Prediction of thermal conductivity of graphene-based adhesives

• Main Author: Κοντογιάννης, Αθανάσιος
• Other Authors: Kontogiannis, Athanasios
• Language: English
• Published: 2022
• Subjects: Thermal conductivity prediction; Composites; Digimat; Abaqus; Πρόβλεψη θερμικής αγωγιμότητας; Σύνθετα υλικά; Αριθμητικό μοντέλο; Προσομοίωση

Polymeric composites, during the last couple of decades, have been widely studied, hugely improved and advanced as it comes to their mechanical, thermal and electrical properties. This has resulted to them being widely used not only in the automotive and aerospace industry, but in every-day items too. While fiber-reinforced polymers and laminates are the most commonly used composites, in the last years, inclusions in the nano-scale have been considered as a great alternative to the more typical inclusions, as they combine great intrinsic thermal properties with more-than-decent mechanical properties. Heat transfer has recently become one of the most important research fields for the design of polymer composites, since the electronic components are constantly becoming faster, lighter and more power-dense, cramming loads of computational power in as little physical space as possible. This creates the need for superior materials that manage to dissipate greater amount of heat per volume. So, the polymeric nano composites provide a great opportunity for the design of the appropriate material for each use, as there is a great number of variables that contribute to the final properties of the material such as: the type of the fillers, their physical properties, their content in the composite, the processing and manufacturing method of the composite. However, there is a great amount of effort involved when it comes to developing such materials since there is no way to know a priori their effective thermal conductivity. This creates the need of having to physically make and measure a great number of materials until the required properties are achieved. That is the problem that this work tackles. More specifically, the goal of this work, is to predict the thermal conductivity of polymeric nano composites. The first idea was to utilize an analytical micro mechanical model. After examining the previous research work, the model of Nan [1], was chosen since it is fairly complex and considers many of the parameters that affect the thermal conductivity of the composite. Although promising, the model did not converge with the experimental data from [2] (which is the benchmark for this thesis). Since no other promising analytical models were found, the construction of a numerical model with a Representative Volume Element (RVE) of the composite, was the next reasonable option. At first, the DIGIMAT FE software was utilized but due to its limitations when it comes to discretizing the RVE and also due to its inability to run parametrically, the decision of constructing the model via a custom Python script using ABAQUS, and simulating Carbon Nano Tubes (CNTs) instead of Graphene Nano Platelets was made. The different choice of particles was made due to the less complex geometry of the CNTs. To overcome various challenges regarding the required periodic geometry of the RVE and its meshing, a series of powerful tools were developed that add to the functionality of ABAQUS in a way that they can be used in many different applications (not limited to RVEs and composites, but to actually mesh any complex geometry). The model showed great results and converged with the experimental data of [2]. Through this work and the current thesis, a very interesting discussion is made on whether the vision of a model that can predict the thermal conductivity of a particulate polymeric nano-composite is attainable.