Accuracy of non-resonant laser-induced thermal acoustics (LITA) in a convergent–divergent nozzle flow

Abstract

Non-resonant laser-induced thermal acoustics (LITA) was applied to measure Mach number, temperature and turbulence level along the centerline of a transonic nozzle flow. The accuracy of the measurement results was systematically studied regarding misalignment of the interrogation beam and frequency analysis of the LITA signals. 2D steady-state Reynolds-averaged Navier–Stokes (RANS) simulations were performed for reference. The simulations were conducted using ANSYS CFX 18 employing the shear-stress transport turbulence model. Post-processing of the LITA signals is performed by applying a discrete Fourier transformation (DFT) to determine the beat frequencies. It is shown that the systematical error of the DFT, which depends on the number of oscillations, signal chirp, and damping rate, is less than $$1.5\%$$ for our experiments resulting in an average error of $$1.9\%$$ for Mach number. Further, the maximum calibration error is investigated for a worst-case scenario involving maximum in situ readjustment of the interrogation beam within the limits of constructive interference. It is shown that the signal intensity becomes zero if the interrogation angle is altered by $$2\%$$ . This, together with the accuracy of frequency analysis, results in an error of about $$5.4\%$$ for temperature throughout the nozzle. Comparison with numerical results shows good agreement within the error bars.