iMechanica - reinforced concrete walls under cyclic (quasi-static) loading in ABAQUS - Comments

Convergence

Zach Thiemann — Wed, 14 May 2008 10:34:33 -0400

I'm running a model with solid elements and concrete damage plasticity model. I'm having problems with convergence as I make my geometry more complex. I've modeled a slab and I-girder as one solid element to model a monolithic connection between the girder and slab. When I run the same model with only a concrete slab the model runs fine and will converge. But once I redo the geometry and include the girder with the slab the model does converge after a fraction of the previous model. Any suggestions on what to change? I'm using the Riks method to solve the iterations. I'm trying to end with Force vs. Displacement diagrams for the model.

Need Help

m_meisam — Mon, 24 Dec 2007 08:14:15 -0500

How U define the concrete plasticity in abaqus?

Explicit vs Standard

Gabriel Sas — Wed, 21 Nov 2007 02:47:20 -0500

Thank you for the interest shown Mr. Wang. I will run an Standard analyses to see the differences.

why do you want to use Explicit for quasi-static?

Mingji Wang — Tue, 20 Nov 2007 22:07:33 -0500

You need a good control to eliminate inertia effects. Any accumulated error will become stress wave in the model. Try Standard which has stringent convergence criterion.

reinforced concrete walls under cyclic (quasi-static) loading in ABAQUS

Gabriel Sas — Tue, 20 Nov 2007 15:33:32 -0500

Hi,

I started modelling the
behaviour of the reinforced concrete walls under cyclic loading using
ABAQUS/Explicit. I used a shell element part with solid homogeneous
section and defined the concrete behaviour using concrete damaged plasticity
material definition. Embedded truss elements were used to define the steel
rebars. In 2 different steps (left and right) I defined the cyclic loading (in
plane) as imposed displacement (using amplitude-smooth step).

My question are:

1. Is it correct to define
the cyclic loading in this meaner?

2. How can I define an
unloading step between the left step and the right step, as the structure
in reality will have the tendency to move (partially) towards the initial
position ? Probably this is the reason why I'm receiving the warning The option *BOUNDARY,TYPE=DISPLACEMENT has been used with a jump in displacements at the nodes in node set WarnNodeWithJump-Step2 at the beginning of the next step; all jumps in displacement across steps are ignored.

If needed I can send the input file to those interested in the subject.

Best regards,

Gabriel