Open Access
Add to BibSonomy
Abstract
Direct femtosecond shadowgraphy and supercontinuum (SC)-illumination imaging methods for diagnosing liquid-gas mixed sprays in the near-field region of spray nozzles were compared. Some big spray structures can be captured using femtosecond shadowgraphy which can freeze the motion of the sprays. But the speckles caused by the interference of multi-scattered photons erode the edges of ligaments and conceal many fine droplets. SC-illumination imaging can not only freeze the motion of the sprays but also significantly suppressing the speckles, presenting a more realistic spray pattern. Based on the SC imaging technology, the effects of various factors such as flow ratio of gas to liquid (GLR), total flow and nozzle size on the spray were studied.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
The importance of the liquid fuels, specifically fossil fuels is unquestionable in the transportation sector of today [1]. Liquid fuels are widely used in various engines because they have significantly higher volumetric energy density than fuels in other phases and have advantages in storage and transportation [1–3]. The conversion from the chemical energy stored in the liquid fuels to mechanical energy relies on the chemical reaction in gas phase between air and fuel vapors. The efficient evaporation demands an increase in the surface area of the liquid by breaking the liquid up into small droplets. These droplets then evaporate and mix with air for combustion. Hence, the complete and efficient fuel combustion needs the well-designed process for the liquid fuels including injection, atomization, evaporation and mixing with air.
It’s very meaningful to research the conversion from the contiguous liquid to dispersed droplets by injecting it into the combustion chamber through a liquid-gas mixed nozzle. However, the dynamics are not well understood because the fuel spray has very high speed (several hundred m/s) and high concentration in two-phase flow which makes it optically impenetrable. In the last few years, a variety of techniques have been proposed to capture the dynamics of high-pressure liquid-gas sprays. The straightforward way to capture the fast moving target could be using a camera fast enough to record the moving target. However, because of the prominent photon degeneracy caused by the high spray concentration, conventional techniques are not able to reveal internal spray structure accurately. And because of the extremely high velocity of the liquid-spray in the near-nozzle region, high-speed cameras fail to avoid blurring of the structures in the spray images due to their inadequate exposure time. For these limitations of traditional optical diagnostic methods, the ultrafast laser are the preferred light sources in modern imaging systems [4]. But the speckles caused by the interference of multi-reflections of such coherent photons conceal the true morphology of the spray, thus significantly degrading the quality of the resulting images [5]. In shadowgraph technologies, the identifiability of some valuable features in images of the research targets could be corrupted [6]. Some digital image processing algorithms have been developed for suppressing the speckle formation [7]. But the digital method is not suitable for imaging the spray patterns in real time. In principle, reducing the coherence of the illumination source is one of the criterions for suppressing speckles.
Over the years, some techniques have been proposed to suppress speckles by using incoherent light sources. It’s demonstrated that random lasers with inhomogeneous and highly irregular spatial modes can be used to provide speckle-free full-field imaging in intense optical scattering conditions [8]. Laser based broadband green illumination source with low temporal coherence was also used to reduce the speckle noise [9]. Speckle patterns can also be avoided by diffusing the spectra of the Nd-YAG laser beam [10]. However, the pulse duration of these illumination sources can only be modulated up to the nanosecond level, which still remains too long to freeze the motion of high-velocity sprays [11–13]. The LED technology can provide high-frequency light pulses with short enough pulse duration which may fit the high-speed capturing requirement. However, the energy output of high-power sub-nanosecond LEDs is still deficient to enable imaging in high-density and high-velocity sprays [14,15]. While, breaking though the limitations of these methods, supercontinuum (SC) may be an ideal illumination source for imaging liquid-gas mixed sprays without producing motion blurring effects and speckle erosion. SC improves the image quality by destroying the coherence of the laser beam [16], keeping the pulse width almost same [17], and keeping high enough peak power of every pulse. The images of spray internal structures with high recognizability and high time resolution can be captured by using SC as the illumination source.
In our previous work, SC illumination with low coherence has been employed to achieve speckle-suppressed full-field imaging of resolution test target through a scattering medium [18]. In this study, we used femtosecond laser-induced SC as a light source to perform optical diagnosis for the gas-liquid mixed spray. The results showed that this method could not only avoid blurring effect of the high-speed moving spray structures but also signally suppress the speckle noise, which significantly improved the signal-to-noise ratio and authenticity of gas-liquid mixed spray imaging. Based on this technology, we further studied the dynamics of the liquid-gas mixed spray by quantitatively changing the parameters affecting the spray patterns such as the flow ratio of gas to liquid (GLR), total flow of gas and liquid, and exit diameter of the nozzle. And some laws about controlling spray morphologies have been obtained.
