Design: Straight Lines Vs Curves

Hello Escobar (hope I'm right, anyway),

Thank you for the insight into issues I raised in "Design: Straight Lines Vs Curves." I actually raised issues, not questions. So I will consider your responses as insights. You addressed the issues very well.

I am only trying to focus on something like, say, "naturalism": "Design with nature", sort of, or "organic design" as you rightly put it. I seem to understand that nature itself (I mean the vast expanse of creation) only supports and tolerates a straigh line concept. In fact, the straight line is simply a starting point in forming the curve. When two points make a straight line, they are only "waiting" for the third, fourth...points to swing in a curve as desired. So, the straight line itself does not exist as a "structure", or an end, but a means to an end.

As to your question about designing a "house's floor in a curved surface", it is always possible, and naturally desirable to do that. The "intensity" of the curve of the floor will only depend on the length of my radius. After all, it can be argued that a straight line is itself a curve with an infinite radius.

I will be much pleased to continue this discussion.

Thank you.

Modibbo J.

Depends what you're designing.

A bow, for instance, will (typically) be curved, while the string, at rest, will necessarily be straight. A curved strut in compression will be pretty useless, unless other design constraints make it unavoidable.

Gaudi's architecture is full of curves and is stunning... a longbow is curved...a woman is curved.

Yup curves do it for me... even my whicskers are curved.

Del

What kind of freaky cat has curved whiskers? But I've got to agree on the curves on women though...

I have a 15 year old tom cat who's only remaining passion is finding a warm spot to rest. His whiskers are quite curved. Something about time and gravity.

Indeed... but watch out, curves must have an acceptable radius and be placed in the right places... if you understand me...

Thanks for the lesson.i never thought of that as "a radius."

I always used the waist to hips ratio.

My bad.

milo

I have been through so many electronic drafting and 3D modeling programs over the last 25 years, they've become one big tangled morass from which I have trouble pulling any accurate memory.

One of them, I suspect CAD-KEY, listed the line command under curves, along with arcs, circles, parabolas, splines, etc, etc.

It puzzled me and still puzzles me now. When I asked why the response was: ". . . because that is where it belongs."

That's not a very satisfactory response, I know. However, if you remember any of Albert Einstein's conversations about time and space, it was he who first argued that light bends and so does space when influenced by bodies with a large mass.

If "Old Al" is correct, then there is no such thing as straight line except where there is no gravity.

L.J.

My recommendation would be not only to consider shape, but also manufacturability. Straight lined objects tend to be easier to manufacture. Curved geometry should be considered when needed.

So next to function, you must also consider cost, and part made up with straight lines are easier to manufacture, produce and check (quality control) than with curved geometry.

If you're into the analytical part of straight lined geometry, curves and higher-order curves, then search the web on "Supershapes" of the "Gielis Formula" by Johan Gielis and Bert Beirinckx. The maths behind it show that all base shapes like squares, circles, triangles etc etc can originate from one equation, based on the equation of the circle.

BBEI Wrote:

"If you're into the analytical part of straight lined geometry, curves and higher-order curves, then search the web on "Supershapes" of the "Gielis Formula" by Johan Gielis and Bert Beirinckx. The maths behind it show that all base shapes like squares, circles, triangles etc etc can originate from one equation, based on the equation of the circle."

Thanks BBEI. After all these years someone has answered my question!

L.J.

Actually, both regular curves and straight lines are unnatural. If you are after a more natural design, it should be fractal (a simple pattern that appears the same at any scale). Nature is essentially fractal...

Hmmm... Have you thrown a pebble into a pond lately?

A link about Gaudi's use of entirely natural catenery curves.
And more about catenery curves.

Del

Hello Modibbo,

Cubist vs Organics.

One is the accent of the other, to use both so they complement is the sign of a master.

Stress can be directed so failure is not catastrophic yet more than strong enough to do the job at hand.

Strait lines have their place just as curves do.

Check out this design master http://www.chipfoose.com/products.aspx?CatID=37&linkId=lfMen Every drawing he can build and build well. Here are a few of them. http://www.chipfoose.com/gallery.aspx?LinkID=tn7

Brad

Hi,

Since the bebinning and end will always be just two points and the fact that a straight line is decribed by two points and most stat models manage data between a beginning and end point, you are correct. Then there is another methodoldgy system that is intended to project like, transorbital trajectories and other dynamics models and the like.

I have found it useful to migrate all of my data to a positive quadrant with the beginning point at more than 1,1,1. I allways start wth at lest +10,+10,+10 as the min for any data in the set.