D 520 Rhino Build

Discussion in 'Gallery & Designs' started by fimdan, Jan 5, 2006.

  1. wunwinglow

    wunwinglow Active Member

    Joined:
    Jan 18, 2004
    Messages:
    1,093
    Likes Received:
    1
    ooofergoodnessakes!!!!!! If you design a shape that won't unroll, that's not Rhinos fault!!! Its your fault for not designing the shape APPROPRIATELY!!! If you are designing a model for constructing in paper and card, you are limited to using shapes that are either flat, or curved in one direction only. Anything that is curved in two directions will NOT unroll, no matter what software you apply to it!

    When you convert a moulded shape to a mesh, you are approximating the shape to a mesh composed of FLAT triangles. When you run this sort of mesh shape through Pepakura, it splits the edges of some triangles, and flattens the remaining joined edges until the whole shape is opened out into one plane.

    When Rhino unrollsrfs a surface, it does a similar thing, geometrically, but across the whole span of the part. It turns the surface into a load of skinny triangles, then opens this out like a paper fan. It does this mathematically of course, there is no mesh object involved. It will ONLY work if the surface has one degree of curveture, and that curve direction is perpendicular to the long skinny triangles.

    Rhino has another command, FlattenSrf, where you can type in the sample spacing (the width of the skinny triangles) so you can try different values to see what changes this makes on the resulting flattened surface.

    Expander does just what it says on the tin; it EXPANDS the surfaces where needed to get them flat! By allowing this distortion, in a controllable way, it gets unrollable surfaces flat. But they are distorted! This isn't something that Rhino is somehow 'failing' to do, rather than Expander giving you the ability to 'cheat' the limitations of geometry. This of course is very useful if you are using materials that will also 'cheat', in other words stretch, but unless you are into the kind of paper moulding that Gil excells in, it is beyond the capabilities of paper modelling. And if this is a problem, sorry, but you should ditch paper and start using plastic sheet and a vacforming machine. Which is PRECISELY what a lot of modellers do when it comes to canopies.

    Tim P
  2. fimdan

    fimdan Member

    Joined:
    Jun 9, 2004
    Messages:
    77
    Likes Received:
    0
    Some more pictures.
    I am still trying to finish the wing area.
    There are so many little things to remember about. Layers are extremely helpful with managing all these different objects.

    [​IMG]

    [​IMG]

    Also, at what point in your design process do you actually print and try to assemble the model?

    FimDan
  3. wunwinglow

    wunwinglow Active Member

    Joined:
    Jan 18, 2004
    Messages:
    1,093
    Likes Received:
    1
    Fimdan

    start test assembly on your frames at the very least right now! You don't want to have to redo a load of skinning work because the frames need adjusting!

    Tim P
  4. fimdan

    fimdan Member

    Joined:
    Jun 9, 2004
    Messages:
    77
    Likes Received:
    0
    Ok, So I started printing. So far, so good. Found one minor error that was easy enough to fix. I also found it beneficial to hide all isocurves before exporting to ai format (Object Properties -> Show surface Isocurve).

    The process of getting things out of Rhino is a little time consuming. Maybe some of you have a patent for doing this efficiently? The problem is that some surfaces must be unrolled; others are already flat but not facing in thesame direction. Nothing difficult, but I would rather invest my time in modeling, not moving things around.

    Also, I am really starting to see the importance of taking material thickness into consideration. This model will not be that large so even 1 mm makes a difference.

    Last thing, scale - 1:33.

    Don't go away... We will be right back...

    FimDan
  5. wunwinglow

    wunwinglow Active Member

    Joined:
    Jan 18, 2004
    Messages:
    1,093
    Likes Received:
    1
    Oo, this is getting good! Don't forget to post pics of the test assemblies, if possible.

    Rhino v4 will have much enhanced unrolling capabilities, although their website seems to have some glitch at the moment regarding logging on, so I can't get a url for you. However, it will onroll a load of adjacent surfaces as one piece; true unfolding! Wahoo!

