メインコンテンツへスキップ

## コメント

10件のコメント

Hi.

It's a contradiction/ confliction in 3D volume i think (edit - i think wrong - see below) The material being created around the curved ends, needs to be less than the material at the middle. To explain another way...The volume of material at the curved ends is having to be compressed as the surface attempts to sweep normal to the spline trajectory around the curve.

Try this...

Design driven and driver curves, now offset these curves for the material thickness. Sweep to produce the external surface, repeat to make the internal surface. Good geometry is vital, offsetting splines needs the spline end/start to be normal (perpendicular) to the offsetting axis. Suggest starting with ordinary lines / regular arcs to get the concept understood...

Hmm - i'm not sure of what i know as i need to contradict myself in the above post.

Sorry...

Mimicking your design, my efforts show 'crumpling' of material ( disproving my hypothesis above).

Note the new surface 'zebra' tools of V5.

Even on 'ordinary' , basic non spline driven geometry, in this circumstance self overlapping/intersecting geometry has been allowed.

Just my thinking, on the other hand it would be great if DSM could calculate the behaviour of a real material being shaped. For example with Aluminium there is a narrow path - a shaping that works well with 2 mm sheets produces cracks with 3 mm. But for this to calculate the software needed more than just one material parameter (like Young modulus), and presumably more than FEM algorithms.

I tried sweeping before with other shapes and it worked fine, as you demonstrated.
So the phenomenon is a bit mysterious.

At least a more verbose error message might be helpful.

Ok, next try from scratch, maybe the problem was in two line segments chosen.
So now I sketched the sp-line in one piece - and copied the same line on the perpendicular plane to use is also as the path.

But dragging the yellow arrow yields - nothing.

Is there a trick somehow to connect the two lines before dragging?

Think i've got it

Splines being pulled by splines get very fussy. - if the resulting pulled surface returns and touches itseff, it stops there. Crumpling / distorting is OK / permitted.

Splines pulled by axis just intersect themselve and SPLIT into 2 surfaces all benignly.

Hope this image explains my observation.

I see your point but it looks like a separate challenge to me - of course there will be some complication if you sweep a line on the inside as well as the outside of the axis.

But my request seems much simpler to me - sweeping a kind of half-circle along the same pattern, no overlapping inside/outside parts.

Now I tried a new variant - and it works.
The splines do not seem to cause any problem.

The trick is to move the path line (right) completely out the space that will be covered by the final shape.
My understanding that the two lines have to touch at the start point was nonsense.

Maybe these conditions should be mentioned in the docs.

No, i have to disagree...

I think the spline drive path can be anywhere as this image.

but what makes a real signifcant difference, is the degree of curvature of the path spline, also the curvature plus the 'return length' of the pulled spline corner rounds - i hope these images show that clearly. Have we cracked it?

It seems to be a complicated matter on the whole.

I can only report about my experience - where I could only create the complete surface after I moved the path line to the side.
So at least my reported problem was not connected to the 'degreee of curvature' of my line or path.

For the records - I finally achieved the desired thickness of my hull by scale-ing it
(whereas my first approach was to sweep the line after thickening it to a surface).

サインインしてコメントを残してください。