18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

 

Thermographic Particle Image Velocimetry in flames: current state of the technique B. Fond1*, Z. Yin2, C. Abram1, W. Meier2, I. Boxx2, and F. Beyrau1 1: Lehrstuhl für Technische Thermodynamik, Otto-von-Guericke-Universität Magdeburg, Magdeburg, Germany 2: Deutsches Zentrum für Luft- und Raumfahrt (DLR), Institut für Verbrennungstechnik, Stuttgart, Germany * Corresponding author: [email protected] Keywords: Temperature Imaging, PIV, Thermographic Phosphors, Reactive flows Session: Scalar Measurements, Combustion

ABSTRACT Optical diagnostics for the simultaneous measurement of fluid temperature and flow velocity are very valuable tools in fluid research and in particular for the study of reacting flows. A recently developed technique based on thermographic phosphor tracer particles seeded into the flow and called Thermographic Particle Image Velocimetry has shown significant promise in non-reactive flows. Since most thermographic phosphors are inert, have a high melting point, and are insensitive to the gas composition, they may be also particularly suitable for combustionrelated applications. In this study, the thermographic phosphor BAM:Eu2+ is seeded in laminar non-premixed flames, to investigate its potential and limitations for flame studies. Simultaneous temperature and seeding density measurements are presented, which show that thermal quenching of the luminescence emission reduces the luminescence signal and imposes an upper temperature limit of 900 - 1000 K, at a local seeding density of 3 1011 particles per m3. Temperature and velocity measurements are also performed in turbulent non-premixed flames to show the potential of this joint velocity-temperature diagnostic, affording high image quality at intermediate temperatures. Since the particles have a melting point above 2000 K, the second part of this study investigates whether the same particles can be used to measure after being exposed to high-temperature combustion. Phosphor particles are seeded into a pure hydrogen diffusion flame and probed 20 cm downstream, after the exhaust gases are cooled down by dilution. Simultaneous measurement of luminescence intensity, temperature and seeding density shows that the luminescence properties of the particles were unaffected by their journey through the flame. BAM:Eu2+ particles can therefore be used for measurements in exhaust gases of reciprocating engines, or in burners with exhaust gas recirculation.

1. Introduction Modern combustion devices are typically designed to burn fuel lean mixtures in highly turbulent flows in order to achieve low pollutant emissions. However what complicates their practical implementation is that these combustion modes are prone to complex instabilities originating from turbulence-chemistry interactions such as local extinction [1,2] and thermoacoustic oscillations [3,4]. A significant research effort is required to understand these

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  interactions and to provide predictive capabilities to aid the design of the next generation of efficient, safe and clean combustion devices. Among the desirable quantities to measure, the temperature-velocity pair is certainly one that offers very valuable information on these turbulence-chemistry interactions. However it is also one of the most difficult to measure. Traditional thermometry approach such as those based on molecular scattering (Raman and Rayleigh scattering) can only be applied in conjunction with Particle Image Velocimetry (PIV) at the expense of great experimental complexities, e.g. Filtered Rayleigh Scattering [5,6], since they require the efficient rejection of the elastic scattering from the PIV tracer particles. Thermographic Particle Image Velocimetry (Thermographic PIV), on the other hand, is a laser-based technique that uses thermographic phosphor tracer particles as a single seeded tracer for the simultaneous measurement of the temperature and velocity field in fluid flows. The laser light scattered by the particles is imaged in a classic PIV approach, while the temperaturedependent laser-induced luminescence emission from the same particles is spectrally filtered to determine the particle temperature using a ratiometric approach. For micrometre-size particles, it can be shown that the particle temperature and velocity rapidly match that of the surrounding gas [7]. The principle of this thermographic PIV technique has been demonstrated for timeaveraged [8] and single-shot measurements [7,9] of temperature and velocity as well as at kHz repetition rates for temporally resolved measurements [10]. The technique has several advantages: it uses simple instrumentation, including solid-state lasers and non-intensified cameras. Also thermographic phosphors are inert, have a high melting point, and are often insensitive to the gas composition [11,12] so they are particularly suitable for reactive flow studies. However, for most phosphors, the thermal quenching of the luminescence emission at elevated temperatures imposes an upper limit on the temperature range of this technique. Therefore, while this technique would be an enormously useful diagnostic tool for turbulent flame research, two questions are yet to be answered: 1. What is the maximum temperature that can be measured with this technique, and how does the technique performance degrade with temperature? 2. Since the melting point of most types of phosphor particles is often in excess of 2000 K, can the particles be used to probe in low temperature, post-combustion regions after they went through a high-temperature flame? There is almost infinite variety of phosphor materials, so the answer to these questions clearly depends upon the choice of tracer particles. The aim of this paper is to develop a methodology to answer those questions for a given tracer material. It is applied to 2.4 µm BAM:Eu2+ particles, which are now a well-characterised Thermographic PIV tracer [12] used in a

