Solid steel section made from steel plates

[formi]

Hello everyone!

I wonder what's your opinion about welding thick plates in order to make a solid section (e.g. five 20"x 4" plates welded together --> 20" x 20" solid steel section)? Should the welds be CJP's or can be partial penetration groove welds?

 

[KootK]

I vote partial if the numbers work. What's the application?

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

 

[formi]

high-rise building, ground floor columns

 

[KootK]

I would expect that arrangement to be terribly inefficient Formi. Have you looked at a biult up HSS, wide flange, or cruciform section?

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

 

[formi]

it's a concept not detailed design. Columns buckling lenght 25 ft., loads: 25,000 kips! max allowable column size 30x30"

 

[KootK]

Awesome! A box has four welds. The five plates will have eight. I suspect much will depend on that and keeping the finished product relatively straight. Got the courage to laminate the plies with slip critical bolts? I'm not sure that I would but it ought to be possible.

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

 

[BAretired]

A square hollow section 30" x 30" with 4" thick walls has an area of 416 in². For a load of 25,000 kips, the averages stress is 60 ksi. That would require a high grade of steel even if the columns carried only axial load.

BA

 

[racookpe1978]

And how will your proposed five plates create a single vertical member?

A normal wide flange has 3x plates
A normal square hollow box has 4x plates.
A normal 5x plates has ????
(Will you make a WF shape, then weld 2x more end plates across flange end to flange end, welding only from the outside with a V-prep? )

Would you gain any resistance if you filled the hollows with concrete?

 

[formi]

Will you make a WF shape, then weld 2x more end plates across flange end to flange end, welding only from the outside with a V-prep?
Then it's a wide flange with cover plates, not a solid section. Thornton Tomasetti successfully implemented this approach (solid steel columns) in The NY Times Building and Chicago Spire. In NYT Bldg they create 30x30" solid columns made out of five 6 inch plates welded together. Filling with concrete don't give so much extra capacity and moreover you gotta take care of differential axial shortening...

 

[KootK]

I found some articles on the NYT building and its solid steel columns. I suppose that, if you're looking to support a given load and the only parameter of interest is footprint, then a solid column makes sense. I'd love to know how the NYT columns were fabricated. I would think that weld restraint issues would come into play as subsequent plies were welded to plies already welded and cooled. Preheat I suppose. Left to my own devices, I'd approach the welding like this:

1) Heavy welding at the ends of the column to get the load in and out from the various plies quickly.

2) light welds or even bolts between column ends to ensure composite buckling. I'd minimize weld size as much as possible then, if it made sense, I'd look at stitch welding.

Another approach could be to make every second ply an inch or two narrower than the others. That might allow fillet welds to be used and avoid the need for costly surface prep.

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

 

[formi]

Another approach could be to make every second ply an inch or two narrower than the others. That might allow fillet welds to be used and avoid the need for costly surface prep.
I found some MEng project (Chicago Spire) done under supervision of Thornton Tomasetti's Engineers and look what they propose:

 

[KootK]

Neat. Thanks for sharing this Formi. I'll try to work some radiator columns into one of my future projects.

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

 

[BAretired]

It's not clear how they would do the welding and the maximum load is nowhere near 25,000 kips.

BA

 

[formi]

It's not my design, it's someone's MEng Thesis - structural design of Chicago Spire...

 

[formi]

Anyway, how would you size a fillet weld for this kind of "radiator" columns If they are axially loaded and there's no bending = no shear? I thought about calculating possible shear due to member imperfection. What do you think? What minimum fillet weld thickness do you recommend?

 

[dhengr]

Formi:
Obviously, you don’t have a normal flange to web welding situation in the last set of sketches you have shown, or with material which is that thick. You are not trying to tie parts together the way we might in building up a normal WF or box shape. But, you must account for any shear flow which might occur btwn. the various parts and you must pay attention to individual pl. buckling, as well as total section buckling, where shear flow comes back into play. Then, you will have some special welding considerations w.r.t. AWS and AISC and min./max. weld sizes vs. plate thickness. Actually this is really a heat input issue, so that the large heat sinks/plates don’t cool the weld so quickly that it cracks or causes inferior HAZ’s. That means pre and post weld heating to the proper temps. and time durations. Then, you have an AISC ‘end bearing’ issue, you must end mill these types of heavy sections for good bearing. Finally, I can imagine some load transfer problems when sections or individual plates change sizes and locations; end splices will be complicated. I don’t think much of this requires full pen. welds, just imagine what a night mare that would be, in terms of prep. and weld costs and in terms of final section deformation from welding. Butt splices (in compression) btwn. pieces of pl. probably do need full pen. welding, and pl. edge prep. will be critical for fit-up and welding. You would weld something like this section from the middle outward, with much welding automation, repositioning for welding and deformation control.

 

[KootK]

You may find this useful: Link

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

 

[formi]

Thanks for paper KootK. One more question: what weld type do you recommend for 'normal' WF's with cover plates. e.g. W14x730 with two 4"x20" plates welded to the side of flanges. Would you specify fillet welds or PJP?

 

[KootK]

I vote for fillet welds if possible. Eliminating the surface preparation that comes with PJP can save some real money. This is kindergarten level welding advice though. Other "heavy steel" practical welding issues may come into play. Hopefully Dhengr will jump back in with some more recommendations. I agree that heat input issues are certainly a big deal with this kind of thing.

Is this a pricing exercise where there will eventually be a "right" answer that you can share with us?

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