Analysis of the configurational temperature of polymeric liquids under shear and elongational flows using nonequilibrium molecular dynamics and Monte Carlo simulations

Abstract

Rheological and structural studies of liquid decane, hexadecane, and tetracosane under planar elongational flow using nonequilibrium molecular-dynamics simulations J. Chem. Phys. 122, 184906 (2005) We present a detailed analysis of the configurational temperature ͑T conf ͒ for its application to polymeric materials using nonequilibrium molecular dynamics ͑NEMD͒ and nonequilibrium Monte Carlo ͑NEMC͒ methods. Simulations were performed of linear polyethylene liquid C 78 H 158 undergoing shear and elongational flows. At equilibrium, T conf is equal to the set point temperature of the simulation. An aphysically large decrease in T conf is observed in the NEMD simulations for both flows, especially at strong flow fields. By analyzing separately the individual contributions of the different potential interaction modes to the configurational temperature, it is found that the bonded modes ͑which constitutes almost 99.5% of the total͒ dominate the total T conf over the nonbonded ones; i.e., bond-stretching ͑Ϸ86.5%͒, bond-bending ͑Ϸ11.8%͒, bond-torsional ͑Ϸ1.2%͒, nonbonded intermolecular ͑Ϸ0.4%͒, and intramolecular ͑Ϸ0.1%͒ Lennard-Jones. The configurational temperature of the individual modes generally exhibits a nonmonotonic behavior with the flow strength and a dramatic change beyond a critical value of flow strength; this is mainly attributed to the dynamical effect of strong molecular collisions occurring at strong flow fields. In contrast, no such behavior is observed in the NEMC simulations where such dynamical effects are absent. Based on the principal physical concept of the configurational temperature, which represents the large-scale structural characteristics of the system, we propose to exclude the dynamical effects exhibited by the individual interaction modes, in obtaining a physically meaningful T conf as the configurational entropy of the system should not be affected by such factors. Since ͑a͒ the main difference between equilibrium and nonequilibrium states lies in the change in the overall ͑global͒ structure ͑represented by the bond torsional and nonbonded modes͒, and ͑b͒ the local, very short structure ͑represented by the bond-stretching and bond-bending modes͒ is barely changing between equilibrium and nonequilibrium states and its contribution to the total system configurational entropy is negligible compared to the large-scale structural changes, in order to accurately describe the structural changes occurring at nonequilibrium states by use of the configurational temperature, we further propose that only the contributions from the bond-torsional and nonbonded modes to ⌬T conf between equilibrium and nonequilibrium states should be taken into account to generate a physically meaningful ⌬T conf . Applying the above hypothesis to the analysis of the simulation data, good agreement between the NEMD and NEMC simulations ͑and between NEMD simulations for different flows͒ is observed. Furthermore, the configurational temperature obtained in such way is found to match remarkably well with the heat capacity...

Challenges in Multiscale Modeling of Polymer Dynamics

“…In addition, the flow curves for η, Ψ 1 and the second normal stress difference, Ψ 2 = σ yy − σ zz , have been obtained and compare well with the NEMD results [299]. Baig and his co-workers [318,319] adopted a similar concept to study the viscoelasticity of polymer melts. They used an expanded Monte Carlo method as the macroscale solver for a family of viscoelastic models, which are built on the "structural" variable, x.…”

Atomistic simulation of crystallization of a polyethylene melt in steady uniaxial extension

“…The configurational temperature expression has been further extended for physical systems involving hard-core or discontinuous potentials [56]. Very recently, a comprehensive analysis of the fundamental aspects of the configurational temperature for nonequilibrium systems has been performed through extensive NEMD and NEMC simulations [57], suggesting that in order to define a physically meaningful configurational temperature for polymeric systems under nonequilibrium conditions, the dynamical effects created by the external fields should be excluded.…”

“…However, a more rigorous argument may be formulated in terms of a generalized ensemble. Following the procedure suggested by Ilg [58], the expression of the configurational temperature can be further generalized by applying a generalized canonical ensemble [57] such as the one adopted here-see Eqs.…”

Non-equilibrium molecular dynamics simulations of the thermal transport properties of Lennard-Jones fluids using configurational temperatures

“…There is no reason a priori that any one temperature should be a better choice than others; in fact, Ayton et al [35] established that in systems experiencing spatially varying strain rates, only Rugh's dynamical temperature is able to correctly account for the resulting heat fluxes. Several articles have reported configurational temperatures in non-equilibrium systems under shear where the temperature can feature spatial anisotropy, and di↵erences between the temperatures obtained from the kinetic and configurational routes have been reported in models of polymer fluids [36]. In a recent investigation of argon confined in zeolite structures [37], it was found that the kinetic and configurational approaches can yield significantly di↵erent results.…”