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  variety of applications [7,10,12,13-16]. Its luminescence emission was previously quantified over a wide range of conditions from 300 to 900 K. One finding particularly relevant to combustion applications was that its luminescence properties are unaffected by the presence of oxygen over the investigated range [12]. Here, the technique is applied to laminar hydrogen diffusion flames with and without N2 dilution, where comparatively shallow temperature gradients allow the clear spatial definition of the region where the technique yields temperature data. Ref [17] had showed that the luminescence signal used for thermometry increases linearly with the seeding density, with an associated improvement in the temperature precision. In answering question 1, a comparison with other studies can only be performed if the seeding density is known too, and so in this study, temperature and seeding density were measured simultaneously. With regards to the second question, several studies have reported evidence of thermal degradation occurring in BAM:Eu2+ powder samples after extended (one-hour) exposures to temperature exceeding 770 K [18,19] in oxygen containing environments. In these “heat treatment” processes, the oxidation state of the Eu2+ atoms increases to Eu3+, resulting in a sharp decrease in the room-temperature emission intensity of the blue broadband luminescence. However, in relevant combustion diagnostics test cases, e.g. flame stabilisation by recirculation, internal combustion engines, particles reside at high temperature for a much shorter time (millisecond-second timescales). Whether degradation would occur at these time scales remains to be investigated. Van Lipzig and co-workers have used BAM:Eu2+ for temperature imaging in a high pressure combustion chamber [13]. They compared the pre-combustion, room-temperature intensity ratio, with calibration curves obtained from the temperature rise of subsequent combustion cycles. Their results seem to indicate that the intensity ratio response to temperature remained unchanged. However, no information was provided on a possible decrease in emission intensity per particle cycle after cycle, which in view of the abovementioned studies [18,19] is a more suitable marker of degradation. If the effect is severe the temperature precision would be greatly reduced in post-flame regions. In this work, a pure hydrogen diffusion flame is seeded with phosphor particles and the particles are probed 20 cm downstream of the burner exit, i.e. after they went through the flame and were subsequently cooled down by the entrainment and mixing of ambient air. The luminescence intensity of the phosphor particles is compared to previous measurements performed in heated jet of air and nitrogen [12].

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  2. Experimental setup 2.1. Non-premixed jet flames An atmospheric pressure axisymmetric-coflow laminar H2/N2 diffusion flame was generated in a Gülder burner, commonly used in soot particle research [20, 21]. The burner consists of a 10.9-mm-inner-diameter tube for fuel delivery and a concentric 100-mm-diameter air co-flow. The co-flow channel was filled with glass beads and capped with a porous ceramic disk to suppress flow fluctuations. The air co-flow was maintained at 60 SLM, while various N2 dilutions in the H2 line were used to achieve different peak flame temperatures. A small amount of CH4 (0.08 SLM) was also added in the fuel stream to visualise the flame. The fuel stream was seeded with 2.4 µm (median volume-equivalent sphere diameter as measured by Coulter counter sizing technique) BAM:Eu2+ particles using a magnetic stirrer seeder. In the laminar diffusion flame cases, the coflow was not seeded. In the turbulent case, the coflow was seeded with non-luminescent TiO2 particles using a fluidised bed seeder (Pivtech PIVSolid8). 2.2. Optical setup

Fig. 1 - Experimental set-up. SO sheet optics, GW glass window, DBS dichroic beam splitter, IF interference filter.

