There is also analytical data that shows that it works.
The triangles and circles in the charts are measured data .... the red dashed line is theoretical from a math model.
The thing to note is the fact that the reversed airfoil, with extremely "sharp" end pointed into the wind, has a very linear passage through zero angle of attack. The rounded end forward has some wanderings around zero before it settles on a slope. Yes the Re number for this data is less than our applications, but I suggest that its still applicable data.
Now you are going to say that, "I know at 10 or 15 degrees a sharp stab is gonna stall like an sob. Who says that the stab ever "sees" an angle of attack that high? Remember in a sharp turn the tail is being rotated very rapidly in a direction that reduces the aoa. Save that for another time.
Empirically, yes I have tried it, .... it works. I went for a year of feeling Paul's stabs at the Nats before I tried it. It doesn't stall. I was somewhat surprised. The feeling I got was just what the charts show, a more rapid initiation of the turn with no degradation in tracking.
Mine looks like this, except it has the front rounded off to about a 1/16" radius and much thinner. The part that I think matters (after having done some experiments myself, and witnessing and analyzing the results from numerous other's tests, like David's documented tests, Ted and David's largely undocumented tests with flat stabs and turbulators, and Paul's various solutions) is that you maintain some amount of normal component of q as far back as you can.
This is
my analysis of the various experiences (back to Gid Adkisson's "Laser", which is where a lot of this started), I emphasize that this is merely
my analysis of others experiences and a few of my own experiments (with various “slippers” over the stab/elevator of several airplanes). I don’t claim it will meet anyone else’s standards for “validity”, however it does seem to cover most examples for which the stab/elevator was the relevant factor and for which the experimenters were known to be sound and objective observers. It does meet
my standards (low or otherwise):
Thin flat stabs of low aspect ratio and radiused LE- thin means the pressure change at the corner is so abrupt that the separation bubble is always present and more-or-less in the same place, and, the vortex from the tips curls around and affects a significant fraction of the stab all the time, and adds normal force, or at least, turbulates the surface
Moderate flat stabs - These have the issue with the separation bubble forming or not forming, because the radius is such that it doesn't always guarantee that it forms on both sides all the time, and either forms or goes away with tiny changes in AoA. If you make it low aspect ratio, the interference from the tip interferes with it some, and maybe helps. Even more so with low aspect ratio and a large LE sweep angle. High aspect ratio allows the "bubble/no bubble" effect because most of the stab is free of any influence from the tip. Even worse with tip plates, that interrupt the main vortex and replace it with a much smaller one that has nearly no effect at the LE
Moderate flat stabs with turbulators - experiements show most of the problems if you have have the "bubble/no bubble" problem with the medium flat stab are eliminated if you put turbulators on it, top and bottom, right near where the radius ends and the flat begins. It is better than nothing to put them further back, but the further forward you put them, the better. My theory on why that is is that it forces those bubbles to form and be stable at the same place most of the time at most AoAs, in addition to deflecting a bit of normal force back due to viscous shear effects.
The “moderate flat stab” comments and analysis are based on very extensive experiments by David, Ted, with me observing/commenting, and with me finding that my similar stabs saw no real problem and no improvement from turbulators, and Paul's observation of the same - not a problem. The big difference was that my stab and Paul's stabs were lower aspect ratio than those Ted and David were using, and also Doug's bear, who with tip plates, came a lot closer to infinite aspect ratio and gave a negative example (gleaned through comments on SSW). This was what lead up to the "De-Tails aticle, after David decided that, based on the analysis above, that he wanted to experiment with further variants and airfoils. It was already posted so I won’t belabor it.
As far as I can tell, David’s results with flat stabs and various LE’s were consistent with the “separation bubble” theory and I was able to pretty reliably predict the results using that theory, particularly the thin and triangle-shaped LE. The “smaller elevator” was no surprise to anyone, but that is a different topic.
So, to the extent I understand it, with Ted’s “pointy” rework of the second Trivial Pursuit stab, which is very close to if not identical to what Paul is currently using, and looks very similar to the one Craig Beswick posted earlier. Same theory works there – it’s *better*, because the abrupt change in shape is less abrupt – a slope leading into the discontinuity at the radius/flat transistion point and less likely for the bubbles to form/disappear with small changes in the AoA, although apparently, still not immune because on both Ted’s model (wire turbulators) and Pauls (zigzag strips) it still helps, but perhaps not as much as turbulators on the flat stab with a blunt radius. The gadgets help by stabilizing the existence and position of the separation bubble.
Also from David’s article – the airfoiled C1 stab. This was at my suggestion and more-or-less straight off the second Infinity plan, with the exception the squared-off corners at the hinge line (another thing I shamelessly ripped off from Paul). This was designed using the theory of separation bubbles from the various flat-stab experiments, trying to keep at least some positive slope until I couldn’t avoid it any more, and eliminating the bubbles at least at low AoA where it matters most to tracking and corner exit. Effectively, this extends the point section that is more-or-less “tacked on” to the front of the otherwise flat stab all the way to the hinge line, so, similar idea just taken to the obvious conclusion. Note it is also very similar to the Green Box and ’52 Nobler plans, particularly once you round off the hinge lines.
I note that this also explains why, so far, I haven't had *any* significant positive effect with tripper strips, VGs, or turbulator wires, turbulator wires being a moderate *negative* and the others a wash. David had very positive effect with tripper/zigzag strips, but he also didn't round of his corners at the hinge line, so that may be a different (also very interesting) issue.
Again, C2 the same thing with the thinner stab, which predictably had poor tracking by running the elevator in turbulence.
The advantage to the flat stab, or variations, is that it is easy to align. The second half the theory is that, after a lot of examples slapping us in the face, there is a fair bit of tolerance in the alignment in one direction (positive) and not the other (negative). However, if you have a flat stab, it is far, far more critical, because you can’t have it running at a tiny positive or negative angle of attack due to the same separation bubble theory – that’s where it is the most critical. So, with an airfoiled stab as shown, you can’t align it as well, but you also don’t have to.
Of course, if make it removable in the most convenient way, you can also experiment with the alignment instead of having to glue it in place at construction and then hope it never changes - in my case by building the fuselage like a bridge girder, and in other cases just crossing your fingers. The alignment (positive/negative) was barely mentioned in David’s article but was a much bigger effect than the stab airfoil, with the flat stabs super-critical and the airfoiled stabs much less so – as long as you never put in even the slightest breath of “negative”. David actually ended up with FAR MORE positive than I ever would have had the guts to build in, inadvertently leading some others (particularly with flat stabs) to grief – the sort of grief that leads to the use of Zona saws.
Frank's data tends to confirm the situation I envisioned, he posted a picture earlier that was nearly identical to what I drew out in the pits at the old Napa field when we were doing the flat stab turbulator experiments.
So, anyone can throw darts if they will, but I was able to predict what was going to happen and what seems to happen in most of the variations.
Brett