Dumb Question

Discussion in 'Gallery & Designs' started by rmks2000, Aug 1, 2006.

  1. rmks2000

    rmks2000 Member

    If you are designing a four-sided tube using a simple drawing program, do you

    a. Create a large rectangle and add three equidistant 'bisecting' lines within it? ex. a 1 inch wide square tube would consist of a 4 inch wide rectangle an lines at the 1, 2, and 3 inch marks


    b. Create a single rectangle and daisychain three more to it?
    ex. create a 1 inch wide rectangle and add another 1 inch wide rectangle adjacent to it and so on. (Of course the two adjacent lines would actually occupy the same space so as not to introduce an additional line thickness)

    The same question applies to creating rudders, wings, etc. Opinions?

    GEEDUBBYA Active Member

    Howdy Rmks,

    Well, you lost this countryboy at "equidistant" lol. How bout I just draw ya a tube? lol
    But seriously if I follow you coreectly, both methods you mentioned would achieve the same result wouldnt they? If so, then the question you are asking would be "what technique do you use?"
    As for myself, I have never really given it much thought, my program, "anim8or" takes care of that type stuff for me. you can search for anim8or on google and download it for free.

    Sorry I couldnt be of more help,

    Greg aka GW
  3. thewoodengraver

    thewoodengraver Active Member

    I draw one rectangle with a black outline , then copy, then paste three times onto the same image. All four sides will be exact.
  4. thewoodengraver

    thewoodengraver Active Member

    oh, daisy chain..(finally sunk in) yes, thats the way.
  5. Willja67

    Willja67 Member

    To intruduce complexity to the issue it depends on what you want. For example in Corel Draw that I use if I want each side to be a different color then the daisy chain method is best if the whole thing is the same color make one rectangle with three lines inside the rectangle. As one signature used to read "There is no right way only what works".
  6. lizzienewell

    lizzienewell Member

    With CorelDraw I would use the graph drawing tool(located on the toolbar with the polygon tool and the spiral tool) with it set to do four columns and one row. Once this is drawn the graph can be ungrouped and it becomes four rectangles.
    I usually overlap by a full side so I'd do 5 columns instead of 4. To keep it symmetrical I might make it out of 2 parts each made of 3 rectangles. This would depend on where I needed strength and on how much symmetry I want. The most symmetrical would be to make 4 parts each with 2 rectangles. Each side then would be two layers thick with lots of glue surface and would be exactly the same. This would avoid the problem for adjusting for the thickness of the paper.
  7. rmks2000

    rmks2000 Member

    Thanks for all the replies. It's interesting to see different techniques even on a simple design. When doing the "daisychain" design, has anyone experienced thicker lines due to the printer printing the overlap lines twice in the same spot?
  8. thewoodengraver

    thewoodengraver Active Member

    No, I use paint shop pro, and defloat the layers when finished.
  9. rmks2000

    rmks2000 Member

    I have PSP also. It's been awhile since I played with it, and I can't recall if I had multiple layers.
  10. Dumb question time. How do you design a 3D card model in a 2D program like PSP or Corel? I'm starting to think in terms of designing (way ahead of my current position on the learning curve) and so MilkShape and Metasequoia are on my agenda for investigation. But I've got PSP, Illustrator and even TurboCAD (2D and 3D). What do/don't I need to do this sort of design work?

    Cheers --- Larry
  11. TheWebdude

    TheWebdude Just a Member

    2D? No software required. Just Pencil, paper and a good knowledge of geometry and math. (Eraser is optional dependant on your confidence level)
  12. 46rob

    46rob Member

    I've never designed a model using anything but Photoshop, paper and pencil and a bit of modeling technique. Math needed is basic fractions, and the ability to accurately lay out rectanles and cones. Look on the FG website to see how I designed the fuselage for the FJ-1 Fury, or in the PARTS BIN section of this forum, under TOOLS.
  13. Darwin

    Darwin Member

    This is reaching back several years (more than 40, actually), but remembering back to the tin-bending lab I had to take in college, when you put a 90 degree bend in a flat surface, you loose a bit of the linear dimension of the parts in making the bend. If I remember correctly, the linear dimension of a "face" part needs to be increased by about 0.7 times the thickness of the material you are fabricating it from to account for the circumferential distance around the outside of the bend. As anyone who has run track knows, the outside lane takes more strides to get around than does the inside lane....this is why runners have a staggered starting point iln order to equalize the number of strides needed to complete a race. As an example, if you want to build a square tube that is 100 by 1000 pixels on each face, and the paper thickness equates to 10 pixels, the width of the part you lay out on the paper should be increased by 7 pixels for each 90 degree bend it abutts. When the part is laid out on the paper, the width should be 114 pixels (two bends to account for), not 100 pixels, in order to compensate for the dimensional "loss" that occurs from going around the outside radius of the bend. This is why, when designing a superstructure of a ship that consists of several boxes laid end to end, the as built length of the completed superstructure is not quite as long as the location outline on the deck. When working with paper, we tend to ignore the affect of the paper thickness on finished size of the model....however, particularly for heavier card (such as 110 lb), the small dimensional inaccuracies of ignoring the paper thickness tend to all add up to produce the glaring 1/4 inch gap in the finished prototype. I've found that a basic textbook for sheet metal fabrication is a good addition to the serious paper model designer's library. In "simple" structures (ones made up from only a few primitive objects), ignoring the structural material thickness usually won't make a noticible difference. The more complex the shape, the greater the error becomes when they stack up on you during the build (Murphy's law says that such errors are all additive).

