How does Turbulence Intensity affect drag and flow separation?

[Ghalih]

Just to the point..

How Turbulence Intensity affect a drag? As far as I know if there is decrease in Turbulence Intensity, the coefficient of drag will decrease to, but I don't know why the reason behind that. Also, the relationship with flow separation too.

I hope someone can help me with that problem.

Thank you

 

[jack action]

The boundary layer is the part of the flow that «sticks» to an object moving through a fluid. When the boundary layer begins to detach itself from the object, it is called flow separation. This creates a low pressure zone (turbulent flow) behind the object that pulls the object backward (think vacuum).

Imagine an object moving through a fluid. Each time you move the object forward, it leaves an empty space behind that must be filled back by the surrounding fluid. This process takes some time to occur. The faster the object goes, the greater this zone will be and the more turbulent the flow will be as it tries filling the void.

But if you induce a tiny turbulent flow just before going behind the object, this tiny low pressure zone created pulls back the flow toward the object, therefore delaying the beginning of flow separation. The delayed flow separation reduces the size of the low pressure zone behind the object, thus decreasing the drag. The most popular example is the dimpled golf ball.

But @boneh3ad is usually the best at explaining this stuff.

 

[boneh3ad]

When you ask about "turbulence intensity", what exactly are you trying to describe? I don't really want to throw out an answer before I make sure I know what you are trying to ask. Typically, turbulence intensity is the Pythagorean sum of the rms fluctuations of each of the three components of velocity ($Tu = \sqrt{u^{\prime 2} + v^{\prime 2} + w^{\prime 2}}$). Is this the quantity you mean? And if so, do you mean turbulence intensity in the free stream? The boundary layer? The wake?

The general rule would be that as turbulence intensity increases, the drag will also increase, but the why for this rule will depend on exactly what you are asking.

 

[Ghalih]

When you ask about "turbulence intensity", what exactly are you trying to describe? I don't really want to throw out an answer before I make sure I know what you are trying to ask. Typically, turbulence intensity is the Pythagorean sum of the rms fluctuations of each of the three components of velocity ($Tu = \sqrt{u^{\prime 2} + v^{\prime 2} + w^{\prime 2}}$). Is this the quantity you mean? And if so, do you mean turbulence intensity in the free stream? The boundary layer? The wake?

The general rule would be that as turbulence intensity increases, the drag will also increase, but the why for this rule will depend on exactly what you are asking.

what i mean is turbulence intensity that state ratio of rms fluctuation to average velocity ($TI = u' / U$), it is the same with the equation above?. Also, the turbulence intensity that I mean is in the wake region, because right know I am studying about reduction of drag with active flow control (AFC). The data that I get are turbulence intensity in the condition without using AFC is 13%, but after using the AFC the turbulence intensity reduced to 12%.

I hope that make my question clearer

Thank You.

 

[Ghalih]

The boundary layer is the part of the flow that «sticks» to an object moving through a fluid. When the boundary layer begins to detach itself from the object, it is called flow separation. This creates a low pressure zone (turbulent flow) behind the object that pulls the object backward (think vacuum).

Imagine an object moving through a fluid. Each time you move the object forward, it leaves an empty space behind that must be filled back by the surrounding fluid. This process takes some time to occur. The faster the object goes, the greater this zone will be and the more turbulent the flow will be as it tries filling the void.

But if you induce a tiny turbulent flow just before going behind the object, this tiny low pressure zone created pulls back the flow toward the object, therefore delaying the beginning of flow separation. The delayed flow separation reduces the size of the low pressure zone behind the object, thus decreasing the drag. The most popular example is the dimpled golf ball.

But @boneh3ad is usually the best at explaining this stuff.

Thank you for the answer

 

[jack action]

I knew I might of needed help on this. Time to get the popcorn and learn something new!

 

[boneh3ad]

Out of curiosity, what is the shape of your object in the flow?

 

[Ghalih]

Out of curiosity, what is the shape of your object in the flow?

The object that I use is in a shape of car. I'm using model from Daihatsu Ayla car.

 

[boneh3ad]

The shape doesn't really matter but I was just curious.

Anyway, if you discount all of the other factors, as the turbulence intensity increases in the wake, you will tend to observe a higher drag on the body. Turbulent fluctuations tend to increase the transport of momentum through a fluid, and in this case, it means that the wake would be pulling more fluid along with it and the body is therefore having to drag more fluid along with it, thus higher drag. Of course, the difference between 12% and 13% is small and probably not going to be a very noticeable difference.

The thing is, though, that none of these things can ever truly be independent of other factors in the flow. In this case, the size and shape of the wake region is determined largely by the location(s) of boundary-layer separation on the body itself. These, points are, in turn, intimately related to the state of the boundary layer, and a turbulent boundary layer is going to remain attached to the surface longer than a laminar or transitional one. Of course, the state of the boundary layer is highly dependent on the level of turbulence in the free stream.

So, what does it all mean? A higher free-stream turbulence level is going to mean that the boundary layers transition to turbulence sooner. This means that the viscous drag is going to be higher (sometimes substantially). It also means that boundary-layer separation is delayed, and the drag as a result of the wake is going to be smaller (often a larger effect than that of viscous drag). Of course, in the real world, you don't get to choose your free-stream fluctuation levels, but this is a real issue in wind tunnel experiments where the results are highly dependent on the state of the boundary layer.

 

[Ghalih]

The shape doesn't really matter but I was just curious.

Anyway, if you discount all of the other factors, as the turbulence intensity increases in the wake, you will tend to observe a higher drag on the body. Turbulent fluctuations tend to increase the transport of momentum through a fluid, and in this case, it means that the wake would be pulling more fluid along with it and the body is therefore having to drag more fluid along with it, thus higher drag. Of course, the difference between 12% and 13% is small and probably not going to be a very noticeable difference.

The thing is, though, that none of these things can ever truly be independent of other factors in the flow. In this case, the size and shape of the wake region is determined largely by the location(s) of boundary-layer separation on the body itself. These, points are, in turn, intimately related to the state of the boundary layer, and a turbulent boundary layer is going to remain attached to the surface longer than a laminar or transitional one. Of course, the state of the boundary layer is highly dependent on the level of turbulence in the free stream.

So, what does it all mean? A higher free-stream turbulence level is going to mean that the boundary layers transition to turbulence sooner. This means that the viscous drag is going to be higher (sometimes substantially). It also means that boundary-layer separation is delayed, and the drag as a result of the wake is going to be smaller (often a larger effect than that of viscous drag). Of course, in the real world, you don't get to choose your free-stream fluctuation levels, but this is a real issue in wind tunnel experiments where the results are highly dependent on the state of the boundary layer.

I see... Thank you for the answer sir...