Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Shear-stress relaxation in free-standing polymer films

G. George, I. Kriuchevskyi, H. Meyer, J. Baschnagel, and J. P. Wittmer
Phys. Rev. E 98, 062502 – Published 4 December 2018

Abstract

Using molecular dynamics simulation of a polymer glass model we investigate free-standing polymer films focusing on the in-plane shear modulus μ, defined by means of the stress-fluctuation formula, as a function of temperature T, film thickness H (tuned by means of the lateral box size L), and sampling time Δt. Various observables are seen to vary linearly with 1/H, demonstrating thus the (to leading order) linear superposition of bulk and surface properties. Confirming the time-translational invariance of our systems, μ(Δt) is shown to be numerically equivalent to a second integral over the shear-stress relaxation modulus G(t). It is thus a natural smoothing function statistically better behaved as G(t). As shown from the standard deviations δμ and δG, this is especially important for large times and for temperatures around the glass transition. μ and G are found to decrease continuously with T and a jump-singularity is not observed. Using the Einstein-Helfand relation for μ(Δt) and the successful time-temperature superposition scaling of μ(Δt) and G(t), the shear viscosity η(T) can be estimated for a broad range of temperatures.

  • Figure
  • Figure
  • Figure
  • Figure
  • Figure
  • Figure
  • Figure
7 More
  • Received 7 September 2018
  • Revised 25 October 2018

DOI:https://doi.org/10.1103/PhysRevE.98.062502

©2018 American Physical Society

Physics Subject Headings (PhySH)

Polymers & Soft MatterCondensed Matter, Materials & Applied Physics

Authors & Affiliations

G. George1, I. Kriuchevskyi2, H. Meyer1, J. Baschnagel1, and J. P. Wittmer1,*

  • 1Institut Charles Sadron, Université de Strasbourg & CNRS, 23 rue du Loess, 67034 Strasbourg Cedex, France
  • 2LAMCOS, INSA, 27 avenue Jean Capelle, 69621 Villeurbanne Cedex, France

  • *joachim.wittmer@ics-cnrs.unistra.fr

Article Text (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand
Issue

Vol. 98, Iss. 6 — December 2018

Reuse & Permissions
Access Options
Author publication services for translation and copyediting assistance advertisement

Authorization Required


×

Images

  • Figure 1
    Figure 1

    Some notations: (a) Simple shear with γ being the strain increment imposed at t=0 and σ(t) the measured shear stress increment as a function of time t. (b) Shear-stress relaxation modulus G(t) (dash-dotted line) and generalized shear modulus μ(t)=μAμF(t) (thin solid line). The affine shear modulus μA=G(t=0)=μ(t=0) is indicated by the dash-dotted line, the thermodynamic long-time limit μeq for G(t) and μ(t) by the bold dashed line.

    Reuse & Permissions
  • Figure 2
    Figure 2

    We study free-standing polymer films with M=768 chains of length N=16 monomers confined in periodic boxes with L being the imposed lateral box size in both x and y directions. The film thickness H1/L2 (to leading order) is operationally defined using the Gibbs dividing surface.

    Reuse & Permissions
  • Figure 3
    Figure 3

    Shear modulus μ as calculated by means of Eq. (1). μ(T) decays continuously in all cases considered. Main panel: Data obtained at a sampling time Δt=104 for three-dimensional bulks (stars) and films of different lateral box lengths L. Inset: μ(T) for film 1 comparing different Δt.

    Reuse & Permissions
  • Figure 4
    Figure 4

    Film thickness and glass transition temperature. Top inset: Number density profile ρ(z) for T=0.5 with z=0 corresponding to the center of mass of each film. The midplane density ρ01 is indicated by the dashed horizontal line. Main panel: H as a function of temperature T for film 1. The glass transition temperature Tg and the film thickness Hg at the transition (bold dashed lines) are operationally defined by the intercept of the linear extrapolations of the glass (dashed line) and liquid (solid line) limits. Left inset: Tg as a function of 1/Hg confirming the linear superposition, Eq. (3).

    Reuse & Permissions
  • Figure 5
    Figure 5

    Comparison of the different contributions to the shear modulus μ=μAμF=(μAμ0)+μ1 as functions of T focusing on data obtained for film 1 and Δt=104. Inset: Double-logarithmic representation of μ0/μA1 vs. T.

    Reuse & Permissions
  • Figure 6
    Figure 6

    Affine shear modulus μA. Main panel: μA(T) for all systems studied. Inset: As shown for T=0.5, μA decreases linearly with 1/H in the liquid limit.

    Reuse & Permissions
  • Figure 7
    Figure 7

    Shear-stress fluctuation μF for Δt=104. Main panel: μF(T) for all systems. Right inset: μF decreases linearly with 1/H in the liquid limit (T=0.5). Left inset: μF increases linearly with 1/H in the solid limit (T=0.1).

    Reuse & Permissions
  • Figure 8
    Figure 8

    Film thickness dependence of μ(T) for Δt=104. Main panel: Scaling collapse of y=μ(T)/μg vs x=T/Tg. Inset: μgμ(Tg) and μpμ(T=0.1) vs. the inverse film thickness 1/H of the respective temperature. As emphasized by the bold line, both shear moduli are consistent with Eq. (3). We shall use μp in Secs. 3e and 3f for the TTS scaling of μ(Δt) and G(t) comparing different ensembles.

    Reuse & Permissions
  • Figure 9
    Figure 9

    Sampling time effects for μ and its contributions focusing on film 1 and T=0.3. Only the simple averages μA and μ0 are strictly Δt-independent. μ1 and (hence) μ decrease monotonically. The solid and dashes lines have been obtained using Eq. (2).

    Reuse & Permissions
  • Figure 10
    Figure 10

    Double-logarithmic representation of μ(Δt) for a broad range of temperatures T focusing on film 1. μ(Δt;T) decreases continuously with both Δt and T. A pseudoelastic plateau is observed in the solid limit with μμp15.5 (horizontal dashed line). The 1/Δt-decay in the liquid limit (bold solid line) is expected from the Einstein-Helfand relation, Eq. (9). Inset: Shear viscosity η(1/H) for T=0.55. The values are used in Sec. 3e to define an absolute scale for τ(T). The line presents a linear fit according to Eq. (3).

    Reuse & Permissions
  • Figure 11
    Figure 11

    TTS scaling for y=μ(Δt)/μp as a function of x=Δt/τ(T) with μp being the plateau modulus defined in Sec. 3d and τ(T) the relaxation time indicated in the inset. We impose τ(T=0.55) according to Eq. (10) to have an absolute timescale. The two asymptotics of the scaling function y=f(x) for x1 and x1 are indicated by dashed and solid lines. Note the broad crossover regime between these limits. Inset: Data collapse of terminal relaxation time τ vs. x=Tg/T for all our ensembles. Arrhenius behavior (bold solid line) is observed around the glass transition (x1).

    Reuse & Permissions
  • Figure 12
    Figure 12

    Unscaled stress relaxation modulus G(t) for film 1 using half-logarithmic coordinates. No indication of a jump singularity with respect to temperature is found.

    Reuse & Permissions
  • Figure 13
    Figure 13

    Successful TTS scaling plot of y=G(t)/μp as a function of reduced time x=t/τ using the same relaxation times as in Fig. 11. The two indicated power laws (bold and dash-dotted lines) are given for comparison. Unfortunately, our production runs are too short to reveal the expected final exponential cutoff even for the highest temperatures.

    Reuse & Permissions
  • Figure 14
    Figure 14

    Shear modulus μ, shear relaxation modulus G and the corresponding standard deviations δμ and δG taken at t=Δt=Δtmax=105 as functions of T. Focusing on film 1 all data are averaged over m=120 configurations without additional gliding averages and logarithmic binning. The observed two inequalities Gμ and δGδμ are both consequences of the stationarity relation Eq. (2).

    Reuse & Permissions
×

Sign up to receive regular email alerts from Physical Review E

Log In

Cancel
×

Search


Article Lookup

Paste a citation or DOI

Enter a citation
×