laser plasma ignition: status, perspectives, solutions

Laser plasma ignition: status, perspectives, solutions

E. Wintner*a,b, H. Koflera, D. K. Srivastavab, A. K. Agarwalb aVienna University of Technology, Photonics Institute, Gusshausstrasse 27, A-1040 Vienna, Austria;

bDepartment of Mechanical Engineering, Engine Research Laboratory, Indian Institute of Technology Kanpur, Kanpur-208016, India

ABSTRACT

Laser ignition can yield certain advantages compared to conventional sparkplug ignition. Among other already frequent-ly discussed reasons due to: i) option for sequential or multipoint ignition which can contribute to more reliable ignition in direct injection engines; ii) ignition of leaner mixtures at higher compression being most relevant for gas engines. A satisfying solution to the above mentioned requirements is the longitudinally diode-pumped passively Q-switched Cr4+:YAG/Nd3+:YAG laser capable of emitting ∼1-ns-pulses of at least 20 mJ . This type of solid-state laser (SSL) con-fectioned in an engine-compatible form can be called a laser sparkplug. Early versions of this concept comprised a high-power diode pump laser (quasi-cw power >500 W @ ∼500 µs duration) which were placed remote from the engine to avoid detrimental influences of temperature, vibrations, pollution etc. In this case only the SSL is exposed to the elevated temperature in the vicinity of the cylinder walls (<100°C). Recently, technical and cost-oriented considerations allow a change of concept from fiber-based remote pumping via edge emitter arrays to the use of newly developed so-called power VCSELs with two-dimensional stacking. Collimation to form a round pump beam thereby becomes much easier. Their temperature resistance allows lower-cost direct mounting although thereby a wavelength shift is induced. The Q-switched SSL in the sparkplug also faces temperature dependent phenomena like reduction of pulse energy and efficien-cy, a change of pulse timing and beam profile which will be discussed in the paper.

Keywords: Laser ignition, passively Q-switched solid-state laser, monolithic laser, optical sparkplug, power VCSELs, temperature effect on VCSELs, temperature dependence of Q-sw-Nd:YAG laser, gas engines

1. INTRODUCTION This paper deals exclusively with non-resonant optical breakdown (NRB, see e.g. review [1]) as one of the possible processes underlying laser ignition (LI). As a matter of fact, it was first discovered on basis of this phenomenon that in-tense light allows to create plasma, i.e. a spark (Terhune et al., 1962 [2]). At that time, only the Ruby laser could enable such strong laser-matter interaction taking place at intensities of the order 1011 W/cm2. Nowadays, there is an economic necessity to employ cost-efficient and compact laser systems for engine ignition, and hence the wavelength choice has to be made out of a number of available discrete ones. One of the best candidates by all aspects is diode-pumped passively Q-switched Nd-laser, either employing YAG as a crystalline host or ceramics as a more recent approach.

Since the first – as seen from today’s perspective – inadequate attempts of igniting engines by a laser (a pulsed CO2 laser in this case [3]), tremendous advances in laser technology as well as technical understanding of the plasma interaction processes [1] have been made. Highlights are i) the development of compact solid-state lasers (SSLs), especially Nd:YAG [4] (e.g. microchip lasers [5], ceramic laser materials [6]) associated with high perfection in Q-switching for the generation of ns-pulses [7]. Monolithic arrangements of Cr4+:YAG [8] and Nd3+:YAG are most elegant and robust; ii) the development of adequate powerful quasi-continuous wave (qcw, pulse duration in the 100-µs-regime) pump diodes with emission @ ∼808 nm is a second most important step towards the realization of a real laser sparkplug which even-tually could replace a conventional sparkplug in specific applications. Since several years, high power diode lasers with peak power >500 W are commercially available (e.g. 600 W JENOPTIK, JOLD-600-QPXF-2P2 as employed by Kofler [9]) and allow the generation of ns-pulses with energies up to 20 mJ [10]. Presently, high-power qcw 2-dimensionally stacked vertical cavity surface emitting lasers (VCSELs) become available and have the potential to revolutionize the concept of diode pumping of such SSLs [11],[12].

*[email protected]; phone +43 1 58801-38712; fax …-38799; http://www.photonik.tuwien.ac.at; present address: b

Invited Paper

Fundamentals of Laser-Assisted Micro- and Nanotechnologies 2013, edited by Vadim P. Veiko, Tigran A. Vartanyan,Proc. of SPIE Vol. 9065, 90650B · © 2013 SPIE · CCC code: 0277-786X/13/$18 · doi: 10.1117/12.2053152

Proc. of SPIE Vol. 9065 90650B-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 01/07/2014 Terms of Use: http://spiedl.org/terms

