Uniform load on Lexan (polycarbonate) circular plate with simple support boundary conditions on edge 5

[epsilonlambdarho]

Hello Eng-tips,

This is my first inquiry, so I hope it to not be too juvenile.

I am tasked with analyzing a circular plate, simply supported on its edge, with a uniform load in the transverse direction. Essentially, it is being used as a catwalk. The material is Lexan, which has further complicated things for me. I have quite a few issues with this problem.

[ol 1]
[li]I am having trouble finding any type of aide/design manual for plastic materials[/li]
[li]I am also having trouble finding out what type of failure modes I need to be checking for such a configuration (bending and shear yes, but anything else? I have a background of steel, so I think of weird cases like block shear, yielding. etc...)[/li]
[li]When looking at the datasheet for Lexan which has all of the mechanical properties of the material, I am guessing these values were from a test which in-plane loading was performed. Even if I wanted to check, all of the ISO standards cost mucho money to look at even one, let alone the several that were used for all the different properties. I have a finite element model with the stresses of the plate, but nothing to compare those values to[/li]
[li]Not to mention that the displacement of the plate is 10x the thickness, so it doesn't even qualify under thin plate theory. Does this null my analysis results completely?....[/li]
[/ol]

So I have quite a few problems. Would it help to contact the manufacturer to get their take, would they know what resources I need to solve this? My thought too is maybe the mech engineering board would be a better place for this thread, as this is a very atypical problem?

 

[JStephen]

There have been several posts through the years on glass or plastic guardrails, that is the closest thing that comes to mind.

I would seek information from the manufacturer, and also would want to find some kind of code or standard for the application before proceeding. From the manufacturer, I would want to know, among other things, whether they would suggest or recommend it for this application.

On the displacement issue, typically, if the deflection is too large, it invalidates a bending analysis, but is often times conservative. However, I would think if you got deflections 10 times the thickness, that's probably not something you'd want to be walking on, either.

On of the questions to consider is how, or if, the material ages- what will it be like 20 years from now?

 

[rb1957]

properties ... if they're not readily available on the web ... buy some Lexan, and perform your own tests.
displacement > 10x thickness ... yes, you've invalidated your assumptions (ie small displacements, in particular for plate design ... zero stress on the neutral axis ... with large displacements, the plate reacts pressure with membrane stresses).

a circular plate is an odd choice for a catwalk ... maybe you mean annular ?

how far apart are the supports for the catwalk ?

another day in paradise, or is paradise one day closer ?

 

[epsilonlambdarho]

JStephen,

thanks for the input. I found out about the Structural Plastics Design Manual released by ASCE in 1984, but it has been discontinued for some unknown reason and not been updated, which makes me suspicious about looking into it. I sent the manufacturer an email, so hopefully they will have some input about codes, or if it should be used for a structural application at all.

To my understanding (which may be totally inaccurate), thin plate theory (which leads to a linear analysis in an FEA program) requires the deflection to be much smaller than the thickness of the plate. Using such an analysis, I get a deflection 10x the thickness. When I initiate a p-delta (first order) analysis, however, I get practically the same results. I don't really understand that. This was obtained assuming a 40 psf live load on the plate (a very typical value for floors in CE).

I'm not sure at all about aging of the material. A great question for the manufacturer, though!

rb1957,

requirement of testing isn't preferred by the client, but i'll keep it in mind if it becomes necessary...See above about my FEA results. The plate is 1.7m in diameter. There is a cylindrical tank in a chemical plant which must be cleaned. The bottom portion of it is taken out, a cleaner climbs into the tank and replaces the hole with the Lexan plate to stand on. That's about as much as I know.

 

[oldestguy]

For a single use thing I'd change the material to something you are OK with. How about plywood? Or reinforce the plate with something that is OK for that environment, such as fiberglass.

 

[rb1957]

as you've described the use ... why lexan ? why not steel grating ? (I thought it was an architectural thing (like that glass walkway in the grand canyon).

is the panel supported on the tank walls and/or on some central spindle ?



another day in paradise, or is paradise one day closer ?

 

[canwesteng]

There are codes for FRP design that would be worth looking at imo. You need to consider creep if there's any kind of dead load on the structure. Depending on the span, I think kirschoff plates may still be good for that deflection, but I may be wrong. You should be looking at some type of non linear modeling to capture the membrane stresses though.

 

[epsilonlambdarho]

oldestguy,

Thing about this is they've been using this for years. I'm guessing someone looked at it one day and was like, "Yeah, we better get this jerry-rigged lexan-stand contraption checked out." So we're analyzing an existing set-up.

rb1957,

I could certainly do a quick separate analysis and suggest that...yeah there's nothing architectural about it. Just a simple plate to make standing inside the cylinder to clean it easier. It is simply supported at about 90% of the radius all the way around.

