Modelling tower

02-Getting started with the modelling
Post Reply
camilocorrea
Posts: 5
Joined: 03 Sep 2019, 14:17

Modelling tower

Post by camilocorrea » 13 Sep 2019, 14:21

Hello, For a month or so I've been modelling a tower in seismostruct and doing a Push down analysis (z direction) when I run it, it often reach a point where it says it can't whithstand the force, but if I set another load factor or number of steps it may reach a higher displacement, also I check the elements reinforcemnt and most of the them havent reach failure so I'm not sure which is the maximum force it can whithstand.

Also another question I have is what is the difference between using links and off rigidity (in element connection), which one is better to use?

The model I'm talking about is:
https://drive.google.com/open?id=1fRORa ... 2pGC_NYSlb

If anyone can check it and have a sugestion, I'll be really greatful, the model is not the typical structure so thats why I have a lot of doubts

Best regards,
Camilo Correa

huffte
Posts: 840
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: Modelling tower

Post by huffte » 14 Sep 2019, 19:21

Hi camilocorrea. I looked the model over this weekend and couldn't identify any problems. I'm not sure what you are trying to simulate. But the loads are enormous. Shear failures are indicated in the performance criteria at a vertical load of about 59,000,000 pounds and the maximum attained load is about 80,000,000 pounds with significant inelastic behavior indicated by plotting hysteretic curves in the post-processor. It does seem unusual to try to displace the top of the tower 31 inches in the vertical direction. But I am not questioning your model.

As to the question of link elements versus rigid offsets in elements, I think the choice really depends on the behavior you wish to simulate. At least in my experience, I have used links where I need to model some inelastic behavior of an isolator or a shear mechanism in an eccentric braced frame. I have used rigid offsets when I have wanted to study the implications of a completely rigid panel zone on behavior. So I am unsure of how you are using the links in your model, even though I see that you have specified linear links with equal stiffness in all three translations and equal stiffness in all three rotations.
Tim Huff

camilocorrea
Posts: 5
Joined: 03 Sep 2019, 14:17

Re: Modelling tower

Post by camilocorrea » 16 Sep 2019, 22:47

Hello, thank you for replying

I'm using this model to replicated a real structure that collapsed so I want to reach a point to determinate for what load and displacement the structure starts to behave unstably.

Also another factor that I'm not really sure is that the middle part of the tower both arms are connected by 12 tendons, but in seismostruct there isnt a way to modelled prestressed elements, so what I did is I put 2 forces compressing the middle element with a steel section (using the equivalent area). I don't know if this was the best way to represent it?


Thanks for the feedback! Best regards,
Camilo Correa

User avatar
seismosoft
Posts: 605
Joined: 06 Jul 2007, 04:55

Re: Modelling tower

Post by seismosoft » 17 Sep 2019, 17:09

Indeed, applying externally compressive forces at the locations where prestressing is applied is a way to model it
Seismosoft Support

camilocorrea
Posts: 5
Joined: 03 Sep 2019, 14:17

Re: Modelling tower

Post by camilocorrea » 07 Oct 2019, 14:55

Hello Tim Huff, after checking the model do you perhaps have any suggestions to connect the several horizontal elements (the screen conformed by beams) between themselfs and also the question i put before about how do I know when a structure starts to behave unstably?





Thanks for all the help,

Best regards,
Camilo Correa

huffte
Posts: 840
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: Modelling tower

Post by huffte » 07 Oct 2019, 15:28

Hi Camilo. I am not sure what you mean by the "screen conformed by beams". As far as how to determine when instability occurs, I would say look at the hysteretic plots. For example plot tower tip displacement versus load factor in the post-processor. For a ductile structure, you will eventually see nonlinear behavior. If things start actually failing, you may see sudden jumps in displacement. Best of luck Camilo.
Tim Huff

camilocorrea
Posts: 5
Joined: 03 Sep 2019, 14:17

Re: Modelling tower

Post by camilocorrea » 28 Oct 2019, 13:57

Thank you for all the support, the other question I got from the model is that when i check strains graffics, in a few elements I'm getting behaviors completely out of normal and I dont understand why because all of othe elements are connected in the same way and its just in a few that i'm getting this response.

https://ibb.co/JHdtNCD
This first image shows the normal and expected behavior
https://ibb.co/brg5pMd
This is the anormal behavior of one of the elements


Thanks again for all the help,
Best regards
Camilo Correa

User avatar
seismosoft
Posts: 605
Joined: 06 Jul 2007, 04:55

Re: Modelling tower

Post by seismosoft » 29 Oct 2019, 18:35

Can the 'abnormal behaviour' be a matter or redistribution of forces? It seems like it. What happens if you apply a more stringent convergence criterion?
Does the analysis run up to the same stage (if not, then the strange plot is because of loose convergence criteria).
Seismosoft Support

camilocorrea
Posts: 5
Joined: 03 Sep 2019, 14:17

Re: Modelling tower

Post by camilocorrea » 30 Oct 2019, 14:16

Thanks for replying, I'm working with defaults so what would be a more stringent convergence criteria?

User avatar
seismosoft
Posts: 605
Joined: 06 Jul 2007, 04:55

Re: Modelling tower

Post by seismosoft » 31 Oct 2019, 13:09

You could try 1e-5m for displacements and 1e-5 for rotations.
Seismosoft Support

Post Reply