DarrenW
Staff Sergeant
I apologize as the link worked fine on my laptop but not on my phone. Thanks for posting it as it seems to include more airplane types than what can be found in the L5A30 report. 
J2M Raiden
- Thread starter Sagittario64
- Start date
Ad: This forum contains affiliate links to products on Amazon and eBay. More information in Terms and rules
My GUESS would be that they would have had to do something else to improve cooling during ground and low speed operation.
Then again, perhaps with a different engine (not that they had many choices), they may not have needed the cooling fan.
Note that even with the cooling fan, the BMW 801 tended to run hot.
Note also that there were other radial engine fighters with similar sized engines that did not need a cooling fan.
I am not convinced that the FW 190A was all that draggy in comparison to American radial engined fighters. Do you remember the discussion about equivalent flat plate area back in your FW 190 thread? Basing the CD on wing area is a bit misleading and makes aircraft with small wings look particularly bad.
For a practical example of this, consider the Unlimited class racing planes. Many of them have reduced sized canopies modified cowls and drastically reduced wing area. I am guessing that their CD values are fairly high in comparison to their unmodified cousins but their overall (equivalent flat plate) drag is much less.
Hello Swampyankee,
Actual testing results do not seem to support your assertion. Also, as I commented to DarrenW, I do not believe that coefficient of drag is always a good comparison.
Hello Shortround6,
My guess would be to reduce airflow through the cowl possibly because the engines had a tendency to run too cool?
Just a guess.......
- Ivan.
DarrenW
Staff Sergeant
Yes I remember that discussion. Basically we were talking about two different concepts. I was looking at zero-lift drag coefficient which gives an indication of an aircraft's aerodynamic refinement, and you were talking mostly about drag area. I think the CD0 formula tries to separate the drag component produced by lift alone in order to see how much parasitic drag the airframe produces. The formula will produce low drag figures for aircraft with small wings if the speed attained is high enough. Apparently the FW-190A just wasn't fast enough to allow this to happen, given all the variables.
But of course until we can get these airplanes together in the same wind tunnel, under similar test conditions in order to take some real world calculations, we are basically stuck with formulas to work with.
I still don't think you understand the point I was getting at.
The meaning of the Coefficient of Drag is that when multiplied by some area value, it gives a number which when multiplied by the aerodynamic force (1/2 Rho V^2) gives the actual drag force. The problem is that the reference area is usually the Wing Area, thus if you have two aeroplanes with the same actual drag but one has a larger wing, the one with smaller wing has a lower coefficient of drag.
My earlier example was intended to illustrate this point.
Imagine that I just got a stock Grumman Bearcat and want to race it.
I do just a few things to clean up the airframe a bit: Reduce the size of the canopy, Align and seal panels and gaps, gun ports, etc.
Let's say that my racing Bearcat new has about a 5% decrease in drag force as shown in a wind tunnel test.
The Coefficient of Drag obviously just went down, right? I should go faster than stock....
But I am not satisfied.
Next, I chop several feet off each wing tip because this will be a pylon racer and not a fighter.
Let's say the wind tunnel test now shows a 8% decrease in drag from stock instead of the prior 5%....
Removing the wing tips also removed some wing area, let's say 12%.
What do you suppose just happened to my Coefficient of Drag if the Wing Area is the reference area?
I have an aeroplane that now has substantially less absolute drag but has a higher Coefficient of Drag.
The induced drag probably went up just a touch, but at high speed, it tends to be much lower than parasitic drag.
These numbers may or may not be realistic, but I believe they illustrate my point about why CD doesn't always tell the full story.
Equivalent Flat Plate Area tells a better story when there is a substantial difference in wing area.
No argument here.
The problem is that sometimes it makes sense to get into the details and sometimes the details hide the overall picture.
- Ivan.
DarrenW
Staff Sergeant
I appreciate your example concerning the modified racing aircraft and I can see where you find problems with how CD0 is calculated. Do you have actual aerodynamic data of modified WWII aircraft used for pylon racing to support your assumptions? Not that I'm dismissing your ideas as I don't know the answer either. You just peaked my curiosity on the subject, that's all.
