Rigid or semi-rigid diaphragms?

[FEPI]

Hi, we need some help.

Usually we analyze this kind of structures. (See a simple model attached, this is just an example, This is NOT a project we are working on).

This kind of structures consist on HSQ beams supporting precast concrete slabs, using vertical concrete cores and vertical bracing to resist horizontal forces. This is the only horizontal stability system. The main system consists in rigid diaphragm floors and vertical concrete cores.

Often our mission consist to deliver:

1)     Reactions to foundations

2)     Steel columns and steel HSQ beams dimensions. (We use a specific program to calculate these beams, so we only take the ULS forces diagrams from ROBOT acting on the beams as an input).

3)     Forces acting on the concrete cores (for the companies working on the precast concrete panels). We don’t dimension any concrete element. (Not concrete slabs, not vertical concrete cores).

4)     General stability check for the structure. Global displacements, structural behavior, etc.

We are trying to find the best way to work with ROBOT because is our main calculation software.

For our models, we use claddings (because it is light, easy to change loads on, no need of shell elements in very large models) and fake bars to simulate the stiffness of the concrete slabs. By doing this we obtain the vertical loads for beams and also the horizontal transfer to the concrete cores, without using shell elements to modeling the floors.

Questions:

1)     Do you think that there is a better way to model this rigid floor effect and at the same time loading the beams, using rigid or semi-rigid diaphragms for example?

2)     Is there a possibility to define a stiffness for a cladding? So we skip the creation of dummy bars. And still we are not using horizontal shell elements.

3)     Which is the best and easier way to present results on the concrete cores? We have lots of trouble trying to extract results from vertical shell elements.

Can you help us with this? Maybe give us some examples?

Best Regards

[Rafal.Gaweda]

Hi

If you have not received my e-mail I will post the answer here also in a minute

[Rafal.Gaweda]

---

@FEPI wrote:

1)     Do you think that there is a better way to model this rigid floor effect and at the same time loading the beams, using rigid or semi-rigid diaphragms for example?

If you want to consider fake beams for applying loads you have to consider them - Geometry \ Additional attributes \ Load Distribution Options

Diaphragms creates rigid links between nodes so in such case (fake beams not loaded but used as a stiffeness elements) you do not need them.

2)     Is there a possibility to define a stiffness for a cladding? So we skip the creation of dummy bars. And still we are not using horizontal shell elements.

Stiffeness for the cladding can be considered only as diaphragms (rigid links). If you want to apply real thickness of panels to claddiings it is possible but it will be considered for selfweight calculation only unless you mesh such claddings (-> Geometry \ Properties \ Panels Calculation Model)

3)     Which is the best and easier way to present results on the concrete cores? We have lots of trouble trying to extract results from vertical shell elements.

 What do you really need?

 You can present results as for every panel - tables, maps ; as Panel Cuts ; table of Reduced Results ; results for Cores;  Diagrams for Building ; also for walls in Story Table.

Best Regards

---

[Robot1985]

I also have this issue on a steel frame building.  I am using Robot 2011.

I have steel beams supporting a concrete floor.  The design philosophy is that Horizontal forces will be transferred by the concrete slab, while vertical forces are carried by the steel alone with no composite action assumed.   One slight complication is the slabs are inclined, hence the diaphragm is not in the global axis.

I would like to use panels to:

a) apply and distribute  horizontal loads to columns (eg. horizontal panel edge load).

b) lock column nodes together (ie rigid link in x-y direction of panel plane)

c) transfer vertical loads to steel beams (ideally one way, but .

However, i do not want the slabs to act in flexure.

I have tested the various panel calculation models but without much success;

The flexible diapragms model  appears to work in composite action with  the steelwork, similar to shells.

The one way slab model was not providing correct distribution of vertical loads to the structure and also appeared to work in  composite action

I would be grateful for any advice on the best way of modelling this using the panel command.  I have a work around using rigid links and manually calculated beam loads but I think there must be most be an easier way!

Many thanks

[Rafal.Gaweda]

Hi

We hardly see the real phisical behavior of this model, but try this:

- model slab as shell with one way loads distribution (-> panel calculation model)

- connection between inclined slab and beams - compatible nodes with proper selection of dofs and correct rotation of compatibilities

- lack of panel flexure: either by Material Orthotropy with very small thickness (h) and big stiffeness = increase of n1\n2 factor(s) or by user defined orthotropy -> play with Flexure Stiffeness Matrix

If it is not what you expect please privide any sketch \ real example how this should work \ look like etc.

[PaulGrimes2282]

Hi,

Sorry to resurect this old thread, but it seems to be the closest to what i'm trying to achieve. I also have a panel acting as a diaphragm but without composite action with the beam, so i have followed the instructions from your final post. In applying the releases i have chosen to release Rx and Ux to avoid any composite action, however this is giving me model instabilities:

 - if I reduce the material stiffness coeficient n2 below 0.1, i get a type 3 instability (Max min values on stiffeness matrix diagonal are greater than 1e12 here, but no short elements - checked wtih add in)

- if i keep it at 1, and release edges as discussed i get Ry instabilities along one side of the panel (i.e. nodes along one of the supporting edge bars rotate around bar x-axis. Bar is torsionally restrained at one end)

Please could you have a look at the model and see what i'm doing wrong?

thanks!

[Pawel.Pulak]

Explanations to questions above are given in another post created by PaulGrimes2282:

http://forums.autodesk.com/t5/Autodesk-Robot-Structural/Releases-to-the-edges-of-panels/m-p/4351410#...

---------------------------------------------
If this post answers your question please click the "Accept as Solution" button. It will help everyone to find answer more quickly!

Regards,