1 Topic by francisco (edited by francisco 2013-04-30 10:40:24)

  • Registered: 2012-03-29
  • Posts: 9

Topic: bond in Reinforcement bars

Hello,

I'd like to model a reinforcement bar with a bonded length in a part, and the rest with a reduced bond (1/10 or something like this, in order to simulate the presence of a PVC sheeting). I've modeled this bar with two bars connected in one point, and I've set up that point to don't slip (flag).

In order to control if this way of modeling is correct, I've first assigned the same bond to the two parts of the rebar. I could observe a strange behavior in corrispondence of the connection point between the two rebars. In particular, I can see a concentration of stresses and cracks around that point.

Is this way of modeling correct? If not, how can I obtain the desired result?

Thank you for your help,

Francesco

Quote Post 1

  • dpryl
  • CC software developer
  • Offline
  • Registered: 2007-07-13
  • Posts: 1,107

Re: bond in Reinforcement bars

Dear Francisco,
there are 2 basic ways of separating a bar from the surrounding volume:

A. create an opening in the volume such that there is some space between the bar and the concrete. The opening can (and should) be simplified to a square or so, but even then, this way of modelling makes reasonable meshing more work.

B. restrict the automatically created connections between concrete and bars only to the Macroelements/Volumes the bar is connected to. This can currently be only done by manually modifying the ATENA Input File (.inp).  When doing this for the first time, send us your model + the correcponding .inp file and we will help you with the modifications.

dpryl's Website

Quote Post 2

3 Reply by francisco (edited by francisco 2013-04-30 18:41:29)

  • Registered: 2012-03-29
  • Posts: 9

Re: bond in Reinforcement bars

And what about my proposal of connecting two bar ends ? The simplest way is really a manual edit of the input files? Your first proposal (the opening) seems a bit expensive to realize. In fact, starting with around 3000 elements with only this little change the number of elements passed 6 thousand.

Quote Post 3

  • dpryl
  • CC software developer
  • Offline
  • Registered: 2007-07-13
  • Posts: 1,107

Re: bond in Reinforcement bars

The problem with connecting bar ends is that in fact, this is done by connecting both bars to the same neighbouring nodes of volume elements. We agree a connection of 2 bars with contiuous bond slip would be useful, but so far we were not able to find a good way of implementing it.

If you are scared by editing text files, we can help you optimize the mesh such that the opening does not introdcue so many additional elements.

dpryl's Website

Quote Post 4