Links between two vertical xlam shells - a problem of offset 1

[serra890]

Hello,

I've to model an xlam wall. This one is divided into two walls: first plane wall (2.95 m) and second plane wall (2.95 m). These walls are connected with non linear multiplastic link (hold downs and angle brackets). There is a problem: when I draw the wall above, Sap2000 merges this one with the wall below and so I cannot insert links. How can I divide 2 walls? How can I insert links between them in such a way as to restore the congruence? I attach an image to to better explain the problem.

 

[SteeliestFire55]

SAP2000 also has a auto merge tolerance (Options>Dimensions/Tolerances)

You need to check that the gap you've specified is larger than the auto merge tolerance.

You could try selecting only the top half of the shells only (without the nodes at the intersection) and move it away from the other half (Cltrl+m).

Btw, are you planning on using a 1 joint link or a 2 joint link to connect your walls?

 

[serra890]

I plan to use "two joint link". How long do the links? Is it important their lenght?

 

[SteeliestFire55]

Unlike frames, the length of a multilinear elastic/plastic link does not affect its axial or shear stiffness. Axial (U1) and Shear (U2 and U3) stiffnesses are defined per unit length.

The only effect is the bending at the ends of the link due to (shear*length of link).

See

https://wiki.csiamerica.com/display/kb/Link+FAQ#LinkFAQ-Doeschanginglinklengthaffectresults?

 

[serra890]

Thank you so much! I have another question about friction isolator on Sap2000. Ive a friction coefficient equal to 0.4. I model a friction isolator (non linear) with these following proprieties. Is it correct? From results I notice that there is no difference between results with isolator and without isolator. It seems I've not inserted it in my model. Why?

 

[SteeliestFire55]

Verify that all the load cases that you are running are set as 'non-linear'.

I'm not too familiar with friction isolators but since F=coefficient of friction*normal force, the link axial (U1) stiffness is required to calculate the normal force. In this case the U1 stiffness should be non-linear such that you have a gap for positive deformation and the compression stiffness for negative deformation.

 

[stressed]

With friction, you have to define a direction with a stiffness in order for the weight to act. Modify your friction link to add U1 and define a stiffness approximating the stiffness of what the link is connecting to. Also, unless you have it restrained, friction would typically act in 2 local directions, local 2 and local 3.

I've never modeled these types of walls, but is modeling with NL links "overkill" for this application?

 

[serra890]

Thank You for your answers. I've to model this xlam wall with friction at the soil . I activates the u1 and u2 direction but there is a problem : what values of friction stiffness I use in u1 and u2 direction? This friction isolator approximates the friction between wall and soil. The friction coefficient is 0.4.