Keel design issues

Discussion in 'Sailboats' started by HeloDriver, Sep 8, 2004.

  1. koen
    Joined: Aug 2004
    Posts: 10
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: the Netherlands

    koen Junior Member

  2. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    Excellent paper! That's the "dillet" shape I mentioned before.

    The paper only covers a strut at zero angle of attack. I think a keel would need to have the the fairing widened so as to provide the same benefit as the stagnation line moves around the leading edge, and to prevent separation of the flow moving crosswise over the dillet.

    Here are some pictures of a T-33 wing root that show the classic expanding-radius type of fairing for the rest of the root:

    [​IMG]
    [​IMG]
     
  3. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    Thanks.

    I don't have any experience with exits. I suspect you're right that adding a modest radius to the aft edge would help somewhat. Obviously, having an exit that turned parallel to the flow so the drained water came out tangential to the skin would be even better. The Laser has such an exit built into it's bottom, although most sailors fill it up with a flush plug that has a bailer built into it.
     
  4. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    Morgan's a great guy! He used to talk about that keel some when I'd meet him at landsailing events. Interestingly, he's actually an architect, not a NA, although he maintains that small boat design is the last discipline in which a designer is truly free to be creative!
     
  5. Banush
    Joined: Sep 2004
    Posts: 2
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Spain

    Banush New Member

    Dear Sir,
    Your article is very interesting.
    Where can I learn more about the 32 foot boat with a 28" wide keel you mention in your article.

    I am now planning on buildig a 32 foot sailing sloop in aluminium. I thought that the foil keel would have to be rather thin - at 28" you have makes it very interesting. Do you have designs and/or pictures of that boat.

    Best Regards
    Banush


     
  6. Banush
    Joined: Sep 2004
    Posts: 2
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Spain

    Banush New Member

    "but over 355 in aluminum,fiberglass balsacore,etc. were built in europe"

    How can asccess information about these boats?
     
  7. Karsten
    Joined: Jun 2004
    Posts: 184
    Likes: 8, Points: 0, Legacy Rep: 33
    Location: Sydney

    Karsten Senior Member

    Tom, what happens if you sweep the keel backwards? This should move the stream away from the hull at least on the side of the keel with the low pressure if you have an angel of attack.

    Karsten
     
  8. jehardiman
    Joined: Aug 2004
    Posts: 3,762
    Likes: 1,152, Points: 113, Legacy Rep: 2040
    Location: Port Orchard, Washington, USA

    jehardiman Senior Member

    Karsten asked...
    If I may interject,

    Actually I believe it is the other way around. For a foil with positive sweep (i.e. the maximum section line raked aft), flow is toward the hull on the low pressure side and away from the hull on the high pressure side. The major reason to extremely sweep (>~ 20 degrees) a foil for lift/drag reasons is to increase the critical Mach number. As most vessels with foils do not operate anywhere near most foil's critical Mach number, it is a moot point as cavitation dominates for most water foils (see Hoerner, Fluid Dynamic Lift, chpt XV).

    Studies done with foils for vessels shows that sweep (+ or -) effects the lift/drag ratio and stall angle. Positive sweep may decrease stall angle and increase L/D while negative sweep may increase stall angle and decrease L/D (see PNA, chpt IX, sect 14). Because sweep angle affects which end of the foil will stall first, taper angle cannot be seperated from sweep effects.

    I personally perfer my foils with about a -5 degree sweep, a taper ratio of 40-50%, and an aspect ratio of 2-3 for practicality reasons.
     
    Last edited: Sep 23, 2004
  9. Karsten
    Joined: Jun 2004
    Posts: 184
    Likes: 8, Points: 0, Legacy Rep: 33
    Location: Sydney

    Karsten Senior Member

    Hang on. Something is fishy here.

    1. The stream going towards the hull at the leading edge of the keel looks good on the plot. But my explanation is very simple. The boundary layer flow along the centreline of the hull will hit the leading edge of the keel and therefore has to slow down. Slowing a flow leads to a higher pressure in the flow direction and at one point separation of the boundary layer. That's what you see on the plot. Slightly off the centre line the flow doesn't have to stop because there is no keel leading edge anymore. It might still slow down and maybe even separate but the effect will be smaller. My guess is that there is only a separation bouble near the keel leading edge and if you bump up the reynolds number the flow will be very nice and perpendicular to the keel leading edge further away from the hull.

