1
$\begingroup$

I'm projecting a wing in XFLR5. The lift at the tip of the wing looks like it's dropping very quickly close to the stall angle, so I think there may be wingtip stall, however, is there any way I can confirm that? I thought about comparing the lift coefficient of the wing with the maximum lift coefficient of the airfoil, and If the local lift coefficient is bigger than the maximum airfoil lift coefficient, I can conclude that region is stalling. Is this thought process correct, or is there any other analysis I could make? If so, which Reynolds number should I use to acquire the airfoil lift coefficient, should I just consider the aircraft speed in the formula?

Here's the image of the lift distribution, if it's useful having a look at it: Lift distribution

$\endgroup$
3
  • $\begingroup$ Is it possible to plot also the distribution of drag and pitching moment? $\endgroup$
    – sophit
    Feb 26 at 19:34
  • $\begingroup$ @sophit Sure. Here's the drag and the pitching moment distributions. If the wingtip is stalled then the drag should have a significant increase on the tip, which isn't the case here? Although there is a very small increase, does this say anything? And for the pitching moment, we should look for a high increase at the tip as well? Comparing with other wings, it seems the pitching moment difference between the root and the tip is higher in this wing, does this indicate anything? $\endgroup$
    – Luan Arita
    Feb 26 at 23:01
  • $\begingroup$ According to the documentation, xflr5 is a simple panels method solver i.e. it definitely cannot predict stall or any other viscous phenomenon. Maybe what you can do is get the plot of the lift coefficient vs. alpha for each airfoil and check if at the local Cl that you get from xflr5 the relevant airfoil is stalled or not. $\endgroup$
    – sophit
    Feb 27 at 6:34

2 Answers 2

2
$\begingroup$

Yes, as plotted the wing will stall at the tips first. Consider to reduce the taper ratio (i.e., make the tip chord larger).

To predict stall, angle of attack, the rate of angle attack increase, wing sweep, wing aspect ratio, Mach number and Reynolds number all have an influence. For a single wing Mach number and rate of increase are the same, but taper will reduce the Reynolds number at the tips, so stall will occur at a slightly lower angle of attack there than at mid-chord. Therefore, your local lift coefficient should have at least a stall margin of 20% at the tip when it reaches its 2D stall value at mid-span.

Increase this margin if you have ailerons and especially if there are flaps at mid-span. Slotted flaps in the center section should be combined with slats on the outer wing to avoid tip stall.

Giving the outer wing panels more dihedral also helps to delay tip stall since the angle of attack will increase there only with the ratio of the cosines of the inner and outer wing panels.

$\endgroup$
2
  • $\begingroup$ Just to verify if I understood correctly, the local stall coefficient at the tip must be at least 20% smaller than the maximum 2D lift coefficient? And what about applying twist to the wing, would that help the tip stall problem as well? $\endgroup$
    – Luan Arita
    Feb 27 at 13:09
  • $\begingroup$ @LuanArita Not stall coefficient, but lift coefficient. Yes, twist will also help. If you hit the maximum lift coefficient near the center while the outer wing is 20% or more below its maximum lift coefficient, stall should be benign. Make sure to pick an airfoil with gentle stall characteristics. 5-digit NACA airfoils are NOT recommended. $\endgroup$ Feb 27 at 19:23
0
$\begingroup$

As far as I understand, XFLR5 is not able to predict wing stall. The lift distribution must return to zero at the wing tips, it will often do so fairly sharply.

That said, this sort of analysis is often used as a crude predictor of CLmax and also the location of the wing where stall occurs. However, instead of plotting the lift distribution (cl*c), you should plot the distribution of the lift coefficient (cl).

In this case, because c tapers at the tips, where cl*c is relatively constant, cl will increase even more sharply. I.e. overall the cl distribution will have peaks in the outboard regions.

You can compare the sectional cl value with an estimate of the 2D clmax you expect fo this airfoil (and Reynolds number). Lets say you expect 2D clmax to be about 1.4 (and the whole wing uses that same airfoil), then you can repeat the analysis for a range of alpha, plotting the cl distribution each time, and when the first section reaches 1.4, you might expect the start of stall.

This also tells you whether a wing will stall first inboard or outboard -- perhaps where the roll control is (bad news).

This general approach is not perfect, but it isn't terrible either. While it has no rigorous basis, it has been used in conceptual design for a very long time. It will work better for a subsonic, un-swept wing like this one.

$\endgroup$
0

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .