New platform on circular metal chimney

[LannyBudd]

I am trying to expand an existing platform on a 14'-0" diameter metal chimney. Currently, the chimney has an octagonal platform which is support with knee braces. At the platform elevation the platform beams frame into a rolled C8x11.5 with the C8 flanges pointed outwards and at the chimney/knee brace interface (@ minus 4'-0" from Top of Steel) the double angle KB gusset plates frame into a rolled L4x4x0.25.

My question is how should I analyze the metal chimney for the local and global effects of the additional DL, LL, Wind Loads?

 

[JStephen]

There is a standard for steel stacks. I don't recall what all is in it (allowable stresses, for example?):

https://www.asme.org/products/codes-standards/sts1-2011-steel-stacks

It would really really help to have the original design calculations for the stack, or to be very familiar with the design of stacks in general. For example, does that rolled angle or rolled channel act as a stiffener for the top of the stack, or are they just there for the platform? Or does the whole platform framing act as a truss-type stiffener?
If it's a very tall stack, the additional vertical loads may be negligible compared to wind bending loads.
Note that Roark's Formulas for Stress and Strain has stress equations for circular rings with uniform radial loads, which may be helpful for some of that.

 

[KootK]

Ick. For local stresses, it's hard to imagine that there would be any reasonable closed form solution to such a problem. To assess it in detail, one would likely have to resort to FEM. If it were me, I'd try to detail the connections in such a way that the concentrated loads were spread out such that the incremental flexural stresses in the chimney were very small relative to the flexural capacity of the steel plate comprising the chimney.

For the global effects, I would first try to demonstrate the the incremental load effects were small enough to be ignored. Failing that, I'd fire up my calculator for a workout.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[JStephen]

Note that if you had local (point) loads, they can be analyzed using methods for nozzle analysis on pressure vessels- but if they tie into circular stiffeners, you don't need to go that route.
Also, it would be expected that plate stresses at the top of the stack would be low, unlike at the bottom.