Regarding applying accepted Oswald Efficency factors in high G turns

[drgondog]

drgondog: I'd be happy to send it to you but in this case I don't have it .  I wish I did.

LOL - For a long time I had access to NAA 'stuff' via Al White... now it is nigh impossible to get access to NAA archives from Boeing..

Thanks for posting the June 46 report on the effect of AoA in CDp - it is the first time I have seen it and I am curious whether that was a NACA result or perhaps some arcane Schmeud methodology.

The June 1946 Report No TSCEP5E-1908 by Gentile has the most representative flight values that I use for the 51D
so far..

http://www.wwiiaircraftperformance.org/mustang/p51d-15342.html

[dtango]

lol I know what you mean with about Boeing.  I tried to get some docs from them but they no longer had a way to transfer from microfiche to any other format.  I almost bought them a digital camera as a work around.  Pretty pathetic.  The Boeing NAA history dept is really crappy.  I don't know, maybe I should try connecting them up with the NASM to get some gov't funds to get something that can read microfiche into PDF or something.  Given all the $ Boeing is sinking into the Dreamliner I doubt it's going to improve anytime soon.

As to P-51 drag polars NACA ACR 4K02 has a compilation of flight test and wind tunnel data.
http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19930092458_1993092458.pdf

Flight Test Drag Polar P-51B


Wind Tunnel Drag Polars P-51B


I haven't ever attempted comparing NAA-46-130 with the NACA data though for various reasons.

[drgondog]

lol I know what you mean with about Boeing.  I tried to get some docs from them but they no longer had a way to transfer from microfiche to any other format.  I almost bought them a digital camera as a work around.  Pretty pathetic.  The Boeing NAA history dept is really crappy.  I don't know, maybe I should try connecting them up with the NASM to get some gov't funds to get something that can read microfiche into PDF or something.  Given all the $ Boeing is sinking into the Dreamliner I doubt it's going to improve anytime soon.

As to P-51 drag polars NACA ACR 4K02 has a compilation of flight test and wind tunnel data.
 I haven't ever attempted comparing NAA-46-130 with the NACA data though for various reasons.

IIRC, NAA a.) stated the flight test results were in close agreement with the wind tunnel tests, but the approximation of the desert 'dust' was not as close to an attempt to reproduce a similar surface roughness to get true RN equivalency?

[Stoney]

It's significant at high aoa Stoney.  For example just look at the P-51 graph above.  AHT has CD for the P-51D at .0176 for level flight.  According to Pyro's data at Cl of 1.0 the incremental drag due to viscous separation is about .016 by itself so the viscous effects by itself nearly doubles CD.

No, I understand the relationship of AoA to Cd...I was curious about the original post regarding estimating "e".  Perhaps I should have worded it differently.  Raymer states that Oswald's technique is only valid at moderate AoA.  He recommends the suction method as more accurate for high-speeds and to account for viscous separation...etc...

My question was how much variation in "e" do we see, away from the oft used estimate of .8, in high AoA maneuvers and whether or not it was significant.  Perhaps I'm misunderstanding something.

[dtango]

No, I understand the relationship of AoA to Cd...I was curious about the original post regarding estimating "e".  Perhaps I should have worded it differently.  Raymer states that Oswald's technique is only valid at moderate AoA.  He recommends the suction method as more accurate for high-speeds and to account for viscous separation...etc...

My question was how much variation in "e" do we see, away from the oft used estimate of .8, in high AoA maneuvers and whether or not it was significant.  Perhaps I'm misunderstanding something.

Ah I see.  Using the same P-51 data from Pyro above I get a value of e~ .6 at Cl=1.0 vs. e=.87 at level flight.

For reference here's a comparison of e variation for modern jets at load factors of up to 3 g's so the variation of e from low aoa can be significant as evidenced.

[drgondog]

Ah I see.  Using the same P-51 data from Pyro above I get a value of e~ .6 at Cl=1.0 vs. e=.87 at level flight.

For reference here's a comparison of e variation for modern jets at load factors of up to 3 g's so the variation of e from low aoa can be significant as evidenced.

(Image removed from quote.)

Interesting Tango - I find it interesting that L/D as a function of g is linear, particularly with relatively low aspect wings..  but then at .6M I guess I should expect Induced Drag to be relatively low.

[drgondog]

No, I understand the relationship of AoA to Cd...I was curious about the original post regarding estimating "e".  Perhaps I should have worded it differently.  Raymer states that Oswald's technique is only valid at moderate AoA.  He recommends the suction method as more accurate for high-speeds and to account for viscous separation...etc...

My question was how much variation in "e" do we see, away from the oft used estimate of .8, in high AoA maneuvers and whether or not it was significant.  Perhaps I'm misunderstanding something.

Stoney - here is Raymer's doctoral thesis.  I am going to spend more time understanding his approach to Design Optimization but have gone far enough to see that he will not attempt a 'closed form solution' via Calculus of Variations..  I also noticed he does not define or use 'e' in his Nomenclature/Index of terms but may have overlooked something.

I do recall in past cursory reading of one of his design texts that he used 'e' strictly as a relation to wing planform w/o taking into consideration a 'wing/body' effect to 'e'

http://www.aircraftdesign.com/RaymerThesisFinalRevLowRes.pdf

[dtango]

A couple of thoughts on Raymer's & Shevell's e estimation methods as I understand them...

