Why are drag polar CD and speed debug window CD diffferent?

Question

question

donstim asked Jul 14, '21 donstim commented Aug 31, '21

Why are drag polar CD and speed debug window CD diffferent?

Another drag-related question -- One can see the drag polar (CL vs CD) with the debug polar window. You can also get the current CL and CD from the speed debug window. However, the CD from the drag polar for the CL shown in the speed debug window can be quite different than the CD in the speed debug window.

See the screenshot below. In the speed debug window CL is 1.77436 and CD is 0.17123. If you go to the drag polar for that CL, the CD is well above 0.175. Digitizing that plot, you will find that CD from the drag polar is 0.18236. Why is the CD from the drag polar so much different than the CD shown in the speed debug window?

aircraft flightmodel

conf-full.png (2.0 MiB)

10 |10000 characters needed characters left characters exceeded

Attachments: Up to 5 attachments (including images) can be used with a maximum of 4.8 MiB each and 23.8 MiB total.

Answer 1 · 2021-08-31T09:18:08Z

FlyingRaccoon answered Aug 31, '21 donstim commented Aug 31, '21

Hello

Yes I was talking about propeller slipstream. Ok that's not the culprit if you're talking about the A320.
In fact there are a lot of possible explanations for this, I'll try to give a bigger picture.

The value printed in the Debug Aircraft Speed is the final value measured on the aircraft in game while the polar is the result of the normalization process. The normalization process happens in a very controlled and theoretical (simplified) environment, while the final value is the result of a more complex calculation.
During the normalization process, there's no wind, turbulences, side slip, engine side effects, control surfaces are in a perfectly neutral position, pressure is homogeneous, etc... this is basically a plane gliding in perfect conditions.
The final values are computed using the final aircraft FM in its final environment so any of these will have an influence and that's the difference you're observing.

For example, the dynamic pressure is homogeneous in the wind tunnel but for the final FM, we perform dynamic pressure measures all over the aircraft (thousands of those). Typically the top of the aircraft is in a less dense air than the bottom. I'm not saying this is having a big influence, just demonstrating the difference between the theoretical model and the final more complex FM.

Another more relevant example would be the the way the lift is computed. The normalization computes the lift at AOA0 and just before the stall and considers a straight line between the two while in reality (and in our end FM) this is more of a parabolic. On the A320, that can account in some cases for a 5% difference in the drag value between the normalization and the final FM.

We double checked with the A320 and flying in the appropriate conditions, the difference is really negligible. (We've added Min/Max/Avg for the final values so you have a better idea how it changes across the aircraft)

Note: a bug has been fixed with debug values being incoherent when too many debug panels are opened.

I hope this answers your concerns.

Regards,
Sylvain

1630399570947.png (36.1 KiB)

1630401261650.png (1.6 MiB)

1

10 |10000 characters needed characters left characters exceeded

Attachments: Up to 5 attachments (including images) can be used with a maximum of 4.8 MiB each and 23.8 MiB total.

donstim · Aug 31 at 09:09 PM

Thank you. This makes sense and does provide a complete answer to the question. I had done a number of comparisons of the drag coefficient from the speed debug window compared to the drag polar. The difference seemed to be configuration dependent for our FBW A320. In most cases, the difference was on the order of a percent or two, with some cases being around 6%. The worst case I saw was pretty high though, 13.5% in CONF 3.

1 ·

Answer 2 · 2021-07-14T22:06:10Z

boufogre answered Jul 14, '21

I am interested too in knowing the answer to this one.

10 |10000 characters needed characters left characters exceeded

Attachments: Up to 5 attachments (including images) can be used with a maximum of 4.8 MiB each and 23.8 MiB total.

Answer 3 · 2021-08-27T15:28:38Z

FlyingRaccoon answered Aug 27, '21 donstim edited Aug 27, '21

Hello

The displayed values depend on if the engine slipstream is taken into account or not.

Once again, we improved the debug panel to make it more explicit:

Regards,
Sylvain

1630077992674.png (450.7 KiB)

2

10 |10000 characters needed characters left characters exceeded

Attachments: Up to 5 attachments (including images) can be used with a maximum of 4.8 MiB each and 23.8 MiB total.

FlyingRaccoon ♦♦ · Aug 27 at 03:28 PM

@Nocturne FYI

1 ·

donstim · Aug 27 at 07:29 PM

What do you mean by "engine slipstream?" Are you talking about propeller slipstream? My question was relative to the A320, which has wing-mounted jet engines. What slipstream effect is being added? Any engine effect (with both engines operating) should already be included in the basic lift curve and drag polar. There should be identifiable windmill drag and yaw (control) drag in the case of shut down engine. (At least that's the way it is done for airplane performance modeling IRL.)

0 ·

question