The output of a frequency tripled, single cavity, double Q-switched Nd:YAG laser (Innolas, SplitLight 600, 5 Hz repetition rate) was formed into a 50 mm high and 0.6 mm thick light sheet in the measurement plane. The thickness of the light sheet was determined as the full width half maximum using a beam profiling camera, and a filter wheel. This light source was used for both the luminescence excitation and the Particle Image Velocimetry based on the Mie scattered light

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  from the particles at 355 nm. The fluence was estimated to be 38 and 55 mJ/cm2 for the first and second pulse respectively, using a combination of a power meter and a photodiode. The delay between the two laser pulses was set to 50 µs. The luminescence from the particles following the first laser pulse was imaged using two non-intensified CCD cameras (Imager Intense) equipped with 50 mm/#f=1.4 lenses positioned in a beamsplitter configuration. A dichroic 45° beamsplitter (455 nm longpass) in combination with two 0° interference filters (a bandpass filter with a central wavelength and bandwidth of 466 nm and 40 nm respectively and a notch filter with a central wavelength and bandwidth of 355 nm and 17.8 nm respectively) were used to separate and filter the two detection channels, as shown in Fig. 2. The choice of filters was optimised for measurements at high temperatures as described in [Fond et al. 2015 b]. The exposure was set to 5 µs, starting 1 µs before the first laser pulse, in order to capture the whole luminescence decay (1/e lifetime of BAM:Eu2+ is 1 µs at room temperature) while avoiding possible camera-laser jitter effects. The camera/beamsplitter arrangement was carefully adjusted to minimise displacement resulting from misalignment and distortion, as well as difference in the light collection path. Note that, as in Ref [12] the recording rate of the two camera was set to 10 Hz, while the laser was operated at 5 Hz, to allow background images to be evaluated from every second frame (see section 2.3.1.).

Fig. 2 BAM:Eu2+ emission spectrum as obtained from Fond et al. 2015 b. and filter curves (355 nm notch filter (red), dichroic beamsplitter (green) and 466nm bandpass filter (blue))

The laser light scattered by the particles was imaged on a sCMOS camera operated in double frame mode and positioned on the opposite side of the measurement plane, and equipped with a 355 nm bandpass filter positioned at an angle to prevent reflection of luminescence light on this filter to reach the luminescence cameras. 2.3. Post processing

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  2.3.1. Intensity Ratio/Temperature imaging Background images were first subtracted to the recorded luminescence images. The background images were determined based on non-illuminated images, recorded every second frame. As reported in Ref [12], this procedure reduces errors due to camera offset fluctuations, which are significant for the Imager Intense cameras used. Time-average images were also recorded without seeded flow for each burner position to evaluate possible flare or luminescence from deposited particles. The background-subtracted images were then corrected for linearity based on camera response determined using an Integrating sphere. A cutoff filter of 15 counts was then applied to the corrected images, to avoid establishing statistics from region with insufficient signal, followed by a 3 x 3 moving average filter, for a final in-plane resolution of 600 µm which matches the laser sheet thickness. Image pairs were divided and resulting ratio images were themselves divided by an average ratio image obtained at room temperature to account for nonuniformity in the light collection efficiency [7,12], and in the laser excitation fluence as described in Ref [22] and for normalisation. In the absence of a high temperature in-situ calibration, ratio images were then converted to temperature using a calibration curve derived from powderbased spectroscopic measurements performed in an optically accessible furnace [12]. The calibration curve was also normalised to room temperature. High-temperature gas-phase calibration is an area of further investigation. 2.3.2. Seeding density measurements The Mie scattered light from the PIV images was used to determine the seeding density. Using this approach, the local seeding density is determined in two dimensions, allowing the mapping of the intensity per particle in individual non-isothermal images. For this, the Mie scattering images were corrected for the laser intensity profile, and converted to seeding density using a factor determined at room temperature. This factor is determined the following way. The number of luminescence photons emitted by a single particle at room temperature has been measured in Ref {12] from 1 to 400 mJ/cm2. From the number of photons per particle, evaluated at 37 mJ/cm2, the conservation of radiance through the imaging optics, and the measured responsivity of the sensor (using a calibrated light source), the recorded luminescence intensity on the sensor can be converted into seeding density. By recording simultaneously the luminescence intensity at room temperature, yielding the seeding density, and the Mie scattering signal, the Mie scattering cross section of the particle is estimated. A conversion factor in number of particles/m3 per Mie scattering counts is then

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  derived at fixed laser fluence. Since the Mie scattering signal on a per particle basis is independent of temperature, this factor can be used to convert directly Mie scattering signal into seeding density at higher temperatures. 2.3.3. Estimation of the intensity per particle at elevated temperature. For determination of the intensity per particle, the background subtracted luminescence images were corrected for non-uniformity in the laser fluence. For this, a vertical profile was derived from time-averaged luminescence images recorded in large stream of air seeded with particles where the seeding density is assumed to be uniform. The corrected luminescence images were then divided by the seeding density field to yield the luminescence intensity on a per particle basis. When necessary for the evaluation of the quantities of interest, additional details regarding to spatial and time averaging are provided in the respective Results sections. 3. Results and discussion 3.1. Laminar non-premixed flames.

