Are humming foils slow?

[Nigel Hale]

Hi,

I sail an Australian cherub (www.cherub.org.au), which is a 12 foot skiff type dinghy, in a similar style to a 29er.

When planning at high speeds downwind we find that the centreboard makes a humming noise. The same noise can be heard on many skiffs, but ours seems particularly loud. So much so that nearby boats have commented on the sound. It could also be the rudder, but I don't feel anything in the tiller. Both foils are a very tight fit in their cases.

Is this noise slowing us down, and if so what can be done to fix it? I figure that the energy to create the sound must be comming from somewhere, and that it can't be great for speed.

Many people have suggested that it is all about the tip of the centreboard, which seems feasible. The main length of the centreboard is a high aspect shape with a straight leading and trailing edge. However the tip of our centreboard is a J shape with a curved, raked front edge leading down to the deepest point which is at the bottom of an almost straight trailing edge. The trailing edge also has a flat back, probably 1 or 2mm wide along it's total length.

Any suggestions would be much appreciated. I have also heard a similar noise, although much quieter, on a laser.

Thanks in advance,

Nigel.

[water addict]

yes, humming foils are slow.
Try bevelling the trailing edge of your blade so that is a few degrees from perpendicular to the x axis (fore-aft axis). The hum is usually from alternating vortices at the trailing edge of the foil.

[jehardiman]

I would beg to differ with Water Addict.

First of all, all foils will hum to some extent. It is, as stated, a function of the trailing vortex street (which is always present and linked to the trailing edge treatment) and the amplification of the vibration by the structure. Laser's can be notorious for humming, because the board causes the deck to radiate the noise. Do not underestimate the ability of the foil to radiate fluid borne noise. I have owned a Ranger 26 that, if down below, you could hear motorboat and ship props through the keel long before before you could hear them on deck.

The question to ask is what is the frequency of the hum. A low pitched hum can be bad as it indicates seperation well up the foil span (i.e. a wide vortex trail that rolls over the trailing edge and swaps sides). I have seen it so bad that the board will actually "chatter" in the case as it rocks back and forth.

A high pitched hum is not so bad, as it indicates a thin, stable, well attached vortex. Give me a nice stable manageable vortex anyday, which is the reason for the bevel that water Addict mentions. Indeed I commonly "tune" foils by selecting trailing edge treatments. I can then look at noise cuts against the expected frequency to determine if the after section of the foil is thick enough and not seperating well up the cord.

[MikeJohns]

Originally Posted by jehardiman

............all foils will hum to some extent. ...................

I've been looking at a noisy foil recently on a racing strut/bulb design. When the vessel gets over 15 knots the foil bursts into an unpleasant higher pitch principle around 800Hz harsher sound not a hum.

I modelled the natural frequency (and harmonics) and it's not structural resonance . The strut is not a very hydrodynamic one but more rectangular section. I figure in this case it is cavitation.

[DGreenwood]

Turn the sound up and listen to the first part of this video. Talk about singing foils!!

http://video.google.com/videoplay?do...arch&plindex=1

[Petros]

You are correct in assuming that a humming foil causes more drag than a non-humming foil. You are better off to find the cause and eliminate it (even if it is considered "normal"). It is not just the noise energy, but also if the centerboard is vibrating it takes energy out of the forward motion to cause the flexing. This may not be insignificant.

There could be several causes, one is alternating unstable vortexes alternately coming off the trailing edge due to inaccurate profile shape (it pulls the surface in alternate directions causing the vibration), it also could be that the foil is ventilating and alternately shedding the trapped air off each side of the foil. The trapped air also means that the "shape" the water sees in not the shape of centerboard, which is almost certainly more drag than the foil shape of the centerboard.

I would check the shape of the centerboard section all the way down to the tip. It must be correct, especially at the tip, or you could get the vibration. You are better off with a strait taper centerboard that is simply cut off square at the tip than with a fancy tip shape but with a poor foil section. A quick test would be to tape some vortex generators on each side of the centerboard down near the tip. These will trap and stabilize the vortexes and stop them from pulling the surface in alternate directions that cause the vibration. If this is the case you need to reshape the section. Simply squaring off the tip with a sharp edge should help.

I did not see anything in the rules that require a certain shape centerboard, so you should experiment with better shapes. When the aft profile of the foil closes off too quickly the risk of ventilating air is large (the flow does not want to stay "attached" to the surface). So you might concider a foil section with the max thinkness more forward to allow a more faverable pressure gradient along the aft part of the foil section.

Also making it stiffer will reduce the tendency to vibrate, and it will reduce the size of the oscillation. That is presuming there is no ventilation.

I would also check how good the "seal" of the centerboard slot at the bottom of the hull, it is possible the air is ventilating down the centerboard case. If there is any practical way to seal this better against the centerboard (rubber blade seals perhaps?) then you would reduce the tendency to suck air down along the trailing edge of the centerboard.

Let us know what you find out and how you fix it. Good luck.

