Develop the Full Tension Capacity of the Member 1

[Althalus]

I've heard this many times in the past. All welds must be designed to develop the full tension capacity of the member. The implication being "regardless of actual load".

But I've read through all of Section J2 of the Manual. I can't find such a requirement. Is it somewhere else? Is it an old requirement that isn't used anymore? Is it only for a seismic design?

 

[Enable]

Not sure if there ever was a code provision that used this kind of language. However, back in the day connection design was radically different than it is now. We try and economize connections (given their expense) but previously it was thought: "why skimp on the connection? Why should this be the weak link?" Hence they were designed for the max load able to be developed by the member (and in a lot of cases much MORE). It's kind of a burr in my saddle at times because when I repair old structures I cant infer the demand from the connections because it likely was designed for load it was never intended to see!

It's also a tact taken by engineers who don't want to think too hard about the actual demands at the connection (more common in restoration than in new construction). It's safe for them and less work.

EDIT - These days the demand you design a connection for will depend on purpose of the member. For example, braces acting in both compression and tension are usually designed for the greater of 50% of tensile capacity or 100% of compression capacity. Simple shear connections for beams are often designed for 0.5UDL. Etc. It'll also depend on if the members are sized for failure at ULS or more likely, if SLS governs the design.