NRB is the best suited breakdown mechanism for a successful commercial system of LI, and it is also most investigated [1]. It offers clear advantages: i) as already mentioned, one may choose the wavelength with the best availability within a range like the near infrared (as judged by all aspects like laser material properties, efficiency, complexity, size, opera-tional lifetime, cost). ii) Due to the nature of non-linearity, the relevant process of NRB will only happen in a high inten-sity region, i.e. in or near the focus which may be placed at any position according to engineering aspects as long as op-tics (focal size for instance) and power allow it. This is most important for internal combustion engines, i.e. Otto engines in this context. iii) There might be a third advantage relevant for direct fuel injection gasoline engines: due to the differ-ence of plasma formation threshold of gases and condensed matter (∼500 versus ∼1 GW/cm2) the laser spark can “search” for fuel droplets, even if the location slightly changes according to a change of engine parameters. iv) Further-more, in case of lean mixtures the higher temperature of a laser plasma compared to an electrical spark becomes a useful feature for more reliable ignition [13],[14], associated with the fact that a ns-plasma cannot be extinguished easily via turbulences. The advantages of LI vs. conventional spark ignition (SI) have been discussed frequently [15],[16]. The favorable pressure dependence certainly is one major motivation followed by the absence of electrode erosion and flame kernel quenching, earlier onset and decrease of combustion time as well as more stable combustion and increased engine performance [17],[18],[19],[20]. To realize the full potential of laser ignition, especially in gas engines fuelled by com-pressed natural gas (CNG) with its remarkably high knocking threshold, special engine development for LI application would be required which, to our knowledge, never anywhere has been carried out so far, neither in case of automotive engines nor in stationary gas engines of MW power range. The goal of creating a long-lived ignition device of acceptable price which clearly reduces service cost going along with the mentioned advantages has not been demonstrated in any respect until now.

Although the focus of this paper lies on the construction, performance and problems with optical sparkplugs (“local la-ser”, or “local resonator”) and their consequences, it has to be mentioned that a large amount of work on LI has been carried out with commercial mainframe lasers, either for preliminary application or following the concept of “central laser”. Fig. 1 depicts the 3 possible versions of laser concept schematically: i) (left) a central laser can be located off en-gine thereby remaining unaffected by its detrimental influences like heat, vibrations and pollution. In many cases, espe-cially for stationary large gas engines, space availability is no big deal. An important feature, of course, is the method of transportation of the ignition pulses to the individual cylinders. This can be done via open beam propagation with mirrors and lenses (just protected by tubes), potentially involving multiplexers of any type, the simplest being electro-mechanic mirrors as they are widely employed in show laser scanners [21]. Fiber transportation looks much more elegant and has been widely investigated [22]. Generally, it is a major problem to propagate pulsed beams with powers suitable to cause breakdown in gases in any kind of fiber [23]. Due to the ∼100 times lower breakdown threshold in condensed matter, solutions involving hollow fibers have been attempted with varying success. The scope stretches from dielectric capilla-ries, silver polymer interior coating to highly sophisticated photonic bandgap fibers. Bending remains an unsolved prob-lem in all cases. The success of A. Yalin’s group at Colorado State University remains remarkable [24]: they employed thick (≤ 1 mm) step-index fibers with thick coatings, and low numerical aperture optics to couple the laser pulses in. Hence only low order modes with large diameter were stimulated to propagate, and within a few meters distance no coupling to other modes takes place. Radii of curvature in the ½-m-range are acceptable in the context of large gas en-gines. A specially constructed multiplexer is indispensable. ii) (Fig. 1, center) In this approach, the laser resonator is lo-cal and hence attached to every cylinder, while the pump source is positioned off engine with similar advantages like in i). This specially applies to diode pump sources which have an emission wavelength dependency on temperature (λ in-creases with T). SSLs like Nd:YAG do not follow this shift with respect to their absorption as will be shown in section 3. However, they suffer from other T-dependent phenomena. As a matter of fact, fiber-pumped SSL generally follow the longitudinal pumping scheme (comp. Section 2). The optical sparkplug developed at the first author’s group at TU Vien-na followed this scheme, although pump multiplexing was not applied. Some researchers claim that there is no advantage in this kind of multiplexing if lifetime of sources (edge emitters are believed to be able to emit ∼109 pulses) is compared to additional cost of such a device. iii) (Fig. 1, right) This reveals an optical sparkplug in the strictest sense of the term: a complete diode-pumped SSL attached to each individual cylinder. According to the state of the art, pumping can be lon-gitudinal or transversal. The former is addressed in Section 3 comprising the approach employing power VCSELs. There are several potential benefits to this concept: power VCSELs are monolithic two-dimensional arrays which eventually could be produced very cheap. Due to their close to circular shape collimation can be easy. Compared to fiber incoupling (which in most cases leads to over-confinement of the beam size), cost can be reduced considerably. The drawbacks are the thermal influences which are the main topic of this paper.



1.3MW

Central Laser

Laser

Optical Fibres

Energy

Cooling

Central Pump Source,local Resonator

0,5kW Laser Energy

Cooling

Local Laser

CoolingQ- SwitchFocusing Lens

optical Window

Focusing Lens

Cooling

Q- Switch

Focusing Lensoptical Windowoptical Window 1

Focus

Combustion Chamber Combustion Chamber Combustion Chamber

18

16T=150°C T =275 °C T =400 °C

14-

12-

10-

8-

6-

4-

2- / minimum plasma

0..._=

1 2 1.4 1.6 1.8

energy' I I I

2.0 2.2 2.4 2.6 2.8

air equvalence ratio 2.

