two-camera dual-band collection toluene plif thermometry...

16th Int Symp on Applications of Laser Techniques to Fluid Mechanics Lisbon, Portugal, 09-12 July, 2012

- 1 -

Two-Camera Dual-Band Collection Toluene PLIF Thermometry in

Supersonic Flows

Victor A. Miller1*, Mirko Gamba1, M. Godfrey Mungal1, 2, and

Ronald K. Hanson1

1: Department of Mechanical Engineering, Stanford University, Stanford, California, United States

2: School of Engineering, Santa Clara University, Santa Clara, California, United States * correspondent author: [email protected]

Abstract Toluene PLIF has been applied to image temperature in supersonic flows generated in an expansion tube. Toluene has high fluorescence quantum yield and high temperature sensitivity compared to similar aromatic tracers, and these qualities of toluene make it ideal for imaging supersonic flows where large temperature gradients can exist. Toluene fluorescence exhibits a red shift with increasing temperature, enabling temperature imaging through a single-excitation wavelength, two-camera imaging strategy. A single pulse of 266nm laser light excites the tracer, and two cameras with different optical filters simultaneously image different spectral regions of toluene fluorescence; the ratio of these images is converted to temperature. In this work, the design of the diagnostic is described, the diagnostic is validated through imaging of flow behind incident shockwaves, and the technique is demonstrated on two canonical flow fields, M = 2.3 flow of toluene-seeded nitrogen over a wedge and cylinder.

1. Introduction This work presents the application of two-camera dual-band collection toluene planar laser-induced fluorescence (PLIF) temperature imaging in an expansion tube flow facility. Toluene exhibits a high fluorescence quantum yield in oxygen-free environments compared to similar aromatic tracers, such as 3-pentanone or acetone, and toluene fluorescence is exceptionally sensitive to temperature. These qualities make toluene an ideal tracer for imaging compressible flow fields in which large temperature gradients may exist. Furthermore, it has been shown that toluene fluorescence has a red shift with increasing temperature [Koban et al., 2004], enabling temperature imaging in flow fields with non-uniform pressure or tracer seeding via a two-camera, single-wavelength excitation imaging scheme, for which each camera images a different spectral region of the fluorescence, and the ratio of these images is converted to temperature.

In this work, an expansion tube is used to generate supersonic flow fields. An expansion tube is an impulse facility typically used to generate moderate- to high-enthalpy flow. An expansion tube allows freedom of choice over the free stream test gas, and also allows for the seeding of tracers into the test gas. In this work, the expansion tube is operated at low- to moderate-enthalpy conditions, such that toluene does not pyrolize and it emits sufficient fluorescence for imaging. Toluene (0.5% by volume) is seeded into nitrogen test gas, a 266nm pulsed light source is used to acquire single-shot images, and intensified CCD cameras are used to collect fluorescence over different spectral regions.

This work is organized as follows: first, a description of toluene spectroscopy and design of the


- 2 -

two-camera imaging strategy are presented. Then, a description of the expansion tube flow facility and test conditions are provided. The thermometry strategy is validated by imaging flow behind incident shockwaves, where temperature is inferred from measured shock speeds. Finally, two canonical flow fields are imaged: supersonic flow over a wedge and cylinder. A wedge offers both a uniform-pressure field upstream and downstream of an attached oblique shock, as well as a well-known solution across the shock. The cylinder offers variable temperature and pressure fields downstream of a bow shock, enabling further evaluation of the technique in a controlled but complex flow field, and to assess its suitability in more general flow configurations.

