Biopolymer Non-Parametric Analysis: A Degradation Study under Accelerated Destructive Tests.

The degradation of biopolymers such as polylactic acid (PLA) has been studied for several years; however, the results regarding the mechanism of degradation are not completely understood yet. PLA is easily processed by traditional techniques including injection molding, blow molding, extrusion, and thermoforming; in this research, the extrusion and injection molding processes were used to produce PLA samples for accelerated destructive testing. The methodology employed consisted of carrying out material testing under the guidelines of several ASTM standards; this research hypothesized that the effects of UV light, humidity, and temperature exposure have a statistical difference in the PLA degradation rate. The multivariate analysis of non-parametric data is presented as an alternative to multivariate analysis, in which the data do not satisfy the essential assumptions of a regular MANOVA, such as multivariate normality. A package in the R software that allows the user to perform a non-parametric multivariate analysis when necessary was used. This paper presents a study to determine if there is a significant difference in the degradation rate after 2000 h of accelerated degradation of a biopolymer using the multivariate and non-parametric analyses of variance. The combination of the statistical techniques, multivariate analysis of variance and repeated measures, provided information for a better understanding of the degradation path of the biopolymer.

Biopolymer Degradation Analysis: Accelerated Life Testing Study to Characterize Polylactic Acid Durability.

While the degradation of Polylactic Acid (PLA) has been studied for several years, results regarding the mechanism for determining degradation are not completely understood. Through accelerated degradation testing, data can be extrapolated and modeled to test parameters such as temperature, voltage, time, and humidity. Accelerated lifetime testing is used as an alternative to experimentation under normal conditions. The methodology to create this model consisted of fabricating series of ASTM specimens using extrusion and injection molding. These specimens were tested through accelerated degradation; tensile and flexural testing were conducted at different points of time. Nonparametric inference tests for multivariate data are presented. The results indicate that the effect of the independent variable or treatment effect (time) is highly significant. This research intends to provide a better understanding of biopolymer degradation. The findings indicated that the proposed statistical models can be used as a tool for characterization of the material regarding the durability of the biopolymer as an engineering material. Having multiple models, one for each individual accelerating variable, allow deciding which parameter is critical in the characterization of the material.