. I hope we will get other version as a collector plane, because, frankly, the D is worst of the bunch. The handling and damage resilience were compromised, too, and it's clear P51Ds were needed for range mostly.

Oversimplification

  • Upvote 1

Share this post


Link to post
Share on other sites

Oversimplification

 

And low price and light logistic, too - the Jug or the Lighting costed twice as much per unit, needed much more fuel and were harder to service.

Share this post


Link to post
Share on other sites

Not only speeds but unusually fast cruise and fuel efficiency. I hope we will get other version as a collector plane, because, frankly, the D is worst of the bunch. The handling and damage resilience were compromised, too, and it's clear P51Ds were needed for range mostly.

I think you should assume less.V1650-7 was optimised for low level. That canopy design went through a lot of wind tunnel tests. P-51D with 67' achieved during a test at SL 375 mph with wing racks on. Hardly slower than B.

  • Upvote 3

Share this post


Link to post
Share on other sites

It's hard to say exactly what made the biggest difference, but I doubt it was what'ever little thrust generated by the radiator as the outlet went almost fully open once running on max boost. I'd attribute the P-51's great straight line speed to the overall extraordinarily clean design of the aircraft, which includes the sculpting of the radiator inlet, as well as the drag bucket of the laminar flow airfoil at low AoA. 

Edited by Panthera

Share this post


Link to post
Share on other sites

It seems that much like the Yak1 s69 transitioned to s127 you see a speed increase in all the little things combined. Thats probably what we see in the B and C mustangs to the D only on a much greater level, because engine also played a big part.

 

I never thought of a boundary layer splitter either. I really want to take some engineering courses after a lot of what I read on this forum. 

Share this post


Link to post
Share on other sites

...B and C models could hit 445+ mph at altitude.  The draggier D models lost about 20mph from the razor back types because of drag from the bubble canopy.

Everyone says this, but if you actually look for the differences between the razorbacks(P-47 too) compared to their bubble canopy variants you'll find a marginal difference. You'd probably be more concerned about the loss of longitudinal stability.

 

6mph difference between the D-22 and D-25 at 30,000ft.

 

http://www.wwiiaircraftperformance.org/p-47/p-47-tactical-chart.jpg

Edited by DSR_T-888

Share this post


Link to post
Share on other sites
On 2/27/2018 at 4:25 AM, Panthera said:

It's hard to say exactly what made the biggest difference, but I doubt it was what'ever little thrust generated by the radiator as the outlet went almost fully open once running on max boost. I'd attribute the P-51's great straight line speed to the overall extraordinarily clean design of the aircraft, which includes the sculpting of the radiator inlet, as well as the drag bucket of the laminar flow airfoil at low AoA. 

For the engineer, who had a major part in designing P51 it was clear what made the biggest difference:

http://www.historicracer.com/aviation/p51-mustang-meredith-effect-lee-atwood/

"it was the attention to detail and optimising the Meredith Effect"

How much the P51' nozzle opens depends on a flight envelope: on a takeoff it may be fully open - at altitude and high speed it will be almost closed.

An another nice article:

http://skyviking.net/svwp/?p=450

Share this post


Link to post
Share on other sites
On 26.02.2018 at 8:29 PM, AeroAce said:

But you want turbulent air inside a radiator as it increases mixing and hence heat transfer.

 

Laminar air flow sucks for heat transfer as it stays in the layers having an insulating effect.

 

Defo don't want the rad itself causing turbulence elsewhere though.

A bit late to the party here, but...

... in aforementioned case, not quite, they wanted first and foremost uniform velocity flow field in the intake, as the lip was immediately followed by a big diffuser. Removing boundary layer is critical for getting the most out of diffuser's pressure recovery, whether it's needed later for cooling or charging is irrelevant. That's why today, in both subsonic airplane and racing car design, if NACA duct is not enough for required airflow, offset scoops are the most efficient alternative.

Share this post


Link to post
Share on other sites
3 hours ago, Ehret said:

For the engineer, who had a major part in designing P51 it was clear what made the biggest difference:

http://www.historicracer.com/aviation/p51-mustang-meredith-effect-lee-atwood/

"it was the attention to detail and optimising the Meredith Effect"

How much the P51' nozzle opens depends on a flight envelope: on a takeoff it may be fully open - at altitude and high speed it will be almost closed.

