Michell v CFD and EFD: Round 2007

Discussion in 'Software' started by Leo Lazauskas, Dec 5, 2007.

  1. johnhazel
    Joined: Jun 2008
    Posts: 250
    Likes: 5, Points: 18, Legacy Rep: 60
    Location: Michigan

    johnhazel Senior Member

    Hey Leo, Thanks for posting the beta. I'm trying to find the Aleg42.dll to get the beta to run.

    John
     
  2. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 155, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    Thanks for the reminder, John. Other people who have recent variants should already have that dll.

    I've put up a new version with the alleg42.dll file at:
    http://www.cyberiad.net/michlet_betas.htm
    Cheers,
    Leo.
     
  3. johnhazel
    Joined: Jun 2008
    Posts: 250
    Likes: 5, Points: 18, Legacy Rep: 60
    Location: Michigan

    johnhazel Senior Member

  4. Guest625101138

    Guest625101138 Previous Member

    Leo
    I have the 20 series running. I will take some time to see what it tells me.

    Rick W
     
  5. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 155, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    It will probably tell you that you could have used a series with fewer parameters in the first place! :eek:

    I just noticed that Series 9 is also included in the beta. That series is like Series 8, but with an additional parameter that allows flared sections.

    Have fun,
    Leo.
     
  6. johnhazel
    Joined: Jun 2008
    Posts: 250
    Likes: 5, Points: 18, Legacy Rep: 60
    Location: Michigan

    johnhazel Senior Member

    Leo,

    Looking at the discription of the 20 parameters I'm having trouble understanding how a triangular transom can appear.

    from your comments in the manual:

    "f15: Section shape of control section 4 (min. 0.0, max. 1.0).
    ... As with Hull Series 1,7,8, and 9, the sections are rectangular if the parameter is equal to 0.0, elliptical if it is equal to 0.5, and parabolic if equal to 1.0."

    But my first try at godzilla run with the 20 parameter(which was a deliberate attempt to get a weird shape being 3.0m/s design speed in 1m water) gave a triangular transom.

    Maybe its the value of cut-off ratio from f19? I'm unsure of the meaning in the f16 to f19 parameters.

    Also how do I get the values f1-f16 after an optimization run?

    Thanks
    John
     

    Attached Files:

  7. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 155, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    The "optimal" shape parameters are at the end of the out.mlt file.

    The triangular transom shape is a result of the way the sections are "lifted" out of the water by the cut-off ratios. It might be best if you play around with Series 9 to see how the extra parameter "lifts" sections out of the water. For example, if you raised a section of a Wigley hull, the portion below the waterline would look more triangular than parabolic.

    BTW, I'm not claiming any particular advantages for the 20-parameter series. It is just one of many that I have. However, it is a good one to play around with because it is consistent with the other families I have released in the public domain versions of Michlet.

    I don't think you guys are ready for the 400 parameter series quite yet ;)
    Leo.
     
    1 person likes this.
  8. yipster
    Joined: Oct 2002
    Posts: 3,486
    Likes: 97, Points: 58, Legacy Rep: 1148
    Location: netherlands

    yipster designer

    i have the statons 20 running too, for tailored and wobly hulls that must be more accurate i guess? al commands working in vista x64 mode exept the maximise and close bittons.. still checking things and a freeschip hull later tonight out... sh*t now i forgot how ctrl alt del is new in vista... but very interesting thread, thanks!!!
     
  9. johnhazel
    Joined: Jun 2008
    Posts: 250
    Likes: 5, Points: 18, Legacy Rep: 60
    Location: Michigan

    johnhazel Senior Member

    Ah! So given f15=1.0 and f19=0.5, what appears to be triangular is actually the bottom half of a parabolic section?

    That's going to be very helpful for me in modeling USCA hulls. thanks

    John
     
  10. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 155, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    Glad to be of help.

