Development of intersection of two cones and two planes.

Discussion in 'Boat Design' started by pdmclean, Nov 19, 2014.

  1. TANSL
    Joined: Sep 2011
    Posts: 7,369
    Likes: 699, Points: 123, Legacy Rep: 300
    Location: Spain

    TANSL Senior Member

    Gonzo, I do not know what you can (you are able to) obtained from a computer program. What I do know, because I have experienced, is that since back in 1980, small and medium shipyards stopped working at 1/1 or 1/10 scale and started doing the "fairing" with computer. Maybe all of them are wrong.
    Computer programs do not work with approximations and they do come as close to fairing, or even more, as full size lofting does.
    I have worked with both systems (scale 1/10 and computer) and I know of what I talk.
     
    Last edited: Dec 2, 2014
  2. DCockey
    Joined: Oct 2009
    Posts: 5,229
    Likes: 634, Points: 113, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

  3. gonzo
    Joined: Aug 2002
    Posts: 16,790
    Likes: 1,714, Points: 123, Legacy Rep: 2031
    Location: Milwaukee, WI

    gonzo Senior Member

    My lofting board was 70'X12' (21.3x3.7meters). A small change of less than 1/2" in the whole length may mean the difference between a beautiful line and an indifferent one. Computer lofting is not capable of distinguishing beauty. The huge amount of ugly boats from the 80's on is a testament to that.
     
  4. NoEyeDeer
    Joined: Jun 2010
    Posts: 983
    Likes: 32, Points: 28, Legacy Rep: 300
    Location: Australia

    NoEyeDeer Senior Member

    Computer lofting is still controlled by humans. Humans, or at least some of them, are capable of distinguishing beauty.

    The modern trend for ugly boats started before computer lofting became widespread. Beautiful boats have been built since then, from computer lofting. There are several examples on this forum.

    Computer lofting is just a tool, like any other. It can be used to produce good results or bad results, depending on how it is used.
     
    Last edited: Nov 28, 2014
  5. TANSL
    Joined: Sep 2011
    Posts: 7,369
    Likes: 699, Points: 123, Legacy Rep: 300
    Location: Spain

    TANSL Senior Member

    The computer neither increases nor decreases the taste, originality, genius of a designer, but gives him many options that did not exist before, eg compare different models or test different versions of a base model. Before he could also do this but at the cost of investing a time generally not available. It seems incredible that we have to clarify these things.
     
  6. NoEyeDeer
    Joined: Jun 2010
    Posts: 983
    Likes: 32, Points: 28, Legacy Rep: 300
    Location: Australia

    NoEyeDeer Senior Member

    The other thing it can do is perspective views, which are not available via traditional lofting. Perspective views are very handy if you are trying to design something aesthetically pleasing. They can enable you to catch things that are harder to pick on basic plans, elevations and profiles.

    Also, 1/2" over 70 feet is a bit over 1/8" in 20 feet. When drafting a 20 foot boat in Delftship, I can certainly tell the difference when a line is moved that far.

    None of this is meant to denigrate traditional lofting*, which is a perfectly good way to go if you want to do it. However, not having to loft at full scale can be a good thing too, and any decent computer app will generate offsets that are fair beyond a level that any human could build to.


    *ETA: For that matter I wouldn't even denigrate people who don't even bother to loft. I admire traditional Norwegian boats and their builders, and those guys just grab some planks and start propping them off the roof of the shed.
     
  7. gonzo
    Joined: Aug 2002
    Posts: 16,790
    Likes: 1,714, Points: 123, Legacy Rep: 2031
    Location: Milwaukee, WI

    gonzo Senior Member

    I do a lot of building "on the eye" too. It is not different from lofting full size. I think that the new generation that is used to computer 3D, can't translate 2D to 3D in their heads. They have a different set of skills though. I chomp at the bit when a program makes jump through hoops to do something that is really easy with a pencil, while they are comfortably designing away.
     
  8. NoEyeDeer
    Joined: Jun 2010
    Posts: 983
    Likes: 32, Points: 28, Legacy Rep: 300
    Location: Australia

    NoEyeDeer Senior Member

    Oh sure, you have to be able to think in 3D if you want to build 3D things, and you have to be able to correlate that with the 2D views since they are still the basis of the drawings.

    And I totally get the jumping though hoops bit. On the other hand, if the app will spit out plate developments so you don't have to figure them out, that's not a bad thing either.
     
  9. TANSL
    Joined: Sep 2011
    Posts: 7,369
    Likes: 699, Points: 123, Legacy Rep: 300
    Location: Spain

    TANSL Senior Member

    The computer does not give you 3D vision. If you do not have it you will not be able to draw with the computer, but the perspective views, etc ,. help a lot. Having 3D vision means you can seamlessly switch from 2D to 3D and vice versa.
    I do not think anyone is capable of designing with computer if he does not know crafting. The real problem is that there are people who know to design with pencil and paper but is unable to use the computer.
     
  10. gonzo
    Joined: Aug 2002
    Posts: 16,790
    Likes: 1,714, Points: 123, Legacy Rep: 2031
    Location: Milwaukee, WI

    gonzo Senior Member

    My main problem is with aesthetics. Hydrostatics or other boring calculations I can happily give to the computer.
     
  11. TANSL
    Joined: Sep 2011
    Posts: 7,369
    Likes: 699, Points: 123, Legacy Rep: 300
    Location: Spain

    TANSL Senior Member

    What other boring calculations?, because that is precisely the most interesting part of the boat designer.
    No computer can perform calculations if you have not previously introduced a model of what you want to calculate. And that's where the difficulties begin from those without spatial vision or those who are terrified by the computer.
     
