How to input acceleration? 2

[OscarPacheco]

Hello to all!
I am trying to simulate a frontal crash and see the reactions of this to the cervical spine.
I have some doubts about how to input the acceleration resulting of the crash in my model. At this moment I am trying two options but it seems they are not working somehow.

1- Input with value 4g as "Gravity" load in the direction of the crash. It seems that no motion is being obtained...
2- Input "Acceleration/rotation acceleration" 4g as a boundary condition.

Please find hereunder a image of my model; the acceleration in both cases is applied to the whole model except for the lower intervertebral disc (in grey). I do this because there is an encastre there and I dont want the model to move as a whole, but keep this disc fixed.

I am acting ok? Do you have any othe ideas? They are all welcomed!!

Thank you very much.
Regards,
Oscar

 

[Mustaine3]

Both should work, but a BC is something different than a load and would create a different results.

 

[rrg1016]

What kind of analysis are you doing? Static, dynamic?
To start i will do a static analysis. If you create the load "Gravity" , you should create the BC to restrict the movement of the spine.
The disc is another part? how is it conected to the spine?

 

[OscarPacheco]

Thnaks both for your response!
The analysis performed is dynamic.
I have done it with gravity load and with the boundary condition of acceleration (two diferent jobs) and obtained similar results in both of cases. At this moment I am trying the load "Body Force".
In all the cases I have applied the load to all the vertebrae and discs except for the intervertebral disc at the bottom that is only used as an encastre.
All the discs are connected to the vertebrae they are in touch with by a Tie Constrain.
A normal gravity load (-9,81) is also included.

@rrg1016 which BC do you think I should use to restrict the movement of the spine?

I submitted the job to a 0,250 seconds step (this is according to the literature the time of a car impact) Actually, at the end of the job, the spine moves a little bit, but not enough; the spine doesn't make the movement of hiperextension it is suposed to after an acceleration of this magnitude.

I am thinking it may be a problem with the properties of the materials... Any ideas?

Thank you for your help! It is very useful!

 

[rrg1016]

Body force have a different definition than gravity. I suppose that you account this difference.

I have a few possible causes for the limited displacement of the spine:
- Maybe you have a problem with the units (units of loads or materials).
- I don't know anything about spines, but maybe the tie that you used between a disc and a vertebra increases spine's rigidity out of normal values.
- If you know the rigidity of the spine you can check the tie interaction calculating the rigidity of your model.
- Also you can try to check the magnitude of the reaction forces and if it correspond to the load aceleration.

 

[IceBreakerSours]

What I would do for this problem, at first, is make all vertebrae rigid because it would speed up the analysis time and allow for iteration without a big hit on the results. I would not spend too much time running Static analyses - with this model - mainly because it is most definitely a dynamic problem with inertial forces playing a major role. Most spine modelers end up using Explicit for this kind of a problem but I suspect *Dynamic, Implicit should work well too.

Coming to applying accelerations, you have the option of assigning acceleration in the Boundary Condition manager. You will need to know the time history from literature and use *Amplitude with your boundary condition to provide a time-dependent acceleration to your cervical spine model (at multiple locations, perhaps?).

*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083

 

[OscarPacheco]

Thank you for your answers!
According to the literature this is the movement the cervical spine must have after a 7g acceleration impact:

As you can see the impact last around 250 ms.

I am using abaqus dynamic explicit.
Step time is 0.250
I am using gravity load of 98.1 m/s^2 and no amplitude. My reason for not using amplitude is because the acceleration in a colision is received all of a sudden, if I use amplitude, I would only get the 10Gs at the end of the job and that is not a real simulation dont you think?
The gravity load is applied to the whole model but the disc of the bottom.

I have also tried boundary condition of acceleration but the movement was very similar to the one got by the gravity load.

Any suggestions?

Thank you very much.

 

[IceBreakerSours]

You do not need to specify *Amplitude with gravity; there is no time-dependence for this body load! What you need is to specify displacement/velocity/acceleration profiles at one or more points in the cervical spine. Read works by Panjabi/Yang/King/Yoganandan/Hayashi for experimental evidence and finite element simulations.

*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083

 

[OscarPacheco]

Hi! Thank you!
I have tried some new things. I transformed all the vertebrae into discrete rigid bodies to reduce computational time. I assigned the mass and inertia and defined reference points for each vertebrae. However Abaqus says thta gravity loads are not applied to rigid bodies! ?????? so how can I do it?
In addition, in order to not allow parts to penetrate each other, in the previous model I used interaction surface to surface, but it seems this interaction is not working with rigid elements. I have tried general contact but it seems it is not working very good either.