화학공학소재연구정보센터
Fuel, Vol.214, 528-537, 2018
Prediction of growth of jet fuel autoxidative deposits at inner surface of a replicated jet engine burner feed arm
A low thermally stable jet A-1 fuel was heated up to 115 h under engine representative condition in a simulated burner feed arm with cylindrical shape using "Aviation Fuel Thermal Stability Test Unit (AFTSTU)". The local growth of surface carbonaceous deposits was measured by the use of three K-ntype thermocouples which were inserted in discrete locations along the inner surface of heated tube. Subsequently, the deposited tube was sectioned and prepared for " Scanning Electron Microscopy(SEM)" to visualise the circumferential profile of the carbonaceous deposit. This helped to identify the axial profile of deposit thickness by integration of the average deposit thickness at five cross sections along the tube. Using the temperature rise data, a one dimensional, analytical heat transfer model was used to calculate local deposit thickness as a function of time. A transient fuel dependent, two stage chemical kinetic model was developed to simulate the growth of deposit as a function of time, temperature, and fuel chemical composition. In this model, the growth of deposit was assumed to be equal as a geometrical displacement in radial direction. Such a displacement is explicitly calculated using an initial rate of deposition(steady state at t = 0) and accelerating deposition rate(at t = t +Delta t) to account for the non linear rate of deposition over the time period of thermal exposure. The concentration of insoluble species and initial rate of deposition as well as temperature at the adjacent layer to the heated surface were obtained through a steady state " computational fluid dynamics(CFD)" simulation using pseudo detailed mechanism of fuel autoxidation by Kuprowicz et al. (2007). The rate parameters of the accelerating deposition were optimised by the application of pattern search method with respect to the calculated deposit thickness from one dimensional heat transfer model. Eventually, the optimised model was used with a transient CFD simulation with dynamic mesh using Ansys Fluent commercial package to predict the growth of deposit in a reactive flow medium. The predicted results are in good agreement with the experimental data obtained from visualised deposit.