Why do you feel that flat plate drag calculations are a better way to understand parasitic drag figures? Wouldn't larger aircraft be at an automatic disadvantage using this approach when looking at overall aerodynamic refinement? I would say aircraft with similar wing areas could be compared with such calculations but as the differences grow it becomes more problematic.
Case in point. Which aircraft would you say is more aerodynamically refined, a Lockheed Constellation or a Sopwith Camel?
World Heritage Encyclopedia:
....In another comparison with the Camel, a very large but streamlined aircraft such as the Lockheed Constellation has a considerably smaller zero-lift drag coefficient (0.0211 vs. 0.0378) in spite of having a much larger drag area (34.82 ft² vs. 8.73 ft²).
So you can see why I have my reservations with flat plate drag calculations when discussing aerodynamic refinement of a particular aircraft, just as you are suspect of zero lift drag calculations. If the wing areas of two aircraft are close than drag area would be more ideal, but this is not the case with all comparisons.
Why do you feel that flat plate drag calculations are a better way to understand parasitic drag figures? Wouldn't larger aircraft be at an automatic disadvantage using this approach when looking at overall aerodynamic refinement? I would say aircraft with similar wing areas could be compared with such calculations but as the differences grow it becomes more problematic.
Case in point. Which aircraft would you say is more aerodynamically refined, a Lockheed Constellation or a Sopwith Camel?
World Heritage Encyclopedia:
....In another comparison with the Camel, a very large but streamlined aircraft such as the Lockheed Constellation has a considerably smaller zero-lift drag coefficient (0.0211 vs. 0.0378) in spite of having a much larger drag area (34.82 ft² vs. 8.73 ft²).
So you can see why I have my reservations with flat plate drag calculations when discussing aerodynamic refinement of a particular aircraft, just as you are suspect of zero lift drag calculations. If the wing areas of two aircraft are close than drag area would be more ideal, but this is not the case with all comparisons.
Last edited:
KiwiBiggles
Senior Airman
Because it's not a beauty contest; what matters is how much parasitic drag is generated by the non-lifting parts of the aeroplane. And for comparing this, the equivalent flat plate area is far more relevant than the Cd. Comparing drag coefficients related to an irrelevant reference like wing area, means that fatties like the P-47 or F6F end up looking similar to planes as svelte as the Bf 109. You can easily lower your Cd by increasing your wing area, but it's not going to help you at all.
Parasitic Cd is a useful measure for examining changes to a particular design, or for comparing aeroplanes of similar size. Using it to compare planes of different sizes is an exercise in navel-gazing.
DarrenW
Staff Sergeant
Yes, but with increased wing area usually comes reduced speeds, all other things being equal. This will effect the CD0 calculations as well, increasing it incrementally. But like what I posted concerning the comparison of the Camel to the Constellation (and I believe we are in agreement about this) you can use flat plate area calculations to help compare overall parasitic drag (or aerodynamic refinement) IF the wings are of similar area. The larger the difference gets the less accurate the comparison becomes.
Last edited:
I was getting a little too fancy with my example I guess. How about we just consider the same stock Bearcat but with no other changes than chopping a couple feet from each wing tip? I doubt it would reduce drag by much but certainly would reduce CD.
I can think of a couple real life examples but need to think of where I would get the aerodynamic data to illustrate.
First of all, it is not "Flat Plate Drag", but "Equivalent Flat Plate Area". An actual Flat Plate of the Equivalent Flat Plate Area would have much more drag (around 1.3x to 1.5x IIRC). It is a good measure because it is representative of the drag of the actual aircraft that you are trying to move through the air. The fact that a larger aircraft would have more drag is exactly what we are missing with a simple expression of CD.
Your Sopwith Camel versus Lockheed Constellation is actually a great example. Imagine that we have the two airframes sitting side by side in a hangar but without engines. I give each aeroplane a 200 HP radial engine. Which one do you suppose will be flyable with that kind of power and which would not? Why would the Constellation with such aerodynamic refinement not ever get off the ground?
As I see it, when you take two aircraft that can both accomplish the same mission, then one compares other factors such as straight line performance and maneuverability, durability, and other abilities. If one aircraft has measurably better performance and maneuverability, do we really care that it has more or less wing area or that has a different coefficient of drag?
