Abstract

Temperature measurements at the inlet of a gas turbine inlet have proven challenging due to the high flow velocities, limited space, and high temperatures. This work investigates the use of the Integrated Spectral Band Ratio (ISBR) method to measure temperature using a new set of gas bands and a new method for collecting, measuring m and converting raw signals to digital temperature data at low cost and high speed (4 Hz). The Pressurized Oxy-Coal (POC) reactor was refurbished and used in this work to simulate turbine inlet conditions. The POC was modified to produce a stable natural gas and air flame up to 8.35 atm (122.7 psi). A water-cooled probe was designed and fabricated to house a sapphire rod and a ZBLAN optical fiber to collect irradiation signals from the inside of the POC at path lengths varying from 25.4 mm (1 in) to 203.2 mm (8 in). The light passing through the fiber was incident on four photodiodes located behind band pass filters. The detectors, amplifiers, and CPU were mounted on a printed circuit board where the incoming radiation was converted to a current and then a temperature. Three tests or "Runs" were conducted where measurements were obtained at 8 optical path lengths varying from 25.4 mm to 203.2 mm in 25.4 mm increments for each of three reactor pressures: 6.98, 7.70, and 8.35 atm (102.6, 112.7, and 122.7 psi). The Optical probe, located 1523 mm downstream of the burner, was used to take Optical temperature measurements of the gas which were compared to aspirated thermocouple measurements located 1219 mm downstream of the burner. The wall surface temperatures were also measured by the Optical probe and were compared to B thermocouple temperatures at the same distance from the burner but embedded 3.17 mm beneath the surface, inside the wall of the POC. The temperature trends of the optical wall, aspirated thermocouple, and embedded thermocouple showed a decrease in temperature with increasing time in all three test runs. The decrease was caused by the need to turn off the flame each time the probe path length was changed. Typically, the data were less likely to converge on a solution at lower pressures and shorter path lengths although converged data were obtained at 50.8 mm (2 in) and 6.98 atm (102.6 psi). The trends of the optical gas temperatures matched more closely with the optical wall temperatures than the aspirated thermocouple temperatures. There were times, typically at longer path lengths, when the optical gas temperature measurements were higher than the aspirated thermocouple temperatures, which is illogical, indicating large errors existing in the optical gas temperature. The ISBR method uses a set of four measurements to solve for three unknowns and is therefore over specified. It was found that the temperatures produced by the ISBR method could not be used in a model to match the absolute intensities of the measured gas bands. In particular, the C-band measured value appears to be too low compared to the other bands. The non-linear calibration produced by the C-band and a signal from the C-band detector even when it is blocked by an optical filter indicate the C-band was not being measured correctly and is the likely cause for the error. It is also possible that the wall emissivity of the C-band is not the same as the E-band which could also contribute to the observed results. A new C-band filter, detector, and circuit board are required to correct the issue.

Degree

MS

College and Department

Ira A. Fulton College of Engineering; Mechanical Engineering

Rights

https://lib.byu.edu/about/copyright/

Date Submitted

2025-07-10

Document Type

Thesis

Keywords

ISBR, Turbine Inlet, gas, temperature, measurement, combustion

Language

english

Download

Included in

Engineering Commons

Share

COinS