Great thread this, thanks @silas this was very interesting, and interesting that they start with the result and then work backwards.
With an additional one being to avoid major low pressure peaks on the upper side of the foil as this will make it have bad ventilation characteristics.
This is not forward to back but transverse loading?
Fwiw for the rest of us, this article explained things in terms that I could just about follow, (this clip helped make sense of the plots - note the reverse y axis on the screenshots. They are worth understanding if youâve got this far!)
Imagining that the HA combines these in some clever way, something like below, but then I presume you start encountering the âTrue reflex entails a very aggressive pressure recovery that can cause quite some stall issuesâ
Front to back, not transverse. Only talking sections here so strictly 2D. A negative Cp means pressure below ambient pressure. When such a negative pressure peak gets near the water surface, it causes the surface to deform and when the deformation is big enough the foil sucks air and down you go without even getting a tip out.
Here is a NACA 0012 at 10 degree AoA to show what I mean:
Huge negative Cp spike right near the leading edge, falling quickly and then more gently down to zero. That nice gentle slope from 20% chord aft means a very gentle pressure recovery back up to ambient pressure, and thus a modest adverse pressure gradient (going from low pressure to high pressure as we proceed from leading edge to trailing edge on the upper/suction side of the foil surface). With a modest adverse pressure gradient, risk of stall (trailing edge stall to be precise) is low and it will be very gentle and progress slowly and predictably when it occurs. But, if that pressure peak comes anywhere near the free surface it will suck air suddenly and violently lose lift.
Contrast that with the first image you posted (forget the differences in angle of attack for now).
The suction pressure peak is much more modest near the leading edge and the pressure coefficent stays at something like 70% of the peak value all the way aft to 50% of the chord length. For the same lift coefficient (area under the Cp curve) as the 0012 section above, this more spread out pressure distribution will be able to get much closer to the free surface before it ventilates and the ventilation wonât be nearly as violent, giving better chances of recovering from it without crashing. But, look at the slope in the pressure coefficient from about 55% to 75% of chord - very steep, meaning a very strong adverse pressure gradient, and much bigger risk of stall - more violent, sudden, and unpredictable stall affecting the last 30-40% of chord all at once. This is the fundamental challenge with these types of so called roof-top pressure distributions (ones that avoid big peaks and stay high over more of of the chord length) in general, and foil reflex in particular.
That is super interesting. Making much more sense now, this tradeoff is super interesting. As with all of this, these tradeoffs seem insanely complex, and often non-linear, to the downside.
vs
On the sucking air point, I have been riding a 1201 with a repair on one tip, and I could swear it is sucks air on that side. The other side I can do a flat breach and it will pop out and I can catch it, but the repaired side is like a sniper shot when it gets near it seems to just drop out entirely, and it is on my stronger side too.
I thought I was crazy to suggest, but based on this maybe it is possible can see it being entirely possible that for that wingtip, it went from something like the first image to the second? (the front pressure profile)
What you describe is totally possible, itâs really easy to sand in a flat/bump/sharp on the LE and changes that are hard to see can have significant effects on pressure distribution.
Yes, especially in the tips where the foil sections are thinnest, 0.1mm can be a significant percentage change.
Iâve had the same problem before and just gone at it a few gentle strokes of 400 or 600 grit wet paper at a time until it seemed to get tolerable. Pretty much blindly, just trying to round and smooth as best as possible. Fingers much better than eyes for this work.
Interesting. Any idea what kind of profile I should be aiming for to minimise Cp spike possibility? From the image, maybe thicker is better, but seems unlikely to achieve that (or anything else really) by hand⌠Seems unlikely to achieve much, maybe chopping would help>
Note the pressure peak near the nose. A thicker section would have a less prominent peak.
This maybe along one of the reasons why the Armstrong HA V1 suggested the chop. Would chopping the tips help I wonder?
Ehh, then you start messing quite heavily with the spanwise loading (transverse lift distribution). If that was designed with some care to begin with chopping is unlikely to improve things, depending on your goals and priorities of course. Iâd say recent gen foils have had a lot more attention paid to all of these design aspects than earlier gen. Like I said, I went at it blind and in the end any percieved improvement may have been entirely placebo, but it got me to stop being so tentative on my backhand so main goal accomplished.