An another nice article:

http://skyviking.net/svwp/?p=450

 

Yeah I've read that, but I'm not so sure about it. Remember other fighters also took advantage of the effect (the 109 also featured a variable outlet design for example), however seeing as the temperatures involved aren't ever very high the thrust gained can't have been that significant. Also keep in mind that it was the P-51's speed at low alt which was particularly impressive, where'as its high alt performance had a lot to do with engine performance being quite abit above average up there.

In short I am convinced it was the overall cleanliness of the design which was the major reason behind it being so fast, in particular at low altitude, and not just because the radiator design took advantage of the meredith effect which was quite well understood at the time.  

Edited by Panthera

Share this post


Link to post
Share on other sites
46 minutes ago, Panthera said:

 

Yeah I've read that, but I'm not so sure about it. Remember other fighters also took advantage of the effect (the 109 also featured a variable outlet design for example), however seeing as the temperatures involved aren't ever very high the thrust gained can't have been that significant. Also keep in mind that it was the P-51's speed at low alt which was particularly impressive, where'as its high alt performance had a lot to do with engine performance being quite abit above average up there.

In short I am convinced it was the overall cleanliness of the design which was the major reason behind it being so fast, in particular at low altitude, and not just because the radiator design took advantage of the meredith effect which was quite well understood at the time.  

Other planes didn't have proper parts to make it work well - not a proper diffuser, nor intake in the 109. For the temperatures, you have to consider the ambient air temperature, which is under -40C at 6000m, and compression done by the diffuser.

Look at the P51 cutaway - there is no other ww2 plane, with such an elaborate ducting inside for the cooling subsystem. So, no - making a plane's rads to "happily eat" turbulence is not a sign of understanding this well enough.

 

 

 

 

Share this post


Link to post
Share on other sites
14 minutes ago, Ehret said:

Other planes didn't have proper parts to make it work well - not a proper diffuser, nor intake in the 109. For the temperatures, you have to consider the ambient air temperature, which is under -40C at 6000m, and compression done by the diffuser.

Look at the P51 cutaway - there is no other ww2 plane, with such an elaborate ducting inside for the cooling subsystem. So, no - making a plane's rads to "happily eat" turbulence is not a sign of understanding this well enough.

 

I've seen the cutaway, but again the temperature differential simply isn't that high,  esp. at low altitude where the P-51 really proved the most impressive speed wise.

Here's a cutaway of the 109 radiator design, which really isn't that much different and works on the same principle:

f_airflow.jpg

 

Share this post


Link to post
Share on other sites

Thanks for interesting schematics.

However the diffuser part is very short and that boundary layer bypass doesn't help pressure.

Share this post


Link to post
Share on other sites
3 hours ago, Ehret said:

Thanks for interesting schematics.

However the diffuser part is very short and that boundary layer bypass doesn't help pressure.

The boundary layer bypass shouldn't effect the airflow on any side of the radiator as they are seperated, also it wasn't present past the F series according to what I've read.

As for the inlet being short, that shouldn't be a big issue, the important part is the temperature differential between front and back as well as outlet being variable down to a small opening.

Edited by Panthera

Share this post


Link to post
Share on other sites
2 hours ago, Panthera said:

As for the inlet being short, that shouldn't be a big issue, the important part is the temperature differential between front and back as well as outlet being variable down to a small opening.

Not an inlet being short but the diffuser - it's a very important part as it increases pressure and slows down flow, so the radiator core can transfer heat more efficiently. The P51's radiator isn't split into two devices like in the 109 - not good for damage resilience, but good for thermal efficiency.

This is a good video showing it inside:

https://youtu.be/jzKFkYLHq3Q?t=866

 

Looking at the 109' outlet flaps they don't seem to be proper nozzles - sides are partially open, thus will leak left and right.

Edited by Ehret

Share this post


Link to post
Share on other sites
5 hours ago, Ehret said:

Not an inlet being short but the diffuser - it's a very important part as it increases pressure and slows down flow, so the radiator core can transfer heat more efficiently. The P51's radiator isn't split into two devices like in the 109 - not good for damage resilience, but good for thermal efficiency.

