High-peak-power passivelyQ-switched Nd:YAG/Cr4+:YAG composite laser with multiple-beam output

T. Dascalu; G. Croitoru; O. Grigore; N. Pavel

doi:10.1364/PRJ.4.000267

1. INTRODUCTION

Ignition induced by laser is currently viewed as an alternative technique to ignition realized by a spark plug, especially for vehicles with internal combustion engines ^[1–4]. Among the advantages of this new ignition method, one could mention: (i) the absence of the quenching effect of the developing flame kernel due to the lack of a spark plug electrode, resulting in a shorter burn duration; (ii) the possibility to ignite at an optimal position into the engine chamber for efficient combustion; (iii) lowering fuel consumption and decreasing exhaust gas emissions under conditions of normal engine operation, i.e., near the stoichiometric air–fuel ratio, or (iv) the possibility to ignite lean air–fuel mixtures and thus to further reduce engine impact on the environment.

A ${CO}_{2}$ laser was used to obtain in 1978, for the first time, the ignition by laser of a single-cylinder engine ^[5]. Furthermore, a four-cylinder engine was ignited, in 2008, by a $Q$ -switched Nd:YAG laser ^[6]. Because of the experimental conditions, the bulky lasers were placed outside of the engines and the laser beams were directed into the engine cylinders by adequate optics.

A solution for realization of a compact spark-like laser device was proposed in 2007 by Kofler et al. ^[7]. Thus, it was shown that a Nd:YAG laser medium that is passively $Q$ -switched by a ${Cr}^{4 +} : YAG$ crystal with saturable absorption (SA) can deliver pulses with few-millijoule energy and nanosecond-order duration, suitable for laser ignition. The first high-peak-power passively $Q$ -switched $Nd : YAG - {Cr}^{4 +} : YAG$ spark-plug laser was realized in 2010 by Tsunekane et al. ^[8]; furthermore, the same research group has investigated the characteristics of a $Yb : YAG - {Cr}^{4 +} : YAG$ discrete media combination for laser ignition applications ^[9]. Several other approaches, like the passively $Q$ -switched diffusion-bonded $Nd : YAG / {Cr}^{4 +} : YAG$ laser with wedged ${Cr}^{4 +} : YAG$ SA ^[10] or the diffusion-bonded $Nd : YAG / {Cr}^{4 +} : YAG$ laser pumped laterally through a YAG prism ^[11], could also be considered and optimized in order to be used for ignition by a laser. It is worth mentioning that the majority of these lasers were pumped longitudinally with fiber-coupled diode lasers. On the other hand, side pumping with a diode array enabled the realization of a high-energy $Nd : YAG - {Cr}^{4 +} : YAG$ laser (the so-called HiPoLas laser) ^[12]; the same pumping scheme was employed to build a passively $Q$ -switched $Nd : YAG - {Cr}^{4 +} : YAG$ pulse-burst laser for laser ignition ^[13].

For the first time, in 2013 an automobile gasoline engine was ignited with only laser sparks by Taira et al. ^[14]. Recently, we have also operated a four-cylinder engine with laser sparks ^[15]. Measurements at average speeds (below 2.000 rpm) concluded that ignition by laser improves engine stability and decreases emission of CO and HC in comparison with the same engine that was ignited by classical spark plugs. In these experiments, laser sparks that were built with diffusion-bonded $Nd : YAG / {Cr}^{4 +} : YAG$ composite media and that delivered single-beam output have been used ^[14,15]. On the other hand, a laser offers the possibility to obtain ignition at multiple points, thus better and more uniform combustion in comparison with ignition by electrical spark plugs. For example, increased pressure and shorter combustion time were measured for laser ignition in two points of ${CH}_{4} / air$ and $H_{2} / air$ mixtures in a static chamber ^[16]. Also, experiments realized with $H_{2} / air$ mixtures (in a static chamber) concluded that ignition at two points assures faster combustion in comparison with single-point ignition ^[17].

