Dears,

I want to know how we can model the 2 initial nodes for modeling the link element. when I had done it I have this warning

"zero diagonal term"

Thanks,

Dears,

I want to know how we can model the 2 initial nodes for modeling the link element. when I had done it I have this warning

"zero diagonal term"

Thanks,

I want to know how we can model the 2 initial nodes for modeling the link element. when I had done it I have this warning

"zero diagonal term"

Thanks,

- bavafa612006 offline
**Posts:**12**Joined:**04 Dec 2013, 11:25

bavafa612006,

The two nodes in question (nodes 1 and 2 for a link element) are coincident - they have the exact same coordinates.

The structure must still be stable under the defined stiffness values for the link.

It sounds as though you have an unstable structure. This may or may not be related to your link.

You could try large (artificial) values for all link properties (all 6 DOF) and see if that takes care of the instability. If it does, then you know the problem is not your link. If it doesn't then there is an instability in the structure due to your modeling of the link.

That would be a starting place anyway.

Best of luck, bavafa312006.

The two nodes in question (nodes 1 and 2 for a link element) are coincident - they have the exact same coordinates.

The structure must still be stable under the defined stiffness values for the link.

It sounds as though you have an unstable structure. This may or may not be related to your link.

You could try large (artificial) values for all link properties (all 6 DOF) and see if that takes care of the instability. If it does, then you know the problem is not your link. If it doesn't then there is an instability in the structure due to your modeling of the link.

That would be a starting place anyway.

Best of luck, bavafa312006.

Tim Huff

- huffte offline
**Posts:**779**Joined:**22 Jul 2011, 10:19**Location:**Cookeville, Tennessee, USA

thanks huffte,

Another thing I want to know....is that how can we connect 2 buildings while in defining the gap element the nodes 1 and 2 for a link element) are coincident - they have the exact same coordinates.If you have any paper about that and if you have any idea please help me.

thanks,

Another thing I want to know....is that how can we connect 2 buildings while in defining the gap element the nodes 1 and 2 for a link element) are coincident - they have the exact same coordinates.If you have any paper about that and if you have any idea please help me.

thanks,

- bavafa612006 offline
**Posts:**12**Joined:**04 Dec 2013, 11:25

It sounds as though you could simply define two coincident nodes midway between the buildings at the desired floor level(s) with a rigid link (or very stiff axial element) between one of the nodes and Building "A" and a second rigid link (or very stiff axial element) between the second node and Building "B".

The link would have the two subject nodes as nodes 1 and 2 with appropriately defined nodes 3 and 4 (see the Help system for defining link nodes). The link would be assigned properties to model your gap.

Best of luck bavafa612006.

The link would have the two subject nodes as nodes 1 and 2 with appropriately defined nodes 3 and 4 (see the Help system for defining link nodes). The link would be assigned properties to model your gap.

Best of luck bavafa612006.

Tim Huff

- huffte offline
**Posts:**779**Joined:**22 Jul 2011, 10:19**Location:**Cookeville, Tennessee, USA

Thanks for your answers...they are so helpful. I have done it in that way, but I think there is a problem remain.....while I check the mode shapes ( to check the periods and run Eigenvalue analysis) it seems that the 2 structures are separated and not connected.

I do not know why?

I do not know why?

- bavafa612006 offline
**Posts:**12**Joined:**04 Dec 2013, 11:25

bavafa612006,

Naturally, the two structures are, as you say, unconnected for the eigenvalue analysis. An eigenvalue analysis is linear and elastic in nature, so it is not possible to incorporate nonlinear gap elements in the eigenvalue analysis.

If you perform the dynamic time history analysis it will be clear that the gap elements are behaving as expected. You may run with real-time depiction of the deformed shape and observe times of separation and engagement. And if you observe link force/displacement time histories in the post-processor, you will clearly see times of engagement and separation there as well.

So I do not see this as a problem.

To perform an eigenvalue analysis with the structures connected, you could simply create linear links (instead of gap links) with high axial stiffness and temporarily re-assign the links to have the linear properties.

So, conceptually, I see nothing wrong with your model. Of, course, it is up to you to make decisions as to whether or not your model accurately depicts the structure whose response you wish to estimate.

Best of luck bavafa612006.

Naturally, the two structures are, as you say, unconnected for the eigenvalue analysis. An eigenvalue analysis is linear and elastic in nature, so it is not possible to incorporate nonlinear gap elements in the eigenvalue analysis.

If you perform the dynamic time history analysis it will be clear that the gap elements are behaving as expected. You may run with real-time depiction of the deformed shape and observe times of separation and engagement. And if you observe link force/displacement time histories in the post-processor, you will clearly see times of engagement and separation there as well.

So I do not see this as a problem.

To perform an eigenvalue analysis with the structures connected, you could simply create linear links (instead of gap links) with high axial stiffness and temporarily re-assign the links to have the linear properties.