    Tim P
  6. Willja67

    Willja67 Member

    Joined:
    Feb 12, 2005
    Messages:
    907
    Likes Received:
    0

    Why not print straight out of Rhino until you get to the painting stage?
  7. fimdan

    fimdan Member

    Joined:
    Jun 9, 2004
    Messages:
    77
    Likes Received:
    0
    Quick update,

    I have printed some parts and it turns out that the glue I was using stretched the paper so much that nothing fits the way it should. I have to investigate this more.
    Besides that I have also started a new job and that keeps me busy now. This will change once I know what I am doing.

    FimDan
  8. wunwinglow

    wunwinglow Active Member

    Joined:
    Jan 18, 2004
    Messages:
    1,093
    Likes Received:
    1
    Fim, I found photo papers were pretty good at resisting stretching, although they are a bit more expensive than normal inkjet papers. Using a solvent based glue rather than a water based one helps too. The glossy papers give a great finish for gloss painted models, such as racing cars or sports aircraft.

    Good luck with the new job! Although I've been at mine now for eight years, I still don't have a clue what is going on....

    Tim P
  9. fimdan

    fimdan Member

    Joined:
    Jun 9, 2004
    Messages:
    77
    Likes Received:
    0
    Weekend is here, work continues

    Hello,

    Weekend is here. I was able to sit down and print those parts again. I used a different glue this time, Elmer's Glue All. I was happy to see that everything fits so far.
    Here are some pictures.

    [​IMG]
    [​IMG]

    FimDan
  10. fimdan

    fimdan Member

    Joined:
    Jun 9, 2004
    Messages:
    77
    Likes Received:
    0
    Ok, major setback.

    I have not been careful enough with my formers and it turns out that I have to redo them along with pretty much everything. It will not take that much time but discovering things like that is pretty discouraging. Great learning experience though.

    FimDan
  11. Willja67

    Willja67 Member

    Joined:
    Feb 12, 2005
    Messages:
    907
    Likes Received:
    0
    Do you mind letting us in on what went wrong so we don't duplicate your mistakes?
  12. fimdan

    fimdan Member

    Joined:
    Jun 9, 2004
    Messages:
    77
    Likes Received:
    0
    Explain

    Ok so here it goes.

    I did not design my formers carefully enough. I made the curve go below the end point which resulted in something like the bottom curve in the picture. The picture has it a little exaggerated but even a little of this makes a difference when you offset this curve. The top curve is the bottom curve offset by some distance. The whole problem is that the end of the offset curve is not aligned vertically with the original curve. That is wrong. Even a little bit of misalignment makes it difficult to properly modify objects (trimming, joining, etc).

    [​IMG]

    In order to avoid this problem, make sure you design your former so that the end point of your curve is the lowest (highest) point on your curve. The picture below shows what the curve should look like. The key here is to get the last control point to be exactly aligned with the end of the curve.

    [​IMG]

    As you can imagine, redesigning all the formers requires me to fix the skin. Redesigned skin means redesigning everything :(

    I hope I explained it well. If you have any questions, feel free to ask.



    FimDan
  13. wunwinglow

    wunwinglow Active Member

    Joined:
    Jan 18, 2004
    Messages:
    1,093
    Likes Received:
    1
    Phew! Glad you spotted that so soon. This illustrates exactly how you HAVE to be precise about what you are doing, because while Rhino will work very accurately, if you ask it to do something that you don't actually want, you will get very accurate mistakes! I'd always recommend a short design/print/check assemble cycle, because with Rhino you can't go back down the design tree and change a feature and have everything update. This is a feature of parametric modelling programs, which are all much more expensive than Rhino, with the exception of Alibre, that free program that caused such a stir last year. https://www.alibre.com/xpress/software/alibre-design-xpress.asp for more details.

    Tim P
  14. Maurice

    Maurice Member

    Joined:
    Jan 16, 2004
    Messages:
    509
    Likes Received:
    2
    FD (or whatever)

    In view of the limited comprehension shown in some of the responses around here I hesitate to offer a further contribution, but anyway here's an attempt to define the source of the problem.

    It arises because the end point of an offset to a curve will always lie at a point which is on a line perpendicular to the tangent at the endpoint of the curve.
    Indeed to be a proper offset any point on it must lie at a point which is on a line perpendicular to the tangent at the equivalent point on the curve.

