is there any acceleration in your implicit analysis? At time 0 of the explicit analysis, it just shows the states of your previous/imported model. If you don't need those stress/strain from previous analysis, you can change the parameters, update and state, in the *import command.

Hello,

Thank you for your reply.

Since my implicit calculation is static, there are no accelerations. But after transferring the results to Abaqus/Explicit, the accelerations are very large at t=0.  I did need the stress/strain from the previous analysis. It seemed to me that the nodal forces due to the stress are not zero at the beginning of explicit analysis, as we know, they have to be none in implicit analysis. However, I have kept the same material properties and the two analysis are geometrically nonliear. I can't find the reason for the acceleration sources.  

Best regards

what your loading conditions, disp, force  or temperture?  Could you please show the analysis step in .inp file format?

In my implicit analysis, the load is applied by contact conditions as I perform a 4 point bending simulation on a plate. I m sure that there is no acceleration since it is prescribed by the static formulation. 

what's the loading or boundary condition in explicit analysis?

It's better that you can upload those .inp files so that I can check it on my computer

I think those initial accelerations are due initial displacements you imported from the static analysis.

Numerical dynamic analysis methods need to calculate initial disp, velocity, and acc so they coud perform the dyn analysis.

These init values are calculated from the equilibrium equation of motion (beyond the specifity of each method).

Mu..+Cu.+Ku=p(t)

at t=0 let us simplify C=0, and P(0)=0

you left with Mu..+Ku=0

if u is different from 0 =>  u.. must be different from 0.

Hello Jue,

I'm sorry, the odb file is very heavy, about 30GB, it's impossible to pass to you by internet. In my explicit analysis, the loading (i.e. the contact condition) is the same with the implicit analysis.

Hello, Djamel

Thanks a lot for your reply. But I don't agree with you entirely. Since in the static analysis, we have Ku=p, and at some node positions where there is no external loading, the equation is Ku=0, this refers to the fact that the nodal forces induced by the stress are none at these nodes. So at the beginning of explicit analysis, we have Mu..+Ku=0, knowing that Ku=0, u..=0, where the nodes libre of external forces.

I don't know if my argument is correct or not. Hoping for more discussions. 

As I understand you have performed a static analysis. so you get displacement at the end of analysis. This analysis and its results (displacement) are imported to an explicit dyn analysis. So, at the begining of explicit analysis you have an initial displacement that is different from 0. As a consequence initial acceleration wil be different from 0.

To chek, try to change u and control if u is changed linearly. Or try to add damping...

Hi Zhao,

You don't need to send me .odb files. If you don't use .inp inputs, you can send me or upload that file with .pes extension. '.pes' file contains all the info for the calculation, so I can run it on my PC