3.0 3 2

Figure 1. Schematic drawings of ignition laser-cylinder concepts: (left) a central laser (any suitable type) can be employed with beam transportation to every cylinder, potentially involving a multiplexer; (center) a passively Q-switched solid-state laser is mounted to every cylinder just like a sparkplug, however being pumped remotely via (standard) optical fiber; multip-lexing of the pump being possible; (right) a local laser represents a real optical spark plug, and each cylinder is equipped (Graphics: AVL)

2. REQUIRED SPARKPLUG PARAMETERS In several research groups, among them TU Vienna Laser Ignition Group [16],[25] as well as IIT Kanpur Engine Re-search Laboratory [26], initial testing of LI was performed in constant volume chambers. Nearly all parameters like type of fuel, pressure p, air-equivalence rate λequ, temperature T, focal geometry, wavelength λ, combination of two λ, pulse duration τp, multiple sequential or spatial pulses have been varied and their influences on the minimum ignition energy (MIE) have been evaluated (MIE represents the optical energy applied to the focal spot to generate the plasma; the laser pulse energy required to enter the container with the compressed air/fuel mixture is much higher – sometimes called min-imum pulse energy MPE – mostly according to reflection and transmission losses; however, MPE also stands in many cases for minimum plasma energy!). Obviously, in order to apply LI, especially when designing a laser sparkplug, the required MPEs have to be known. Therefore, already long ago, at TU Vienna such specs have been identified as is shown in Fig. 2 for air/methane mixtures at p = 20 bar vs. λequ at 3 different T realistic for engine operation.

Figure 2. Minimum ignition energy MIE vs. equivalence ratio λequ at three ignition temperatures T. MIE and MPE are main-ly defined by the ignition pressure (p = 20 bar) and the focal volume. The minimum plasma energy is independent of T [25].



50

LPE = 9.70 mJ/PulseLPE = 18.6 mJ/Pulse

40 - LPE = 24.0 rriJ/PüTse

30

20 -

10-

10

-20

120

80

-80 -60 -40 -20 0 20 40

Crank angle (deg.)

. 2060 80

18

15

12

9

6

5 bar_A

10 bar

= 1.1

A=11

1

._.,_/,7

50 100 150

Focal Length (mm)

200 250

Engine tests at IIT Kanpur with a Diesel engine converted to burn natural gas (naturally aspirated, water-cooled, four-stroke, single cylinder, compression 9.9, variable ignition timing and air-fuel ratio) allowed to evaluate the required laser pulse energies (τp = 6-9 ns, λ = 1064 nm) [26]. Fig. 3 reveals results acquired for cylinder pressure rise (left scale) and heat release (right scale) for λequ = 1.2, when ignited with laser pulse energies of 9.7, 18.6 and 24.0 mJ. As it can be clearly recognized, results do not differ significantly, and hence the conclusion can be made that pulse energies around 15 mJ are sufficient, especially if τp is shortened significantly using laser sparkplugs based on shorter resonators (τp ≤ 1.5 ns).

Figure 3. Cylinder pressure variation with crank angle for different laser pulse energies at λequ = 1.2 for CNG operated gas engine at 1500 rpm; focal length of focusing lens 30 mm [26].

In Fig. 4 three different focal lengths of the focusing lens were used to ignite air/methane mixtures in a constant volume chamber at p = 5 and 10 bar and λequ = 1.1. The increase in MPE is around 50 %. This has to be considered in the attempt to achieve a focal spot lying rather in the center of the compressed mixture volume inside the cylinder.

Figure 4. Minimum ignition pulse energy vs. focal length of the focusing lens for λequ = 1.1 of air/methane mixture [26].



3. OPTICAL PUMPING OF LASER SPARKPLUGS As implicitly expressed by the title of this paper, other ignition concepts based on plasma sparks besides the optical sparkplug are not treated here. This should not devaluate the above mentioned approaches of Colorado State University and Argonne National Laboratory. However, these results do not represent advances in miniature laser technology which is considered to be more universally applicable, e.g. in vehicles. Despite the bulkiness of the systems with rather large fiber radii of curvature or inflexible laser beam propagation tubes, they represent viable approaches being highly robust and unaffected by temperature changes with respect to laser performance. Nevertheless, according to our opinion, they cannot represent solutions for automotive implementation!

Laser sparkplugs, in general, require pump diodes whose temperature dependent emission has to match with the laser material absorption. Therefore the engine environment represents a major problem with respect to elevated, and even worse, varying temperature as well as vibrations and pollution. Only in case of longitudinal diode pumping, this negative influence can be ruled out to a major extent if solely the SSL is exposed to the elevated temperature in the vicinity of the cylinder walls (<100°C) and the other mentioned influences. Kofler et al. and Taira et al. have successfully followed this concept [27],[28]. The latter group has recently published the first automotive operation of a laser ignited engine [29].

