Reynolds number and JAVAfoil

[daiquiri]

Perfect explanation Mr. Speer, as always, in all its parts. Thanks.
It is always important to know both strong points and weak points of any analysis tool and what can/cannot be expected from it.

May I ask you please to tell us few words about the XFOIL's (or other IBL-method software) accuracy of prediction of airfoil moment coefficients, particularily when some amount of separation is present? Does it match the accuracy of force coefficients, as far as you could verify?

And, since you have mentioned Profili as a GUI version of XFOIL, I'd like to point you out to another, imho well-done, implementation of XFOIL. It is called QFLR5, is distributed under GNU licence and it's home page is http://xflr5.sourceforge.net/xflr5.htm
It has an additional (and very useful) built-in module, which can analyse wings and wing-tail-body combinations with either non-linear lifting line method, a Vortex Lattice Method, or a 3D panel method.

[Guest625101138]

Originally Posted by tspeer

I.....
If you are working at low Reynolds numbers with single element sections, it would be worthwhile to learn to use XFOIL (or the graphical user interface Profili). The reason is XFOIL is the only inexpensive program I know that properly handles laminar separation bubbles and small amounts of turbulent separation.

.......

Tom
I will make the effort to get to know Xfoil to explore the limits of JavaFoil around my conditions of interest.

Rick

[peterraymond]

I've been messing with foil modification at the design tab in JavaFoil. It seems easy to use, once I figured it out, but maybe doesn't do everything I was hoping.

I was able to take a 2 element combination and modify element 1, while the software kept track of the influence of the second element. It's also pretty easy to grab the points and drag them. I typically set the drop down for symmetric Cp modification to keep the foil symmetric. It's interesting that with non-zero angle of attack, as you drag control points on one side of the foil it also moves them on the other. If you knew what you were doing, you could modify a single element, duplicate it to form element 2 and then further modify element 1. Not direct, but possible.

I wanted to particularly play with the ends of the elements. At high lift coefficients its hard to get the forward stagnation point on the elements to be at the leading edge, so I figured I'd try to make the radius more uniform there and accept higher drag. Also, stall at the trailing edge seems common and it can start on either surface, so I figured I could maybe play around there too.

The control points are close together and hard to separate at the foil ends and with the "smooth Target Cp" button clicked, sometimes moving a point seems to add waves to the curve. If you move points one at a time it's easier to get the points where you think you want them, but you don't maintain symmetry. Actually, as I played with asymmetric point by point modifications, the 1st element seemed to maintain more thickness towards the trailing edge and to begin to form a channel around the leading edge of the flap. Not unlike some of the high lift multi-element wings you see on race cars sometimes.

As you click the design button it's not uncommon for the Cp's at the trailing edge to start diverging in spikes. The convergence process seems to start near the middle of the foil and move towards the ends and as the part of the design curve that matches the goal curve moves out, the variation in the unconverged portion increases. However, reducing the relaxation from 10% to 5% seems to help this.

I've gone through the process several times, but haven't created anything of value. I tried to design a mast and wing at 50% cord each, working at a 20 degree flap angle and at a reasonably high CL, but didn't manage to come up with one that wasn't stalled somewhere. Getting rid of stall at the trailing edge of element 2 seems hardest. There are big velocity spikes at the the leading edge of the flap in opposite directions on the two faces, which makes it seem like it has to be harder for the velocities to converge at the trailing edge.

I also notice that JavaFoil shows both velocities spiking downward towards zero at the trailing edge of every element. This seems like an aberration, as if JavaFoil was treating the trailing edge as tightly radiused, not sharp.