    2. If the keel (or wing) is swept back (like on most airliners) the flow on the low pressure side goes to the tip. Imagine you are an air molecule on the low pressure side exactly where the lowest pressure is in one section. Where are the low pressure points next to you? The one towards the hull is upstream (you are already past it) and the one towards the tip downstream. So where do you go if you are a typical air molecule? Towards the hull against a high pressure gradient or towards the tip where the pressure gradient is much smaller? Let me tell you. Air and water molecules are pretty lazy ******** who don't like to klimb steep hills.

    Cheers,
    karsten
     
  10. jehardiman
    Joined: Aug 2004
    Posts: 3,762
    Likes: 1,152, Points: 113, Legacy Rep: 2040
    Location: Port Orchard, Washington, USA

    jehardiman Senior Member

    Remember; we are in a flow field which includes time dependent fluid momentum, not just a static pressure tensor. I'm going to pick up a big hammer and hit your mind right now. While we test foils in tunnels where the moving fluid flows "aft" remember that in real life an advancing foil causes the fluid to move "forward" from an "undistrubed" state.

    If we take a Y(stbd)Z(down) cut plane in the fluid and look at what happens as a positively swept infinite foil (measured clockwise from Z in the XY plane) moves through it in the +X direction we will see that on the low pressure side the maximum low pressure occurs just after the maximum change in curvature of the foil section (approximately just ahead of maximum section for symeritic foils at low angles of attack without too sharp entries). Therefore, as our foil passes through the cut plane at the section of minimum pressure, the low pressure will be on the leading side of the foil and the higher pressure will be on the trailing side of the foil, causing the flow to move up (-Z) and forward (+X). Conversely, the opposite is true for the pressure side; i.e. the point of maximum pressure is on the leading side, and a lower pressure is on the trailing side causing flow to move down (+Z). Notice that this cause the streamlines to seperate between the face and back of the foil and that higher lift occurs at the trailing "tip" vice the leading "root" because of the increased pressure difference.

    Now comes the bugaboo, the short span 3D foil in an incompressable real fluid :D . At the root, as was stated before, you get an increase in pressure due to flow effects around the foil. Depending on the magnitude of this pressure, flow can be outward and down the foil causing the horseshoe vortex. Similary at the tip, the flow on the pressure side is down and off the foil from a high pressure and that flow on the suction side should be up the foil away from the tip. This in a real fluid attempts to cause a very low pressure on the suction side (increased lift at tip, remember) which can reverse the spanwise flow and generates the tip vortex as momentum of the pressure side fluid is changed to fill the pressure void. In aeromechanics the vortex is tight and high energy. In hydromechanics at sailboat speeds, the momentum cannot be changed as easily so a slow low energy tip vortex is generated.

    Do some streak tests and convince yourself
     
  11. Karsten
    Joined: Jun 2004
    Posts: 184
    Likes: 8, Points: 0, Legacy Rep: 33
    Location: Sydney

    Karsten Senior Member

    O.K. I'm going to get out the hammer. If you have a low aspect ratio and a low reynoldsnumber I can imagine that the streamlines are going towards the hull on the low pressure side. As you said (and I agree) the flow goes from the high pressure side to the low pressure side around the tip and therefore pushes the flow on the low pressure side towards the hull. If you have a low aspect ratio wing the influence will be probably up to the wing root.

    I'm from the aeronautical engineering mob and we deal with high reynolds numbers and high aspect ratios. I'm quite sure that on a high aspect ratio swept wing the flow on the suctions side goes towards the tip because the tip vortex has very little influence on the flow. I'm going to search for a streak photo to show it. I've one for a low aspect ratio delta wing (used for military aircraft) at high angle of attack but I guess that's not very relevant here. It has a vortex above the leading edge which makes things even more complicated.

    Would be interesting if soembody could find a streak photo for a keel.

