Christoph Knappe

Limitations of ICCD detectors and optimized 2D phosphor thermometry

This paper presents investigations on the limitations of intensified CCD (ICCD) cameras when performing quantitative 2D temperature measurements using thermographic phosphors. The temperature is evaluated using a two-color ratio technique, which demands high signal linearity of the detector. However, this work shows that the ICCD cameras used suffer from nonlinearity at levels of irradiance that do not necessarily produce particularly high count levels, i.e. well below the saturation limit of the A/D converter. For the phosphor thermometry, this results in false absolute temperature readings and decreased measurement precision. The results show that there is a useful workspace in the irradiance/camera gain combination where the detector should preferably be used. A response function for the ICCD camera used is constructed. Furthermore, in order to perform quantitative measurements, there is an upper limit of counts that should not be exceeded. This limit must be investigated for each experimental configuration since it depends on the characteristics of the specific light source and also varies between different cameras.

Investigation and compensation of the nonlinear response in photomultiplier tubes for quantitative single-shot measurements.

A concept for time-sensitive optical detectors is described that shows how to confirm whether the detection device is operating in the linear response regime. By evaluating the recorded time decay of a thermographic phosphor, even weak saturation effects far from obvious situations can be identified and further related to either optical or electrical saturation. The concept has been validated by running a PMT detector close to saturation and exposing it to the optical signal decay of two different thermographic phosphors, La(2)O(2)S:Eu and CdWO(4). It was confirmed that short but intense light exposures at the beginning of an individual time decay influence the detector response for the rest of the decaying signal including temporal areas, where the anode current has dropped well below the manufacturer specified current limit. Such situations are common when applying, e.g., phosphor thermometry where it is necessary to retrieve the full decay curve from a single-shot event, i.e., standard techniques based on single-photon counting are omitted. Finally, means of compensation are introduced in order to facilitate the retrieval of useful information from the measurement data when operation in the non-linear response regime is inevitable.

Comparison of photo detectors and operating conditions for decay time determination in phosphor thermometry

This work compares the extent of linear response regions from standard time-resolving optical detectors for phosphor thermometry. Different types of photomultipliers (ordinary and time-gated) as well as an avalanche photodiode were tested and compared using the phosphorescence decay time of cadmium tungstate (CdWO(4)). Effects originating from incipient detector saturation are revealed as a change in evaluated phosphorescence decay time, which was found to be a more sensitive measure for saturation than the conventional signal strength comparison between in- and output. Since the decay time of thermographic phosphors is used for temperature determination systematic temperature errors in the order of several tens of Kelvins may be introduced. Saturation from the initial intensity is isolated from temporally developed saturation by varying the CdWO(4) decay time over the microsecond to nanosecond range, resultant of varying the temperature from 290 to 580 K. A detector mapping procedure is developed in order to identify linear response regions where the decay-to-temperature evaluations are unbiased. In addition, this mapping procedure generates a library of the degree of distortion for operating points outside of linear response regions. Signals collected in the partly saturated regime can thus be corrected to their unbiased value using this library, extending the usable detector operating range significantly.

Precision in 2D temperature measurements using the thermographic phosphor BAM

Investigation of optimized spatial precision for surface temperature measurements is performed. The temperature is measured by means of two-color ratio imaging with ICCD cameras, using the thermographic phosphor BAM. The precision in temperature is put in relation to the spatial resolution, two quantities which involve a trade-off in this case: the more spatial smoothing the better precision, but also the worse spatial resolution. Two different setups are used in order to investigate the influence of laser shot-to-shot variations, the flat-field correction and image registration process on the precision. In order to achieve high precision it is crucial to operate the ICCD cameras with a gain setting that does not introduce nonlinearity effects at the present level of irradiance. The results provide guidance on the precision to be expected from surface temperature measurements using the two-color ratio technique in combination with thermographic phosphors and also confirm the importance of highly stable and linear ICCD detectors. At room temperature and low spatial resolution the precision is evaluated to 0.4%.

