MATHEMATICAL AND STATISTICAL MODEL OF CLEANING IN A HIGH-GRADIENT MAGNETIC FIELD OF HYDROGENATED FAT FROM SUSPENDED CATALYST

The introduction specifies the object of the study - hydrogenated fat subjected to purification from suspended nickel-containing catalyst in a high-gradient magnetic field. The nickel-containing catalyst, due to the presence of nickel salts in it, is paramagnetic and can be effectively extracted from hydrated fat in the magnetic field by a method of highly gradient magnetic separation. In the experiment technique the magnetic field strength was not less than 600 kA/m, the gradient not less than (20-30)∙104 kA/m2. Based on the results and their discussion, the analysis of empirical dependencies on each factor basis is carried out separately, as well as the construction of classical types of one-dimensional non-linear regression models for each of them.  It is shown that the most statistically reliable are the exponential models. It was found that after purification the content of suspended nickel-containing catalyst in the hydrogenated fat has an inverse proportional dependence on the magnetic field strength, height (length) of the filtration zone and filtration rate. A two-dimensional exponential regression model of the content of the nickel-containing catalyst in the hydrogenated fat in the complex dependence on the rate and height of the filtration zone is built. Based on the experimental data at a certain height of the filtration zone and the selected filtration rate, a mathematical-statistical model of change in the concentration of nickel-containing catalyst depending on the magnetic field strength is made. Using the resulting two-dimensional exponential regression model, it is possible to predict the degree of purification of the salomas from the suspended catalyst depending on the specific tension of the magnetic field and to determine the tension of the magnetic field, at which the concentration of nickel as a result of cleaning will reach an advance value.