1. Introduction

Ignition of a flammable premixture using a laser beam has several advantages compared to ignition using a conventional spark plug. For example, the ignition location can be chosen freely and the flame kernel is not affected by heat loss to the plug. Laser ignition is expected to be a reliable technology to ignite fuel-lean mixtures, which are important for improving the thermal efficiency and decreasing the emission of nitrogen-oxides in gas turbines and internal combustion engines. Beduneau et al. investigated the emission spectra of sparks from laser-induced breakdown to specify the key parameter for the transition to a self-sustaining flame [1]. The results showed that sufficient radical levels were required for the self-sustaining flame. However, the details of the flame kernel development process have not yet been sufficiently clarified.

The basic behavior of laser-induced breakdown and ignition has been summarized in ref. [2]. When laser pulsed light with sufficient energy is focused on a gas, high-temperature and high-pressure plasmas are generated by multi-photon absorption and successive cascading processes. Then, toroidal vortexes are formed at the tips of the plasmas when a rarefaction wave passes through the plasma region. In addition, a third lobe is formed in the upstream direction of the laser light [2]. The generation processes of the toroidal vortexes and the third lobe were demonstrated by numerical calculations [3, 4], in which an asymmetric plasma distribution causes the characteristic gas motions.

As seen in OH-PLIF images obtained by Lackner et al. [5], for breakdown ignition in a flammable mixture, the third lobe enhances the initial flame development; however, in the case of a lean mixture close to the flammability limit, it can lead to misfiring due to flame stretch. In fact, according to Phuoc et al., laser-induced spark ignition fails to ignite a methane–air mixture of less than 6.5% methane, which is richer than the lower flammability limit [6]. It is thought that a leaner mixture can be ignited stably by appropriately controlling the flow of the hot gas kernel.

In this study, we attempted to control the plasma formation process and the flow by providing electrons from the small breakdown of a femto-second Ti:Sapphire (TiS) laser focused near the focal point of a Nd:YAG laser. By using TiS laser, electrons can be provided with little energy, and almost all energy in a plasma is provided only by the YAG laser. The changes to plasma distribution and successive flow were discussed. Furthermore, their effects on the ignition of a flammable mixture were investigated.

2. Experimental setup

Laser-induced breakdown plasmas were generated using a nano- and a femto-second pulsed laser beam, and the plasma and flows were visualized. The schematic of the optical arrangement is shown in Fig. 1. The main laser was a Q-switched Nd:YAG laser at a wavelength of 532 nm, and the beam was focused with a lens of focal length 100 mm. A mode-locked TiS laser with pulse duration 150 fs at wavelength 745 nm was synchronized with the YAG laser and focused with a lens of focal length 70 mm. The TiS laser pulse was shot approximately 10 ns before the YAG laser pulse to supply initial electrons to the YAG laser path around the focal point of the YAG laser beam. Here, the lifetime of the plasma produced by the TiS laser is tens of nanoseconds and the timing of TiS laser pulse is selected so that the absorption energy of YAG laser becomes maximum. The pulse energy of the YAG laser was varied using a half waveplate and a polarizer. Plasma images were captured by an image-intensified ultra-high-speed camera (nac ULTRA Neo) at a frame interval of 5 ns. For visualizing the gas flows, schlieren photography was employed with a high-speed camera (Photron FASTCAM SA1.1) at a frame rate of 75000 fps and an exposure of 3 μs. Approximately 10% of incident laser light was reflected by the quartz glass to measure the incident laser energy for every shot. The relation of the reflected energy and incident energy into the vessel was calibrated in advance. At the same time, the transmitted energy was also measured. Then, absoption energy was calculated by substracting the transmitted energy from incident energy. The pulse energy of the TiS laser was 4–8 mJ, which does not influence the result shown in this paper.

