Open Access
Add to BibSonomy
Abstract
We demonstrate a non-contact spatiotemporally resolved comprehensive method for gas flow velocity field measurement: Filamentary Anemometry using femtosecond Laser-extended Electric Discharge (FALED). A faint thin plasma channel was generated in ambient air by focusing an 800-nm laser beam of 45 fs, which was used to ignite a pulsed electric discharge between two electrodes separated over 10 mm. The power supplier provided a maximum voltage up to 5 kV and was operated at a burst mode with a current duration of less than 20 ns and a pulse-to-pulse separation of 40 μs. The laser-guided thin filamentary discharge plasma column was blowing up perpendicularly by an air jet placed beneath in-between the two electrodes. Although the discharge pulse was short, the conductivity of the plasma channel was observed to sustain much longer, so that a sequence of discharge filaments was generated as the plasma channel being blown up by the jet flow. The sequential bright thin discharge filaments can be photographed using a household camera to calculate the flow velocity distribution of the jet flow. For a direct comparison, a flow field measurement using FLEET [J. B. Michael, Appl. Opt . 50, 5158 (2011)] was also performed. The results indicate that the FALED technique can provide instantaneous nonintrusive flow field velocity measurement with good accuracy.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Velocity is one of the key parameters that characterize a flow field. Hence, it is an important task to develop velocimetry techniques, among which molecular tagging velocimetry (MTV) has received considerable attention owing to its less intrusiveness to the flow field, compared to the probe methods such as hot wire and pitot tube, and its more accurate tracking ability of the flow field, compared to the particle-seeding methods such as particle image velocimetry [1].
In an MTV measurement, velocity is determined based on the movement within a known period of time of the tagged molecules, where tagging technique is the core issue. Laser and electron beam are two typical tagging sources and have been broadly used. Various laser-based MTV techniques have been developed and applied successfully, e.g. RELIEF [2], VENOM [3,4], OTV [5], APART [6], and HTV [7]. Detailed description of these methods could be found in Ref [8]. These optical methods are experimentally complicated, involving frequently seeding molecules to the flow field and using often two lasers with one tagging the molecules and the other “reading” the tagged molecules after the displacement. As an attempt to lift those limitations, femtosecond (fs) laser electronic-excitation tagging (FLEET) was developed [8], followed by an updated version called STARFLEET [9]. Both FLEET and STARFLEET use the special properties of the ultra-high peak power of a focused femtosecond laser to generate the so-called filament-induced nonlinear spectroscopy (FINS) [10] for molecular tagging. Like other MTV techniques, FLEET and STARFLEET suffer from weak signal, and thus an intensified (ICCD) camera is needed to follow the movement of the tagged volume; besides, the decay of the FLEET and STARFLEET signals are fast, typically on the order of microseconds [11], which makes it mainly suitable for flow fields of high velocity. Electron beam is another prominent tagging source. Velocimetry technique based on electron beam-induced fluorescence [12] has been invented long time ago, but being mostly limited to low-density and high-speed flows, e.g. the studies of vehicles’ re-entry through the upper atmosphere.
Electric discharge might be an alternative for molecular tagging in atmospheric pressure. In our previous works [13,14], we found that the plasma channel generated by a gliding arc discharge, which could be considered as tagged volume in the concept of MTV, could follow the gas flow with its decay time of hundreds of microseconds. However, the discharge cannot be accurately controlled neither spatially nor temporally, since the path and time of the discharge is highly susceptible to the flow field. Besides, breakdown of the air of a relatively long distance imposes high demands to the power supplier, and the discharge under this condition might be so drastic that it makes the non-intrusiveness premise of the measurement vain. Hence, in order to apply this concept to the flow field for velocity measurements, those issues are to be addressed. Recently, it was found that fs laser-induced filament could be used to guide a discharge [15,16] and could even mitigate the discharge [17]. In this way, the discharge could be accurately regulated both spatially and temporally.
In this work, we developed a new velocimetry technique for gas flow field measurements: Filamentary Anemometry using femtosecond Laser-extended Electric Discharge (FALED). Femtosecond laser filamentation was used to guide and control the electric discharge, and the thin streamer of the faint plasma channel was used to visualize the flow field.
