Stress output of a polymer melt subjected to oscillatory shear

Dear LAMMPS users,

I am subjecting a polymer melt system (Kremer Grest model) to oscillatory shear using fix deform wiggle along with NVT/SLLOD. I am attaching a small portion of the script here for reference.

dt 0.001
fix f1 all nvt/sllod temp 1.0 1.0 0.1
fix 2 all deform 1 xy wiggle 0.1 6 remap v
compute press all pressure f1_temp
variable strain equal xy/lx
variable xy equal xy
fix f3 all ave/time 1 1 1 c_press[4] v_strain v_xy file response.txt start 360000

I further wish to evaluate G’ and G" from the well known relationship: (stress_xy) = G’sin(omegat) + G"cos(omegat). To do so, I compute pressure first, then print out the c_press[4]. Further, I average over many cycles (around 5000) and then curve fit the final averaged cycle to obtain G’ and G". I am attaching the averaged response cycle for your reference.

response_output

My issue is, when I curve fit the above equation, the G’ and G" both are turning out to be negative values. I am not sure where the error is. Also, as seen in the above diagram, the strain is lagging the stress (shouldn’t it be the other way around, since strain is the applied/forcing variable?) I am attaching the averaged response for your reference too.

response.txt.averaged (139.6 KB)

Awaiting suggestions, thank you.

Hi @Sameer_Kalghatgi,

This is not so much a question about LAMMPS usage, but more about polymer simulation and results analysis. It might be better fitted in the Science talk section.

That said please see my comments on nvt/sllod in this answer. I do not think they are suited for constant temperature oscillating shear given the equations steady-state assumptions.

Concerning the fitting of your values, there can be several sources of error in your script and it is very hard to give proper insight without following your full procedure. A direct colleague, advisor or friend might give your better insights seeing what you have actually done. However, from the few elements you provide, you are trying to fit noisy data to your equation and I find it weird that you have non-zero average stress at zero initial strain.

Concerning the lag, polymer melts have several regime below their T_g and behave as complex molecular fluids. Their mechanical response to strain might not be immediate depending on the frequency, mechanical properties and temperature of your system. This is a question you should sort by digging the literature.

Hi Germain,

Thank you for your reply. I will recheck my script for possible errors.

Regards
Sameer