numerical drag modeling

Discussion in 'Hydrodynamics and Aerodynamics' started by philSweet, Oct 21, 2013.

  1. philSweet
    Joined: May 2008
    Posts: 2,350
    Likes: 210, Points: 63, Legacy Rep: 1082
    Location: Beaufort, SC and H'ville, NC

    philSweet Senior Member

    A question for the numerical modelling crowd.

    Regarding drag modelling of ships during the early design phase, Does anyone know of any work done where drag is calculated using 'generalized functions' or 'linear functionals' based on basic hull geometry parameters and the control model of the appendages?

    It occurs to me that if you know the form stability, the velocity, and the control algorithm that governs each of the various gadgets which control stability, you should be able to calculate the power required for any seastate (a distribution functional). This would require the parent model to optimize the gain of each controller (ie, design the rudder based on the ship and the rudder's control algorithm and account for the other gadgets' cross coupling effects).

    The effect would be that at the conceptual design phase, all you need to tell the modelling program is the particulars of the rudder's control algorithm and its location. Same with any other stability gadget such as foils, gyros, tanks, prop thrust regulation, thrust vector control, bow thrusters, etc.
    There are limits to the complexity of the control algorithms that can be analyzed this way, but, that's probably ok at the early phases of design. The point being that with respect to a particular sea state description, I believe the control algorithms, and the efficiency with which the boat can get power from the prime mover to the actuator, and basic hull geometry completely determine the tradeoffs between the way different appendages should be employed to regulate stability. And it would also determine the amount of form stability that should be provided by the hull.

    The point of the exercise is to have a fairly simple and robust modelling environment that can receive control algorithms as inputs and determine the relative appendage and gadget sizes for minimum drag. I think it can be built fairly easily using functional analysis. You would still need the usual mass, moments and hull geometry numbers for the hull that feed standard dynamic models.

    One possible advantage in doing this would be to identify control systems that can be scaled in the way a known hull is scaled to a similar one. Currently, I get the impression that active stability and helm controls have to be reanalyzed when you scale a known boat. Functional analysis is good at finding ways to combine things so that they can be scaled smoothly.

    The result might be a set of particular control algothims that are independent of the boats size for a given sea state, and a then all the mother computer has to do is tune the individual controllers for the sea state, and that would be practically the same for all boats of a given class.

    For all I know, they've been doing this for thirty years, but I haven't seen anything about it.
     
  2. NavalSArtichoke
    Joined: Oct 2013
    Posts: 431
    Likes: 8, Points: 0, Legacy Rep: 83
    Location: GulfCoast

    NavalSArtichoke Senior Member

    The resistance generated by an arbitrary ship hull moving through the water is a little more complicated than you describe. Froude started model testing to determine the resistance of hull designs using scale models in the 1870s and this is still an important tool. However, in recent years, computational fluid dynamics has advanced to the point where one can compare the effects of various hull features on resistance, but the final predictions of resistance for large vessels are usually confirmed by model tests.

    Test series have been developed over the years whereby data, gathered from model tests of a series of the same hull having certain form parameters varied, has been collected in the form of charts or formulas which allow one to estimate the resistance of a new design.
     
  3. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 149, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    Of course, it should be added that even the most sophisticated CFD
    codes makes many simplifying assumptions, whether at model scale or
    full scale. For example, most CFD codes are unable to estimate the
    effects of splash and spray, especially if they land back on the
    water surface and thereby perturb the (oncoming) upstream flow.

    Many of us operate according to George E.P. Box's splendid advice:
    "Essentially, all models are wrong, but some are useful."
    Many more people should! :)
     
  4. philSweet
    Joined: May 2008
    Posts: 2,350
    Likes: 210, Points: 63, Legacy Rep: 1082
    Location: Beaufort, SC and H'ville, NC

    philSweet Senior Member

    In the above, I'm mostly concerned with accounting for the increase in drag due to a seaway over what a smooth water test would show. And also the increase in drag due to the requirement to limit the vessel's excursion from it's course and other state parameters. If you have six degrees of freedom and a linear model, you have twelve components which are of interest - three position variables, x,y,z, three orientation angles theta, phi, psi, and the rate of change of each of those six.