 

Fig. 1 Optical arrangement

Download Full Size | PDF

Experiments were performed in a vessel equipped with five quartz windows. The vessel was filled with air or a CH₄–air mixture and the laser pulses were focused on gas at room temperature without any flow. For the ignition tests, the pressure in the vessel was measured with a strain gauge pressure transducer attached on the top of the vessel and the success of the ignition was determined based on the pressure change.

3. Results and Discussion

3.1. Plasma formation by nano- and femto-second lasers

To investigate the effect of the TiS laser on laser-induced breakdown, the breakdown threshold energy of the YAG laser was first examined. Figure 2 shows the breakdown threshold energy in air for several pressures in the vessel. The breakdown threshold energy for the YAG laser without the TiS laser was determined based on a ICCD camera image. A minimum energy for which the ICCD camera captured a plasma image is the threshold. In the case of the combination of the TiS and YAG lasers, the breakdown threshold energy was determined based on a absorption energy of YAG laser pulse. With breakdown, absoption energy of YAG laser is positive, and without breakdown, it is zero. The focusing position of the TiS laser was set at the focal point of the YAG laser. In the case of the breakdown by the YAG laser without the TiS laser, lower air pressure requires higher YAG laser energy, which is consistent with other works [2, 7]. In the case of the combination of the TiS and YAG lasers, the tendency for the pressure is the same as in the case without the TiS laser, and the threshold energies drastically decrease. This indicates that the TiS laser influences and promotes breakdown by the YAG laser pulse.

 

Fig. 2 Breakdown threshold energy of air at different pressures

Download Full Size | PDF

Next, the relation between the incident energy and absorption energy of the YAG laser beam at a pressure of 0.1 MPa is shown in Fig. 3. Here, the absorption energy was derived using the incident energy and transmitted energy. For shots without the TiS laser, breakdown does not occur for incident energy values less than 40 mJ. The absorption energy increases with values of incident energy larger than 40 mJ. With the TiS laser, absorption can be observed even for very low incident energy; absorption energy increases with the incident energy, which is the same tendency as the case without the TiS laser.

 

Fig. 3 Relation between incident energy E_in and absorption of YAG laser pulse E_ab

Download Full Size | PDF

With the TiS laser, because a small amount of plasma is already produced, a high energy density for multi-photon absorption and successive cascading processes is not required. Thus, the breakdown threshold becomes much lower than in the case without TiS laser; in addition, the YAG laser beam is effectively absorbed. In the case without the TiS laser, a lot of YAG laser energy passes through the focal point before breakdown occurs.

To discuss the plasma development process with the TiS laser, ultra-high-speed direct imaging data are shown in Fig. 4. The plasma distributions vary depending on the YAG laser incident energy E_in and TiS laser incident position x_TiS. In the case of YAG incident energy of 10 mJ, plasma is formed in with-TiS cases, while it is not formed in the without-TiS case because of the very low incident energy. The plasmas are produced around the TiS incident position, which means that the TiS laser successfully promotes and controls the breakdown. For the higher YAG laser energy of 40 mJ, plasmas are also initially generated at the TiS laser incident position. After that, plasma expands around the initial plasma in the x_TiS = −1 mm case and extends to the upstream direction of the YAG laser in the x_TiS = 1 mm case. For an even higher YAG laser energy of 70 mJ, as observed in other cases, plasmas initially form at the TiS laser incident position. In the x_TiS = −1 mm case, plasma expands around the initial plasma and a weak plasma is focused toward the focal point of the YAG laser. For both the x_TiS = 0 and 1 mm cases, plasmas extend to the upstream direction of the YAG laser; however, there are different plasma distributions: bimodal distribution can be observed in the x_TiS = 1 mm case. The right side of the plasma is presumed to be generated by the TiS laser, while the left side of plasma would be formed independent of the TiS breakdown, because the center of the plasma is weak and the YAG laser energy is sufficient to enable breakdown without the TiS laser.

