Upper Surface Transition at Negative Angles of Attack

[Thusithatck]

I'm trying to validate some experimental data on a NACA 643-418 wing section from XFoil. Below are some basis data,

Number of panels around the airfoil - 43
Mach number - 0.0487
Reynold number - 1107872
Velocity - 16.67 (not important)

During the analysis I have set the upper surface transition (tripped) location as 0.1 and free transition at lower surface for positive angles of attack.

But when it comes to negative angles of attack, is it acceptable to have upper surface transition position as 0.1 ??

 

[boneh3ad]

While I don't know the specifics of the NACA 643-418 off the top of my head, even at positive alpha having a transition point of 0.1 seems extremely conservative. At negative angles of attack on many airfoils, there is often no transition, especially at such low Re.

Perhaps I am just a bit confused by your question, but in my experience, the natural transition point on the upper surface of similar 6-series airfoils seems to be as far back as 80% in quiet wind tunnels with small negative alpha.

 

[RandomGuy88]

Why are you only using 43 panels? When I use X-Foil, which I admit is not often, I usually use over 200 panels.

I know that doesn't answer your question, but I was just curious.

 

[Thusithatck]

Why are you only using 43 panels? When I use X-Foil, which I admit is not often, I usually use over 200 panels.

I know that doesn't answer your question, but I was just curious.

I have used 43 panels as I tried to replicate experimental data. 43 panels gave me the closest fit to the experimental data. I did tried using 200 panels, but it over predicted massively.

 

[boneh3ad]

I have used 43 panels as I tried to replicate experimental data. 43 panels gave me the closest fit to the experimental data. I did tried using 200 panels, but it over predicted massively.

That just tells me something is wrong. You can't make your simulation "dumber" to match the data. Instead, that means something about your initial conditions, boundary conditions or assumptions was invalid.

 

[Thusithatck]

That just tells me something is wrong. You can't make your simulation "dumber" to match the data. Instead, that means something about your initial conditions, boundary conditions or assumptions was invalid.

I was told to vary the number of panels around the airfoil to get a good agreement with the experimental data. Where the panels in XFoil are similar to those cells in Fluent.

Free transition, and input parameters (Re, and M) are correct. Therefore the initial conditions would be valid. Only the Reynolds number has confused me. Where the XFoil Reynolds number is calculated using freestream velocity, density and viscosity and an implied unit chord.

My Reynolds number is 380,000 (based on the chord length which is 0.343m). I am confused whether to input the Reynolds number in XFoil as 380,000 or 1107872 (implied unit chord obtained by dividing 380,000 by 0.343). The input coordinate airfoil has a unit chord.

If my approach is wrong please correct me. Especially the Reynolds number situation.

 

[boneh3ad]

Your Reynolds number should remain constant. It is the proper scaling term. In other words, your free stream velocity in XFOIL will differ from experiments.