34th America's Cup: multihulls!

The linear downwash is a beautiful theory but hard to tackle in practice, at least in case of a multielement foil with variable camber & twist & sweep. Usually associated with lifting line theory which cannot handle complex geometries.

Is there a way to measure downwash in practice - in the wind tunnel or real world? What exactly would you be looking for in a CFD analysis, velocities in a plane perpendicular to the trailing edge exit or what? How far behind the wing/lifting system?

As sidenote, downwash is a bit misleading in case of sails - they would be washing the air rather upwind than down...

 

Stored power was strictly forbidden, except for small springs and shock cords. No hydraulic accumulators on the supply side were allowed - only small accumulators on the return side to prevent cavitation.

The only time I know of that stored power was allowed was the 33rd America's Cup Match, when the Defender waived the racing rules that dealt with the design of the boats (jibs tacked on centerline, no stored power, no specially textured surfaces, no releasing of substances that modified the flow in the boundary layer). That's why USA 17 and Alinghi 5 used engines to run their hydraulics. But no stored energy in the 34th AC.

 

"Downwash" is a theoretical quantity of the idealized wing vortex wake in the downstream Trefftz Plane. It cannot be measured directly, or even extracted directly from a CFD solution. But the spanwise downwash distribution can be computed from the measured or computed spanwise load distribution (circulation distribution to be more precise) without too much difficulty. Its existence does not depend on the lifting line approximation.

 

Lifting line theory is still very useful for applying the principle to design situations. The target values of wing twist to lower the center of effort with minimum drag can be generated this way.

As for measuring downwash, that's a subject that deserves its own thread.

 

Hi Everybody,

Yesterday, I just read Dimitri Despîerre interview in a french sailing magazine.
He confirms what Mr Speer has explained above, and it seems that their system advantage was to move the AoA of the board by only 0.5° each time with 2 buttons for Jimmy.
It is also mentionned that the magnitude of the change was managed by mechanical feedback systems, which makes the change in AoA independent from the pressure level in the system, only the speed of the change is linked with the pressure available (my guess).
Among other, he mentionned:
They add some "flaps" at the bottom of the transom (he says jupes in french), which aimed to deflected some water downward.

But their most decisive modification, was to disconnect the load alarm on the bottom parts of the wing, in order to increase power in the bottom, move the wing center of effort aft, and therefore put more load on the rudder blades.

AFAI understand, they tried to minimize global twist on the element 1, but with more twist up and less down.
They open a bit more the tab (element 1 moving trailing edge) to increase air flow on the element 3 (the big flap) in order to have more load here and move aft the center of effort of the wing with more load on the rudders.

They also change the connection between vertical rudder blade and horizontal flap to decrease cavitation drag.

So no little or big Herbie, just plain vanilla stuffs which fit the rule-box.

The cost of their mechanical feedback system around 5000$

The 3 dudes at the origin of this " Race Winner" are (Alphabetic order):
Eduardo Aldaz.
Dimitri Despierres.
Neil Wilkinson.

Regarding induced drag:
Should a CFD rookie consider the basic equation of
Induced drag coefficient and other related concepts with some cautious?
Or does it still provide a good proxy of real word?

Best Regards


Erwan

 

Sorry for the off-topic question about induced drag,
I did not see the new topics on the Hydrodynamics forum.

 

First Foiling Americas Cup

Interview with Pete Melvin:
https://www.youtube.com/watch?v=VRGNR1O5mtA#t=1564

 

Lift force control

I did not have the time to look into the details of Oracle's hydraulic control system until now. What I can see is a mechanical feedback for a closed loop control system for the geometric foil angle. The automatic control system will keep the prescribed geometric angle constant. Why is such a system needed?
Because the foil is flexible and bends under load.
If the foil were not flexible, one could set the angle and lock it hydraulically, no mechanical feedback would be needed and no constant grinding.
If the foil is flexible, the geometric angle is a direct function of the lift force. Keeping the geometric angle constant is therefore identical to the task of keeping the lift force constant. With the pushbuttons at the steering wheel it is therefore possible to preset a certain lift force and the system will keep this lift force constant, better than any human being can do it. There are no additional sensors or even electronics needed. You only have to grind like hell to keep the pressure up. My old post http://www.boatdesign.net/forums/multihulls/34th-americas-cup-multihulls-34612-247.html#post654467 is still valid.
I am wondering if in the neutral jury, who examined this system, were any experts on aero-elasticity and sophisticated control systems. Did they understand the system? It is indeed no "Herbie", because the lift control is hidden in the physics of hydro-elasticity.

