Application of multicriterial optimization for determination of crystallinity degree
of semicrystalline polymers
Polimery 2003, No 4, 288
On the basis of a literature review, an additional criterion of resolution of diffraction curves into crystalline peaks and amorphous component, in generally used Hindeleh-Johnson method of the polymers crystallinity degree (SK) determination, has been proposed. Besides two classical optimization methods by Powell and Rosenbrock, new multicriterial optimization procedures (hybrid system) have been used. They were applying the genetic algorithms which are less sensitive to initial conditions choice. All that allowed to obtain much more credible solutions. The results of resolution of diffraction curves of poly(ethylene terephthalate) and polypropylene samples, using three above mentioned methods, have been compared (Tables 1 and 2, Figs 3-6). As well statistical analysis concerning numerical SK values obtained was presented (Table 3). Scatter of these values in the case of multicriterial optimization appeared to be much narrower than when classical optimization methods were used.
Keywords: semicrystalline polymers, degree of crystallinity, distribution of X-ray diffraction curves, classical optimization methods, multiobjective optimization in hybrid system, statistical analysis of the results
Application of multicriterial optimization for determination of crystallinity degree of semicrystalline polymers