P-51D, Tempest, LA-7 Airfoil Comparisons

[gripen]

Tilt,
It's just that using the polars of the entire airframe give quite bit different results than the wing section data. Below is the polars determined in the wind tunnel with a 1:3 scale model of the P-51. Note that I admit right away that scale models tend to give lower drag values than the real one in the full scale tunnel like the La-7 in the T-101 in your data so following is just an presentation how to use the polars:



To get the normal form of the drag polar:

Cd = Cd0 + (CL^2/(pi*AR*e)

I quickly calculated the value of the e being 0,87 for the P-51 (using just three samples from the above polar) and from your data I got 0,84 for the La-7 (I used the better one of the two between Cl 0,2 and 1,0, I don't know what's the difference between these). So polars can be presented:

P-51 => Cd = 0,02 + (CL^2/(pi*5,8*0,87)

La-7 => Cd = 0,025 + (CL^2/(pi*5,5*0,84)

And graphically:



And further as lift to drag ratio:



So in practice there was no dramatic difference between these planes but these are just examples how to use the polars; my readings are just quick samples. In real life the Cd0 was around 0,022 for the P-51B in the full scale tunnel and value of the e was probably a bit lower (I tend to use numbers around 0,75 in my own calculations).

[Stoney]

Stoney,
I'm not claiming that you have stated otherwise, sorry if you got such impression. However, these are more or less theoretical values; in practice in flight measured profile drag coefficients of the P-51 were far higher than wind tunnel measured...Notable thing is that in the case of profile of the Mustang, so called "laminar flow bucket" disapeared with standard finish ie the polar shape was in practice very similar with older profiles...In other words my opinion is that your comparison is more or less theoretical and valid only in the perfect conditions. Tilt's approach by using polars of the entire airframe is much better for comparing airframes.

OK.  To be honest, since we're simply comparing airfoils, I could have made the comparison to any of the NACA 23000 users out there, as it was the predominant airfoil used for the period.  By using the Reynolds numbers that were associated with the respective geometries of the two aircraft, I thought it would be a little more realistic than simply a side by side comparison at theoretical reynolds numbers.

Since the drag coefficient for the aircraft includes more than simply the wing, obviously there are other factors involved.  What can be said though is, comparitively, the Pony's airfoil is more efficient at its design lift coefficient than the NACA 23015, trim drag comparison not withstanding.  (I was going to factor in pitching moments in the comparison, but didn't trust the XFoil data I got from the analysis).

But, like I said, I wasn't comparing the airframe--I tried to make that clear in the disclaimer posted early before the data.

[Tilt]

I wish I understood this................

However is there now enough data to dial in a comparison between the well documented P51 and the little documented La7 in particular with respect to acceleration and high speed dive characturistics.


btw re the two curves

The data is accompanied with photographs showing the ac at AoA's 2, 6, 8, 12, 16, 18, 20 degrees however there are 3 photographs per AoA

1 shows a standard La7 with engine cooling vanes fully feathered.
2 shows an La7 with additional booster bulges (on either side of the fuselage) again with engine cooling vanes fully feathered.
3 shows the same ac as 2 but now with cooling vanes open. (more drag)

[Stoney]

I wish I understood this................

However is there now enough data to dial in a comparison between the well documented P51 and the little documented La7 in particular with respect to acceleration and high speed dive characturistics.

What Gripen is talking about is the difference between a 2 dimensional airfoil comparison (what I'm showing), and the 3 dimensional wind tunnel tests.  I'm looking at merely a slice of the wing at the mean aerodynamic chord.  Obviously, since a wing is 3-D, there are other factors that the 2D comparison won't show.  Gripen is correct in saying that an airfoil comparison doesn't take into account the 3D flow characteristics of an entire wing.  Usually, a wind tunnel test will give a coefficient of lift/drag for the whole aircraft, while what I'm showing is merely that slice of the wing that exists at the MAC.  That's why the maximum coefficients of lift are lower on the wind tunnel documentation.

The P-51D, as you alluded to, has plainly demonstrated documentation that shows the whole aircraft coefficient of drag.  I cannot, through mathmatical terms, create that.  Wind tunnel, flight testing, or computational fluid dynamics (CFD) software can.

What I was merely trying to do was show a comparison between the two airfoils only.  I don't know if there's a published Cd for the LA-7.  If there is, then we could plug them into the acceleration formulas for comparison.  But, I can't create one--we'd have to find some documentation that shows one for a more precise comparison.

[gripen]

However is there now enough data to dial in a comparison between the well documented P51 and the little documented La7 in particular with respect to acceleration and high speed dive characturistics.

Infact your data on the La-7 is very good for the airframe analysis; it's the real thing in the wind tunnel and the data shows large Cl range.  It's tested just at one Mach and Reynolds number but it's still rather valid up to about Mach 0,4-0,5 because Cl/Cd ratio is quite constant up those speeds.

The data is accompanied with photographs showing the ac at AoA's 2, 6, 8, 12, 16, 18, 20 degrees however there are 3 photographs per AoA

1 shows a standard La7 with engine cooling vanes fully feathered.
2 shows an La7 with additional booster bulges (on either side of the fuselage) again with engine cooling vanes fully feathered.
3 shows the same ac as 2 but now with cooling vanes open. (more drag)

I don't really know if the vanes are the reason for the difference between the curves because the effect of the vanes should show up at all speeds as increased drag. However, the dragcurves are the same up to Cl (Cy) 0,4 or to the AoA of 4deg. My first impression was that the difference is caused by the automatic slots but that is just a quess. Perhaps some one who can Russian could translate graphs?