2. Diagnostic design 2.1 Toluene spectroscopy Electronic transitions in toluene exhibit broadband absorption and fluorescence. In the weak excitation limit, fluorescence signal Sf collected by an imaging system is modeled by

S f =Ehυ

PtolRT

σ (λex,T )φ(λ,λex,T,Ptol,P)η , (1)

which is the product of the incident number of photons Ehν

, where E is the fluence of the excitation

light, h is Planck’s constant, and ν is the frequency of excitation light, the fraction of photons

absorbed PtolRT

σ , which is the number density of the fluorescing species multiplied by the

absorption cross section σ, the fraction of photons reemitted as fluorescence φ, the fluorescence quantum yield (FQY), and an efficiency or response of the imaging system η. The absorption cross section and FQY are both functions of temperature T and excitation wavelength λex, and the FQY is a function of wavelength of fluorescence and both partial pressure of the fluorescing species Pi and total pressure P.

Koban et al. [2004] have characterized toluene absorption cross section and FQY from 298 K to

300 400 500 600 700 800 9001

2

3

4

5

6

7

Temperature, K

Abs

orpt

ion

cros

s sec

tion m

*1019

,cm2

266 nm248 nm

300 400 500 600 700 800 900

10ï3

10ï2

10ï1

100

Temperature, K

Rel

ativ

e flu

ores

ence

qua

ntum

yie

ld, q

(T)/q

(298

)

266 nm248 nm

(a) (b) Fig. 1: (a) Toluene absorption cross section as a function of temperature and excitation wavelength; (b) integrated relative fluorescence quantum yield as a function of temperature and excitation wavelength


- 3 -

1000 K. Data were collected for both 248nm excitation via a KrF excimer laser, and at 266nm via a frequency quadrupled Nd:YAG laser. The absorption cross section for 266nm excitation increases with temperature, from nearly 2 × 10−19cm2 to 6.5 × 10−19cm2 from 298K to 900K. For 248nm excitation, the absorption cross section is nearly constant at 3.1 × 10−19cm2 (Figure 1a). Koch et al. also measured FQY as a function of temperature and fluorescence wavelength λ for both 248nm and 266nm excitation. The integrated FQY

φ dλ∫ ��for both 248nm and 266nm decreases

exponentially as a function of temperature, dropping nearly three orders of magnitude from 300 to 900K (Figure 1b). Absolute FQY of toluene at 298K and 1 bar for 266nm and 248nm excitation is 0.19 and 0.036, respectively [Cheung, 2011]. With increasing temperature, the fluorescence spectrum also shifts to longer wavelengths for both 266nm excitation (Figure 2a) and 248nm excitation [Koch et al., 2004].

To maximize fluorescence signal level, we only consider 266nm excitation in this work due to both the larger FQY and larger absorption cross section of toluene with 266nm excitation. Furthermore, it has been shown (Yoo et al. [2010], Cheung [2011]) that toluene fluorescence quantum yield is nearly independent of pressure above 0.8 bar of total pressure in an oxygen-free environment. Under this condition, the fluorescence signal Sf for toluene at conditions in this work is a function of toluene number density, fluorescence wavelength, and temperature

S f (λ,Ptol,T ) =PtolRT

σ (T )φ(λ,T )η(λ) . (2)

2.2 Two-camera thermometry Toluene PLIF thermometry using a two-camera, single-wavelength excitation scheme was first proposed by Koban et al. [2004]. As mentioned, toluene fluorescence exhibits a redshift at increasing temperature. Temperature can be inferred by imaging the spectral region insensitive to temperature (i.e., near 280nm), and the spectral region sensitive to temperature (i.e., wavelengths greater than ∼ 305nm). The ratio of signals from these two bands changes with increasing temperature, and in this way, temperature can be imaged. This technique has been demonstrated through bench-top imaging of a heated plate by Cundy et al. [2011]. Toluene fluorescence also exhibits a similar red shift with increasing oxygen partial pressure at constant temperature, enabling imaging of oxygen mole fraction; this has been demonstrated in a free jet by Mohri et al. [2011]. The present work has been conducted in an oxygen-free environment, so oxygen quenching is not an issue.

