Finish to bear connection

[HanStrulo]

Hi Everyone,

I have a weird problem and would like your input.

I am checking the connection between a beam and a horizontal brace in compression. I designed the connection to be welded and checked the weld length necessary.

In the AISC "30 good rules for connection design, It says to "finish to bear" which gave me the idea that it might not be necessary to design the weld and since the brace is always in compression, it will be more economical to design a small weld just to hang the brace and the connection will be working by bearing.

Is my understanding correct?

Thanks for your input

 

[jayrod12]

I'm a big fan of belts and suspenders. So finish to bear, and welds designed for the entire load, give you two modes of load transfer in case one doesn't end up working as intended.

 

[DrZoidberWoop]

Seconded. I haven't run across a horizontal brace connection I didn't evaluate for both tension and compression, yet. But maybe you're designing a rocket launch platform or something and the loads will truly not be reversible.

 

[dhengr]

HanStrulo:
That’s a pretty short quote from some AISC document someplace, but "finish to bear" probably means that when you have a bearing detail it should be finished smooth, square, perpendicular and in a plain, in all directions, so that it does really bear. Then, a weld just holds it in place, but does not really take any/much of the load. And, this is an acceptable condition. A common example would be a heavy column section splice or its connection to a base pl. If it is cut or milled to a certain min. tolerance of planar surface, square, and surface finish, it can be assumed to be acting in bearing to take the compressive loading. Then, the weld just seals the joint and holds the pieces together, more or less. Alternatively, you might mill the end of a heavy compression member for bearing, to take most of the loading, but if it has high shears (or lateral loads) or moments across the joint, or meets the other member 20° out of perpendicular, where it has a high lateral load component, then the welding must account for these non-direct load bearing components.

 

[HanStrulo]

i do have a purely compression member (not going into tension at all in its life) but the force is really high (300T). so i end up with alot of weld because it was designed for tension.

Since i want to reduce the weld, i would like to design it for bearing only. no shear and no no moment will be experienced. is designing for bearing alone safe?

 

[Enable]

If I have a compression brace I'd usually design it for 50% tension or 100% compression, whichever is greater. But even if the compression load is higher I'd still design it in such a way that 50% of the tensile capacity could be achieved.

Can you imagine having fitup issues with a brace carrying 300T that only become apparent under load? The moans and bangs you are going to hear...I wouldn't do it IMHO.

EDIT - What are you bracing that you get 300T of compression? Can you offer a schematic of the assembly? I'm just a bit curious

 

[MIStructE_IRE]

At 300T I’d be looking for belt, braces and a spare pair of jocks!!

What are you designing that has such a high force?

 

[BridgeSmith]

We specify "finish to bear" on the bearing stiffeners for bridge girders. Typically, they're carrying 50-100 tons. It's apparently a labor-intensive process, to the point that some fabricators propose to substitute a CJP weld instead, which we typically allow. (Fillet welds don't meet the allowable fatigue stress limits we usually need require)

Rod Smith, P.E., The artist formerly known as HotRod10

 

[HanStrulo]

Hi Everyone.

the brace with the 300T is not part of a building. it's a "strut" for a braced excavation. basically a compression member between two beams in bending under the effect of lateral earth pressure.

I have not done a belt before. Could anyone propose a schematic for how it works?

@Enable, I did design the brace connection for 50% tensile capacity. I just needed some reassurance that not designing for the full axial load was the right call to make.

 

[BridgeSmith]

Since it sounds like this connection may be made on-site, I should mention the economy of a finished to bear connection vs. a CJP or other weld may be different in the field than it is in the shop. From what you describe, it seems either would meet the design requirements, so I'd suggest talking with the fabricator, if possible, to see which is preferred, or show one (probably the weld) and note the other as an alternative.

Our construction specs state:

"For bearing and other stiffeners designated as “finish to bear,” mill or grind to achieve at least 75 percent contact with the flange and no gap elsewhere greater than 1/32 in [1.0 mm]."

That can require numerous trips back to the grinder to get it right.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[jayrod12]

When I do shoring/excavation bracing they cut everything with a torch on site to fit. And then provide completely oversized welds. Pretty much a 3/8 or 1/2" weld all around everything they can reach. I think specifying finish to bear in that application is a pipe dream. They aren't that exact when putting up the solider piles or driving the sheet piles. Therefore you can't pre-fabricate the walers and struts the same way you would with building steel and tight tolerances.

Maybe PEInc will weigh in with his opinion since he's the resident shoring expert. But the above is my experience.