Modelling and optimization of the mechanical and other material properties of a polymer nanocomposite using statistical design of experiments

Polymer compounds are complex systems that typically involve many additives that tend to interact with each other. The system is further complicated by the fact that the additives tend to have an effect on multiple material properties. Hence, the effect of a particular ingredient on a certain material property should not be quantified in isolation. For instance, an important consideration in evaluating the effectiveness of an ingredient is not only how it effects the property it was designed to effect but how it effects other properties, such as the mechanical properties of the compound, in the context of the proportions of the other ingredients. This can be achieved by using the principles of statistical design of experiments. In this investigation the mechanical properties of a polymer nanocomposite, a PVC compound including a Layered Double Hydroxide (LDH) nano-additive, are modelled using 2nd degree Scheffe polynomials. The proportions of all the ingredients (7 in total) are varied in a space filling experimental design. The mechanical properties of each formulation are tested using a tensile test on samples manufactured using injection molding. Injection molding is crucial because it produces homogenous test samples that give an accurate representation of the inherent mechanical properties of the material. The models are determined using k-fold cross validation. The mechanical property models, in conjunction with models of other important material properties, allow for an analysis of the effects and interactions of all of the ingredients. For instance, the analysis shows the negative effect that the LDH has on the elongation at break which needs to be taken into account when considering the positive effects it has on the thermal stability of the compound. Importantly the models can also be used to optimize the system.