To generate multiple points of ignition, several solutions have been proposed. Thus, end pumping with three independent lines of a passively $Q$ -switched diffusion-bonded $Nd : YAG / {Cr}^{4 +} : YAG$ composite ceramic medium allowed realization of the first spark-plug-like laser device with three focusing points, at fixed locations ^[18]. In another approach, a spatial light modulator was applied to a single laser beam to obtain multipoint laser-induced spark generation with arbitrary geometrical location in three dimensions or at different distances on the central axis of propagation ^[19]. The energetic beam (with long, 10 ns duration at 1064 nm) was, however, delivered by a flash-lamp-pumped $Q$ -switched Nd:YAG laser. Also, a $2 \times 2$ lens array was employed to transfer the pump beam from a fiber-coupled diode laser to a $Nd : YAG - {Cr}^{4 +} : YAG$ laser and thus to realize a device with four-beam output ^[20]. Each laser beam operated in burst mode, and every pulse burst contained two to five pulses of low (0.12–0.22 mJ) energy and long (10.5–11.5 ns) duration at a kilohertz-order repetition rate. In this case, further increase of the laser pulse energy is necessary to obtain air breakdown.

In this work we report on the design, realization, and performance of a high-peak-power passively $Q$ -switched $Nd : YAG / {Cr}^{4 +} : YAG$ composite monolithic laser that yields four independent beams, each beam having appropriate characteristics for laser ignition. A compact optical system was made to transfer every pump beam into $Nd : YAG / {Cr}^{4 +} : YAG$ , keeping the compactness of the device and assuring millimeter-order distance between laser beams. Typically, the energy of a laser pulse was higher than 3 mJ with duration shorter than 1 ns. The focusing system was designed for various purposes, aiming focusing and air breakdown in four points (in the same plane or at different distances from the laser) or in a single point. This laser can be used for multipoint ignition of an automobile engine; it could be also of interest in space propulsion systems ^[21,22] or for ignition in turbulent conditions specific to aeronautical combustion ^[23].

2. THE LASER CONFIGURATION

A. Design: Single Line, Experiments, and Modeling

The laser configuration is shown in Fig. 1. The laser medium is an $Nd : YAG / {Cr}^{4 +} : YAG$ composite structure of 10 mm diameter; it was made of an 8.5 mm long, 1.1 at. % Nd:YAG medium that was bonded to a 2.5 mm thick ${Cr}^{4 +} : YAG$ SA with initial transmission ( $T_{i}$ ) of about 40%. Such an $Nd : YAG / {Cr}^{4 +} : YAG$ medium assures robustness and compactness, as well as a high resistance at vibrations of the laser ^[14,15,18].

Figure 1.Laser configuration (patent pending) is shown. Each pump beam (①,②, ③, and ④) is directed from the corresponding optical fiber (F) to the composite $Nd : YAG / {Cr}^{4 +} : YAG$ medium through a lens (L) and a folding prism (P). Four (1, 2, 3, and 4) laser beams are obtained.

Download full size

View all figures

A monolithic resonator was obtained by coating the surface S1 of Nd:YAG with a high-reflectivity mirror (HRM, reflectivity $R > 0.999$ ) at the laser wavelength $λ_{em} = 1.06 μm$ and the outcoupling mirror (OCM) with transmission $T_{OCM} = 50 %$ ( $\pm 5 %$ ) on surface S2 of ${Cr}^{4 +} : YAG$ . Surface S1 was also coated with a high transmission layer ( $T > 0.98$ ) at the pump wavelength $λ_{p} = 807 nm$ . It is worth mentioning that, in this design, the pump-beam absorption efficiency in Nd:YAG was evaluated to be around 0.95. Therefore, the ${Cr}^{4 +} : YAG$ SA properties were not influenced by the residual pump beam. We mention that $Nd : YAG / {Cr}^{4 +} : YAG$ single crystals as well as $Nd : YAG / {Cr}^{4 +} : YAG$ fabricated by ceramic techniques were used in experiments. The best results were obtained with $Nd : YAG / {Cr}^{4 +} : YAG$ ceramic (or polycrystalline) media from Baikowski Co., Japan.

