Laminar vs attached airflow

[raymondu999]

Someone was preaching to me recently that attached airflow is basically laminar airflow (ie they're both the same thing). As far as I understand it - he was wrong. And that you can have attached but turbulent airflow, ie not laminar, but he spoke with such confidence that I was unsure of my own proposition. Could someone help me clarify?

[NoDivergence]

Your conceptualization is correct. Flow does not stay laminar on a surface for very long due to the viscous effects with the surface where there is considered to be no slip of the air. The flow becomes turbulent due to the shear. You are correct in the fact that the turbulent boundary layer actually delays the separation of the flow from the surface. However, if the adverse pressure gradient is high enough, the flow inverses upon itself and becomes completely separated.

[No Lotus]

Yes, they are distinct. Turbulators, for example, serve to trip the boundary layer from laminar to turbulent for the purpose of keeping the flow attached.

[Ogami musashi]

Strictly speaking laminar flows are flows with no vertical or transversal velocity component but only longitudinal component in a layer i.e all layers flow parallel to each other (but not module, that is they can have different speeds due to shear stresses).

Turbulent flow is thus when the local velocity have compenents in both vertical/transversal and longitudinal directions.

Separated flow is when the boundary layer velocity is reversed.

As you can see there's no correlation between the two.

[KeithYoung]

You are correct, the guy that disagreed with you is clearly mistaken.

[Blanchimont]

I think the most common example is the golf ball with and without dimples, which change the roughness of the surface.

This comparison shows that both the laminar and the turbolent flow around the golf ball separate from the surface, but at different angles. In case of the golf ball, the turbolent flow is attached for a longer time and therefore reduces the drag.

Search the forum for golf ball aerodynamics, it was discussed several time.

[riff_raff]

I think the most common example is the golf ball with and without dimples, which change the roughness of the surface.This comparison shows that both the laminar and the turbolent flow around the golf ball separate from the surface, but at different angles. In case of the golf ball, the turbolent flow is attached for a longer time and therefore reduces the drag.

I'm not an aero expert, but I don't believe your explanation of why the dimples on the surface of a golf ball are effective is entirely correct. It is the combined effect of the dimpled surface and reverse spin about the ball's pitch axis that creates both reduced aerodynamic drag and increased aerodynamic lift. Thus if the ball did not have reverse spin about the pitch axis when hit, the surface dimples would not help much.

[Kiril Varbanov]

Someone was preaching to me recently that attached airflow is basically laminar airflow (ie they're both the same thing). As far as I understand it - he was wrong. And that you can have attached but turbulent airflow, ie not laminar, but he spoke with such confidence that I was unsure of my own proposition. Could someone help me clarify?

This generalization is rather misleading, but it's a good topic. Flow cannot stay attached for enormous amount of time because of the viscous effect, as @NoDivergence said.
The flow reversal is essentially caused by an adverse pressure gradient imposed on the boundary layer by the outer potential flow. Then, in a perfect world, we should be able to calculate the Reynolds number of the local flow conditions and thus determine whether the boundary layer will be laminar or turbulent.

[Blanchimont]

I'm not an aero expert, but I don't believe your explanation of why the dimples on the surface of a golf ball are effective is entirely correct. It is the combined effect of the dimpled surface and reverse spin about the ball's pitch axis that creates both reduced aerodynamic drag and increased aerodynamic lift. Thus if the ball did not have reverse spin about the pitch axis when hit, the surface dimples would not help much.

I do not know what the effect of spin on the distance achieved by the golf ball is, as i have only little experience in golfing and didn't look this topic up. But i imagine it definitely has an effect on the curve the ball takes once it was hit.

I can say i'm sure my explanation of the effect of the rougher surface of the dimpled ball is correct. The rougher surface leeds to a more turbolent flow around the ball and therefore the flow can stay attached at the surface for a bigger area and reduces the wake as a result.