2. Experimental details
A schematic of the imaging system is shown in Fig. 1(a). An amplified Ti:sapphire femtosecond laser system (Libra-USP-HE, Coherent Inc., USA) was used as the initial light source. It can generate ∼120-fs, 3-mJ, and 800-nm laser pulses at a repetition rate of 1 kHz. A plano-convex lens L1 (focal length, f1 = 150 mm) was used to focus the femtosecond laser pulses into a 5-cm-thickness quartz cuvette filled with distilled water to generate SC. A continuously variable attenuator F1 was used to adjust the power of incident light in order to ensure the stability of SC. The power of the beam used to induce the supercontinuum was 40 mW in our experiment. The transmitted SC was collected by a plano-convex lens L2 (focal length, f2 = 150 mm). Then the SC was modulated by the imaging object. The imaging object could be a resolution test target (RT, RT-MIL-TP2001, Beijing Reallight Technology Co., Ltd., China) or liquid-gas mixed spray. An achromatic lens L3 (focal length, f3 = 200 mm) was used to image the object and was placed at twice the focal length behind the object. To perform SC-illumination imaging and femtosecond shadowgraphy simultaneously, the imaging beam was split into two parts by a beam splitter (BS). The transmitted part formed the SC-illumination imaging, in which the short-pass filter (SPF) filtered out the light with a wavelength longer than 800nm and reserved the supercontinuum components. The reflective part formed the femtosecond shadowgraphy, in which the band-pass filter (BPF) only passed the 800nm fundamental light through. The cut-off wavelength of the short-pass filter is 800 nm. The transmission bandwidth and transmittance of the short-pass filter are 400-790 nm and over 91% respectively. The central wavelength and full width at half maximum (FWHM) of the band-pass filter are 800 ± 1 nm and 5 ± 1 nm respectively. Both imaging beams were imaged onto charge-coupled device cameras (CCD1 and CCD2, INFINITY3-1 M-NS-TPM, Lumenera Corporation, Ottawa, Canada). Two continuously variable attenuators F2 and F3 were used to adjust brightness of images on the CCDs. The spectrums of supercontinuum and fundamental-frequency laser after the SPF and BPF were shown in Fig. 1(b).
Fig. 1. (a) Imaging system for simultaneously performing the femtosecond shadowgraphy and SC-illumination imaging — M1, M2: reflector; F1, F2, F3: continuously variable attenuator; L1, L2, L3: lens; BS: beam splitter; SPF: short-pass filter; BPF: band-pass filter. All the three lenses L1, L2 and L3 have a diameter of 25.4 mm; (b) Spectrums of supercontinuum and fundamental-frequency laser.
In the experiment, the liquid-gas spray was provided by a specially designed spray generation system with an imitative rocket injector modified from a gas-centered swirl-coaxial injector configuration [19]. The infector could produce a liquid flow with a range of 0.2 to 6 liters per minute (lpm) and ± 2.5% full-scale accuracy. It could produce a gas flow with a range of 10 to 200 liters per minute (lpm) and ± 2% full-scale accuracy. The exit diameter of the nozzle could be changed to 1.5mm, 2.5mm, 4mm and 4.5mm.
3. Results and discussion
First, the performance of the imaging system was tested with a resolution test chart as the imaging object. The femtosecond shadowgraphy and SC-illumination imaging for the object were respectively shown in Figs. 2(a) and 2(b). From Figs. 2(a) and 2(b), we found that the background of the picture taken with SC-illumination imaging was more uniform. Because the interference caused by defects of optical components in the optical path was reduced when the supercontinuum with low coherence was used as a light source. To evaluate the performance of the imaging system quantitatively, the Modulation Transfer Function (MTF) and Point-Spread Function (PSF) curves of each experimental configuration were then calculated as shown in Figs. 2(c) and 2(d). The calculation methods of MTF and PSF are derived from the Ref. [20]. The results showed that the both imaging methods had almost the same ability to reproduce the imaging target, in which the maximum resolvable spatial frequencies were almost 50 line pairs per millimeter (lp/mm) and the full width at half maximum (FWHM) of the PSF was about 2 mm.
