Dis-continous hysteresis

[hasanozkaynak]

Thanks alot to both of you,

I will strictly follow your suggestions and let you know the latest situation.

Thanks and best regards...
Hasan

[hasanozkaynak]

Hi,

I passed through your suggestion, thanks for your attention again..

Actually the most efficient proposal was to change the convergence tolerance to the 1.0E-5. Also I changed the solution time interval to 0.005 sec. The erratic behaviour mostly recovered however steel there are some inconviniencies..

I intend to baseline correction the input data but, there is significant elimination in the residual displacement how ever the original data did not seriously changed.

The most popular filtration method is butterworth but What kind of filtering (lowpass/highpass) do you prefer in which frequency ?

Do you think the filtration data should be applied to input data or out put data ??

Regards
Hasan

[huffte]

I would not recommend filtering of any kind without significant efforts to research the instrumentation which recorded your input motion, Hasan. Baseline correction is fairly straightforward, so I would specify baseline correction with no filtering. The proper filter specification depends upon the instrumentation.

I would also examine the reinforcement details, particularly for the t-beams. This may be the actual reinforcement pattern in a constructed facility, but it does seem to be quite light to me.

Also examine your choice of ground motion records. If this is an actual facility, did you select and modify records appropriately for the site?

Best of luck, Hasan.

[hasanozkaynak]

Dear Timothy,

Thanks for your commends,

Actually the input data is the a artificial program which is adopted to a site specific code based elastic spectra by using the Oasis program from England.

Regarding the beam actually the flexural behaviour is not dominant because I performed equal SDOF which makes the frame deform in shear frame (single curvatured columns)

If you do not suggest filtering I only apply base line correction, at this point I doubt there is no significant change in the source of data

By the way could you explain a bit more about the instrumentation ?

Regards
Hasan
-----------------------

[huffte]

Hasan, I am certainly no expert on filtering of ground motion records, but the following web site may shed some light on the issues of baseline correction and filtering for you.

nsmp.wr.usgs.gov/processing.html

Further, as long as you are looking at element - level response, it is not uncommon for baseline correction to have little effect on results. I do believe it is good practice to use corrected accelerograms because the real differences can show up when you begin to look at displacements. Say you have 2 adjacent structures and want to find out if they collide during the earthquake. If you have a 2 meter drift in the record and the roof displacement shows up as 2.2 meters, you might conclude serious collision forces when, in reality, the real displacement was only 0.2 meters.

Enjoy what remains of the weekend, Hasan.

[hasanozkaynak]

Thanks for your comments.

I think this much of convinient loops are enough and now I will pass to the real case which is 6 storey RC building.

Thanks for your supports...

Hasan

[hasanozkaynak]

Hello again,

I have just send you an e-mail regarding new arrival questions related with 6 storey RC building pushover analysis.

Aprriciate if you mind replying it
Regards
Hasan

[huffte]

Hi Hasan. I wonder if you could e-mail the model (not the results, just the model file)? I'll try to take a look at it as I can find time. It is difficult to make any conclusions based solely on the error message.

[hasanozkaynak]

Hi Tim,

I am immediately sending it to you now
Just for now I am trying to perform push over analysis

Regards
Hasan

[hasanozkaynak]

Hello,

I have defined three pcs. beams or columns in the model with displacement based plastic hinge and the push over analysis were not succesful, later I changed the beams and columns to be only one psc. element and actually nothing changed. I have already send the model to seismosoft support team.

Would you kindly propose some thing to do ?

Regards
Hasan