Open Access
Add to BibSonomyAbstract
Modern diesel injectors operate at very high injection pressures of about 2000 bar resulting in injection velocities as high as 700 m/s near the nozzle outlet. In order to better predict the behavior of the atomization process at such high pressures, high-resolution spray images at high repetition rates must be recorded. However, due to extremely high velocity in the near-nozzle region, high-speed cameras fail to avoid blurring of the structures in the spray images due to their exposure time. Ultrafast imaging featuring ultra-short laser pulses to freeze the motion of the spray appears as an well suited solution to overcome this limitation. However, most commercial high-energy ultrafast sources are limited to a few kHz repetition rates. In the present work, we report the development of a custom-designed picosecond fiber laser generating ∼ 20 ps pulses with an average power of 2.5 W at a repetition rate of 8.2 MHz, suitable for high-speed imaging of high-pressure fuel jets. This fiber source has been proof tested by obtaining backlight images of diesel sprays issued from a single-orifice injector at an injection pressure of 300 bar. We observed a consequent improvement in terms of image resolution compared to standard white-light illumination. In addition, the compactness and stability against perturbations of our fiber laser system makes it particularly suitable for harsh experimental conditions.
© 2015 Optical Society of America
1. Introduction
High-velocity, high-pressure liquid jets and sprays are complex multiphase flow phenomena involved in many industrial applications such as fuel injection systems, thermal and plasma spray coating and liquid-jet machining [1–4]. In the case of high-pressure fuel injection, understanding the structure and dynamics of fuel sprays in the near-orifice region is critical in optimizing the injection process to increase fuel combustion efficiency and reduce pollutants emission. Indeed, most of the instabilities leading to the liquid jet atomization arise in the near-orifice region. Moreover, most of high-fidelity numerical simulations, for example based on the level-set method, are currently limited to the near nozzle region as they would require too heavy computational resources for the far-field of the nozzle [5]. To better understand how these instabilities originate and propagate, it is now crucial to record high-resolution images with high-repetition rates and compare them quantitatively with the images obtained through numerical simulations. However, liquid sprays are often difficult to study using optical techniques because of intense multiple light scattering from surrounding liquid droplets which hinders the exploitation of conventional techniques, such as phase Doppler measurements [6], direct imaging [7] or light scattering-based techniques [8], near the nozzle outlet. During the last decade, great efforts have been devoted to the development of advanced optical techniques featuring ultrafast lasers and X-ray radiation to mitigate multiple scattering [9]. Indeed, the exploitation of X-ray imaging to characterize the near nozzle region of fuel sprays has enabled remarkable advances by studying several features such as fuel mass fraction [10], spray-induced shock waves [11], fuel dynamics and velocity fields [12, 13], as well as the influence of the injection conditions and the inlet geometry on the spray dynamics in single and multi-hole injectors [14–16]. The main constraint with X-ray imaging is its limited potential for wider engine studies given the need for a synchrotron as an X-ray source and also interpretation of the X-ray images may be difficult [17]. The second relevant technique which is well adapted to investigate the optically dense region of the spray is ballistic imaging [18, 19]. This technique is based on the use of ultra-short light pulses and exploits the combined actions of an ultrafast shutter and a spatial filter to eliminate scattered light and capture only ballistic photons that have passed through the optically-dense spray. Ballistic imaging has been widely used to study different liquid jet systems, from water sprays to high-pressure diesel sprays [20–25]. Most of these studies are based on femtosecond Ti-sapphire chirped-pulse amplifiers [20–22] or picosecond Nd:YAG lasers [24, 25]. However, these commercial ultrafast laser sources operate at low repetition rates and for instance, for most Ti-sapphire crystal-based ultrafast lasers the repetition rates are limited to a few kHz. Q-switched Nd-YAG lasers may seem appropriate for high-repetition rate imaging of high-velocity diesel sprays, but speckle patterns induced by the interference of multi-reflections of such coherent monochromatic light conceal the underlying morphology of the spray, thus significantly altering the quality of the resulting images [26]. Speckle patterns can be avoided by spectrally diffusing the Nd-YAG laser beam [27] but the nanosecond pulse duration still remains too long to enable blur-free imaging of high-velocity sprays. The LED technology can provide high-frequency light pulses that may fit the high-speed imaging requirement [28]. However, the energy output of high-power sub-nanosecond LEDs is still insufficient to enable imaging in high-density high-velocity sprays [29]. In this contribution, we report for the first time to the best of our knowledge the development and exploitation of a high-repetition rate picosecond fiber laser to capture backlight images of high-pressure fuel sprays revealing a strong potential for high-velocity diesel spray metrology.