This is a good video showing it inside:

https://youtu.be/jzKFkYLHq3Q?t=866

 

Looking at the 109' outlet flaps they don't seem to be proper nozzles - sides are partially open, thus will leak left and right.

Well again I'm not sure how important that length really is, but I assume you have a report on it?

That said keep in mind that I'm not saying the 109's radiator necessarily does as great a job as the P-51's at producing the meredith effect, but in principle it works the same and the effect is there. Also the outlet flaps are actually shielding off the sides so that the air doesn't go left & right, just as on the P-51.

Excellent little 3 part video on the Pony btw.

 

Edited by Panthera

Share this post


Link to post
Share on other sites

According to Atwood, both the 109's and Spitfire's inlets were too short, consequently too abrupt, and stalled at high speed causing both increased drag and reduced cooling.  This is what led both companies to deduce, incorrectly, that Meredith Effect wasn't practical and the Mustang's secret to speed had to be the wing.  According to Lee Atwood (again) the Mosquito was the only other WW2 combat plane to successfully take advantage of Meredith.  Supermarine's later experience with the Spiteful showed them how tenuous and unreliable the gains with laminar wings were. 

  • Upvote 2

Share this post


Link to post
Share on other sites
3 hours ago, chuter said:

According to Atwood, both the 109's and Spitfire's inlets were too short, consequently too abrupt, and stalled at high speed causing both increased drag and reduced cooling.  This is what led both companies to deduce, incorrectly, that Meredith Effect wasn't practical and the Mustang's secret to speed had to be the wing.  According to Lee Atwood (again) the Mosquito was the only other WW2 combat plane to successfully take advantage of Meredith.  Supermarine's later experience with the Spiteful showed them how tenuous and unreliable the gains with laminar wings were. 

Hmmm that's interesting considering the Mosquito also featured quite a short inlet:

Mosquito.jpeg

 

Also post war designs didn't make a whole lot of use of long inlets, like for example the supermarine designs which seem to have largely incorperated the 109's radiator configuration.

Edited by Panthera

Share this post


Link to post
Share on other sites

The Mosquito's radiators were positioned in wings' leading edges - probably to reduce drag in the 1st place. The intakes aren't much smaller than frontal area of radiators' core, so maybe a very short diffusor was enough to match. Then what was considered a success here? - perhaps a modest improvement in just one flight regime was enough.

 

For the minimal diffusor length I read a (indirect) note about it in one article. Now I have read in a post above (thanks chuter!) it was because of stalling. Indeed - making it short you are increasing internal angles, so at some air velocity it will stall.

 

I have scanned Meredith's article and the curious thing is that what we have in the Mustang isn't yet a complete device... Engine exhausts should be directed there as well to boost efficiency further.

 

Share this post


Link to post
Share on other sites
On 3/3/2018 at 1:02 AM, Panthera said:

 

Yeah I've read that, but I'm not so sure about it. Remember other fighters also took advantage of the effect (the 109 also featured a variable outlet design for example), however seeing as the temperatures involved aren't ever very high the thrust gained can't have been that significant. Also keep in mind that it was the P-51's speed at low alt which was particularly impressive, where'as its high alt performance had a lot to do with engine performance being quite abit above average up there.

In short I am convinced it was the overall cleanliness of the design which was the major reason behind it being so fast, in particular at low altitude, and not just because the radiator design took advantage of the meredith effect which was quite well understood at the time.  

 

The theory behind the Meredith effect was understood, but not how to make it work in practice.  The reason why we should take Atwood's word for the effect of the radiator installation on the top speed is that the radiator design was changed during development and testing and the results of the tests observed. This is far more convincing than looking at schematics. 

 

I posted earlier an Article in which Atwood's presentation to an Aeronautical society was given - I will see if I can find it and add here if I can work out the new forum. It is very clear that the improvement in the speed came from specific changes to the radiator design, including getting the intake out of the boundary layer. 

