Time history analysis for bridges

[Rafamed]

Can anyone point me some instructions or links on how to perform a time history analysis for a vehicle crossing the bridge span.I need to Know which time functions are recommended for this case as well as the whole procedure.
It's a very special bridge.There's a 240m arch crossing over the bridge span from one side to de other, with a height over the road of 40m.From this arch 16 cables give intermediate support to the 180m span.
I'm using Robot millenium 14.2.

 

[Qshake]

Are you doing a live load analysis? If so you don't need a time history for that. Incidentally a time history is a measured response (funtion of time) to some excitation. For example, a time history might be the displacement due to the vehicle. But the vehicle should not be the focus of the time history.

Please tell us what you're looking for and we should be able to better answer your question.

 

[Ron]

Ditto on Q's comments. Also, I have done dynamic strain monitoring of live loads on bridges, so if you are looking for strain peak/decay info, that would be one place to get it. Most state DOT's have info like that.

One of the professors at Clemson University is studying bridge testing and rating, and he probably has tons of info. Unfortunately I cannot recall his name at the moment, but if you contact the Civil Engineering Department at Clemson, I'm sure they can put you with him.

 

[Rafamed]

Qshake,
I'm interested on knowing if the passing of the vehicle causes a significant increase on the cable stays tension forces,or in the opposite side ,if it can totally unload the cable.The static analysis results showed me that cables are ok.
What I've done so far is :
I created several load cases with a single unity concentraded load.For each load case this unity load is applied 4m ahead of the previous load case, simulating the passing of a moving load, through the whole span.In the time history menu, I used tringular time functions associated with each of these loads.Then I tell the program when each of these load cases will act.Depending on which vehicle speed I want to simulate, I set each the loads to be applied on its correct time window,one after the other.
Then , comparing the bars internal forces from the time history analysis with those found on the static analysis(of these load cases),I know if there's an amplification factor or not.
If there is an amplification ,then I apply the factor only to the internal forces due to moving load case.
Is this correct?

 

[Qshake]

The key here is the vehicular speed. As the forcing function, of a time dependent or dynamic system the speed is critical. Refer to general structural dynamics text. And I'm not sure that you want or need an analysis based on something (vehicle speed)that is not easily quantified and which would amount to statistical data at best.

I rather think you want and need the influence of the vehicle irrespective of speed. A moving load analysis. In other words where can I place the load to determine the most critical response to the element I am investigating.

How you set that up might be easy or difficult depending on how you've modelled the bridge or section thereof.

Based on your explanation I think this is what you're attempting but confusing it with the dynamic capabilities of the program. You should have three loads, simulating the three axles of the HS20 vehicle that are 14' apart. Move these three loads around to find the maximum response for the element you're looking for. Most programs have the capability to do this automatically for predetermined axles and or user defined axles.

 

[Rafamed]

I've already done the moving load analysis for various moving load cases, and statically the structure is solved.
I want now to analyse the dynamic bridge response to a moving load entering the span and crossing it at some speed.
What's the best method to do that and if what i've done is correct?

 

[Qshake]

You will need to obtain a time-history from some source. Check the Universities for more information. In particular check those universities currently doing bridge research, Iowa State University, Georgia Tech, University of Illinois, University of Maryland - College Park, University of Missouri at Columbia to name a few.

Also, I recommend that you check the websites of ASCE and FHWA. At the ASCE website you will be able to access the abstracts for the Structural Engineering and Bridge Engineering Journals. At the FHWA website you will be able to access the following libraries or databases: Transportation Research Board (TRB), National Highway Research Cooperative Program (NHCRP), and TRIS.

Once you have an actual time history you can read it into the program you are using or manually input the data. Please note that some commercial programs have serious limitations on the size of the time history that may be used. Run the analysis and evaluate the results. Keep in mind that the time history is very specific and should be combined with others to approximate the vehicle behavior. Any scaling of the time history should be checked in both the time and frequency domain for plausibility. In other words you can't just factor the data -- it is a mathematical representation and must make sense in either domain.

Good Luck!

 

[Rafamed]

Thanks for the tips Qshake!!!
Rafael

 

[butelja]

Mechanical ans Structural Vibrations by Demeter G. Fertis (ISBN 0-471-10600-3) outlines a procedure in chapter 7 using modal analysis for calculating the dynamic response of a structure to moving loads. If this procedure is used, you would need to analyze for many different moving load speeds to find the worst case. Of course this is applicable only to linear systems, since modal analysis is used. I hope this is helpful.

 

[Qshake]

Thanks BUTELJA, I will check that reference out for myself and see what, if any, comparisons I can make with the many structural dynamics texts that I have. With every discipline of engineering there is a little different emphasis on material presented.