Strand 7 help Modelling a vehicle impact with a bridge

[cbat]

Hi guys,

My current project involves modelling an excavator arm impacting a series of cross-beams on a steel bridge.
I have little to no experience in Strand 7 and am looking for a few pointers on how to go about the model! If anyone has some tips to direct me along that would be fantastic.
So far I have modelled a single I-beam, and am currently trying to determine how to model the impact of the vehicle colliding through the lower 3rd of the beam.

The I-beam was modelled using beam elements to represent the cross-section, then extruded as plates and fixed at each end.

Should I be using contact elements and a load factor table to represent the impact over a short time? I am looking to find the permanent dispacement and nodal reactions caused by the colliding vehicle, and then translate those through the rest of the bridge structure.

Cheers,
Cam

 

[IDS]

How much time have you got?

To do a detailed impact analysis as you descrbe is a major undertaking, even for someone familiar with dynamic FEA, which I'm guessing you are not.

Practical options are:

[ol 1]
[li]Apply an equivalent static load specified in a design code or from research on impacts under similar conditions.[/li]
[li]Do a "pushover" type analysis, looking at the deflection required to absorb the energy of the load.[/li]
[/ol]

This really isn't a software related question. I think you need to do some research on the specific impact type and decide on a design and analysis approach, then look at how it might be implemented in a specific software package.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[cbat]

Hi Doug, thanks for the reply.

Correct you are, I'm not familiar with FEA.

I have about 4 weeks to finalise my model. Thanks for the suggestions above, I did a bit of research into both and it will probably suit me to run with option 1).
I will use the equivalent static load, as I have a report from the incident detailing equivalent loads determined by an external commission.

I think my workflow here is to apply the load over an area, with a time vs load factor graph, in the nonlinear transient dynamic load, using an appropriate number of time steps?
What is a simple method to spread the impacting load over an area of say, 50cm on the lower flange and lower section of the web? Create a contact element from beams, or create some springs that apply the load in compression over that area?

Sorry for the complete lack of understanding on my behalf. I appreciate your time and knowledge on this one!

Cheers,
Cam

 

[GregLocock]

If you aren't familiar with FEA why are you using it on this job? Crash type work uses a completely different solver to linear work. If this is a court case then then the opposing expert will have an absolute field day with an FEA sim prepared by an inexperienced user with zero correlation data run in the wrong environment.

Sorry about that, but there is more to FEA than producing pretty plots for managers to look at.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[IDS]

If this is for a forensic investigation I agree with Greg, you should advise your managers that this sort of investigation needs an expert.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[cbat]

Thanks for your concern, if it was a case of any real lift relevance I would agree with you. Fortunately though, it's actually just a research project with no definative impact on anyone. Hence me out of my depth and looking for a few basic pointers.
If your willing to help that would be great! However I understand that there are real world problems that deserve more attention..

cheers!

 

[GregLocock]

Phew, that's better. OK take a look at LS Dyna which is a one stop shop for crash type work (many of the other FEA suites will handle it but it is the only one I know by name). I'm sure they have case studies that will be very educational.

One approach that might work with Strand7 is to use it to ascertain the force vs deflection curve up to full plasticity at the point of impact. You can then use a spreadsheet to model the behaviour of a rigid mass as it strikes this non linear spring.

That is more or less how we used to do crash before non linear FEA was available and accurate, that is build a series of non linear springs and then cascade them.




Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?