An exercise: DRAWING A PARISIAN FENCE

simon le bon

Dear Taff and Elisei

I keep going after all: I think I have a good idea for the final curved sinusoid but without help of a bending tool.

The last development leads me to find (as you) the mathematical size of 6 semi-waves diamond's width.

So the Diamond becomes a mathematical object perfectly defined in its own dimensions: 6X width, 5x for the sides.

We only have to say that SketchUp loose the preciseness to calculte the angle, and don't return two times the same value.

But the preciseness is two point better that what gives the tables of sinus:

Nasa table of sinus

simon

pilou

It's maybe not a good idea to use the Shape Bender for a long parisian fence roll!
Better is make a texture from an image result!

(and if it's Nasa who give Sinus'waves that will be even more precise

TaffGoch

@simon le bon said:

"I think I have a good idea for the final curved sinusoid but without help of a bending tool."

While I didn't use it, half of an ellipse looks pretty good:
Half ellipse

@simon le bon said:

"We only have to say that SketchUp loose the preciseness to calculte the angle", and don't return two times the same value.

I reliably & repeatedly get 73.740°, when using SketchUp's protractor:
Angle
You may need to change your angular settings, in the "Model Info > Units" dialog. (Turn off snapping, and change the precision.)

-Taff

simon le bon

Dear Taff,
Sorry again for the time.
The main work which consuming time is to report: prepare Su, take snapshots, comment them, make thumbnails, upload,,

Also I'm not a very brilliant student: which is a good thing to show all the difficulties we have to pass over

I am stucked again.

As we havn't in Su a Bezier curve tool which is able to deal a curve of defined number and length of segments between two fixed points,

%(#FF0000)[I am actually not able to draw the curve needed.!! stupid isn't it ]
(I had first tried to bend (fredoscale) a line exactly segmented and welded as a curve, using a classic bezier curve as a leading shape. but this bending operation had made the segments length changing .)

Please Taff, Can you give your solution!!??

Now for the angle of the diamond and the ability of SketchUp into such a preciseness, it seems that I

@unknownuser said:

You may need to change your angular settings, in the "Model Info > Units" dialog. (Turn off snapping, and change the precision.)
that was done

have done something wrong
First I have taken your angle of 73,740° to verify -> ok!

Second, I have made the construction again, and I find pretty the same as yours.

*s

TaffGoch

@simon le bon said:

"As we havn't in SU a Bezier curve tool which is able to deal a curve of defined number and length of segments between two fixed points,...

I am actually not able to draw the curve needed.!!...

Please Taff, Can you give your solution!!??

Simon, I kept the 5-segment and 7-segment curves separate. (I didn't try to complete the entire hoop as one piece.)

For the 5-segment section, I set the SU "Arc" tool to 5 "sides," then started an arc from the bottom (tangent) upwards, toward, and ending on, the centerline (where it will eventually meet it's mirror image.)

I watched the length in the "Entity Info" box, and kept zooming in, on the endpoint, until the length was very close to 5.0 (so that each segment will be 1.0 unit long.) When I "exploded" the curve, I found that the length changes a little bit. By drawing curve-after-curve, I was able to "sneak up" on the correct length, by trial-and-error.

I kept the one curve that came the closest, and deleted the others. After exploding the retained curve, each line segment is accurate, to within 3-decimal-places ("1.000")

Once I had the curve, as a guideline, I could further subdivide it into smaller segments, to be used to construct the sine curve.

I constructed the 7-segment curve, using the same technique. I suspect that the Fredo-curved arc will provide comparable accuracy, as long as you don't try to make the entire hoop at one time. SketchUp's "tangent" inference made my method possible, even though I think other methods will work just as well.

@simon le bon said:

"Now for the angle of the diamond and the ability of SketchUp into such a preciseness,...

While your measurement is now "precise," I never used the angle to construct anything. Interesting intellectual exercise, but not important to making the model. If you want to rotate a 5-segment-long line, to meet a triangle centerline, you can use the technique described in this model, by Jean Lemire:

Rotating an edge to another edge

I hope my explanation is clear enough to help you pick up a "trick" or two; to polish your modeling skills.

-Taff

TaffGoch

@unknownuser said:

"It's maybe not a good idea to use the Shape Bender for a long parisian fence roll!
Better is make a texture from an image result!"

