# Separation bubble

Discussion in 'Sailboats' started by Fidippide, Sep 15, 2009.

1. Joined: Sep 2009
Posts: 3
Likes: 0, Points: 0, Legacy Rep: 10
Location: Italy

### FidippideNew Member

How can I display separation bubble with XFOIL on the surface of a NACA simmetric 4 digit profile? Thanks.

2. Joined: Feb 2002
Posts: 2,319
Likes: 295, Points: 83, Legacy Rep: 1673
Location: Port Gamble, Washington, USA

### tspeerSenior Member

The boundary layer calculation in XFOIL is an integral method, so it doesn't actually calculate the details of the flow in a separation bubble the way, say, a Navier Stokes code would do. Instead, it calculates parameters that determine the overall characteristics of the boundary layer.

You can often spot the separation bubble in the pressure distribution. It will be characterized by a nearly flat spot that extends out from the inviscid distribution, then rapidly drops to the inviscid distribution after transition occurs within the bubble. Here's an example:

From the OPER menu, you can go to the VPLO menu and plot the skin friction. The skin friction will be negative or zero in the region of separated flow. The shape factor (H) plot can also be useful.

The BL command from the OPER menu will plot boundary layer velocity profiles. Or use the BLC to plot a specific velocity profile on the airfoil of the CPX plot. Look for profiles with inflection points.

3. Joined: Oct 2007
Posts: 2,936
Likes: 148, Points: 63, Legacy Rep: 1593
Location: Arlington, WA-USA

### PetrosSenior Member

Many years ago when I worked in aerodynamics with the early 2-D and 3-D main frame programs, we would create a foil shape by combining the foil with the attached separation bubble, estimated from the stream line data and the method Tom outlined above. this gave strange, really thick sections, but it worked.

The program would build the boundary layer over the outside of the combined separation bubble and airfoil shape to approximate the actual airflow. It gave fairly good approximations of the lift and moment coefficients, the drag was never very accurate. If I recall, X-foil is a fairly simple program that might work the same way.

4. Joined: May 2004
Posts: 5,372
Likes: 255, Points: 93, Legacy Rep: 3380
Location: Italy (Garda Lake) and Croatia (Istria)

### daiquiriEngineering and Design

It sounds to me like you have created Stratford pressure distribution over the airfoil. If so, the final result you had obtained must have been something like Liebeck's airfoil shape?

#### Attached Files:

• ###### Liebeck.gif
File size:
13.8 KB
Views:
2,021
5. Joined: Oct 2007
Posts: 2,936
Likes: 148, Points: 63, Legacy Rep: 1593
Location: Arlington, WA-USA

### PetrosSenior Member

That looks about right, but there was a think surface trailing way behind the TE from the separated flow causing a very thick boundary layer.

6. Joined: Feb 2002
Posts: 2,319
Likes: 295, Points: 83, Legacy Rep: 1673
Location: Port Gamble, Washington, USA

### tspeerSenior Member

XFOIL has an inviscid panel code coupled with an inverse integral boundary layer method. The boundary layer method specifies the skin friction at the surface and calculates the resulting pressures. The boundary layer and the outer flow solution are iterated together to converge on the solution. The boundary layer method is basically the same as used in the ISES & MSES codes. It is capable of handling limited amounts of separation, where direct boundary layer methods (that calculate the skin friction given the pressures) cannot. This is a much more sophisticated capability than the earlier methods of loosely coupling the boundary layer and outer flow solutions.

ISES/MSES use an Euler method for the outer flowfield instead of potential flow, which gives them the ability to handle compressible flows better. But at low speeds, XFOIL is state of the art.

7. Joined: May 2004
Posts: 5,372
Likes: 255, Points: 93, Legacy Rep: 3380
Location: Italy (Garda Lake) and Croatia (Istria)

### daiquiriEngineering and Design

Can someone please tell me if there is a GUI version of XFOIL available?
Also, do you know if any research has been done in order to verify the degree of error given by XFOIL when treating flows separated at leading/trailing edge, and flows with separation bubbles?

8. Joined: Feb 2002
Posts: 2,319
Likes: 295, Points: 83, Legacy Rep: 1673
Location: Port Gamble, Washington, USA

### tspeerSenior Member

9. Joined: Feb 2007
Posts: 547
Likes: 166, Points: 53, Legacy Rep: 160
Location: California, USA

### Doug HalseySenior Member

I've recently started using Xfoil & am interested in seeing validation cases for foils in the Reynolds number range 500,000 to 2,000,000. The previous post is 5 years old now & none of the links are still active. Can anyone recommend anything current?

10. Joined: Jul 2012
Posts: 1,358
Likes: 92, Points: 48, Legacy Rep: 152
Location: United States

### SkyakSenior Member

Doug,
I am not sure what your background or needs are but it sounds like your request needs some work. There is nothing in your question that indicates Xfoil would be a bad tool, but it sounds like you are looking for something that will tell you how accurate it is when the truth is that xfoil is a tool who's value depends on the skill of the user.

Xfoil uses the panel method -which is simply proven mathematically. The error and estimation comes from the boundary layer development, and this is dependent on the case and conditions. If you just want case studies google
"xfoil boundary layer thickness estimation"
This is where you will find help applying xfoil to your real case -which is not an infinite 2D foil section in an infinite steady parallel flow. The panel method and Xflow are simple enough to be understood and computationally efficient so it is the first tool you use to develop your profile.

To understand the theory behind aerodynamic modeling I would recommend you take one of the (MIT) beginning aerodynamics classes available on EdX.

11. Joined: Feb 2007
Posts: 547
Likes: 166, Points: 53, Legacy Rep: 160
Location: California, USA

### Doug HalseySenior Member

Skyak : I ask for some references for code validation & you suggest I should take a beginning course in aerodynamics? Give me a break !

12. Joined: Jul 2012
Posts: 1,358
Likes: 92, Points: 48, Legacy Rep: 152
Location: United States

Doug,