| In order to meet the increasing demand for electricity and electrical power's efficient transmission, power ratings and operating voltages have seen a continuous rise over the years, which inevitably brings greater electro-thermal stresses to power cable insulation. Polyethylene (PE), an insulating material widely used in underground power cables, is susceptible to electrical degradation and exhibits a low thermal conductivity. Consequently, there is a compelling need to improve its dielectric and thermal performances.;During the last decades, considerable attention has been given to a new class of dielectric material---nanodielectrics, i.e. dielectric composites containing nanometric fillers. They are frequently reported to have superior dielectric properties as compared to neat polymers and microcomposites, and thus have great potential to serve as the insulating materials for highvoltage (HV) power cables. Nonetheless, such property enhancements can only be achieved when the nano-filler has a good size dispersion and spatial distribution within the host dielectric. However, due to nano-fillers' strong tendency to agglomerate, and their generally poor compatibility with polymers, their dispersion is often compromised with aggregates of micrometric sizes. Furthermore, hydrophilic nano-fillers attract water to the filler-matrix interface, not only impairing the crucial role of the interphase, but also causing property and material degradation. In order to facilitate a homogeneous nanoscaled filler dispersion and prevent water absorption, inorganic nano-fillers are commonly treated with dispersants and coupling agents. However, this adds extra work to material fabrication. Moreover, little is known about the long-term stability of these surface modifications under electro-thermal stresses. Furthermore, a thorough removal of hydrophilic groups may not be straightforward, and so is the complete prevention of filler aggregation and water absorption.;In this context, the objective of this PhD research is to develop PE-based nanodielectrics with enhanced dielectric and thermal performances, as insulating materials for HV underground power cables.;In order to avoid the aforementioned problems concerning surface modifications, polyhedral oligomeric silsesquioxanes (POSS), which are by nature nanoscaled molecules bearing builtin functionalities, were used. POSS selection, POSS loading and fabrication method all play an important role in developing PE/POSS nanodielectrics with enhanced performances. In this project, three types of POSS with different alkyl substituents were studied---solid octamethyl- POSS (OmPOSS, OM), solid octaisobutyl-POSS (OibPOSS, OIB) and viscous-liquid isooctyl- POSS (IoPOSS, IO); 1 wt% and 5 wt% POSS loadings were investigated; and three fabrication methods were attempted---ball milling (BM), xylene solution blending (XSB), and E). The obtained composites were examined concerning their dielectric and thermal properties. Additional characterizations such as scanning electron microscopy (SEM) and differential scanning calorimetry (DSC) were also performed to obtain information for possible performance explanations.;The results showed that none of the utilized fabrication methods were effective in producing PE/solid-POSS nanodielectrics. Regarding thermal properties, POSS was able to increase PE's thermal conductivity up to 8%. Moreover, it did not have any significant effects on PE's thermal stability in a nitrogen atmosphere, and it caused little changes in PE's melting and crystallization behaviors, as well as degree of crystallinity. As far as dielectric properties were concerned, all PE/POSS composites had enhanced resistance to corona discharges, from a minimum of 6% to a maximum of 61%. This strengthens insulation against cumulative damage brought about by electrical degradation under service conditions. Moreover, POSS had only a weak effect on the dielectric response of PE at ambient temperature: the dielectric constants and the dielectric losses of the composites being, respectively, 2.25--2.35 and generally 10-5-- 10-4, were similar to that of PE. This is praiseworthy for their applications related to electrical insulation which require low losses and low dielectric constants. At elevated temperatures, the influence of POSS on PE's dielectric response remained small, except for some loss increase in PE/IO1 (E) and PE/OM5 (E) at frequencies below 10 Hz, and for enhanced charge transports (diffusion and conduction) in PE/IO5 (E). In contrast to the laudable performance in erosion resistance, POSS was barely able to improve PE's short-term dielectric breakdown strength; in the best-case scenario, dielectric breakdown strengths of PE/POSS composites were similar to that of PE.;Among the three methods used, ball milling was more effective in dispersing 5 wt% OibPOSS, extrusion and xylene solution blending were more effective in dispersing 1 wt% OibPOSS, and extrusion was more effective in dispersing IoPOSS. Fabrication methods had little effect on thermal conductivity and dielectric response, except when they induced critical POSS distributions, as in the cases of PE/IoPOSS composites: possible filler alignment in PE/IO1 (XSB) and possible filler quasi-connection in PE/IO5 (E) may be the main reasons for PE/IO1 (XSB)'s enhanced thermal conductivity and PE/IO5 (E)'s intensified charge transports. The method of material fabrication had an influence on erosion resistance: composites obtained by xylene solution blending generally had higher erosion resistances than those obtained by extrusion. This may be related to material precipitation and xylene evaporation processes, which somehow brought POSS to the surface of samples, forming an erosion-resistant layer that hindered erosion from further progression. Fabrication method influenced dielectric breakdown strength through filler dispersion. Good filler dispersions with relatively small POSS sizes resulted in roughly maintained breakdown strengths, as in the cases of all 1 wt%-POSS composites, all 5 wt%-IoPOSS composites, and the 5 wt%-OibPOSS composite obtained by ball milling. On the other hand, compromised filler dispersions with relatively large POSS sizes resulted in reduced breakdown strengths, as in the case of all 5 wt%-solid-POSS composites produced by extrusion and xylene solution blending. Among all the composites studied, PE/OIB1 (E) was found to be the best material for HV power cable insulation, thanks to its 23% enhancement in erosion resistance, its low dielectric constants and dielectric losses under utility frequencies (50--60 Hz) at both ambient and elevated temperatures, its unreduced breakdown strength, and its 6% improvement in thermal conductivity. PE/IO5 (E) had good performances at ambient temperature but its dielectric properties at elevated temperatures were much less appealing. PE/IO5 (XSB) had good dielectric performances, especially its 61% improvement in erosion resistance. However, compared to PE/OIB1 (E), it had a reduced thermal conductivity; it required five times more filler content; and its fabrication method (i.e., xylene solution blending) was more demanding than traditional extrusion. PE/OIB1 (BM) and PE/OIB5 (BM) had low dielectric permittivities and well-maintained breakdown strengths. However, more characterizations were needed to better evaluate their potentials as HV insulation materials. Both PE/OM1 (XSB) and PE/IO1 (XSB) had a 40%-enhanced erosion resistance. However, PE/OM1 (XSB) had a decreased thermal conductivity and a reduced breakdown strength, and PE/IO1 (XSB) had a diminished shape parameter and thus a greater breakdown-strength scattering. PE/OIB1 (XSB) had an 8%-increased thermal conductivity. However, its shape parameter was much lowered and its erosion resistance was exceptionally smaller than other XSB-produced 1 wt% composites. As for PE/OM5 (XSB), PE/OIB5 (XSB), PE/OM5 (E) and PE/OIB5 (E), their breakdown strengths were largely reduced. |
