Aerospace Wing transition to turbulent flow

  1. Jul 27, 2010 #1
    To preface, I know about ~nothing about aerodynamics, but have been reading through wiki pages half the morning, but am still not quite sure how to answer a question.

    Basically my question is concerning the turbulent flow that happens behind an aircraft, maybe just focusing on the wing for now. My understanding is that laminar flow typically takes place on regular (say commercial) aircraft wings, and a colleague of mine (who a long while back majored in Aerodynamics) stated that the turbulent flow *behind* the aircraft was very important.

    Specifically he was citing the pressure differential between the front and end of the wing, and stated that it was a considerable factor relative to thrust. My understanding is that turbulence generates lower pressure than laminar flow behind the wing (wrong maybe), such that in that regard turbulent flow would effectively slow down a plane more than laminar flow.

    Alternatively there is the effective cross section of a wing, maybe also described as the flow displacement extending beyond the physical wing. (I can't remember what your term is for it). Turbulent flow in general should lower this is my understanding. Creating early turbulent flow is one way, and turbulent flow from behind the wing the other possibility.

    So the heart of my question is actually which of the two processes (or maybe something else), is the dominant force on planes (obviously drag is dominant, but next leading in order). I'd guess it would have to be geometry and speed dependent, so any guidelines would be useful.

    Thanks in advance. Hopefully my post is comprehensible -- I don't feel comfortable using a lot of the standard terminology.
  2. jcsd
  3. Jul 27, 2010 #2
    Not quite, turbulence is a measure of unsteadiness in flow. This happens inside the boundary layer. There is always a transition point at some point along the length of the wing chord where the flow transitions from laminar to turbulence. This has nothing to do with 'turbulent flow behind the airplane.' Naturally, you want to minimize turbulence and increase laminar flow as much as possible, with the provision that laminar flow has enough energy within its boundary layer to prevent separataion (see Golf Balls for why turbulent boundary layers are sometimes better).

    Wings never generate thrust, only lift and drag.
  4. Jul 27, 2010 #3
    Yes obviously the wing doesn't generate thrust.

    So your statement is that the only meaningful turbulence is the turbulence that occurs along the wing (and in general we try and minimize that turbulence). So the turbulence behind the plane is effectively meaningless, as far as flight dynamics go.
    Last edited: Jul 27, 2010
  5. Jul 27, 2010 #4
    Sorry, I missed the word 'relative' in that sentence, which changes things. That being said, of course! In steady wings-level flight thrust = drag. The drag is specifically due to disturbances in the air, which will manifset itself as a trailing vortex downstream of the aircraft. Turbulence is a word that is often abused and misused, so let's be clear on three things: Vorticity (think swirl), Turbulence (deviation in the flow from a straight streamline), and separation (large components of unsteady flow due to high angle of attack at values of stall). They are interrelated, but its important to recognize the differences in terminology.

    For the most part, when you say turbulence, think boundary layer.
  6. Aug 21, 2010 #5
    the reason you want turbulent flow is to push back airfoil seperation on the cross section of your wing. turbulent flow has higer energy than laminar flow which prevents seperation happening too early.

    turbulance flow generates more friction drag (than laminar flow) but friction drag ist very small in comparison to pressure drag from seperation. So turbulance flow is much more the choice.
  7. Aug 25, 2010 #6
    I'm glad you mentioned this, as "a measure of unsteadiness" indicates some sort of smooth function between laminar and turbulent flow. You've already indicated the juncture tends to be fairly abrupt, which is a bad thing. Most airplane wings built between the 1970s and the 1990s had tabs across the top which were alternatingly canted slightly left or right to the flow over the wing. Their function was induce turbulent flow along the wing.

    Quite right.

    True... However, attempting to achieve laminar flow on golfball would require a different shape. Turbulent flow on a round object is indeed the best way to achieve the longest flight (min drag/distance). But laminar flow, if it can be achieved, is indeed the best way to achieve the min drag/distance for a wing, a fact embodied by the lack of such tabs on top of more modern wings.

    Another thing about the tabs - they''re also positioned only along the interior of span of the wing, which also sports a chord with a steeper angle of attack. Thus, the secondary purpose of the tabs for slow flight was to ensure stalls developed near the root of the wing, rather than the tips, which contain the ailerons. This helps to do two things. First, it allows the stall to propogate slowly from the root towards the wingtips, giving the pilot time to recognize and recover from the stall. Second, it helps pilots to maintain control of the aircraft while in the beginning part of the stall.

