FEA concrete analysis torsional stiffness

[asixth]

Hi Guys

I'm unsure here about what torsional stiffness to use for a slab using RAM Concept. The options are:

1. Two-way slab - (KMrs is set to 100%, when I perform the analysis the results give 11.4mm longterm deflection).
2. No torsion two-way slab - (KMrs = 0.1%, longterm deflection = 25.4mm).

When I modify the KMrs manually, I get the following
3. KMrs = 10% ; longterm deflection = 20.8mm
4. KMrs= 25% ; longterm deflection = 18.4mm
5. KMrs= 50% ; longterm deflection = 15.6mm

Reading thru previus posts, RAPT suggests 30% (thread507-279687).

Is there an industry standard and has anyone had any real-life comparisions between FE and actual performance. This is a two-way reinforced member and there is nothing regular about it in terms of loading and support.

Any suggestions and further reading material will be appreciated.

Regards

 

[rowingengineer]

Asixth,
what method are you using for column stiffness?

http://www.nceng.com.au/

"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning."

 

[asixth]

25% for slab design.

 

[slickdeals]

You mean 75% Ig?

It’s no trick to get the answers when you have all the data. The trick is to get the answers when you only have half the data and half that is wrong and you don’t know which half - LORD KELVIN

 

[asixth]

No, for slab design I use a column stiffness of 25% Ig for the floor design (deflections). Punching shear design I put the stiffness to 100% and take the moment at the face of the column. But that is a different topic.

I'm not learned with FE formulations for plate elements but I am thinking that it is the stiffness which results in Mxy.

See attached.

 

[rowingengineer]

Asixth,
Normally I assume a value between 10-40% (if the fea program doesn't handle the reduction for me) depending on the loading situation and the thickness of the slab. However I do assume a different column stiffness than you (fair bit higher).

I do find it surprising with such a low column stiffness value you get such differing results, what other stiffness are being reduced during the analysis by the computer and you?. I have to assume you have some walls in your model, I would suggest you that model the walls stopping about 400mm from the corner intersection and see what happens. I would also like to know if the short term deflection has the same upscale multiplier?. I would also suggest you get out the book that I know you have called "finite element design of concrete structures".

http://www.nceng.com.au/

"Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning."

 

[SethGuthrie]

If your long term deflection is using the Load History method, then cracking of the slab is part of that analysis and I would tend to use the uncracked slab in the initial analysis. If you are not using Load history methods, but rather you are estimating long term deflections using other methods, then you could initially reduce the slab torsional stiffness like you were suggesting.

Be sure to review the calc log and/or the load history calc log for details in case there are any instabilities or section inequilibrium warnings for the case with the no-torsion slab in particular.

 

[rapt]

Asixth,

Mxy moments will still exist even if the torsion stiffness is reduced to zero. They are a result of the adjustment from principal moments to Mx and My moments. To describe the conditions fully Mx, My and Mxy are required to match the Principal moments.

In RAM Concept, you have to select to have the Mxy included in design. By default RAM Concept ignores them, incorrectly in the view of most experts on use of FEM for concrete structures.

 

[asixth]

Hi

Thanks for the input. I had a flick thru the Rombach book and it had some transformation of Mxy and so on but not a stiffness multiplier. What I did find was a paper from Concrete International in 2009 titled "Twisting Moments in Two-Way Slabs".

It says "Twist-free slab analysis will result in larger calculated deflections than those predicted by the analyses with non-zero twist stiffness". It goes on to say, "If the deflections from the twist-free analysis are significantly greater than those of the from the analysis with twist stiffness (for example by a factor of 2), it should be investigated whether excessive twist-related cracking might occur".

I'm going to go the middle ground here and use 0.3.