For dual-band thermometry, two cameras are positioned to image the same region of interest, and optical filters are placed in front of each camera to pass only a specific band of fluorescence. Therefore, the signal collected by each camera Sf1 and Sf2 is

S fi (λ,Ptol,T ) =PtolRT

σ (T ) Fi (λ)φ(λ,T )ηi(λ)dλ∫ , (3)

where Fi(λ) is the transmission curve for each spectral filter, and ηi is the response of each imaging system. Neglecting the spectral response of each camera, the fluorescence signal at each pixel is

only a function of local number density PtolRT

, temperature T, and the spectral filter Fi.


- 4 -

By taking the ratio of two images, where each pixel represents the same volume in space and

therefore has the same number density and temperature, the ratio of signals Sf 1Sf 2

is

Sf 1Sf 2

=F1(λ)φ(λ,T )η1(λ)dλ∫F2 (λ)φ(λ,T )η2 (λ)dλ∫

, (4)

in which the only unknowns are the temperature and the response of each imaging system. Imaging system response parameters that are independent of wavelength, such as vignetting or CCD spatial uniformity, can be removed from the integral; however, wavelength-dependent aspects, such as the quantum efficiency of the ICCD camera, cannot be cancelled in a ratio. The absolute quantum efficiency of each camera varies by approximately 5% over the spectral regions imaged, and if taken to be constant, it can be removed from the integral.

Then, for a given set of imaging settings (gain, aperture, etc.), the flow field images are normalized by a set of images at known temperature, which both calibrates the imaging system and eliminates the dependence of each systems response relative to each other, provided all imaging system response terms can be removed from the integral.

Sf 1(T )Sf 1(298)

Sf 2 (T )Sf 2 (298)

=

F1(λ)φ(λ,T )dλ∫F1(λ)φ(λ, 298)dλ∫

F2 (λ)φ(λ,T )dλ∫F2 (λ)φ(λ, 298)dλ∫

(5)

240 260 280 300 320 340 3600

0.2

0.4

0.6

0.8

1

Wavelength !, nm

Norm

alize

d LIF

sign

al, a.

u.

BP280WG305

298 K

669 K496 K

372 K

300 400 500 600 700 800 9000

0.2

0.4

0.6

0.8

1

T, K

S f1/S

f1

WG305WG320WG335

(a) (b) Fig. 2: (a) Toluene fluorescence spectra as a function of temperature at 266nm excitation. Data courtesy of Koban et al. (2004); (b) expected signal ratios as a function of temperature for dual-band thermometry, for a BP280 filter paired with a WG305 (yellow), WG320 (orange), and WG335 (red) filter. Data points correspond to original spectroscopic data provided by Koban et al., and solid curves correspond to third-order polynomial fits.


- 5 -

2.3 Filter combination performance Selection of imaging filters was carried out to maximize temperature sensitivity and measurement accuracy and precision, defined in terms of absolute signal level and signal-to-noise ratio, SNR. Temperature sensitivity of equations 5 is determined by the filters selected to isolate the temperature-sensitive and -insensitive bands of fluorescence. Toluene fluoresces between 265 and 340nm, with relatively more fluorescence at higher temperatures from 305 to 340nm compared to 260 to 305nm (figure 2a). Twelve filters were considered for this work, yielding 132 different combinations; filters considered were seven high-transmission (> 80%) interference band-pass filters (denoted as “BP###” filters, where ### is the center wavelength), four Schott Glass long-pass filters (WG###, where ### is near the cutoff wavelength), and one Schott glass UG11 short-pass filter. The band-pass filters ranged in center wavelength from 270nm to 330nm, each with a full-width half max of 10nm. Using spectroscopic data and manufacturer provided filter curves, the signal ratio as a function of temperature can be computed for a given filter combination. Figure 2b presents this curve for three different long-pass filters paired with a band-pass filter centered at 280nm. The slope of each curve is the temperature sensitivity of the imaging scheme.