The pump was made at $λ_{p} = 807 nm$ with fiber-coupled (diameter of 400 μm and numerical aperture $NA = 0.22$ ) diode lasers (JOLD-120-QPXP-2P, Jenoptik, Germany) that were operated in quasi-continuous-wave regime; in the experiments, the pump pulse duration was set at 250 μs and the repetition rate was increased up to 60 Hz. To keep the pump line simple, the pump beam delivered through a fiber (F) was focused into Nd:YAG with only a single lens (L). A small prism (P) was used to fold each pump line and thus to accommodate four fibers around surface S1 of Nd:YAG.

The pump-line configuration was decided first. For this purpose, the $Nd : YAG / {Cr}^{4 +} : YAG$ composite ceramic medium was pumped directly from fiber F with only one lens, L. The distance between F and L was noted by $d_{1}$ and the distance from L to the $Nd : YAG / {Cr}^{4 +} : YAG$ was denoted by $d_{2}$ . Several aspherical lenses L, with focal length $f$ between 3.1 and 7.5 mm were employed; for each lens, the laser pulse energy $E_{p}$ was measured as a function of $d_{1}$ and $d_{2}$ .

Figure 2(a) shows several results obtained with two lenses, the first with focal length $f = 4.0 mm$ and a second one with $f = 6.2 mm$ . In these investigations the repetition rate was set at 5 Hz. When the lens with $f = 4.0 mm$ was placed at $d_{1} = 3.4 mm$ from optical fiber F, laser pulses with energy $E_{p} = 2.9 mJ$ were obtained for $Nd : YAG / {Cr}^{4 +} : YAG$ positioned at distance $d_{2} = 7.5 mm$ [Fig. 2(a)]. The corresponding pump pulse energy was $E_{pump} = 26 mJ$ [Fig. 2(b)]. Increasing $d_{2}$ at 10 mm improved $E_{p}$ at 5.2 mJ, whereas $E_{pump}$ reached the level of $\sim 45 mJ$ . In the case of lens L with $f = 6.2 mm$ placed at $d_{1} = 4.9 mm$ , laser pulses with $E_{p} = 3.0 mJ$ were measured for $Nd : YAG / {Cr}^{4 +} : YAG$ situated at distance $d_{2} = 15 mm$ ; the requested pump pulse energy was $E_{pump} = 31.8 mJ$ . For this lens, the laser could be operated up to a distance $d_{2} = 18.5 mm$ ; at this point, the laser pulse energy increased to $E_{p} = 5.9 mJ$ and the available pump energy $E_{pump} = 47 mJ$ delivered by the diode was reached. The colored points for $E_{p}$ shown in Fig. 2(a) were used in modeling (as will be explained below). We have also measured the laser beam $M^{2}$ factor using the knife-edge method (10%–90% level, ISO 11146/2005). For laser pulses with energy $E_{p}$ around 3 mJ (a minimal value of interest for realizing air breakdown in our experimental conditions), the $M^{2}$ factor was around 4.1; the beam quality decreases with increasing the laser pulse energy, with $M^{2}$ factor of about 5.2 for pulses with $E_{p} \sim 5 mJ$ .

Figure 2.(a) Laser pulse energy $E_{p}$ versus distance $d_{2}$ between lens L and surface S1 of Nd:YAG. Data for two lenses, first with focal length $f = 4.0 mm$ placed at distances $d_{1} = 3.4$ and 4.4 mm from F, and the second with $f = 6.2 mm$ positioned at $d_{1} = 4.9$ and 5.9 mm from F, are shown. The inset is a sketch of the single pump-beam arrangement. (b) The corresponding pump pulse energy, $E_{pump}$ , is plotted.