The following picture from http://en.wikipedia.org/wiki/Golf_ball (origin http://www.grc.nasa.gov/WWW/k-12/airpla ... phere.html)
shows the cd number of two cylinders, a smooth one and a rough one(dimpled). You can see the cd number isn't constant over the different speeds, it changes from 1.5 to 0.1, depending on roughness and speed (expressed by Reynolds number).
At a certain Re number (10^5 for rough cylinder, 3*10^5 for the smooth cylinder) there's a drop in cd. The drop for the rough cylinder appears at a lower speed compared to the smooth surface, between Re 3*10^5 and 10^7 the cd number of the smooth surface is smaller. If our cylinder (ball) reaches a top speed close to Re 3*10^5, then the cylinder (ball) will reach larger distances than the smooth cylinder. If top speed is in the region of Re 10^6, the smooth cylinder (ball) will be the better choice.
This is the same for a smooth golf ball or a dimpled one, although the graph can differ from the one shown here in case of where the drop in cd takes place.

From the wikipedia article we can also read that the
"diameter of the golf ball cannot be any smaller than 1.680 inches (42.67 mm). The maximum velocity of the ball may not exceed 250 feet per second (274 km/h) under test conditions"

This gives an Re number of Re = 42,67/1000 m * (274/3,6) m/s / 15,3*10^-6 m^2/s = 2,12*10^5. The real golf ball should be designed for this Reynolds number to achieve long distances.

[Tommy Cookers]

the above rather 'bluff body' case usefully shows the dramatic Re number effects (over a big Re range)
but is unrepresentative of the case more valuable to us
that is, drag and lift/DF of the car-shaped and car-sized, or wings on these

our cars (and wings) operate in region 4, where the drag and lift coefficients are constant (relative to 'V squared')
(except historics, where eg the low Re of wire wheel spokes and gauze intake protection etc causes some higher region 2 drag)

there is little laminar flow or detached flow on aircraft because of their form (and size)
only active boundary layer control (suction or blowing) can have substantial effects on drag and lift

the practical form of cars makes detached flow inevitable, so surface effects have almost no value
the practical size of cars keeps them in region 4

BTW if dimpling was the 'magic beans', wouldn't propellor manufacturers be the first to use it ?

[olefud]

Someone was preaching to me recently that attached airflow is basically laminar airflow (ie they're both the same thing). As far as I understand it - he was wrong. And that you can have attached but turbulent airflow, ie not laminar, but he spoke with such confidence that I was unsure of my own proposition. Could someone help me clarify?

It’s a little bit true. Attached flow results from a fluid flow over a stationary solid surface. Initially, the fluid in contact with the surface is attached by surface roughness, electro, Vander Waal’s forces etc. Between the surface and the free fluid flow the fluid shears into layers or laminations of progressively greater flow velocity until the free-flow velocity is reached. Thus laminar flow is always attached flow.

However, as the shear forces between the lamination exceed the cohesion there between, the layers become turbulent but still within the attached layer. But this condition tends to be somewhat unstable and is often the precursor to detachment.

[Raptor22]

spot on.

[Wicker Bill]

However, as the shear forces between the lamination exceed the cohesion there between, the layers become turbulent but still within the attached layer. But this condition tends to be somewhat unstable and is often the precursor to detachment.

Detachment in a laminar flow yes but if the boundary layer develops as fully turbulent it will become more stable and delay detachment.

[cwb]

When fluid flow along a surface encounters an adverse pressure gradient it will detach at some point, regardless of whether it is laminar or turbulent. The only difference between laminar and turbulent flow is that turbulent flow can resist an adverse pressure gradient for slightly longer.

An adverse pressure gradient exists on a bluff body when the cross sectional area starts decreasing in the downstream direction. Dimples on golf ball work because the make the flow over the ball more turbulent, enabling the flow on the downstream side to stay attached for slightly longer, resulting in a smaller wake (as has been stated a few times already above)

Drag reduction at a gross level is all about avoiding detached flow. This is why teams wont be fussed about whether they have a stepped or vanity nose. Because the nose is in the "bluff body cross sectional area increasing in the direction of flow zone" , it wont have a significant effect on the aero. As long as they ensure no detachment occurs in the transition to the bulkhead, the effect will probably be negligible.

[olefud]

Drag reduction at a gross level is all about avoiding detached flow.

For drag reduction this optimally means filling the "hole" left by the form which, as said,avoids detachment. Bu tif it does detach it needs to do so cleanly with complete separation from the surface, preferable with as little turbulance as practical.