Pilou,

I agree. I look at this model as an exercise, or challenge, to improve modeling skills. Such a detailed fence would make a model very "heavy," because it introduces too many line and face entities.

For inclusion in a larger model, an image-texture, with transparent background, and some SU face trimming (for proper shadows) would be a much better option. If trimming is too complex, just turn off shadow-cast and shadow-receive checkboxes (in "Entity Info" dialog box.)

Fence as texture
-Taff

Textured_fence.skp

simon le bon

Dear Taff, first thank you so much for this documented reply. I Know so well how to produce documented pictures is time consuming.

I was seriously beginning to see myself with a too bad limited mind. After various false tries, I was turning around like a dervish. I was thinking of decompose (as you) the curve, but adjusting several portions of mathematical circular curves (ellipsoid) for which I had not the theoretical power to make any calculation..

So a great thank to you to have showed me what I was missing: modesty!!. When we haven't the theory nor the tool, it remains the " bricolage", But this can be a clean "bricolage" as you have done.
Since Su doesn't give the curve length in addition to the radii in real time into the VCB, going near by near as you do with simple portions of circle is the elegant solution.

One of my interessant tries was to use Fredoscale's bend tool applied on a line with the exact length, bend it a little with the correct number of slices, and take bend tool again on this new well segmented curve but with a slice of One and bend it step by step. May be it is possible but I haven't reach to go to the end without bugs..

interessant tools:
Didier Bur Dim-Angle

Fredo Polyline Divider

I'm going for work until far in the night, so no try before tomorrow.

++simon

simon le bon

I progress.
Thank you Taff for all the tracks you have left.
Very good exercise indeed.
Still some problems to resolve.
Not at home this week, hope a final point on Monday...

- simon

TaffGoch

Simon,

Sine wave count and braiding looks good. You are to be commended for persevering, until achieving success!

-Taff

simon le bon

Finally I've got mine

*s

TaffGoch

@simon le bon said:

"Is that the waves have to be regularly curved in the curved area?"

Simon, good catch.

Once I had established the curve, of 5 and 7 straight-line segments, I could then use the "arc" tool, to create new curves, of 35 and 49 straight-line segments (by multiplying by 7.)

I did this, only to more "smoothly" model the curve. It's not absolutely necessary, but isn't difficult, since the sine-wave amplitudes are already established. Incorporating the amplitudes into the new, smoother curve is relatively easy, using SketchUp's inferences.

-Taff

simon le bon

Dear Taff,

I was to bring a pair of my tips for this exercise and was waiting for your validation and:::

@unknownuser said:

You are to be commended for persevering, until achieving success!
I thought I had finished my exercise??

I open your skp to get an idea of what you mean!
Is that the waves have to be regularly curved in the curved area? I don't see what can you mean else

simon

simon le bon

Dear Taff,

Re: Doing this way, we can't maintain the sine segments along the top of the fence exactly the same length!

Anyway, I would like to propose you some of the tricks I had previously talking about to achieve easily the round top of the fence as you wish .
YouTube SketchUp_ParisianFence_Tricks 1

[flash=640,505:2smet3c3]http://www.youtube.com/v/EwExCv2ZTRw[/flash:2smet3c3]

using:
Component Stringer
BezierSpline
Projections
Lines to tubes

Now I just have to do it for real
*s

pilou

It was the trick N°1, what will be the n°2

The curious thing is there is no precise documentation on the Net about this sort of garden materail

simon le bon

Hi Dear Pilou,

You always keep a vigilant and friendly eye

I guess you have noticed we all have followed your slices:

(The second trick after about a week to film "Rock'n Roll!!!!")

Dear Taff,
Here is my "final-final" parisian fence...

*s

TaffGoch

@simon le bon said:

I havn't discovered what was the mistake Taff was talking about

Simon, look here:
Fence mistake fixed
-Taff

TaffGoch

Simon,
I found that the key characteristic of the fence is the horizontal-to-vertical spacing. The diamond shape has to be exactly the correct proportions, or the sine waves of the top and bottom "rails" don't overlap properly. After a couple of false starts, I recognized my error, and got the diamonds the right shape.

I hope that I've provided enough detail, in the model, to demonstrate how I constructed the sine wave paths.

-Taff

This model contains a mistake. Download corrected model from 3D Warehouse.