Design a beam by using Shell elements!!! 1

[jackyyau]

I am studying shells elements and need your help. How to design a beam with the SAP's shell element outputs, F11, F22 etc? see

http://home.i-cable.com/jacky.yau/BeamDesignbyShell.tif

The model:
12m span
1m x 1m x 0.6m thick shells
area loads = uniform 100 kPa (global gravity)
fixed at coners

Results:
F11=3600.45kN/m at mid-span bottom chord
F22=720.09kN/m
.... all moment=0 kNm
reaction at supports =300kN per corner, ok (agree with theory)
displacement at mid-sapn=0.00445m, ok (agree with theory)

The BS8110 design procedures (mid-span):
1. find the mid-span moment, M=100*12*12/24=600kNm
2. find K-value .....etc
tension steel required, As=1610mm2

But here, using shell elements we don't have moment, (no M33 in this case), then how to use F11, F22 etc for reinforcement design and check concrete section capacity?

if we use F11=tesion force=3600.45kN/m at this point,
the required As=3600.45x1000/0.87/460=8996.6mm2>>1610mm2 by moment method!

 

[ugoveri]

Hi jackyyau...

It depends a bit of what version of SAP you have...

If you have SAP2000 version 8 or 9, I would recommend you to use a very tool those version have.

For that you have to define groups and then section cuts.

The results given by SAP will then be integrated, which means that the forces will include the effect of the torsional moment.

What I mean is that m1 = m11 + |m12| and m2 = m22 + |m12| and the same with the forces f1 = f11 + |f12| and so on.

Of course you can use the forces obtained directly from the forces output in sap, but if the moments that produce flexion perpendicular to the plane of the shell are easy to combine, the bending paralell to the shell is a bit more difficult to obtain manually since you have to integrate the forces along the height of the element. That's why section cuts are very helpfull (version 7 doesn't include this tool).

There are 2 types of section cuts, the drawn and the defined by means of groups... I recommend you use the second because the first usually gives incompatible results...

I also would recommend you to refine reasonably the shell mesh because otherwise, the bending effect will be neglected in favour of the shear effect (You won't get the exact value for bending moments).

Hope I have helped.

 

[shakildor]

Jackyyau,
First of all I am confused that you have mentioned there are four supports. Are you placing the shell or horizontal plane so that it actually acts like a slab of 0.6 m thick and 1.0 m wide.

If I am right that you have placed it flat on horizontal plane, then M11 or M22 of shell will give the required BM to design the slab.

If you want to place it as deep beam, then you have to place it vertically and integrate F11 (or F22 ) force on the desired section to get BM as UGOVERU has said

 

[IYIeI2cY]

guys,
I'm glad you mentioned the section cut tool, however, i'm very confused using it, is there any solved Step by Step example of implementing that tool?

About the original question, i guess we shouldn't just take the value at the point, we could take the resultant force over the area (it should be some thing like a a triangle or parabola with). Just guessing, correct me if i'm wrong.

Thanks