Lateral Pedestrian Traction Loading 1

[phamENG]

I'm designing a maintenance platform - interior, no wind, seismic is negligible. I know there have been a few threads on this, but none seem to reach a rational conclusion and none seem to bring up an idea that I have, so I want to float it to see what others think.

Assume a 300# pedestrian (big guy carrying a big case of tools). If he's walking at 3.1mph (Wikipedia's average walking speed), or 4.55fps, is taking one step per second, and stops in one step, the force to stop him his:

F=ma
F=(300#/32.2ft/s2)*(4.55ft/s^2) = 42.4#, or about 14% of the applied gravity load.

This agrees pretty closely with the results found here, so I'm confident in the validity of my idea.

Now the question is how to apply it. In the linked article, they're looking at decks/balconies where a group of party-goers may do some sort of big, coordinated dance. This is an industrial application. No dancing, no synchronized marching, etc. Anyone have a suggestion for developing a rational application of this lateral load? If I design the platform for 100psf to account for personnel and material, a 14psf traction load on the whole surface feels like a bit much. Maybe a 60*0.14=8.4psf traction load based on a 60psf personnel loading for maintenance platform?

What do you guys think?

 

[jayrod12]

I could've sworn I've seen 5-10% somewhere in our Canadian codes/standards as a notional load on interior type structures. I always operated under the assumption that in most general cases this provides sufficient lateral stiffness to avoid occupant discomfort.

 

[PEinc]

You said that the walker's velocity is 4.55 fps. but what is his deceleration (negative acceleration)? To calculate the stopping force (F = ma) you need the negative acceleration, not the velocity.

http://www.PeirceEngineering.com

 

[phamENG]

Correct. I say he's walking at 4.55fps with 1 step per second and stops in 1 step. Therefore, his deceleration time is 1 second. Change in velocity (4.55fps to 0fps => -4.55fps) divided by time (1s) = -4.55ft/s2. And that's what I used in the equation above - though I did neglect the negative sign.

 

[phamENG]

There is a flaw, though - nobody has legs long enough to pull that off. Average pace is closer to 4-5 ft, , so you'd be looking at 2 steps per second. That means everything has to be doubled...

 

[phamENG]

Even so, that still puts me right at the article's peak cyclic loading reaction for a flexible structure. (12.1psf lateral load for a 40psf vertical load)

 

[BridgeSmith]

I would think that a reduction in loading would require a conservative assumption of the realistic possible lateral loading, which will be dictated by the dimensions and use of the platform. For instance, if it's 20' long and 3' wide and used as a 'catwalk', then 7 or 8 guys, with tools, walking briskly across it and all stopping suddenly would seem like a realistic possibility. If however, the platform is 20' by 20', the possibility of there being 20 guys on the platform may be unlikely. The chances of even half of those guys walking briskly (with their tools) in the same direction and all stopping at the same time, is probably nil. I suggest looking at it using a total lateral loading for connection of the platform to the structure, which I assume is what you're trying to assess the loading for, right?

Rod Smith, P.E., The artist formerly known as HotRod10

 

[phamENG]

Essentially. It's free standing, so I'm designing the bracing, etc. and connections to the foundation. It lands more in the 20' x 20' category, and I agree with your assessment - the question is what conservative assumption can be safely made? Realistically, there will never be more than 2 or 3 people on here actively doing the maintenance, and as you said - the odds of their motions being synchronized in a way that would develop a unified lateral load is hardly worth considering. But then, I don't want to fall into the "probable vs. reasonable" trap - it is reasonable that everyone will be so impressed with my design they'll all come stand on it at once!

I'm asking the question because it doesn't seem to be addressed in the codes anywhere - at least not in the US (thanks, jayrod!) - and I don't want to kill the design with an excessive load, but more importantly I don't want to kill a person with an inadequate one.

 

[BridgeSmith]

I understand your dilemma, but I'm not sure I can help you with the assessment of the possible/reasonable loading. I'm a bridge guy, so what little we do with pedestrian loading doesn't include lateral. We do braking force for vehicle loads - 25% of the weight of 1 truck or 5% of the truck plus lane (one design truck + other general traffic), but that's not really analogous.

The other consideration that I would take into account before I got too far 'into the weeds' on what loading is realistic, is what is the cost difference to design it overly conservative? Sometimes, there's very little difference (bigger bracing angles, larger anchor bolts, etc.) between designing for a realistic loading, which could take significant time to do, document, and defend, than it's worth compared to just throw the unrealistically large loading at it and being done, not having to defend it, and not losing sleep wondering about how it might be used.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[jayrod12]

I think if you've got a remotely competent lateral system for general stability, that you likely wouldn't be in the kill someone range, but rather in the make someone sick from motion sickness while on the platform. (I speak from experience on this one) So perhaps that can lower the hair on the back of your neck a bit.

What is the intended lateral system?

 

[EZBuilding]

Take a look at ASCE 7-10 1.4.3 which provides a minimum lateral force to be utilized in the analysis of the any structure. Could serve as a frame of reference against the loads you are coming up with.

 

[phamENG]

I guess I was using a bit too much hyperbole with the "kill someone" remark - jayrod nailed my goal for this case. You guys bring up valid points, but I'm stubborn and don't like taking the conservative approach without understanding why it's conservative - and I've always thought a big part of that is understanding why the other alternatives are not. While a specific design brought this question (back) to me, I brought it here for a general discussion to help bring about a better way of thinking about the problem in general as it seems so neglected by the literature. Thank you all for your thoughts so far.