2. Experiments
Figure 1 illustrates a schematic of the FALED experimental setup, which consists of a gas jet, an fs laser, an electric discharge system, and a detection system. The jet nozzle, as shown in the photo, is a glass tube with an inner diameter of 2.5 mm and was supplied with dry air regulated by a mass flow controller. The laser source was the fundamental output from an fs Ti:sapphire laser (Spitfire Ace, Spectra-Physics), 800 nm in wavelength, 45 fs in pulse duration, 6 mJ in pulse energy. The repetition rate of the laser is adjustable up to the highest frequency of 1 kHz. The laser was focused by a spherical lens (f = 500 mm), and a visible laser-induced filament (or, rather, a weak visible plasma channel) as long as 15 mm was formed.
Fig. 1 A schematic of the experimental setup, along with a typical FALED photo taken by an SLR camera with an exposure time of 1/10 second to cover the burst of 6 sequential discharges with 40 μs the pulse separation time.
The electric discharge system consists of a high voltage direct current (DC) pulsed power supplier (MINISTROBOKIN 20, HSPS), a current limiting resistor, and a pair of electrodes. The peak output voltage is 5 kV. The interval between two consecutive discharges is adjustable from 40 µs to longer (in burst mode), and the number of discharges in a burst can be varied up to 8 pulses. The high voltage terminal and the ground terminal of the power supplier were connected with two electrodes. The ground terminal connected in series with the current limiting resistor to control the discharge current. The electrodes are cone-shaped, as shown in the photo, and the distance between the two tips was set at 10 mm to prevent any free discharge at 5 kV. When we synchronize the laser pulse to the first discharge pulse of the burst and place the laser-induced filament close to the two tips, discharge streamer can be initiated by the weakly ionized plasma channel along the filament.
The detection system is as follows. The scattering of the laser pulse and the emission of the discharge pulses were detected by a fast photodiode. The voltage output from the power supplier and the current detected by a current probe (CT-1, Tektronix) were simultaneously recorded in a 600 MHz oscilloscope. The emission spectra of the plasma were measured using a spectrometer (Acton 2300i,Princeton Instrument). The emission from the plasma was imaged onto the entrance slit, which was vertically oriented to allow a spatial resolution along the gas flow direction. The photo in Fig. 1 was taken using a single lens reflex (SLR) camera (D90, Nikon) with an exposure time of 1/10 second. It can be seen that the path of the first filamentary discharge (labelled as “1”) can be strictly defined by the laser filament. The plasma channel of the filamentary discharge can sustain for a period of time when drifting downstream following the flow field. If the plasma channel survives until the arrival of the next high voltage peak, another discharge can be ignited through the same thin path. That was the case as shown in the photo in Fig. 1, where the initial plasma channel was discharged repeatedly for another 5 times. As the displacement can be accurately measured using this photo, flow field velocity can be evaluated.
3. Results and discussion
The pulsed power supplier was externally triggered and ran at its burst mode. A pre-trigger TTL signal from the laser system was used to trigger the burst of the high voltage pulses to achieve the synchronization with the laser pulse. The temporal sequence of various events is shown in Fig. 2, i.e. the formation of laser-induced filament, the emission of the discharges between the electrodes, the voltage waveform applied to the electrodes, and the current during a FALED measurement. A typical FALED photo consisting 4 sequential discharges is also shown in Fig. 2(a) as an inset, and each bright wire is sequentially numbered (labelled from 1 to 4) to the corresponding discharge.
Fig. 2 Temporally resolved voltage (black lines), current (blue dotted lines) and optical emission (red lines) profiles, recorded during a typical FALED measurement including four sequential discharges (a). The inset photo was taken with an exposure time covering all four discharges, and each bright wire was sequentially numbered to the corresponding discharge. A zoom in (b) shows details of the laser scattering and the first induced discharge.