Fig. 3 Single-shot measurements in a 33% H2 / 66 % N2 diffusion flame (U=0.7 m/s). Measured temperature (left), luminescence intensity (centre) and seeding density (right) fields. The adiabatic flame temperature is 1739 K.

A single shot temperature image obtained in a nitrogen diluted diffusion (34% H2, 66% N2) flame is shown on the left in Figure 3. This image shows that the technique captures the broad

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  gradient expected from a diluted laminar diffusion flame well. This allows the characterisation of the maximum temperature that can be measured in a practical application. White areas correspond to locations where temperature measurements were not possible due to the lack of luminescence signal for thermometry. For this flame, the adiabatic flame temperature is estimated to be 1739 K. A map of measured measured luminescence intensity and seeding density are also presented in Figure 3 in order to gain insight into the temperature limit in this configuration. It shows that, as the temperature increases with distance from the burner exit, the intensity per particle drops. The seeding density is also reduced as the gas density decreases, but it is clear that the thermal quenching is the dominating effect. In other words, particles are present in the flow, but they do not emit sufficient light to be detected for thermometry.

Fig. 4 Single-shot single row temperature, luminescence intensity and seeding density profiles obtained at 18 mm from the burner exit. The points were sampled to reflect the actual spatial resolution of the measurements (600 µm)

As an example of quantitative analysis, a set of single-shot single-row radial profiles was extracted at a position 18 mm downstream of burner exit (Fig. 4). Here only statistically independent measurement points are plotted, therefore reflecting the actual spatial resolution of the measurements. It is important that the temperature gradient of the flame is captured with sufficient spatial resolution to better resolve the temperature limit but also to prevent averaging over quantities which have a non-linear dependence on temperature, e.g. intensity ratio and luminescence intensity. In this respect, this nitrogen diluted hydrogen/air laminar diffusion flame offers a significant advantage over premixed flames. The maximum temperature

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  measured at this distance downstream is 900 K, corresponding to the location x = - 2.5 mm. In the next measurement location to the left there is not sufficient luminescence intensity to measure the temperature, at a seeding density of 3x1011 particles/m3. We conclude that the temperature limit using BAM:Eu2+ is 900 K at this seeding level. It should be noted that in situations where there are significant differences in luminescence intensity, as the case here because of thermal quenching, it is possible that multiple scattering of light between particles can interfere with the measurements. This is predominantly the case where, for example, a surrounding large-diameter coflow is seeded with particles, which at room temperature emit a lot of light. Here, in order to avoid any possibility of multiple scattering to interfere with the measurements, the coflow stream was not seeded with particles. For this reason, the seeding density field no longer reflects an inverse of the temperature field as it also decreases by dilution with the unseeded coflow stream. 3.2. Velocity and temperature imaging in a turbulent non-premixed flame Simultaneous temperature and velocity measurements were acquired in the same Gülder burner but operating under turbulent conditions, as shown in Figure 5. In addition, the coflow stream is seeded with non-luminescence TiO2 particles. These single shot images shows the high level of high image quality afforded by this technique in low-medium temperature turbulent flows. The white areas are zones where either no thermographic phosphor particles are present, or the temperature is too hot for sufficient luminescence light to be emitted, therefore no measurements are obtained. The presence of sharp transitions from cold zones to “no measurement” zones in some parts of the images can be explained by the absence of transverse transport of the particles while hydrogen molecules are transported by molecular diffusion. While seeding the co-flowing stream with phosphor particles could help removing this ambiguity, light emitted from cold particles in the co-flow may in this case be “re-imaged” in the hot temperature regions by multiple scattering (as described above) and lead to the incorrect interpretation of the data.

