Accumalated strain under consecutive accelerogram

[fakharifar.mostafa]

Hello,

For the sake of clarity, in "Dynamic Time History Analysis" if a series of consecutive scaled increasing ground motions are input in one analysis, are the "Output Stresses and Strains" in the "Analysis Output" tab the ACCUMULATED STRAINS of all the input motions?

In other words, if input motions are adequately separated in time, would the Strains extracted for different "Sectional Point Coordinates" be ACCUMULATED value of strain at the end of all motions? I know SStruct does this type of analysis, but was not sure about Strains associated with Sectional Point Coordinates. Thanks.

[ruipinho]

fakharifar.mostafa,

Yes, of course, the results from a single nonlinear dynamic analysis cannot be arbitrarily reset to zero midway through the analysis; that would make no sense.

Hence, you always obtain cumulative strain values from a single nonlinear dynamic analysis.

Rui

[TUC]

When you say "...series of consecutive scaled increasing ground motions are input I n one analysis..." you are talking about IDA, aren't you?

[fakharifar.mostafa]

Thanks Rui, I wanted to confirm this.

TUC thanks for the comment to make this question more clear. By "a series of consecutive scaled increasing ground motions are input in one analysis" I meant one input motion including two to three dynamic excitations which are shifted apart in time, not like the IDA. Thanks again.

[fakharifar.mostafa]

As a follow up post for this topic, if a repair scenario (like steel jacketing) on an earthquake damaged bridge column under dynamic time history analysis is required for after shocks:

Is using the "Activation / Deactivation time" in the element classes tab an appropriate tool?

I modeled two column cross sections for the plastic hinge:
1- the intact as-built column cross section
2- the steel jacketed column cross section (with repair grout concrete properties and degraded reinforcing steel)

And synthesized one ground motion including the MAIN ground motion and a LOW amplitude after shock adequately spaced in time. Accordingly, used the appropriate activation and deactivation time for each of the column cross sections for before / after repair under main and after shock, respectively.

However, this approach seems not to be applicable (can't assign different elements simultaneously and get the cross section updated at a specific time step). If this is not the right approach, then how can I import the accumulated strain/damage level from an existing analysis to an updated model?

When I change the cross section the model resets to the virgin model and analysis starts from scratch. Thanks for the potential advice.

[huffte]

Hi fakharifar.mostafa.

I believe that your approach will work.

1. Create an element class with a section using the original properties and a deactivation time equal to somewhere in the 'quiet' range between the main shock and the aftershock.

2. Create a second element class with the section using the revised properties and an activation time equal to the deactivation time in part 1 above.

3. The original member and the retrofitted member will be coincident in the geometry of the model.

It is an interesting problem. Your premise seems to be that the retrofit would be applied to damaged members quickly after the main shock, with no other remedial measures taken. So it seems that you are working on a way to protect structures from collapse during aftershocks rather than a permanent retrofit?

Best of luck fakharifar.mostafa.

[fakharifar.mostafa]

Thanks for the concise answer as usual.

You're correct, indeed I am working on an emergency repair scenario where the damaged column is just temporarily repaired with jacketing and would like to see the repaired behavior under the after shocks.

I am still using this approach and hopefully could solve the numerous divergence errors so far (might be real collapse of the bridge).