Modeling of splat particle splashing data during thermal spraying with the Burr distribution

Abstract

Splashing of splat particles is one of the most important phenomena in industrial processes such as thermal spray coating. The data relative to the degree of splashing of splats sprayed with a normal angle are commonly characterized by the Weibull distribution function. In this present study, an effort has been made to show that the Burr distribution is better than the Weibull distribution for presenting the distribution of the degree of splashing. For this purpose, the Burr Type XII distribution and Weibull distribution are compared using different criteria. Furthermore, because of the great importance of statistical prediction of censored data in reducing costs and improving quality of the coating process, we consider different predictors of this data based on a progressively censored sample. For computing the prediction values we obtain the maximum likelihood estimates using the Expectation-Maximization (EM) algorithm. An important implication of the present study is that the Burr Type XII distribution more appropriately described the degree of splashing data. Therefore, the Burr Type XII can be used as an alternative distribution that adequately describes the splashing data and thereby predicts the censored data.

Keywords

• Burr Type XII
• Censored Data
• splat particle
• Splashing
• Progressively censoring
• Thermal Spray