2. Design of the picosecond fiber laser
The picosecond laser system consists of a fiber master-oscillator power-amplifier (FMOPA) as shown in Fig. 1. It is based on a passively mode-locked ytterbium-doped fiber laser oscillator operating in the all-normal dispersion regime [30]. The oscillator consists of a Fabry-Perot cavity comprising 15 cm of heavily ytterbium-doped single mode fiber (YDF) with a core mode field diameter of 6 µm and a numerical aperture of 0.15. The YDF presents peak absorption of 1200 dB/m at 975 nm and its group velocity dispersion (GVD) has been estimated around +24 ps2/km at 1040 nm. It is pumped through a 980/1040 wavelength division multiplexer (WDM) using a single-mode laser diode. A fiber-Bragg grating-based filter (FBG) centered at 1040 nm is included in the cavity to limit the spectral bandwidth to less 140 pm [31]. The laser output is taken through the FBG which presents a low reflectivity of 20%. Mode-locking is achieved by using a resonant saturable absorber mirror (R-SAM) which constitutes the end mirror of the cavity. It exhibits a high modulation depth of ∼ 50% and a relaxation time of 9 ps. By adjusting the focusing conditions on the R-SAM to meet the saturation criteria, a sustained single-pulse mode-locking regime is achieved at a relatively low pump power level of about 150 mW. The core-pumped single mode laser generates nearly transform-limited pulses with ∼ 20 ps duration and less than 1 nJ energy, at a repetition rate of 8.2 MHz.
Fig. 1 Experimental setup of the fiber master-oscillator power-amplifier (FMOPA). Nomenclature - R-SAM: resonant saturable absorbed mirror; {L1, L2}: biconvex lenses; P1: polarizer; H1: half wave-plate; WDM: wavelength division multiplexer; SM: single mode; LD: Laser diode; FBG: fiber Bragg grating; MM: multi-mode; OC: output coupler; COM: pump and signal combiner; LMA: large mode area; HP-ISO: High power isolator.
A two-stage fiber power-amplifier is then used to boost pulse energy to some hundreds of nano-joules. The first stage is based on a core-pumped single-mode YDF and delivers more than 8 nJ energy. The final stage includes a double clad step-index YDF with a core diameter of 35 µm from INO Inc. The fiber is based on the selective gain amplification concept to discriminate high-order modes and thus promote quasi-single mode operation. The measured M2 value of less than 1.15 confirms the very good quality of the guided beam. The fiber is cladding-pumped through a signal-pump combiner by two fiber-coupled multimode diodes emitting at 976 nm. The fiber end is angle cleaved to avoid any parasitic reflection into the power amplifier. Moreover, a high-power isolator is added to protect the laser system from any parasitic feedback. Pulse energy as high as 300 nJ is achieved after the power amplifier without any temporal distortion. This corresponds to more than 2.5 W of average power. The auto-correlation trace measured at the FMOPA output is shown in Fig. 2 and corresponding pulse duration assuming a Gaussian pulse shape is 21 ps. It is worth noting that at this energy level the output pulses endure some spectral broadening due to self-phase modulation as evidenced in the inset. For energies higher than 400 nJ, nonlinear effects can cause severe temporal distortions on the output pulse.
Fig. 2 Autocorrelation trace measured at the FMOPA output. Inset: Optical spectrum measured after the amplifier.
The amplitude stability of the mode-locked pulse train was evaluated by radio frequency (RF) measurements using the power spectra obtained with a microwave spectrum analyzer via a high-speed photo detector (8 GHz bandwidth). It exhibits very satisfying stability with a contrast of more than 70 dBm between the fundamental harmonic and the background as depicted on Fig. 3. This indicates that laser operation is very stable and free from any Q-switching instability. Only one noise band is observed at very low frequencies (< 4 kHz). The energy fluctuations associated with this structure are estimated to be lower than 0.25%.
3. Spray backlight imaging
3.1. Experimental setup
The fiber laser (λ = 1040 nm, pulse width ≈ 20 ps, average power ≈ 2.5 W) was used for recording high-speed spray videos using a Phantom v12.1 high-speed camera with 300 ns of exposure time. The experimental setup is shown in Fig. 4. An electro-optic modulator from Conoptics Inc. was used to decrease the repetition rate of the fiber laser from 8.2 MHz (one laser pulse every 120 ns) to less than 100 kHz to get only one pulse in each frame of the high-speed camera. The electro-optic modulator is designed for pulsed lasers up to 100 MHz with wavelength range from λ = 350 − 1600 nm. The output pulses features (pulse shape and duration) are not affected by the electro-optic modulator due to their relatively narrow spectral width. Note that, the repetition rates used in these experiments are limited by the high-speed camera and not by the laser.