Download full size

View all figures

To explain the laser pulse performance, we have applied a model developed in our previous work ^[15,18]. Thus, we remember that the laser pulse energy can be written by ^[24,25] $E_{p} = \frac{h ν}{2 γ_{g} σ_{g}} A_{g} \times \ln \frac{1}{R_{OCM}} \times \ln (\frac{n_{g f}}{n_{g i}}),$ (1)where $h ν$ is the photon energy at $λ_{em}$ , $γ_{g}$ is the inversion factor, and $σ_{g}$ represents the Nd:YAG emission cross section. $A_{g}$ is the laser beam cross-section area in Nd:YAG and the OCM reflectivity is $R_{OCM} = (1 - T_{OCM})$ . The density of the initial inversion of population is $n_{g i} = β / (2 σ_{g} λ_{g})$ , with $λ_{g}$ the Nd:YAG length. The density of the final inversion of population $n_{g f}$ was obtained from the transcendental equation ^[15,18]: $(1 - r_{n}) + (1 + \frac{(1 - δ) \times \ln T_{i}^{2}}{β}) \times \ln (r_{n}) + \frac{1}{α} \times \frac{(1 - δ) \times \ln T_{i}^{2}}{β} \times (1 - r_{n}^{α}) = 0,$ (2)where $r_{n}$ is the ratio $r_{n} = n_{g f} / n_{g i}$ . The element $β = (- \ln R_{OCM} + L_{t} - \ln T_{i}^{2}) / [1 - \exp (- 2 a^{2})]$ includes the parameter $a$ , defined as the ratio between the pump-beam radius $ω_{p}$ and the laser-beam radius $ω_{g}$ in Nd:YAG, $a = ω_{p} / ω_{g}$ . $L_{t}$ is the double-pass residual loss of the monolithic $Nd : YAG / {Cr}^{4 +} : YAG$ laser, accounting for the Nd:YAG losses ( $L_{i}$ ) as well as for the final transmission ( $T_{f}$ ) of ${Cr}^{4 +} : YAG$ SA. Also, $δ$ is the ratio $δ = σ_{ESA} / σ_{SA}$ , with $σ_{ESA}$ the excited-state absorption cross section and $σ_{SA}$ the absorption cross section of ${Cr}^{4 +} : YAG$ ; $α$ is $α = (γ_{SA} σ_{SA}) / (γ_{g} σ_{g}) \times (A_{g} / A_{SA})$ , with $γ_{SA}$ the inversion reduction factor for ${Cr}^{4 +} : YAG$ . The laser beam area in ${Cr}^{4 +} : YAG$ , $A_{SA}$ , and $A_{g}$ were considered equal, a fair approximation due to the short length of the $Nd : YAG / {Cr}^{4 +} : YAG$ medium. For the simulations, the Nd:YAG emission cross section was taken as $σ_{g} = 2.63 \times 10^{- 19} {cm}^{2}$ ; the absorption cross section and excited-state absorption cross section of ${Cr}^{4 +} : YAG$ SA were $σ_{SA} = 4.3 \times 10^{- 18} {cm}^{2}$ and $σ_{ESA} = 8.2 \times 10^{- 19} {cm}^{2}$ , respectively. Because of the high value of the $M^{2}$ factor, the pump beam as well as the laser beams were considered to have uniform (like top-hat) distributions.

Figure 3 presents modeling of laser pulse energy $E_{p}$ function of laser beam radius $ω_{g}$ ; the value of $ω_{g}$ was determined for each configuration by an ABCD formalism using the data recorded during $M^{2}$ factor measurements. In simulations, losses $L_{i}$ were taken as 0.01 (a typical value for Nd:YAG with the doping level and length used in our experiments). Also, values of 0.94 and 0.98 for the final transmission $T_{f}$ of ${Cr}^{4 +} : YAG$ were considered. One could observe that, for a fixed value of parameter $a$ , the influence of total losses $L_{t}$ on laser pulse energy is not too significant; more important is the value of $a$ .

Figure 3.Laser pulse energy $E_{p}$ versus laser beam radius $ω_{g}$ . Losses $L_{i} = 0.01$ , and, for ${Cr}^{4 +} : YAG$ , final transmission $T_{f}$ of 0.94 and 0.98 were considered in simulations.