  12. John Perry
    Joined: Nov 2003
    Posts: 308
    Likes: 53, Points: 28, Legacy Rep: 129
    Location: South West UK

    John Perry Senior Member

    I am coming a bit late into this discussion, but perhaps I can describe how I have gone about determining the shapes to cut from sheet material to form a panel that is part of a boat hull.

    Think of a wire basket defining a model of a chined hull so that each chine is a length of stiff curved wire. Coat the chine wires with paint or tar or something then rock the model back and forth on a flat sheet of paper so that two adjacent chine wires are always in contact with the paper as the hull rocks. Intuitively, I think you can imagine that the shape marked out in paint or tar on the paper is the shape that you need to cut a sheet material panel to fit between those two chines. Of course this is just a 'thought experiment' - this probably wouldn't be a practical method in reality.

    As you try to rock the hull on the paper you may find that three points come into contact with the paper simultaneously, in which case the hull will not rock further. In that case you have a non-developable panel and you will not be able to fit that panel without stretching the sheet material as well as bending it. In general sheet materials will bend a lot more easily than they stretch, so non developable panels are difficult or impossible to fit, although with material such as thin plywood and suitable application of brute force and clamps it may be possible.

    Now come back to the thought experiment, and assume that we have a developable panel. At the start of the rocking movement we could mark the two points on the paper where the chine wires touch the paper. The tangents to the chine wires at these two points both lie in the same plane - i.e. the plane of the paper. This gives a clue as to how a computer program could be written. I first did this back in the 1970s using a fortran program on a main frame computer which did not even have any way of visualising the shapes, input was by a keyboard on a teletype machine and output was by a line printer that output text only. The first requirement is to have a mathematical method of defining the chines as curves in 3D space. There are many possibilities for this, I looked in some maths books and chose the cubic spline method to fit curves through points that I had entered into the program as an input data file. It is best not to input any more of these defining points than are really necessary, otherwise you get a 'lumpy' curve - much the same as when using 'ducks' on a traditional drawing board. Take a starting point on one chine, find the tangent to that chine at that point then iteratively work back and forth along the adjacent chine until you find a point that has a co-planar tangent. I did this by taking two points very close together on the first chine, then moving two points back and forth along the second chine until I had four points that formed a flat quadrilateral in 3D space - you know it is a flat quadrilateral if the two diagonals intersect (or if two opposite sides intersect) - i.e. the formula for the closest distance between the two diagonals (or two sides) gives you an answer close enough to zero. Now move a very little way along the first chine and repeat the process to find an adjoining flat quadrilateral. Keep doing this and you end up with a chine panel defined as a large number of flat quadrilaterals in 3D space and it is then a simple matter for the computer to join these up in 2D space to determine the shape you need to cut out the panel from your plywood or other sheet material.

    I used that Fortran computer program to build my boat. Something went a bit wrong in the bow area - the boat has some bumps there to this day! Sails and rows ok though. Something like 15 years later I re-wrote that program with Pascal and included the hydrostatic calculations, it was easier then because we had computer screens which could display graphics. Then another 10 or so years later I re-wrote it again with MS Visual C++ and attempted to give it a much more user friendly interface, you could even move weights around the hull with the mouse and watch the hull settle to the new waterline and you could roll the boat through 180 degrees and see how the underwater shape and righting moment changed. After that I gave up with that rather nerdy hobby - apart from anything else, general purpose design software had by then become so advanced that there seemed to be little point writing software purely for working with boat hulls.
     
  13. pdmclean
    Joined: Nov 2014
    Posts: 19
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Hobart, Tasmania

    pdmclean Junior Member

    Back on topic

    After a bit of mucking around I got it to this developement:

    http://dl.dropboxusercontent.com/s/0l9bxm1l1f29m82/image.jpg

    http://dl.dropboxusercontent.com/s/vcv8obuy252tc66/u3NdI2r_Ru6s6KnjINQuLQ.png

    The left line is the bottom of the hull and the right will be level/straight when the shape is joined to another along the bottom.

    Note that this is just one out of a multitude of hull forms paramaterised in my design.

    Also, by cutting this by an arc parallel to the right arc I can get a flat-bottomed boat. I'm working on matching this shape with the intersection of two other cones to get a v-shaped bottom.
     
  14. NoEyeDeer
    Joined: Jun 2010
    Posts: 983
    Likes: 32, Points: 28, Legacy Rep: 300
    Location: Australia

    NoEyeDeer Senior Member

    Ok. I have to wonder, what is the point? Both pictures are the same, and don't look very interesting.
     

  15. wolle
    Joined: Jan 2012
    Posts: 19
    Likes: 3, Points: 0, Legacy Rep: 36
    Location: berlin, germany

    wolle Junior Member

    i missed the thread, same thoughts some years ago.

    not advanced enough for specific purposes, i think. I struggled with delphtship to get a simple developable hull, but this sotware is not accurat in this respect nor is it straightforward for this purpose. With low number of controlpoints, best for a simple hull, the surfaces on the bow become inevitebly curved inward. with more controlpoints, the meaning of "control" is vanishing.

    my approach: the keelline (curve of any kind) is segmented in a polygon (smaller segment size, more accurate results), start with a choosen angle of the first face aft outside the hull. a controlline (curve of any kind) is added to determin the "conicality" betwen two single faces. the chineline (curve of any kind in planview), will cut the faces, a line in profileview will do so to cut the transom (only necessary for a transom hull). the resulting surface is completely developable.
    same procedure with the second chineface (if a double chine is the wish).
    the sheerface smilar, except that a sheerline (curve of any kind) in profileview will cut the face to shape.

    sounds complicated, but it is very convenient to use, particular to get the bowsections to get shaped. (if there is interest, i will describe more in detail)

    its a script written (in fact pascal and still full of bugs) in a CAD-sofware not used for boatdesign.


    wo
     
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.