Purpose of fixing few layers of atoms

LAMMPS LAMMPS Mailing List Mirror
1

Hi All,
In LAMMPS simulation of nanoindentation of metallic materials, we are fixing the bottom 3 or 4 layers of the substrate material and stating that it will resemble bulk property of the material. Is it right? Is there any literature that substantiate the proof?

2

Just run a test keeping those moving. You will see that those layers will expand and reconstruct like on any free surface and thus impact the atoms on the other surface. It is pretty self-evident that you need to preserve those atoms at bulk geometry.
Axel.

3

If you are interested in nanoindentation this is an excellent paper to become familiar with . It doesn’t specifically address your question because, like Axel said, if you run your simulation without an immobilized bottom layer you will see the problem rather quickly. To help with your specific question though, the thicker your model the less effect that immobilizing the bottom layer will have on surface measured properties.

For geometry and setup I will refer you to Figure 1 of that paper.

Hope these comments help.
Best,
Dylan

4

Thank you Axel and Dylan.
I have two more doubts -

  1. Is there a thumb rule stating how many layers needs to be fixed for the substrate material?

  2. Is the substrate thickness a major factor in hardness determination or the number of fixed layers? For example, if I have a soft material like Aluminium (Al) as the substrate and a hard material like Nickel (Ni) as coating, to determine the hardness of the nickel coating for varying coating thicknesses (4nm to 50nm) on Al substrate using nanoindentation, whether a 10nm thick substrate is fine for simulation with bottom 3 layers fixed or as in experimental study, the thickness of substrate should always be higher than the maximum coating thickness irrespective of number of layers fixed?

5

Thank you Axel and Dylan.
I have two more doubts -

  1. Is there a thumb rule stating how many layers needs to be fixed for the substrate material?

you need as many as adding one will change the inter-layer interactions above. the maximum is determined through the cutoff for non-bonded interactions. there is no value in going beyond that.

  1. Is the substrate thickness a major factor in hardness determination or the number of fixed layers?

again. you can determine this empirically by adding layers until the results are converged. for the immobile atoms, the limit is theoretically bounded by the interaction cutoff. please avoid the term “fixed” in the context of LAMMPS, as “fix” has a special meaning, and thus things can be easily confused. also immobilized is the more accurate term.

axel.