If there is no difference from a performance standpoint, and there is a significant difference in size and weight, is there an advantage of one over the other? I believe the answer is that for equivalent capabilities, choose the smaller, lighter one because it will be easier to transport and store.
-Ivan.
DarrenW
Staff Sergeant
Hi Ivan,
Yes, I understand the simple concept that a larger aircraft will produce more overall drag, thus requiring more power to gain the desired performance. But again other factors such as weight should be considered too (using the Camel/Constellation comparison again). Is that what you've been trying to explain all along? If it is than I agree with your points. And thanks for correcting my verbiage concerning flat plate area. I was being a bit sloppy in my choice of words.
I'm also positive that you will agree that a biplane built in 1917 is less aerodynamically 'clean' than most if not all modern built monoplanes. I've only looked to the CD0 formula to derive refinement of an aircraft design, not how much overall drag it may have (which takes in to affect the wing design of course). Newer aircraft may have larger aerodynamic 'foot prints' but they are still far more advanced designs nevertheless. This is the only point that I was trying to make concerning zero-lift drag coefficient calculations.
So my next question would be is how do fans, such as what was used on the BMW 801 radial, affect cooling at higher speeds? Could there be a point where the blades themselves block airflow that otherwise would pass though and around the cylinder heads?
Yes, I understand the simple concept that a larger aircraft will produce more overall drag, thus requiring more power to gain the desired performance. But again other factors such as weight should be considered too (using the Camel/Constellation comparison again). Is that what you've been trying to explain all along? If it is than I agree with your points. And thanks for correcting my verbiage concerning flat plate area. I was being a bit sloppy in my choice of words.
I'm also positive that you will agree that a biplane built in 1917 is less aerodynamically 'clean' than most if not all modern built monoplanes. I've only looked to the CD0 formula to derive refinement of an aircraft design, not how much overall drag it may have (which takes in to affect the wing design of course). Newer aircraft may have larger aerodynamic 'foot prints' but they are still far more advanced designs nevertheless. This is the only point that I was trying to make concerning zero-lift drag coefficient calculations.
So my next question would be is how do fans, such as what was used on the BMW 801 radial, affect cooling at higher speeds? Could there be a point where the blades themselves block airflow that otherwise would pass though and around the cylinder heads?
DarrenW
Staff Sergeant
Sometimes it's nice to get two exceptionally well engineered aircraft engines side by side and marvel at the technology of the day:
KiwiBiggles
Senior Airman
Wouldn't that increase the Cd, as the wing area has decreased with no real decrease in drag?
Last edited:
I am glad you understand the difference. In this case wording is important because the drag of a flat plate and Equivalent Flat Plate Area are not really the same thing and I am sure there are folks that don't know the difference.
Agreed regarding aerodynamic quality difference between Constellation and Camel. I am guessing that the Camel simply could not go nearly as fast as the Constellation regardless of power that was installed.
The point I was trying to make was that actual performance beats the heck out of minor differences in "Aerodynamic Quality".
You kind of answered your own question with the photographs of BMW 801 and R-2800.
My own opinion is that the R-2800 is superior to the BMW 801 in most ways.
Hello KiwiBiggles,
Thanks for finding my error. What I actually typed would not be very coherent and not even support the point I was trying to make.
Chopping Wing Tips would obviously reduce wing area and INCREASE Coefficient of Drag.
- Ivan.
The last production variants of the 801 was the F and minor variations of that . Those engines were to have a 14 blade cooling fan which would be an external identifier but this fan was later abandoned (and the 12 blade reverted) to when it was found that it interfered with high speed cooling. (Due to the small inlet size on the 801 cowl the fan was apparently only for ground and climb cooling.) Most American radial cooling relied on a large opening and cowl flaps which, when closed, built up a high pressure bubble in front of the engine forcing most air around the cowl achieving a similar, if much less aesthetic, result as the small inlet but still had much better ground cooling and was ridiculously simpler. Small gains could be had the BMW way but large aircraft formation ground and flight ops made the extra effort too much ... I guess.
Hello Chuter,
My understanding was that the BMW 801F was never actually produced.
Perhaps you are discussing the BMW 801TS version that was fitted to the A-9 and F-9 series?
Those engines did have a greater number of fan blades.
Do you happen to know what versions came after those that reverted to the 12 blade cooling fan?