Figure 2(b) shows a zoom in of the narrow time window in Fig. 2(a) (indicated by a green dashed box), covering the laser pulse and the first discharge. It can be seen that after receiving the pre-trigger signal from the laser, the voltage began to increase gradually. When the voltage increased to ~2 kV, the laser pulse arrived (the first sharp peak in the signal profile from laser scattering detected by the photodiode), generating a filament. The filament did not induce the discharge immediately. After the laser pulse the voltage kept increasing, and about 0.6 μs later, when it reached 5 kV, the first discharge happened (the second peak of the photodiode signal). The discharge peak monitored by the photodiode had a duration of less than 100 ns, which means that the strong visible emission from the discharge lasted for less than 100 ns. Along with the discharge, the voltage dropped sharply, and on the contrary the current rose steeply. The following drastic wrinkling of the voltage does not contribute to the plasma emission and can be regarded as electronic noise from internal reflections introduced by the sharp current/voltage variation. In order to optimize the first discharge, the delay time between the laser pulse and the voltage pulse had been adjusted. It was found that the optimal value was 0.6 μs for generating the most stable discharge. The phenomenon is rather interesting and is believed to be stemmed from the property of the weak plasma generated from the fs laser-induced filament. The governing mechanism of gas discharge in this regime is too complicated to explain the details. One reasonable interpretation might be: following the laser excitation, free electrons generated from multi-photon ionization in the filament are consumed shortly; part of the absorbed laser energy storing in the excited particles will start to thermalize through collisions; at a delay time of 0.6 μs, a hot thin filamentary volume is formed containing negative/positive ions and metastable electronically excited nitrogen molecules, which forms a pathway easier to be discharged through; by that time, the voltage just approach the maximum to enable an efficient discharge.
Although from the discharge peak it seems that the strong light emission from the plasma channel lasts only ~100 ns, our previous studies [13,18–20] indicate that the lifetime of the plasma channel is much longer than that: the faint detectable emission from the thin plasma column was observed to last hundreds of microseconds after the electric current shut-off. Wang et al. [21] also reported a lifetime of an fs laser-guided glow discharge to be a few microseconds. Hence, in our case the wire-shaped plasma channel generated by the first discharge will survive long enough to host another discharge that, in this case, will arrive 40 μs later. As the plasma channel drifts following the flow field, each discharge will “re-emerge” in the space showing its instantaneous position. After a burst of discharges, the movement of the channel will be depicted, and the velocity information of the flow field can be extracted. A series of short and bright flashes of the burst enables the detection of the flow speed with un-intensified cameras.
A spatially resolved spectral image of a burst of FALED discharges (three discharges in total) was recorded and shown in Fig. 3(a). The exposure time of the ICCD was 150 μs and the gain was 95. The discharges share the same spectral profile, though with a small variation in intensity. The spectral curve of the first discharge was extracted and presented in Fig. 3(b). As comparisons, the FINS (fs laser filament-induced nonlinear spectroscopy) plasma emission was recorded with the same optical system while turning off the electric discharge and is presented in Fig. 3(c). The exposure time of the ICCD was 1 μs and the gain was 95. It can be seen that the integrated emission of FALED is one order of magnitude stronger than that of the FINS. The current limiting resistor was set at 3 kΩ in this case, and the FALED emission intensity can be sensitively controlled by vary the impedance of the resistor. The spectrum of a free discharge between the electrodes (the distance between the two electrodes in this case was 1 mm) driven solely by the high voltage was also recorded and presented in Fig. 3d. Since the free discharge was much stronger, a stack of neutral filters was used to reduce the emission intensity.
Fig. 3 Emission spectra from the filamentary column of different plasmas collected with spectrometer slit vertically orientated perpendicular to the thin plasma columns. a) Image spectra of FALAD of three consequential discharges along the jet flow; b) spectral curve integrated over the first discharge shown in a); c) spectral curve of emission from the filamentary plasma introduced by the fs laser only (FINS); d) spectral curve of a free electric discharge (with closely placed electrodes, but no fs laser ignition).
