strengthening a deficient steel beam by welding a steel plate to the bottom edge

[lubos1984]

Hello everyone,

Hope you are all enjoying the warm weather for everyone living up north!

I have a project where the client wants to add a hot tub and additional dead load onto a flat roof. Because of this the current W530x74 beams are not sufficient (fail by moment). it is not practical in this circumstance to add additional beams or posts, so I was looking on what can be done to strengthen the current beam since the fail ratio for moment is about 1.15. The scenario that seems to work based on SAP modelling is to weld a 3/8" plate to the bottom flange through the entire length.
My question however is does the plate need to be continuous or can you weld it in 3 piece lengths ? The steel beam is over 36' long! I'm having trouble modelling this. Anyone have experience doing this ? Would welding it in 3 pieces be an issue ?

Thanks again!

 

[dik]

Often done... and easy. You might want to consider adding some round BAR stock in the top flange-web fillet. One of my earlier ocncerns about using an oxy-acetylene torch for cutting out a bunch of 5" pipes from our old gravity hot water heating system was their proximity to wood framing. Same problem with welding.

The plate on the bottom can be in 2 or 3 pieces, but they should be welded 'end to end' in high moment areas.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[structSU10]

You need it the full length? more than likely just a middle zone will need reinforcing for moment unless you are trying to improve stiffness as well.

Just extend the reinforcing zone over where the beam is deficient and provide the proper anchorage distance beyond the cut off point. You may only need to add a 12' or so long plate.

 

[XR250]

I rarely find any steel beam in residential controlled by strength. Have you checked deflection? !5% is within pencil sharpening territory - especially residential where it will never see the design load.
Also, just curious - how did you model that in SAP? For me, I would have done this by hand using basic mechanics of materials.

 

[wth]

I've modelled beams like this in another FEM software to check deflections. What I do is split the beam in 3 parts. Unmodified - custom section with reinforcement - unmodified.

Makes it easy tweaking the reinforced length against deflection criterias.
Then you calculate the necessary welds by the shear flow. I've never split the plates, but shouldn't be a problem as long as the welds between them ensures continuity?

 

[Ron247]

Make sure you do not use the last allowable bending formula (F1-8 in the ASD method) since it requires compression flange to have as much area as the tension flange.

 

[Once20036]

The plate on the bottom can be in 2 or 3 pieces, but they should be welded 'end to end' in high moment areas.

I`m curious about this statement.
Assuming a simple span beam and assuming that the reinforcing is only being applied where required for strength purposes (per structSU10, above), the bottom chord will be in tension.
In areas where two plates meet, the section will behave as unreinforced, due to the lack of tension continuity in the reinforcing plates. That means that each splice point will behave like a hinge. Sounds like a collapse mechanism to me.
How is is possible to skip the "end to end" welds in low seismic regions?