sail aerodynamics

Discussion in 'Hydrodynamics and Aerodynamics' started by Guest, Mar 21, 2002.

  1. CT249
    Joined: May 2003
    Posts: 1,449
    Likes: 191, Points: 63, Legacy Rep: 215
    Location: Sydney Australia

    CT249 Senior Member

    Gust response is one obvious reason. And if the masthead rig was so much more effective why is it unknown in open-design unrestricted inshore racing classes (apart from those that use masthead genoas just for extra area in light winds)
     
  2. markdrela
    Joined: Jun 2004
    Posts: 307
    Likes: 30, Points: 28, Legacy Rep: 324
    Location: MIT Aero & Astro

    markdrela Senior Member

    Euler and Navier-Stokes CFD methods can certainly account for a wind gradient. Gentry was undoubtedly referring to potential-based methods, specifically Full-Potential, Panel, or Vortex-Lattice. These assume an irrotational freestream (uniform freestream in practice), and hence cannot exactly represent a wind gradient.

    But Gentry was too pessimistic I think. The potential-based methods can account for a nonuniform freestream if one makes the assumption and approximation that the vorticity in the freestream is not significantly redistributed by the sail's own flowfield. Another way to say this is that a puff of smoke introduced into the airmass anywhere upstream of the sail will not change its vertical position very much (compared to the sail height) by the time it arrives at the sail.

    Anyway, with this assumption, one can incorporate a wind gradient into a potential-based method, like Vortex-Lattice for example, by simply defining a "local freestream velocity" which is a function of height.


    This is equivalent to what's called the "frozen gust" approximation in the aero business. It's routinely assumed when people compute or estimate gust loads on an airplane.
     
  3. Doug Halsey
    Joined: Feb 2007
    Posts: 640
    Likes: 212, Points: 53, Legacy Rep: 160
    Location: California, USA

    Doug Halsey Senior Member

    I think it's likely that Gentry knew about the possibility of representing the velocity gradient, in theory, but was referring to the capabilities of the programs that he was using.

    When he was at Douglas Aircraft, he worked closely with John Hess (one of the pioneers of 3-D panel methods), who tried to explain to me how superimposing the non-potential onset flow with a potential perturbation field would be a legitimate thing to do.

    I never really understood at the time, so thank you (markdrela) for clearing that up.
     
  4. Mikko Brummer
    Joined: May 2006
    Posts: 574
    Likes: 83, Points: 28, Legacy Rep: 258
    Location: Finland

    Mikko Brummer Senior Member

    You are certainly right, Tom, although although I'm not sure if you should call it "starting vortex"... Anyhow, taking another look at the original file and with the sails "see-through", it would seem that the vortex leaving the jib upper leech would consist of left overs from the windward side turbulence and some turbulence from the upper leech stalling as the jid head swings back. But surely there's an abrubt change in circulation when the jib head hits its aftmost point and starts swinging forward, causing a change in vorticity, too.

    I rotated the axis system into the lift & drag reference, and it appears that at the moment the vortex is being shed from the jib leech (the bow is starting to go down after pitching up to the max), the jib lift is at its minimum. The jib lift reaches it's maximum when the bow is starting to go up after it has been max down. This is consistent with what we found in the Star simulation, but with the Star you don't seem to get a similar puff of vorticity from the jib as you do with the X-35 model. Maybe it's the higher aspect ratio of the jib, but on the other hand the Star motions are double as much.

    The motion in the simulations is forced motion, it's not simulated. For the Star, the motions were recorded on board when actually sailing, so they should be fairly realistic. For the X-35, I simply halved the amplitude of Star motions but kept the period, as it would seem something you could have onboard a bigger boat, but it could be that the period, for instance, is short. In both cases, the waves in the simulation are "fake", a solid moving with the true wind, to add realism to the flow around the hull and thus hull drag. The waves are not making the boat pitch, and in case of the X-35 they could be a little out of phase, too.
     
  5. Jamie Kennedy
    Joined: Jun 2015
    Posts: 541
    Likes: 10, Points: 0, Legacy Rep: 117
    Location: Saint John New Brunswick

    Jamie Kennedy Senior Member

    Great work Mikko.
     