There are interesting details in the recorded spectra as shown in Fig. 3, which reflects the different mechanisms of the observed thin plasma columns. FALED emission is generated by the pulsed electric discharge. Free electrons are accelerated by the external electric field, and they collide mostly with N2 molecules and generate fluorescence from the N2 C-B and B-A transitions, where the individual vibration bands of the C-B transition are resolved and the wavelength are marked in Fig. 3(b). FINS emission is generated by the high peak power fs laser through multiphoton processes. Both the observed spectral curves from the FALED and the FINS are dominated by N2* and N2+ bands with the emission of a typical non-thermal plasma. The N2 B-A transitions in the FALED is much stronger than the C-B transition compared with FINS, and the N2* and N2+ bands in FALED spectrum are broader also. The FALED signal is strong enough to be imaged by cameras without an intensifier.
The spectrum of the free discharge is completely distinct from FALED and FINS. The ionization and dissociation are more drastic, resulting in a much stronger plasma. The spectrum is dominated by atomic lines from N+, and molecular bands hardly stand out. The spectrum also features a strong continuum background of Bremsstrahlung radiation, a typical feature of thermal plasma. During a free discharge, according to the manufacturer of the high voltage power supplier, a maximum electric pulse energy as much as 25 mJ can be released, and this amount of energy will cause intrusiveness to the flow field. For FALED, however, the intrusiveness is substantially reduced. By comparing the integrated absolute optical emission between Fig. 3(b) and Fig. 3(d), we roughly estimated that for a single discharge in FALED the energy deposition to the flow field is about tens of μJ. Besides, based on the voltage and current profiles as shown in Fig. 2(b), we estimated that even if we ignore the energy consumption by the limiting resistor in the circuit, the energy deposition for a single discharge in FALED to the flow field is about 400 μJ at most. We also measured the energy deposition of an fs laser pulse in FINS by measuring the laser pulse energy before and after the forming of filament, which is ~10% of the excitation laser pulse energy, ~600 μJ in our case. By employing a deep UV laser beam, STARFLEET [9] can be achieved using laser pulses of only 30 μJ. The heating effects are clearly inevitable for any MTV techniques when tagging the molecules, both for FALED and for FLEET. The influence to the accuracy of the flow speed measurements is to be considered for different cases. It is noteworthy that FALED provides much brighter tracer with similar energy deposition, which provides potential space for reducing the intrusiveness.
To evaluate the performance of FALED for velocimetry, a direct comparison between FALED and FLEET was carried out. The FLEET signal in an air jet after a delay of 40 μs is too weak to be detected. To enhance the FLEET signal [22], the gas jet was supplied with pure nitrogen with 500 ppm CH4 additives instead of air for the FLEET measurements. Using the ICCD camera (gate delay = 40 μs, gate width = 1 μs), a typical single-shot FLEET image is shown in Fig. 4(a).
Fig. 4 FLEET and FALED image with a gas flow speed of 35 m/s: (a) FLEET image with a delay time of the ICCD camera gate of 40 µs after the laser pulse; (b) FALED photo taken by the SLR camera of four sequential discharges with 40 μs separation; (c) a direct comparison between the FLEET and FALED.
A typical FALED photo measured under the same flow field condition is shown in Fig. 4(b). A direct comparison between them is demonstrated in Fig. 4(c) where an excellent match was revealed. It indicates that the FALED can provide flow speed measurement similar as FLEET. Furthermore, the wire-shaped channel of FALED is much thinner (due to the less than 100 ns blinking time as shown in Fig. 2(b)) than that of FLEET, which might be useful in reducing the uncertainty in velocity evaluation. Besides, FALED needs neither an intensified camera, nor a camera with precise control of the exposure gate. Even a household camera is good enough for FALED measurements, which largely reduces the burden on the detector side.
Proof-of-principle measurements using FALED were carried out under various conditions, and the results are shown in Fig. 5. The photos in the upper row were taken with the separation time between two consecutive discharges being set to 40, 80, and 120 μs, while the gas supply speed was kept constant at 15 m/s. The results indicate that the plasma channel has a memory as long as 120 μs at least, which give the possibility for slow flow field measurement. Those in the middle row were recorded with the gas supply speed being set to 15, 25, and 35 m/s, while the separation time was kept constant at 40 μs. The recorded images show the difference in the flow pattern as the flow speed increase. Those in the lower row were recorded with the height above the nozzle being set to 3, 7, and 11 mm, while the separation time was kept at 40 μs, and gas supply at 35 m/s. The recorded images show that the jet waist broadens as moving away from the jet nozzle. It can also be seen that with longer separation time or higher gas supply speed, the distance between two consecutive discharge wires gets bigger, and that with higher height above the nozzle the distance gets smaller due to the momentum loss of the gas jet at downstream.
