Best Marine Design Software for Hull Modeling? (2007)

Discussion in 'Software' started by Admin, Jan 25, 2007.

?

Which program(s) do you use as your primary hull design software?

  1. Autoship (Autoship Systems Corporation)

    13 vote(s)
    8.1%
  2. Catia

    4 vote(s)
    2.5%
  3. DefCar (DefCar Engineering)

    2 vote(s)
    1.3%
  4. Delftship

    20 vote(s)
    12.5%
  5. Fastship (Proteus Engineering)

    7 vote(s)
    4.4%
  6. HullCAO (HullCAO)

    1 vote(s)
    0.6%
  7. Hull Form (Blue Peter Marine Systems)

    1 vote(s)
    0.6%
  8. Maxsurf (Formation Design Systems)

    51 vote(s)
    31.9%
  9. MultiSurf (Aerohydro)

    10 vote(s)
    6.3%
  10. Naval Designer

    1 vote(s)
    0.6%
  11. Prolines (Vacanti Yacht Design)

    9 vote(s)
    5.6%
  12. ProSurf (New Wave Systems)

    9 vote(s)
    5.6%
  13. Rhino (Robert McNeel & Assoc.)

    55 vote(s)
    34.4%
  14. SeaSolution

    0 vote(s)
    0.0%
  15. TouchCAD

    5 vote(s)
    3.1%
  16. Other (please post below)

    14 vote(s)
    8.8%
Multiple votes are allowed.
Thread Status:
Not open for further replies.
  1. Andrew Mason
    Joined: Mar 2003
    Posts: 397
    Likes: 18, Points: 18, Legacy Rep: 206
    Location: Perth, Western Australia

    Andrew Mason Senior Member

    Leo

    You don't seem to be distinguishing between a table of offsets and a NURBS surface - they are not the same. If you supplied me with a table of offsets for the DTMB 5415 destroyer I would ask you why you did not supply me with the NURBS surface definition for the same hull which can be downloaded from the web (see rendering below).

    The NURBS surface definition can have its surface area calculated extremely accurately directly from the NURBS surface - we don't use or recommend using sectional data for this, so grid spacing is irrelevant.

    If you supplied me with a hull that you had only as a table of offsets, then it would be necessary to fit a surface to it to accurately calculate hydrostatic data. The accuracy of the fitting would be subject to the quality, quantity and spacing of the offset data and the fitting is not a trivial problem. However it is a separate problem from the question of whether a NURBS surface area can be calculated to any desired precision, which it can.


    If you define a 1 metre radius sphere in Maxsurf and calculate the surface area at the standard high precision(i.e. 64 x 64 isoparametric panels), you get an answer of 12.558565, to 8 significant digits. The correct answer to the same precision is 12.566370. So to four significant digits Maxsurf gives the correct answer, to six significant digits an error of 0.06%.

    This level of precision is bettered by most Maxsurf surfaces as they typically contain a lot less curvature than a sphere. The flatter the surface the more accurate the area estimation. So no, I don't see the estimation of surface area by Maxsurf as a problem.
     

    Attached Files:

  2. Andrew Mason
    Joined: Mar 2003
    Posts: 397
    Likes: 18, Points: 18, Legacy Rep: 206
    Location: Perth, Western Australia

    Andrew Mason Senior Member

    Hi Martijn

    We don't use sectional data for surface area in Maxsurf because it is not at all accurate for surfaces that are inclined relative to the vessel centreline. For cylinders it is accurate as the distance between stations is the same as the surface length between stations, however if the surface was crossing the stations at 45 degrees the surface length would be 1.4142 * the distance between sections. As you can see this will produce a significant error in the surface area calculation. Many hydrostatics programs seem to ignore this error, however it can be significant in vessels that are very round in the bow and/or stern.

    A simple test case for a hydrostatics program is to define a cone with its axis of symmetry running longitudinally and calculate the surface area. Programs using sectional integration will get significant errors, programs that integrate the surface directly will approximate the analytical answer. For a cone of 1 m radius and 1 m length, Maxsurf gives the correct answer to 4 significant digits.
     
  3. 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 estimates, Andrew.

    I'm not disputing that Maxsurf can calculate areas very well given
    good input data, but my point is still that it is difficult to
    assess accuracy objectively and scientifically, which is what the
    original question was about.

