School Floor Vibration Values

[Galambos]

I wanted to do a survey of those of you that have designed a school composite steel framed floor for vibration and see what values you are using for the Design Guide 11 method.

Which Beta, Dead Load, Live Load, and Damping Ratio values are you using and why?

From my own estimate, I would use Beta= 0.50%g, Dead Load = 4psf, Live Load = 4 psf, Damping = 0.20 for churches, but this is killing my weights.

Thoughts?

 

[lkjh345]

Recently did a composite deck/composite beam school building. I used the same Accleration limit, Dl and LL values as you but used a damping ration of .025 as I felt my building was a little closer to an 'open office' than a church becasue I have full height drywall partitions about 40 feet apart.

I ran into a vibration problem too. I ended up using 3" deck with 4 1/2" normal weight concrete topping (7 1/2" total). The last 1" of concrete was becasue of the vibrations.

 

[jike]

I have added one extra inch of concrete to many "open concept" school and office floors to help dampen transient vibration.

 

[271828]

Let's get some terminology straight first. Beta is the damping ratio in DG11. Also, we don't "dampen" vibrations unless we're trying to make them wet. It's "damp" or perhaps "damp out." Adding mass does not "damp out" anything--it's just that a=F/m, so adding concrete decreases the response. More mass doesn't increase damping.

Now that I have that off my chest, LOL:

0.5%g is ok.

DL = 4 psf is probably OK. The bottom line is that you need your absolute best estimate of the mass.

LL = 4 psf seems low. DG11 recommends 8 psf for an electronic office which contains almost nothing. Might be OK--might need increased to 8 psf depending on what you know is there.

Damping = 0.020 (definitely not 0.2!) is probably OK assuming you don't have full-height partitions. Might be a little on the low side. In reality, nobody anywhere, period, can tell you that 0.02 is right and 0.03 is wrong. The data doesn't exist.

 

[Lion06]

I would use an acceleration limit of 0.5% (because it's given as the limit in DG#1). I would use a damping of 0.025 (mostly because I think that pughs, and pulpets count for something), I would use a DL of 4 psf, but I would no LL (mostly because you can have one person praying when someone comes walking in with their footfall).

 

[Lion06]

woops, I gave my values for the church in lkjh's post. I would use the same values, except I would use probably 10 psf for LL.

I have found sometimes that adding mass makes vibrations worse. Has anyone else seen this?

It is true that a=F/m, but the a/g equation given in DG#11 has the natural frequency in it. I have found that sometimes the natural frequency is a bigger player and can trump the mass, such that adding mass makes things worse.

I would be interested in hearing if anyone else has ever come across this.

 

[msquared48]

Just have a ban posted on the user of any woofers within 100 yards of the school. That will help more than anything.

Mike McCann
MMC Engineering

 

[271828]

"It is true that a=F/m, but the a/g equation given in DG#11 has the natural frequency in it. I have found that sometimes the natural frequency is a bigger player and can trump the mass, such that adding mass makes things worse."

The natural frequency ends up in that equation by the following:

1. ap/g is a modified steady-state response to sinusoidal force. If you have Chopra, see the section for "Response to Harmonic and Periodic Excitations" with omega = omegan -- 3.2.2 in my 1995 model. If you take the steady-state maximum amplitude and multiply it by omega^2 to get to acceleration and then cancel out some things, you end up with a = po / (2*zeta*m) which looks suspiciously like DG11 Eq. 2.3. The po and 2 get absorbed along with reduction factors into the numerator of Eq. 2.3.

2. The sinusoidal force amplitude is the amplitude of whatever footstep force harmonic matches fn, often the 3rd or 4th harmonic. See Eq. 2.1.

3. The sinusoidal force is less for higher harmonics than for lower ones. This decrease in the sinusoidal force results in Fig. 2.2 and is where the e^(-0.35fn) comes from. Note that higher harmonics are required to match higher fn, so that's how fn gets in there--we're matching h*fstep with fn. h*fstep is what really needs to be in the e^(), but fn is there because we're matching them and it's more convenient (and more confusing).

Po*e^(-0.35fn) is the harmonic force combined with some fudge factors. See the paragraphs above Eq. 2.3. Therefore, if you have a higher fn, then the input force is lower, so it makes sense that the response is lower. Because fn is inversely proportional to the sqrt of mass, one might jump to the conclusion that lower mass is better, but this is not always the case. If the mass is low enough, then the impulsive response from individual footsteps causes high accelerations like everyone's felt on a cheap wood floor at some point.