Fig. 5 FALED images obtained at varied conditions: upper row, different pulse separation time at constant flow of 15 m/s; middle row, different flow speed at constant pule separation time of 40 μs; lower row, different height above the nozzle rim at constant pulse separation time of 40 μs and flow speed of 35 m/s.
4. Summary
We found that filamentary discharge can be initiated by laser-induced wire-shaped plasma column through synchronizing the outputs of the fs laser and the high voltage power supplier. The path and time of this filamentary discharge can be strictly defined by the laser. The output power of the discharge can be regulated by varying the impedance of the current limiting resistor. The wire-shaped channel can be discharged repeatedly while drifting downstream. The burst of discharges emits strong light at relatively low energy deposition, and an SLR camera without being precisely synchronized with the laser or high voltage pulse can be used to record the signal. The results indicate that this technique, named FALED (Filamentary Anemometry using femtosecond Laser-extended Electric Discharge) can find general applications in flow velocity measurements. We are presently working on finding a proper algorithm and programming to convert the FALED image to velocity field by dealing with the vertical and tangential part of sequential thin filaments.
It is worthy to note that the upper limit of the flow velocity measurable with the present FALED setup is restricted by the shortest pulse separation between the pulses in the high voltage burst. With the availability of a faster burst, FALED is even in favour for high speed measurements. Similar to all the other MTV techniques, the perturbation of FALED to the flow due to the energy deposition when tagging the molecules is to be quantitatively analysed when applied it to different practical applications, regarding the scale to be resolved and accuracy to be achieved.
Funding
National Natural Science Foundation of China (NSFC) (91741205, 51776137).
References
1. R. Huffman and G. Elliott, “An experimental investigation of accurate particle tracking in supersonic, rarefied, axisymmetric jets,” 47th AIAA Aerospace Sciences Meeting and Exhibit, AIAA 2009–1265, Oriando, Florida. [CrossRef]
2. R. B. Miles, J. Grinstead, R. H. Kohl, and G. Diskin, “The RELIEF flow tagging technique and its application in engine testing facilities and for helium-air mixing studies,” Meas. Sci. Technol. 11(9), 1272–1281 (2000). [CrossRef]
3. A. G. Hsu, R. Srinivasan, R. D. W. Bowersox, and S. W. North, “Molecular tagging using vibrationally excited nitric oxide in an under expanded jet flow field,” AIAA J. 47(11), 2597–2604 (2009). [CrossRef]
4. R. Sánchez-González, R. Srinivasan, R. D. W. Bowersox, and S. W. North, “Simultaneous velocity and temperature measurements in gaseous flow fields using the VENOM technique,” Opt. Lett. 36(2), 196–198 (2011). [CrossRef] [PubMed]
5. L. A. Ribarov, J. A. Wehrmeyer, F. Batliwala, R. W. Pitz, and P. A. Debarber, “Ozone tagging velocimetry using narrowband excimer lasers,” AIAA J. 37(6), 708–714 (1999). [CrossRef]
6. S. Nandula, R. Pitz, J. Bominaar, C. Schoemaecker, N. Dam, and J. T. Meulen, “Kinetics of NO tag formation in air for unseeded molecular tagging velocimetry,” 42nd AIAA Aerospace Sciences Meeting and Exhibit, AIAA 2004–390, Reno, Nevada. [CrossRef]
7. M. D. Lahr, R. W. Pitz, Z. W. Douglas, and C. D. Carter, “Hydroxyl-tagging-velocimetry measurements of a supersonic flow over a cavity,” J. Propuls. Power 26(4), 790–797 (2010). [CrossRef]
8. J. B. Michael, M. R. Edwards, A. Dogariu, and R. B. Miles, “Femtosecond laser electronic excitation tagging for quantitative velocity imaging in air,” Appl. Opt. 50(26), 5158–5162 (2011). [CrossRef] [PubMed]
9. N. Jiang, B. R. Halls, H. U. Stauffer, P. M. Danehy, J. R. Gord, and S. Roy, “Selective two-photon absorptive resonance femtosecond-laser electronic-excitation tagging velocimetry,” Opt. Lett. 41(10), 2225–2228 (2016). [CrossRef] [PubMed]
10. H. L. Xu, Y. Cheng, S. L. Chin, and H. B. Sun, “Femtosecond laser ionization and fragmentation of molecules for environmental sensing,” Laser Photonics Rev. 9(3), 275–293 (2015). [CrossRef]
11. M. R. Edwards, A. Dogariu, and R. B. Miles, “Simultaneous temperature and velocity measurements in air with femtosecond laser tagging,” AIAA J. 53(8), 2280–2288 (2015). [CrossRef]
12. A. K. Mohamed, J. Bonnet, S. Larigaldie, T. Pot, J. Soutadé, and B. Diop, “Electron beam fluorescence in hypersonic facilities,” AerospaceLab1–9 (2009).