So, conceptually, I see nothing wrong with your model. Of, course, it is up to you to make decisions as to whether or not your model accurately depicts the structure whose response you wish to estimate.

Best of luck bavafa612006.

Tim Huff

- huffte offline
**Posts:**779**Joined:**22 Jul 2011, 10:19**Location:**Cookeville, Tennessee, USA

Dear huffte,

thanks for your answers.

I have an other question...

in order to define the damping for gap elements. As this element has not mass we have to use stiffness-proportional damping method. now I want to know what is exactly the stiffness parameter....can we use C/K as the stiffness parameter? C is damping the dash-pot and k is a axial stiffness of the colliding slabs. if you have any idea please tell me.( I used this value; with initial stiffness; and after the analysis there was no change in impact forces!!!)

Thanks,

thanks for your answers.

I have an other question...

in order to define the damping for gap elements. As this element has not mass we have to use stiffness-proportional damping method. now I want to know what is exactly the stiffness parameter....can we use C/K as the stiffness parameter? C is damping the dash-pot and k is a axial stiffness of the colliding slabs. if you have any idea please tell me.( I used this value; with initial stiffness; and after the analysis there was no change in impact forces!!!)

Thanks,

- bavafa612006 offline
**Posts:**12**Joined:**04 Dec 2013, 11:25

bavafa612006,

I am certainly no expert on damping levels induced into a system from opening and closing of gap elements.

That said, I would leave the local-specified damping equal to zero and let the global damping take precedence. This too is often set equal to "none", thereby making all energy dissipation due to hysteretic behavior of inelastic elements in the structure.

Without some sort of damage caused by the engaging of the gap element, I don't imagine that significant damping would be induced by the impact. Once the gap element engages, it would seem that hardly any additional displacement occurs in the element (assuming you have used a very large stiffness of the gap element) so that the area beneath the force-displacement curve would be very small.

Perhaps you had better search the literature on this one, but these are some initial thoughts.

Best of luck bavafa612006.

I am certainly no expert on damping levels induced into a system from opening and closing of gap elements.

That said, I would leave the local-specified damping equal to zero and let the global damping take precedence. This too is often set equal to "none", thereby making all energy dissipation due to hysteretic behavior of inelastic elements in the structure.

Without some sort of damage caused by the engaging of the gap element, I don't imagine that significant damping would be induced by the impact. Once the gap element engages, it would seem that hardly any additional displacement occurs in the element (assuming you have used a very large stiffness of the gap element) so that the area beneath the force-displacement curve would be very small.

Perhaps you had better search the literature on this one, but these are some initial thoughts.

Best of luck bavafa612006.

Tim Huff

- huffte offline
**Posts:**779**Joined:**22 Jul 2011, 10:19**Location:**Cookeville, Tennessee, USA

hi

i modeled a 2D Structure(3bay+3story).i defined linkes element in all the conections(forexmple for a beam between 2 columns, i define 2 linke at start and end of the beam in conection to columns). it means i have 42 linkes in this model.when i run the program, it stop by this massage "sol_prb", but when i delete just one of the linkes (specialy linkes in beam), then i can run the model without any problem.could you please help me to find the problem? is any limitation to defining the number of linkes? this fram defin in X-Z direction and all the columns and beams define linear.

NOTE: i added just a column to this frame without adding more linke(it mean i have 3stry+3bay with 42 linke and added a column on top of the frame.i run the program without problem.

i modeled a 2D Structure(3bay+3story).i defined linkes element in all the conections(forexmple for a beam between 2 columns, i define 2 linke at start and end of the beam in conection to columns). it means i have 42 linkes in this model.when i run the program, it stop by this massage "sol_prb", but when i delete just one of the linkes (specialy linkes in beam), then i can run the model without any problem.could you please help me to find the problem? is any limitation to defining the number of linkes? this fram defin in X-Z direction and all the columns and beams define linear.

NOTE: i added just a column to this frame without adding more linke(it mean i have 3stry+3bay with 42 linke and added a column on top of the frame.i run the program without problem.

- amir_nayebe65@yahoo.com offline
**Posts:**9**Joined:**29 Nov 2013, 13:20

amir_nayebe65@yahoo.com,

If you would like to send your model to me via e-mail, I will do my best to try and find time to look it over. I will be away for some of the holiday season, so I cannot make any guarantees. But I will try if you like.

Kind regards.

If you would like to send your model to me via e-mail, I will do my best to try and find time to look it over. I will be away for some of the holiday season, so I cannot make any guarantees. But I will try if you like.

Kind regards.

Tim Huff

- huffte offline
**Posts:**779**Joined:**22 Jul 2011, 10:19**Location:**Cookeville, Tennessee, USA

Users browsing this forum: GalenWak and 1 guest