Mark Thompson said:Quoted message said:Turbulent is 'better' [1]by the way it results in less drag
hence "turbulators" on the wings of gliders.
Why's that then? (in near baby language please)
You can have laminar or turbulent boundary layers. Laminar are
smooth, thin, low friction and not very sticky. Turbulent are
turbulent (you don't say), thick, high friction and sticky.
There are two components to drag - the drag arising from teh boundary
layer, and the drag arising from boundary layer separation.
Separation is where the boundary layer gets pulled away from teh
surface, and still air creeps in behind. For example, if you have a
flat plate moving through the air, flat face-on, the air strikes teh
front face, flows across teh face to teh edge, but then it doesn't
manage to get round the edhge and flow across teh back face - the
boundary layer separates at teh edge of the plate and there's stil air
behind.
Dare I try ascii art? The [censored] is the plate, and I've drawn two
'stream lines':
-----------------------
/ X
| X
| X still air
===========< X in here
| X behind plate
| X
\ X
-----------------------
By bernoulli (or, as observed and codified by Bernoulli), teh still
air has higher pressure than teh flowing air, so there's a pressure
differential on each side of teh plate. This gives a force, which is
part of the drag force.
IF separation occurs at all, it is almost invariable that the pressure
differential is the major part of teh drag force - the friction of teh
boundary layer on teh front of teh plate is very very small compared
to teh effect of the pressure differential. The friction of the
boundary layer is small regardless of teh type of boundary layer.
I don't know actual numbers, but a laminar flow on teh front plate
might be 1 unit, if it were turbulent it might be 2 units, but the
pressure differential is 100 units.
Now, consider a sphere ('cos the plate is not so good for the next
bit). If you have a sphere with a laminar boundary layer, you get
separation of teh boundary layer near the widest point (maybe even
just ahead of teh widest point). There'll be a full pi r^2 of area
for teh relatively high-pressure still air inside the separation to
act on - high drag.
If you have a turbulent boundary layer, it's stickier, and stays
attached to teh surface - it might get 20 or 30 degrees past the
widest point before it separates. Now there's less area - maybe half
as much area for the pressure differential.
So using similar number as the plate - suppose laminar had 1 unit on
teh front of teh sphere, and 100 units due to pressure differential,
giving 101 units of drag.
Turbulent is twice the friction, and it goes further round teh surface
- it might have 3 units of drag. However, the pressure differential
is halved - so we have 53 total drag.
I'm not doing that in ascii - try this:
http://www.princeton.edu/~asmits/Bicycle_web/pictures/R_combined.GIF
picture on the left - separation near the widest point, wide wake.
Picture on teh right - a trip-wire turbulator is added, forcing a
sticky turbulent boundary layer that stays attached further round teh
sphere, and a narrow wake.
In fact, you could look at the whole of
http://www.princeton.edu/~asmits/Bicycle_web/bicycle_aero.html for an
on-topic discussion of aerodynamics! (now why didn't I find that when
I started typing).
regards, Ian SMith
--
|\ /| no .sig
|o o|
|/ \|