 

Actually, it was because the supply pressure was highly variable. The prohibition against stored energy did not allow an accumulator on the supply side of the hydraulic system. All of the hydraulic power had to come from the grinders at that instant. When something else in the hydraulic system was moving at the same time, the pressure would drop even if the grinders were keeping the pumps going.

As a result of the variable supply pressure, for a given movement of the valve controlling the rake cylinder, the response of the board would be different. This made it virtually impossible to achieve precise, consistent control of the board. Every time Jimmy pressed a button, he got a response he couldn't anticipate.

The feedback resulted in the same board movement for each press of the buttons. The buttons drove an electromechanical servo, much like you'd see in a model airplane. The servo provided the mechanical command signal to the actuator servo loop. The loop would drive the board to match the servo, regardless of the supply pressure. If the pressure was high, it would get there quicker than if the pressure was low, but the board would stop at the same angle.

Board flexibility didn't really play into it. Board flex would effectively change the cant angle, but not the incidence of the wing because there was no sweep to the geometry.

The Measurement Committee that evaluated the system did have hydraulic experts. They understood what the system was and how it worked.

 

Thank you. I guess I should believe you.

 

I don't know if this has been posted yet but I found it a useful summary of the Oracle Team USA system.

http://www.cupexperience.com/blog/2013/11/ac72-foil-control-system

 

Thanks for posting. That's what I guessed it was back on September 28 in post #3755. Not clear to me from the illustration if the piston in the tube functioned as a damper.

It's a vey clever, simple and elegant solution to a complicated problem. Perhaps much simpler than some folks thought was possible.

 

First Foiling America's Cup

===============
Thanks,Corley-great stuff!

 

Thanks Corely. That led me to this (via a search for the spool valve part number ).

http://noticeboard.americascup.com/wp-content/uploads/2011/09/PI_52.pdf

Now that document answered quite a few on my questions from earlier. But the measurement committee doesn't seem to have a hydraulics guy on it. Talk about letting the camel's nose into the tent.

lets take a look at a couple snippets -

"The effect of direct contact" is all but meaningless. The effects can propagate through the entire ship if you desire them to. And in the context here, we do

I like the wording in red. It sounds like they really mean it. But the committee somehow failed to see that their idea of black-box-hydraulics was being sorely abused by this system.

This is a stability augmentation device that happens to also be the prime mover for foil rake. The rake can be adjusted hydraulically based on a manual input to a stepper (allowed) or, from an input that is a function of all of the degrees of freedom of the board case (not allowed, hydraulics can not have inputs other than manual). By choosing the exact point in space to mount the valve, a wide range of response options are available (exactly how much space they had to work with probably restricted things a bit). The valve can receive input from case yaw, case rake, case cant, case heave, case sway, and case surge. Some of these will be designed out - the system will be insensitive to some of them. It will try to be as sensitive a possible those perturbations that represent a height excursion - case yaw, case pitch, case surge.

This is a data acquisition system. And it is a control system. It senses ride height excursions in the form of board case deflections and initiates the appropriate corrective action. It's also very old school and works splendidly. There are rocket motor steering controls that work the same way. Lacking a control input, they correct for perturbations passively. It's just what they do when they are doing nothing.

The protester was looking at the wrong end of the machine. The issue really wasn't an energy storage issue or how the device responds to manual control. The issue was what it is doing in the absence of manual input.

 

A Clockwork Oracle

Somehow, I missed the last fifteen pages of discussion

Petereng beat me to it by 10 pages.

The typical problem in hydraulic position control is to eliminate the slop in the system from feeding back as described above. One way to do this is to use the ram itself as the extension measure, not the structure being rammed. If the entire mechanism that they built was strapped to the ram, instead of the case, they have a black-box system. As it is, they don't.

I still have some more reading to do. I tried to find some examples of the two different systems, ram-feedback and frame-feedback, but no luck so far. Most modern systems perform this feedback control, for instance, keeping the forks level on a shooting boom forklift as the load changes, using fancier valve bodies that respond to pressure changes in a manner similar to what petereng was describing with the pressure relief setup. Those fancy valves may or may not be allowed under the rules, but they basically won't work to do what Oracle is doing here anyway. What Oracle is doing isn't feedback (because it's not related to the control input). It is passive control.

I believe the net effect of the system is to improve stability and lessen the amount of foil cant necessary to achieve stability. If all you want to know is the ram position, you do something along these lines.

http://www.youtube.com/watch?v=S0subOEmdx8