Applications Of The Intelligence Methods In The Study Of Polymer/Inorganic Nanocomposites

With the quick development of electronic information, the functional requirement on the macromolecule memory material is rising. As matrix materials, the limitation of traditional materials, such as including Ethylene-vinylacetate (EVA) and linear low density polyethylene (LLDPE), is revealed gradually. Due to this reason, it is necessary to improve the property of matrix material. Aiming at this purpose, a lot of experiments should be done. However, it will need much more regulations groped with experiments, manpower and material resources and waste a plenty of time. Therefore, it is important and urgent for us to find a way to forecast the material properties with reduced times of experiments.Nanomaterial possesses a series of novel and fascinating physical and chemical properties. Polymer/inorganic nanocomposites can be made by inorganic fillings distributed in organic polymer matrix with nanometer dispersion state. Due to the rigidity, dimension stability and thermostability of inorganic material and the toughness of organic polymer, the properties of polymer/inorganic nanocomposites were improved effectively. Now, as the matrix materials, the application of macromolecule materials in polymer/inorganic nanocomposites becomes a hotspot in the researches. Since it possesses important significance in the development of the high property and functional composite material, we believe that polymer matrix/inorganic nanocomposites with nano-particle is of more value than that with traditional fillings, which is the best way to prepare new matrix. In order to resolve the limitation problem of the single material and traditional composite material, and to exert cooperation effect of composite material and nanomaterial, which possess some excellent property that single component material does not have, the way of adding inorganic nano-particle into EVA and LLDPE or into their copolymer was adopted to prepare nanocomposite material.In this dissertation, by using orthogonal experiment method, a series of EVA/TiO2, LLDPE/ZnO and EVA/LLDPE/ZnO nanocomposite materials were prepared with banbury mixer molten blend method. The banbury mixer molten blend method were used because this method was easy to operate and industrialize, and has formed and become one well-rounded way in the molecule memory matrix material production. Based on these finite experiments, and instructed by scientific and intelligent method, the aims to the request material were approached quickly. This can resolve the problem of wasting time, manpower and material resource in the modified modification of polymer-based nanocomposites. Some examples were shown here. The first one was that by using deduced means of finite element method to simulate the macro-effective elastic modulus of the polymer-based inorganic nanocomposites. The second was that the properties of polymer-based inorganic nanocomposites were predicted by artificial network model and its modified model. And the third was that the preparation technology of polymer-based inorganic nanocomposites was improved by two methods. Samples were prepared through the improved technology and the formation mechanism and their mechanical properties were studied. Much useful information was obtained, and the properties were improved highly. This would possibly be the master batch of the pyrocondensation tube material.The main contributions of the dissertation are as follows:1. By using orthogonal experiment method, EVA, with LLDPE and their blend composite material as matrix, a series of EVA/TiO2 nanocomposite material, LLDPE/ZnO nanocomposite material and EVA/LLDPE/ZnO nanocomposite material were prepared with the banbury mixer molten blend method, under the condition of filling nano-particle with different proportion and different preparation technology. The properties of nanocomposite materials were tested, which provide the foundation for further research.2. The means of finite element method was deduced to simulate the Macro-effective elastic modulus of polymer/inorganic nanocomposites material. The effective elastic modulus was simulated for EVA/TiO2 nanocomposite materials, and the effects of effective elastic modulus were studied by the amount and the shape of nanoparticle. The effective elastic modulus of nanocomposites material increases upon increasing the concentration of nano-TiO2, no matter the filled particles was quadrate or round. The effective elastic modulus of quadrate particle-filled composites was bigger than that of round particle-filled composites, when equivalent amount of nanoparticle was filled in matrix. The results of the simulation and the test were accordant basically, and the results of round nanoparticles filled in matrix were much better than that of quadrate nanoparticles. It was showed that the method of simulating the effective elastic modulus of polymer-based nanocomposites materials by means of finite element method was effective, practical and feasible.3. Single target properties of EVA/TiO2 nanocomposite material, double target properties of LLDPE/ZnO nanocomposite material and multi-target properties of EVA/LLDPE/ZnO nanocomposite material were forecasted. It was showed that forecast errors of single target property and double target properties were very small. However, the forecast errors of multi-target properties were distributed nonuniformly, and a BP-Markov chain model were then structured to resolve the problem with a perfect result.4. The process parameters of EVA/nano-TiO2 composite materials were optimized based on orthogonal experiments and also on the neural network of BP. Samples were prepared by the best one-step method. The dispersion morphology and the mechanical properties of material were studied by FESEM and SEM. It was showed that the optimizing method based on the neural network of BP and genetic algorithms was better than that of optimizing method based on orthogonal experiment. It was found that TiO2 particles with pore size of about 20~60 nm dispersed in EVA matrix, and Nano-TiO2 particles were well-distributed in EVA matrix. The tensile strength, fracture elongation rate and modulus of elasticity of nano-composite material were improved, which results in the effect of reinforcing and toughening. The tensile strength was the highest when nano-TiO2 amount was 5%. The fracture elongation rate was the highest when nano-TiO2 amount was 1%. The modulus of elasticity tended to increase, along with the rising of nano-TiO2 amount.5. The tensile strength and fracture elongation rate of nanocomposite material were tested. A relatively optimal group of process parameters were obtained by orthogonal experiment and another group of process parameters were also obtained by the neural network of BP based on genetic algorithms. With analysis, it was showed that the group of process parameters obtained by BP neural network and genetic algorithms based on the orthogonal experiment data were much better than that group obtained by orthogonal experiment. LLDPE/nano-ZnO composite materials were prepared by melt blending under the better process parameters. The dispersion morphology and the mechanical properties of material were studied by FESEM and SEM. The tensile strength and fracture elongation rate of nano-composite material were tested. It was found that ZnO particles with pore size less than 100nm were dispersed in LLDPE matrix, and nano-ZnO particles were well-distributed in LLDPE matrix. The tensile strength and fracture elongation rate of nano-composite material were improved, which results in the effect of reinforcing and toughening. The ZnO particles linked with LLDPE through chemical bond were transformed from brittleness to toughness when the bond ruptured. The tensile strength and impact strength were the highest when nano-ZnO amount was 3%. The fracture elongation rate was the highest when nano-ZnO amount was 5%. The property of nanocomposite materials were much improved when the amount of nano-ZnO was 3%, and it was the acceptable result when the content of nano-ZnO was 2%. Some reasons should be responsible for the degradation, such as the search of algorithm or the design of experiment method. The group of process parameters was not the best when nano-ZnO amount was 2%, but it is the second best, which also showed that the method was effective, practical and feasible, though it needs furhter studies.In conclusion, the matrix modification idea of the macromolecule memory material was put forward. Some intelligent methods were used in the study of modification material. A Markov chain model based on the BP artificial neural network was integrated to forecast the properties of composite material. The conventional orthogonal experiment, intelligent artificial neural network and genetic algorithms were combined together, and process optimization model was constructed and properties of polymer-based nanocomposites were studied. These studies will provide new ways for the study of composite material and enlarge the field of our vision.