Download full size

View all figures

B. Four-Beam Output Laser

A typical laser module made in this work is shown in Fig. 4. The four prisms, P [Fig. 4(a)], were used to fold and redirect each pump beam from the corresponding optical fiber F toward the $Nd : YAG / {Cr}^{4 +} : YAG$ medium [Fig. 4(b)]. Several geometries were considered and realized. For example, a first one was built with identical aspherical lenses L of focal length $f = 4.0 mm$ ; distance $d_{1}$ was 3.4 mm and the $Nd : YAG / {Cr}^{4 +} : YAG$ medium was positioned at $d_{2} = 8 mm$ (we remember that, as shown in Fig. 1, on each folding arm, $d_{2}$ accounts for the distance from a lens L to $Nd : YAG / {Cr}^{4 +} : YAG$ ). For a second arrangement (which will be discussed below) we used aspherical lenses with focal length $f = 6.2 mm$ ; distance $d_{1}$ was 4.9 mm and the distance $d_{2}$ was chosen as $d_{2} = 15.5 mm$ . Uncoated, 3 mm right-angle prisms were employed in this scheme, each prism being nearly 5 mm away from the corresponding lens L.

Figure 4.(a) The module with the four folding prisms is presented. Each lens (L) position is indicated. P: prism. (b) A photo of the $Nd : YAG / {Cr}^{4 +} : YAG$ laser with four-beam output is shown. The directions of laser beams 1, 2, 3, and 4 are given by the colored lines with arrows.

Download full size

View all figures

At 5 Hz repetition rate, the laser pulse energies were 3.25 mJ for beam 1, 3.30 mJ for beam 2, 3.60 mJ for beam 3, and 3.20 mJ for beam 4. The small differences from the expected value of 3.4 mJ [according to Fig. 2(a)] could be attributed to some length differences in every pump line, as well as the fact that each pump was not made in the center of the $Nd : YAG / {Cr}^{4 +} : YAG$ medium but on its circumference (in this case, small differences in Nd concentration or coatings can affect the results). Moreover, differences in the four diode lasers used for pumping can influence the laser pulse energy. The pump pulse energies were 30.5 and 32 mJ for beams 1 and 2, respectively; $E_{pump}$ amounted to 33 mJ for beam 3 and to 30 mJ for beam 4. Thus, the overall optical-to-optical efficiency (defined as the ratio between the laser pulse energy and the pump pulse energy, $E_{p} / E_{pump}$ ) was in the range of 0.103 for beam 2 up to 0.109 for beam 3. The pulse duration (which was determined with an ultrafast InGaAs photodetector, rise time shorter than 35 ps) was around 0.9 ns for each beam, and therefore the laser pulse peak power was in the range of 3.55 MW for beam 4 to 4.0 MW for beam 3.

Important for stable laser emission is the $Nd : YAG / {Cr}^{4 +} : YAG$ temperature. We have used an FLIR T620 thermal camera ( $- 40 ° C$ to $+ 150 ° C$ range, $\pm 2 ° C$ accuracy) to measure the temperature of surface S2 of the $Nd : YAG / {Cr}^{4 +} : YAG$ medium. As shown in Fig. 5, when the laser medium was kept in air and with no cooling [as shown in Fig. 4(b)], after a few minutes of operation the maximum temperature of surface S2 at the location of beam 1 increases from 27°C at 2 Hz repetition rate up to 64.5°C at 60 Hz repetition rate. Similar results were recorded for all lines (and Fig. 5 presents also the temperature for the central axis of beam 4). Longer time operation could not be performed, especially at high repetition rate, as the medium temperature increased further and the laser emission eventually ceased. Therefore, in the next experiments, the $Nd : YAG / {Cr}^{4 +} : YAG$ medium was wrapped in indium foil and then clamped in a copper holder; a fan was used to cool the metallic holder. The temperature of surface S2 was much lower in comparison with the previously described operation regime. Thus, as illustrated in Fig. 5, only a small increase in temperature, from 24.5°C at 2 Hz operation rate up to 30.8°C at repetition rate of 60 Hz was measured in the central position of beam 1. The laser pulse energy $E_{p}$ was also stable at high rates of repetition. For operation at 60 Hz, the increase of $E_{p}$ in comparison with the values measured at 5 Hz were below 3.7% for beams 1, 2, and 4; a slightly higher increase of 5.5% was measured for the energy of beam 3.