The Phantom high-speed camera was being operated at a resolution of 800 × 600 px and a frame rate of 11000 fps. A Micro-Nikkor lens from Nikon with focal length f = 200 mm and an adjustable aperture ranging from f/32 to f/4 was used to form an image of the fuel spray at the camera with a magnification of 1.1. The synchronization between the laser, fuel injection and camera was adjusted using a DG-645 (Stanford Research Systems) electronic delay generator.
The high-repetition fuel spray images obtained when the spray was illuminated using the ps fiber laser are shown in the bottom half of Fig. 5 and are compared qualitatively with the images obtained using continuous white-light from a Xenon lamp. The experimental setup in Fig. 4 was modified accordingly - the electro-optic modulator was removed and the illumination source was changed to a Xenon lamp without disturbing rest of the setup including the high-speed camera. To avoid accidental combustion as well as pollution of the optical devices, the fuel spray was injected inside a closed chamber maintained at atmospheric pressure and temperature. Injection pressure was set to ∼ 200 bar, injection duration to 5.0 ms and the spray was issued using a single orifice nozzle with a diameter of 185 µm. The liquid used for injection was ISO 4113 calibrated oil (see Table 1) with very similar properties as that of diesel fuel. Figure 5 includes the first few frames of the high-speed video captured using the high-speed camera showing the formation of the spray. The time for each frame with respect to the start of the injection (SOI) is indicated below the frame. The last images (e and j) recorded 1.7 ms after the SOI for both cases show the spray in its stationary state – when the injector needle is completely lifted. The optical resolution of these images is 18.2 µm per pixel. First qualitatively, one can immediately notice that the images obtained with our custom-designed ps fiber laser are sharper than with white-light illumination.
Fig. 5 Spray backlight images obtained using continuous white-light (top, a to e) and ps fiber laser (middle, f to j) illuminations. Optical magnification: 1.1. A zoomed section from the (d)th and (i)th frames are shown at the bottom - left image corresponds to frame (d) (with continuous illumination) and the image on the right corresponds to (i) (with pulsed illumination). The colored curve in images (e) and (j) correspond to the spray boundary extracted using standard image processing tools.
Table 1. Physical properties of the diesel-like liquid used for recording spray images.
The theoretical fuel velocity estimated from Bernoulli’s equation is of the order of 200 m/s for an injection pressure of 200 bar. In order to prevent motion blur, liquid fuel must travel a distance shorter than a pixel size (i.e. 20 µm), corresponding to a travel time of ∼ 0.1 µs. If continuous illumination is used, the exposure fixed by the camera’s shutter (300 ns) is three times larger, which leads to a blurring effect of about 3 pixels. On the other hand, displacement of fuel with ps laser is only ∼ 4 nm, which is much shorter than the pixel size. This is one of the reasons for the difference in image quality obtained with continuous white-light and ps fiber laser illuminations.
3.2. Static spatial resolution measurement
The spatial resolution of the optical setup was measured using USAF 1951 chart with continuous white-light and pulsed fiber laser illuminations. Images of USAF chart with the two illumination sources are shown in Fig. 6. The measured value of the static spatial resolution (target at rest) with continuous white-light was found to be 16 line pairs/mm, corresponding to the 1st element of the 4th group, where the three lines in the chart are distinguishable (Cmin > 0.1). The same with ps fiber laser illumination was found to be 27.5 line pairs/mm, limited by the pixel size of the camera. This means that with our optical setup it is possible to resolve an object of 31 µm or more using the broadband white-light illumination whereas with ps fiber laser even smaller objects can be resolved.
Fig. 6 (Top) Image of USAF 1951 chart with (left) continuous white-light and (right) ps fiber laser illuminations. Optical magnification: 1.1. (Bottom) Contrast function, C = (Imax − Imin)/(Imax + Imin) with respect to spatial resolution in line pairs/mm. measured for all elements of groups 2, 3 and 4 of the USAF chart.
Note that the image of the USAF chart when illuminated with fiber laser is much sharper than that obtained using white-light. The wide spectrum of white light source can give rise to chromatic aberration making the image appear as it is out of focus, as can also be deduced from the contrast function (C) shown in Fig. 6. The overall higher value of the contrast for all groups with fiber laser illumination shows a significant improvement in the contrast compared to standard white light illumination.
