Hi everyone,
I am using LAMMPS to simulate corrosion/interdiffusion at an Fe / liquid LBE (Pb-Bi eutectic) interface, and I am comparing the behavior under static LBE and flowing LBE conditions.
Model setup:
Lower part: solid Fe
Upper part: liquid LBE（Pb/Bi）
Interface normal direction: z
A vacuum layer is left above the liquid
Interatomic potential: EAM/alloy
Boundary condition: P P P
A few bottom Fe atoms are fixed, and the remaining atoms are defined as mobile
Under the flowing condition, I apply an x-direction body force (parallel to the interface) to the liquid LBE to drive the flow. In the production stage, I use a single mobile nvt ensemble, and temp/profile is used to remove the x-direction streaming bias:
compute Tflow mobile temp/profile 1 0 0 z 20
region LBE_flow block INF INF INF INF 58.86 124 units box
variable fx equal 5.0e-5
fix 3 LBE addforce v_fx 0.0 0.0 region LBE_flow
fix 4 mobile nvt temp 973 973 0.1
fix_modify 4 temp Tflow
Quantities analyzed
1、the number of Pb/Bi atoms penetrating into Fe
2、the maximum penetration depth of Pb/Bi into Fe
3、the number of Fe atoms dissolved into LBE
Compared with the static LBE case, under the flowing condition I obtained the opposite of what I expected:
1、the number of penetrating Pb/Bi atoms decreases
2、the maximum penetration depth becomes smaller
3、the number of dissolved Fe atoms also decreases
In other words, in my current simulation, the flowing condition does not accelerate corrosion/penetration. Instead, it seems to suppress both the penetration number and the penetration depth, which is opposite to the common expectation that flowing LBE should enhance corrosion compared with static LBE.
I would like to ask whether this result is physically reasonable for this kind of MD setup.or whether there may be some other issue.
Any comments or suggestions would be greatly appreciated. Thank you
Wang zhao le