 

edit: here is the thread with the article. I see that you posted in that thread too - so not sure why you think that your opinion after looking at a few diagrams trumps that of one of the design team.   

 

Edited by unreasonable

Share this post


Link to post
Share on other sites
18 minutes ago, unreasonable said:

 

The theory behind the Meredith effect was understood, but not how to make it work in practice.  The reason why we should take Atwood's word for the effect of the radiator installation on the top speed is that the radiator design was changed during development and testing and the results of the tests observed. This is far more convincing than looking at schematics. 

 

I posted earlier an Article in which Atwood's presentation to an Aeronautical society was given - I will see if I can find it and add here if I can work out the new forum. It is very clear that the improvement in the speed came from specific changes to the radiator design, including getting the intake out of the boundary layer. 

 

Redirecting the boundary layer would've helped reduce drag, which is also what I said in regards to the sculpting of the radiator being important (the Germans did the same in the 109F series), however what we're debating here is wether P-51's speed was mostly down to the meredith effect which I am not convinced of at all. Instead I am saying that the P-51's impressive top speed was mostly down to drag reduction via the careful shaping of the fuselage in addition to the drag bucket of the laminar flow airfoil in low AoA flight. You see the large effect of the drag bucket on the Hawker Tempest as well, which despite a rather hefty wing area & weight was probably the fastest piston engined aircraft at low altitude during the war.

 

As for the radiator design changing with development, this is normal and happened with many designs during the war in an effort to improve cooling and reduce drag.

Edited by Panthera

Share this post


Link to post
Share on other sites

The Tempest is a bit heavier but not that much bigger than the Mustang, and has 500-700HP more than P51D's WEP.

Edited by Ehret

Share this post


Link to post
Share on other sites
40 minutes ago, Panthera said:

 

Redirecting the boundary layer would've helped reduce drag, which is also what I said in regards to the sculpting of the radiator being important (the Germans did the same in the 109F series), however what we're debating here is wether P-51's speed was mostly down to the meredith effect which I am not convinced of at all. Instead I am saying that the P-51's impressive top speed was mostly down to drag reduction via the careful shaping of the fuselage in addition to the drag bucket of the laminar flow airfoil in low AoA flight. You see the large effect of the drag bucket on the Hawker Tempest as well, which despite a rather hefty wing area & weight was probably the fastest piston engined aircraft at low altitude during the war.

 

As for the radiator design changing with development, this is normal and happened with many designs during the war in an effort to improve cooling and reduce drag.

 

I know what you are saying. I am saying that one of the men in the design team says otherwise, on the basis of the work done in designing and modifying the actual plane.

 

I am not interested in trying to change your views: believe what you like.  Since you have access to this information but still value your own opinions over those of Atwood, it seems unlikely that anything could possibly influence your opinion.

 

My post is for the benefit of others reading the thread who have not yet read Atwood's words.

  • Upvote 2

Share this post


Link to post
Share on other sites

Lets compute equivalent HP rating of Me262 at her maximum level speed: 2*8800N at 900km/h equals 5900HP - an amount going 100% to thrust.

When a V12 in the P51D is doing 1500HP, considering a low efficiency of a piston engines, a radiator has to dissipate (roughly) around 4500HP. That's a lot of heat to work with and a comparable (nomally wasted) power to an early jet engine(s).

Share this post


Link to post
Share on other sites
46 minutes ago, Ehret said:

Lets compute equivalent HP rating of Me262 at her maximum level speed: 2*8800N at 900km/h equals 5900HP - an amount going 100% to thrust.

When a V12 in the P51D is doing 1500HP, considering a low efficiency of a piston engines, a radiator has to dissipate (roughly) around 4500HP. That's a lot of heat to work with and a comparable (nomally wasted) power to an early jet engine(s).

 

Not really sure where you want to go with that? The temperature of the radiator, as well as compression of the air, never becomes high in relative terms compared to what is seen in an actual jet engine.  Thus what'ever amount of thrust gained through the meredith effect should be small.

 

Had the meredith effect been THE major reason behind the P-51's impressive straight line speed, then IMHO we would've had more examples of post war piston engined designs following a similar radiator design path, however few appear to have done so.

 

Edited by Panthera

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...