Fig. 2. Resolution test chart images obtained by femtosecond shadowgraphy and SC-illumination imaging methods and their MTF and PSF — (a) femtosecond shadowgraphy; (b) SC-illumination imaging; (c) MTF of the imaging system; (d) PSF of the imaging system.
Then, the femtosecond shadowgraphy and SC-illumination imaging methods for measurements of flow structures in liquid-gas mixed sprays were compared. In the experiment, the exit diameter of this nozzle was 3.5 mm, and the flow rates of water and air were 0.4 lpm and 170 lpm, respectively. Figures 3(a)-I and 3(b)-I show images of liquid-gas mixed spray simultaneously obtained by using femtosecond shadowgraphy and SC-illumination imaging. It’s noted that the size of the imaging area is 7.39 mm × 5.52 mm. The Figs. 3(a)-II – 3(a)-IV and Figs. 3(b)-II – 3(b)-IV were the magnified views of the selected areas in Figs. 3(a)-I and 3(b)-I respectively. These magnified views were processed to two-level images using the maximum entropy threshold segmentation method in order to enhance the comparison effect [21]. From the Fig. 3(a), we can see that some macro information of the liquid-gas spray structures, such as curvature of the big liquid structure, ligament length, and macro distribution of the droplets, can be captured by using the femtosecond shadowgraphy due to freezing the motion of the spray structure. But from the magnified views in the Fig. 3(a) we can see that speckles caused severe noise which seriously affected the authenticity of the spray structure if using femtosecond laser as the illumination source. In the femtosecond shadowgraphy, some edges of large and medium ligament structures were blurred by speckle erosion. For the same structures captured by SC-illumination imaging as shown in Figs. 3(b)-II – 3(b)-IV, the edge structures were significantly sharper. For the small ligaments and tiny droplets, speckles appeared on both the edges and the interior in femtosecond shadowgraphy, which caused these structures difficult to distinguish from the background noise. In SC-illumination imaging, false black and bright dots were signally reduced. These tiny structures were imaged more authentically and more clearly because the noise caused by speckles was significantly suppressed.
Fig. 3. Images of liquid-gas spray obtained by two methods of imaging and their partial enlargements: (a) femtosecond shadowgraphy and the magnified views of the selected areas; (b) SC-illumination imaging and the magnified views of the selected areas.
Thus, we further analyzed the dynamics of the liquid-gas mixed spray by quantitatively changing the parameters affecting the spray patterns by using the SC-illumination imaging methods. First, we quantitatively changed the ratio of gas to liquid flow (GLR) to observe the patterns of liquid-gas mixed sprays. It’s noted that the nozzle with an exit diameter of 2.5mm was used in this part of the experiment. We set the liquid flow to 0.4 liters per minute (lpm) and set the gas flow to 10, 50, 90, 110 and 130 lpm respectively. The changes of appearance of the spray were shown in Figs. 4(a). As the GLR increased, the liquid core at the nozzle exit was broken up to ligaments faster and further broken up to tiny drops more completely. When the liquid and gas flow were set as 0.4 lpm and 10 lpm respectively (GLR = 25), we can see that liquid core was only broken into some big liquid sheets as shown in the Fig. 4(a)-I. When the gas flow was increased to 50 lpm (GLR = 125), we can see that the liquid sheet only existed near the bottom of the nozzle and the ligaments were abundantly present in the imaging area as shown in the Fig. 4(a)-II. As the GLR was further improved, airflow further ruptured the liquid sheets and ligaments. The proportion of ligaments gradually decreased while the primary droplets were significantly increased. When the liquid flow reached 130 lpm (GLR = 325), the liquid was broken quite thoroughly, making the spray optically dense as shown in the Fig. 4(a)-V. The similar law for sprays occurred when the liquid flow was set to another value. To quantitatively describe the effect of GLR on the optical depth of the spray, we calculated the average intensity of each subgraph in Fig. 4(a) and plotted the line graph as shown in Fig. 4(b). We can see that the intensity of the images decreased with the GLR’s increasing. It proved that higher gas flow resulted in denser primary droplet distribution and made the optical depth of the spray field area deeper.
Fig. 4. Spray morphology and average strength under different gas-liquid ratio conditions: (a) Spray morphology under different GLR; (b) average strength curve of Fig. 4(a).