In addition to temperature sensitivity, absolute signal levels must also be taken into account when selecting filters and estimating the performance of filter combination; a filter combination very sensitive to temperature may result in signals so low that images are of low quality (i.e., low SNR). Using linearized error propagation, sensitivity and expected signal level, and therefore expected signal-to-noise ratio the uncertainty in temperature, ΔT , is predicted for a given filter pair

ΔT 2 =dTdSf 1

"

#$$

%

&''

2

σ f 12 +

dTdSf 2

"

#$$

%

&''

2

σ f 22 , (6)

which can be reduced to

ΔT 2 =dTdψ"

#$

%

&'

2 Sf 1Sf 2

"

#$$

%

&''

21

SNR12 +

1SNR2

2

"

#$

%

&' , (7)

where subscripts correspond to each camera and filter, ψ = Sf1/Sf2, and dT/dψ is the inverse of the slope of the curve in figure 2b. A signal-to-noise ratio of 25 with a WG280 and UG11 filters at 520K was used as a baseline for comparing all other filter combinations under the shot-noise limit; for example, a filter combination that resulted in 1/2 the signal level as Sfwg280+UG11 has 1/√2 × SNRWG280+UG11.

Performance of each combination of filters was estimated at temperatures ranging from 298 to 900K. Figure 3 shows a colored matrix illustrating the performance of each filter combination; the color of each square in the matrix corresponds to the expected uncertainty for a filter combination given for the (i,j) location in the table at a given temperature. Combinations above the diagonal were computed at 450K, and below the diagonal were computed at 650K. If the expected uncertainty exceeds 150K, the square corresponding to that combination is colored white. A BP270 filter was included in our analysis for completeness, but due to the inability of a BP270 to adequately filter reflections of incident 266nm laser light, it was not considered as a candidate for use.

Two filter pairs outperform all the rest, the BP280/WG305 and BP280/WG320. The BP280/WG305 exhibits the lower sensitivity to temperature (i.e., this combination is the shallowest curve in figure


- 6 -

2b), yet this combination has similar performance to the BP280/WG320 combination, resulting from the larger pass band of the WG305 compared to the WG320. The larger signals passed by a WG305 results in higher quality images (i.e. greater SNR) than for the WG320 while offering similar performance to the WG320, and therefore, we elect to use the BP280/WG305 combination in this work. Henceforth, the camera outfitted with the BP280 filter will be referred to as the ’blue’ camera, and the camera with the WG305 will be referred to as the ’red’ camera.

3. Expansion tube facility and set up 3.1 Expansion tube operation An expansion tube is used to generate test gas conditions of interest; the facility is schematically depicted in Figure 4. Expansion tubes are impulse facilities, capable of producing high-enthalpy, short-duration flows with low levels of dissociation in the test gas [Trimpi, 1962]. These characteristics make such a facility useful for studying supersonic combustion, for which the composition of the test gas accurately represents conditions experienced by a hypersonic vehicle. In this work, the expansion tube is operated at relatively low free-stream enthalpies such that toluene yields large fluorescence signals and does not pyrolize.

For a thorough description of the facility, its design, operation, and capabilities, see Heltsley et al. [2006]. The expansion tube can be operated in “expansion mode” to generate low- to high-enthalpy supersonic flows; we use the tube in expansion mode for imaging supersonic flow over the wedge and cylinder. In expansion mode, an unsteady expansion accelerates and cools shock-treated test gas to supersonic speeds.

Fig. 3: Matrix of filter combinations and predicted performance. Squares above black diagonal correspond to expected temperature uncertainty at T = 450K, and below the diagonal T = 650K.


- 7 -

The expansion tube can also be operated in “shock mode,” for which there is no unsteady expansion to accelerate and cool shock-treated test gas. In “shock mode”, the tube is operated like a conventional shock tube, and imaging is done behind the incident shockwave, rather than after the reflected shockwave. Operation in “shock mode” enables the imaging of flows with well-known post-shock temperature, which is inferred from shock-speed measurements.

The test gas for all conditions tested is 0.5% toluene by volume in a balance of nitrogen. A 12 L mixing tank is used to make test gas mixtures. In “expansion mode”, test gas duration is approximately 3ms, and is steady in Mach number to within 10% as determined from high-speed schlieren imaging of an oblique shock over a wedge. In “shock mode”, post-shock temperatures and pressures range from 450K to 900K and approximately 0.5 bar to 1.5 bar.