13. Z. W. Sun, J. J. Zhu, Z. S. Li, M. Aldén, F. Leipold, M. Salewski, and Y. Kusano, “Optical diagnostics of a gliding arc,” Opt. Express 21(5), 6028–6044 (2013). [CrossRef] [PubMed]
14. J. J. Zhu, J. L. Gao, A. Ehn, M. Aldén, Z. S. Li, D. Moseev, Y. Kusano, M. Salewski, A. Alpers, P. Gritzmann, and M. Schwenk, “Measurements of 3D slip velocities and plasma column lengths of a gliding arc discharge,” Appl. Phys. Lett. 106(4), 044101 (2015). [CrossRef]
15. P. Rambo, J. Schwarz, and J. C. Diels, “High-voltage electrical discharges induced by an ultrashort-pulse UV laser system,” J. Opt. A, Pure Appl. Opt. 3(2), 146–158 (2001). [CrossRef]
16. P. Polynkin, “Multi-pulse scheme for laser-guided electrical breakdown of air,” Appl. Phys. Lett. 111(16), 161102 (2017). [CrossRef]
17. E. Schubert, D. Mongin, J. Kasparian, and J. P. Wolf, “Remote electrical arc suppression by laser filamentation,” Opt. Express 23(22), 28640–28648 (2015). [CrossRef] [PubMed]
18. J. J. Zhu, Z. W. Sun, Z. S. Li, A. Ehn, M. Aldén, M. Salewski, F. Leipold, and Y. Kusano, “Dynamics, OH distributions and UV emission of a gliding arc at various flow-rate investigated by optical measurements,” J. Phys. D Appl. Phys. 47(29), 295203 (2014). [CrossRef]
19. J. J. Zhu, J. L. Gao, A. Ehn, M. Aldén, Z. S. Li, A. Larsson, and Y. Kusano, “Spatiotemporally resolved characteristics of a gliding arc discharge in a turbulent air flow at atmospheric pressure,” Phys. Plasmas 24(1), 013514 (2017). [CrossRef]
20. J. Zhu, A. Ehn, J. Gao, C. Kong, M. Aldén, M. Salewski, F. Leipold, Y. Kusano, and Z. Li, “Translational, rotational, vibrational and electron temperatures of a gliding arc discharge,” Opt. Express 25(17), 20243–20257 (2017). [CrossRef] [PubMed]
21. T. J. Wang, Y. Wei, Y. Liu, N. Chen, Y. Liu, J. Ju, H. Sun, C. Wang, H. Lu, J. Liu, S. L. Chin, R. Li, and Z. Xu, “Direct observation of laser guided corona discharges,” Sci. Rep. 5(1), 18681 (2016). [CrossRef] [PubMed]
22. B. Li, D. Y. Zhang, X. F. Li, Q. Gao, Z. F. Zhu, and Z. S. Li, “Femtosecond laser-induced cyano chemiluminescence in methane-seeded nitrogen gas flows for near-wall velocimetry,” J. Phys. D Appl. Phys. 51(29), 295102 (2018). [CrossRef]