In the next part of the experiment, we designed four sets of spray conditions with almost the same GLR (350∼380) but different total flow: liquid flow rates were set to 0.2, 0.3, 0.4 and 0.5 lpm respectively; gas flow rates were set to 70, 110, 150 and 190 lpm respectively. It’s noted that the nozzle with an exit diameter of 2.5 mm was used. The results were shown in Fig. 5. It’s obvious that the density of the spray increased with the growth of total flow. What was meaningful was that the distribution of liquid form had a relationship with the total flow. When the total flow was low, the liquid mainly existed in the form of liquid sheets and big ligaments as shown in Fig. 5(a). When the total flow was high, the proportion of small-sized liquid structures significantly became higher.
Fig. 5. Spray morphology under different total flow conditions, the flow rates of liquid and gas are (unit: lpm): (a) 0.2 and 70; (b) 0.3 and 110; (c) 0.4 and 150; (d) 0.5 and 190.
Furthermore, we studied the effect of nozzle diameter on spray morphology. In this part of experiment, we set the liquid and gas flow to 0.3 and 50 lpm respectively (GLR = 167). The exit diameters of the nozzles were 4.5, 4, 2.5 and 1.5 mm. The results in Fig. 6 showed that when the total flow and GLR were the same, the smaller the nozzle diameter, the more complete the liquid breaks. When the exit diameter of the nozzle was 4.5 mm as Fig. 6(a) showed, the liquid column flowed down and hardly cracked. Only a few large droplets appeared around the liquid column. When the exit diameter of the nozzle was 4 mm as Fig. 6(b) showed, the central liquid column broke into several small liquid columns and some large continuous structures. There were still not many droplets. When the exit diameter of the nozzle was reduced to 2.5 mm as Fig. 6(c) showed, there was a large degree of rupture in the liquid column. The small ligaments became the main fluid structure. Many small droplets appeared on the edge of the spray. When the exit diameter of the nozzle was 1.5 mm as Fig. 6(d) showed, the liquid broke into smaller structures. The spray was full of dense and tiny droplets. Because when the nozzle diameter was small, the liquid flow rate was large. In this case, the influence of inertia on the flow field was greater than the viscous force, and the fluid flow was so unstable that easy to be broken by the gas flow. Small changes in flow rate caused by the gas flow developed and enhanced, creating the irregular turbulence.
Fig. 6. Spray morphology under same flow conditions but different exit diameter of the nozzle (D): (a) D = 4.5 mm; (b) D = 4 mm; (c) D = 2.5 mm; (d) D = 1.5 mm.
4. Conclusion
In conclusion, femtosecond shadowgraphy and supercontinuum (SC)-illumination imaging methods were compared to image liquid-gas sprays. The Modulation Transfer Function (MTF) and Point-Spread Function (PSF) of the two imaging systems were quantitatively calculated for imaging a resolution test target. The results showed that for static targets, these two methods had almost the same imaging capabilities. From the simultaneous comparisons in liquid-gas mixed sprays, it can be seen that the edges of many spray structures were eroded by speckles and lots of tiny droplets were submerged in the speckles by using femtosecond shadowgraphy. While the finer droplets, shaper edges of ligaments, and more-realistic liquid sheets were captured using SC-illumination imaging method by suppressing speckles. The dynamics of the liquid-gas mixed spray were further analyzed using SC-illumination imaging. By quantitatively changing the parameters we found that the degree of rupture of the spray was affected regularly by the flow ratio of gas to liquid (GLR), the total flow and the exit diameter of the nozzle. When other conditions were the same, the higher GLR, the larger total flow and the smaller nozzle diameter, the more completely the spray were broken.
Funding
National Natural Science Foundation of China (61427816, 61690221); Fundamental Research Funds for the Central Universities (xzy012019039); Natural Science Basic Research Plan in Shaanxi Province of China (2018JM6012).
Disclosures
The authors declare no conflicts of interest.
