drag, sail, induced and other.

Discussion in 'Hydrodynamics and Aerodynamics' started by Anatol, May 23, 2015.

  1. Anatol
    Joined: Feb 2015
    Posts: 191
    Likes: 0, Points: 16, Legacy Rep: 10
    Location: los angeles

    Anatol Senior Member

    cardinal sin

    I guess I've committed the cardinal sin by not reviewing other threads past and present which might relate to my concerns. Hopefully if I say 'sorry' I will not be permanently outcast.
    http://www.boatdesign.net/forums/hydrodynamics-aerodynamics/sail-aerodynamics-457-11.html
    is very long and I'm about halfway through it, but my question is addressed by Tom Speer and responded by CT249 and Brian Eiland. Very interesting - copied below. AFAIK, Tom did not respond.

    Tom says a knife-like leading edge is not as good as one with some cross sectional volume - this is consistent with airplane wing shapes. But clearly, big mast sections are also undesirable, especially when they're not aerodynamic.

    So there is a sweet spot. If one could build a leading edge that was ideal in shape, (assuming it could have infinite strength as a mast) how 'fat' would it be?


    http://www.boatdesign.net/forums/hydrodynamics-aerodynamics/sail-aerodynamics-457-11.html
    TSpeer The theoretical effectiveness of the headsail is not due to its cleaner leading edge. It's due to its interaction with the mainsail. The mainsail gives a boost to the head sail, and the headsail reduces the lift of the mainsail. It's exactly the same as lee-bowing your competitor - the jib gets a lift from the main, and the main gets a header from the jib. But the effectiveness of the whole combination is increased, so there's a net gain.

    The thin leading edge of the headsail is actually a disadvantage because there's no forward-facing area upon which leading edge suction can act. A mast with no separation zones would be a better leading edge than the thin leading edge of the jib - hence the effectiveness of wingmasts and pocket luffs.

    There are two separated zones behind the mast - one on the lee side and one on the windward side. The jib reduces the effect of the leeward surface separated zone in a couple of ways. One is by reducing the wind speed at the mast, so the mast effectively sails in a lighter wind, producing less drag. The other way is by creating a favorable pressure gradient to the jib trailing edge that encourages the separated flow to reattach to the mainsail.

    In high winds when the mainsail is operating at a lower angle of attack, the backwinding of the main by the jib, causing reverse curvature in the luff of the main, also reduces the windward separation zone. This is anagous to under-rotation of a wingmast.

    So the jib reduces the drag of the mast over what you'd have with the same rig bare headed. This is in addition to the lee-bow effect.



    CT 249
    Tom, I didn't realise the mast improved the main's efficiency, I was going off information like Marchaj's notes about things like "the adverse effect of the mast on the sail" and Bethwaite's words "if we take the mast away, and consider the performance of a sail set on a wire, such as a headsail, the situation becomes much better..."

    Why has the view changed? Is it that the testing techniques have improved, or where some of the earlier theorists just using wrong principles? My Marchaj has him talking about air speeding up in the slot, which as you and Gentry point out, is the opposite of what happens. Or so I hope, as I feared (and I think Gentry says) aerodynamics is not an area for laymen.

    As you say "the jib reduces the drag of the mast over what you'd have with the same rig bare headed. This is in addition to the lee-bow effect" why do A Class and C Class perform better without jibs? Were you just using the case of a mainsail of the same size and adding a jib, rather than redistributing a given area between the two sail? Is the drag reduction fairly minor? I assume it's actually fairly minor (although "minor" is indefinable) because otherwise boats with low foretriangles (and hence a smaller % of the mainsail with the advantage of the jib's effects) would be less effective than masthead rigs in some conditions, wouldn't they?


    TSpeer
    Take a look at Polhamus' leading edge suction analogy, expecially figure 2. Although the paper is concerned with planforms that have more sweep than a typical jib, the flow picture at a jib's leading edge is quite similar.

    A sharp leading edge basically has no forward facing area except for the slope of the sail's camber. The low pressure peak at the leading edge gets pointed more to leeward than would be the case for a more rouded leading edge. Here's a physical way of looking at it. The air has to turn through a certain angle in order to follow the lee surface of the jib aft of the leading edge. If the air does most of the turning curving around the leading edge, that orients the suction more forward. But if the air flows past a sharp leading edge and does its turning around a vortex or leading edge separation bubble, then the low pressure of the turning is oriented at right angles to the sail's surface instead. This is still pointed forward relative to the boat, but you'd rather have it pulling forward and to windward instead of forward and to leeward.

