Modifying Tower Compression Ring for Increased Wind? 2

[ConstantEffort]

A client decided ex post facto to increase the wind design criteria. The tower has already shipped to the client, but no foundation has been poured yet. What miracle solutions can anyone recommend for this situation?

The vessel has a compression ring baseplate arrangement which is the only component which fails our normal criteria due to too much compressive stress. The options we're currently floating around:
1. Increase the compression ring design stress based on the materials minimum specified yield strength compared to the higher value indicated by the CMTR.
2. Increase the compression ring thickness

A. by cutting off the old and installing a new​

B. by adding an additional loose ring on top as a sort of washer​

 

[JStephen]

I'm not visualizing the exact geometry from the description.
On the actual wind loading, it'd be worth looking at that in detail to see if there are any areas that can be reduced or adjusted, any overly conservative assumptions that were made, etc. Importance factor, gust factor, exposure, etc.
Make sure they (or you) are not confusing ASCE 7-10 wind speeds (strength level) with ASCE 7-05 wind speeds (service level) or something like that.
If the loading in the base plate is a combination of bending and hoop force, you may be able to come up alternate ways to combine the two.
If there are bolt holes in the ring, you may be able to neglect the effect of holes in calculating compression.

 

[TomBarsh]

The problem seems to be that the tensile force of the anchor bolt (uplift of the vessel on that side, due to wind/seismic) overstresses the compression ring. The bolt force applied to the plate (compression ring) produces bending stresses in the "quasi-rectangular" region between the adjacent gusset plates (assuming they are there), the edge welded to the skirt, and the free edge of the ring. There are various analyses for this plate; it's very similar to analysis of the top plate of an anchor bolt chair.

Some texts treat this as a rectangular plate with various edge conditions, and the applied load. The Jawad and Farr text "Structural Analysis and Design of Pressure Vessels" provides a solution for this based on what they call a yield line theory (based on plastic or limit analysis, rather than an elastic analysis); see Chapter 12 in that text.

I don't think that providing a washer will do much. It "might" if it's continuous, then you'd have a layered construction with both lamina in bending. I don't think I would want to analyze that myself.

The analysis provides only a few variables that have effect on the stress. (1) Obviously, the bolt force, so investigate that to see if it can be reduced. (2) Next is the allowable stress, you can try using yield stress from the ctr's. (3) Finally, there are several dimensional variables that you could adjust in order to reduce the stress. Of course, this would involve welding on the base ring components. The goal would be to reduce the span of the region between the gussets, or between the skirt OD surface and the edge of the compression ring. I imagine that some bars could be welded into place to reduce the span; but this seems like an awful lot of welding.

 

[TomBarsh]

By the way, I have always wondered why these things were called "compression rings". Anyone have any idea?

 

[TomBarsh]

What about sending out a fabricated base ring assembly per a new design, cut the old base ring off with a short section of the skirt, then butt weld the new assembly's skirt to the existing skirt? Back-charged to whoever elected to change the specs.

 

[dhengr]

ConstantEffort:
If you really want a meaningful discussion here on E-Tips, you would probably do well to show a good, well dimensioned, details of the tower base. I’m having a little trouble seeing it from here. Show were the max. stresses are, and any other important design info. too. Otherwise, we’re just guessing exactly what you’re talking about.

 

[marty007]

The suggestions above are good suggestions.

I had a similar issue with a distillation column recently when the client decided half-way through fabrication to mention some very significant pipe loads right at the top of the tower. The base ring/compression ring had already been fabricated and welded to the skirt.

Our solution was to double the number of anchor bolts. I forget, but I think we had 8 anchor bolt holes specified, so the calculations were redone with 16 anchor bolts and that did the trick. As for fabrication, this meant drilling 8 more holes in the base ring and in the continuous compression ring, and adding 16 more gusset plates.

I don't know how many anchor bolts you're talking about or how big your tank is, but could this work in your case?

 

[JStephen]

I normally hear compression ring or tension ring used where a cone transitions to a cylinder, and the stresses will put the ring in compression or tension depending on the geometry, but the description didn't match that.

 

[ConstantEffort]

Thanks all! Got called away and couldn't get back to this post to follow up.

I need to finally read Jawad and Farr for myself, I think. I've read summaries and extracts, but never the real thing.

The increased anchor bolts is a great idea and I'll run that by the project team to see if that is feasible. We're a bit unsure of field labor quality.

 

[Duwe6]

". . . bit unsure of field labor quality"

You will only really need one good welder. Not EXCELLENT, or SUPERB, merely Good-to-Very Good. Vessel work does not require perfection, just "good enough". But with that 'merely good' welder, you will need a good welding inspector in attendance during the welding. May end up cheaper to fly in a superb welder and skip the inspector IF not Code work.