On the automation of thermographic phosphor calibration

Författare

Fahd Jouda Abou Nada
Christoph Knappe
Xin Xu
Mattias Richter
Marcus Aldén

Summary, in English

Thermographic phosphors can be robust temperature remote sensors. The accuracy of the temperature measured by the phosphor is highly dependent on the quality of the phosphor calibration used. Conventionally, thermographic phosphors are calibrated by measuring a series of decay curves at known stable oven temperatures. The process is then repeated covering the thermal sensitivity range of the phosphor chosen. Heating and cooling rates of high temperature ovens are usually low. Also, thermal equilibrium of the system is required at each calibration temperature before acquiring luminescence decay curves. Thus, the process is usually time consuming and the number of calibration points achieved is limited to a couple of dozen points. This study presents and validates the development of an automatic routine for the calibration of thermographic phosphors. It was designed to continuously and simultaneously acquire phosphor decay curves along with their corresponding thermocouple temperatures. The developed routine required software and hardware improvements. An updated design of the calibration substrate was implemented to improve the thermal conditions during calibration. Thermal gradients were further studied using a heat transfer model. The routine implemented a specially designed sparsing algorithm that reduced the sampling rate of the decaying luminescence curve without influencing the calculated decay time. The upper heating rate is set at 4 K.min-1 due to limitation imposed by the ceramic calibration oven. The phosphors CdWO4 and Mg3F2GeO4:Mn were chosen to validate the finalized routine. After the completion of the calibration process, a library-based calibration is created as the final product. The automated calibration routine delivered an overall accuracy improvement of 1-2 K, reduced calibration duration by factor of four and provided the possibility of deriving signal recognition algorithms. The condensed calibration dataset produced by the proposed calibration routine was further employed to develop a novel signal shape recognition algorithm for temperature evaluation.

Avdelning/ar