Development of an automatic routine for calibration of thermographic phosphors

An automated routine for the continuous calibration of thermographic phosphors was developed as a replacement for the conventional calibration scheme that relied on fixed temperature points. The automated calibration routine was validated using Mg3F2GeO4:Mn as a calibration phosphor. Hardware and software aspects of the calibration process were addressed in this development. The hardware aspect included a new substrate design using a high performance alloy, the Hastelloy-C alloy, whereas the software aspect included an automated acquisition system which was capable of acquiring simultaneous thermocouple temperatures and phosphor decay waveform in real time. The design of the calibration process eliminates the need for a system in thermal equilibrium during a phosphor calibration measurement. Temperature ramping rates of up to 4 K min −1 were employed in the oven without a delay in the temperature response being measured between the phosphor and the thermocouples involved. In addition, the automated calibration setup allowed for detailed investigations on the effect of heat being delivered to the phosphor coating by the laser. These findings were confirmed by a simple heat transfer model, based on lumped system analysis. In comparison to the data acquisition performed at several fixed points with the conventional calibration scheme, the experiment duration was shortened by a factor of 4 with the overall accuracy improved by 1-2 K.

A first application of thermographic phosphors in a marine two-stroke diesel engine for surface temperature measurement

Phosphor thermometry is applied for the first time in a large-bore two-stroke diesel engine. The work proves the practicality of phosphor thermometry in large-bore engines. The experiments were conducted on the MAN 4T50ME-X marine research engine equipped with an optical cylinder head. By employing a thin surface coating of CdWO4 phosphor, cycle resolved temperature measurements of the cylinder wall were obtained. Motored and fired engine operations were tested at engine loads covering the low and medium engine load range. Phosphor thermometry proved to be successful in retrieving the temperature with standard deviations ranging around 1-8 K. Experimental considerations like detector linearity, coating thickness and an automated phosphor calibration routine will be addressed. (Less)

Enhanced color ratio calibration approach for two-dimensional surface thermometry using laser-induced phosphorescence

An alternative concept for calibrating spectral image intensity ratios is described, which provides simple, but effective means of compensation for systematic errors, caused by nonlinearities in signal detection. The method relies on image segmentation by means of signal thresholds, where pixels are organized into different subgroups according to their corresponding signal count value. Instead of defining one global spectral ratio per calibration temperature, the phosphor-coated target surface is illuminated inhomogeneously and resulting image ratios are calibrated individually for each pixel intensity subgroup. This allows the exploitation of high intensity regions on the camera chip which offer great precision advantages, but suffer from systematic errors caused by signal nonlinearities. Temperature calibration data of BaMg2Al16O27:Eu in the temperature range between 270 and 470 K was used to assess and compare the potential of both calibration approaches. In comparison to the conventional, e.g. global calibration approach, accuracy improvements of up to 39% were gained even while keeping average signal intensities below 15% of the detector’s full dynamic range. Image ratio evaluations, based on segregated pixel subgroups, could help improve measurement accuracy also for other techniques, relying on the calibration of measured quantities. In two-dimensional phosphor thermometry, it helps bridge the current precision gap between two-color ratio methods and more elaborate lifetime-imaging approaches. (Less)

Response Regime Studies on Standard Detectors for Decay Time Determination in Phosphor Thermometry

This work compares the extent of linear response regimes from standard time-resolving optical detectors for phosphor thermometry. Different types of Photomultipliers (ordinary and time-gated) as well as an Avalanche Photodiode are tested and compared using the phosphorescent time decay of CdWO4 that ranges from 10 mu s down to a few ns within a temperature span of 290 to 580 K. Effects originating from incipient detector saturation, far from obvious to the operators eye, are revealed as a change in evaluated phosphorescence decay time. Since the decay time of thermographic phosphors itself is used for temperature determination - systematic temperature errors up to several tens of Kelvins may be introduced by such detector saturation. A detector mapping procedure is suggested in order to identify linear response regions where the decay-to-temperature evaluation can be performed unbiased. Generation of such a library is highly recommended prior to any quantitative measurement attempt. Using this detector library, even signals collected in the partly saturated regime can be corrected to their unbiased value extending the usable detector operating range significantly. Further, the use of an external current-to-voltage amplifier proved useful for most applications in time-based phosphor thermometry helping to limit saturation effects whilst maintaining a reasonable bandwidth and signal outputs.