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

 

Fig. 5 Single shot temperature and velocity images obtained in a 24% H2 / 76 % N2 turbulent diffusion flame (U = 7 m/s). Note that the time-averaged velocity field was subtracted to the instantaneous velocity fields. The central jet was seeded with thermographic phosphor particles, while the coflow was seeded with non-luminescent TiO2 particles.

3.3. Post-flame investigations The intensity per particle was measured in the post-flame region of a pure hydrogen flame to investigate the survivability of the particles after experiencing the high temperature region of the flame. For this, the burner was translated downwards, so that the field of view was located 20 cm above the burner exit. At this location, small regions (streaks) with a moderate seeding density could be observed which seem to correspond to regions where exhaust from the seeded flame intersects the measurement plane. Some light at these locations was detected in the luminescence cameras. The absence of signal in background images recorded after each frame rules out that this signal may be flame chemiluminescence, particle blackbody radiation or flame-induced particle luminescence. For each streak, the luminescence intensity, temperature and seeding density were evaluated simultaneously. Statistics were then compiled based on about 150 streaks to evaluate the mean intensity per particle and the mean temperature. The result is shown in Fig. 6. The uncertainty bars are evaluated as the standard deviation divided by the square root of the number of

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  samples. Note while there is little variation in the measured temperature, the intensity per particle varies significantly from streak to streak. This may be due to variations in the Mie scattering signal from non-spherical particles due to changes in their orientation. The particles aligning with the flow in a position of maximum drag, their orientation will follow flow direction changes. This is a disadvantage of deriving the seeding density from the Mie scattering intensity as opposed to deriving it from the particle counting technique described in [17], which counts the number of particles in one image to derive the seeding density. However, 2-D mapping of the seeding density was necessary in this study. This result shows that the intensity of these post-flame particles at about 500 K correlates well with measurements performed on particles, which have not encountered high temperature regions (as in medium temperature heated jets). Were these particles degraded in the high temperature region, their luminescence would be dimmer according to existing studies on thermal degradation of this material. This finding indicates that the particles can be used for thermometry in cooled post-flame gases. This is a significant advantage over organic LIF tracers which would be burnt, thus, phosphor particles can aid the study of burners with recirculation or of exhaust gas recirculation in an engine.

Fig.  6  Intensity per particle as measured in a heated jet (from [12]) and downstream of a pure hydrogen laminar diffusion flame (this study).  

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  5. Conclusions In this study a method was developed to establish the performance and limitations of the thermographic particle image velocity technique in reactive flow applications for a given thermographic phosphor tracer material. Highly diluted non-premixed hydrogen/air jet flames stabilised on a Gülder burner offered a very gradual temperature increase across the flame, which was captured by the temperature measurements. Simultaneous seeding density measurements show that while particles are still present, their luminescence signal is no longer sufficient for measurements above 900-1000 K at a local seeding density of 3 x 1011 particles/m3. Joint-temperature and velocity measurements are also presented in a turbulent diffusion flame case, as an illustration of the current state of the technique for turbulent reacting flow research. The diagnostics offer high image quality but with a limited temperature range. In addition, particles were probed far downstream of a non-diluted hydrogen diffusion flame in order to study their survivability in high temperature flames. It is shown that the signal from these particles actually reappears after the burnt gases are cooled, and that their luminescence intensity remains unaffected by their journey through the flame. This feature offers a unique capability for measurements in burners and engines with recirculation of burnt gases. There is a wide range of phosphor materials and the vast majority of them have not yet been studied. The limit identified here will now serve as a reference point and these tests as a reference method for the future development of the technique using other phosphor materials, which potentially have a delayed onset of thermal quenching and can be used to measure at higher temperatures. References [1] M. Stöhr, I. Boxx, C. Carter and W. Meier, “Dynamics of lean blowout of a swirl-stabilised flame in a gas turbine model combustor,” Proc. Combust. Inst., 33 (2011) 2953-2960 [2] M. Stöhr, C.M. Arndt and W. Meier, “Transient effects of fuel-air mixing in a partiallypremixed turbulent swirl flame,”Proc. Combust. Inst., 35 (2015) 3327-3335 [3] W. Meier, P. Weigand, X.R. Duan, and R. Giezendanner-Thoben, “Detailed characterization of the dynamics of thermoacoustic pulsations in a lean premixed swirl flame,” Combust. Flame, 150 (2007) 2-26 [4] I. Boxx,, C. Arndt, C. Carter, W. Meier, “High-speed laser diagnostics for the study of flame dynamics in a lean premixed gas turbine model combustor.” Experiments in fluids 52.3 (2012): 555-567 [5] J. N. Forkey, N. D. Finkelstein, W. R. Lempert, and R. B. Miles, “Demonstration and characterization of filtered Rayleigh scattering for planar velocity measurements,” AIAA J. 34(3), 442–448 (1996). [6] D. Most and A. Leipertz, “Simultaneous two-dimensional flow velocity and gas temperature measurements by use of a combined particle image velocimetry and filtered Rayleigh scattering