Another approach has been successfully performed by Carynthian Tech Research (CTR): they followed the concept of transversal pumping [30],[31]. According to their reports, the setup is extremely unaffected by engine vibrations [32]. However, the temperature inherent problem exists, associated with rather high cost of this solution. Hence it is no won-der that broad commercial application still is missing. Fig. 5 depicts the mentioned optical sparkplug solutions.

Figure 5: Different laser sparkplugs developed worldwide (counterclockwise): (a) Concept Vienna University of Technolo-gy, Vienna, Austria [9], (b) Institute of Molecular Science, Okazaki, Japan [28], Carynthian Tech Research Villach / AVL Graz, Austria [32].



105

104

2

etched glasssubstrate substrateVCSEL VCSEL

102

10'500 480 460 440 420 400 380 360 340

Junction Temperature (K)

760...945nm singlemode VCSELs

0.7eV

50 °C ambienttemperature

540510 480 450 420 390 360

Tluncuon (K)

330

1000000

100000

L10000 F

1000

300

I-

3.1 High-power qcw VCSELs as pump sources for laser sparkplugs

Recently, mainly cost-oriented considerations, but also technical advances, led to a change of concept from fiber-based remote pumping via edge emitter arrays to the use of newly developed so-called power VCSELs ([11],[12]; these home-pages of Princeton Optronics and ULM Photonics, respectively, give comprehensive evidence for many following state-ments). Thereby, among other advantages, the symmetry of pump laser arrays is in the focus of attention: edge emitter arrays usually consist of one row of ∼20 emitters (aperture around 1×5 µm2) with ∼50−100 µm periodicity, i.e. a very elongated arrangement of light sources. Collimation to form a round pump beam represents some challenge applying the laws of classical optics. Opposed to that, VCSELs are 2-dimensionally arranged; hence the problem of collimation optics is much more relaxed. Additionally, direct mounting of the pump laser onto the laser sparkplug will improve simplicity and eventually cost, provided operation at elevated or, even more problematic, variable operation temperature T1 due to direct mounting on the cylinder head of an engine can be tolerated.

An increase of T1 has at least a threefold impact on the performance: the output power will be reduced; the lifetime of the device will shrink and the emission wavelength will become slightly longer. The mode characteristics might also vary, however this does not follow such a clear pattern. In [11], as an example, the L-I curves for a qcw array at 808 nm at different T1 are depicted. Changing from 20°C to 85°C @ 50A current is reported to result in a linear drop of the maxi-mum output from ∼50W to ∼25W, i.e. to one half! In good agreement with this measurement, an empirical rule for the reduction of output power can be found in the literature [33]: 0.8% per degree C (20° to 90°C).

The lifetime of VCSELs can lie above 1 Mio. h (corresponding to >100 yrs) as illustrated in Fig. 6 decreasing exponen-tially, however, with rising T1 [33]. This may lead to a value <1/10 of this timespan for a realistic T1 rise, i.e. >10 yrs, and hence can be considered to be still sufficient.

Figure 6: Lifetime dependent on operation temperature [33]. (MTTF ‘mean time to failure’).

The dependency of the mean emitted wavelength for a VCSEL typically lies in the range of 0.06 nm/°C [34],[35]. This means, a diode emitting @ 804 nm at RT will emit @ 808 nm at 77°C. It depends on the specific application, if this change can be estimated small or large. In any case, the T1 dependence of the VCSEL with respect to emitted wavelength is clearly smaller than for the edge emitter (“classical laser diode“), amounting 0.3 nm/°C [36].

If a certain narrow absorption line should be met, e.g. like the 808 nm-line of Nd:YAG, the T1 dependence has to be con-sidered in any case. For a high power VCSEL sufficient cooling not only has to be provided for the sake of the diode conservation (lifetime!), but also a decent temporal stability of operation temperature, otherwise the wavelength may leave the desired emission range and, for instance, cannot be absorbed any more by Nd:YAG. As a matter of fact, cool-ing stringently must be not only efficient, but also constant over time!

4. INFLUENCE OF TEMPERATURE ON Q-SWITCHED ND:YAG LASER PERFORMANCE Under engine-like operation conditions, the laser sparkplug being in immediate contact with the cylinder will work at crystal temperatures T2 up to 100°C and hence the question of T2 influence on laser performance has to be investigated.

For this purpose, a heatable and adjustable laser crystal holder was designed. In order to reduce any T2 related effects of the pump beam absorption, the emission wavelength of the diode was set 0.6 nm over the maximum absorption to a T2



300 ps

Illpppulse

um

laser

pulse

t [ps]

laser at room temperature laser shifts to later pulse timing no pulse formation

0.50

0.45-

0.40-

E 0.35

0.30-

0.25-

0.20-

(a)

Nd YAG - temperatures-.-22°C--150 °C--200 °C

0.1518 20 22 24 28 28 30

diode temperature ]'C] (- 0.3nmIC)

20

18

18

14

É̂12

P 10

em 8

6

4

2

0

/

(b)

30 °C

100 °C

125 °C

150 °C

175 °C

200 °C

20 40 60 80 100 120 140 160

pump energy rind] @ 300ps

7 5-

m 5-

8 4

0

(c)

012

010

- 0.07 c

0.05

o

- 0.02

0.00

50 100 150 200 250 300

temperature ]°C]

insensitive point (Fig. 7 (a)). A clear dependence of the Q-switched output pulse energy could be measured in different open and monolithic setups [27]. Consequences of this and other results will be discussed further down.