Figure 5.Maximum temperature of surface S2 of $Nd : YAG / {Cr}^{4 +} : YAG$ at central locations of beams 1 and 4 versus the pump pulse repetition rate.

Download full size

View all figures

The laser device described in this work can be used (after further optimizations, testing, and integration) for multipoint ignition of an automobile engine, especially for ignition of lean air–fuel mixtures. Furthermore, it could be of interest for ignition of propellants in space applications or for ignition in turbulent conditions that are specific to aeronautics. For this purpose, different focusing schemes that allow access of a defined volume of fuel can be designed.

For example, as shown in Fig. 6(a), all beams could be focused in the same point; an aspherical lens with focal length of 8.1 mm ( $NA = 0.50$ ) was used in this scheme. In another arrangement, each beam could be focused in a point on its axis of propagation, using identical lenses. Such a layout is presented in Fig. 6(b), where aspherical lenses with focal lengths from short, $f = 7.5 mm$ ( $NA = 0.30$ ), to long, $f = 18 mm$ ( $NA = 0.15$ ), could be used to obtain air breakdown. We mention that the distance between the central axes of beams 1 and 3 and that between beams 2 and 4 was 7.0 mm. A larger volume could be accessed by focusing the beams at different distances from the optics. For example, Fig. 6(c) illustrates the air breakdown obtained with beams 1 and 3 from the horizontal plane that were focused in front of beams 2 and 4 (that are in vertical plane). A similar case is shown in Fig. 6(d), where beams 2 and 4 were focused close to the optics. Other combinations of optics could allow the focusing of each beam outside (or inside) its propagation axis, thus offering more possibilities for location of the ignition points in space. Furthermore, the timing between each beam could be controlled; in this way a method for avoiding misfiring, an event that usually happens in the case of igniting lean air–fuel mixtures, or for fuel in turbulent conditions. In addition, one more laser beam could be easily obtained by pumping the $Nd : YAG / {Cr}^{4 +} : YAG$ ceramic medium on its central axis. These approaches are under investigation and will be reported later.

Figure 6.Several combinations for focusing the laser beams are shown: (a) focusing in a single point; (b) focusing all beams in the same plane, at equal distances from the optics; (c) beams 1 and 3 (from the horizontal plane) are focused before beams 2 and 4 (from the vertical plane); (d) beams 2 and 4 are focussed close to the optics, in front of beams 1 and 3.

Download full size

View all figures

3. CONCLUSION

We have proposed a high-peak-power passively $Q$ -switched $Nd : YAG / {Cr}^{4 +} : YAG$ laser with four-beam output. Each pump line has a simple design, containing a lens and a folding prism. The laser yielded four beams, with energy per pulse of 3.25 mJ for beam 1, 3.30 mJ for beam 2, 3.60 mJ for beam 3, and 3.20 mJ for beam 4 at 5 Hz repetition rate. The pulse duration was about 0.9 ns. When the repetition rate was increased to 60 Hz, the laser pulse energy increased, in the range of 2.4% for beam 2 up to 5.5% for beam 3. Different focusing arrangements were presented to obtain air breakdown in variable volume. Some applications of such a laser device include ignition of lean air–fuel mixtures in automobile gasoline engines, ignition of a propeller under high-altitude conditions and even in space propulsion, or ignition in turbulent conditions that are specific to aeronautics.

Category: Lasers and Laser Optics

Received: Jul. 26, 2016

Accepted: Sep. 25, 2016

Published Online: Dec. 23, 2016

The Author Email: N. Pavel (nicolaie.pavel@inflpr.ro)

DOI:10.1364/PRJ.4.000267