[water addict]

Originally Posted by jehardiman

I would beg to differ with Water Addict.
......

Alright - I was going for the short simple answer.
Humming most of the time indicates a structural vibration induced by the flow on the blade. This is robbing energy from the forward motion, and is slow. The bevel at the trailing edge wil cause the vortex to stay to one side of the blade instead of alternating. The alternating of the trailing edge vortex tuning with localized natural frequency of the blade causes the vibration. Inducing vibration takes energy from the system that could otherwise be used overcoming other components of drag.

I think that's pretty much the same thing you said jehardiman. I thought a succinct answer that directs to almost all cases of blade vibration would be of more help.

[Paul B]

Originally Posted by MikeJohns

I've been looking at a noisy foil recently on a racing strut/bulb design. When the vessel gets over 15 knots the foil bursts into an unpleasant higher pitch principle around 800Hz harsher sound not a hum.

I modelled the natural frequency (and harmonics) and it's not structural resonance . The strut is not a very hydrodynamic one but more rectangular section. I figure in this case it is cavitation.

What sort of a "racing strut/bulb design" would use a "more rectangular section"? That sure doesn't sound like anything any raceboat designer would do.

[clanning]

One technique used in aircraft wing design (hollow aluminum section construction) is to use a concave trailing edge, i.e., instead of having the trailing edge come to a sharp point where the upper and lower surfaces meet, there is a "U" channel in between (with the open part of the "U" facing back). This gives you 2 nice sharp exit points, and has the added benefit of easing riveted construction in aircraft.

I've never heard of airplane wing "humming" but I would think that the objective is similar, that is a cleaner exit of the air past the trailing edge. (Airplanes normally use a non-symmetrical foil, so the trailing edge loading would be different and maybe less susceptible to differential vortex generation and flow separation.)

[jehardiman]

Originally Posted by water addict

Alright - I was going for the short simple answer.
...snip...I think that's pretty much the same thing you said jehardiman. I thought a succinct answer that directs to almost all cases of blade vibration would be of more help.

I don't know , I think it is important to differentiate between flutter, which is an energy robbing structural problem, unconstrained seperation, which is an energy robbing flow problem, and and the unavoidable bound seperation, which is a result of using a foil in water. All of those cause "noise","humming" or "singing". It is important to determine which one you have before trying to say how to fix it. Just absolutely beveling the trailing edge may actualy make the situation worse if the foil is properly designed and the noise is just the expected trailing edge bound vortex. And more importantly, I think it is important to point out that lower frequencies indicate more energy loss. A foil that has seperation below the acoustic threshold actually has more loss than one you can hear (and in response to clanning's comment... you don't "hear" aircraft wings because they are too high to hear unlike the older biplane wing wires, but it can be measured by accelerometers). See Hoerner, FDD, Section III-3

It is also important to remember that stiffer, lighter, structures transmit more acoustic "noise" than more massive flexible ones. Two identically shaped foils, one of polyester-glass over wood and the other of carbon-epoxy over rigid foam, will have different sounds in response to the same flows.

[jehardiman]

Originally Posted by MikeJohns

I've been looking at a noisy foil recently on a racing strut/bulb design. When the vessel gets over 15 knots the foil bursts into an unpleasant higher pitch principle around 800Hz harsher sound not a hum.

I modelled the natural frequency (and harmonics) and it's not structural resonance . The strut is not a very hydrodynamic one but more rectangular section. I figure in this case it is cavitation.

800Hz @ 15 knots would call for a seperation width of ~.07 inch (1.7mm). Unless you have a VERY thin plate foil I would look else where, such as cavitation as you mentioned. You will need to calculate the cavitation number or look in Blevins, Applied Fluid Dynamic Handbook, pp362-366, for a sizing answer

[Petros]

Originally Posted by clanning

I've never heard of airplane wing "humming" but I would think that the objective is similar, that is a cleaner exit of the air past the trailing edge. (Airplanes normally use a non-symmetrical foil, so the trailing edge loading would be different and maybe less susceptible to differential vortex generation and flow separation.)

actually a similar effect does occur on aircraft, it is called flutter and it is very dangerous. When it happens the airplane will usually (not always) experience an in-air break-up unless you can slow down in a hurry. So a lot of effort is taken to eliminate it from aircraft. Nothing quite so dramatic occurs on boats, the speeds are slower so the forces are smaller, so we live with it on boats. But it can and does occur aircraft. You simply see the effect on both centerboards and rudder, and even on sails, and live with them.

Also many aircraft will use symmetrical airfoils on the tail surfaces and on aerobatic aircraft where it is desirable to fly as well inverted as well as wheels down.

Whether symmetrical foil or not flutter will occur where unsteady and cyclic pressures changes coupled with cycle airframe flexing. The solution is the same on them as on boats, stabilize the unstable pressure variations (change the shape or use vortex generators), and/or stiffen the structure. Sometimes you have to sacrifice light weight and minimal (theoretical) drag to eliminate flutter. And it will have less drag and be more controllable than when you have a flutter condition.

