Formulation and Experimental Verification of Alternate Calibration Techniques of the Temperature Dependent Response of Phosphor Fluorescence

The overall goal of this work is to provide an alternative approach to the thermographic phosphor (TP) time constant calibration method for temperature recovery. In this work two techniques are proposed that retain the pulsed source input used in the standard TP time constant calibration approach but reinterpret the phosphor response taken a fixed distance such that the single-exponential decay assumption is removed. The methods do not require knowledge of key parameters prior to data processing, nor do they involve complicated numerical schemes that attempt to fit data in the low signal-to-noise region of the phosphor response. The approaches do involve integrating the full phosphor response signal to arrive at a single value related to the integrated intensity trajectory. This value can be calibrated to temperature. One method uses the slope of the integrated intensity in the rise portion of the phosphor emission as the calibration parameter, while the second technique uses the total integral of the emission as the calibration parameter.

Both techniques are validated as an effective means of TP calibration by experimental data. First, the phosphor emission response is recorded at different steady-state temperatures in order to form a calibration curve. Different regression models are investigated to determine the functional relationship that best fits the observed calibration data. Second, the phosphor is heated under transient conditions and both calibration techniques are applied to resolve the temporal temperature history of the test sample. From the experimental results, it is found that a bi-exponential based calibration curve or a rational function based calibration curve accurately predict the temperature measurements of the transient tests for both calibration procedures. However, it is suggested that the total integrated intensity method is more reliable compared to the slope calibration method since smaller error estimates are observed using the total integral in the transient sense. Another attractive feature of the integral method is that the only numerical manipulation of the raw physical experimental data to resolve the calibration parameter involves integrating the signal. The outcome of this work is highly encouraging and indicates that these techniques could be found useful in certain applications.