  6. Mikko Brummer
    Joined: May 2006
    Posts: 574
    Likes: 83, Points: 28, Legacy Rep: 258
    Location: Finland

    Mikko Brummer Senior Member

    Many more good comments that I'd like to reply to, but this time of the year is too busy... For the record however, lift & drag numbers:

    The Jib max CL is 1,99, CD at this point is -0,02 (negative drag)

    The Jib minimum CL is 1,35, CD at this point is 0,08

    The absolute min drag of the jib is CD -0,115

    I'll post the main numbers and the vertical distribution of lift & drag later when I get the time.
     
  7. philSweet
    Joined: May 2008
    Posts: 2,682
    Likes: 451, Points: 83, Legacy Rep: 1082
    Location: Beaufort, SC and H'ville, NC

    philSweet Senior Member

    Mikko-

    I think this is worth a look at. I have some doubts that you can transfer six-DOF motion from one boat to another quite so blithely.

    The canonical response to dynamic motion is dependent on three terms and equated to a perturbed angle of attack for that instance. From Mark Drela's book, the equation is -

    alpha = h'/U + Theta + Theta' * chord / 4 / U

    The point of reference on the airfoil is taken at the 3/4 chord point, not midpoint or 1/4 chord.

    h' is the rate of change of translation of the foil, basically sway and roll in this case. Theta is the pertubed angle of attack, and Theta' is the rate of change of Theta.

    In the model, h', Theta, and Theta' are each a cyclic function of all six degrees of motion (assumed to be a linear combination), so in order to maintain the relative importance of the terms in which they appear, you must do the following -

    Change the wind speed to match the change in scale. That brings the third term in line with the second one.

    Looking at the first term, the roll would automatically scale the h' by the scale factor, so h'/U would be constant after the above was done. So would sway. Yaw rate might need looking at also. This gives the happy result that you can in fact scale a motion such as pitch by 1/2 as long as you scale all six of the motions by the same proportion, which isn't obvious until you do the math.

    However, there is still the matter of the response to these dynamics. The response is expressed in nondimensional time - Tbar = U * T / (chord/2).

    When you scale the model up as you did, the response is slower. So if you keep the encounter frequency the same, the response function will have a different phase lag than in the original. This is something a helmsman will notice and adapt to. So you also need to change the encounter frequency to preserve the phase relationship, ie freq/Tbar needs to be constant. (added: but if you scale velocity along with linear distance, you can keep frequency the same)

    At first glance, it looks like you can scale frequency and velocity as above and scale the size of the motions independently and preserve the quality of the helmsman's actions, which are supposed to be making the boat go faster. If the you don't scale things as above, It looks like the helmsman's actions are going to be out of phase due to the phase change in the sail's response, and will probably be counterproductive.
     
    Last edited: Jun 24, 2015
  8. Jamie Kennedy
    Joined: Jun 2015
    Posts: 541
    Likes: 10, Points: 0, Legacy Rep: 117
    Location: Saint John New Brunswick

    Jamie Kennedy Senior Member

    Negative drag on jib is interesting. I presume this is due to the deflection of the wind around the main, so the wind is effectively coming from a better angle than the boats apparent wind used for the lift/drag datum. I guess this demonstrates how the main and jib must be considered together for evaluating performance, but can be analyzed individually for better understanding.
     
  9. philSweet
    Joined: May 2008
    Posts: 2,682
    Likes: 451, Points: 83, Legacy Rep: 1082
    Location: Beaufort, SC and H'ville, NC

    philSweet Senior Member

    And also from the motions of the sails which change the apparent wind direction as well. The variation of the jib's drag would be caused by changes in the main's circulation and the unsteady motion of the jib.


    Edit: thinking about this a bit more, the change in the circulation of the mainsail would be about proportional to the change in the velocity over the jib, which would keep the uplift angle about the same, so I think the variation in drag would be due mostly to changes in transverse motion (roll rate, sway, and yaw rate), and only very weakly a function of pitching/surge.
     
  10. Jamie Kennedy
    Joined: Jun 2015
    Posts: 541
    Likes: 10, Points: 0, Legacy Rep: 117
    Location: Saint John New Brunswick

    Jamie Kennedy Senior Member

    I always thought the main reason to try and concentrate weight to reduce your radius of gyration was only to reduce energy losses through the water, but this thread has helped me realize the other reason, perhaps the bigger reason, is to cause less disturbance to the air flow. Makes sense. Still I wonder if there is any mechanisms to improve performance by storing and releasing energy, like with a bendier mast, stretchier rigging, more fractional rig for more whip and opening action on the top of the main, maybe less tension on the forestay. Does the Star and Finn simulations include rig dynamics?
     