Fig. 7 (b) shows results attained with the monolithic crystal setup being placed in the heater. The graphs represent energy slopes at several T2. It is remarkable that all slopes seem to originate in zero, which would mean that the laser had no lasing offset. This is explainable by the pump light adjustment to the respective conditions which results in the effect, that always the same intensity was achieved in the pumped volume. In other words, the pumped volume scaled with the energy.

The graph in Fig. 7 (c) traces the optical efficiency as a function of T2 showing an accelerated decrease beyond 150°C. This finding clearly recommends cooling for higher T2 ranges.

Figure 7. (a) Transmission of pump light through a 5 mm long Nd:YAG crystal at different crystal temperatures T2 vs. pump diode temperature T1. At T1 = 27°C, pump transmission showed minimum T2 dependence (⇒ diode operating point); (b) Output versus input energy slopes of monolithic laser setup for different T2; (c) Influence of T2 on the optical efficiency expressed via pulse energy; (pump duration 300 µs, pump power 278 W). [9]

Generally, while heating up the crystal, the pulse formation time shifted continuously to later timings. Since the working point of the laser was already close to the falling edge of the pump phase, the incoupling geometry had to be readjusted to maintain laser operation (Fig. 8). This means, in terms of the pump light intensity distribution in the crystal that the radiation needed to be focused to smaller volumes, similar as if the pump power would be reduced.

Figure 8. Pulse formation timing at fixed incoupling geometry depending on crystal temperature T2. With increasing T2 it shifts to later times, first leading to increased pulse energy (more optical pump energy stored in resonator). Eventually, after the pump phase is exceeded, pulse formation will fail. [9]

A closer look on the beam profiles via a beam profiler reveals that the decrease in efficiency is accompanied by a de-crease in beam diameter as well as in mode order, which can be ascribed to the reduced pump volume. Fig. 9 depicts these findings in dependence of T2. At RT (1) the laser was running in a nearly perfect cylindrical TEM10 mode yielding a wide and energetic beam. A distinct decrease in beam cross section was already observed at 100°C (4), however, the laser was still operating in the same mode. Around 125°C (5), the beam profile started to jump into the fundamental mode which strongly constricted the beam diameter. On basis of the non-normalized profiles taken by the profiler it is



7

r'

l beam diameter

2

3

_4 4`

pulse energy `t

8

7

8

4.0 oo25 50 75 100 125 150 175 200

temperature [t]

discernible, that the maximum intensity in the center stayed at a constant level for all temperatures. Additionally, an in-crease in pulse duration was observable from 1.3 ns at room temperature to 2.0 ns at 200°C.

Figure. 9. Temperature (T2-) dependence of pulse energy and corresponding beam diameters. The numbers between the graphs indicate discrete points of measurement of beam profiles (@ 50 cm distance) from which values of beam diameter were taken (monolithic resonator). [9]

It is a fact that the gain in the crystal decreases with T2. The reason for this efficiency drop may lie in several competing mechanisms: i) Florescence lifetime: Kaminskii reports no changes in the radiative lifetime for a 0.3 at.%-doped Nd:YAG within T2 up to 600°C [37], however, he also points out that for doping levels near or in the concentration quenching regime the lifetime does not remain constant with increasing T2 [38]. Minor changes over a range of 160°C (−80°C to +80°C) were observed by [39] and [40] assuming a homogeneous lifetime reduction with increasing T2; ii) Stimulated emission cross section: [39],[41] report a linear decrease in the emission cross section of Nd:YAG in a range between −80°C and +80°C with a slope of ∼0.3×10-19 cm2/100°C; iii) Changes in the crystalline field due to thermal ex-pansion → changes in the Stark splitting; iv) Thermal (homogeneous) line broadening [42]; v) Cr4+:YAG: the above mentioned considerations are also valid, moreover T2 might be more influential since Chromium is a transition metal ion; vi) Coated dielectric mirrors: the thickness and the refractive index of the layers changes with T2 which also leads to thermal stress. This may affect the resonator; vii) Boltzmann occupation of the lower laser level can be neglected within that T2 range, since a 1% occupation will not be reached below 350°C (for comparison, in Yb:YAG, a typical quasi-three level system, a population of 1 % is reached already at -100°C). On the other hand, at the excited state, it may cause sta-tistic population differences between the 4F3/2 state and the several close positioned multiplets (4F5/2, 2H9/2,…).

As a matter of fact there is little, sometimes even controversial sounding literature dealing with lasers operating at higher T2. Bass et al. and Kimmelma et al. deal with a flashlamp-pumped resp. a longitudinally diode-pumped laser [43],[44]. Both groups report an increase in pulse energy with increasing T2. This results from their fixed incoupling setup, which probably leads to longer pump durations under thermal load and, therefore, to higher energy storage in the crystal.

