Seismosoft | Earthquake Engineering Solutions

Hi,

I've read the topics about cables and traction-only elements,
I've been trying to model SMa wires but with no success.
Basically i'm trying to do this basic model:
http://tinypic.com/r/ok3ds/8
I'm using truss element class for the SMA wires and link element with gap_hk on F1 and lin_sym with a high value of stiffness on the other directions (k=90000000).
1) I'm defining gap_hk with:
-Displacement d+ = 9999
-Stiffness k+ = 999E+012
-Displacement d- = -0.0000001
-Stiffness k- = 1.00E+012 (standard value)
On the stiffness k- value should i use the value (EA/L) of the sma wires ?
2)Defining the nodes, i'm using 5 nodes, one of them non-structural:

NODE| x| y| z| TYPE
1| 0,| 0,| 0,| structural
2| 700,| 0,| 0,| structural
11| 699,9| 0,| 0,| structural
22| 699,9| 0,| 0,| structural
4| 700,| 0,| 0,01| non-structural

The sma wire is connected from NODE1 to NODE22
The hook element is defined as:
Element End nodes
Node1: 22 ; Node2: 11
Element Orientation
Node1: 2 Node2:4

Restraints
Node Restraints
1 x+y+z+rx+rz
2 y+z+rx+rz
11 y+rx+rz
22 y+rx+rz
4 non-structural

I apply a time history (bi)triangular force (i.e. one cycle traction and one cycle compression) on node 2 with different values but still got no results.

What can i do to get the results that i want ( SMA with tension-only behavior) ?

Thanks in advance

Hi rgduarte,

The tension-only behaviour that you trying to get can be effectively achieved by using the lin_asm curve with a large stiffness for tension (e.g. 1e12) and zero stiffness for compression.
What the gap_hk element does is to restain the deformation after a certain displacement, which is why your model possibly diverges (you apply displacements that the gap_hk curve is preventing).

SeismoSoft Support

Thank you for the explanation. I've replaced gap_hk by the lin_asm suggested but i still have compression stress on the truss.
I've insert the link elements in both ends of the truss, one element on each end, and since i'm doing a simple 2D analysis i'm restraining all the structural nodes - y, rx and rz (And the end nodes are fixed) ! Besides that when i run the processor, a message appears stating that some of the nodes have zero elements though they're fixed by the restrains.

Is there any example of a successfully cable time-history analysis that i can get ? (time-history or any other type)
Or any way to get some help on the model (posting the model here or something) ?
Thank you.

Dear rgduarte,

Exampple 7 in chapter 2 of SeismoStruct verification report might be helpful for you.

Zoi Gronti
Seismosoft srl

Thank you very much, I've successfully corrected the model.