 

Fig. 4 Plasma visualization images by ultra-high-speed ICCD camera

Download Full Size | PDF

Next, the gas motion generated after breakdown is discussed. The high-speed schlieren images are shown in Fig. 5. Each image set is taken for the same shots as the direct image set shown in Fig. 4. Without the TiS laser, a jet, which is called a “third-lobe”, and toroidal rings are formed in the upstream direction of the YAG laser in both the E_in = 40 and 70 mJ cases. According to Bradley et al., these flows occur due to the displaced plasma center to upstream of the YAG laser [2]. High density plasma is distributed on the left side of the plasma and the centers of the shock wave and rarefaction wave are upstream of the YAG laser; then, the pressure gradient from the right edge of the plasma to the center becomes large when the rarefaction wave passes through the plasma. This pressure gradient produces the gas motion.

 

Fig. 5 High-speed schlieren images (Each image set is the same shot as Fig. 4)

Download Full Size | PDF

When using the TiS laser for a low incident energy of E_in = 10 mJ, no jet flow is observed. Also, by the setting TiS laser incident position to x_TiS = 0 mm, no jets are generated. According to Fig. 4, which corresponds to these cases, the plasma distributions are almost symmetric. It is thought that the pressure gradient at the left and right tips of the plasma is almost the same in the symmetric plasma, and accordingly, jet generation does not occur. On the other hand, in the case of (x_TiS, E_in) = (−1, 40), (−1, 70), (1, 40), a jet flow can be observed: the jet is formed leftward in (−1, 40) and (−1, 70) and rightward in (1,40), with high density plasma formed in the left side and right side, respectively. Then, the pressure gradient is produced when a rarefaction wave passes through the plasma, which leads to the jet flow. In (x_TiS, E_in) = (1, 70), the jet is not formed, but the hot gas kernel spreads in the perpendicular direction to the YAG laser path. A possible explanation of this motion is as follows. Long plasma distribution as seen in Fig 4 produces a large pressure gradient in the left and right tips when a rarefaction wave passes through the tips, and jet flows toward the center of the plasma is generated. Then, the flows from both the left and right sides collide at the center and spread in a vertical direction.

3.2. Ignition characteristics of CH₄-air mixture

As shown above, the breakdown threshold energy decreases and the absorption energy increases by using the TiS laser. In addition, the plasma distribution and flow can be controlled by changing the TiS laser incident position. The flow with jet toward the upstream and downstream directions can be generated as well as without jet. If an appropriate flow is produced in a flame kernel, the ignition potential achieved would be higher than that of a normal laser ignition without the TiS laser. In this section, the minimum ignition energy and the ignition ability for a CH₄–air lean mixture are discussed.

First, to investigate the minimum ignition energy for laser-induced breakdown, lasers were focused on CH₄–air mixtures having various equivalence ratios ϕ with various incident YAG laser energies E_in. The TiS incident position was set to x_TiS = 0 mm, and the results are compared with the without-TiS case. In this experimental setup, incident and transmitted energies of TiS laser pulse were not measured for every shot. However, the average energy of incident and transmitted energies were measured in advance, and it is confirmed that the absorption energy of the TiS laser beam is less than 1 mJ, which is sufficiently low to influence ignition. Figure 6 shows the results for both the fire and misfire cases, which were plotted for ϕ and YAG incident energy E_in in (a) and for ϕ and YAG absorption energy E_ab in (b). The determination of a successful ignition was made on the basis of the pressure rise in the vessel. In this study, a pressure rise of larger than 0.01 MPa, which is three times of the resolution of AD converter of pressure sensor, indicates that the mixture successfully ignited. For lower ϕ, higher incident energy is required to successfully ignite a mixture for breakdowns both with and without TiS. In particular, the required E_in becomes much higher in less than approximately ϕ = 0.55. For all values of the equivalence ratio, a much lower E_in is sufficient to ignite a mixture in the with-TiS case than in the without-TiS case. On the other hand, by comparing the required absorption energy in both with- and without-TiS laser cases, no clear difference can be observed, and a minimum ignition energy is represented by the dashed line in Fig. 6(b) for both cases. This indicates that the ignition ability is mainly determined by the energy supplied in the gas. Because the absorption energy becomes much higher by using TiS laser, as shown in Fig. 3, the incident energy of the YAG laser is effectively used for ignition purposes.

