Development of A Multi-Spectral Pyrometry Sensor for High-Speed Transient Surface-Temperature Measurements in Combustion-Relevant Harsh Environments

Abstract

Accurate and high-speed transient surface-temperature measurements of combustion devices including internal combustion (IC) engines, gas turbines, etc., provide validation targets and boundary conditions for computational fluid dynamics models, and are broadly relevant to technology advancements such as performance improvement and emissions reduction. Development and demonstration of a multi-infrared-channel pyrometry-based optical instrument for high-speed surface-temperature measurement is described. The measurement principle is based on multi-spectral radiation thermometry (MRT) and uses surface thermal radiation at four discrete spectral regions and a corresponding emissivity model to obtain surface temperature via non-linear least squares (NLLS) optimization. Rules of thumb for specifying the spectral regions and considerations to avoid interference with common combustion products are developed; the impact of these along with linear and non-linear MRT analysis are assessed as a function of temperature and signal-to-noise ratio. A multi-start method to determine the MRT-solution global optimum is described and demonstrated. The resulting multi-channel transient pyrometry instrument is described along with practical considerations including optical-alignment drift, matching intra-channel transient response, and solution-confidence indicators. The instrument demonstrated excellent >97% accuracy and >99% 2-sigma precision over the 400−800 °C range, with ~20 µs (50 kHz, equivalent to 0.2 cad at 2000 RPM IC-engine operation) transient response in the bench validation.

Keywords: combustion; multi-spectral radiation thermometry (MRT); pyrometer; surface temperature.

Figure 1

Measured emissivity (dotted curves) of a stainless-steel IC-engine valve and fits (solid curves) using a first-order (m = 1) (a) exponential and (b) polynomial emissivity model.

Figure 2

Wavelength parameters including number (n), spectral width or bandwidth (Δλ_j), and adjacent spacing (λ_j–λ_j−1), where j: 1 to n and total spectral range (λ_n–λ₁) of the different regions must be selected and influence MRT analysis performance.

Figure 3

Additional spectral emission from a +20 °C surface temperature change in (a) relatively low-T (300–600 °C) and (b) high-T (600–900 °C) ranges. Calculated via Planck’s law.

Figure 4

Absorbance of major combustion products and spectral regions defined by available bandpass filters to avoid related interference in the pyrometry instrument. The C₂H₄ region is generally representative of hydrocarbon fuels.

Figure 5

Pyrometry instrument setup picture and schematic detailing part numbers.

Figure 6

Solution optimization schematic for NLLS MRT.

Figure 7

Schematic of laboratory bench setup for measuring surface temperatures of an LNF exhaust valve, and optical-chopper assembly for creating synthetic surface-emission transients to assess transient response of the pyrometry instrument. The furnace assembly houses the valve sample in an N₂-purged quartz tube; insulation (gray) shields the sample and optical probe from the furnace heating elements (red); a thermocouple measures environment temperature near where the optical probe captures surface emission. The chopper assembly is inserted in the optical path between the probe and pyrometry instrument.

Figure 8

(a) Synthetic 2 kHz surface-emission transient from the LNF-valve sample at 716 °C using the chopper assembly shown in Figure 7. Individual signal transient from the four spectral channels and related transient MRT analysis results. Dashed lines are the signals at a steady state (without chopper). (b) Individual signal transient from the four spectral channels, with the slow 2100 nm channel replaced by a scaled version of the fast 1500 nm channel signal transient, and related transient MRT analysis results.