Drag is the result of air ‘pushing’ the opposite direction of flight, if you want to think of it that way. The skin friction drag is made from the air trying to pull the wing opposite of the flight direction. My question is, why would turbulent flow change this?

Say you somehow have air on a wing that is going exactly 90 degrees to the direction of flight. Would this create drag for the wing, acting opposite of the flight direction? I would think not because the air is pushing the wing sideways not backward.

Basically turbulent air is air moving in random directions. So say it randomly moves left (spanwise) on a wing. Why would that make drag? Or if it moved right?


4 Answers 4


Short answer:

turbulent flow causes vigorous mixing of the flow nearest the wing with flow further away from the wing. this tends to share the energy imparted from the wing to the airflow nearest the wing with airflow further away from the wing. This speeds up the air further from the wing while slowing down the airflow nearest the wing and this effect increases the drag asserted on the wing by the airflow nearest it.

  • $\begingroup$ Ah okay, thanks a lot! This is exactly what I was thinking happened. $\endgroup$
    – Wyatt
    Mar 30 at 2:43
  • $\begingroup$ Yes, this is the right simple explanation. The turbulent mixing brings much more bigger momentum fluxes from the wing surface. $\endgroup$ Mar 30 at 12:29
  • 2
    $\begingroup$ Vladimir F, are you a slavic hero? $\endgroup$ Mar 31 at 17:17
  1. Stop thinking about the direction of the velocity as a force pushing on a wing. Velocity is not a force. Stop it, you're making yourself confused.

  2. Turbulence has different meanings.

Normal people (not engineers) think of turbulence as 'rough air' that an airplane encounters in flight. It can cause the aircraft to bump around violently and is scary and potentially dangerous.

This kind of turbulence is caused by updrafts and downdrafts from thermal activity, mountain ridges, and other weather and terrain issues. In a small aircraft, you can also encounter turbulence caused by a larger aircraft in front of you.

Engineers also think about turbulent flow vs. laminar flow. Here, turbulence is microscopic tiny random oscillations in the flow. They do not have a predominant net direction that can cause the aircraft to heave or jerk.

A turbulent boundary layer has fundamentally different characteristics from a laminar boundary layer.

It is almost like the fluid itself changes -- from air to water (not really, but the important part is that properties of the flow change and the differences we see is due to the changes in those properties. The changes are not due to some bulk velocity or major thing).

Of first importance, at the same Reynolds number, the skin friction coefficient of a turbulent boundary layer is higher than of a laminar boundary layer. Note this is on a log scale.

enter image description here

At Re=5e5, the turbulent skin friction coefficient is about 3x higher than laminar.

There are other situations where turbulent flow is good.

Turbulent flow also has a much higher heat transfer coefficient than laminar flow. So, if you're designing a heat sink or radiator to pull heat out of something, you would much rather have turbulent flow.

Also, turbulent flow is less likely to separate than laminar flow. So a golf ball uses dimples to force transition to turbulence and therefore keep the flow attached to reduce the separation drag on a golf ball and allow it to fly much further. Separation drag is almost always a bigger deal than skin friction drag. Only worry about skin friction drag after you've taken care of separation drag.

  • $\begingroup$ I see. So is there an acknowledged reason why the properties of turbulent flow change so much, like from air to water? $\endgroup$
    – Wyatt
    Mar 29 at 20:09
  • $\begingroup$ Sorry if I missed something obvious. $\endgroup$
    – Wyatt
    Mar 29 at 21:09
  • $\begingroup$ Recall how we've talked about how a gas is a bunch of random particles bouncing around. That behavior helps establish certain properties of the gas (perfect gas law P=rho R T, speed of sound a=sqrt(gamma R T), etc.). However, at larger length scales, we don't really need to worry about the tiny particles, we instead treat the air as a continuum. Laminar vs. turbulence is a similar thing, but at an intermediate length scale. Large compared to the size of molecules, small compared to the size of airplanes. $\endgroup$ Mar 29 at 21:13
  • $\begingroup$ At the length scale of laminar vs. turbulent flow, we either have nice orderly flow that is all parallel to the velocity of the net flow. Or, we have random little eddies that are chaotic and random -- but, their average velocity is still going with the flow. Imagine busy highway traffic -- on the one hand, you have everyone staying in their lane, going the same speed, generally well behaved. On the other hand, you have cars all going random speeds, changing lanes randomly, total chaos -- but the average traffic flow is the same. $\endgroup$ Mar 29 at 21:16
  • $\begingroup$ I think the change in heat transfer is the easiest to understand. In laminar flow, the wall heats the first layer, but each orderly layer must heat the next layer. Air is a pretty good insulator, so there is only so much heat that can get pulled out this way. However, in turbulent flow, particles near the body pick up heat, but then they move away from the body, where they transfer heat to other particles, or mix with those particles. Because of this, turbulent flow is a much better conductor of heat. $\endgroup$ Mar 29 at 21:19

Let's slightly change your concept and reframe your thinking, if that is ok. Think of drag as the wing dragging air in the direction of flight. Now think of the process by which that would happen. Momentum from the aircraft is imparted to that air. For the air that has passed over the wing, we can actually measure that momentum.