Further, to quantify the difference in the spray image quality obtained with continuous white light and ps fiber laser illuminations, we monitored the length of the edge of the spray as it evolved with time.
4. Evolution of spray edge length with time
The length of the edge of the spray was estimated by using standard image processing tools as available in most commercial softwares. The recorded grayscale spray images with the two kinds of illuminations were divided by the background image (an image without the spray) and then the pixel values were re-scaled from 0 to 1. These normalized images were then converted to binary images based on a threshold value (Th) determined independently using the algorithm proposed by N. Otsu [33]. These values of Th are shown in the bottom part of Fig. 7 for each spray image obtained at different time after the SOI with the two kinds of illuminations. The overall different values for binarization threshold with two illuminations is due to the difference in background intensities whereas the fluctuations in Th values is due to frame to frame intensity fluctuations (negligible for continuous source).
Fig. 7 The length of the spray edge (top subplot) estimated from the spray images recorded at different time from the start of the injection (SOI) for continuous white light and ps fiber laser illuminations. The consecutive plots correspond to different threshold values (Th) used for the images binarization, as displayed on the legend. The symbols Thc and Thp indicate different binarization threshold values for continuous and pulsed illuminations respectively, determined independently for each spray image and shown in the bottom subplot.
From one of the two spray side boundaries (left boundary chosen arbitrarily) the length of the spray edge was estimated by simply counting the number of pixels on the spray contour in the binary spray image. One detected spray edge after binarization is highlighted in Figs. 5(e) and 5(j) for demonstration. The top part of the Fig. 7 shows the spray edge length for different thresholds for the spray images obtained with continuous white-light and ps fiber laser illuminations, showing the low sensitivity of edge length with threshold (Th).
The overall higher value of the spray edge length for the images obtained with ps laser illumination confirms that these spray images are more detailed with well defined boundaries compared to the images obtained with continuous white-light illumination. The larger fluctuations in the edge length over time indicate that the spray is changing continuously and as one can see these changes are not so profound when the spray is illuminated using the continuous white light due to the fact that the spray images are not well resolved due to significant amount of motion blur.
5. Ultrafast high-repetition recording
Backlight spray images are recorded at a frame rate of 80575 Hz (one pulse every 12.41 µs) using the same optical setup as shown in Fig. 4. As the throughput of the high-speed video cameras is fixed, image recording size was reduced to 512×128 pixels to reach this high frame rate. The injection pressure for these measurements was 300 bar (corresponding to Bernoulli velocity of ∼ 250 m/s) for a duration of 2.5 ms. A single orifice nozzle with orifice diameter 350 µm was used here.
Figure 8 shows a few frames from the high-speed spray video. With such high-repetition acquisitions, entire duration of fuel injection can be analyzed where only one image-pair is recorded with commercial ultrafast lasers [34]. This frame rate is high enough to determine fuel velocity at the beginning (or end) of the fuel injection, as can be seen in Fig. 8. Fuel velocity can as well be estimated during the stationary phase of injection by further increasing the frame rate of the high-speed camera (and reducing image size).
Fig. 8 Spray backlight images obtained using ps fiber laser illumination (a to t) at a repetition rate of 80.5 kHz (time between consecutive frames ∆t = 12.41 µs). Optical magnification: 0.63.
6. Conclusions
A monolithic fiber laser system based on a highly stable all-normal dispersion fiber laser and a Yb-doped large mode area fiber has been demonstrated for the first time. The custom-designed fiber laser generates ultra-short pulses at 1040 nm wavelength with 21 ps duration and 2.5 W average power at a repetition rate of about 8 MHz. This corresponds to more than 300 nJ pulse energy. Energy scaling to the microjoule level is yet possible by shortening the length of the amplifying LMA fiber or by increasing the pulse duration using a FBG with narrower bandwidth.
It is shown with the help of a USAF 1951 test chart that the images obtained with our picosecond fiber laser have higher contrast compared to the images obtained using continuous white-light illumination. The spatial resolution of our optical setup at a magnification of 1.1 was limited by the resolution of the high-speed camera to a value of 25.4 line pairs/mm for the ps laser.
The fuel spray images obtained with continuous white light and pulsed illuminations were compared quantitatively by estimating the spray edge length and its evolution over time after the start of fuel injection. The overall higher value of the spray edge length with picosecond fiber laser compared to continuous white-light shows that the spray images obtained with ps fiber laser illumination are better resolved. This is due to the very short pulse duration of fiber laser, which prevents spray images from any blurring effect and, due to its narrow spectral range, effects of chromatic aberration are avoided.
