Steel beam splice at midspan

[mck26]

Can anyone suggest guide/code reference for flange plated beam splice that affects beam stiffness/deflection? I have a 90 ft roof purlin span with a midspan beam splice (at request of fabricator). Beam design was governed by deflection.

 

[JoshPlumSE]

Bolted, or welded connection? If it's welded, I'd say there is almost no difference in stiffness. For bolted, the important thing is using fully pretensioned bolts. Though you have the potential for slip increasing deflections. You could calculate that based on bolt holes and depth of beam.

 

[mck26]

Bolted. Yes it will be a slip critical connection. The question is should the plates and bolts be designed for just the required moment at point of splice, or a percentage of beam capacity (50%, 75%, etc.)

 

[jayrod12]

If the beam design was governed by deflection, I would argue for a Full Moment Splice to ensure you have participation from the entire section across the splice. but I'm always conservative. In this case it would maybe result in extra bolts and thicker plates, but likely not the end of the world.

 

[CANPRO]

I suggest investigating the effect of the splice stiffness on the overall deflection of the purlin. Try modeling the 90ft purlin, but add in a short section of a different beam size at your splice location. Then you can adjust that section to be 100%, 75%, 50% of the stiffness of the purlin. I think you'll be surprised at the results. Getting the effective stiffness of the connection might be a bit tricky, but it is comforting to know you can take a bit of a hit on the stiffness without exceeding deflection limits. All of this of course depends on your section size, loading, and how close the original section was to your deflection limit.

With that said, my first attempt at this would be to size the splice for the full moment capacity of the purlin, and if the sizes were reasonable (or I wasn't comfortable using anything lighter) I'd call it a day with the full moment splice.

 

[canwesteng]

End plates or splice plates? For end plate moment connections, there are minimums in the relevant AISC design guide to ensure the connection is stiff enough. For splice plate moment connections, I would count on a 1/16" displacement of the top and bottom flange (the amount the holes are oversized) at the location of the splice, and see if that is acceptable. If not design the plates to resist all forces as slip critical. The actual deformation of the splice plates is going to be negligible, I wouldn't try and design for some % of capacity.

 

[kipfoot]

It might help to consider doing it in three pieces, with splices at less critical locations, or putting the splice away from the center of the span.