Seismosoft | Earthquake Engineering Solutions

I've to calibrate simplified bilinear Takeda curve for capturing the hysteretic behavior of my bridge pier (UCSD Blind Prediction Contest Full-Scale Bridge Pier). For this, i performed static pushover analysis of the bridge pier to obtain the capacity curve (Base shear~ top displacement).From Global Response Parameters tab, i got the capacity curve but i need to convert this curve to an idealized trilinear and later to bilinear curve. I think that the last branch of the trilinear curve (branch from yield point to end point) will be obtained simply by joining the yield point and the end point and the slope of this line is the post-yield stiffness. Is this a correct practice?
Please check my model
https://drive.google.com/folderview?id= ... sp=sharing
My querry may be very simple to most users but I'm new into the subject and software also.

adeelfaisal, your model seems to match the UCSD blind prediction contest column with the exception that:

1. In the experimental analysis for dynamic time history loading (SeismoStruct Verification Model Ch4-blind-01), the concrete tensile strength was set equal to zero whereas you have a value of 3221 kPa. This should not have a large impact on results.

2. You do not auto-compute the confinement factor for the core concrete, but specify a value of 1.2. Auto-calculation would result in a value of about 1.1.

I would say that your number of steps (440000) seems inordinately large and creates an enormous output file for such a very simple model. I think you could use a much smaller value for the number of steps in your pushover analysis.

The conversion of a load-displacement curve to a bi-linear model requires some engineering judgment. Generally, when such a procedure is done, the post-yield line 'adds' a bit of area beneath the load-displacement curve near the yield point and 'cuts off' a bit of area beneath the load-displacement curve near the ultimate point. It is desirable to 'add' the same area at the beginning as you 'cut off' at the end. hence, the required judgment as there may not be a single, unique line.

Best of luck adeelfaisal.

Thank you very much Huffte for considering my problem.

1. I set some value of tensile strength for obtaining the Cracking Point by monitoring the tensile strain through Performance Criteria Check.

2. In the Verification Report Model Ch4-blind-01, I found the confinement factor specified as 1.2 (and Not auto-computed). I think this was there in the Verification Model before i made any changes to it (assuming I have not mistakenly changed this). You may please confirm it by taking a look at the verification model you have.

After I'm sure about confinement factor value to use, I'll reduce the no. of steps and find the optimum no. of steps for reasonably accurate and fast results.

I think that if the starting point of the curve (origin) and the ultimate point are allowed to remain fixed and the yield point is shifted untill the areas added and cut-off are equal, the original load-deflection curve can be converted to a bilinear curve. I could find some more guidelines in the section 3.3.3.2.4 of FEMA 356. FEMA 356 can be obtained from
https://drive.google.com/file/d/0B1zFr_ ... sp=sharing