    As I said, there are no hydrodynmaic advantages to my hull series, however they are reasonably easy to imagine given the parameters. For example, if you told me that your optimum (3 parameter) hull had shape parameters (0.9,7 0.55, 0.75) I'd know that you are talking about a canoe-like body with roughly parabolic waterlines, elliptical sections and a fair bit of rocker.

    Leo.
     
    1 person likes this.
  11. Guest625101138

    Guest625101138 Previous Member

    Leo
    The series 20 gives an interesting stern shape as I reduce the maximum hull length for a given design speed. The resulting stern looks like it would cause heavy drag.

    Have you ever seen one of these fish tail hulls tested? Are there any published papers on it. They make an interesting option for a length constrained hull but they defy my basic understanding. I am wondering if it is a modelling anomaly or if it actually works.

    Rick W.
     
  12. johnhazel
    Joined: Jun 2008
    Posts: 250
    Likes: 5, Points: 18, Legacy Rep: 60
    Location: Michigan

    johnhazel Senior Member


    I got several wide transom hulls too. This was pretty consistent even when letting Godzilla run to 1e6 Evals.
    I'm playing with Froude numbers of :
    3.0m/s / SQRT(9.8m/s^2 * 5.64m) = 0.40
    Also I am using depth = 1m so the depth Froude number is near critical (1.0) :
    3.0 / SQRT(9.8*1.0) = 0.958
    Maybe the whale tail hulls are a result of optimizing for this maximum squat and drag speed?

    Having any computer model address these critical Froude values might be asking a bit too much. However the plots Leo gave for experimental vs computed values are very encouraging.

    Changing the special hull constraints to a value of 0 did not stop the whale tail either.

    Maybe we are setting a beam min limit and the wide transom is the lowest drag configuration that meets the min beam criteria?
     
  13. Guest625101138

    Guest625101138 Previous Member

    If the hull is not length constrained the tail goes away. It produces a conventional canoe stern. As the length is reduced for a given speed the tail gets wider. I have trouble envisioning the boat in water. I need to see it. Interestingly, when there is no length constraint, the 20 series produces a longer hull than the 7 series but the drag is the same.

    Rick W.
     
  14. johnhazel
    Joined: Jun 2008
    Posts: 250
    Likes: 5, Points: 18, Legacy Rep: 60
    Location: Michigan

    johnhazel Senior Member


    So there is some critical whale tail froude number above which 'Zilla will create a wide stern. This would be easy for me to understand if the hull were allowed to squat and trim in the Michlet. The whale tail would reduce both.

    But Michlet calculates with ship fixed unless you change the trim and sinkage values in the input file. Even then the trim and sinkage are forced and not calculated as a result of the foces from waves on the hull.

    Maybe in the integration of the waves done to calculate resistance, the whale tail tends to flatten the far field surface. I can imagine the boundary condition of the whale tail forcing the removal of verticle components from the water as it flows off the back of the ship.
     

  15. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 155, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    Looks like you're having fun!

    I'm not surprised that you are finding unusual hulls using Series 20. That's just a consequence of allowing the hull too much freedom.

    With even more free parameters (e.g. where each offset is allowed to vary) you can get weird hulls like Ward's Optimum Symmetric Ship. I've also seen hulls that are like narwhals, with a great big spike at the bow and the stern. That is one way of "cheating", i.e. increase the hull length with little additional penalty for surface area.

    The "whale tails" you are getting are probably a new way that Godzilla is finding to cheat. Remember that it can trade off combinations of hydrostatic drag, skin-friction and wave drag which sometimes makes it hard to understand why the algorithm has found a particular combination of parameters. That's one reason I started with a fixed hull shape and just tried to find the optimum length when I started this type of research 10 years ago. The results were unusual, but I could at least understand why the optimal length varied in the way it did. With more free parameters it can become a nightmare.

    Personally, I don't have a lot of confidence in estimates for hulls with large transom sterns: Michell's small hull slope condition is violated, and there is a lot of uncertainty regarding the hydrostatic drag.

    BTW, there is no "integration of waves" in wave resistance calculations.

    All the best,
    Leo.
     
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.