Seismosoft | Earthquake Engineering Solutions

I'am using the gap_hk link element to model the pinned support with vertical movement prevented and downward movement allowed (dependent to the stiffness of the column bellow).

The support is of course at the coulmn top where i have placed two nodes with same coordinates of which one is connected to the column, other just restrained from any movement. Other nodes needed are on the structure.

I'have valued and iterpreted gap link parameters for directon considered (F3) on following manner:
d(+)=0 - gap activated from beginning,
k(+)=10e15 - vertical movement prevented
d(-)=0 - gap activated from beginning,
k(-)=0 - no gap stiffness in downward direction i.e. lefted for the column bellow

Please tell me if my interpretation is wrong and what should be the right one.

Thank you!

Provided that you have used the default orientation nodes for the link, your model seems correct to me.
In any case you can test your model by applying a permanent vertical load to the lower node of the column. If the load is applied downward, you should get vertical dispacement, while in the oposite case you should not. Note that the rigidity of the link k(+) should be about 1000 times greater than the axial rigidity of the column in order to practically achieve prevention of vertical movement.

What if the one point of the gap link element is constrained, while the other of the same coordinates is on the top of the column.

Should then the stiffness of the gap be valued also zero in vertical direction? I have tried the analysis and it seems that gap doesnt recognize that one of his points is constrained.

Also, even though the force acts downwards and gap is defined to be downward free movement while vertical movement is constrained, part of the applied vertical load is taken by the gap and other by the column. The whole load is applied by the coulmn after some horizontal displacement of bare frames that i'am analyzing, occured.

Thank you!

Dear dpenava,

We looked into your model, changed the gap curve parameters to '0 1e15 -1e15 0' and the analysis ran without problems.

We also take this opportunity to note that:
- using 600 increments in a pushover analysis is hardly justifiable and renders the analyses unnecessarily slow; 30 steps will be more than sufficient for your model
- using short infrm elements (in your case, the column ends feature a length of only 10 cm, half of the cross-section dimensions) may lead to unexpected concentration of inelastic deformations; do consider these issues with care (see discussions in Help System and in other Forum posts)

Seismosoft Support