Period Difference between SAP and OPENSEES

silvia · Post by **silvia** » Fri May 15, 2009 6:37 pm

how many displacement-based bc elements did you use?
why one the force-based element?

basir · Post by **basir** » Fri May 15, 2009 9:23 pm

[quote="silvia"]how many displacement-based bc elements did you use?
why one the force-based element?[/quote]

in sac-3 i used 15 displacement-based bc elements
sac-9 i used 60 displacement-based bc elements
in sac-20 i used 132 displacement-based bc elements

i will try force base elements and send the result subsequently

but dont you think all these diffrences in modal analysis results are just because of fiber sections?

silvia · Post by **silvia** » Fri May 15, 2009 9:50 pm

yes, the discretization of the fiber section and of the elment.
you should try the force-based element, as you only need one element per frame element.

basir · Post by **basir** » Sat May 16, 2009 10:10 am

dear dr silvia
i used force-based (elementelement nonlinearBeamColumn) with fiber sections and hysteretic material( also elastic perfectly plastic material ) and the modal responses like perioads of vibrations are the same as the dissplacement-base element with fiber section and hysteretic material( also elastic perfectly plastic material ) for all structures

silvia · Post by **silvia** » Sat May 16, 2009 10:18 am

excellent, it is a much more efficient model with only one element.
be careful with the epp material not to reach an overall zero tangent in your system.

basir · Post by **basir** » Sat May 16, 2009 10:22 am

[quote="silvia"]excellent, it is a much more efficient model with only one element.
be careful with the epp material not to reach an overall zero tangent in your system.[/quote]

excuse me i ddnt get what you said
my structure doesnt have just one element it made of several elements
but i hcanged all these elements from diplacement base to force based element

basir · Post by **basir** » Sat May 16, 2009 10:44 am

what i got through my investigation on diffrent structures demonstrates that fiber section increase periods of vibrations and i dont know why and i dont know how we should interpret its modal results or maybe how to modify these fiber section to give the responses which are reasonable in comparative with other softwares

results of (elastic sections or elastic element) are the same as what we get with other softwares

idont either know wich results are correct
results of (elastic sections or elastic element)

or

results of fiber sections

please help me dr silvia

silvia · Post by **silvia** » Sat May 16, 2009 11:25 am

yes, the numerical integration across the fiber section softens your system.

basir · Post by **basir** » Sat May 16, 2009 12:02 pm

[quote="silvia"]yes, the numerical integration across the fiber section softens your system.[/quote]

ok
so what should we do?
do you think the results obtain from modal analysis of models which has fiber sections is correct?
do we have to do something ?

silvia · Post by **silvia** » Sat May 16, 2009 12:22 pm

you might want to research this topic or talk to someone in person about this type of numerical simulation/integration.

basir · Post by **basir** » Sat May 16, 2009 12:56 pm

[quote="silvia"]you might want to research this topic or talk to someone in person about this type of numerical simulation/integration.[/quote]

tanx dr
may i ask you which element and sections you yourself prefer for pushover analysis?also nonlinear dynamic transient analysis?

silvia · Post by **silvia** » Sat May 16, 2009 8:00 pm

please look at the examples manual.

basir · Post by **basir** » Sat May 16, 2009 11:30 pm

i tried checking period with some modification in elements and sections

1-when we increase the number discritization of fiber sections in both orientation it affect the modal analysis and decrease period of structure

2-period of structure decrease when we change the number of integration points along the diplacement base nonlinear elements from 5 to 2

whats your idea about this

sac-3 structure period :
1-elastic element = 1.01sec

2-displacement base bc element with 5 integration points along the element+ fiber section which number of subdivisions of fibers are 16 and 4 and hysteretic material=3.01sec

3-displacement base bc element with 2 integration points along the element+ fiber section which number of subdivisions of fibers are 90 and 30 and hysteretic material=1.63sec

silvia · Post by **silvia** » Sun May 17, 2009 6:59 am

i really think you should use the force-based element, with 4 integration points.

basir · Post by **basir** » Sun May 17, 2009 11:37 am

[quote="silvia"]i really think you should use the force-based element, with 4 integration points.[/quote]
hi dr
i checked my model using forcebased element too
i checked lots of model and i solved most of my problems except one of them
when i use eigen command it give negative eigen values in sac-20 structure
i dont know why?