[SgtPappy]

Does anyone know if the AFDU trials or any turning performance trials during the war actually tested aircraft with their flaps down at varying degrees?

[Stoney]

However, these are more or less theoretical values; in practice in flight measured profile drag coefficients of the P-51 were far higher than wind tunnel measured.

Notable thing is that in the case of profile of the Mustang, so called "laminar flow bucket" disapeared with standard finish ie the polar shape was in practice very similar with older profiles.

Well, if I ratchet up the roughness used in the calcs, both airfoils show roughly the same comparison, albeit with much higher drag counts.  The shape of the graphs are not quite as drastic, but maintain the same basic shape.  I think it is safe to say that even though it doesn't match the theoretical values shown above, it still points to the P-51 profile being much more efficient on-design.  The NACA 23XXX series were the most prevalent airfoil of the war, and certainly performed Yeoman's work for many aircraft of all sides in the war.  Interestingly enough, as alluded to above, this doesn't take into account trim drag.  The 23XXX was popular for many reasons, chief among them the lack of high pitching moments, while the P-51D airfoil had much higher pitching moments.  Perhaps if a trim drag comparison was conducted, the gap between the two on-design would narrow appreciably.

<S>

Stoney

[gripen]

Well, if I ratchet up the roughness used in the calcs, both airfoils show roughly the same comparison, albeit with much higher drag counts.  The shape of the graphs are not quite as drastic, but maintain the same basic shape.

Below is one set of Naca 66 series section polars from "Theory of wing sections" by Abbott and von Doenhoff. I choosed this one because it just happened to be quite similar as profile used in the Mustang. Notable thing is that in the case of the smooth profile the air flow stays laminar at quite large part of the wing at low Cl values (say below 0,1). However when the Cl increases, the transition point from laminar to turbulent starts to move forward at certain point increasing drag as can be seen between Cl value 0,1-0,3 and at higher Cl values transition point stays again quite constant causing slower increase of the drag. This is the phenomena which causes so called "laminar bucket".

In the case of the Standard roughness, the airflow is allready turbulent at most of the wing even at low Cl values so there is no sudden movement of the transition point and no "laminar bucket" either. Therefore the basic shape of the polar should not be the same as in the case of the smooth surface.

[Stoney]

In the case of the Standard roughness, the airflow is allready turbulent at most of the wing even at low Cl values so there is no sudden movement of the transition point and no "laminar bucket" either. Therefore the basic shape of the polar should not be the same as in the case of the smooth surface.

Among most resources I've read, the NACA standard roughness figures are not "standard" with respect to service conditions of wings.  Service condition is certainly worse than the theoretical values, but not nearly as rough as the standard roughness.

[gripen]

Among most resources I've read, the NACA standard roughness figures are not "standard" with respect to service conditions of wings.  Service condition is certainly worse than the theoretical values, but not nearly as rough as the standard roughness.

Well, we have measured data on the XP-51 (posted above), the normal condition profile drag values being near 50% higher than in the case of the smooth surface. There is also a lot of discussion on issue in the "Theory of Wing Sections".

[Charge]

"Below is one set of Naca 66 series section polars from "Theory of wing sections" by Abbott and von Doenhoff. I choosed this one because it just happened to be quite similar as profile used in the Mustang. Notable thing is that in the case of the smooth profile the air flow stays laminar at quite large part of the wing at low Cl values (say below 0,1). However when the Cl increases, the transition point from laminar to turbulent starts to move forward at certain point increasing drag as can be seen between Cl value 0,1-0,3 and at higher Cl values transition point stays again quite constant causing slower increase of the drag. This is the phenomena which causes so called "laminar bucket"."

The effect of that is also well visible in Stoney's first set of plots with AoA to Cd and Cl as the effect of "the bucket" extends there too.

Gripen, could you post one of the 2300 series just for comparison. I have the book but not a working scanner...

-C+

[gripen]

Here is the 23012.

[Stoney]

Well, if you look at the standard roughness plots for almost every airfoil in the entire book, you'll see a Cdmin of something around .010, +/- about .2.

Given that, I have a hard time believing that the NACA standard roughness is representative of the "service condition".  I agree that any airfoil in the service condition will fail to represent the theoretical values, but NACA standard roughness is anything but standard.  Lednicer, Ribblet, etc. agree.

[gripen]

Hm... I'm not refering on Standard roughness but on measurements on the XP-51 posted above. The service condition profile drag being about 50% higher than the ideal case so large amount of laminar flow is unlikely and therefore also appearance of the "laminar flow bucket" is unlikely as well. I posted the pages from Abbott and von Doenhoff just to show the phenomena, absolute values are irrelevant.

[Stoney]

Hm... I'm not refering on Standard roughness but on measurements on the XP-51 posted above. The service condition profile drag being about 50% higher than the ideal case so large amount of laminar flow is unlikely and therefore also appearance of the "laminar flow bucket" is unlikely as well. I posted the pages from Abbott and von Doenhoff just to show the phenomena, absolute values are irrelevant.

I thought that chart was from a wind tunnel model of the entire aircraft?  Given that, I don't think you'd see a "bucket" that resembles the airfoil "bucket".  I'll do a chart showing the effects of roughness on the P-51D root and we'll see how it plots.