Aerodynamics and transoms...

[Errol]

Just a quick question on the new aero model. Does it simulate turbulence caused by flat faces facing away from the air stream? Any edge near the back of an object causes turbulence, which causes lots of drag, and the larger the flat face, and the sharper the angle between the face and the side of the vessel, the worse the turbulence. I'm wondering specifically for the case of fairings. If you are enclosing a very wide payload, is there a benefit to having the longest possible tapered section behind the payload?

[bewing]

The aero model simulates drag using cubes for each part. The cosine of the angle of the face times the drag of that face for that part. Air motion is not modeled. Turbulence is not modeled. Occlusion of the air by one part for another part is modeled, of course.

 

[Kryxal]

My understanding is that open nodes leaving a flat surface WILL cause an increase in drag, though.

[diomedea]

Turbulence is somewhat modeled since KSP 1.0. Back then (and up to version 1.1.3) the debug window included a Physics/DragProfile tab that allowed to see what contribution to drag coefficient (Cd) the different faces would generate, variable with speed (in Mach). The profile with Tail (surfaces facing away from the air flow) was representative of the multiplier applied (or rather the divider, as it always seemed to go in reverse to me); that value changes mainly to represent the onset of turbulence approaching and beyond Mach 1. Of course nothing as accurate as using fluid dynamics models (which would be quite expensive to compute, though Ferram's Aerospace Research add-on makes a very good job doing as close as possible).

(Have no idea why DragProfile was removed in KSP 1.2, but it was mainly used as a debug tool for the early versions of the aerodynamic model).

[Foxster]

I don't know the drag cube model stuff for it but, yes, the effect is there. 

If you taper the rear of a stack as well as the front then the drag is a lot lower. 

[Errol]

So a long tear drop shaped payload does perform better then a mushroom shape, even if they are the same width and height?

[diomedea]

50 minutes ago, Errol said:

So a long tear drop shaped payload does perform better then a mushroom shape, even if they are the same width and height?

I'd say, if mushrooms had to withstand serious aerodynamic flow, they would have evolved in a drop shape too . More seriously, yes, long thin objects work best (due to limited cross-section and minimal drag from surfaces parallel to the flow), and having tear drop shape at the extremities greatly reduces front and tail drag coefficient (from the dragcubes geometry with the parts, cosine of the flow incidence angle with each exposed face, and the drag multipliers that mimic Reynolds dependency as a function of speed for different attitude). Flat surfaces worked the same as properly curved ones up to KSP 0.90, while the aerodynamic model was very primitive.

[Errol]

Just want to say thanks for all the replies in here!