Raymer uses a semi-emperical approach to estimate K from leading edge suction.  The issues with this approach are: a) you use a look up chart for S based on Cl but this is based on knowing the design Cl of the wing, b) the technique was geared more toward thin wing, low aspect ratio jets, c) it doesn't account for the entire plane including the fuselage.

Shevell's approach is also a semi-emperical approach based on data obtained from data from Douglas aircraft and accounts for fuselage affects.  However in the past when I've run the calculations using Shevell's estimating method it estimates e much higher than what the P-51 data posted by Pyro shows.

I believe both of their estimation methods are intended for estimating initial designs.  They are great for that purpose.  However to reverse engineer and accurately estimate lift dependent viscous drag for an airplane I think we're stuck with getting into nasty Navier-Stokes world of CFD...

...or go find some good drag polars for specific aircraft based on flight tests or wind tunnel tests .

[Stoney]

How much impact does the fuselage have at high AoA?

[dtango]

How much impact does the fuselage have at high AoA?

It's all about what happens in the boundary layer and the boundary layer surrounds the surface of the entire plane, fuselage and all.  The details are very important.  Something simple as an open tail wheel well, the angle of the windscreen, curvature of the fuselage, etc. make big differences on the pressure gradient in the boundary layer.  So qualitatively it has a signifcant impact.  Quantitatively it's hard to estimate what that impact would be without some serious computation.

[gripen]

Well, Oswald's original idea was that Cl^2/Cd relation is usually roughly linear (ie the Cl/Cd is parabolic) between Cl 0.1-1.0 and as this Cl range covers about pretty much entire flight envelope, it's been proven to work pretty well. As example Pyro's data from NAA report gives almost linear line at this Cl range. However, its probably a calculation.

[drgondog]

A couple of thoughts on Raymer's & Shevell's e estimation methods as I understand them...

Raymer uses a semi-emperical approach to estimate K from leading edge suction.  The issues with this approach are: a) you use a look up chart for S based on Cl but this is based on knowing the design Cl of the wing, b) the technique was geared more toward thin wing, low aspect ratio jets, c) it doesn't account for the entire plane including the fuselage.

Shevell's approach is also a semi-emperical approach based on data obtained from data from Douglas aircraft and accounts for fuselage affects.  However in the past when I've run the calculations using Shevell's estimating method it estimates e much higher than what the P-51 data posted by Pyro shows.

I believe both of their estimation methods are intended for estimating initial designs.  They are great for that purpose.  However to reverse engineer and accurately estimate lift dependent viscous drag for an airplane I think we're stuck with getting into nasty Navier-Stokes world of CFD...

...or go find some good drag polars for specific aircraft based on flight tests or wind tunnel tests .

We appear to be in violent agreement..

[drgondog]

Well, Oswald's original idea was that Cl^2/Cd relation is usually roughly linear (ie the Cl/Cd is parabolic) between Cl 0.1-1.0 and as this Cl range covers about pretty much entire flight envelope, it's been proven to work pretty well. As example Pyro's data from NAA report gives almost linear line at this Cl range. However, its probably a calculation.

Go look at the charts (Oswald's Report 408, charts 24) on page 25 which are flight test results for various German a/c.  Note the rapid change in CDp with respect to CL @ ~ 1.0 with no further extrapolation (other than a huge gradient) past CL=1 .

For the turning CL's experienced in high G manuevers, the CL's approach CLmax, which for most WWII fighters is between 1.4 and 1.7 for zero flap manuever.  

It is clear from the charts that CDp is changing far more rapidly than CDi above CL=1.0

[Stoney]

Ok, but I guess where I'm hanging up on this is that it would seem that as the wing approaches extreme AoA, the drag from the fuselage isn't the rub, its the fact that the boundary layer on the wing is almost totally separated.  Is the Cdi the only indication of the separation on the wing?  Is the increase in drag primarily from the fuselage/empenage?

[drgondog]

Ok, but I guess where I'm hanging up on this is that it would seem that as the wing approaches extreme AoA, the drag from the fuselage isn't the rub, its the fact that the boundary layer on the wing is almost totally separated. 

Is the Cdi the only indication of the separation on the wing? 

No. CDi is solely associated with the Induced Drag of the wing and expressed as CL>>2/pi*AR

Is the increase in drag primarily from the fuselage/empenage?Mostly yes - in the form of incremental Parasite Drag, and further referred to as viscous drag related to lift - more below. It affects the fuselage and tail by immersing those components into a larger wake - but also affects friction drag on the wing aft of the separation

Stoney - The boundary layer growth and separation on a wing  is caused by adverse pressure gradient.  When it 'separates' the wake/form drag and even friction drag increases dramatically.  These effects are Not limited to the wing but affects the fuselage/tail, etc which are immersed in the expanded wake.These components are frequently further defined within parasite drag as 'viscous drag related to lift' - but not part of induced drag.

 'e' theoretically accounts for these incremental 'delta' drag components related to lift, and as such, is built into the Induced Drag component of Total Drag. Initially CDi was expressed only in context of wing, but the more sophisticated developments of 'e' also accounted for the wing/fuselage combinations (high/middle/low wing for example)

Induced drag, is the drag due to the lifting line discontinuity of the wing at the wing tip - which creates an infinite vortex and is a function of the variation from elliptical wing planform as well as aspect ration to further refine from two dimenional airfoils to 3-d realities..