Ok... Now let's rewind for just a second and look at the drag associated with flight. To lift the aircraft, we have to expend energy, develop power, and generate thrust. This drag is associated with lift. Just set that aside; we are not talking about that kind of drag.

We are talking about drag associated with the boundary layer that develops in the flow of air over the surface of the wing. For the ideal conditions generally discussed in conceptualizing flight, we think of the aircraft passing through its parcel of air. There is no wind, this air is still. Consequently, the wing is passing through still air. The turbulence produced by the wing passing through that still air does not move anywhere except by the momentum imparted to that air through the boundary layer on the wing of the aircraft passing through that air. In producing that turbulence, energy is transferred into the boundary layer through drag consequent to, or from the aircraft passing through that air. That drag happens through skin friction resulting from the movement of the turbulent flow over the surface of the wing, itself. Momentum is imparted to that turbulent flow, mostly as motion of the turbulent wake in the direction of the aircraft, but also by momentum in the sustained turbulence that is left behind. Let's think about this for a moment...

That turbulent boundary layer moves downstream over the surface of the wing, and through skin friction, energy is imparted into, and in development of, that turbulent boundary layer. Consequently, in that still air, we can actually see the movement of the turbulence that remains, as the turbulent wake left behind the wing moves gently in the direction of flight of the wing, itself. If we know the properties of that air, and the width and average velocity of the flow in the wake (remember, it has general velocity in the direction of flight), then we can determine the momentum of that air, and the power expended in producing its motion.

A common point of confusion is the conception that turbulence is the random motion of air moving in random directions. For the wing in flight, airflow over the wing, whether laminar or turbulent, moves in a direction opposite the direction of flight. There is no net motion to the left or right; all of the motion in a relative sense is downstream over the surface of the wing.$^1$ If we view the wing as passing through still air, the turbulence left behind in the wake is truly of random motion. Careful observation of the air in that wake will show general motion in the direction of flight, that motion determined by the energy imparted to it through momentum lost from the aircraft, itself. This can be measured.

How do we make these measurements? If our wing is in a wind tunnel, we can attach a scale to the wing and measure the force of drag. We can put a velocity-survey rake in the wake behind the wing and measure the velocity of the airflow in that wake. We can even make this measurement on a wing in flight. Nevertheless, you may say, the wing in a wind tunnel is still and the air is moving. No matter, conceptually everything is essentially the same as the wing in flight through still air.

Now let's consider something that seems totally strange yet is perfectly natural - an aspect of practice in aircraft and wing design, a fundamental principal that is applied in wing design. In this case, we consider the wing as still and air as moving over the wing surface. Turbulence forms in an adverse pressure gradient. This is a physical fact that cannot be changed. What we can change is the drag. The shape of the wing section can be adjusted so that air flowing over the surface through this adverse pressure gradient is more gently slowed. The consequence is to thicken the boundary layer and reduce skin friction. Further, less motion is imparted to the turbulent wake as the wing drags this flow in the direction of flight. This process can be determined in producing a turbulent boundary layer that has no skin friction whatsoever. B S Stratford was the first to demonstrate this more than 70 years ago in a wind-tunnel experiment where the shape and concavity of one wall of the tunnel could adjusted to produce a pressure gradient resulting in a turbulent boundary layer having zero skin friction. As mentioned elsewhere, my term for this is infinitely separating...

$^1$ In this case, as noted previously, we are not discussing lift induced flow which may be skewed to the left or right over the surface of the wing.

  • 1
    $\begingroup$ I make my living by simulating turbulence and teaching about it (teaching mostly in atmosphere, not mechanical engineering) and I have hard time understanding these paragraphs. If they are supposed to be understandable by a layperson, then these persons may have some hard time as well. $\endgroup$ Mar 30 at 12:27
  • $\begingroup$ @VladimirFГероямслава Noted, and also exactly understood. Taken to heart and will look into text modifications, simpler explanation, and examples (as time allows). $\endgroup$ Mar 31 at 20:56

One simple explanation is just math, though it is possible that this is already accounted for as part of what makes laminar flow a different regime.

You mentioned air moving in random directions, but that doesn't actually cancel out when you're moving against friction, because of the velocity squared term. If you're going 10 units on average, that makes for a factor of 100. But if one section is moving at 9 and another at 11, that makes for (121+81)/2 = 101 units of effective drag.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .