Design load to check lateral stiffness of walkway 1

[CTW]

I'm designing a walkway that spans 18 ft with a 100 psf load. I want to ensure that the walkway is stiff and does not sway too much side to side. I've seen and walked on too many long span walkways that have significant lateral movement. Although it's structurally safe, the movement makes people nervous. What load should be applied in the lateral direction to check the stiffness in this direction?

 

[JStephen]

Have you checked wind loading on it?

 

[CTW]

Indoors. No wind load. I could apply an arbitrary wind load to it, but there is not that much surface area to apply this to, thus the overall effect will be small. The walkway is simply two channels or beams with grip strut planks as the walking surface and handrail along the sides.

 

[jheidt2543]

The Steel Joist Institute has studied floor vibration extensively and developed a proceedure for checking the "bounce" in joist supported floor slabs. SJI Technical Digest #5 (1988) and AISC/CISC Steel Design Guide 11 (1997) discuss in detail the methods. While they apply to joist supported floors, you could adapt the method to beam supported floors too. You can check the SJI website for more info.

 

[IFRs]

OSHA requires that the handrail system resist a 200 pound load anywhere in any direction. Perhaps this could be the basis for a side load. You might also want to determine the resonant frequency of the assembly and make sure that it is well out of the frequency of normal walking / running.

 

[Grizzman]

How about seismic loads? Those would induce lateral movements.
What loads do you think will generate lateral movement? Perhaps a percentage of the live load. Maybe 5%.

 

[PSlem]

The 50#/lineal foot ICC code for handrails is probably a more accurate model. A person swaying might generate that. I would think a zig zag diagonal tie on the underside should take care of it.

 

[steve1]

Provide a horizontal bracing scheme between your two beams. Attach the web members to the underside of the top flange of your bems. You don't need a design load for this, just do it!

 

[UcfSE]

What kind of main supports do you have? Instability in those members could cause it to sway. Even if the beams are connected together, assuming you have beams, they can still buckle together and cause sway under load. If it were a wide flange for instance you could design it as unbraced. That would give you a stiff fat beam that shouldn't sway. The horizontal bracing that steve1 suggested sounds like a good idea to me, maybe with a certain percentage of the gravity load acting as the lateral load to give you somthing to design with. That's probably cheaper than a giant beam depending on exactly what you need.

 

[CTW]

I always provide lateral supports for the grating that are a minimum of 3/4 the depth of the main members, but is this enough to provide a stiff system? I've considered diagonal braces, but I was hoping there was a reasonable load I could apply to check the system without the diagonals to avoid unnecessary costs. The percentage of the live load, the 200 lb OSHA handrail load, and the 50 plf load sounds like a good starting point.

 

[Qshake]

At the end of the day, you could apply several imaginary loads and still not ensure you've met the requirements that you've specified for yourself.

If you've noted lateral displacement to be an issue then what you really want to do is limit lateral displacement. You need a lateral deflection limit, not more load.

I'm sure there are several articles written on the subject as there are deflections limits for beams in flooring as well as limits on vibration characteristics.

Another example is that highway bridges with pedestrian traffic are limited to L/1000 in deflection.

Look to the codes for examples of where deflection limitations are provided and see how they can be applied in your case.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[CTW]

From Qshake:
"At the end of the day, you could apply several imaginary loads and still not ensure you've met the requirements that you've specified for yourself."

BINGO!!! This is what I was hoping to avoid. I don't guess there is a known load that can be applied to check whether or not the system has enough lateral stiffness.

Another thought I had was that if I have two main channels oriented like so ][ with a 3 foot spacing between them and lateral members (say channels) between these two channels that are the same depth as the main channels (spaced 3-4 feet apart), wouldn't this form one big beam? If so, the 3 foot depth would make this walkway have sufficient lateral stiffness.

 

[jhardy1]

CTW,

You need to consider the possibility of dynamic response of the bridge to pedestrian-induced excitation.

Footfall frequency of pedestrians is typically in the range of 1.5 to 4 Hz, depending on their walking speed. Due to the mechanics of body balance, pedestrians also create lateral forces at half the footfall frequency (in the range from 0.75 to 2 Hz). If any of your significant mode shapes have natural frequencies anywhere in this range, you need to seriously consider a dynamic check. In particular, any walkway with a lateral sway frequency less than about 1.3 Hz is possibly prone to “synchronous lateral excitation”.

A key factor is whether the bridge is potentially subject to crowd loading. Loosely scattered pedestrians tend to walk with random, uncorrelated steps (as long as we are not dealing with a squad of marching soldiers or similar). However, if the crowd density gets high enough, but not so high that movement effectively ceases, people tend to subconsciously fall into step with each other, to avoid tripping over each other’s feet. (This phenomenon has been called “syncopated shuffle”.)

If the frequency of footfall closely matches a natural frequency of the bridge, especially a lateral mode shape, the bridge will start to sway in response, and then the frequency of the pedestrians’ footfall will tend to “lock in” to the bridge lateral sway frequency, as people subconsciously match their footfall to the movement of the bridge. This phenomenon can lead to dramatic swaying of the structure, as was exhibited on the opening day of the London Millennium Bridge in June 2000, for example.

The subject is very complex, but the magnitude of the lateral force induced by a walking pedestrian is typically in the order of 4% to 10% of the person’s weight, depending on whether the bridge is “static” or swaying noticeably in response to the pedestrian excitation.

Here are a couple of papers to get you started:

http://go.rwdi.com/technotes/t15.pdf

http://www.arup.com/MillenniumBridge/indepth/pdf/IStructE.pdf

Hope this helps!

 

[jhardy1]

Something went wrong with my secondf link - try this:

http://www.arup.com/MillenniumBridge/indepth/pdf/IStructE.pdf

 

[Grizzman]

CTW:
In order for the channels to act as a singular 3ft member, you'd either have to provide diagonal bracing or the strut to channel connection would need to transfer moment. In this circumstance, you're bettor off using WF's and letting the erector field weld diagonal L's to the underside of the top flange. This method of stiffening a walkway should not raise the price very much. fabrication cost is not increased, and the erector is given some leeway in how he puts it up. Of course there could be asthetic reasons for not doing it this way.
If the struts are the same depth as the beams, the eccentricity of the L's won't shouldn't much of a problem.