The high pressure bubble you are describing would certainly vary with dynamic pressure and thus would become less effective as speed increased, or is there something I am not seeing here?
If this lack of a spinner worked so well, then why did the P-47J and P-72 use spinners for their installations?
Note also that the radial versions of P-60 also wore spinners.
Why would the BMW method affect large aircraft formations? Are you describing the single engine control lever from the Kommandogerat?
If so, there was also a control for fine tuning RPM which could be used for formation flying.
- Ivan.
Shortround6
Major General
For the Boeing 307 the flat disk behind propeller helped the cooling in ground running. It was found that with the aircraft stationary (or moving slowly/taxing?) some of the air, instead of going reward through the cowl/baffles actually flowed forward around the propeller hub and blade roots, mixed with the incoming air and recycled. a fair amount of churning going on right in front of the propeller.
At some point the discs were dispensed with. I have no information on why or what other modifications were done.
SB2C Helldivers have pictures both with and without spinners. Please note a large spinner may perform a similar function to the flat disk and help prevent air from bleeding out around the prop hub in ground running.
However I would be very leery of making much in the way of generalizations.
Early SB2C.
Later version.
Difference in engine was 200hp for take-off and less at altitude. The higher powered engine had better fining and should have cooled better at idle/warm up settings. What we don't know is what (if any) changes were made to the cowl flaps and what changes were made to the internal baffles of the cowl or between cylinders.
Please note the Martin PBM-3D patrol bomber used fan cooled 1900 hp R-2600s instead of the non fan 1700 R-2600s of earlier versions. However the PBM-5 used non-fan 2100hp R-2800s which pretty much solved the overheating on take-off problem.
There are a lot of variables in cooling an air cooled engine and some designers attack the problem with different approaches.
Also spinners were used for several reasons. Including shrouding the blade roots and those large blade attachment sleeves from simply churning the air.
How much was theory and how much was actual tested fact I don't know.
A lot of spinner use seemed to involve trial and error in actual test flights. I would also note that even propeller 'theory" was far from exact as early trials aircraft using contra-rotating propellers on high powered aircraft (like the P-47J and P-72) often failed to show any improvement in performance and often slight decreases (6-8mph?).
At some point the discs were dispensed with. I have no information on why or what other modifications were done.
SB2C Helldivers have pictures both with and without spinners. Please note a large spinner may perform a similar function to the flat disk and help prevent air from bleeding out around the prop hub in ground running.
However I would be very leery of making much in the way of generalizations.
Early SB2C.
Later version.
Difference in engine was 200hp for take-off and less at altitude. The higher powered engine had better fining and should have cooled better at idle/warm up settings. What we don't know is what (if any) changes were made to the cowl flaps and what changes were made to the internal baffles of the cowl or between cylinders.
Please note the Martin PBM-3D patrol bomber used fan cooled 1900 hp R-2600s instead of the non fan 1700 R-2600s of earlier versions. However the PBM-5 used non-fan 2100hp R-2800s which pretty much solved the overheating on take-off problem.
There are a lot of variables in cooling an air cooled engine and some designers attack the problem with different approaches.
Also spinners were used for several reasons. Including shrouding the blade roots and those large blade attachment sleeves from simply churning the air.
How much was theory and how much was actual tested fact I don't know.
A lot of spinner use seemed to involve trial and error in actual test flights. I would also note that even propeller 'theory" was far from exact as early trials aircraft using contra-rotating propellers on high powered aircraft (like the P-47J and P-72) often failed to show any improvement in performance and often slight decreases (6-8mph?).
MiTasol
2nd Lieutenant
Good find Shortround
It appears that copy is sold but the NASM has a copy of the 1945 edition and they will provide quality photocopies for a modest price.
Aviation fuels and their effects on engine performance / / prepared for U.S. Army Air Forces on purchase order (33-038) 44-5908-E and Bureau of Aeronautics, United States Navy on requisition P.D. ENII-28528-45.
There is a google copy of the 1951 edition at Aviation fuels and their effects on engine performance, prepared by Ethyl Corporation. Supplied to U. S. Air Forces on purchase order AF-33(600)5312; ... - download using the Hathi Download Helper - 277mb so it is actually a reasonable copy tho as always with google copies foldouts are butchered.
Last edited:
