Predicting the electric field distribution under dc stress within polymeric insulations remains a challenge, as space charge behaviour is still difficult to understand in these materials. Charge generation is often thought to arise from injection at the electrodes. Hence, surface roughness should be taken into account, as it strengthens the electric field locally and therefore promotes charge generation at some points. A charge transport model has been developed in 2D to account for surface roughness. The model is first validated with the help of previous results, obtained with a one dimensional charge transport model. Then, with simple shapes accounting for roughness, the simulated results show that surface roughness has a significant impact on the net space charge behaviour. The impact of shape, and size of protusions is presented, as well as a more realistic case were a large surface of the electrode is considered as rough.

No abstract

One of the difficulties in unravelling transport processes in electrically insulating materials is the fact that the response, notably charging current transients, can have mixed contributions from orientation polarization and from space charge processes. This work aims at identifying and characterizing the polarization processes in a polar polymer in the time and frequency-domains and to implement the contribution of the polarization into a charge transport model. To do so, Alternate Polarization Current (APC) and Dielectric Spectroscopy measurements have been performed on poly(ethylene naphthalene 2,6-dicarboxylate) (PEN), an aromatic polar polymer, providing information on polarization mechanisms in the time- and frequency-domain, respectively. In the frequency-domain, PEN exhibits 3 relaxation processes termed β, β* (sub-glass transitions), and α relaxations (glass transition) in increasing order of temperature. Conduction was also detected at high temperatures. Dielectric responses were treated using a simplified version of the Havriliak-Negami model (Cole-Cole (CC) model), using 3 parameters per relaxation process, these parameters being temperature dependent. The time dependent polarization obtained from the CC model is then added to a charge transport model. Simulated currents issued from the transport model implemented with the polarization are compared with the measured APCs, showing a good consistency between experiments and simulations in a situation where the response comes essentially from dipolar processes.