References
1. S. Chu and A. Majumdar, “Opportunities and challenges for a sustainable energy future,” Nature 488(7411), 294–303 (2012). [CrossRef]
2. M. Fischer, M. Werber, and P. V. Schwartz, “Batteries: Higher energy density than gasoline?” Energy Policy 37(7), 2639–2641 (2009). [CrossRef]
3. S. Chu, “The science of photons to fuel,” in AIP Conference Proceedings (AIP, 2008), pp. 266–282.
4. H. Purwar, H. Wang, M. Tang, S. D. Idlahcen, C. Rozé, J. Blaisot, T. Godin, and A. Hideur, “Ultrafast high-repetition imaging of fuel sprays using picosecond fiber laser,” Opt. Express 23(26), 33396–33407 (2015). [CrossRef]
5. J. C. Dainty, Laser Speckle and Related Phenomena (Springer-Verlag, 1984).
6. L. Fabbrini, M. Greco, M. Messina, and G. Pinelli, “Improved Edge Enhancing Diffusion Filter for Speckle-Corrupted Images,” IEEE Geosci. Remote Sensing Lett. 11(1), 99–103 (2014). [CrossRef]
7. H. G. Cui, G. M. Huang, T. Liu, and J. Gao, “Speckle Filtering Algorithm of PolSAR Imagery Based on Two-Dimensional Polarimetric Subspace ICA,” in Proceedings of the 6th International Asia Conference on Industrial Engineering and Management Innovation (Atlantis Press, 2016), pp. 343–353.
8. B. Redding, M. A. Choma, and H. Cao, “Speckle-free laser imaging using random laser illumination,” Nat. Photonics 6(6), 355–359 (2012). [CrossRef]
9. N. E. Yu, J. W. Choi, H. Kang, D. Ko, S. Fu, J. Liou, A. H. Kung, H. J. Choi, B. J. Kim, and M. Cha, “Speckle noise reduction on a laser projection display via a broadband green light source,” Opt. Express 22(3), 3547–3556 (2014). [CrossRef]
10. 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]
11. P. Karra, T. Rogers, and P. Lappas, “Air entrainment in gaseous fuel jets using particle image velocimetry and high speed schlieren photography in a constant volume chamber,” in Proceedings of SAE 2015 World Congress & Exhibition (SAE Technical Paper, 2015).
12. J. Manin, M. Bardi, L. M. Pickett, and R. Payri, “Boundary condition and fuel composition effects on injection processes of high-pressure sprays at the microscopic level,” Int. J. Multiphase Flow 83, 267–278 (2016). [CrossRef]
13. O. A. Kuti, J. Zhu, K. Nishida, X. Wang, and Z. Huang, “Characterization of spray and combustion processes of biodiesel fuel injected by diesel engine common rail system,” Fuel 104, 838–846 (2013). [CrossRef]
14. C. Willert, S. Freitag, and C. Hassa, “High speed imaging of fuel spray using a low-cost illumination source,” in Proceedings of 22nd European Conference on Liquid Atomization and Spray Systems (ILASS, 2008), pp. 8–10.
15. C. D. McGuinness, K. Sagoo, D. McLoskey, and D. Birch, “A new sub-nanosecond LED at 280 nm: application to protein fluorescence,” Meas. Sci. Technol. 15(11), L19–L22 (2004). [CrossRef]
16. M. Nakazawa, K. Tamura, H. Kubota, and E. Yoshida, “Coherence degradation in the process of supercontinuum generation in an optical fiber,” Opt. Fiber Technol. 4(2), 215–223 (1998). [CrossRef]
17. H. Zhang, Z. Zhou, A. Lin, J. Cheng, L. Yan, J. Si, F. Chen, and X. Hou, “Chirp structure measurement of a supercontinuum pulse based on transient lens effect in tellurite glass,” J. Appl. Phys. 113(11), 113105 (2013). [CrossRef]
18. Y. Zheng, J. Si, W. Tan, Y. H. Ren, J. Tong, and X. Hou, “Speckle-suppressed full-field imaging through a scattering medium using a supercontinuum,” Opt. Express 24(23), 26338–26343 (2016). [CrossRef]
19. G. Cheng, R. Cohn, C. Johnson, R. Davis, and J. Muss, “Development of GOX/kerosene swirl-coaxial injector technology,” in Proceedings of 39th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit (AIAA, 2003), p. 4751.
20. M. Paciaroni and M. Linne, “Single-shot, two-dimensional ballistic imaging through scattering media,” Appl. Opt. 43(26), 5100–5109 (2004). [CrossRef]
21. J. N. Kapur, P. K. Sahoo, and A. K. Wong, “A new method for gray-level picture thresholding using the entropy of the histogram,” Comput. Graphics Image Process. 29(3), 273–285 (1985). [CrossRef]