    These days, wing design philosophy is a bit different, in large part because supercomputers have given us much better insight into various airfoil designs and the characteristics of laminar flow in general, not to mention spanwise flow, upon which we capitalize by means of winglets. In the meantime, controlling stall characteristics is done by means of varying airfoils between the root and the tip, along with warp (twist, ie angle of attack between the root and the tip). Most airfoils in general use today simply morph from one at the root to another at the tip. But for both lifting bodies as well as high-performance airfoils where every nuance of energy must be exacted from the wing during its primary profile, multiple airfoils can be used, along with differing rates between them.

    This is true for level flight and climb, but during descent they generate thrust just fine. :)
  8. Aug 25, 2010 #7

    And laminar flow drag (it takes energy to accelerate laminar flow, as well).

    Let's think span-wise flow, instead, which is somewhat mitigated in most aircraft today by winglets.


    Your definition of "turbulence" and "separation" are more than related - they're aspects of the same thing.

    For wings, yes. For fuselage...

    ...yes. It's all boundary layer, and attempting to minimize boundary layer separation.

    Have you ever seen either the wake or the boundary layer separation turbulence trailing behind a supertanker steaming 18 kts out of port? I have - a child could swim in it. We liked that problem in the early 80s (courtesy of NASA computers in Slidell), and the airflow boundary layer issues soon followed (a bit more computationally difficult, given they were somewhat compressible, even at low velocities).

    Ever since, whether wings, fuselage, or lifting body, we've treated the results much the same. The only reason we differentiate at all is because structurally, it's still far less complex and less costly to build a tube with wings than it is to build a lifting body. The Dreamliner? Tube with wings. The only one's who're doing lifting bodies are those who need it for the stealth purposes (B-2, F-117). But the additional manufacturing costs are justified by the increased stealth. Hey, even the non-stealthy drones are tubes with wings. The stealthy drones which look very much like the B-2, revealed in prototype at Farnsworth in 2006 and spotted in Iraq in 2009 are black ops for a reason.

    Then you have the microdrones... :) Personally, I think these little puppies are the wave of the future.
  9. Aug 28, 2010 #8
    Hey mugaliens, have not seen you in a while. :smile:

    I'm not sure what part of my standard description of turbulence made you question it, so I can only ask that you reread it again. Also, there is a very important difference between turbulence and separation. As you well know, turbulence is deviations of the flow as it goes along a streamline. On the other hand, separation occurs when the flow separates from a body. You can have turbulence inside an attached boundary layer, that is the distinction I was trying to show.
  10. Aug 31, 2010 #9


    User Avatar
    Science Advisor
    Gold Member

    I would say that mugaliens probably questions your definition of turbulence because it wasn't really correct. If you use your definition, "deviation in the flow from a straight streamline," then you could say that any time a flow encounters a body, it is turbulent because it no longer follows a straight streamline, but rather a curved one. The definition of turbulence that you gave in this more recent post, "turbulence is deviations of the flow as it goes along a streamline," is better.

    Still, no definition of turbulence is really correct without giving a nod to its chaotic, stochastic nature characterized by increasingly tiny eddies that dissipate the flow's energy. Streamlines have little or no meaning once the flow is turbulent since the flow no longer follows them and instead intermixes quite a bit (thus the reason for the greater potential for thermal convection in turbulent flows).

    That isn't correct either. Transition does not occur at a fixed point, but rather over a fairly roughly defined region. At a certain point in the flow where instabilities in the laminar boundary layer reach a critical magnitude, you start seeing small turbulent pockets. These pockets grow and combine until eventually the entire boundary layer is turbulent. The region between the point where the pockets originally form and where the flow is fully turbulent is usually called the transition region and can vary in size greatly with flow conditions.
  11. Aug 31, 2010 #10
    Ah! My misunderstanding, then!

    Good to see you around here, as well. :)
  12. Aug 31, 2010 #11
    Hey boneh3ad, thanks for the reply and welcome to PF. Nice to have more Aero people around! You are right, my answer was from a basic Young, Okiishi and Munson text book. I'm a dynamics/control guy, not Aero/CFD, so I only know enough to be dangerous. I think there is a "q criterion plots" that show you the unsteadiness of the flow, but I could, and probably am, remembering that incorrectly.
  13. Aug 31, 2010 #12


    User Avatar
    Science Advisor
    Gold Member

    I suppose I could grab Okiishi off of my bookshelf and find the plot you are speaking of just for grins, but I officially shut my brain off right around 90 minutes ago.

    At any rate, thanks for the welcome. I have browsed through here on occasion in the past but never really bothered to register, but then I saw a thread on transition over a wing and just had to drop in my two cents. I am in a research group whose primary focus is boundary layer transition, so how could I resist? Unfortunately, that means I have now nearly expended all of my current knowledge on turbulence, as the bulk of my work is done to prevent it from ever occurring. :wink:
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Similar Discussions: Wing transition to turbulent flow