TUBE - is this normal?

[rocorona]

I have a very rimple TUBE command.
The section is a SQUARE, centered on the path.
In the syntax of the command there is the "angle" parameter, that allows for rotating the section along the path...
The angle can be different at each node of the path, but I use the same value.
Well, when the section is rotated 0.0 degrees, the TUBE has always the same width, but if the section is rotated 90.0 degrees (being a square it is identical to the previous one) the tube width is changed at the path "curves".

It seems that ONE of the section dimensions (X or Y) is retained at the bisector plane, but this is the behavior of the TUBEA command, not of the TUBE. Anyway, the section at the curved node is a rectangle.

Is this normal?
(verified with AC 7.0 & 8.1)

s = 0.20
r = 90.00

TUBE 4, 6, 115,
-s, -s, 0,
+s, -s, 0,
+s, +s, 0,
-s, +s, 0,

-2, 0, 0, r,
-1, 0, 0, r,
0, 0, 0, r,
0, 0, 1, r,
0, 0, 2, r,
0, 0, 3, r

[Vitruvius]

Yes.

I only do TUBE with the following path nodes:

x,y,z,0

If you provide an angle other than '0' for the last value you get some very wierd results as the tube tries to twist its profile whilst following the path.

It's far simpler to just use the x,y,z, coordinates to define where you want it to go and let TUBE take care of the mitres.

[tsturm]

If I am correct, the rotation command would allow you to create a spiraled fluted baluster as the rotate portion of the TUBE command takes the shape and rotates it on the axis of the path.

Therefore if you wanted to create a rope swag, you could use the tube command to rotate the shape along the path.

Vitruvius is right on the use of rotation for the picture frame. Only use the rotate to do a spiral shape or a ramp, like GS uses in the example.

[david nicholson-cole]

rocorona wrote:
In the syntax of the command there is the "angle" parameter, that allows for rotating the section along the path...
The angle can be different at each node of the path, but I use the same value.

when I last looked into this the angle was not relative to the prevous point (as in Sweep) but to the start. therefore a square section rotated 90 will look identical to the one not rotated. If you use PUT&GET you can 'accumulate' a changing twist angle as you progress, in a non parametric way you would say,
.................
x1,y1,z1,15,
x2,y2,z2,30,
x3,y3,z3,45,
x4,y4,z4,60,
..........
and so on. the twist angle relating to where it started from. eg, in this case, if the first point was twisted zero, you would make the phantom point twist minus fifteen, so that the twisting motion looked correct.

>>david