 

[epsilonlambdarho]

canwesteng,

that's a good idea, thanks for the tip. the plate is used for very brief amounts of time (only during cleaning of the cylinder) and then taken off until it's needed again. so no dead load. I am using RAM Elements (the p-delta analysis option) and it seems to be spitting out approximately the same answers as linear. So yeah, I guess the difference is negligible.

 

[BAretired]

The simplest method is to apply load gradually until it reaches service load multiplied by an appropriate safety factor as specified by OSHA. If the plate fails, then it was no good anyway, so replace it with something that works.

BA

 

[canwesteng]

A caveat with BA's test to failure approach is that depending on temperature/UV exposure/age the plastic may behave differently, so use an appropriate safety factor. As far as if RAM is doing the non linear analysis correctly, make sure there are membrane stresses in the plate. I know a P delta analysis with STAAD will not correctly capture these effects.

 

[BAretired]

Does it need to cover the entire 1.7 diameter circle? Why not get a 4' x 6' piece of 2" deep FRP grating and cut the ends to fit the circle? That should provide sufficient room to do the cleaning and is easier to store. See the load tables in the following link:

http://www.fibergrate.com/media/3725/MoldedProductsBrochure.pdf

BA

 

[rb1957]

does it have to fit inside the tank ? maybe if it needs to travel up the tank ... but then supporting it would be difficult ?

maybe keep the lexan if you need to collect the wash water ?

another day in paradise, or is paradise one day closer ?

 

[epsilonlambdarho]

Excellent suggestion, BAretired. I think I will give that a try at them.

 

[structSU10]

RAM elements will show no difference with the P-delta analysis for a plate like analysis you describe. The plates will not induce any membrane action due to a vertical load on it, and P-delta is a iterative analysis that only will produce different results when you have an axial force with bending.

Also, is 40 PSF realistic for what they are using it for? Do they have more than one guy on this thing at a time? a point load of a few hundred pounds in the center may be a good place to check whats going on as well. If you are just checking this thing, you can analyze it to what works, and say "you can have (1) person within the tank for cleaning (or whatever its used for)" or something like that.

 

[MikeHalloran]

Uh-oh. CLEANING. Not good.

Lexan has one well known limitation; environmental stress cracking.

If someone spills a few drops of almost any chemical on a piece of Lexan, then immediately wipes it up, no harm done, right? Wrong; at some later time, the Lexan will suffer a brittle failure under a very small applied load, perhaps as small as a small dog standing on it. The typical mechanical properties supplied for new, clean, Lexan do not apply after it has been exposed to ... well, the list of aggressors is very long, and I have never seen data that would be useful for estimating a capacity after chemical exposure; I'm not sure there is any useful strength at all.

Example: A 1" x 4" x .010" coupon of Lexan is incredibly tough. You can fold it in half with your hand, and it will fight you hard, then spring back to nearly straight. Now dip it in alcohol, wipe it off, wash it off with soap and water, do any kind of cleaning you like, and try to fold it; it will shatter like polystyrene before you bend it much at all.

I personally would not use Lexan, or any other polycarbonate, or any plastic alloy containing polycarbonate, as a catwalk, or in any other structural capacity where life safety is involved and exposure to any chemical other than water is even remotely possible. But that's just me; I know of no statutory restrictions on its use.

It's starting to smell like the magic polycarbonate catwalk has already failed, and someone got hurt, or worse, and you are being sucked into a litigational vortex.
Tread, er, carefully.





Mike Halloran
Pembroke Pines, FL, USA

 

[epsilonlambdarho]

One more thing. I am thinking of the failure that the ladder leg would punch through the material. So all I am trying to do is find the transverse shear strength of the material. Remember this is a circular plate which is being loaded orthogonal to its plane.

I am having a surprisingly impossible time finding out such a simple thing. I already know what the stress will be, i.e. what the load is. I need to find what the material is able to resist, or its strength. Am I missing a really obvious equation here?

 

[BAretired]

I have never used Lexan and don't know its properties. Would they be as shown in the attached link? That is for Lexan 9034. Shear strength is given. Are there other Lexan designations? What is the thickness of the existing Lexan plate?

I don't think a uniform load of 40 psf is representative of the actual loading conditions. There will likely by one man on a ladder, maybe two but that is something your client can tell you about their normal practices when cleaning these tanks.

If you're worried about punching shear, you can always put a larger base plate on the ladder to get the shear stress down to an acceptable value.

BA

 

[BAretired]

I have used FRP gratings for a variety of decks and platforms in a Dow Chemical plant near here. The ones I used were extremely resistant to chemical attack of all kinds, much better than steel grating, but this is information you should be able to get from the supplier.

BA