    The seaway itself, on average, will increase drag on a vessel. Any attempt to control the response of the moving vessel must also add drag and consume power. This includes passive geometry such as form stability and bilge keels, as well as active control gadgets. A reasonable question to ask early in the design process is what sort of active and passive controls should be looked at, how should control power be apportioned, and how much form stability and damping should be provided to meet the mission requirements. I believe the energy costs and tradeoffs can be modeled at a very early point in the design and that the control algorithms for the various gadgets are the primary determiners of how a stability requirement is best achieved. That still leaves the cost factors to be looked at though.
     
  5. NavalSArtichoke
    Joined: Oct 2013
    Posts: 431
    Likes: 8, Points: 0, Legacy Rep: 83
    Location: GulfCoast

    NavalSArtichoke Senior Member

    In my experience, unless there are unusual control issues with keeping a vessel directionally stable, the power losses due to active control elements are small in comparison to those required to keep the vessel traveling at a certain speed.

    Things like bilge keels and active stabilizer fins are designed and positioned on the hull so that they catch streamlines and their parasitic drag is reduced to a minimum.

    Yes, sea state has a significant effect on power required to maintain speed, but, absent model testing, there is currently no universal predictive model which can be applied to determine the additional power requirements for a given speed and sea state. Some studies were made back in the 1960s I believe which determined curves of additional resistance for general cargo ships given a sea state, but I haven't seen any similar studies of a more recent nature.

    When doing power predictions for a new design, a 'service allowance' is usually built into the final power number to account for the myriad things which might add drag but which cannot be predicted with reasonable certainty, a 'fudge factor' if you will.

    For vessel motion, prediction of directional stability of a given design is still more art than science. Although there are theoretical models out there which describe the directional characteristics of a general vessel, many of the parameters of these theoretical models cannot be easily determined by calculation and must be derived from turning tests performed in a model tank.

    In designing a new vessel, many aspects are competing for the naval architect's time and attention. Some factors like stability, power, and cargo capacity receive more attention than others, if only because these are better understood than some other factors.
     
  6. dskira

    dskira Previous Member

    And probably more financially obvious.
    After all the end-use of a commercial ship is financial rentability. And that is a very well understood factor.

    I enjoy artichokes. Mostly the heart, with a olive oil and parsley sauce on top.
    How did you get the idea of your user name?
     
  7. NavalSArtichoke
    Joined: Oct 2013
    Posts: 431
    Likes: 8, Points: 0, Legacy Rep: 83
    Location: GulfCoast

    NavalSArtichoke Senior Member

    At the place I worked once, the secretary had brought a large artichoke in for lunch. I and another engineer got to teasing her about her lunch and she started calling us 'naval artichokes' instead of 'naval architects'.
     

  8. NavalSArtichoke
    Joined: Oct 2013
    Posts: 431
    Likes: 8, Points: 0, Legacy Rep: 83
    Location: GulfCoast

    NavalSArtichoke Senior Member

    There was an old joke that the only reason marine engineers tolerated naval architects was because the marine engineers needed someone to keep the machinery dry.

    A vessel is more than a vehicle for carrying around a control system at sea.
     
Loading...
Similar Threads
  1. sinmania
    Replies:
    0
    Views:
    1,377
  2. Inquisitor
    Replies:
    6
    Views:
    353
  3. myszek
    Replies:
    5
    Views:
    495
  4. Peaky
    Replies:
    24
    Views:
    1,391
  5. Mikko Brummer
    Replies:
    62
    Views:
    3,847
  6. waterbear
    Replies:
    3
    Views:
    1,061
  7. dustman
    Replies:
    53
    Views:
    5,025
  8. Konstanty
    Replies:
    32
    Views:
    3,732
  9. rwatson
    Replies:
    8
    Views:
    2,274
  10. Slip
    Replies:
    3
    Views:
    1,137
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.