    This is the basic idea behind Polhamus' analogy. There has to be a low enough pressure over a large enough area to turn the air and produce a given amount of lift. You get the same lift in either case, but the sharp leading edge has a drag component that the rounded leading edge with attached flow doesn't.

    The ideal shape would be thin everywhere except near the leading edge, like this Liebeck section:

    As a practical matter, jibs with head foils have proven to be every bit as competitive as jibs with wire luffs or hanks. They are an example of a jib with a thick leading edge.

    Brian Eiland
    You're right Tom, that 'very thin foil' leading edge can be a real problem in the practical world as well. In gusty, shifty conditions it would be real difficult to maintain an ideal flow over the headsails, particularly at this leading edge. Originally I had thought the big round furling foils of Profurl units were less efficient than the oval or foil shaped sections of some other manufacturers. But as I explored the situation further I became more inclined to utilize a round-shaped furler foil on my twin-headsailed rig. This round shaped foil rotates more evenly, and could be made more robust. And it would act to cut down some of the 'sensativity' associated with a very thin leading edge. Some 'bluntness' at this leading edge actually transmits an advanced signal to the incoming flow.

    To suggest that a mast section at the leading edge of a sail is not that detrimental is not to my liking. Granted mast sections have steadily declined over the years either by material advances and/or creative rigging support. But most are still pretty big 'obstacles' at the leading edge of an airfoil. On most cruising boats with fixed mast I'd still be willing to wager that close to the first foot of sail area behind the mast does not contribute to driving the boat forward. Okay lets say this might be as little as only 6 inches in the very best cases. One half a foot along the entire luff of the conventional mainsail is a lot of lost sail area, and particularly towards the top of a fractional rig where there is no help from a headsail. If one doesn't utilize a fat-head mainsail you might as well write off the top 10-15 percent of the mainsail area as a contributor of forward drive to the boat.
     
  2. Mikko Brummer
    Joined: May 2006
    Posts: 574
    Likes: 83, Points: 28, Legacy Rep: 258
    Location: Finland

    Mikko Brummer Senior Member

    Anatol,

    You may also want to take a look at http://cdn2.hubspot.net/hub/209338/news/Ad_aerodynamics/index.htm?t=1428319882242

    In the attachment, a more recent article about Finn mast aerodynamics. According to this study, the mast shape is not very important - what you gain in the mast efficiency you tend to lose in the sail, and vice versa. The cat rig is different from a sloop in that the mast rotates with the sail - it makes the mast more efficient.
     

    Attached Files:

  3. Anatol
    Joined: Feb 2015
    Posts: 191
    Likes: 0, Points: 16, Legacy Rep: 10
    Location: los angeles

    Anatol Senior Member

    Mikko
    thankyou for this most informative link and doc!
    on page http://cdn2.hubspot.net/hub/209338/news/Ad_aerodynamics/index.htm?t=1428319882242#mastdrag
    you say -
    "the mast drive can be negligible or it can be even negative, real drag".

    "where the drag is oriented against the direction of the motion of the boat, the drag is often negative."

    I'm sorry but these two sentences seem contradictory and confuse me. " it can be even negative, real drag" seems to mean 'bad'. But "the drag is often negative" seems to suggest 'good' Am I to assume that 'negative drag' is lift?

    Also on this page, I get 404 for the 'close up of star mast' animation. :(

    Regarding cat rig, you seem to assume rotating mast - I'm not sure why. Given your 'the mast shape is not very important' remark, my thoughts on avoiding the complexities of rotating foil/wing masts are confirmed.

    From the pdf "������ ������������������ �������� " Wait, that doesn't look right :) So much for copy/paste. Anyway, you say - all the masts, attached to the same sail, perform similar." Very interesting! but these are all aerofoils of some sort. The big rectangular section of the spruce mast on my 60s sloop can't be a help :)

    Your doc supports my intuition is that a smooth curve between mast and sail on the lee side is important. How much worse would a simple round mast be, if the sail attachment was on the leeside and relatively smooth?

