How do I run XFOIL?

[Norman]

Norman here on the boards is REALLY good at XFOIL.

I've been using computers since DOS 3.2 (actually 1979 if you count that bad day I had with a UNIX dumb terminal in highschool). I'm much older now and those command line interfaces, which I always hated, just put me to sleep now. I use Profili for airfoil Analysis and design. It doesn't let me use all of the advanced design features of Xfoil but I'm a hobbyist and exercising my brain, instead of letting the software do all the heavy lifting, is part of the fun. If you really must have the full functionality of Xfoil, with a GUI, get XFLR5 or QFLR5. They both contain all the Xfoil program code plus AVA all wrapped up in a graphical interface.

FLZ Vortex also looks interesting. By default the interface is in German but there is an option under the "Sprache/Language" menu to switch to English.

The documentation with all of these programs say that they're not for full scale use. My personal opinion is that this is just the standard legal disclaimer that comes with all software. It's obviously not as high quality as some other programs but unless you have the budget of Boeing you should never consider software output an actual design until you've done some other form of verification (like build a test piece and actually test it). Comparisons with tunnel results shows the airfoil data from software is ususly good within the linear range but may be optimistic at the extremes ie CLmax is sometimes too high and CDmin is sometimes too low (these errors can be as high as 10%). Also PC generated post stall can be a complete crap shoot.

One thing I like to point out is that this software is relatively new whereas the code that Harry Riblett used to create his airfoil sections was written in the early 80s and if it didn't have the same disclaimer it should have. So, is it irresponsible to use free or cheep software? Only if you go directly from the screen to the plane with someone's butt in it.

[HumanPoweredDesigner]

The main reason I want to run the simulations is to get data for the range of Re across the span of the prop, most of which are above the 100,000 in the data base. And a lot of airfoils in the databases have jagged surfaces, which makes me wonder if it messed up their L and D results. And I could develop intuition faster if I could alter parts of airfoils without changing the rest of it.

[Norman]

Xfoil works fine at Re=100,000. The problem with "smoothnes" is that in different contexts it means different things. Xfoil needs a lot of coordinates for its calculations. If there aren't enough points (especially around the LE) the polars will be spikey. There is a <smooth> command on the menu that will divide up the spline into more points so that Xfoil can converge on more answers thus producing smoother polars, usually. If the spline wasn't smooth in the first place all <smooth> will do is produce more data for your lumpy airfoil. If that's the case you have two options that I know of:

First you can zoom in really close in the graphical editing screen (oops, that's in XFLR5 and Profili) and manually move the coordinats around until you've gotten rid of the bumps, valleys and flat spots. Then run <smooth> and check the results in the graphical editor. Sometimes just deleting coords that you're not sure of works.

Second is <smooth Qspec> If you use this one you're making a new airfoil because the resulting shape will be completely deferment so don't save it with the same name

[adawy1985]

The main reason I want to run the simulations is to get data for the range of Re across the span of the prop, most of which are above the 100,000 in the data base. And a lot of airfoils in the databases have jagged surfaces, which makes me wonder if it messed up their L and D results. And I could develop intuition faster if I could alter parts of airfoils without changing the rest of it.

Dear HumanPoweredDesigner,

See this tutorial.
I think you will like it.
http://cobweb.ecn.purdue.edu/~aae333/XFOIL/Tutorial/Tutorial%20for%20XFoil.htm
It's perfect for a start...

I was totally frustrated and it made me very comfortable with it.

Ahmed