Walking Pace/Speed for Floor Vibrations 2

[blondie_3]

I am looking for input on walking pace/speed for floor vibration analysis. The structure I am working with is a composite steel framed floor with typical 33'-4" square bays, so fairly long spans. We are using RAM Structural System for our floor vibration analysis. The criteria we are trying to satisfy is for sensitive equipment and is 6000mips.

In Table 6.2 from the First Edition of AISC DG 11, the following walking parameters are listed:
Fast = 100 steps/min
Moderate = 75 steps/min
Slow = 50 steps/min

In Table 6.1 from the Second Edition of AISC DG 11, the following walking parameters are listed:
Fast = 2.10Hz (126 steps/min)
Moderate = 1.85Hz (111 steps/min)
Slow = 1.60Hz (96 steps/min)
Very Slow = 1.25Hz (75 steps/min)

RAM's analysis shows a typical framing bay of W27x84 girders and W24x55 beams satisfies the criteria for a 75 steps/min walking pace. This is a moderate pace per the older DG 11 and while the framing members are fairly large, it seems reasonable to our team given the floor vibration criteria. The newer DG 11 would have us upsize members to satisfy a 111 steps/min pace for only moderate pedestrian traffic. The framing members for that criteria are astronomical and unreasonable (+W36's everywhere).

Using the older DG 11 walking parameters feels like bad practice, but showing W40's and W36's for the floor framing feels like a sure way to laughed out of a job. Has anyone else come across similar issues when trying to apply the walking parameters listed in Table 6.1 of the Second Edition AISC DG 11? Does anyone have any recommendations?

 

[phamENG]

This article discusses some of the changes between the two versions. They touch on that change briefly as to the reason why.

If that doesn't point you in the right direction, you could always try reaching out to Dr. Murray. I had a co-worker who studied under him at VT and he was supposedly pretty approachable if people had questions about his work.

 

[EdStainless]

We had some sensitive equipment on an upper floor.
The only way that we could make it work was to have the floors of those rooms structurally separate from the hallways.
People will often walk 100-120spm down a hall but rarely more than 50-75 within a room.
It was a pain, but the only other option was to provide isolators for all of the equipment.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[JoshPlumSE]

I am looking for input on walking pace/speed for floor vibration analysis. The structure I am working with is a composite steel framed floor with typical 33'-4" square bays, so fairly long spans. We are using RAM Structural System for our floor vibration analysis. The criteria we are trying to satisfy is for sensitive equipment and is 6000mips.

I'm not sure that RAM is going to help you all that much with the "sensitive equipment" aspects of Design Guide 11. I could be wrong, but my impression is that RAM merely implemented a "dumbed down" version of FloorVibe (Dr Murray's software, I believe) that could evaluate a floor panel for the walking excitation chapters (chapters 3 and 4 in the 2nd edition).

I say this because I worked for RISA around that time, and I remember that RAM's implementation was mostly an attempt to "feature match" some of the things that RISAFloor was doing at the time. It was the one aspect of RISAFloor that really encouraged RAM users to switch. RISAFloor could not do a good job on the sensitive equipment chapter either.

If you look at chapter 6 (the one for sensitive equipment), it is more concerned with analyzing the floor for varying speeds of walking. And, tends to use a velocity based acceptance criteria rather than an acceleration based criteria (as is done for chapter 3/4). I'd suggest you take a good look (you probably already have) at section 6.1.7 which concerns non-structural mean reduce response.

Lastly, I will 2nd the suggestion to contact one of the authors. Dr Murray is an excellent one for certain. But, Brad Davis is also a very good person to contact. Especially, if you're considering doing a slightly more advanced analysis of your situation.

Caveat: I work for CSI (which sells SAP2000 and ETABS) so it could be considered a competitor to these programs. Therefore, my comments should NOT be viewed as those of an impartial observer, but rather, as those from a potentially biased source.

 

[blondie_3]

Did some digging - RAM's user manual states that it uses the Second Edition of DG 11 (Section 10.1 of the Steel Beam Manual), but if you dig into the outputs it is actually following the First Edition's provisions and equations. Bentley has stated they are planning to release an update that incorporates the Second Edition design provisions (not sure why their user manual states this is already incorporated). I'm not sure how much this is affecting the results RAM is producing, but it does pretty much eliminate RAM as a possible software for this analysis. We are looking into possibly buying FloorVibe and carrying out our analysis with that.