    Again, intuition tells me that some volume at the leading edge is desirable, as long as curves are fair. I note that fish tend to have a foil shape, the head is usually the widest part.

    thanks!
     
  4. Erwan
    Joined: Oct 2005
    Posts: 460
    Likes: 28, Points: 28, Legacy Rep: 97
    Location: France

    Erwan Senior Member

    Hi,
    The ongoing discussion sparkS a candid question regarding proxy equations for induced drag coefficient and the relation between 2D Lift Coefficient & 3D Lift Coefficient
    The question is how accurate/and or reliable are the following relations :
    Cdi=(Cl^2)/(PI*AR*e) ?
    3D Cl= 2D Cl*(AR/AR+2) ?
    I guess it's a good proxy to establish relative values, isn't it?

    Thanks in advance

    EK
     
  5. Mikko Brummer
    Joined: May 2006
    Posts: 574
    Likes: 83, Points: 28, Legacy Rep: 258
    Location: Finland

    Mikko Brummer Senior Member

    In
    "Note that in aerodynamic terms, the mast does produce drag, as the aerodynamic drag is defined in the direction of the apparent wind. But in hydrodynamic terms, where the drag is oriented against the direction of the motion of the boat, the drag is often negative."

    I refer to the body fixed coordinates of the boat: "negative drag" is not lift, but positive drive, in the direction of the motion of the boat.

    Cat rigs are often unstayed and rotating with the boom, at least on smaller boats.

    A round mast is not bad, but you don't want to attach the sail on the lee side - that's like overrotating the mast, which is not good.

    Intuition is great but often leads you wrong when it comes to aerodynamics.

    I cannot edit those old webpages anymore, but here's screen copy of those broken pages offline.
     

    Attached Files:

    • DD2.jpg
      DD2.jpg
      File size:
      199.9 KB
      Views:
      681
  6. brian eiland
    Joined: Jun 2002
    Posts: 5,067
    Likes: 216, Points: 73, Legacy Rep: 1903
    Location: St Augustine Fl, Thailand

    brian eiland Senior Member

    Just found this subject thread, and ran across this posting.

    Tom, I have a question about "the lift is actually carried by the forward element"? Are you saying the lift is generated by the forward element??
     
  7. brian eiland
    Joined: Jun 2002
    Posts: 5,067
    Likes: 216, Points: 73, Legacy Rep: 1903
    Location: St Augustine Fl, Thailand

    brian eiland Senior Member

    Interesting observation.



    I was making a similar observation observation over here:
    http://www.boatdesign.net/forums/hydrodynamics-aerodynamics/sail-aerodynamics-457-48.html#post752087

    Which monohull one was that you speak of??

    I have a multihull rig design with a fwd-raked mast:
    http://www.boatdesign.net/forums/sailboats/aftmast-rigs-623-7.html#post198605

    And here was an interesting wind tunnel test on an aft-mounted mast with 2 headsails on a monohull:
    http://www.researchgate.net/publication/262797205_Wind_tunnel_and_CFD_investigation_of_unconventional_rigs
     
  8. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    Here's a typical section pressure distribution from that wing. The flap is deflected 24 deg, and and the apparent wind is aligned with the forward element. The forward element and the flap have equal chord lengths.

    The peaks of the pressure distributions are clipped, but you can tell that the area inside the pressure distribution of the forward element is substantially larger than the area inside the pressure distribution on the flap.

    [​IMG]

    For this section, if you hold the forward element fixed and only vary the flap deflection, the change in lift is approximately 80% of what you'd get by rotating the whole section. If you hold the flap at a fixed angle to the airstream and rotate just the forward element, you get about 20% of what you'd get by rotating the entire section. So would you say that means the lift is generated by the flap or by the forward element?

    The lift is generated by the total deflection of the airstream, and the air comes off the wing largely parallel to the direction of the trailing edge. That's why I say that although the lift is carried by the forward element, it is determined by the aft element.
     

    Attached Files:

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

    tspeer Senior Member

    BTW, it would be a huge mistake to think this means the forward element is in some way more effective at producing lift than the aft element, and therefore it should be made substantially larger at the expense of the size of the aft element. It wouldn't be able to produce as much lift as it does if it only had a small flap.

    The forward element gets all the glory, but the flap does the unglamorous grunt work of pressure recovery. What matters is how the combination works together.
     