    Unfortunately, sometimes offset data is all that a user is able to
    supply and, as you say, the fitting accuracy is then not a trivial
    problem.

    And you are absolutely correct to say that surface area estimates
    are more accurate for most real hulls (as opposed to spheres etc)
    because they have extensive flat regions.

    By the way, here are some estimates of the surface area (and
    some other coefficients at Fn=0.28) for the DTMB 5415 in
    non-dimensional form reported at the Gothenburg 2000 Workshop on
    Numerical Ship Hydrodynamics. As you can see, there are some
    significant differences between the various codes.

    CODE S/L^2 CT*10^3 CF*10^3 CP*10^3
    EXPERIMENT 0.1490 4.23 - -
    FINFLO 0.1558 4.42 2.81 1.61
    MGSHIP 0.1485 4.47 3.23 1.24
    CFDSHIP 0.1550 4.36 3.06 1.30
    ICARE 0.1486 4.39 3.18 1.21
    UNCLE 0.1520 4.08 3.05 1.04
    CFX 0.1517 3.94 2.95 0.98

    (Sorry if the table is not aligned well. I'm not sure how to do it on this board).

    Incidentally, I have seen a report that used 1 billion elements
    for this particular hull. I seem to remember that the hydrodynamics
    were not all that impressive when compared to experimental results.
    This raises the uglier question, how reliable are the experiments
    against which ship researchers are supposed to compare their work?

    You might be aware that this hull is being tested at about 30 towing
    tanks around the world. Hopefully the accumulated data will enable
    some better comparisons in 2008/2009 when all the tests have been
    completed, compiled and published.

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

    Leo Lazauskas Senior Member

    Thanks. I didn't realise that some commercial codes used that system. I have seen some published hydrodynamics papers a long time ago that justified ignoring the y-component because it was consistent with other simplifying assumptions.

    Regards,
    Leo.
     
    1 person likes this.
  5. Martijn_vE
    Joined: Apr 2005
    Posts: 254
    Likes: 24, Points: 0, Legacy Rep: 401
    Location: Netherlands

    Martijn_vE Marine software developer

    I couldn't agree more Andrew. From your reply to Leo I understand Maxsurf uses surface paneling too to calculate the area of surfaces. Provided you use enough panels you can indeed calculate the area up to any given accuracy. The problems with sectional data is not restricted solely to the inclined surfaces you mentioned. Also in specific areas of the hull where the contourline changes very rapidly, such as the bulb or the ending of (twin) gondolas, it can cause errors (with respect to the exact surface area) if not enough stations are provided. This is simply caused by Simpson formula.

    But this problem is not restricted to calculating surface area. The same goes for calculating volumes. Does this mean you use the surfacepaneling for that too?
     
    1 person likes this.
  6. Andrew Mason
    Joined: Mar 2003
    Posts: 397
    Likes: 18, Points: 18, Legacy Rep: 206
    Location: Perth, Western Australia

    Andrew Mason Senior Member

    Martijn

    The major potential error is with wetted surface, the volumetric values don't suffer from the same problem, so conventional sectioning is sufficiently accurate. With Maxsurf and Hydromax we use a large number of sections, 200 at high precision, along with trapezoidal integration longitudinally. Each section is refined to a defined minimum chordal error, typically 0.2 mm, so error values for each section are very low.

    As an example, for a 1metre radius sphere, which again is a worst case due to its curvature, Maxsurf calculates a volume of 4.18762, whereas the correct value is 4.18879. So the error is about 0.03%. Errors for ship shapes would be lower.
     
  7. Andrew Mason
    Joined: Mar 2003
    Posts: 397
    Likes: 18, Points: 18, Legacy Rep: 206
    Location: Perth, Western Australia

    Andrew Mason Senior Member

    You are talking about CFD and tank testing here, so we are straying from talking about pure geometric properties into the realm of black magic. I agree that there is still a lot of scope for improvement both these areas.
     
  8. Leo Lazauskas
    Joined: Jan 2002
    Posts: 2,696
    Likes: 155, Points: 63, Legacy Rep: 2229
    Location: Adelaide, South Australia

    Leo Lazauskas Senior Member

    The surface area estimates (S/L^2 in the table) of some of the codes could do with improvement. I didn't expect to see a 5% spread between the estimates by what were then considered as the best codes in their field.

