Publications /
International Journals
2020-Numerical Prediction of DC Breakdown Characteristics in LDPE With Current Profile as Critical Index
posted 2021-01-26 16:52:51

Minhee Kim, Su-Hun Kim, and Se-Hee Lee, "Numerical Prediction of DC Breakdown Characteristics in LDPE With Current Profile as Critical Index", IEEE ACCESS, Nov 2020.

 

Abstract:


Understanding the DC breakdown characteristics of polymeric insulators is essential for stable operation of high-capacity DC electrical equipment. To predict the breakdown characteristics of low-density polyethylene (LDPE), we propose a numerical methodology with a new critical index in which the internal current varying with temperature, thickness, and injection barrier height. To evaluate this current-based index, we applied the fully coupled bipolar charge transport (BCT) and molecular chain displacement (MCD) models to analyze the influence of each variable on breakdown phenomena. The results of this analysis revealed that the amount of space charge accumulation within the insulator has a maximum value at approximately 50 ¡ÆC, which corresponds to the known morphological transition temperature of LDPE. The breakdown strength calculated using this numerical model was found to decrease with increasing temperature and thickness. Although injection barrier height at the electrode was found to be negatively correlated with breakdown strength, its effect was not as significant as that of the other variables. The breakdown strength values obtained using this numerical method were found to be in close agreement with values reported in the literature. Based on these results, we newly suggest the physical quantity to predict the breakdown strength, the current relaxation speed, which is the slope of the Boltzmann sigmoid function, as a positively correlated index. Finally, we determined that the breakdown phenomena are initiated when the amount of impact accumulated in the insulator changes discontinuously and analyzed the contribution of factors affecting the breakdown using the Pearson correlation coefficient and the Sobol sensitivity index.