PDA

View Full Version : crash sim software?


number77
10-07-2006, 10:27 PM
I was wondering if any of you have used some type of crash sim software.
number77
10-07-2006, 10:27 PM
I was wondering if any of you have used some type of crash sim software.
Jersey Tom
10-07-2006, 11:15 PM
.. for what? Modelling crush zone stuff is highly nonlinear plastic defomation. IMO physical testing is the way to go.
Nima
10-08-2006, 02:12 AM
If you access to the ANSYS, it is the most powerfull software for the crash testing.
Dave_Dal
10-08-2006, 05:16 AM
Although ANSYS is a very powerful FEA package, I don't think it's the best for crash testing. If you want to do crash testing I would suggest using LS-Dyna (http://www.lstc.com/), or something similar that is designed for contact/large deformation problems. It's less expensive (but harder to use) than ANSYS but has more robust contact algorithms. Also, if you plan on doing any crash testing you need to have a good material model, because as Jersey Tom stated, crash test simulation is a highly non-linear operation.
Timmay!
10-16-2006, 05:11 AM
Madymo

We use it for our crush zone...

Our advisor uses it with his ferro arm to analyze automotive crashes as well... Its very good crash testing software...

check out http://www.automotive.tno.nl/smartsite.dws?id=648
number77
10-21-2006, 09:40 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Dave_Dal:
Although ANSYS is a very powerful FEA package, I don't think it's the best for crash testing. If you want to do crash testing I would suggest using LS-Dyna (http://www.lstc.com/), or something similar that is designed for contact/large deformation problems. It's less expensive (but harder to use) than ANSYS but has more robust contact algorithms. Also, if you plan on doing any crash testing you need to have a good material model, because as Jersey Tom stated, crash test simulation is a highly non-linear operation. </div></BLOCKQUOTE>
for material model, are you suggesting the use of a full car?
Dave_Dal
10-22-2006, 07:38 AM
By material model I mean that you need to make sure to specify the material properties of what you are crashing properly for a non-linear analysis. If you assume constant material properties then your results will not be correct. There's no need to model the entire car, you just need to specify the appropriate boundary conditions and make the correct assumptions. If you are not familiar with non-linear FEA then physical testing is the appropriate way to go. After you do the testing you can try building a model that will replicate your test results. Good luck.