I am not necessarily interested in doing an analysis beyond the DG 11 provisions, I just want to be as sure as possible we are using appropriate criteria. I asked about the walking speed specifically because it makes a huge difference in the design results and there is a large discrepancy from the First to Second Edition of DG 11. It sounds like our best option may be to do manual calculations for the vibration analysis rather than using a software.

 

[JoshPlumSE]

One of the things that 6.1.7 suggests is ways to limit the walking speed. Alternatively, that probably means there is criteria you can use (in the papers upon which DG-11 is based?) to justify using a walking speed less than 111 steps per minute. Based on the length of hallway, or various turns / obstacles you may not need to use that higher walking speed.

 

[JLNJ]

Your spans are right in that problematic 30'-35' range. Best to change that if you can.

 

[271828]

How are you choosing what walking speeds to use? These have to do with walking paths. If wide open like corridors then fast clearly applies. If the paths are cramped, then very slow. It's hard to sharpen the pencil enough to decide when to use slow or moderate.

What damping are you using? Are there lots of partitions?

Are you computing fn from the dunkerley formula in Chapter 3 or the min of the beam and girder frequencies?

 

[ThomasH]

In my experience the natural frequency of the floor is the main parameter for this type of analysis. If there is a risk for resonance and you don't include it due to a presumed difference between walking speed and the floors frequency, then you might have a problem.

 

[FE_struct1]

A couple of quick questions OP -
[ul]
[li]You mentioned that the vibration requirements are for sensitive equipment. How strict are these requirements ? Are you able to recommend that all sensitive equipment be mounted on vibration isolation devices (if they're not already) ? Or are you also required to reduce human-perceptible vibrations down to the stated requirement of 6000 mips ?[/li]

[li]What values of damping are you using ?[/li]
[/ul]

In my experience, trying to reach vibration criteria by increasing stiffness can produce very bulky sections. Additionally, making your beams stiffer could actually make vibration response worse, since it will reduce mass participation of slabs from adjacent bays (think of it as your vibration getting locked into your bay due to very stiff beams). It's usually much easier to simply add mass to your system as this will increase your inertia, but meeting minimum natural frequency requirements could get tricky.

I'm not based in the US so I don't know what limits DG-11 provides, but to my knowledge achieving a natural frequency of 6-8Hz is usually sufficient.

 

[blondie_3]

FE_struct1,
The criteria is 6000mips is for booms and other equipment that will be in the occupied spaces. We are providing springs and neoprene pads where we are able to, but most of the equipment cannot utilize neoprene pads, springs, or other means of damping.
We are using a 5% damping ratio right now. We have discussed increasing this some since the building is going to be packed with MEP equipment and has a high number of partitions. A concern with this is damping is hard to quantify and if we use a non-conservative value, occupants may experience vibration issues.

We are using a 4 1/2" concrete slab over 3" deck (7 1/2" total thickness) which is fairly thick. We discussed possibly using high density concrete but that introduces so many other issues that we nixed it.

 

[blondie_3]

271828,
The walking speed is the crux of my question. We are using walking speeds based on the DG 11 recommendations. We do have corridors within the same bays as the rooms that will have equipment, but those corridors are curved and have many doors (likely more stop-and-go traffic than one long walking path). Perhaps a moderate speed can be justified considering this, but even then we are left at 111steps/min (per Second Edition recommendations) which blows up the framing sizes.

As previously mentioned, we are using 5% damping and there are many partitions and lots of MEP equipment. The same concerns about using a non-conservative value if we increase our damping apply.

Our current analysis used the Dunkerly approximation since RAM still does vibration analysis per the First Edition of DG 11, but this will be revised to the minimum beam and column mode frequencies. I'm not sure how much of a difference in the results we should expect to see based on this.

 

[271828]

Are the corridors through the middle of the bay? The predicted velocities ara at the middle unless you apply the phi factors.

Going from the Dunkerley fn to the min of beam and girder fn should make a big difference.