 

Fig. 6 Results of ignition tests in CH₄-air lean mixture. ‘○’ shows successful ignition and ‘×’ shows misfire.

Download Full Size | PDF

As shown above, the absorption energy is one of the main factor for the successful ignition. In addition, the flow generated after breakdown may affect the ignition characteristics in a flammable mixture. We showed that the flow can be controlled using the TiS laser in the previous section for air. The generated flow were visualized by schlieren high-speed photography for CH₄-air mixture and confirmed that almost the same flows are produced with air breakdown cases. To investigate the effect of the flow on the ignition lean limit, the pressure rise in the vessel was investigated to clarify the effect on the ignition of the TiS laser incident position, x_TiS = −1, 0, 1 mm. Figure 7 shows the average and standard deviation of the maximum values of the pressure rise after breakdown, which are represented by p̄ and σ_p, respectively. The test was performed 3–15 times for each equivalence ratio ϕ. The YAG laser incident energy was set to (a) E_in = 40 and (b) 70 mJ, where the absorption energy is E_ab = 30 and 60 mJ, respectively. The dependence of the absorption energy on x_TiS is less than 5%, which is not large compared with shot-to-shot variations. Commonly, p̄ is nearly zero for less than a certain equivalence ratio, which means that the ignition fails for a leaner mixture, and becomes higher for a higher equivalence ratio. According to σ_p for x_TiS = 0 mm in (a), the shot-to-shot fluctuation is large for ϕ = 0.55–0.56 because there are successfully ignited cases and misfire cases, and the gas pressure does not reach an adequate value in some shots because of partial extinction. σ_p is almost zero for a leaner mixture owing to the absence of pressure rise, and it decreases for richer mixtures, which means that ignition has been stably achieved. This tendency is common in all TiS laser incident positions. The incident position x_TiS = −1 mm can ignite leaner mixtures than other incident positions, and that of x_TiS = 1 mm shows the lowest ignition ability. In the case of x_TiS = −1 mm, as seen in Fig. 5, a weak jet and toroidal vortex are thought to stabilize the flame and lead to successful ignition for a leaner mixture. However, at the same time, the jet flow could lead to extinction because of cooling by mixing with an unburnt mixture and flame stretch. In this case, the jet flow is very weak compared with the without-TiS case, and the flame holding effect is thought to be dominant. Although a similar flow structure can be observed in the case of x_TiS = 1 mm, it has a lower ignition ability. In this TiS incident position, because the plasma, or hot gas kernel, is distributed over a wide range and has a long shape, the absorbed laser energy is likely to diffuse without growing to a stable flame.

 

Fig. 7 The average and standard deviation of the maximum values of the pressure rise, p̄ and σ_p in CH₄-air mixture.

Download Full Size | PDF

According to Fig. 7(b), x_TiS = −1 and 1 mm have higher and lower ignition abilities, respectively, as with (a). Compared with (a), in the case of x_TiS = −1 and 0 mm, the pressure rise can be seen in a lower equivalence ratio, which is attributed to the higher absorption energy than in (a). However, in the case of x_TiS = 1 mm, the ignition lean limit is not extended to lower ϕ by increasing absorption energy. In this condition, the plasma has a long and narrow shape, and as shown in Fig. 4, the energy is divided into left and right, which is believed to lead to the diffusion of the hot gas of the flame kernel. Furthermore, the flame kernel spreads in a vertical direction, as shown in Fig. 5, which could lead to extinction by flame stretch. Therefore this type of plasma distribution is not effective for ignition purposes.

4. Conclusion

We attempted to control the plasma formation process and the successive flow by providing electrons from the small breakdown of a femto-second TiS laser focused near to the focal point of a Nd:YAG laser. The main findings are as follows.