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

  technique,” Appl. Opt. 40(30), 5379–5387 (2001). [7] B. Fond, C. Abram, A.L. Heyes, A.M. Kempf, F. Beyrau, Simultaneous temperature, mixture fraction and velocity imaging in turbulent flows using thermographic phosphor tracer particles. Opt. Express 20, 22118–22133 (2012) [8] A. Omrane, P. Petersson, M. Alden, M.A. Linne, Simultaneous 2D flow velocity and gas temperature measurements using thermographic phosphors. Appl. Phys. B 92, 99–102 (2008) [9] N.J. Neal, J. Jordan, D. Rothamer, Simultaneous measurements of in-cylinder temperature and velocity distribution in a small- bore diesel engine using thermographic phosphors. SAE Int. J. Engines 6, 300–318 (2013) [10] C. Abram, B. Fond, A.L. Heyes, F. Beyrau, High-speed planar thermometry and velocimetry using thermographic phosphor particles. Appl. Phys. B 111, 155–160 (2013) [11] J. Brubach, A. Dreizler, J. Janicka, Gas compositional and pres- sure effects on thermographic phosphor thermometry. Meas. Sci. Technol. 18, 764–770 (2007) [12] B. Fond, C. Abram and Frank Beyrau, Characterisation of the luminescence properties of BAM:Eu2+ as a tracer for Thermographic Particle Image Velocimetry, Appl. Phys. B 121, 495-509 (2015) [13] J.P.J. van Lipzig, M. Yu, N.J. Dam, C.C.M. Luijten, L.P.H. de Goey, Gas-phase thermometry in a high-pressure cell using BaMgAl10O17: Eu as a thermographic phosphor. Appl. Phys. B 111, 469–481 (2013) [14] M. Lawrence, H. Zhao, L. Ganippa, Gas phase thermometry of hot turbulent jets using laser induced phosphorescence. Opt. Express 21, 12260–12281 (2013) [15] A.O. Ojo, B. Fond, B.V. Wachem, A.L. Heyes, F. Beyrau, Thermographic laser doppler velocimetry. Opt. Lett. 40, 4759-4762 (2015) [16] P. Schreivogel, C. Abram, B. Fond, M. Straußwald, F. Beyrau and M. Pfitzner, "Simultaneous kHz-rate temperature and velocity field measurements in the flow emanating from angled and trenched film cooling holes," submitted to International Journal of Heat and Mass Transfer, 2016. [17] B. Fond, C. Abram, F. Beyrau, On the characterisation of tracer particles for thermographic particle image velocimetry. Appl. Phys. B 118, 393–399 (2015) [18] Y.H. Wang, Z.H. Zhang, Luminescence thermal degradation mechanism in BaMgAl10O17: Eu2+ phosphor. Electrochem Solid State 8, H97–H99 (2005) [19] G. Bizarri, B. Moine, On BaMgAl10O17:Eu2+ phosphor degradation mechanism: thermal treatment effects. J. Lumin. 113, 199–213 (2005) [20] Schulz, C., Kock, B. F., Hofmann, M., Michelsen, H., Will, S., Bougie, B., ... & Smallwood, G. (2006). Laser-induced incandescence: recent trends and current questions. Applied Physics B, 83(3), 333-354 [21] Hadef, R., Geigle, K. P., Zerbs, J., Sawchuk, R. A., & Snelling, D. R. (2013). The concept of 2D gated imaging for particle sizing in a laminar diffusion flame. Applied Physics B, 112(3), 395-408. [22] C. Abram, B. Fond, F. Beyrau, High-precision flow temperature imaging using ZnO thermographic phosphor tracer particles. Opt. Express 23, 19453–19468 (2015)