Modelling of continuous plate on beams

Post questions and ideas about modelling and the content in the Structure-tab

Moderators: Martynas Sudzius, StruSoft Fredrik L, StruSoft Bocskai, StruSoft Richard, StruSoft Iwona Budny

Anders S
Posts: 1
Joined: 08 Feb 2017, 08:41

Modelling of continuous plate on beams

Postby Anders S » 08 Feb 2017, 09:10


I have a question concerning how to model a quay structure where we have a continuous concrete slab with crossbeams that connects to three main beams. The deck is supposed to transfer the loads to the cossbeams whom transfer the loads the main beams. The crossbeams are 1200 mm high (from the top of the deck) and the slab is 400 mm. The deck crossbeams and slab is similar to a T-crossection.

How should I modell the slab and crossbeams to be able to get accurate results för Min. moment in the crossbeams and deck?

At the moment I use bar elements as crossbeams that are 1200 mm hich and 800 mm wide and then a continuous slab of 400 mm (I use eccentricity in my calculations as well). Is it a better idea to use bar elements that are 800 mm high (1200 mm - 400 mm) with a continuous plate of 400 mm on top of this?

Or use the bar elements 1200 mm high and plate of 400 mm in between with fixed boundary end connection to each crossbeam (so that the slab isn´t continuous over the beams)?

Thank you in advance!

//Anders S

Strusoft Hanif
Posts: 1
Joined: 21 Oct 2016, 11:17

Re: Modelling of continuous plate on beams

Postby Strusoft Hanif » 16 Feb 2017, 15:07

Hi Anders,

We do not recommend to use eccentric slab. If you want to use slab, then everywhere centric slab should be used. If it is down-standing beam, then why eccentric beam is not used? Using eccentric, it will get better results.


Andy Beeton
Posts: 10
Joined: 05 Dec 2016, 15:19

Re: Modelling of continuous plate on beams

Postby Andy Beeton » 02 Mar 2017, 11:22

I think what he is asking is something that bothers me a great deal.

You have a slab and beam that is cast monolithically. To correctly recover the internal forces for beam design, you need to have the slab joined to the beam. If you apply any sort of beam, rectangular or attempt to use an eccentric tee shape, you will have element volumes that overlap and that will affect the analysis (stiffness and internal actions). If you try to avoid an overlap, the only way to do it presently is to have a rectangular section and make it eccentric. If you do that, you have the choice of making the beam depth only equal to the downstand part, and give it the correct eccentricity, but you will find that this does not allow you to correctly design for shear as the beam is now not deep enough and the links/stirrups will be desiged on the lesser depth and some of the shear will go through the slab (how much of which is difficult to estimate). So, it is better, from the point of view of the beam, to use full depth but, to avoid the element overlap, you are forced to cut back the slab to the face of the beam and to connect the slab each side with line elements. This works quite well but is time-consuming to do. You can fill in the spaces between the beam and slab with covers and that helps to pick up distributed loads so you can still apply them fairly easily.

It's a pity that there is no way to make components that combine and integrate the internal forces of beams and attached elements so you could get 'T' and 'L' beam design forces for use with more conventional design methods. This would have to involve assumptions about effective flange width and so on so that the model knows what width of slab to combine with the beam.

There is some interesting observations on this issue in 'Finite Element Design of Concrete Structures' by G A Rombach, who gives comparissons of various strategies.

Return to “Structure”

Who is online

Users browsing this forum: No registered users and 1 guest