Experimental analysis and numerical simulation of ignition delay time of diesel fuel using a shock tube

Abstract

The present work consisted in developing a computational routine for prediction and characterization ignition delay time, pressure, temperature generated and energy released in the combustion process in a shock tube. Mathematical modeling considered the high-pressure region of the shock tube known as the conducted section. The Lazarus code compiler was used to generate the unstructured triangular meshes and to implement the computational routines to simulate the propagation of shock waves inside the shock tube. The mathematical modeling used the geometric parameters of the shock tube of the combustion laboratory of the Federal University of Minas Gerais and was based only on the displacement of the shock wave after the diaphragm rupture. The main objective of the work was the development of computational routines to characterize the shock wave parameters of the ignition delay times of conventional diesel to validate the experimental tests carried out in the shock tube conducted by Santana et al. [3] and compare it with the experimental tests and numerical simulations carried out by other authors available in the literature. A linear regression of the experimental tests conducted by Santana et al. [3] with conventional diesel was performed to obtain the Arrhenius equation for numerical simulation of the ignition delay time of diesel under the following conditions: temperatures from 880 to 1300 K, pressures 24 bar and equivalence ratio 1. The results show good prediction between experimental tests and numerical simulations. Were found delay times ranging from 425 to 1890 μs. Considering all temperature range, the difference between the experimental and simulated test was approximately 15%, this difference also can be explained by the measurement in the shock tube, that the ignition delay time was calculated by the time difference between the passage of the shock wave by the pressure sensor and the start of the ignition detected by the luminosity detection sensor. This work is relevant because it was developed computational routine in open software and the results achieved in the simulation were considered satisfactory, since they are in the same order of magnitude as the results found in the experimental tests carried out by Santana et al. [3] and other works.