Error Estimation of Measured Exhaust Gas Temperature in Afterburner Mode in an Aero Gas Turbine Engine

In a turbofan engine, thrust is a key parameter which is measured or estimated from various parameters acquired during engine testing in an engine testbed. Exhaust Gas Temperature (EGT) is the most critical parameter used for thrust calculation. This work presents a novel way to measure and correct the errors in EGT measurement. A temperature probe is designed to measure EGT in the engine jet pipe using thermocouples. The temperature probe is designed to withstand the mechanical and temperature loads during the operation. Structural analysis at the design stage provided a strength margin of 90% and eigenfrequency margin of more than 20%. Thermal analysis is carried out to evaluate maximum metal temperature. Errors are quite high in high-temperature measurements which are corrected using the available methodologies. The velocity error, conduction error, and radiation error are estimated for the measured temperature. The difference of 97 K between the measured gas temperature and calculated gas temperature from measured thrust is explained. The estimated velocity error is 1 K, conduction error is 3 K, and radiation error is 69 K. Based on the error estimation, the measurement error is brought down to 24 K. After applying the above corrections, the further difference of 24 K between measured and estimated value can be attributed to thermocouple error of +/-0.4% of the reading for class 1 accuracy thermocouple, other parameter measurement errors, and analysis uncertainties. The present work enables the designer to calculate the errors in high-temperature measurement in a turbofan engine.

Influence of the Thickness and Frequency in the Infrared Hybrid Method in Consideration of Heat Conduction

The thickness dependency of the temperature image obtained by an infrared thermography was investigated using specimens with three kinds of metal materials of different heat conduction and four kinds of thickness of the specimens. Then, the infrared hybrid method was developed to separate each stress components. However, it contains the influence of heat conduction in the infrared stress measurement method. Therefore, heat conduction error will arise in the infrared hybrid analysis. Then, the new system which corrects the error by an heat conduction inverse analysis was developed. Thereby, the accuracy of the stress intensity factor was able to be raised using heat conduction inverse analysis. Furthermore, the accuracy of hybrid method taking heat conduction into consideration was discussed in comparison with 3-D finite-element analysis and the 2-D infrared hybrid method.

Experimental Studies on the Design of Thermometer Pockets for Steam Mains

The theory of thermometer pocket errors previously developed by the author has been tested by experiment using steam flows under steady-state conditions. When an appreciable temperature difference exists between steam and pipe wall, whatever the reason, errors arise from the temperature gradient in the steam and the loss of heat by conduction and radiation from the pocket. The results confirmed that the net error is a linear proportion of the steam/pipe wall temperature difference and can be substantial if the pocket is not of a suitable length. The effect of plating the pockets with rhodium was not as expected since the accompanying reduction in radiation error was offset by a larger increase in conduction error. In all other cases, the measured conduction plus radiation errors were in good agreement with the theory, while the effects of the steam temperature profile were also broadly as predicted. The basic theoretical approach was satisfactorily confirmed and the accuracy shown to be better than 5 per cent of the steam/pipe wall temperature difference.

Experimental Studies on the Design of Thermometer Pockets for Steam Mains

The theory of thermometer pocket errors previously developed by the author has been tested by experiment using steam flows under steady-state conditions. When an appreciable temperature difference exists between steam and pipe wall, whatever the reason, errors arise from the temperature gradient in the steam and the loss of heat by conduction and radiation from the pocket. The results confirmed that the net error is a linear proportion of the steam/pipe wall temperature difference and can be substantial if the pocket is not of a suitable length. The effect of plating the pockets with rhodium was not as expected since the accompanying reduction in radiation error was offset by a larger increase in conduction error. In all other cases, the measured conduction plus radiation errors were in good agreement with the theory, while the effects of the steam temperature profile were also broadly as predicted. The basic theoretical approach was satisfactorily confirmed and the accuracy shown to be better than 5 per cent of the steam/pipe wall temperature difference.