  10. brian eiland
    Joined: Jun 2002
    Posts: 5,067
    Likes: 216, Points: 73, Legacy Rep: 1903
    Location: St Augustine Fl, Thailand

    brian eiland Senior Member

    Thanks Tom,
    I thought that was what you were saying, but I got a little confused by the initial posting that the forward element was 'carrying the lifting load'.

    So its the aft element that creates most of the lifting force, ....at the expense of lots of extra drag, is that correct?

    And would I be correct in saying that most of this lifting force is oriented in a upward direction (lifting airfoil), rather than in a forward direction.

    So from the aircraft designers perspective he would not see much reason to rotate the whole section,...he is getting much more lifting force by deflecting the rear flap.

    But the sailor is looking for more forward drive, so he might well choose to rotate the whole section??
     
  11. DCockey
    Joined: Oct 2009
    Posts: 5,229
    Likes: 634, Points: 113, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    The maximum lift coefficient is generally higher when a flap is deployed than by just rotating the entire airfoil.
     
  12. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    That is not correct. The two elements work together. The main element couldn't create the lift that it does without the influence of the flap.

    In so far as the flap lets the section create the lift it does without separated flow, and in so far as the slotted flap allows the lift to be created without the leading edge suction peak that would cause more skin friction, it would be more accurate to say that the flap reduces the drag over trying to do the same job with a single element section.

    Here's the dissipation coefficient distribution for the same case I showed above. This shows where the profile drag comes from around the section. The main element has more drag on it than the flap.

    [​IMG]

    No. By definition, "lift" is the component that is perpendicular to the apparent wind vector and perpendicular to the axis of the lifting surface. For an airplane, the wing and apparent wind vectors are generally a horizontal plane, so lift is directed upwards. For a sailboat, the apparent wind vector is horizontal and the axis of the rig is vertical, so the lift vector lies in a horizontal plane.

    Whether the lift is directed forward with regard to the longitudinal axis of the boat depends on the apparent wind angle. If the apparent wind is at 90 deg to the axis of the boat, the lift points toward the bow. If the boat is head to the apparent wind, the lift is pointed along the beam.

    Given the same apparent wind vector, all rigs will produce lift in the same direction - that's the definition of "lift".

    No. My point was to look at the sensitivity of orienting different parts of the section to the apparent wind in order to determine which one could be said to "produce" the lift.

    Aircraft designers are just as interested in configuring their wings for best lift/drag ratio as sailors are.
     

    Attached Files:

  13. Erwan
    Joined: Oct 2005
    Posts: 460
    Likes: 28, Points: 28, Legacy Rep: 97
    Location: France

    Erwan Senior Member

    Thank you for the Cdiss coef plots and for the wing sail sections and comments

    It provides many answers to old questions I had, as a result I'd like to check some assumptions to be sure my understanding is correct, and hope this candid approach could help other wing rookies.

    1-The respective job of E1 & E2, and the tunnig strategy

    If the E1 job is mostly the "rooftop" job and if E2 is mostly the "recovery" job
    Can we assume that best L/D ratio is likely to occur with E1 @ 0° AoA all along the span, in other words, with a twist that should meet the apparent wind twist ?

    While E2 twist should try to meet the elliptical lift distribution for the required aero load?

    2-E2 wing section
    If E2 job is focused on recovery with a transition almost at the LE, then a wing section with max thickness positionned as much as possible near the LE in order to keep max chord lenght for the "recovery" job, should be better, as long as it does not create more drag everything else equal)?

    In addition, the more the max thickness is in the front of the section, the shorter the LE and the lighter the E2)

    The shape of the E2 section displayed above, seems to confirm this candid intuition, is it correct?

    (BTW, as an alternative to NACA009 orNACA 0012 for E2; I was playing with XFLR5 and the S 8025 which has its max thickness around 22% chord.)

    Hope it's not off topic

    On another thread I saw an ZFOIL graph,
    I didn't know XFOIL had a kid

    Cheers

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

    tspeer Senior Member

    No. There is a range of flap deflection and angle of attack combinations that result in the same lift from a given section. A trimming strategy for a wingsail might be to first of all control the total lift with the sheet, to the limit of either the maximum lift (downwind in light wind) or heeling moment. Then adjust twist so as to achieve the minimum induced drag for the given heeling moment. Finally, trade off wing rotation vs flap angle to trim for minimum profile drag.

