Element classes and damping

02-Getting started with the modelling
Post Reply
mohamad_nazari
Posts: 9
Joined: 08 May 2015, 20:58

Element classes and damping

Post by mohamad_nazari »

Hi my dear friends
I have 2 questions:
1-how can i recognize an element is a DB class or FB class for example a column of a bridge in which group will be classified infrDB or infrFB?
2-in setting and in part damping I assign rayleigh damping.when i determine element class i should determine its damping type.here if i define none damping is not it a problem?does it make trouble?
thank you for your attention my friends
huffte
Posts: 978
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: Element classes and damping

Post by huffte »

Either DB or FB inelastic elements could me used to model a bridge column. FB elements offer the advantage of typically requiring no discretization, or subdivision, into smaller elements. Chord rotations are immediately available for a non-subdivided element. These issues are covered in the help system and I highly recommend you view the excellent help system prior to finalizing your model. In my experience I have found FB inelastic elements very useful in bridge columns, and DB element solutions which I have done seem to converge to the FB element solutions as the number of subdivisions in the DB element is increased. Again, rely on the Help system for more information particular to your situation.

As for damping, it really depends how much inelastic behavior is expected and there is no clear cut answer. You will have to use your judgment and read up in the help system and in the literature in the final analysis. There are strong arguments in favor of tangent-stiffness-based damping, and SeismoStruct offers that option in the settings. The default is no damping, but a small amount is often required to obtain a stable solution.

Best of luck, mohamad_nazari.
Tim Huff
mohamad_nazari
Posts: 9
Joined: 08 May 2015, 20:58

Re: Element classes and damping

Post by mohamad_nazari »

Dear Tim Huff
Thanks for your complete responding.
but in damping I mean the difference between defining damping in setting and element class which I do that.is it a paradox?because I define a damping type in setting( rayleigh) and determine another damping in element class.is it make unstable solution or establish an error in analysis process?
huffte
Posts: 978
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: Element classes and damping

Post by huffte »

As noted in the excellent Help System, "damping defined at element level takes precedence over global damping". It is certainly possible for one damping formulation to produce numerical instabilities, while another might not.

I am not sure if this answers your question or not, mohamad_nazari. While I seldom employ element-level damping, there certainly may be specific cases where element-level damping would be preferred, and perhaps you are studying such a case.

Best of luck in any case.
Tim Huff
Post Reply

Return to “02-Getting started with the modelling”