    Cheers,
    Karsten
     
  12. jehardiman
    Joined: Aug 2004
    Posts: 3,762
    Likes: 1,152, Points: 113, Legacy Rep: 2040
    Location: Port Orchard, Washington, USA

    jehardiman Senior Member

    Here is a photo I was able to find at http://adg.stanford.edu/aa208/modeling/empiricalmethods.html. Now this is not a symetric foil of low aspect ratio in an incompressible fluid at high Renyolds and low Mach, but what the hey, it's something.

    The photo is of a single aircraft model with streaks on the right wing and tufts on the left wing. The streak clearly shows that the flow turns toward the body approximately perpendicular to the quarter cord line, while the tufts show the slowing in the basic direction of flow of momentum layer in the area of low pressure near the leading edge. Notice however, that the horseshoe vortex soon dominates the flow near the wing root both in the streamlines and momentum layer.
     

    Attached Files:

  13. HeloDriver
    Joined: Sep 2004
    Posts: 19
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Portsmouth, NH

    HeloDriver Junior Member

    I had the opportunity recently to the look at the keel root / hull junction of an Etchells, which closely approximates your description. Though, I see little evidence of such on the latest AC designs, from the few pictures I’ve seen. Any thoughts as to why?
     
  14. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    I think you're confusing a couple of different phenomena. First of all, oil flows show the direction of the shear stress at the surface. That's not necessarily the same as the direction of the flow outside the boundary layer.

    The turning of the flow to be more perpendicular to the axis of the wing happens because of the increased velocities on the upper surface. Contours of equal velocity magnitude run at essentially constant percentages of the chord. If you consider a section of the wing taken parallel to the freestream, then on the forward portion of the wing the velocity on the inside of the section is higher, and the pressure lower, than the velocity on the outside of the same section. This makes the flow bend inboard.

    Another way to look at is to take a section perpendicular to the leading edge or quarter chord line. This is approximately perpendicular to the contours of equal velocity. If you take the vector sum of the added velocity (compared to free stream) in this direction, and add it to the freestream velocity, the resulting vector will point inboard compared to the freestream.

    This is the same phenomenon that makes the wind turn towards being perpendicular to a cliff, because the wind accelerates over the crest of the cliff or ridge. Hence the predictable lift you get by working the shore.

    On the aft portion of the wing, outboard of the wingroot area, you are getting the opposite sign of the same phenomenon because the flow is slowing down in the pressure recovery region of the airfoil. Near the trailing edge, the pressure inboard is higher than the pressure outboard of a given section and the flow turns in the outboard direciton.

    I think what you're seeing in that wedge region near the root is an area of trailing edge separation that gets more severe toward the root. This isn't quite the same phenomenon as the horseshoe vortext that the dillet addresses. Both are separation near the wing/body junction, but they arise from different causes. The horseshoe vortex is driven by differences in stagnation pressure along the leading edge, due to the body boundary layer. The separated region shown in the photo is due to the adverse pressure gradient in the wing boundary layer aft of the leading edge.

    I think this is more the kind of drag that's addressed by the fairing shown in the pictures of the T-33.
     

  15. jehardiman
    Joined: Aug 2004
    Posts: 3,762
    Likes: 1,152, Points: 113, Legacy Rep: 2040
    Location: Port Orchard, Washington, USA

    jehardiman Senior Member

    Tom; Your response clearly shows the problem I was trying to point out to Karsten. You're turned around backwards.

    Climbing on heretical soapbox:

    This is only true for a stationary foil in a flow, not for a moving foil in a stationary fluid. The flow is not slowing down, it is speeding up from zero as energy is added to it, not subtracted from it. At the trailing edge the pressure is lower (see the speeding up comment previously) inboard and higher outboard so the flow in to an inboard direction (consistant with stall, i.e. no lift, outboard for a positively swept foil and inboard for a negatively swept foil). Which direction the flow is in is a function of where in the universe you are standing to observe and an incorrect application of circulation theroy. Circulation theroy is only a mathmatical contrivance...read Lanchester, Prandtl, and Lerbs...that falls out of the Kutta-Jaukowski condition; it is not a true representation of physics.

    :(

    Off my soapbox now.
     
    Last edited: Nov 6, 2004
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.