As far as thicker sections having a smaller pressure peak, it isnt that simple. The main correlation is how much the stagnation point moves with angle of attack. A very rounded and blunt leading edge sees the stagnation point move much more than a sharper, narrower. But, the radius to be mavigated is tighter with the narrower leading edge so its all a balancing of tradeoffs game again.
For that repaired tip, I would take a close look at the last few inches of the tip and look/feel for any imperfections, bumps, flats on the front 1/2 of the top surface and LE. Wetsand using a hard block to take down any bumps and then a softer block and finally handsand for the finish. I usually start with 180-220 and finish with 600-1000. A 600 finish might be a bit less prone to ventilating. The main thing to remember is that you want to always be moving and rolling the block along the chordwise curvature to create a fair curved surface any time you are working on or near the leading edge.
I just tested a âsmall batchâ foil designed by a professional foil designer, and I think I can say two things
The aft camber seems to be quite interesting. Iâve ridden the army 880 and this somewhat similar foil, and it really does make a difference. I saw a KT proto, and between them, the KT has by far the most camber, almost crazy looking, then the 880 and then this test foil. (no test on the KT sadly). The Army 880 definitely doesnât feel very fast though, so will be interesting to see if the KT can do both fast and slow.
Foil performance has likely now reached the point where there can be meaningful and probably now necessary contribution from professional engineers and scientists. Take for example Martin Fisherâs involvement in Starboard. Chief scientist for Ineos, which is spending Formula 1 amounts of money.
Yes, we have been open about our relationship with Americaâs Cup foil designers since we started designing masts back in 2016. We have always âoutsourcedâ our section shape studies to experts who have the design, experience, and resources in fluid dynamics because we are experts in composites, structural optimization, manufacturing, and design. Weâre not going to pretend that we can also run CFD (nor spend the many thousands necessary for the software to do so). In the past we have worked with Tom Speer. Tom designed the first rigid wing sails for Oracle and pretty much all lifting foils for the big catamarans. Sadly, he passed from cancer a couple years ago. Now weâre working with designers from Alinghi and American Magic. Foil design is incredibly multi-disciplinary. You canât just focus on one aspect of design (eg. foil section/CFD). We work with these experts to find the necessary balance between structures, materials, manufacturability, cost, high speed performance, low speed performance, stability, weight, and more. Honestly itâs what makes the job fun.
Kane does a great job explaining the design tradeoffs. Interestingly, to get the most out of the top end, you need a symmetrical leading edge, which sacrifices the low end. The aft camber compensates for that. So it might be that the symmetry on the leading edge is the innovation, and the aft camber is to compensateâŚ
Compare the Fireball, with regular camber. Adrian explains camber. So while Adrian seems right in his explanation of camber, Iâm more convinced of Kanes design
Fwiw, very hard to find much about aft camber online.
I found drawing it to help it click, the right version without the foil shows the camber and chord lines
Axis on reflex and camber, while technically being correct about the definition of camber, is possibly missing the point that a lot of downwind foils are adding aft-camber/reflex/whatever
I donât think this is a fad. The aft camber helps immensely with low end but does come at a significant cost. This keeps getting overlookedâŚ
Drag increases for a given size wing, and drag goes up a LOT for low speed even though it is possible to ride at low speed.
So what this does is allow riders to get up early on smaller wings. So even if the drag is a little more for a given size, when we ride a smaller size overall the drag is less.
Herein lies the tradeoff, if we try to ride this small high camber wing on swell that isnât moving fast, we are stuck in a very draggy mode and transiting between swell takes a ton of energy.
Iâm sure you know better than me. But for any given speed and weight of rider, there is a most efficient profile. For slow speeds, Iâm pretty sure it wonât be a small wing with rear loaded section versus a slightly larger wing with a more efficient profile.
My experience comes from riding all the sizes of the CODE R series at a bunch of speeds. I donât have the foil section and havenât run the analysis, but I have made careful observations.
The code R wings are able to ride to incredibly low speeds. But you go to high angle of attack and they become draggy the slower you go. At those low speeds, just for example, the bigger of the ART Pro series have way less drag.
It is certainly possible that other designs perform differently, I have not tried the Armstrong, the new Axis wings nor the new KT wings.
