Separation occurs theoretically at the point where Cf is zero. So, if you have a cf(x) graph, you can find this point easily.

However, it is not so easy to find it from a Cp(x) graph. A separation may be expected normally after a minimum pressure point along a surface, but it is not certain, and if there is a separation you can not predict the separation point exactly from the Cp(x) graph, instead you must calculate the boundary layer and find the point where Cf is zero

I learn that (dp/dx) will be positive and large for the flow is without separation. But if its stated vice-versa its true ?

You can divide the pressure distribution around a body such as an airfoil into two region beginning from the leading edge.

In the first region near the leading edge the pressure begin with a maximum value at the stagnation point then decrease up to a minimum value, while the airflow velocity increase along the surface. This is the favorable pressure gradient region (dp/dx0). In this region there is a risk of separation if the pressure gradient is high enough. After a certain point on the cp(x) graph in this adverse pressure gradient region if there is a constant pressure region, this region may be an indication of separation. However this is not so clear. Therefore it is better to calculate the boundary layer on the surface, then evaluate if there is a point where Cf is near zero.

You have to identify point corresponding to the condition du/dy=0. This comes from shear stress value being zero at that point. As mentioned earlier it is easy to find on cf(x) curve but not on cp(x) curve.

• From my analysis I learn that the Flow separation starts when dp/dx= 0.( in other words.. look for the straight line parallel to X axis)
• And the flow separated when Cf=0

When you see Cf(x) equal to zero value, the C/X value show the separation point, and when the pressure gradient (dp/dx) is positive indicate capable region for separation phenomena. so in this region static pressure also is uniformly constant because we stay on wake zone. in physical view at wake zone mixing is very high, so state values is uniform. 

On the suction side of an airfoil, the flow separates if the pressure rise is too high and the outer streamline of the separation bubble adapts such that pressure stays approximately constant. Since within the bubble pressure normal to the wall is fairly constant and you have a reverse flow near the wall, you will have cf=0 at the separation point (where tangential velocity near to wall is going from positive to negative) and cp ~ constant directly downstream of the separation.

The flow separates where Cf=0.

Hi Roham,

I am conducting an experiment to find out flow separation points. Initially, I tried to find flow separation by using pressure tappings and not so happy with the results.

In the second go, I am thinking to do traverse with Hotwire just before the cylinder(with in 0-2mm) to the mid point( I can only do traverse to the middle of the cylinder with out disturbing the flow). So, that I can record the velocity(y).

I am also interested to find out potential velocity from the above experiment. I was searching for way to do it and I come across your answer.

Can you please explain little more about how to find flow separation from velocity(y) information? any suggestions?

Thanks & regards

Kiran Chalasani