Design,Fabrication,Performance Analysis And Prediction Of Polymeric Thermal Conductive Materials

The polymeric thermal conductive material as a thermal management material effectively reduces the thermal resistance in the process of heat transfer from the target heat source to the heat dissipation components and the external environment and helps the new technologies to break through the thermal bottleneck.In view of the current problems in the research of polymeric thermal conductive material,such as the lack of systematic understanding of classic filler surface engineering methods,the low intrinsic thermal conductivity of the polymer matrix,insufficient generalization ability of thermal conduction theoretical models,and scarce data labels on thermal conductive material performance,both theoretical and experimental methods were used in the dissertation to carry out innovative scientific explorations on the new thermal conductive material type,new filler surface engineering method,and new thermal conduction model.The effects of both the thermal conductivity of polymer matrix and filler and the filler network structures on the thermal conductivity of the composites were analyzed by the thermal resistance network model.The results expound the contribution of the polymer matrix to the thermal conduction performance of the composites and support the thermal conduction path theory to provide a theoretical basis for the explanation of heat transfer mechanism and the design of high-performance polymeric thermal conductive material.Subsequently,liquid crystal epoxy resin was synthesized and the curing process was determined to be170?/2h+190?/1h using DSC thermal analysis technology.Compared with common epoxy resin,the thermal conductivity of the cured liquid crystal epoxy resin is increased by 32.6%,successfully improving the intrinsic thermal conductivity of thermosetting resin.Through the performance characterization of the new thermal conductive material type fabricated by the liquid crystal epoxy resin,the experiments verified the influences of thermal conductivity of polymer matrix on the thermal conductivity of composites and enriched the investigation on the performance of liquid crystal epoxy resin-based thermal conductive material.In order to improve the systematic understanding of classic filler surface engineering methods,the modified BN fillers with different coating thicknesses were prepared via the sol-gel method and then incorporated into epoxy to systematically investigate the factors and mechanisms which influence the thermal conduction enhancement of classic filler surface coating modification engineering.It was found that increasing the filler content will reduce the thermal conduction enhancement of classic filler surface engineering.On this basis,a new concept of effective filler volume fraction was proposed.Within the effective filler volume fraction,the classic filler surface coating modification engineering will enhance the filler-matrix interfacial interaction and improve the filler dispersion to achieve thermal conduction enhancement.However,the thermal conductivity of the coating material is too low which results in a technical defect of the classic filler surface coating modification engineering and brings large interfacial thermal resistance between the filler-matrix and filler-filler interface.With the increase of the filler content,the filler thermal conduction network gradually becomes the main factor affecting the thermal conductivity of the composites.But the interfacial thermal resistance between the modified fillers will weaken the heat transfer capacity of the filler thermal conduction network.Therefore,increasing the coating thickness and filler content will reduce the thermal conduction enhancement of the classic filler surface coating modification engineering.Reducing the coating thickness will improve the heat transfer capacity of the modified filler network and expand the effective filler volume fraction,but the upper limit of effective filler volume fraction will not exceed the maximum random packing volume fraction of the filler.Subsequently,in view of the technical defect of classic filler surface engineering,electroless deposition filler surface engineering was proposed.The new method deposits copper nanoparticles with high thermal conductivity on the filler surface which makes the modified fillers possess good thermal conduction performance and can effectively bridge the thermal conductive fillers through copper nanoparticles to strengthen the interfacial interaction and reduce the interfacial resistance between the fillers.When the filler volume fraction increases from 5%to 15%,the thermal conductivity enhancement ratio of electroless deposition filler surface engineering expands from 8.45%to 35.32%.The new filler surface engineering method significantly promotes the formation of filler thermal conduction network and improves the heat transfer capacity of the filler thermal conduction network.When the filler content is higher and the filler thermal conduction network is better structured,the thermal conduction enhancement of the electroless deposition filler surface engineering is stronger.The electroless deposition filler surface engineering provides a new method for the design and preparation of high-performance polymeric thermal conductive materials.In response to the insufficient generalization performance of the thermal conduction theoretical model and scarce data labels on thermal conductive material performance,a co-training style semi-supervised artificial neural network model(Co-ANN)was designed to predict the thermal conductivity of polymeric thermal conductive material filled with BN filler.The artificial neural network model(ANN)was employed as the learner and regressor to label the thermal conductive material unlabeled data via a single-view co-training style semi-supervised learning process.Compared with the optimal ANN model based on supervised learning,in the test,the MSE,AARD%,and R~2 improvement rates on the full labeled dataset of the optimal Co-ANN model based on semi-supervised learning are 20.22%,22.24%,and0.74%,respectively.The designed model successfully improves the regression and prediction performance and can provide useful guidance for the design of thermal conductive material.The research of the dissertation has deepened the systematic understanding of the effects of matrix,filler,interface,and structure on the thermal conductivity of composites and carried out innovative work from the design,fabrication,analysis,and prediction of polymeric thermal conductive materials,which has theoretical guiding significance and experimental value for the development of high-performance polymeric thermal conductive materials.