Ballroom over Ballroom Steel Truss Vibration

[calvinandhobbes10]

First Eng-Tips post. Visited these forums for years.

Steel truss vibration problem. We use DG11 and FloorVibe.

We have a conference center that features a grand ballroom over another smaller ballroom space.
6-ft deep steel trusses span 81 feet over the small ballroom, and support the grand ballroom floor.
The max trib to a truss is 31 ft. All the trib area is part of the ballroom.
6 1/2" normal weight floor slab (2" deck + 4 1/2") floor, with composite beams framing to the trusses.

We're designing for dining/dancing vibration, 2.0%g limit.
Truss depth is set.

I used FloorVibe (knowing that trusses may be a different animal) to ballpark this truss, especially its frequency.
Used 81' x 31' bay. In the program, trusses are User-Defined shape, using 0.85 * Ig(chords) and Area(chords) to approximate the trusses.
Truss natural frequency as-is comes in around 5.3 Hz, the whole bay around 4.8 Hz.


My questions:

1) I'd expect an additional damping or diffusing mechanism kicks in when you are looking at a larger tributary area (that won't ever see full dancing), but I'm not familiar enough to know how this might be applied. My question 2 is my best-guess. Does anyone know of anything like this? I.e, difference between ten and three hundred people doing the cupid shuffle--there's certainly a break point below which the sensitive occupancy doesn't feel it.

2) Can a form of live-load reduction be applied to vibration? My hypothesis is that I can reduce wp, weight of participants, same as live load (by up to 50%), which reduces the applied acceleration accordingly. DG11 provides 12.5 psf, FloorVibe suggests 6 to 12.5 psf. I take 12.5 (DG11) and apply an analogous live load reduction.

3) Is FloorVibe reasonably accurate to approximate a steel-truss frequency? We have not done an FE model for this. We verified our moment of inertia with our actual truss design (Ram Elements) and I don't really suspect an FE model will end up being that much different.

4) Adjusting the truss to meet the acceleration limit in FloorVibe (using my 6.25 psf live load from question 2), at first check yields a truss about three times the moment of inertia required by gravity. Is this expected or normal for an application like this?

5) The truss top and bottom chord will be horizontally braced. Disbursement cross-bracing is normally just a retrofit option for joist vibration issues, but does this have any effect on design calcs? Doesn't show up in DG11 but am curious if anyone knows much about the effects. FE model might be the only way for this.


We are under pressure to keep steel tonnage in a tight budget and don't want to beef up the trusses unless we absolutely have to. Deepening the trusses or adding more of them would be my next recourse.

Appreciate any insight.

David

 

[BAretired]

It's a problem we don't see very often as structural engineers. To me, your truss depth seems too shallow for a span of 81 feet but I have not gone through any calculations.

This link to a Virginia Tech presentation may contain some helpful information:

http://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&ved=0CE8QFjAD&url=http://

http://www.eac.com.co/files/memoria...-meqr5mfSu9ufyA&bvm=bv.66699033,d.aWw&cad=rja

BA

 

[Archie264]

calvinandhobbes,

Welcome to the forum. I think you've asked a fairly challenging series of questions there, which is probably why you haven't been flooded with responses. Hopefully BA's link was helpful, though it didn't come up for me. I'll agree, though, that a 6' truss sounds shallow for an 81' span but I haven't run any numbers either.

 

[racookpe1978]

Is the lower floor damped - on ground directly; or can it move as well independently or at the same time (frequency) of the upper ball room floor?

 

[calvinandhobbes10]

Thanks for the replies. Understand this is pretty complex. Ba's link was direct to a ppt file and worked for me, might be a downloading restriction.

Under the lower ballroom it transitions to concrete structure. there are four levels of parking garage total height 45 feet. Flat slab w drops PT concrete slab structure. Bays are 23 x 27 typical. No carried columns. Driven pile foundations. Thus far i have been assuming that the contribution of concrete garage to vibration is negligible compared to the steel trusses since its all directly through hefty concrete columns.

 

[bookowski]

C&H - Since you're not getting many replies I'll throw in my two cents which is minimal, as others have said it's a fairly complicated problem.

I had a similar situation, it was a roof play field over a gym over a 2nd gym with something a little over 100' spans. We had to keep the trusses shallow (I think 8ft but it was a while ago) to meet the clearance requirements for the sports in the gym. I had very similar trib area and slab on deck composition to what you have.

I went down a similar path via floorvibe and chasing the vibration criteria starts to seems crazy when you get to a size that works. In our situation we told the client that it did not meet the recommended vibration criteria and the only reasonable way to get closer was depth. In our case the gym could be used for bball, dances, pep rallies etc. The acoustic consultant also advised that there was likely to be an issue. The client ended up deciding to go ahead and was made to understand that a retrofit would be more expensive. We added bottom chord bracing beyond what we normally would and attempted to make the entire panel as rigid as possible, modeling the entire thing in Risa including the slab as compositely attached to the trusses. The school has been open now for several years and they have not had any problems (we asked). Who knows - maybe they'll have a particular song or dance that is the right frequency, but overall it's been fine. The acoustician also did some testing after it was built and it performed much better than their analysis predicted.

 

[JLNJ]

Is it possible to use more trusses with lesser trib area and tie them together (similar to bridging)? Not the cheapest solution but you might get more stiffness and better damping.

 

[JLNJ]

Is it the main truss or the secondary or the combination which is giving you the trouble?

 

[calvinandhobbes10]

The ballroom floor area of concern consists of main trusses (81' span) and secondary composite beams (31' span).
By FloorVibe's numbers, the trusses control the bay by far (5.3 Hz). The secondary beams are not problematic (15-20Hz).

 

[SAIL3]

I would second JLNJ's suggestion to use more trusses with less tributary area if possible....having the outcome depend on one main truss would cause me concern...

 

[Agent666]

I think most classic vibration analyses use/recommend 10% of the live load, I would be wary of reducing this further (which it sounds like you are doing). It may pay to do some sensitivity analysis given the criticality of the problem.

Vibration is a funny beast, sometimes the theory doesn't align with observations, I have seen floors which meet all theoretical requirements, but then there is one floor bay which occupants complain about the liveliness of the floors. Sure enough when you get someone to measure the vibrations there is an issue that is really hard to explain via calculations. But the issue is real.

If it's an issue after working through the calcs, have you looked at tuned mass dampers, they can be quite effective in reducing vibrations.

Check out a product called Resotec as well, you put it between the concrete and beam/truss flange and it can increase damping considerably. It was invented/developed by Arup.