Vibration Mitigation 1

[Boiler106]

I've been tasked with creating design drawings from a vibration mitigation study performed by a specialty engineer on a composite framed steel floor with slab on metal deck for sensitive equipment. The specialty engineer used finite element modelling to determine the repairs which consist of adding posts and beams. Project is in the USA.

Beams (in red) and posts are to be added to the floor framing.

The beams parallel to the deck have us concerned. Although I'm told the FEA model took into account some orthotropic deck properties, it's strange to me that these are effective in a one way slab system. Ignoring all of this (because its not my design) we know that loads associated with vibrations are very small, such that even bolting on reinforcing is not sufficient per the literature. Reinforcing needs to be welded.

I would prefer not to discuss the effectiveness of this reinforcing layout

For those with experience, is it enough to place the beams tight to the underside of the low flutes for these to be effective? Or would they need to be engaged with the deck with a cored slab/headed stud connection?

 

[271828]

That's true for a spot welded bottom chord of a truss, or any bolted bracing, and especially true with posts, but continuously welded reinforcement doesn't have any slack to give up.

Such a scenario might work, but I wouldn't chance it. Retrofit adventures are disruptive and expensive. Doing this twice on the same floor would start to look like a clown show pretty quick.

The problem is we're dealing with unbelievably low stresses from a typical structural engineering perspective.

Vibration at the typical 0.5 %g limit corresponds to a vertical displacement amplitude on the order of 0.001 in. In the OP's case, the vibe limit is much tighter, so the displacement amplitudes will be more like 0.00001 in. Imagine what a small stress that causes in the new welds and new reinforcement. It takes almost nothing for the system to behave as if the new components aren't there.

The retrofit is a lot more likely to work if the stresses oscillate about 2 ksi (say) rather than 0 ksi.

I once did detailed shaker tests on a floor specimen and measured the frequency response function. Force and accelerometer at midbay. Went below the specimen and installed a screw jack at midbay with 100% of the force that I could muster. Repeated the tests. The new frequency response function was indistinguishable from the first one. I wish I would've thought to jack it up, install the screw jack again, and repeat. Get 4-5 kips in it and I bet the FRF would've changed significantly.

 

[BMISL]

In my opinion, if it is only for walking force vibrations, the main question would be: what amount of mass can mobilize the footwalk forces before the reinforcement, and what amount of mass can mobilize the footwalk forces after the reinforcement. If you don't have a problem with local vibration in the floor slab between the beams, the mass of the new beams will have more impact than the increase in stiffness. Additionally, if you don't have a problem with horizontal vibrations, the connection with the deck will have an insignificant effect. The main change here is the addition of the posts. You must ensure that the new posts can counteract the forces that cause the problem (footwalk forces). The posts will have a big impact on the fundamental frequency of the structure, the modes of vibration, and the damping, but they must be working properly.

P.S., sorry for my mistakes, English is not my native language.