5. GAS ENGINE RESULTS AND PERSPECTIVES The proven advantages of laser ignition of gas engines lie, according to the joint experience of the two cooperating groups and in correspondance with the literature [21],[45], in some increase in efficiency (comp. Fig. 10, where higher in-cylinder pressure and earlier heat release are the favorable features compared to SI), clearly better engine performance expressed by the coefficient of variation COV of the IMEP as shown in Fig.11, also allowing lower idle speed, better ignition of lean mixtures, lesser exhaust of NOx and HC.

Presently, to a large extent it is cost-related arguments which prohibit the introduction of LI to gas engines at a commer-cial level. Longterm performance also is still on the stake. With the abundance of natural gas in the present world, a strong motivation is set to employ gas engines on a large scale and simultaneously develop specific models optimized for LI. Exploitation of the high knocking resistance of this fuel paired with the special capability of LI to ignite high pres-



r P

ress

ure

(bar

)

NW

AU

IO

oo

o'

I

''

II

II

I

010)

VC

O(O

OO

O O

OO O

;ase

(J

/deg

.)

c

U10

0

- 20

- 10

0

-10 ¡ I ¡ ¡ i 10-80 -60 -40 -20 0 20 40 60 80

Crank Angle (deg.)

3.

2.

2.

.-.c*"

E 1.2

OU

1.

5

5

0.5

-0--SI_21CA-0r L1_21 CA

- SI_25CA-0- LI_25CA

-r - SI 29CAA LI 29CA

-

0.00.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40

Figure 10. Comparison of in-cylinder pressure and heat release rate for LI and SI for wide open throttle (WOT) at 1500 rev/min, λequ = 1.2 [46].

Figure 11: COV of IMEP for LI vis-à-vis SI w.r.t. λequ at ignition timings of (a) 210 BTDC, (b) 250 BTDC, and (c) 290 BTDC; (IMEP…indicated mean effective pressure, BTDC…before top dead center).[26]

sure gas mixtures would allow impressive further improvements of performance. The upcoming of a mixture of hydro-gen and methane, named Hythane, would even more stimulate such an approach. Hydrogen [46],[47] allows for lower MIE and higher λequ and goes along with faster flame front propagation. All these features are favorable towards exploit-ing the best properties of LI.

6. CONCLUSION Based on a literature review concerning VCSELs and specific experiments on a longitudinally diode-pumped laser sparkplug, it was demonstrated that the temperature effect on both major sparkplug components, the pump diode array (according to our opinion preferentially VCSELs) as well as the Q-switched SSL, is non-negligible. Both will have to be operated at elevated temperature to be kept constant for the diodes, while for the laser a compromise between cold-starting properties and equilibrium temperature has to be found. Additional work based on temperature-stabilized LDs



[48] or hot-band pumping in SSLs [49] should not be neglected, although before practical application there might be a long development effort necessary.

ACKNOWLEDGEMENT

The first author wants to express his gratitude to IIT Kanpur, and especially to A. K. Agarwal, for organizing and financ-ing his visiting professorship in Kanpur. Very generous hospitality is appreciated. Simultaneously, thanks should be ex-pressed to Vienna University of technology for allowing the necessary leave. The funding for developing a laser fired CNG engine from Science and Engineering Research Board (SERB), Department of Science and Technology (Grant No. SR/S3/MERC-0101/2010 (G)) is gratefully acknowledged.

REFERENCES

[1] Phuoc, T. X., “Laser-induced spark ignition fundamental and applications”, Optics and Lasers in Eng. 44, 351-397 (2006).

[2] Maker, P. D., Terhune, R. W. and Savage, C. M., “Optical third harmonic generation”, 3rd Int. Conf. Quant. Elect., Paris 2, 1559-1572 (1963).

[3] Dale, J. D., Smy, P. R., Clements, R. M., “Laser ignited internal combustion engine – an experimental study,” in Proceedings of SAE Congress Detroit, SAE paper 780329 (1978).

[4] Koechner, W., “Solid-state laser engineering”, 6th ed., Springer Series in Optical Sciences, New York, Springer, pp. 764, ISBN: 0-387-29094-X, (2006).

[5] Sakai, H., Kan, H. and Taira, T., “>1 MW peak power single-mode high-brightness passively Q-switched Nd3+:YAG microchip laser”, Opt. Express 16, 19891-19899 (2008).

[6] Dubinskii, M., Goff, J. R., Merkle, L. D., Castillo, V. K. and Quarles, G. J., “Q-switched laser operation of 8% ceramic Nd:YAG”, in Conference on Lasers and Electro-Optics/Quantum Electronics and Laser Science and Photonic Applications Systems Technologies, Technical Digest (CD) (Optical Society of America), paper CTuQ3 (2005).

[7] Zhang, X., Zhao, S., Wang, Q., Zhang, Q., Sun, L. and Zhang, S., “Optimization of Cr4+-doped saturable absor-ber Q–switched lasers”, IEEE J. of Quantum Electronics 33, 2286-2294 (1997).