Lateral system is simple knee braces - there's pipe all around that inhibits full height bracing - and I'm only about 9 feet off the ground. So it's pretty flexible, but not terribly so. Actual dimensions are closer to 11x11.

EZBuilding - thanks. I'm familiar with that provision. My concern with that in this type of application is that it is a function of structure self weight, not the actual usage. These platforms are all light structural steel sections (26plf at most in a lot of cases) with bar grating for walk surfaces. This results in a very low self weight, and a very low minimum lateral force. In fact, if you apply it to the decks in the article I linked to, you'd end up designing for a lateral load roughly 0.0083 times what the author ends up recommending (assuming a 10psf self weight, which is probably pretty high for a simple deck). So while it's valuable for a general stability check, I tend to discount it for most practical applications. After all, even here where we don't get earthquakes (to speak of) I don't think I've ever had a Cs as low as 0.01 - though I could be mistaken there.

 

[Denial]

I had a similar, but perhaps simpler, problem quite a few years ago when I was involved in the design of a very wide pedestrian bridge.[ ] One of my roles was a statistical investigation into the LATERAL load for which the bridge should be designed.[ ] I could find no such investigation in the literature, so I had to roll my sleeves up, grab a machete, and dive into the statistical jungle.

At the end of the exercise I wrote an informal description of my results.[ ] I attach this in its entirety.[ ] The sections that might be of help to you are "Dynamic loads applied by multiple independent pedestrians" and "Appendix A".

My pedestrians were restricted to walking north-to-south or south-to-north.[ ] Yours are able to walk in any compass direction they choose.[ ] However I think that, with a few simplifying assumptions, you should be able to frame your problem in a way that my formulae can be applied.

 

[3DDave]

Isn't the larger consideration that the natural frequency of the structure not match typical gaits of walkers. As I recall that Millennium bridge in London was far from structural failure but would become essentially unwalkable as pedestrians were forced to synch up or get knocked down, eventually reaching an amplitude that walkers were putting most of their effort into avoiding falling while being compelled to add energy doing so.

No one stepped onto the bridge with an intention to synchronize; the bridge design forced them into it.

https://en.wikipedia.org/wiki/Millennium_Bridge,_London#Resonance

 

[Settingsun]

The Australian standard for walking tracks is the only place I've seen a lateral live load based on floor area: 0.25 kN/m2. Otherwise the barrier loads or some percentage of total load is used.

 

[Lomarandil]

Construction work platforms, which are very similar to your application, use the greater of 50lbf/person (assuming a 250lbf worker with tools, 20%) or 2% of the total vertical dead load. (ASCE 37). These are very rarely checked for dynamics.

If you really want to bulletproof your design, elite rugby players are able to generate peak forces of about 2x body weight, and sustain 1x body weight. Competitive recreational players achieve about 75% of those figures.

----
just call me Lo.

 

[phamENG]

Denial - thank you! That's some good work and it is useful. In case this thread attracts others with similar questions/interests, I also found this paper: Modeling of Longitudinal Human Walking Force Using Self-Sustained Oscillator.

3DDave - that is a good point to bring up as it pertains to the general discussion. It also makes me wonder if the goal really should be to solve a stiffness problem rather than a strength problem. I think others have been saying this, but it took your phrasing for me to see it. In a bridge structure, it manifests itself noticeably as vibrations - or even resonant vibrations. In a larger platform structure where walking is more random, the strength in any given direction is not really a question, and the vibrations aren't likely to compound one another, but non-periodic accelerations resulting in random deflections in random directions could compromise the serviceability. This is where Denial's random walking formula can come in handy...

steveh49 - thanks. That converts to about 5.5psf. That's almost dead on my initial calculation (which was based on an impossible walking pattern) for a vertical live load of 40psf. I would assume, then, that it takes the probability of synchronized motion into account which is where the reduction comes from. With respect to my comments above, the 5.5psf is likely to provide sufficient stiffness to minimize the serviceability problems. Maybe that's a good place to land for surfaces where movements are most likely to be truly random. If this were a mezzanine in an office where a large staff announcement may draw a crowd to the rail or something in an area of public assembly, an increase in the direction of the stage/center of the room may be in order.

Lo- thanks. I've used 37 for wind loading on temporary structures before, but never found that provision. Very good to know, and it does seem to fall in line with what everyone else is saying. As far as a the rugby players are concerned - yikes. Though it seems that the 2x reaction is based on pressing against something (such as another rugby player). That would mean they'd have to be pushing against the handrail, so much of the reaction would be canceled out in the overall structure.

 

[spats]

Have you considered Chapter C of the AISC 360 Code?

 

[Phil1934]

Handrail has to take a 200# load (someone may fall against it) so deck should do that, too.

http://beta2.engrain.com/pie-merged/guardrail-deflection-limits/

 

[phamENG]

spats - of course. Though my question isn't about stability for stability's sake, it's about actual lateral loading that is likely to be applied. Simply designing for some notional loading (whether chapter C from AISC 360 or Chapter 1 of ASCE) may be suitable for most situations, but my goal is to better understand which situations don't fall in that category and need a more in depth look. I think that understanding is about all that stands between us and robots designing buildings for us.

Phil - thanks. That always seems to come up when this sort of question is asked. It's a load case to be sure, but probably not the controlling load case unless you have a very small platform. As we've discussed above, the loading should probably be somewhere in the realm of 5-15psf depending on use, so anything bigger than about a 6x6 would have more than 200#.