Creep behaviour anisotropy

The most common technique employed to date has been that of creep in uniaxial tension. It was shown above that with the inclusion of lateral strain measurements this is a powerful technique giving access to up to 6 independent creep compliance functions. This is more than for any other known method. It further has the overwhelming advantage over many methods, such as say torsional or flexural creep, that the stress is sensibly uniform over the working volume of the specimen. This advantage is paramount in studies of materials displaying non-linear behaviour in creep since analysis of the non-uniform stress situation in non-linear systems is not well developed. Attempts to overcome the non-uniform stress situation in torsion, by recourse to, say, torsion of thin walled tubes, lead to severe difSculties in specimen preparation in oriented materials, when anisotropy of behaviour is to be studied. 

A major problem in data presentation for anisotropy of creep behaviour arises from the dependence of behaviour on many parameters. In Section 10.2 the modulus for a given degree of molecular orientation under given environmental conditions for a single material was given as a function of angle, time and either stress or strain whilst the material compliance functions for these conditions were functions of time and stress or strain. Changes of temperature, composition, orientation, structure, etc., will of course affect the whole pattern of behaviour. 

In studying time dependence, Le. creep behaviour, it is necessary to carry out tests at several stress levels in each direction of interest in the oriented material. I his can involve a prohibitive amount of experimental work and, in practice, little is generally lost by reducing the tests to, say, one creep curve and one isochronous stress-strain curve in each direction. The problem then becomes one of selection of the absolute value of stress for each of the creep curves and is most severe when non-linearity and its anisotropy are well developed. The choice of stress levels is arbitrary but interesting special cases are (a) equal stress levels at all angles and (b) equal strain ranges at all angles. 

From the dynamic mechanical measurements of Stachurski and Ward this process is seen to occur at 50°C in the region of 150 Hz. Clayton et al. concluded therefore that the tensile creep behaviour observed at low draw ratio could be the result of the cone easy shear process occurring in their time scale at 20°C. The observed anisotropy of time dependence of S33 and S22 dS33/dt > dS22/dt) leads to a low value of (0) compared with (90), which is consistent with the above conclusion. Furthermore, calculation of the variation of volume strain with time during tensile creep at 0° and 90° (made possible by the measurement of both lateral and axial strains) showed that at 0° the... 

The anisotropy of creep behaviour of oriented amorphous thermoplastics appears to have received even less attention than that of oriented semicrystalline thermoplastics. This may well be associated with the fact that early measurements based on standard tests demonstrated a marked... 

There appears to be no rigorous theoretical scheme for describing anisotropy of creep behaviour in these materials. However, simple extensions of linear viscoelastic theory are presented and shown to be useful though not completely rigorous. Further development is clearly desirable. 

Oriented thermoplastics can show large anisotropy in creep behaviour, expecially in partially crystalline polymers. Significantly different patterns of behaviour occur in different materials. Not only is there anisotropy of isochronous stiffness, but also of creep rate and non-linearity. If stiffoess is regarded as a function of time, direction and stress or strain, the behaviour is such that the variables are not normally separable. 

Anisotropy of creep behaviour in oriented glasses seems to be less well developed than in partially crystalline materials, but the anisotropy increases with time and presumably temperature. The low level of anisotropy in the time/temperature region investigated may well be a consequence of the polymer being well below its glass transition temperature. Qearly there is scope for systematic investigation of the contributions of the various relaxation phenomena to anisotropy in oriented glassy polymers. 

The value of the simplified technique for producing isochronous stress-strain curves for non-linear isotropic materials by successive loading and unloading of a single sample have been amply demonstrated over many years and fully described elsewhere.These techniques become even more valuable in studies of anisotropy, where samples may be difficult to obtain in large numbers and where the scope of the problem is much larger. A considerable proportion of work on oriented materials reported in the literature is essentially confined to this measurement and does not include studies of time dependence of behaviour. Detailed work has been carried out validating this procedure for oriented materials by comparison of the isochronous stress-strain data with isochronous sections from families of creep curves. ... 

The variation of the 100 second tensile creep modulus with 100 second tensile strain is presented in Fig. 16. The behaviour of a specimen cut from an isotropic sheet (which had been subjected to the same thermal cycle as the drawn sheet) is included for comparison. It is apparent that all specimens exhibited non-linear viscoelastic behaviour, but there is little anisotropy of non-linearity. Furthermore the degree of non-linearity exhibited by the specimens from the drawn sheet is similar to that of a specimen from the isotropic sheet. At any chosen creep strain the anisotropy of modulus for the drawn sheet is relatively low. 

Creep testing of oriented polymers, intended to fully characterise the anisotropy of stiffness behaviour, presents formidable difficulties. For materials with fibre symmetry, techniques are now available which allow complete characterisation. These techniques are considerably more sophisticated than simple creep testing in isotropic materials. For lower symmetries it is still not possible to achieve full characterisation. 

In terms of the strain-hardening modulus, this has been developed by the use of Kuhn and Griin models and Kratky models to relate the development of molecular orientation and meehanical anisotropy (see Section 8.6.3). With regard to the strain rate sensitivity the strain rate-dependent viscosity has been developed by studies of creep and yield behaviour (see Sections 11.3 and 12.5.1). 

---