    In short....old school card model designing is every bit as complicated as designing the real thing. Fortunately, the medium we work with is somewhat more forgiving of small tolerance errors than is sheet metal.
  14. 46rob

    46rob Member

    You're on the money on that post. Thirty years of my life were spent as an aircraft sheetmetal mechanic and I'm currently up to my old tricks again in the restoration hangar at the National Museum of Naval Aviation here in Pensacola. A basic knowledge of sheetmetal layout is incredibly useful....in fact...I doubt if I'd be able to design as well without that training.
  15. I"m not sure what happened here. I posted a msg in an existing thread, where guys were talking about designing parts using PSP and Corel (not pencils, not erasers, and not prototyping the parts and rescanning the results). That was the basis for my question.

    Sure, and prototyping your parts, followed by a rescan sidesteps the use of a computer for the true design aspects of the process. By the way, thanks for your Design without CAD thread. I read it a few days ago and it made perfect sense.

    But I would contend that you're not using Photoshop to do the design. You're using paper to determine the shape of complex curvatures. You're using paper to work out fit problems. All great and probably necessary if you're using a 2D graphics program for layout. Nothing wrong with that. I was just curious about what the other guys were talking about as it seemed that they were actually generating parts and I've never seen enough tools in either PSP or Corel to do the math precision required.

    Thanks for the responses. Sorry for the screw up on my part, though I'm not sure what I did to hijack the thread heading :)

    Cheers --- Larry
  16. TheWebdude

    TheWebdude Just a Member

    My answer still holds whether it's a 2D computer program or "by hand no tech".

    You'll need the math skills. (what's the circumference of that circle for the cylinder I need to make in my 2D program?:grin:)
  17. 46rob

    46rob Member

    My F2Y-1 design ws done strictly by laying out cones of the proper size, and removing the appropriate sections. Nothing was ever done on paper first--although minor corrections were made on initial builds. There's lots of ways to skin a cat--just remeber that drawing programs, like Paint or PSP won't do anything automatically--it takes some knowledge of drawing--be it CAD, drafting plus a good handle on geometry and basic metal layout techniques--to produce a realistic model. Photoshop and other programs have rulers built in--and sma4rt folks can figure out ways to use them to advantage in layout. Photoshop CS2 has a measurting tool, that measures not only length, but angles and x-y coordinates of a line. (GIMP has similar tools). These will go a long way, along with some basic geometry. Kell Black has another "Designing wothout CAD" article on the FG website, that goes into detail on layoing out cones and segments.

  18. And that's how you're taking photos of Pixar characters and turning them into card models (grin)? How many 'math skills' are you using? I don't mean to be argumentative here but you're the guy that made me aware of MilkShape with you comments elsewhere.

    As for math skills, I've got them. As for 2D CAD skills, I've got those too. I've also scratchbuilt a lot of models by a cut-fit method. But guessing my way to how a fillet blends a wing to a fuselage without either doing it with material or with software that can blend faces or vectors seems hard to fathom. Guess I just don't have enough imagination for that :)

    Cheers --- Larry
  19. 46rob

    46rob Member

    Actually--I do a lot of that--guessing--from experience in building card models, (I've built several hundred), you know what the basic shape should be--dimensions are the easy part-- and I can come pretty close to the shape I need. A bit of trial fit---and that's about it.

    the whole idea is to start with something simple, design and build it as a confidence excersise, and go on from there. Each time I start a new design, i discover new tricks, shortcuts, ways to accelerate the process....to me the dull and boring part is coloring the finished model..By the time I get to that point--I've pretty well lost interest and am already designing the next one in my head. That's an important part of the process, btw. You need to know in advance, how the sections will fit together, how the wing will mount, whether the fin will be integral with the fuselage or a separate bit---on and on--Once those things are thought out--the desing process usually flows smoothly. I say usually, because sometimes things take a lot of experimentation to get parts that not only fit well, but are easy to assemble and realistically depict the shape. I spent a whole summer playing with the T-33's nose, for example. In between, I worked on other things...but kept revisiting it until I got the shape I wanted, that would build easily.

    Attached Files:

  20. I bought a copy of your FJ-1 so I can compare it to your 'design without CAD' tutorial. Have only looked at it briefly but will study it later tonight. I guess what has me puzzled is why people prefer to use PSP, Corel, or Photoshop rather than using an actual CAD program to do this stuff. You're all suggesting that arithmetic is important and clearly so is measurement. Both are accommodated much better in CAD. I must be missing something...probably a lot of somethings :)

    I couldn't find that. It must come with one of the kits like yours does.

    In the end, people are spending a lot of money on programs like Rhino. They must be doing it for a reason. Personally, I'm more comfortable with the methods you and Webdude talk about than 3D surface generation via software.

    Cheers --- Larry

Share This Page