Reconstruction method considering the electrostatic signal waveform of velocity distribution in flame based on EST

Abstract

The velocity distribution of flame is an important combustion dynamic parameter that is used to analyze the stability, heat transfer and propagation of combustion. Digital particle image velocimetry and laser Doppler velocimetry are commonly used to measure the velocity distribution of flame. However, these methods work by adding tracer particles into the flame. A dual-section electrostatic tomography (EST) system is designed to detect the induction signal of the charged particles generated by the ionization process of combustion. The velocity distribution of flame is reconstructed through the electrostatic direct velocity tomography (EDVT) method. However, the different distance between the movement trajectory of the charged particles and a pair of electrodes for the cross-correlation calculation causes the difference in the induced signal waveform. It leads to an error in the cross-correlation velocity calculation. By analyzing the influence of the single charge induction signal waveform on the calculation of cross-correlation velocity, the reason for this error is clarified. Therefore, an electrostatic signal-waveform-based direct velocity tomography (ESDVT) method is proposed. The experimental results from a belt-type electrostatic device show that the average value of the L2-norm error of the ESDVT method is reduced by 40.9% compared with the EDVT method. Finally, using the method proposed in this paper, the velocity distribution at different sections of the Meker flame is reconstructed without adding tracer particles.