    Leo.
     
  9. mbdezign
    Joined: Jan 2007
    Posts: 3
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Sarasota, Fl

    mbdezign New Member

    The software we use is Unigrapics NX in combination with multisurf. I've tried Rhino & Pro-E for engineering purposes along with Studio Max, Maya, Cinema 4D and a small host of various 3D Modelling programs for other graphic solutions. Unigraphics is the best I've used. However, the price is too much. For the money, buy Rhino, unless you're loaded.
     
  10. Gooitzen
    Joined: Dec 2005
    Posts: 2
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Australia

    Gooitzen New Member

    Just a few quick answers for you here:
    The Maxsurf suite is what used to be MacSurf. When the program was converted over to Windows OS, the name changed to Maxsurf about 12-15 years ago. Maxsurf and all modules in the suite have a consistent graphical user interface which makes it very easy to use. Maxsurf is no longer developed for the MacIntosh computer but many Maxsurf users run the software on MacIntosh in Windows emulation mode or virtual windows.

    The Maxsurf suite is developed in Australia where the majority of boatbuilders and designers use the software. Since the majority of the Australian design projects are multihulls, Maxsurf has developed good support for multihull vessels. For example: in Maxsurf you can define the demihull centreline location which is then used in Hullspeed to correctly interpret the hull for resistance (using the slender body method). Hullspeed also has an algorithm that estimates the form factor coefficient (1+k) for Catamarans (Molland). Hullspeed also allows you to calculate the free surface wave pattern which gives good feedback on the demihulls wave pattern interference at any speed.

    Structural analysis of Multihulls can be done using Multiframe Marine also from the developers of Maxsurf. See their latest newsletter on their website for more information: http://www.formsys.com/maxsurf/newsletters/maxsurf---aug-2006
     
  11. tri - star
    Joined: Feb 2007
    Posts: 87
    Likes: 1, Points: 0, Legacy Rep: 3
    Location: Vancouver, B.C., Canada

    tri - star Junior Member

    Thankyou Gooitzen.

    A most full and complete answer to my request.
    Particularly happy, that I can still maintain a foot in the Mac universe.

    The Maxsurf guy(s) related similer info. to us - although, due to
    diplomacy or modesty. Had not emphasized, this easy interface with
    the Kingdom of Gates.

    Your positive review - will be presented at the next meeting with
    our persimmonous, tight fisted, accountants.....
    Swear it was their own money they are spending !

    Cheers !
     
  12. MauroB
    Joined: Mar 2007
    Posts: 6
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Argentina

    MauroB Junior Member

    Maat is a good software too.
     
  13. dincerd
    Joined: Aug 2004
    Posts: 17
    Likes: 0, Points: 1, Legacy Rep: 10
    Location: Turkey

    dincerd Junior Member

    I have prepared my projects as below;

    Hull design: Maxsurf
    Engineering: Hydromax
    Projects & Plans: Autocad
    3D Modelling: Rhino
    Rendering: 3D Studio Max

    All programs have different abilities for solving problems, so in my opinion, it is a must to use correct programs for getting best results.
     
  14. schrambo
    Joined: Jun 2007
    Posts: 3
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Western Australia

    schrambo New Member

    I use autoCAD just to get a quick 2D drawing done then import it into Rhino 3 and model things from there.

    Admittedly I have only just started and i'm still learning as I go along. I plan to purchase rhino v4 soon. Its worth it for all the new extra features it has. Can't wait :)
     

  15. barthautala
    Joined: Jul 2007
    Posts: 22
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Traverse City, MI

    barthautala Junior Member

    Couldnt you just pop your hull surfaces into FEA like COSMOS(Solidoworks) or Ansys and get all your structural data from that? You can make your surface mesh as fine as your computer can possibly handle..........you can even use Floworks to determine Cd.

    Ive got a Q.....Im familar with 3d modeling and engineering processes. But Im just an amateur boat builder and I want to design my own runabout. Which software is good for that? I just want to model a boat and have it crap out some prints for me such as ribs and dimensional data........

    Like Van Dams boat Alpha Z was modeled but in what?
     
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.
Thread Status:
Not open for further replies.