[clanning]

Originally Posted by Petros

actually a similar effect does occur on aircraft, it is called flutter and it is very dangerous. When it happens the airplane will usually (not always) experience an in-air break-up unless you can slow down in a hurry. So a lot of effort is taken to eliminate it from aircraft. Nothing quite so dramatic occurs on boats, the speeds are slower so the forces are smaller, so we live with it on boats. But it can and does occur aircraft. You simply see the effect on both centerboards and rudder, and even on sails, and live with them.

Also many aircraft will use symmetrical airfoils on the tail surfaces and on aerobatic aircraft where it is desirable to fly as well inverted as well as wheels down.

Whether symmetrical foil or not flutter will occur where unsteady and cyclic pressures changes coupled with cycle airframe flexing. The solution is the same on them as on boats, stabilize the unstable pressure variations (change the shape or use vortex generators), and/or stiffen the structure. Sometimes you have to sacrifice light weight and minimal (theoretical) drag to eliminate flutter. And it will have less drag and be more controllable than when you have a flutter condition.

While new to the kind of Reynolds numbers dealt with here, my first job out of school was running flutter and dynamic loads analyses for deHavilland Aircraft in Toronto. My recollection is that you'll always have some vibration in an elastic structure (which airplanes definitely are) -- flutter happens when the structural or systems (i.e., control surface and system) oscillations start interacting and gaining energy from the aerodynamic ones. Gain some energy and it keeps the vibrations going -- gain too much and they grow bigger until something breaks. It's all about "tuning" a structure to make sure that the natural frequencies of the structure and the aero aren't too close together.

Key aircraft oscillations start around 15 Hz I think -- lower frequency = bigger amplitude = more danger (in general). For a carbon CB, say, the frequencies are high enough to hear and maybe disturb flow a bit, but they shouldn't be structurally significant unless there's a big bulb on a slender fin wobbling about.

REGARDING VORTEX GENERATORS, has anyone seen these used on a boat, used to stabilize flow at the keel root? Even very clean aircraft can have them in strategic positions like the fin-stab junction on a T-tail. (I remember the "shark skin" used on the AC 12 meter Stars and Stripes '87 -- where else has this popped-up?)

Chuck
Ottawa

[jehardiman]

Originally Posted by clanning

While new to the kind of Reynolds numbers dealt with here, ... snip...
REGARDING VORTEX GENERATORS, has anyone seen these used on a boat, used to stabilize flow at the keel root? Even very clean aircraft can have them in strategic positions like the fin-stab junction on a T-tail. (I remember the "shark skin" used on the AC 12 meter Stars and Stripes '87 -- where else has this popped-up?)

Chuck
Ottawa

Aye, there's the rub! Vortex Generators (i.e turbulance stimulators) are not needed at the Rn's that most boats sail at and likewise the size of vortices that are need are about 3 orders of magnitude smaller due to kenimatic viscosity differences. A lot of things that are standard/can get away with in aerodynamics and "inviscid irrotational" CFD work poorly in real water. It is amazing the subtlety that is needed/can be used to manage flow energy gradinents in hydrodynamics.

[Nigel Hale]

Thank you all for your replies. This has turned into a most interesting thread.

I have a few experiments to go and try now.

As a guess, I would say that the noise, starts at about 12knts at about 300Hz. It then increases to probably 800Hz at 20knts which is close to our top speed. These are fairly wild guesses, I have listened to some tones using a sound generating program and tried to imagine if the frequency we hear on the water is similar. The "volume" of noise definitely increases with speed, and maybe this is affecting my judgement of the frequencies involved. Next time I sail, I will listen carefully because I think that it "cuts in" at a certain speed, and is already a clearly audible frequency. I'll speed up and slow down a few times, and listen carefully for the behaviour.

Our board is a very stiff carbon/epoxy board with a timber core, probably stiffer than the others in our class. It's average weight, but very stiff with hugh unidirectional fibers down each side at the thickest part (recessed into the timber so as to be flush).

The board is also very tight in the case (we have to stand on it to get it down) which is packed with carpet. I don't think that it would be air tight, but I could experiment by putting heaps of tape around the top of the case, to stop air getting down the case from the top. I don't think that the board is ventilating, it's not that bad a shape. It was hand crafted, but by a very skilled tradesman (not me!).

However, now that I think about it, I do think that the percentage thickness at the bottom would be too high. Where the J shape comes in, the chord lengh is rapidly shrinking, and the thickness only tapers relatively slowly. This could be causing trouble at high speed???

I have some questions before I go and change the trailing edge though. If I bevel the trailing edge, why won't there still be an osilation in the vortexes, but now alternating around an "off-centre" average. And secondly, won't that cause an asymetric shape, which would produce more lift on one tack than the other, all be it a small difference?

I'll post my findings from this weekend's racing for those who are interested.

Thanks again,

Nigel.