Hysteretic Beahvior of Bridge Pier

[adeelfaisal]

I'm working on effect of earthquake duration on hysteretic behavior of bridge pier. I'm not doing any experimental work. I've to analyze the bridge pier under various time histories using SeismoStruct. I've to find stiffness degradation, strength degradation and pinching as the output parameters and relate them with duration and also other characteristics of time histories.
1. Kindly suggest me the hysteretic model I should use (I guess it should be smooth curve).
2. How to find the 22 parameters that need to be defined in smooth curve. Do I have to perform experiment to find those parameters or can they be calculated ?

[fakharifar.mostafa]

There are plenty of information regarding your question in the existing literature. You have two options to implement is SStruct.

You can either use the Takeda hysteresis behavior (thin or thick Takeda accordingly) to simulate the deteriorating behavior.

Or you could use the link elements associated with Sivaselvan and Reinhorn link elements (multilin or smooth link elements).

Both of the aforesaid link elements are included in SStruct. The calibrating parameters, SStruct help has recommended values, however, one should obtain their calibrating parameters according to their specific model. I recommend using adequately calibrated link elements in series, then you'll get the deteriorating behavior. Below paper gives you an idea on numerical values for Takeda model. Hope this helps.

Tehrani, P., & Mitchell, D. (2013). Seismic Risk Assessment of Four-Span Bridges in Montreal Designed Using the Canadian Bridge Design Code. Journal of Bridge Engineering.

[huffte]

Hi adeelfaisal.

In addition to selecting an appropriate hysteretic model, you will probably want to take care in selecting ground motions to use for your analysis. There is actually a lot to think about regarding this aspect.

Which duration do you use as criteria? Total, D5-75, D5-95, . . .?

Which magnitude ranges do you use?

Which site class criteria do you use?

What source-to-site distance ranges do you employ?

I am not asking you to answer these questions in the forum, I am simply relating to you that you should carefully consider questions such as these in selecting your time histories in order to make the study most meaningful.

Interesting project! Best of luck.

[adeelfaisal]

Thanks mostafa. Can you please send me the paper on adeelfaisal40@yahoo.com . I'm not able to download it.
quote:Originally posted by fakharifar.mostafa

There are plenty of information regarding your question in the existing literature. You have two options to implement is SStruct.

You can either use the Takeda hysteresis behavior (thin or thick Takeda accordingly) to simulate the deteriorating behavior.

Or you could use the link elements associated with Sivaselvan and Reinhorn link elements (multilin or smooth link elements).

Both of the aforesaid link elements are included in SStruct. The calibrating parameters, SStruct help has recommended values, however, one should obtain their calibrating parameters according to their specific model. I recommend using adequately calibrated link elements in series, then you'll get the deteriorating behavior. Below paper gives you an idea on numerical values for Takeda model. Hope this helps.

Tehrani, P., & Mitchell, D. (2013). Seismic Risk Assessment of Four-Span Bridges in Montreal Designed Using the Canadian Bridge Design Code. Journal of Bridge Engineering.

[adeelfaisal]

Thanks a lot huffte for drawing my attention towards important aspects i must pay attention to [:)] Can you please say something about hysteretic model i should use. I think the best is Sivaselvan & Reinhorn's Smooth Curve model but i'm confused regarding finding the 22 input parameters needed to define this curve. I've studied the first 02 sections of the related paper by Sivalselvan & Reinhorn. But i'm still confused about finding the input parameters. I was getting this feel that perhaps these input parameters specially the ones related to hysteretic behavior could not be found without experimental work, specially when i came to the equation 2.1 of the paper (eqn 2.1 gives equation regarding stiffness degradation parameter). I'm sending you the paper on your email.
quote:Originally posted by huffte

Hi adeelfaisal.

In addition to selecting an appropriate hysteretic model, you will probably want to take care in selecting ground motions to use for your analysis. There is actually a lot to think about regarding this aspect.

Which duration do you use as criteria? Total, D5-75, D5-95, . . .?

Which magnitude ranges do you use?

Which site class criteria do you use?

What source-to-site distance ranges do you employ?

I am not asking you to answer these questions in the forum, I am simply relating to you that you should carefully consider questions such as these in selecting your time histories in order to make the study most meaningful.

Interesting project! Best of luck.

[huffte]

Hi adeelfaisal. Thanks for the reference. I appreciate it.

I have replied with additional references on the smooth hysteretic model which may help you in defining the parameters.

My guess is that if you get the parameters right in the smooth hysteretic model you can successfully mimic the behavior of reinforced concrete bridge piers. Hopefully the additional references will help you get the proper parameters in SeismoStruct.

As noted by fakharifar.mostafa, you could also look at the Thin Takeda hysteretic model as it is the one recommended by Priestley, Calvi, and Kowalsky for bridges in "Displacement Based Seismic Design of Structures", IUSS Press, 2007.

Best of luck adeelfaisal.

[fakharifar.mostafa]

You can download the same authors' conference paper from the link below. That's almost the same paper using the modified Takeda hysteresis model (Otani, 1981).

Tehrani, P., & Mitchell, D. (2012). Seismic performance assessment of bridges in Montreal using incremental dynamic analysis. In Proc., 15th World Conf. on Earthquake Engineering.

http://www.iitk.ac.in/nicee/wcee/articl ... 2_3096.pdf

Based on my own experience using SStruct, to capture the deteriorating behavior, I recommend you using the Sivaselvan and Reinhorn link elements in SStruct. Although thin Takeda works just fine. Also try using FB elements rather than DB elements, as element discretization increases the chances of divergency.

Use zero length link element for shear deformation and bond-slip (if your reinforcement is continuous). SStruct link element has slip parameter as well. If you have lap-spliced column you need to calibrate the link elements to consider pinching hysteresis as well.

The Sivaselvan link elements in SStruct are great versatile tools for this use, and you can get highly accurate results. Good luck.

[adeelfaisal]

In "Simplified bilinear Takeda Curve", the unloading rules have been taken from Emori & Schonobrich [1978] "Analysis of reinforced concrete frame-wall structures for strong motion earthquakes". But this paper contains only the parameter "Outer Loop Stiffness Degradation Parameter" (called as a= "unloading stiffness constant" in the paper).I didn't find any discussion on "Inner Loop Stiffness Degradation Parameter". Can anyone please share any paper(s) in which this parameter has been discussed. Does "inner Loop Stiffness Degradation Parameter" mean "Reloading Stiffness Degradation Parameter" ?

[z.gronti]

The inner loop degradation factor refers to the unloading stiffness in inner hysteresis loops, i.e. for every other hysteresis loop apart from the largest one, which is the one with the largest deformation. Note, also, that the reloading stiffness is the same with the unloading one, the only difference is between the stiffness in the outer loop and in the inner ones.

[adeelfaisal]

Thank you very much gronti. I've got it