3.1 Imaging system An injection seeded Coherent Powerlite 8000 Nd:YAG laser is externally quadrupled to 266nm. Three fourth harmonic dichroics separate the fundamental and second harmonic from the 266nm beam. The beam is collimated via f = −100mm and f = 400mm cylindrical lenses, and is focused into a sheet with a f = 1000mm cylindrical lens. In the field of view, sheet thickness of (FWHM) was 400μm was measured using a scanning knife-edge. Before the sheet forming optics, a beamsplitter diverts a fraction of the beam onto UV ground glass, which is monitored by a photodiode to provide a measure of shot-to-shot energy variation and aid in timing the laser pulse with respect to the camera gates.

Two intensified cameras are used to collect fluorescence, an Andor iStar 712 and a LaVision Dynamight. The iStar has a 512×512 pixel CCD array with 24×24μm pixels, an S20 photocathode, and a P43 phosphor. The Dynamight has a 1024 × 1024 pixel CCD array, with 13 × 13μm pixels; this camera is 2 × 2 hardware binned, effectively resulting in a 512 × 512 pixel array with 26 × 26μm pixels. The Dynamight, too, has an S20 photocathode and a P43 phosphor. The Dynamight is fitted with a Nikkor f/4 f = 100mm UV lens, and the iStar with a Sodern f/2.8 100mm UV lens. The Dynamight is mounted in a fixed position on one side of the test section, and the iStar is mounted on the opposite side of the test section on translation stages to allow for x−, y−, and z−translation. Cameras were aligned to match magnification and overlap the field of view as much as possible. Spatial resolution for both cameras is 95 μm per pixel.

The Dynamight was outfitted with two 2.5cm diameter BP280 interference band-pass filters and one UG11 Schott glass filter, and the iStar was fitted with a 5cm square by 3mm thick WG305 filter and one 2mm thick UG11 filter. Filters were added until reflections of excitation laser light off the

Fig. 4: A schematic of the expansion tube facility.


- 8 -

flow models were not saturating the cameras.

Two Stanford Research Systems (SRS) Delay Generators and one Berkeley Nucleonics Corporation (BNC) 555 Delay/Pulse generator are used to control timing of the cameras and laser. The BNC delay generator is triggered off of the first shock counter of the expansion tube, and then the delay generator subsequently triggers the cameras and laser.

Each camera is gated for 400ns, and delays are set such that the laser pulse is 100ns inside the gate of each camera. Fluorescence lifetimes for toluene are less than 50ns at 298K and 1bar, and decrease with increasing temperature Faust et al. [2011]. Gain on each camera was such that the maximum signal recorded was ∼ 25% below the saturation level of the ICCD cameras during test conditions.

3.3 Image processing Images are processed in two steps: 1) the image from each camera is corrected for background signal, CCD response, and laser sheet uniformity; 2) the images are dewarped and the corrected image from one camera is mapped onto the other so that the ratio of the images can be computed. For brevity, a general description is provided and intermediate processing results are omitted.

To correct each individual camera image, average background and dark noise images are first acquired and subtracted from raw images. Then an average “sheet” image is acquired at room temperature, from which background and dark noise are also subtracted. Background and sheet images are acquired on the two cameras simultaneously. For each camera, the corrected raw images are divided by the corrected “sheet” images to correct for spatial response of each camera under the assumption that the laser energy profile and seeding are uniform.

A grid of dots is imaged by each camera, and the resulting image is used to dewarp and map the images onto one another. The grid is a square matrix of 4mm diameter dots spaced 1mm apart printed on transparency and mounted in an aluminum frame. This grid is mounted onto the flow models, backlit, and imaged by both cameras with all filters and test section windows in place. Dot centers are located in each image via a Gaussian fitting routine, and a third order polynomial is used to map each pixel into physical space. Using these mappings, images are mapped on to each other, and then to further improve the overlap between images and account for any small movement of each camera during the run, fiducial marks in the laser sheet are located via another fitting Gaussian routine and used to shift one image to match the other. Finally, the corrected images are divided to yield the ratio of the images, normalized by the signal ratio at 298K (equation 5).

