Thank you all for taking the time to help me. I am new to molecular dynamics and may make simple mistakes.
I am trying to exfoliate a 2D MXene sheet using water flow with the ReaxFF force field. After full NVT equilibration, I began the exfoliation step as shown:
I used fix spring/self 1.0 to restrain the 3D MXene ,fixed mxene and prevent them from sliding.
I applied a force to the water (group “flow”) to drive its flow.
I expected the flowing water to peel off the 2D sheet inside the white box in Figure 1.
However, after running (Figure 2), the 2D sheet in the white box did not exfoliate, and the bonds in the spring‑fixed 3D region (blue box) became distorted.
Please read the guidelines and fix your post by editing it. It is not readable.
Thats an incredibly difficult simulation to make, and if you are indeed new to MD I doubt that receiving advices from here will be enough. It would be better to discuss these issues with a local expert.
While @simongravelle is right that, overall, you are attempting a very difficult simulation that needs local advice and you should take more care to format your input post, here is some easy general advice: you are trying to accomplish too much in one go. It is impossible for you, or any of us, to know which exact component of your simulation has gone wrong.
So you should go back to basics and try to simulate:
3 MXene sheets in water with NVT (guess a reasonable density)
3 MXene sheets in water with NPT (check that barostatting does not destroy your system)
3 MXene sheets in water with NPT and fix spring/self on the MXene (check that the spring fix does not destroy your system)
3 MXene sheets in water with NPT and a very small flow on the water
and finally the simulation you have posted about.
You should also try the above steps on a smaller box for quick checking.
Please look up any similar studies done in previous papers and try to replicate them before trying anything new or fancy.
You should discover this after trying 1, but 20,000 timesteps – 20 picoseconds with default settings!! – is far too short for meaningful equilibration or production.
I haven’t found a suitable force field. I only found REAXFF. My ability is not sufficient to design or optimize the force field by myself.
thank you,I will increase the time for balance.
ReaxFF potentials are not very transferable and specifically not very good to model bulk water and they require excessive amounts to CPU time and a short timestep. So unless chemical reactions are at the core of your study, you should not use it. And if it has not been parameterized for your type of system, it will likely produce garbage.