[8] Sennaroglu, A., Demirbas, U., Ozharar, S. and Yaman, F., “Accurate determination of saturation parameters for Cr4+-doped solid-state saturable absorbers”, J. Opt. Soc. Am. B 23, 241-249 (2006).

[9] Kofler, H, “Development of a Laser Spark Plug and Comparative Testing,” PhD Thesis, Faculty of Electrical Engineering and Information Technology, Vienna University of Technology, Vienna, Austria (2009).

[10] Kofler, H., Schwarz, E. and Wintner, E., “Experimental development of a monolithic passively Q-switched di-ode-pumped Nd:YAG laser”, European Physical J. D 58, 209-218 (2010).

[11] Homepage Princeton Optronics: “Introduction to vertical-cavity surface-emitting diode laser (VCSEL) technol-ogy”; http://princetonoptronics.com/technology/technology.php (25.07.2013)

[12] U-L-M Photonics, Lise-Meitner-Strasse 13, 89081 Ulm, Germany; http://www.ulm-photonics.com/. [13] Radziemski, L. J. and Cremers, D., “Laser-induced plasmas and applications”, Marcel Dekker Inc., New York,

(1989). [14] Schwarz, E., Gross, S., Fischer, B., Muri, I., Tauer, J., Kofler, H. and Wintner, E., „Laser-induced optical

breakdown applied for laser spark ignition”, Laser and Particle Beams 28, 109–119 (2010). [15] Paul, D.R., “Laser versus conventional ignition of flames”, Optical Engineering 33, 510-521, (1994). [16] Weinrotter, M., Ast, G., Kopecek, H. and Wintner, E. “An extensive comparison of laser-induced plasma igni-

tion and conventional spark plug ignition of lean methane-air mixtures under engine-like conditions”, SAE Technical Paper Series 1972, 248-256, (2005).

[17] T.X. Phuoc, and F.P. White, “Laser-induced Spark ignition of CH4/air mixtures”, Combustion and Flame 119, 203–216, (1999).

[18] M. Weinrotter, H. Kopecek, M. Tesch, E. Wintner, M. Lackner, and F. Winter, “Laser ignition of ultra-lean me-thane/hydrogen/air mixtures at high temperature and pressure”, Experimental Thermal and Fluid Science 29, 569-577, (2005).

[19] Tauer, J., Kofler, H. and Wintner, E., “Laser-initiated ignition”, Laser & Photonics Reviews 4, 99–122 (2010).



[20] Graham-Rohe, D., “Lasers for engine ignition”, Nature Photonics 2 (9), 515-517 (2008). [21] Gupta, S. B., Bihari, B., Biruduganti, M. and Sekar, R.,“Laser ignition in natural gas engines”, paper LIC9-1,

The 1st Laser Ignition Conference (LIC'13), Yokohama, Japan, Apr. 23-25, 2013. [22] Mullett, J. D., Dearden, G., Dodd, R. D., Shenton, A. T., Triantos, G. and Watkins, K. G., “A comparative study

of optical fibre types for application in a laser-induced ignition system”, J. Opt. A 11, 054007 (2009). [23] Tauer, J., Kofler, H., Schwarz, E. and Wintner, E.,“Transportation of Megawatt millijoule laser pulses via opti-

cal fibers?“, Cent. Eur. J. Phys. 8(2), 242-248 (2010). [24] Yalin, A. P., Joshi, S. and Wilvert, N., “Fiber optic delivered laser ignition systems”, paper LIC5-1, The 1st La-

ser Ignition Conference (LIC'13), Yokohama, Japan, Apr. 23-25, 2013. [25] Weinrotter, M., Schwecherl, B., Kopecek, H., Wintner, E., Klausner, J. and Herdin, G., “Laser ignition of me-

thane-air mixtures at high pressures and temperatures”, Proc. of European Combustion Meeting, Louvain-la-Neuve, Belgium (2005).

[26] Srivastava, D. K., “Experimental investigations of laser ignition of natural gas-air mixtures for engine applica-tions”, PhD Thesis, Department of Mechanical Engineering, Indian Institute of Technology Kanpur, India (2013).

[27] Lackner, M., Kofler, H., Tauer, J. and Wintner, E., “Laser-induced ignition for combustion engines”, Lasers in Chemistry, 969 – 1000, ed. by M. Lackner; Wiley-Vch, Weinheim, Germany, (2008).

[28] Taira, T., “High brightness microchip lasers for engine ignition”, Frontiers in Optics: Advanced Solid-State La-ser Technologies and Applications, Conference paper FM3G, Rochester, New York, Oct. 14-18, 2012.

[29] Taira, T., Morishima, S., Kanehara, K., Taguchi, N., Sugiura, A. and Tsunekane, M., “World first laser ignited gasoline engine vehicle”, paper LIC3-1, The 1st Laser Ignition Conference (LIC'13), Yokohama, Japan, Apr. 23-25, 2013.

[30] Kroupa, G., Franz, G. and Winkelhofer, E., “Novel miniaturized high-energy Nd:YAG laser for spark ignition in internal combustion engines”, Optical Engineering 48, 014202-1-5, (2009).