4. Results 4.1 Thermometry validation behind incident shockwaves To validate the thermometry strategy, we image incident shock waves in the expansion tube and compare measured signal ratio to the known temperature behind the incident shock wave. We operate the expansion tube in “shock mode” by not installing a secondary diaphragm into the tube. By varying driver gas composition (a mixture of helium and argon), we generate incident shocks of different strengths, and therefore different post-shock temperatures. Shock speed is computed as the time of flight between shock sensors at the wall. We trigger the laser pulse and cameras to image the incident shock while it is in the field of view. We then process images as described above.


- 9 -

Figure 5 displays the raw ’red’ image and the ratio of the ’red’ and ’blue’ images; the shock is located near x ≈ 18mm and is moving from left to right. Using a 25x25 pixel region of the image, we compute average and RMS values of signal ratio. These average values of signal ratio are plotted as a function of temperature behind the shock, and error bars represent the RMS on these measures (Figure 6). An expected signal ratio curve is calculated as described in 2.2, using the number of filters installed on our imaging system (3mm WG305+UG11 and 2×BP280+UG11).

Agreement between measured temperature and expected temperature is within 10% at temperatures below ∼ 700K. With increasing temperature, uncertainty in our measurements (i.e., error bars) increases as signal levels, and therefore SNR, decrease. At elevated temperatures, we also observe an increasing discrepancy between our measured and predicted signal ratios. We also attribute this to decreasing signal levels, which results in the final signal ratio being more sensitive to the details of the data reduction techniques (i.e., primarily by the background subtraction step). Further work is being conducted to determine the most appropriate way to correct for background signals at low signal level. This discrepancy may also result from varying spectral response as a function of wavelength of each imaging system (see equation 4), as well as discrepancies between actual and nominal filter transmission curves. In this work, the expected calibration curve is used to convert images of signal ratio to temperature (i.e., the solid curve in figure 6), and further work is being done to quantify the contribution of each cause to discrepancies between expected signal ratio and temperature.

(a) (b)

Fig. 5: (a) ’Red’ raw image of incident shockwave; (b) Ratio of ’red’ and ’blue’ images


- 10 -

4.1 Imaging of supersonic flow over the wedge and cylinder The technique described above is used to image two canonical flow fields, supersonic flow over a wedge, and supersonic flow over a cylinder. The free stream condition for both each cases is, M = 2.3, P = 1bar, and T = 480K, and are generated by the facility operated in expansion tube mode.

’Red’, ’blue’, and temperature images are presented for the flow over the wedge in Figures 7 and 8a. The shock is clearly marked by a sharp drop in signal. The shock is also 2 − 3 pixels wide, which is limited by the resolving power of our optical system. In both the ’red’ and ’blue’ image, we achieve an SNR of nearly 30, and in the temperature image, we achieve an SNR of 27 in the upstream region, and 22 downstream.

0 5 10 15 20 25 30 35 400

0.2

0.4

0.6

0.8

1

1.2

x, mm

Sign

al R

atio

300 400 500 600 700 800 9000

0.2

0.4

0.6

0.8

1

Temperature, K

BP28

0/WG

305

(a) (b) Fig. 6: (a) Profile of measured signal ratio across incident shocks; (b) measured signal ratio (black circles) as a function of temperature behind incident shocks. Yellow curve corresponds to expected signal ratios computed from spectroscopic data and filter transmission data.

(a) (b)

Fig. 7: (a)’Red’ and (b) ’blue’ images of supersonic flow over a wedge


- 11 -

Figure 8b shows a temperature profile, which is averaged over 10 rows of pixels taken about the dotted line in 8a. By taking the average measured free stream temperature, and inferring Mach number from the shock angle, we compute the expected temperature rise across the shock; this is plotted as the dotted line in figure 8b, and we find agreement is within 5% between the expected and PLIF-derived temperature across the oblique shock.