    Where the section curves are supposed to be continuous across the centreline you need to ensure that the ends of the curves defining the half section are everywhere perpendicular to the centreline.
    This sounds simple enough but needs care to achieve given the limitations that computers encounter when drawing curves.
    You need to trim or extend the offset in cases where the section curve is supposed to meet the centre line at other than a right angle.

    Your trouble detecting the mistake highlights a problem affecting both precision and productivity that I'm having with Rhino.
    How do you get it to zoom to a point on the wheel? It says it can but doesent. (I've already updated the driver.)

    Cheers
    Maurice
  15. wunwinglow

    wunwinglow Active Member

    Joined:
    Jan 18, 2004
    Messages:
    1,093
    Likes Received:
    1
    Dumbcluck here. Use the box zoom thingy doodah. Sorry, being stupid, I just forgot its proper name. And my own name, apparently.

    Happy, the idiot...

    Dumbcluck
  16. Willja67

    Willja67 Member

    Joined:
    Feb 12, 2005
    Messages:
    907
    Likes Received:
    0
    What better way to increase your comprehension than by asking a question? The alternative is to remain stupid and ignorant forever.

    I heard this story about Admiral Rickover. He was responsible for the buildup of the U.S.'s nuclear submarine force. A associate of his remembered him from his days way before he got to that lofty position when he annoyed the instructor in the nuclear physics class so badly due to his constant questions(he was the only one asking any) that the instructor finally said that he would set up a time for anyone who wanted tutoring to come and be helped. When that time came half the class showed up and Rickover turned to the guy who related the story and and said "I must not have been the only idiot in the class".

    So I'll keep asking ignorant questions even if I'm snubbed for it and hopefully one day I'll only be moderately ignorant instead of hopelessly ignorant.
  17. Maurice

    Maurice Member

    Joined:
    Jan 16, 2004
    Messages:
    509
    Likes Received:
    2
    Tim

    I never intended it should become a permanent appellation, however you know what they say about " If the cap fits ....".
    There are no known patents or copyright.

    Meanwhile that zoom thingy doodah is a pretty severe limitation in my book. Pity really.

    Cheers
    Maurice
  18. wunwinglow

    wunwinglow Active Member

    Joined:
    Jan 18, 2004
    Messages:
    1,093
    Likes Received:
    1
    Maurice,

    I can't replicate this now! When I zoom with the mouse wheel, I can go from metres across to 0.001mm entirely with the wheel. I am wondering if this changes with the complexity of the model? I have noticed it a couple of times, but box-zooming works instead. Frankly, I think there are more important things in this world to loose sleep over....

    Dumbcluck and proud....
  19. fimdan

    fimdan Member

    Joined:
    Jun 9, 2004
    Messages:
    77
    Likes Received:
    0
    Just a quick update...

    I have decided to diverge a bit from the routine task of drawing curves and, for now, shifted my interest to rhino scripting. Why? This is a very powerful feature of rhino that may prove useful in the future. The more I know about the software, the better.
    Right now, I am trying to write a few VB scripts to automate some mundane tasks (like offsetting curves).
    Questions to all:
    Have you ever used scripts in your Rhino projects? If so, what did they do? Is there a task, or a series of tasks, that you would like automate to simplify the design process of paper models in Rhino?

    Awaiting your comments….

    FimDan
  20. wunwinglow

    wunwinglow Active Member

    Joined:
    Jan 18, 2004
    Messages:
    1,093
    Likes Received:
    1
    Fimdan,

    I did a bit when I was using Rhino 2, for example importing standard aerofoil section data and generating a curvethroughpoints. Worked in the blink of an eye! I love the way you can create your own buttons and link them to your own instructions/scripts too. If you have repetitive commands, this is another example of how truly useful Rhino is as a design tool. There are some excellent scripts and tutorials on the www.rhino3d. com website, apart from the program help files.

    My script stuff was on my www.tjpgraphics.com website, on the http://www.tjpgraphics.com/html/a_wing_in_rhino.htm page, although it is a bit out-of-date now. The principle is much the same however.

    Tim P (Dumbclucks of the World, Untie!)