[31] Kroupa, G., Baumgart, M. and Tortschanoff, A., “A highly robust, miniaturized Nd:YAG laser for spark igni-tion, paper LIC5-4, The 1st Laser Ignition Conference (LIC'13), Yokohama, Japan, Apr. 23-25, 2013.

[32] Kroupa, G., CTR Carynthian Tech Research AG, Europastrasse 4/1, Technologiepark Villach, A-9524 Villach / St. Magdalen, Austria (personal communication, 2013).

[33] King, R. Grabherr, M., Wiedenmann, D. and Jäger, R., “VCSELS address a wide range of applications”, U-L-M Photonics GmbH (a Philips Lighting company), Lise-Meitner-Str. 13, 89081 Ulm, Germany (2007) © 2007 Mi-crooptics Group (OSJ/JSAP).

[34] Helms, C., Aebya, I., Luoa, W., Herrick, R. W. and Yuen, A., “Reliability of Oxide VCSELs at Emcore”, in “Vertical-cavity surface-emitting lasers” VIII, ed. by Lei, Ch., Choquette, K. D., Kilcoyne, S. P., Proc. of SPIE 5364 (SPIE, Bellingham, WA) 183-189, (2004).

[35] Li, T., Ning, Y., Sun, Y., Cui, J., Qin, L., Yan, C., Zhang, Y., Peng, B., Liu, G., Liu, Y. and Wang, L., “Tem-perature dependent characteristics of 980 nm two-dimensional bottom emitting VCSEL arrays”, Chinese Optics Lett. 5, Supplement, S156-S158, (2007).

[36] Seurin, J.-F., v. Leeuwen, R. and Ghosh, C., “High-power VCSELs mature into production”, Laser Focus World, April 2011, 61-65.

[37] Kaminskii, A. A., “Laser crystals − their physics and properties”, 2nd ed., Springer Series in Optical Sciences, eds. MacAdam, D. L., Schawlow, A. L., Shimoda, K., Siegman, A. E. and Tamir, T., Berlin, Springer, pp. 456, ISBN: 0-387-52026-0, (1990).

[38] Kaminskii, A. A., “High-temperature spectroscopic investigation of stimulated emission from lasers based on crystals activated with Nd3+-ions”, Physica Status Solidi (A) 1, 573-589 (1970).

[39] Rapaport, A., Zhao, S., Xiao, G., Howard, A. and Bass, M., “Temperature dependence of the 1.06- μm stimu-lated emission cross section of Neodymium in YAG and in GSGG”, Applied Optics 41, 7052-7057, (2002).

[40] Chen, X., di Bartolo, B., Barnes, N. P. and Walsh, B. M., “Thermal tuning and broadening of the spectral lines of trivalent Neodymium in laser crystals”, Physica Status Solidi (B) Basic Research 241, 1957-1976, (2004).

[41] Zhao, S., Rapaport, A., Dong, J., Chen, B., P. Deng, B. and Bass, M., “Temperature dependence of the 1.064-µm stimulated emission cross-section of Cr:Nd:YAG crystal”, Optics and Laser Technology 38, 645-648, (2006).

[42] Koechner, W. and Bass, M., “Solid-state laser − a graduate text”, Advanced Texts in Physics, New York, Springer, pp. 409, ISBN: 0-387-95590-9, (2003).



[43] Bass, M., Weichman, L. S., Vigil, S. and Brickeen, B. K., “The temperature dependence of Nd3+-doped solid-state lasers”, IEEE J. of Quantum Electronics 39, 741-748, (2003).

[44] Kimmelma, O., Tittonen, I. and Buchter, S. C., “Thermal tuning of laser pulse parameters in passively Q-switched Nd:YAG lasers”, Applied Optics 47, 4262-4266, (2008).

[45] Dearden, G. and Shenton, T., “Laser ignited engines: progress, challenges and prospects”, paper LIC5-5, The 1st Laser Ignition Conference (LIC'13), Yokohama, Japan, Apr. 23-25, 2013.

[46] Srivastava, D. K. and Agarwal, A. K., “Laser ignition of single cylinder engine and effects of ignition”, SAE In-ternational, in press (2013).

[47] Srivastava, D., Weinrotter, M., Iskra, K., Agarwal, A. and Wintner, E., “Characterisation of laser ignition in hy-drogen-air mixtures in a combustion bomb”, International Journal of Hydrogen Energy 34, 2475 - 2482 (2009).

[48] Tsunekane, M. and Taira, T., “Ignition lasers operating for wide temperature range”, paper LIC2-3, The 1st La-ser Ignition Conference (LIC'13), Yokohama, Japan, Apr. 23-25, 2013.

[49] Sato, Y. and Taira, T., “Temperature dependence of the pump absorption efficiency under hot-band pumping of Nd:YAG”, paper LIC2-2, The 1st Laser Ignition Conference (LIC'13), Yokohama, Japan, Apr. 23-25, 2013.



laser plasma ignition: status, perspectives, solutions

Documents