    Minimum profile drag does not occur at 0 deg AoA. When the drag polars for each deflection are overlaid, you can draw an envelope of the drag polars such that at each lift coefficient, the flap is deflected to minimize the profile drag for that lift coefficient. The resulting flap angle vs section angle of attack schedule is what you want to target.

    It turns out that this tends to occur when the stagnation point is just to windward of the leading edge. So a much better guide to tuning the flap deflection is to look for indications of stagnation point location. This is what the C-class "gang of four" wind vanes are all about.

    First of all, the minimum induced drag for a wingsail does not come from an elliptical or semi-elliptical lift distribution. The minimum induced drag is obtained when there is a linear wash distribution in the wake, and this does not result in an elliptical lift distribution because of the interference with the water surface and the nonlinear apparent wind due to the wind gradient.

    But, given that one knows what the target lift distribution is then, yes, the flap twist is used to get that lift distribution.

    The optimum lift distribution will tend to be more bottom-loaded in heavy winds compared to high winds. This allows the wing to produce more thrust for the same heeling moment, because the heeling moment is constrained by the righting moment from the platform.

    This is essentially the same issue as for the design of single-element sections. It depends a great deal on how much laminar flow you think you can get. If you assume the oncoming flow is so turbulent (or the surface so rough) that the amount of laminar flow is minimal, then you'll design for a fully turbulent boundary layer. The minimum drag for fully turbulent sections depends very little on the section shape, and is driven primarily by the maximum thickness.

    If you are trying to achieve significant runs of laminar flow, then you'll use a flat rooftop or somewhat favorable pressure gradient for the laminar run. This will drive the point of maximum thickness aft.

    A conservative approach would be to design for a moderate amount of laminar flow, while also ensuring that if the boundary layer turns out to be fully turbulent, the profile drag is only a few counts higher than what it would be for a fully turbulent design. That way you've hedged your bet - you'll reap some benefits of laminar flow if you can get it, but you won't be unduly penalized if you don't get it.

    Placing the point of maximum thickness very far forward on the flap allows the bulk of the flap chord to be devoted to the job of pressure recovery. Adding thickness to the flap raises the velocity on both the windward and leeward sides, making the job of pressure recovery that much harder, so flap sections tend to be on the thin side. The leading edge pressure peak from the thin sections can be suppressed by making the slot narrower. A flat contour results in a concave pressure recovery that is a good compromise between a Stratford-like pressure recovery and a NACA-like linear pressure recovery. The small thickness combined with the symmetry constraint also tends to drive the flap shape to flat contours. The end result is a flap section that is similar to an elliptical leading edge with a long tangent wedge.

    The maximum thickness for a slotted wingsail tends to be fairly similar to that of a single element section - around 12% to 18% of the total chord. The thickness ratio of the main element is high, especially if the main element chord is short. A thickness of 22% of the total chord is probably going to have excessive profile drag.
     

  15. Erwan
    Joined: Oct 2005
    Posts: 460
    Likes: 28, Points: 28, Legacy Rep: 97
    Location: France

    Erwan Senior Member

    Thank you very much Mr Speer,

    Until now I was sure that only E1 could target a moderate amount of laminar flow, I guess moderate is < 20%.

    I mentionned elliptical distribution, because, I noticed for an A-Cat sailing at 18knts downwind in 10knts TWS, there is a big twist but almost no shear

    Conversly, windward, the boat sailing at 10/12, there is less twist and a big shear, both shear & twist seem to be mostly in the lower part < 10ft height.

    Thanks again

    Erwan
     
Loading...
Similar Threads
  1. Mikko Brummer
    Replies:
    81
    Views:
    15,419
  2. CatBuilder
    Replies:
    28
    Views:
    9,598
  3. andrew spiteri
    Replies:
    67
    Views:
    4,097
  4. Solario
    Replies:
    110
    Views:
    9,751
  5. Alan Cattelliot
    Replies:
    75
    Views:
    6,942
  6. mitch184
    Replies:
    9
    Views:
    1,267
  7. maughanjamesr@gmail. com
    Replies:
    10
    Views:
    1,994
  8. bitfield
    Replies:
    23
    Views:
    2,943
  9. Bishiba
    Replies:
    4
    Views:
    1,730
  10. captncoop
    Replies:
    5
    Views:
    1,454
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.