Since the repetition rate of our fiber source is much higher than that of the most commercial ultrafast lasers used for imaging, this feasibility study paves the way for studying high-pressure fuel spray with over a single injection as the only limitation is now the frame rate of the high-speed camera. It is also worth emphasizing that unlike standard cumbersome ultrafast lasers, our optical fiber-based system can fit in a laptop-size case and is moreover stable against temperature variations and vibrations, making it suitable for industrial applications. Further experiments are now considered to exploit the full potential of this fiber source, for instance the study of the near-field dynamics of the fuel jet using several pulse sequences and image autocorrelation.
Acknowledgments
We acknowledge support from the French Agence Nationale de la Recherche (ANR), through the program “Investissements d’Avenir ( ANR-10-LABX-09-01), LabEx EMC3” and the Region Haute Normandie, through the project MIST.
References and links
1. J. D. O’Keefe, W. W. Wrinkle, and C. N. Scully, “Supersonic liquid jets,” Nature 213, 23–25 (1967). [CrossRef]
2. K. K. Kuo, Recent Advances in Spray Combustion: Spray Atomization and Drop Burning Phenomena (American Institute of Aeronautics and Astronautics, 1996).
3. H. Herman, “Plasma spray deposition processes,” MRS Bull. 13, 60–67 (1988).
4. R. A. Tikhomirov, V. F. Babanin, E. N. Petukhov, I. D. Starikov, and V. A. Kovalev, High-pressure Jetcutting (Amer Society of Mechanical, 1992).
5. T. Ménard, S. Idlahcen, J.-B. Blaisot, C. Rozé, A. Berlemont, T. Girasole, and L. Méès, “Numerical simulation of optical diagnostics and comparisons to experiments,” in “Int. Conf. Multiph. Flow,” (2010), pp. 1–7.
6. W. D. Bachalo and M. J. Houser, “Phase/Doppler spray analyzer for simultaneous measurements of drop size and velocity distributions,” Opt. Eng. 23, 583 (1984). [CrossRef]
7. G. A. Ruff, L. P. Bernal, and G. M. Faeth, “Structure of the near-injector region of nonevaporating pressure-atomized sprays,” J. Propuls. Power 7, 221–230 (1991). [CrossRef]
8. L. Arnone, F. Beretta, A. Tregrossi, A. D’Alessio, and F. Ossler, “Ensemble and time resolved light scattering measurements in isothermal and burning heavy oil sprays,” Symp. Combust. 24, 1549–1555 (1992). [CrossRef]
9. M. Linne, “Imaging in the optically dense regions of a spray: A review of developing techniques,” Prog. Energy Combust. Sci. 39, 403–440 (2013). [CrossRef]
10. Y. Yue, C. F. Powell, R. Poola, J. Wang, and J. K. Schaller, “Quantitative measurements of diesel fuel spray characteristics in the near-nozzle region using X-ray absorption,” At. Sprays 11, 471–490 (2001). [CrossRef]
11. A. G. MacPhee, “X-ray imaging of shock waves generated by high-pressure fuel sprays,” Science 295, 1261–1263 (2002). [CrossRef] [PubMed]
12. Y. Wang, X. Liu, K.-S. Im, W.-K. Lee, J. Wang, K. Fezzaa, D. L. S. Hung, and J. R. Winkelman, “Ultrafast X-ray study of dense-liquid-jet flow dynamics using structure-tracking velocimetry,” Nat. Phys. 4, 305–309 (2008). [CrossRef]
13. K.-S. Im, K. Fezzaa, Y. J. Wang, X. Liu, J. Wang, and M.-C. Lai, “Particle tracking velocimetry using fast x-ray phase-contrast imaging,” Appl. Phys. Lett. 90, 091919 (2007). [CrossRef]
14. S. Moon, Y. Gao, J. Wang, K. Fezzaa, and T. Tsujimura, “Near-field dynamics of high-speed diesel sprays: Effects of orifice inlet geometry and injection pressure,” Fuel 133, 299–309 (2014). [CrossRef]
