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

[gripen]

I'm refering this report (allready linked above):

http://hdl.handle.net/2060/19790073957

It's in flight measurements with the real XP-51 (2D profile drag). The polar (3D, entire airframe) comes from another tests on 1:3 scale model of the P-51:

http://hdl.handle.net/2060/19930092693

Generally the laminar flow bucket is really visible only in the 2D data.

In addition here is another report on profile drag:

http://hdl.handle.net/2060/19930092825

[Stoney]

Generally the laminar flow bucket is really visible only in the 2D data.

Well, that's my point.  The 2D data merely shows the Cd of the airfoil.  The 3D data shows the Cd of the entire aircraft.  Given all the other components of drag, I wouldn't expect to see a "bucket" for the entire aircraft.  The "bucket" should only be present for the airfoil.

Perhaps we're misunderstanding each other, as my only observation was that the P-51 airfoil was more efficient at design lift coefficients than the 23XXX series.  The airfoil certainly is not the only part of the entire airframe to consider with respect to the Cd of the entire aircraft.

[gripen]

The 2D data merely shows the Cd of the airfoil.  The 3D data shows the Cd of the entire aircraft.  Given all the other components of drag, I wouldn't expect to see a "bucket" for the entire aircraft.  The "bucket" should only be present for the airfoil.

My point is that the measured far higher than ideal profile drag in the case of the real XP-51 indicates that there was no large amount of laminar flow in the service condition (see the first and the third linked reports). In other words there was no laminar flow bucket in 2D nor 3D. Basicly similar phenomena as seen in the case of the NACA 66 with standard roughness.

[Stoney]

the measured far higher than ideal profile drag in the case of the real XP-51 indicates that there was no large amount of laminar flow in the service condition (see the first and the third linked reports). In other words there was no laminar flow bucket in 2D nor 3D. Basicly similar phenomena as seen in the case of the NACA 66 with standard roughness.

Well, I looked through all 3 reports again, just in case I missed something.  I noticed the table that listed the single Cd at differing finishes, but that was a single point Cd.  Sure, the test values were higher, but a higher Cd at a single point does not mean that a bucket wouldn't have existed had an entire polar been generated in that roughness condition.  The third report again tests the entire aircraft, and the Cd polar represents the entire airframe, and thus, there would be no bucket either.

Perhaps I'm just misunderstanding, but I'm not able to find in your references that a "bucket" didn't exist in the service condition.

[gripen]

IMHO you don't seem to understand that they measured profile drag coefficient ie 2D in the first and third report:

http://hdl.handle.net/2060/19790073957  (this the report on in flight measurements of one section on the XP-51)

http://hdl.handle.net/2060/19930092825  (this is the report wind tunnel measurements of one section)

The results are directly comparable; the tested section was the same and the Re number in flight tests is close to wind tunnel tested. Given that in flight measured profile drag coefficient is far higher (about 50% higher in service condition), it's safe to assume that laminar flow did not happen in large degree in service condition. Therefore also it's safe to assume that there was no similar laminar flow bucket in the service condition as seen in the wind tunnel data.

Only the second report contains 3D measurements of entire airframe:

http://hdl.handle.net/2060/19930092693  (this is the report on wind tunnel tests on 1:3 scale model)

It's easy to see difference between 2D and 3D measurements because the absolute values of drag coefficient are very different: around 0,0045-0,0070 for 2D and around 0,02 for 3D at given Cl.

[Stoney]

Given that in flight measured profile drag coefficient is far higher (about 50% higher in service condition), it's safe to assume that laminar flow did not happen in large degree in service condition.

How do you make that assumption?  In the first report, they only measured the Cd's at Cl's less than .4 which was the design lift coefficient for the airfoil.  The plot for the "unfinished" profile begins to suggest a "bucket", even though they plotted it with a moving average.  I suspect that if they had completed the plot through say, a Cl of 1.0, we might have enough information to determine whether or not a bucket existed in the service condition.  The wind tunnel plots show a bucket on each and every chart.

[gripen]

We know from the wind tunnel tests (third report) that when air flow stays laminar large part of the wing then the profile drag coefficient is around 0,0045 at Cl 0,15. However, in the flight tests measured profile drag (first report) is far higher in service condition, over 0,006 so we can conclude that the airflow is turbulent most of the wing. In other words there can't be large transition from laminar to turbulent when the Cl increases because airflow is allready turbulent most of the wing even at low Cl values. Therefore there can't be large laminar flow bucket either.

Another way to look the issue is to interpolate a polar using the third report assuming that most of the wing is allways turbulent (ie no large transition thus no large bucket). In that case the profile drag coefficient would had been a bit over 0,005 at Cl 0,15 which corresponds very well specially treated surface in the first report. There probably was some amount of transition even in the case of the unfinished wing, however, even in the case of the specially treated surface the transition was proabably far less than in the case of ideal smooth surface. Notable thing is  that similar transition existed in some degree with with other profiles as well.