Twisted Column

[Zygnoth]

Hello all,

I am wondering whether anybody has come across a twisted column before and what difficulties they ran into. The easiest way to describe this would be to imagine the cross section at the base as a rectangle with major axis on W-E and a similar cross section at the top with the major axis on N-S.

Currently, this has been bothering me more as an exercise in thought than as an actual structure but I'm having trouble finding similar examples.

I've included (if it works correctly) a very rough mock-up in ms paint of two solutions to the problem. The one on the left reminds me of a column from a Gaudi building but don't think the section change was as drastic as that.

Anyway, please let me know if you want anymore information and thanks in advance for any pointers!

 

[youngstructural]

Though I hate Black Box Solver Solutions (BS squared), I would be very tempted to design this in FEA...

Your left hand column should be quite easy to design, just considering the section at several points long the lenght. The right hand column is much more challenging, and will have inherant instabilities... Definately need to have a good look at the stiffness of the column, and I would want to try and ensure all loads were concentric.

It would be an interesting challenge! I'll be keen to see if anyone else has an example, or perhaps further thoughts.

Cheers,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

 

[kslee1000]

The delimma over a twisted column rest on the ugly facts that there is no simple computational solution, and there are very few articles address it if any. In real world, I have seen quite a few of it (twisted noodle). The twisted columns in a steel mill setting have survivied for many years and still standing, a phenomenon that is worth thinking.

 

[Zygnoth]

The noodle analogy is interesting. Thinking on a similar scale, it reminds me of a regular drillhead where I imagine the strength would not all originate from the small, shared circular cross section in the middle.

Here is an example I found of a rotating brick column (square cross section):

http://flickr.com/photos/copacetic/372427163

In this scenario, the cross sections align every 10 brick courses or so (about 4 times by my eye) and there is a substantial cross section shared between all the courses.

Thanks for the pointers though...much appreciated!

 

[kslee1000]

zygnoth:

Very interesting/creative design, beautiful too, in its surrounding.

 

[youngstructural]

I would argue that the bricks are almost all active, given the squat shape of the column...

I think most issue would arrise when you try to make the column slender.

Cheers,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

 

[jheidt2543]

It looks like each course of brick could have been precast and then laid up with the right amount of rotation. For design purposes, I guess I would only take the common center core of the column as effective and assume the rest is cosmetic.