A study of melt spinning, drawing and twisting of PET monofilament

The purpose of this research was to inyestigate the effect of spinline, drawing and twisting variables on the structure and properties of polyethylene terephthalate (PET) fibers. This work is the first of its kind done for PET fiber in the Chemical, Metallurgical and Polymer Engineering Department at The University of Tennessee. As such, it is a contnuation of the work done on other fiber materials in the Department. For the same reason, the study was by design a broad survey rather than a specialized study of any one subject.

This work involved the design and development of several major pieces of equipment, experimental procedures, characterization techniques and computer software.

Equipment design projects included the redesign of the extruder facilities, the design of a density gradient column apparatus, a bench scale drawing apparatus, and a draw-twist texturing machine.

An extensive computer software package with full graphics capability was developed to analyze x-ray diffraction data for polyester and other polymers. The package, known as APLxlo, is interactive, self supporting, and requires no previous computer experience for its use.

The polyester used in this research was donated by the Tennessee Eastman Company (TEc Reference X-14085-63). This polyester is considered to be textile grade with average molecular weight of 37,ooa and contained Q.445 percent of Ti0, delustrant. The as-received material had to be dried and crystallized before it could be used.

The polymer was extruded using a capillary rheometer and a screw extruder. The rheometer studies showed the effects of temperature and resident time on the polymer degradation as evidenced by a decay in spinnability, melt viscosity and molecular weight. The extruder was used to prepare large quantities of homogeneous samples for drawing and twisting studies.

All spun samples were amorphous. Changes in variables which produced greater spinline stress such as increased take-up velocity, decreased extrusion rate and decreased temperature and residence time, produced increases in the amorphous orientation of the fibers.

The amorphous orientation increased linearly with spinline stress. The stress optical coefficient, defined as the slope of this curve, was found to be the same as that observed for the molten polymer, leading to the conclusion that the orientation produced by stressing the molten polymer during spinning is simply frozen in by the vitrification process.

The elastic modulus, tensile strength and yield strength increased while the natural draw ratio and elongation to break decreased with increasing amorphous orientation and spinline stress.

Annealing studjes done on melt spun fibers showed that small amounts of amorphous orientation produced accelerated crystallization rates. The crystallization kinetics were of the Avrami form but show a secondary crystallization zone.

The drawn fibers were optica!ly opaque. X-ray studies showed the development of crystallinity and orientation with increasing draw ratio. Small angle x-ray scattering studies of drawn fibers showed some equatorial scatter masked by a more intense contribution from Ti02 particles. Electron microscope photographs reveal only a smooth skin.

The Hermans C axis orientation factors were measured using the To5 reflection for a variety of samples with different degrees of intial orientation. Measurements were made for increasing draw ratio up to the Break point before and after an annealing treatment. APLxlo was used for this analysis.

At a give draw temperature, the final orientation in drawn samples with different initial orientations were found to correlate with a variable named the standard draw ratio. This variable was defined as the ratio of the percent elongation introduced during drawing to the maximum percent elongation to break for the same starting fiber.

A theoretical model to explain the change in modulus with twist for single monofilaments is proposed. The model assumes that a filament can Be considered a bundle of fibrils and, hence, is analogous to a continuous filament yarn under twist where fibrils take the place of fibers in the yarn analogy. The model allows for the possibility that the individual fibrils may be strengthened or damaged as a result of twisting.

Twisting experiments with polypropylene and PET were limited to an attempt to show the validity of the proposed theoretical model. Polypropylene fibers show twist strengthening while PET fibers are damaged when twisted at room temperature after being drawn to the natural draw ratio. The theoretical model was correct in predicting trends, but due to experimental problems it remains untested in a truly quantitative form.