Downstand beam - Results from bar element and FE Shell

[lawrence_budziszQTYAB]

Hi Everyone,

I was hoping I could get your insides on highly talked about topic here – Downstand (‘T’) beams.

I have seen many post regarding modelling the beams using FE shell as slab and bar element for beam. I agree with the approach where Iyy of the beam is increased by the Iyy ration of rectangular to t-beam. There is a good amount of information on the forum regarding the modelling and analysis but I was unable to find anything on how to read the results.

My guest ion is on reading the results from the model:

Say I want to design the beam in my spreadsheet and just need the Bending Moment from my Robot model. In addition to extracting moments form the bar element (say, mid span moment from a simply supported condition) should the map moment form the FE shell representing slab be manually added to get the correct moment. Should this be taken as a panel cut taken across the with of the bar element?

I have included example below (middle beam):

Load case: 15kPa

Iyy increased by factor of 2 (as example only)

Pinned support conditions.

Moment in Bar element = 205.23 kNm

Moment in shell cut (over 300mm = width of the bar element) = 22.65 kNm

[Simau]

Hi @lawrence_budziszQTYAB 

I changed your example with stiffer beams and a thinner slab to be in the most common cases and to see easily differences. Do a comparison of the forces in both options.
Other approaches are of course possible.

[lawrence_budziszQTYAB]

Hi @Simau ,

Thanks for going into the effort to modify the model and producing a separate offset model. On the map results you have for the Mxx moment you can still see a moment of circa 26 kNm/m over the beam. compared to almost 0 kNm/m in the offset model.  I see that the offset model produces very small sagging moments in the slab (Mxx). Would you consider the offset case more realistic in terms of stiffness behaviour? 

Thanks, 

Lawrence 

[Simau]

Hi @lawrence_budziszQTYAB 

You should consider not only Mxx but also Myy moment.
The 2 methods are correct, but you need to take into account efforts in both beams and slabs.
Its more difficult for the offset method to design beams by sending efforts (My, Fx) to excel.
If you are interested only by beams, you may follow messge 5 of this topic:
https://forums.autodesk.com/t5/robot-structural-analysis-forum/excessive-bending-moments-on-columns/...

[lawrence_budziszQTYAB]

Hi @Simau ,

I dove into many articles regarding this topic on both Autodesk Robot forums and Engtip. The Myy moment for transverse direction is something I am looking into separately. My main focus at the moment is to get a detailed understanding of how to accurately extract Bending moment from a downstand beam modelled as bar element and shell. If you accurately calculate the ration by which the Iyy is to be increased in the beam you will still develop moments in the slab (in the same direction). I wonder whether that moment should still be added to the beam or ignored. 

This was an interesting conversation on the topic. 

https://forums.autodesk.com/t5/robot-structural-analysis-forum/how-to-model-integral-beam-and-slab/t...

[Simau · Accepted solution]

Hi @lawrence_budziszQTYAB 

There are many interesting articles on this topic on the forum.
In order not to push the open doors, I'll let you form your own opinion.