huangyhg

tubular column-girder ring stiffener design
anyone have any reference material on how to design a ring stiffener for a girder welded to a tubular column, such as an offshore deck leg girder connection? do you use roark case 1?
pc engineers has a small program but i want to see the basis for their calculations:
i don't know butkiss about offshore structures or ring stiffeners but i'd be tempted to ditch the stiffeners in favor of a full on circular plate that passes right thru the hss tube. especially if there's a large moment involved.
actually i'm talking about circular columns, like a 48" diameter tube with w-beams framing in.
external rings are preferable to internal diaphragms for ease of fabrication.
you can model the thing in visual nastran. just some extrusions in autocad, have it in a format readable in the program, put restraints and loading and check the vonmises stress and deformation.
thanks but i was looking for an analytical approach like a spreadsheet or hand calculations so you can just change the parameters for different cases.
yeah, i imagine you'd take some flack for a through plate on a 48" diameter tube. i checked out that software that you linked. based on the crazy stress diagrams, it would seem that you've got your work cut out for you.
as an approximation, maybe something like this:
the job of the stiffener is basically to safely transfer the axial loads from the girder flanges to the portions of the tube that your considering to act as the "webs". i dunno, maybe 1/3 of the depth of the tube on either side? maybe you could design the ring stiffener as a laterally braced, simply supported, semicircular beam? the stiffners could have there own stiffeners if necessary to justify the assumption of lateral bracing. i think that this would be most appropriate for a stiffener that isn't too wide relative to the diameter of your tube.

two references i would recommend:
the black aisc design of hss connections
and
blodgett design of tubular structures
(there is also another similar one)
both are good in their own way.
if i would know enough i would set a mathcad worksheet or maybe (is it feasible?) a excel spreadsheet to drive the generation of the 3d model. i have been perusing pilkey's tables and for now imagining roarks gets not much further, one needs a reference (or program) suited to your intent for proper design of the stiffener. i think i have not but may look further. the suitable fem model might be a more perfect solution, anyway.
you may still generate some kind of partially parametrically driven model by modeling a cylinder in risa 3d, for it admits both speedily change in thickness and dimensions, for you can scale within risa itself the x, y and z dimensions separately. the loading shape, may need to be adjusted, however. obviously, the plate segments will be plane in risa and some of the advantages and problems generated by the fact of the initial shape being truly cylindrical won't be captured.
the release 2010 of autocad also allows parametrical design of blocks. i may try to generate a block like this just as a learning exercise, for it is a new feature, and one that will speed up the generation of structural models of this small complexity. at least the initial shape will be perfect, this way.
ddp,
by tubular, do you mean square or round?
csd,
round sections.
are you talking about blodgetts design of welded structures? i can't find a tubular structures book.
is the aisc book only rectangular?

__________________
借用达朗贝尔的名言：前进吧，你会得到信心!
[url="http://www.dimcax.com"]几何尺寸与公差标准[/url]