The 180° peel strength of four ethylene vinyl acetate copolymers with varying V.A. content was measured (Pol A: 28%, Pol B: 18%, Pol C: 12% and Pol D: 9.5% by weight). The polymers were applied as hot-melts with or without an inextensible fabric support on: (a) mild steel etched in hydrochloric acid; (b) chemically polished copper and; (c) oxidised copper with a fibrous topography. The peel loads for a given polymer decreased from oxidised copper to etched steel to polished copper. For a given substrate the peel loads generally decreased from polymer A to polymer D with a peak for polymer C. This trend was correlated to similar trends obtained in mechanical tests like the tearing energy, strain energy density and tensile strength to failure. The fractured parts of the bond were examined with a scanning electron microscope. There was microscopic evidence of an analogy between the peel load and the observed extent of polymer deformation in the fractured surfaces where the ductilty of the particular polymer and the topography of the substrate were clearly depicted. Contact angle measurement showed a cohesive type of failure in all cases. X-ray photo-electron spectroscopy and multiple internal reflection infra-red spectroscopy provided evidence of polymer oxidation by the substrate, mostly significant in the case of etched steel. The energy balance approach was employed in order to analyse the peel test results further. The polymers response to the imposed mechanical deformations during peeling was shown to control the measured peel load. For the unbacked samples the major energy loss mechanism is the stretching of the freed strip. The much higher peel loads of the backed samples were attributed to energy losses around the peel front. Finally, plastic bending of the polymer was found to account for a relatively smaller part of the total input energy for both backed and unbacked samples.

No description available.

No description available.

No description available.

The development of matrix cracking in (0/8/0) glass fibre reinforced epoxy laminates, where 8 = 45°,54 0, 75° and 90°, has been investigated under quasistatic and cyclic loading. All laminates were fabricated using a wet lay-up process to impregnate the fibres before curing. Uniaxial quasi-static tension tests were carried out on samples of each laminate type. Crack initiation and crack propagation in the 8° ply were studied separately using unnotched samples (tested with both as-cut and polished edges) and notched samples, in which the notch introduced is a drilled hole parallel to the fibres in the 8° ply. The ply stresses at which cracking events occurred were calculated from laminate theory and these stresses are discussed for all four laminate types along with investigations of notch depth effects and crack growth patterns. Material non-linearity was also considered and a method presented to account for material non-linearity in shear. Ply stresses were recalculated to include the effects of non-linearity and the presence of thermal stresses. It was shown that crack propagation is controlled principally by the stress transverse to the fibres and that the stress was independent of notch length. The stress state at crack initiation did not satisfy any of the conventional failure criteria. Uniaxial tension-tension fatigue tests were carried out on samples of each laminate type. Samples were cycled to various maximum load levels and crack propagation from a single notch in each sample was monitored as it grew full width to provide crack growth rate data. Crack growth data were presented on Paris-type plots. The crack growth data were compared to experimental data obtained by other authors for 90° cracking in cross-ply samples and 90° and +45° cracking in quasi-isotropic samples. There was good agreement between the sets of data after accounting for differences in transverse ply thickness. A series of compliance experiments were carried out to obtain the change in compliance with crack length for samples of each laminate type. The experimental compliance changes in (0/90/0) samples were compared to theoretical, calculated compliance changes.

published_or_final_version / Mechanical Engineering / Master / Master of Philosophy

No description available.

In this work, two mathematical models for the simulation of the injection molding process were tested and their predictions were validated with experimental data. One of these models is based on the well-known "Hele-Shaw" approximation which, is commonly used by a considerable number of commercial packages. This method utilizes the fact that generally the flow is confined in a narrow gap in which out-of-plane flows may be ignored and, therefore, only a two-dimensional (2-D) solution of the flow field is necessary. One remarkable limitation of this approach is its impossibility of predicting the so-called "fountain flow". Furthermore, this model neglects the role of crystallization kinetics. On the other hand, the other model proposes a methodology that deals with fountain flow and crystallization. It is based on the so-called "2½-D" numerical simulation since it combines a 2-D flow analysis with a 3-D solution of the energy equation. First, a two-dimensional analysis in the gap-wise direction is performed in order to obtain fountain flow information. Then, in-plane two-dimensional flow solutions are coupled with three-dimensional energy results, which incorporate the heat generated by crystallization. Two different thermoplastics were investigated. Polyethylene was selected to characterize the crystalline behavior. Polystyrene was chosen as the amorphous material. In order to obtain insight of the overall injection molding cycle, pressure evolution in the cavity and in the nozzle was examined carefully. More accurate pressure results were computed when using the 2½-D model. This study thus puts in evidence the importance of including fountain flow and crystallization kinetics in the injection molding process.

Polymer melts exhibit some degree of viscoelasticity in most industrial forming operations, and elasticity is particularly important in flows involving an abrupt contraction or expansion in the flow direction. However, the incorporation of a viscoelastic constitutive equation into computer models for polymer processing poses many problems, and for this reason inelastic models have been used almost exclusively to represent rheological behavior for flow simulation in the plastics industry. / In order to explore the limits of viscoelastic flow simulations, we used two nonlinear viscoelastic models (Leonov and Phan-Thien/Tanner) to simulate axisymmetric and planar contraction flows and extrudate swell. Their predictions were compared with those obtained using a strictly viscous model (Carreau-Yasuda) and with experimental results. The models are implemented in a modified Elastic Viscous Split Stress (EVSS) mixed finite element formulation. The viscoelastic constitutive equations are calculated using the Lesaint-Raviart method, and the divergence-free Stokes problem is solved applying Uzawa's algorithm. The decoupled iterative scheme is used as a preconditioner for the Generalized Minimal Residual (GMRES) method. Numerical instability was observed starting at quite low elasticity levels. For the converging flows, the predicted flow patterns were in fair agreement with experimental results, but there was a large discrepancy in the entrance pressure drop. In the case of extrudate swell, the agreement with observation was poor, and convergence was impossible except at the lowest flow rate. / After exploring the limits of simulations using viscoelastic models, we conclude that there are serious barriers to progress in the simulation of viscoelastic flows of industrial importance. The ultimate source of the problem is the melt elasticity, and traditional numerical methods and rheological models do not provide a suitable basis for simulating practical flows. A new approach is required, and we propose that a rule-based expert system be used.

A major obstacle in the prediction of stress profiles in the viscoelastic flow of polymers is a mysterious breakdown of the numerical procedure, which occurs at relatively small values of the Deborah number. Numerous papers have been devoted to analyzing the reason for this failure, but the exact cause remains unclear. Amazingly, all constitutive equations attempted and all kinds of different numerical procedures employed have run into the very same problem. We have investigated several currently popular constitutive models in a viscometric flow field and have found serious limitations in shear flows which may be the source of numerical problems. There is often a lack of appreciation for the computational uses of fluid models in the process of formulating constitutive equations. In the future, the use of the corrotational time derivative, which appears to create this trouble, may be prohibited. Alternatively, it may be possible to avoid the limitation by adding a proper retardation time in constitutive models which use the corrotational time derivative. The well known upper convected Maxwell model does not exhibit limitation but it is believed that a different source of numerical instability may be inherently present. We have adopted the cylindrical axisymmetric 4:1 contraction channel (the die-entry region) to simulate this fluid using the mixed finite element method. The worrisome infinite elongational viscosity predicted by this model in steady extensional flow, indeed, is responsible for the singular behavior of the stress solution field. The most difficult region for convergence, as might be expected, is found to be just after the reentrant corner.