Extending a Top Flange of Steel Beam

[SinStrucEng]

Good day/evening folks,

So what started as a simple load bearing wall removal job has turned into a headache. The original plans of the building show a load bearing wall directly above a spine beam in the basement. Cool. I got a new beam to replace the wall being removed, got new posts, verified that the existing beam can take the loads, etc. All good.

Then they started pulling the place apart and found that the wall actually sat approximately 6" off of the beam, and was bearing onto the floor joists. This gives me the an issue, because the second floor's joists were lapped over this wall. If I am unable to install the new beam to be positioned directly overtop of the spine beam in the basement, I will be forced to add posts all the way down to the basement.

After measuring the joists, I am short 3.25" of bearing length. That's "all" I need.

The current beam spec is a W200x46. Roughly 8" x 8". I used a previous design approach that I did for brick plates, and worked out the plate that I would need to support the factored loads coming from floor joists. The plate would sit on the top flange of the beam (most of it, 7.5) and extend out by 3.25", welded to the top flange. The plate would need to be just over 1" thick to take the loads. That is obviously far from optimal and frankly seems ridiculous.

Does anyone have any advice as how to squeeze out another 3 to 3.5" of bearing length? Or am I reserved to having it moved over and resorting to new posts in the basement? Client doesn't want to open the entire ceiling to rejig joists and sister them. Thanks in advance.

 

[TLHS]

Not quite sure of the situation. Maybe a sketch?

It sounds like you have access to the basement. What's the load level like? Could you feasibly use the floor joists like a transfer beam if you maybe sistered a couple of new ones in there? Bring the posts down onto the joists and then pass the load from the joists into the existing beams At that distance the moment and deflection issues on the joist are pretty small and you have a pretty straight shear or direct compression path through the joist hanger and into the beam you're trying to load.

 

[SinStrucEng]

Hi TLHS, I will do a sketch in the AM, but based on your reply, it sounds like you understand the situation correctly. Yes the distance is minor. That said, the loads are 47 KN per side (factored). I was hoping to have them land directly on the beam and not load the joists instead... Not exactly minor for a residential application. However, I like your idea... Thinking about it, I reckon that sistering a joist or two to the presently-installed joists would work... Perhaps I can match the existing joist depth to an LVL and use that instead of the ordinary SprucePineFir commonly available.

Will run the calcs tomorrow morning. Thanks for the brainstorming.

 

[KootK]

I used a previous design approach that I did for brick plates, and worked out the plate that I would need to support the factored loads coming from floor joists.

I don't love that option. This will induce a torsion in the beam that probably doesn't have a great place to go at the beam ends.

Could you feasibly use the floor joists like a transfer beam if you maybe sistered a couple of new ones in there?

I like this better. It may be that you don't even need to reinforce the joist given that:

1) The load probably lands within "d" of the beam ends anyhow and;

2) If I understand correctly, the offset exists now such that the condition has already been load tested to a degree.

I've found this offset thing to be a pretty common occurrence in older wood frame building. Just a function of tolerances accommodation I imagine.

By the book, I think that we're not supposed to use the "d" thing for concentrated loads. So there's that...

 

[kissymoose]

By the book, I think that we're not supposed to use the "d" thing for concentrated loads. So there's that...

Indeed. Not as helpful as one might hope.
Also, I'm not sure if everyone does this, but I usually include this concentrated load in my bending check as well, even though it's less than "d" from bearing. And for large loads such as that 47kN, it can make or break the design. But I'd love to have someone convince me that the stresses don't work like that and it can be ignored.
2018 NDS: