Richard E. Robertson

The aging of glassy polymers as determined by scanning calorimeter measurements

A model successfully applied to explain the volumetric relaxation of amorphous polymers near the glass transition has been used to predict the behavior in a scanning calorimeter of amorphous polymers annealed or aged below their glass transition. Although the predictions are incomplete because of the uncertainty of several critical parameters, the calculations indicate that caution is needed in interpreting scanning calorimeter data. Possible problems interpreting such data include (1) underestimation of the conformational heat capacity, (2) incorrect comparison of specimens aged at different temperatures, and (3) uncertain energy of the quenched state, against which aging is often compared.

Fracture in epoxy matrix resins

Abstract The occurrence of fracture-energy-enhancing steps and welts on fracture surfaces of crosslinked matrix resins has been studied in an epoxy obtained from a trifunctional epoxy resin cured with an anhydride. It is suggested that the steps and welts arise from an underlying basic longitudinal texture, which was revealed by strongly tilting fracture specimens toward the collector in a scanning electron microscope. A model for the development of the basic longitudinal texture is proposed involving a meniscus instability of the propagating crack front, which gives rise to a series of fingers protruding into the bulk resin ahead of the nominal crack front. The periodicity of the basic longitudinal texture seen in the epoxy specimens studied was roughly 350 nm, which was independent of the epoxy resin: hardener ratio within at least 10% of stoichiometry. Because the periodicity of the basic longitudinal texture is roughly equal to the separation of the fracture surfaces immediately behind the crack, a considerable blunting of the crack by plastic deformation or yielding is suggested, a property that should depend on the matrix resin.

The stacked lamellar texture on the fracture surfaces of fibre composites

A fracture surface texture, which has been variously termed as “lacerations”, “hackles” or “serrations”, is often observed on the matrix surface of fibre composites, most often in resin-rich regions. This texture, referred to here as a “stacked lamellar texture” to emphasize its plate-like nature, was studied in an E-glass/epoxy composite. Scanning electron fractographs of these materials suggest that the stacked lamellar texture arises from crack fingers due to a meniscus instability mechanism interacting with a reorienting stress field.

A MOLECULAR MODEL FOR RELAXATION IN THE GLASS

A useful approach for studying the nature of glass has been the temperature step experiment. The experiment is usually performed near the glass transition, to bring the kinetics into a convenient time-scale for the experimenter, and consists of one or more rapid temperature changes, or a temperature change at a constant rate, during and following which quantities like refractive index, volume, enthalpy, or density fluctuations are measured. One way of presenting the results is to plot the relative rate of approach to equilibrium versus the distance from equilibrium. (Equilibrium is defined by extrapolating the curve for the quantity of interest versus temperature from the liquid region down into the glassy region.) A standard example is that derived from the volumetric relaxation of poly(viny1 acetate) following a single rapid temperature change obtained by A. J . Kovacs.’ This is given in FIGURE 1 . The abscissa is the deviation of the volume from equilibrium, and the ordinate is the effective retardation time 7 (the reciprocal of the rate) derived from the original volume V versus time I data:

On The Strength of Adhesive Bonds Between Rigid, Non-Crystalline Polymers

Abstract The relatively high strengths exhibited by well-prepared, rigid adhesive bonds partly reflect the crack-growth inhibiting properties of the surrounding bulk. Dispersion forces alone are unable to inhibit crack growth significantly, as is indicated, for example, by the low strengths of low molecular weight glassy polymers. The source of crack-growth inhibition of adhesive bonds was revealed by examining crack fronts with a microscope. Examined were crack fronts along the self-bond between pieces of poly(methyl methacrylate) and along the adhesive bond between pieces of poly(methyl methacrylate) and polystyrene. Associated with each crack front were two sets of interference fringes, indicating the presence of a craze preceding the crack. Crazes form in high molecular weight brittle polymers, and their presence along the adhesive bond ahead of the crack indicates the involvement of the high molecular weight bulk polymer adjacent to the bonding plane. Crazes ahead of cracks are known to inhibit fract...

The Strength of an Adhesive Weak Boundary Layer

Abstract The strength of model adhesive joints composed of different thicknesses of low (10,300 narrow distribution) molecular weight polystyrene sandwiched between high molecular weight poly(methyl methacrylate) has been studied. The joints model a polymer-to-polymer adhesive bond across a low-strength boundary layer. As an appraisal of strength, the fracture toughness was measured by driving a cleavage wedge into the specimens along the polystyrene layer and analyzing the results with Kanninens equation. The fracture toughness for both crack growth initiation K Ic and arrest K Ia was essentially that for bulk poly(methyl methacrylate) until the polystyrene layer exceeded 1 μm, at which K Ic and K Ia fell by approximately 85%. Examination of the fracture surfaces of specimens with polystyrene layers less than 1 μm thick revealed that the fracture path was predominantly in the poly(methyl methacrylate). Possible reasons for this are discussed.

Local motion, free volume fluctuations, and the thermodynamic recovery of polymeric glasses

Abstract The now 20-year old experiments of and Kovacs on the volumetric recovery of poly(viny1 acetate) near its glass transition brought to light several unexpected results: although the kinetics of recovery of the equilibrium volume following a downward step in temperature seemed describable by a free volume model, the recovery kinetics following an upward step in temperature clearly was not a function of free volume alone. In other experiments involving a pair of temperature steps, in which the pair of steps and the time interval between was programmed to bring the glass to an equilibrium volume at the instant the second temperature step was made, the volume following the second step drifted away from this equilibrium value and then returned to it later. Although Kovacs, Aklunis, and Hutchinson have shown recently that these early results could be described by a linear irreversible thermodynamic theory with a set of 33 ordering parameters, we believe the molecular origin of these results is fairly sim...