Beam design in RAPT

[structuralex]

I have a 2-span continuous slab supported by walls on each side and a beam in the middle. The distance between the walls is 12m.
For the beam design: do I use a slab panel width of 6m, or should I increase the panel width by 25% to 7.5m? Because I realised that the middle reaction for a 2-span continuous slab is 25% higher than that of a simply supported slab due to moment redistribution.

 

[hokie66]

That is not moment redistribution, but is just the reaction at the centre support of two continuous spans. You should design the beam for the loading which it will see, which is 5/8 of the total length, so 7.5 metres tributary area.

 

[structuralex]

Thanks hokie. What about for a continuous slab with multiple spans of different lengths supported by beams (where there's no general rule of 5/8 etc)...
For the beam design: would you first model the continuous slab itself with transverse beams at the supports, obtain the reactions at each support (but first take out the transverse beams so as not to double-count the self weight), and then apply the reactions as a line load to each beam?

 

[asixth]

If it were 3 walls I would stay yes because the support in the middle will attract more load. 2 walls and a beam I would design with 6m panel because as the beam deflects downwards load will go back to the walls at either end.

 

[asixth]

Actually I would do it with the load distribution 7.5m. Just built up two runs to get the distribution and even allowing 10mm settlement for the beam is still attracts a fair bit of load.

 

[hokie66]

structuralex,
Yes, run the slab first, and use the results to design the beam.

 

[rapt]

structuralex,

Yes, the best option is to do the slab design first. This will confirm the slab depths etc so you get the correct loading on the beam. You could allow for support settlement in this as asixth suggested. That would depend on the expected deflection of the beam.

Then you could determine what width of slab each beam supports and allow RAPT to calculate loads based on this, or over-ride the program calculated self-weights in the general screen, and then define your own line loads for each load case.

 

[structuralex]

Cheers

 

[structuralex]

rapt: at some point along the beam there is a vertical stepdown, but only half the horizontal steps down. In other words the cross-section goes from a rectangle to an "L" shape. Is there any way to model this in RAPT? If not, how would you recommend analysing it? Cheers

 

[rapt]

structuralex,

You can use Trapezoidal Elements to model it. Void to remove the upper part of the slab to create the set down in the top over the width affected and solid to add the lower part below the slab.

For specific help on using RAPT, it would be better if you email me directly. I check email every day. I do not necessarily check eng-tips every day, especially when I am out of Australia as I am now!