Shear Area

[scm256]

At my work we check decks and the stiffeners we fit under them. Up til now we've tended to use the area of the section as the shear area. We've been picked up on this and have hit a bit of a headache.

When you look at for example an I beam you take the shear area of the flange ignoring the webs, similar with square hollow sections etc only using the side in the direction of the load.

If you have a flat plate BS5950 says that you use 90% of the area. But where I'm getting confused is if you check the plate and say it gets utilised 102%, if we add a small flat bar stiffener under the plate the rules say that you only use the vertical part thus ignoring the plate that almost got you there, and then if the flat bar is looked at on its own it is over utilised more than 102%

Can anyone enlighten me as to how we should be looking at this problem, surely you can't ignore the deck plate as it has to be doing something.

Thanks

Scott

 

[rb1957]

i'm confused ... in-plane shear ? but that'd be carried by by the web, but not if it was transverse shear (then the flanges would be more effective) ??

if it's shear in the plane of the deck, surely the deck would handle it ? the stiffener being good for loads along there axis (tension, bending) and out-of-deck shear (since the deck is useless out-of-plane) ...

 

[scm256]

Sorry rb1957 I'll try again

If I am checking a deck plate, treat it as a beam and calc the bending stress based on the BM and shear stress using the sectional area then in many instances the decks are fine. What we are getting told is if we have the situation where we have to fit a stiffener under the deck then when we come to calculating the area we can only use the area of the stiffener and not the plate.

I'm of the opinion that the plate must be doing something but BS5950 says use 90% of area when theres only plate and 90% of the vertical part if the plate were stiffened, which by my mind will give a higher stress in a structure that you would think would be stronger.

 

[rb1957]

ok, but the deck isn't effective in reacting out-of-plane shear.

it sounds like you're talking about cross-section area to react a shear force aligned to the stiffener web. P/A is a very simple approach (it sounds like you want to calcualte the area and apply the force to it). i'd rather look at the shear flows to see the peak shear stress in the web. adding area to the cap of the stiffener (like effective deck would be) wouldn't particularly reduce the web shear stress (IMH un-thought-out O)

 

[scm256]

Sorry to be a pest rb1957 but what do you mean by "the shear flows to see the peak shear stress in the web"? I've always been told shear = P/A (and i've always used CSA) so I'm getting rather confused now that it appears that its not that.

 

[frv]

In the US, we tend to use "P" for an axial load, not a beam load, but assuming it's what we usually denote as "V", what you are referring to is average beam shear stress. Actual beam shear is VQ/It. The reason you usually ignore the flanges, is because Q is very small in the flanges. There is nothing inherently wrong with using the flat part- think about a flat plate in beam bending- it's resisting shear as well.

 

[scm256]

frv thats exactly my point. We've always assumed the plate helps and does something it just seems weird that adding a stiffener would then mean you start to ignore the plate.

 

[BAretired]

What you have is a number of 'T' beams. Check out this article for shear variation in a T beam.

According to CSA S16, the factored shear stress at any location in the cross section shall not exceed 0.66Fy and shall be reduced where shear buckling is a consideration.

BA

 

[hokie66]

Just a word of caution...for the stiffener to contribute in carrying shear into the support, it has to bear on the support or be connected to it. Otherwise, the plate will take the reaction from the support, thus must be capable of carrying the shear. Most floor plate stiffeners are for moment, not shear.

 

[rb1957]

what do you mean, "what do i mean" ?

we're talking about shear along the web for the stiffener, ie out-of-plane of the deck, right?

this is reacted primarily by the stiffener web, carrying the load to the frames, so the stiffener is in bending, right?

so i think it's obvious (and i think you and others have said) that the deck plate doesn't react this shear.

I think adding a strap to the flange, whilst increasing the cross section area, has little effect on reducing the shear flow in the web. maybe you need to review shear, and shear flows to understand that the shear stress in the web isn't constant, as fry posted VQ/It. I think adding the strap increasing both I and Q, maybe not by exacting the same amount but certainly by something similar.

There is an argument to be made that in a full on ultimate loading, the section won't fail in shear untill Fsu*As is applied (and it'll probably fail in some other mode before then).

i think another question you're asking is why count the deck when there is no stiffner, and not when there is ? the reason would be relative stiffness ... in the absense of a stiffner, all you've got is the deck (Fsu*Adeck); but when you have a stiffener, well, it is much stiffer out-of-deck plane and would naturally react the out-of deck load (way more than the deck above the stiffener) ... the deck plate above the stiffener is much more compliant and would "go with the flow". Now if you're in a full on ulitmate load, well you could add the 1 or 2 % that the deck might give you, but ... that assumes a lot of load redistribution, and plasticity and it'd sounds like you're really pushing the envelope.