’Red’ and temperature images are presented for a bow shock over a cylinder under the same nominal free stream conditions of the flow over the cylinder (figures 9a and 9b). The resulting temperature field will be compared to a CFD solution in the future. Preliminary analysis, such as stagnation temperature and qualitative appearance of the temperature field, suggests that the measured temperature field is in good agreement with expected values. SNR values for the cylinder flow field are similar to the wedge case, with a temperature SNR of nearly 27 upstream of the shock, and ranging between 10 and 25 downstream of the shock.

0 10 20 30 40300

400

500

600

700

800

900

x, mm

Tem

pera

ture

, K (a) (b)

Fig. 8: (a) Temperature image of flow over a wedge; (b) measured (blue) and expected (black) temperature profile from the dotted line in figure 8a. Temperature profile is average of 10 rows of pixels.

(a) (b) Fig. 9: (a) ’Red’ raw images of supersonic flow over a cylinder; (b) temperature image of supersonic flow over a cylinder


- 12 -

5. Conclusions We have applied two-camera dual-band toluene thermometry PLIF for the imaging of supersonic flows. This technique has been validated by imaging flow behind incident shockwaves, and good agreement is observed between our measured and expected signal values. The technique has been demonstrated by imaging supersonic flow over a wedge and cylinder, and for both cases, the shock is clearly visualized, and temperature images agree with expected temperature fields. Further work is being done to reduce uncertainty in the measurements, and further comparisons will be made between temperature field images and CFD solutions.

Acknowledgements

This work is sponsored by the Department of Energy sponsored Predictive Science Academic Alliance Program PSAAP at Stanford University. The author would also like to thank Dr. Wieland Koban and Dr. Jon Koch for providing the spectroscopic data integral to this work.

References

Cheung B. Trace-Based Planar Laser-Induced Fluorescence Diagnostics: Quantitative Photophysics and Time-Resolved Imaging. PhD thesis, Stanford University, 2011.

Cundy M., Trunk P., Dreizler A., and Sick V. Gas-phase toluene LIF temperature imaging near surfaces at 10 kHz. Experiments in Fluids, pages 1–8, June 2011.

Faust S., Dreier T., and Schulz C. Temperature and bath gas composition dependence of effective fluorescence lifetimes of toluene excited at 266 nm. Chemical Physics, 383(1-3):6 – 11, 2011.

Heltsley W. N., Snyder J. A., Houle A. J., Davidson D. F., Mungal M. G., and Hanson R. K. Design and characterization of the Stanford 6 inch expansion tube. In 42nd AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, number AIAA 2006-4443, Sacramento, California, July 9-12 2006.

Koban W., Koch J. D., Hanson R. K., and Schulz C. Absorption and fluorescence of toluene vapor at elevated temperatures. Physical Chemistry Chemical Physics, 6(2940-2945), 2004.

Koch J. D., Hanson R. K., Koban W., and Schulz C. Rayleigh- calibrated fluorescence quantum yield measurements of acetone and 3-pentanone. Applied Optics, 43(31):5901–5910, 2004.

Mohri K., Luong M., Vanhove G., Dreier T., and Schulz C. Imaging of the oxygen distribution in an isothermal turbulent free jet using two-color toluene LIF imaging. Applied Physics B: Lasers and Optics, 103:707–715, 2011.

Trimpi R. L. A preliminary theoretical study of the expansion tube, a new device for producing high-enthalpy short-duration hypersonic gas flows. Technical report, NASA, 1962.

Yoo J., Mitchell D., Davidson D. F., and Hanson R. K. Planar laser-induced fluorescence imaging in shock tube flows. Experiments in Fluids, 45:751–759, 2010.

two-camera dual-band collection toluene plif thermometry...

Documents