15. S. Moon, K. Komada, K. Sato, H. Yokohata, Y. Wada, and N. Yasuda, “Ultrafast X-ray study of multi-hole GDI injector sprays: Effects of nozzle hole length and number on initial spray formation,” Exp. Therm. Fluid Sci. 68, 68–81 (2015). [CrossRef]
16. S. Moon, Y. Gao, S. Park, J. Wang, N. Kurimoto, and Y. Nishijima, “Effect of the number and position of nozzle holes on in- and near-nozzle dynamic characteristics of diesel injection,” Fuel 150, 112–122 (2015). [CrossRef]
17. M. Linne, “Analysis of X-ray phase contrast imaging in atomizing sprays,” Exp. Fluids 52, 1201–1218 (2012). [CrossRef]
18. M. Paciaroni and M. Linne, “Single-shot, two-dimensional ballistic imaging through scattering media,” Appl. Opt. 43, 5100 (2004). [CrossRef] [PubMed]
19. M. Linne, M. Paciaroni, T. Hall, and T. Parker, “Ballistic imaging of the near field in a diesel spray,” Exp. Fluids 40, 836–846 (2006). [CrossRef]
20. J. B. Schmidt, Z. D. Schaefer, T. R. Meyer, S. Roy, S. a. Danczyk, and J. R. Gord, “Ultrafast time-gated ballistic-photon imaging and shadowgraphy in optically dense rocket sprays,” Appl. Opt. 48, B137 (2009). [CrossRef] [PubMed]
21. M. Linne, D. Sedarsky, T. Meyer, J. Gord, and C. Carter, “Ballistic imaging in the near-field of an effervescent spray,” Exp. Fluids 49, 911–923 (2010). [CrossRef]
22. D. Sedarsky, J. Gord, C. Carter, T. Meyer, and M. Linne, “Fast-framing ballistic imaging of velocity in an aerated spray,” Opt. Lett. 34, 2748 (2009). [CrossRef] [PubMed]
23. H. Purwar, S. Idlahcen, C. Rozé, D. Sedarsky, and J.-B. Blaisot, “Collinear, two-color optical Kerr effect shutter for ultrafast time-resolved imaging,” Opt. Express 22, 15778–15790 (2014). [CrossRef] [PubMed]
24. S. P. Duran, J. M. Porter, and T. E. Parker, “Ballistic imaging of diesel sprays using a picosecond laser: characterization and demonstration,” Appl. Opt. 54, 1743 (2015). [CrossRef]
25. S. P. Duran, J. M. Porter, and T. E. Parker, “Picosecond ballistic imaging of diesel injection in high-temperature and high-pressure air,” Exp. Fluids 56, 1–12 (2015). [CrossRef]
26. C. Badock, R. Wirth, A. Fath, and A. Leipertz, “Investigation of cavitation in real size diesel injection nozzles,” Int. J. Heat Fluid Flow 20, 538–544 (1999). [CrossRef]
27. C. Crua, M. R. Heikal, and M. R. Gold, “Microscopic imaging of the initial stage of diesel spray formation,” Fuel 157, 140–150 (2015). [CrossRef]
28. C. Willert, S. Freitag, and C. Hassa, “High speed imaging of fuel spray using a low-cost illumination source,” in “ILASS 2008,” (Como Lake, Italy, 2008), September, pp. 8–10.
29. D. Birch, C. McGuiness, K. Sagoo, and D. McLoskey, “A new sub-nanosecond LED at 280 nm: application to protein fluorescence,” Meas. Sci. Technol. 15, 19–22 (2004). [CrossRef]
30. C. Lecaplain, M. Baumgartl, T. Schreiber, and A. Hideur, “On the mode-locking mechanism of a dissipative-soliton fiber oscillator,” Opt. Express 19, 26742 (2011). [CrossRef]
31. M. Baumgartl, T. Gottschall, J. Abreu-Afonso, A. Díez, T. Meyer, B. Dietzek, M. Rothhardt, J. Popp, J. Limpert, and A. Tünnermann, “Alignment-free, all-spliced fiber laser source for CARS microscopy based on four-wave-mixing,” Opt. Express 20, 21010 (2012). [CrossRef] [PubMed]
32. C. Schaschke, I. Fletcher, and N. Glen, “Density and viscosity measurement of diesel fuels at combined high pressure and elevated temperature,” Processes 1, 30–48 (2013). [CrossRef]
33. N. Otsu, “A threshold selection method from gray-level histograms,” IEEE Trans. Syst. Man. Cybern. 9, 62–66 (1979). [CrossRef]
34. D. Sedarsky, S. Idlahcen, C. Rozé, and J.-B. Blaisot, “Velocity measurements in the near field of a diesel fuel injector by ultrafast imagery,” Exp. Fluids 54, 1451 (2013). [CrossRef]