  11. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    At the time Gentry was writing his papers, the state of the art in computational fluid dynamics was panel codes. These assume potential flow, which is irrotational (except for the vortex singularities). The wind gradient is a sheared flow, which is rotational. So the codes of the day would have only dealt with a uniform free stream.

    Today, when one talks about CFD, typically what's meant is a finite volume code. These implement the Reynolds averaged Navier Stokes (RANS) equations, or more complex flow models like the detached eddy simulation (DES) that Mikko was showing. With a finite volume code, you can have any kind of flow profile as the input to the computational domain, so they can take into account the wind gradient.

    In general, when you read Gentry's stuff, you'll notice that he only deals with two-dimensional flow and usually only talks about lift. But the real difficulty in sail design or sailing trim is drag, not lift. That's not in any way a criticism of his work - he made huge contributions to the rational discussion of sailing aerodynamics instead of (with apologies to Kipling) the "Just So Stories" that pass for explanations of what's going on with sails.

    As Mikko has shown, we can finally start to get a handle on 3-dimensional flows, flow separation, and unsteady effects. But even so, the computations are so expensive that potential flow is still used for the bulk of sail force predictions.
     
  12. brian eiland
    Joined: Jun 2002
    Posts: 5,067
    Likes: 216, Points: 73, Legacy Rep: 1903
    Location: St Augustine Fl, Thailand

    brian eiland Senior Member

    Don't Use the Main(sail)

    This is the title of a subject thread over on the Cruiser's forum that I responded to, and added a copy of that posting to my aft-mast discussion here on this forum:
    http://www.boatdesign.net/forums/sailboats/aftmast-rigs-623-49.html#post741571

    I utilized some of Mikko's (WB-Sails) material and illustrations in an argument to further justify the need to pay more attention to the headsail rather than the mainsail, but not in lieu of the mainsail.

    ...an excerpt
     
    Last edited: Jun 26, 2015
  13. Doug Halsey
    Joined: Feb 2007
    Posts: 640
    Likes: 212, Points: 53, Legacy Rep: 160
    Location: California, USA

    Doug Halsey Senior Member

    So Tom, are you disagreeing with the comments in post #497. Do you think I would be wasting my time to incorporate a sheared onset flow in my Vortex-Lattice code?
     
  14. Erwan
    Joined: Oct 2005
    Posts: 460
    Likes: 28, Points: 28, Legacy Rep: 97
    Location: France

    Erwan Senior Member

    With less ambition than Doug, I would have the same question regarding LLT & Non-uniform flows like the Wind Gradient.

    I know there is a workpaper from Karman & Tsieh, regarding this issue, I don't purchase yet, as it seems to addresses Propeller flow on the wing.

    As application is only A-Cat rig, I am not a professionnal; I wonder what is the marginal efficiency of extra complexity in this context?
    URANS /VLM/ LLT which can translate into
    OpenFoam/Commercial Sail Software/DIY EXCEL spreadsheet.

    Cheers

    Erwan
     

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

    Mikko Brummer Senior Member

    Thanks - great info, even if a little outside my "comfort zone".. So how would you suggest scaling motions between different sized boats? The Star LWL is 4,75 m and the X-35 9,7 m. Displacements 900 kg vs 4900 kg.

    The Star measurements were done in 12 kn of (true)wind, while the X simulation is for 9 kn. I realize I may only have halved pitch - yaw, roll and heave remain the same as for the Star (I should check this). However, pitching has a huge influence on the apparent wind angle in the top of the sails, so I still believe it's an important (the most important) factor in the variation of the instantanious sail forces.
     
Loading...
Similar Threads
  1. Revboat
    Replies:
    28
    Views:
    5,771
  2. Forecaddie
    Replies:
    5
    Views:
    766
  3. Forecaddie
    Replies:
    9
    Views:
    1,632
  4. Forecaddie
    Replies:
    14
    Views:
    1,953
  5. Remmlinger
    Replies:
    45
    Views:
    5,872
  6. dustman
    Replies:
    25
    Views:
    3,832
  7. JerryWo
    Replies:
    4
    Views:
    1,330
  8. BMP
    Replies:
    12
    Views:
    1,503
  9. Howlandwoodworks
    Replies:
    2
    Views:
    1,601
  10. jmf11
    Replies:
    11
    Views:
    2,934
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.