High Power Continuous Wave Fiber Laser System
Disclaimer: This dissertation has been submitted by a student. This is not an example of the work written by our professional dissertation writers. You can view samples of our professional work here.
Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UK Essays.
Introduction of High Power Fiber Laser
The optical fiber with very high surface-to-volume ratio and a strong waveguide effect provides the fiber based laser source the potential to generate high power laser beam with high quality. In addition to the capacity of generating raw optical power with high beam quality, the fiber laser system has other appealing features, such as supporting robust and compact system designs, allowing ultrashort pulse operation, offering a board wavelength tunability, and providing high gains. Those features stimulate the research on the high power fiber lasers system, and lay the foundation of novel appealing applications, such as remote material processing, aerospace and defense. In the past decade, a remarkable increase of the powers produced by fiber lasers with high beam quality has been achieved (see Fig.1). As a result, the high power laser becomes strong counterpart of the solid-state bulk laser, and penetrates rapidly into areas that formerly other lasers were used.
In the early 1960s, the first fiber laser was demonstrated by Snitzer. The doped fiber's potential for high optical gain was revealed by David Payne and co-researchers' working on Neodymium- doped fibers in mid 1980s . In 2009, the high power fiber laser, which based on a specifically silica-host ytterbium-doped fiber-based laser (YDFL), obtained 10 kW output in the single-mode (SM) regime.
Although architectures are different, the high-power fiber lasers and amplifiers are mostly archived with rare-earth-doped (RE-doped) double-clad fibers. The double-clad fiber, which was initially demonstrated in 1988, provided the option of cladding pumping, and proved to be one of the key technologies for power scaling. The structure of this double cladding is that the active RE-doped core is surrounded by a much larger 'inner cladding' (see Fig. 2), and are encircled together by out cladding. The pump beam emitted by fiber-coupled high-power diode bars or other kinds of laser diodes is coupled into the inner cladding, and confined within it by an outer cladding. The confined pump beam will be absorbed into the core while it propagates along the fiber. The laser light is generated in the central core, and the laser light can have very good beam quality ' even diffraction limited beam. Thereby, by means of double cladding configurations one realized the conversion from low brightness pump to high brightness single-mode fiber laser output. As the spatial and angular pump acceptance [can be expressed as the product of area and the square of the numerical aperture (NA)] for the inner cladding is significantly improved to the core pump, Such conversion is more effective, and close to 5 orders have been demonstrated experimentally.
Among high power RE-doped fiber lasers, the YDFL is notable in term of high power. The Yb's broad absorption band extends from 900 to 980 nm (see fig 3), covering high power pump LD's best performance wavelengths, offers a low quantum defect [energy difference between pump and laser photons] for pumping with 9xx nm LD and lasing above 1040nm. This superior property offers the potential for achieving very high power efficiencies and reducing thermal effects. In addition, lasing at wavelength above 1040nm, the Yd ion shows a simple four 'level structure, that excludes excited state absorption and also a variety of detrimental quenching processes allowing high doping concentrations, which means high pump absorption per unit length. On the contrary, the small quantum defect also has a usually unwanted consequence: the significant quasi-three-level behavior, especially when lasing at short wavelengths (less then 1040nm), that will cause a high threshold and decrease the power efficiency.
Fig. 2. Structure of a double-clad fiber and principle of cladding-pumping
The Nd doped laser emitting at 1060 nm is a four-level system, which means a lower laser threshold. Associated with the relatively advanced state of 808 nm diodes for pumping Nd:YAG, this made Nd the choice for high power fiber lasers in early stage. Today's high power pump diodes in 9 xx nm are sufficiently bright to make threshold unimportant for most quasi-three-level high power fiber lasers. These overcome the obstacle of ytterbium's higher threshold and raise advantages of a lower quantum defect and higher doping concentration with quench-free. The first single-mode Yb doped fiber laser with output power over 100 W was demonstrated in 1999 , and it illustrated that the advantages of Yb doped double cladding structure can support for further increase in the average power by scaling the size of the optical fiber and the power of pump diode source. Soon after that, the power of cladding pumped YDFL obtained the kilowatt level. Thereafter, by investigating the large-area core design and fabrication, the single-mode operation in kilowatt level was realized that would not have been possible for Nd doping.
Figure 3: Absorption and emission cross sections of ytterbium-doped germanosilicate glass, as used in the cores of ytterbium-doped fibers.
Another sophisticated technique which is adopted in all double-clad fiber lasers at 3 kW and above, is tandem-pump [in-band pumping with high-brightness pump sources, such as one or several fiber lasers, or thin disk laser]. The tandem pump makes it possible to pump close to the emission wavelength so that the quantum defect heating will be low resulting in a reduced thermal load. Actually, some advanced solid-state lasers, such as thin disk laser, is well matched with requirements of in-band high brightness pumping source, and 1 KW level output thin disk laser pumped fiber laser have been realized.
Nonlinearities are an issue to further increase the CW output power of the fiber laser. The fiber laser considered above such as in Fig. 1 has operated with linewidths in the 1'10 nm range. In such system with cw operating, excepting at the extremely high powers or long delivery fibers, the stimulated Raman scattering (SRS) is a weak effect and is relatively easy to prevent. However, for output power above 10 kW, the Raman gain can become so high (tens of decibels) that a considerable part of the power is transferred to a longer-wavelength Stokes wave, reducing the power in the signal wavelength. There are some applications need single frequency sources which can provide light power with narrower spectral line width, such as coherent beam combination of multiple single frequency fiber sources with high power. This scheme offers a promising method for further power scaling, and consequently this stimulates interest in single-frequency power scaling. For narrow bandwidth, especially at linewidths less than 10 MHz, the SBS is the dominate nonlinearity and the severe obstacle for high power single-frequency fiber sources. The SBS can be suppressed with shorter fiber and larger mode field area, and output power of hundred watts has been reported with such schemes .However, this power is still less comparing with the bulk solid state laser. There are several options for SBS mitigation, including straining the fiber in order to broaden the SBS gain bandwidth, and reducing overlap between the optical and acoustic fields. The highest power high-gain fiber amplifier can archive 1.7KW. It was realized by combination of the modest spectral broadening with phase modulation and the fiber with enlarged effective mode area.
The most effective way to mitigate nonlinearities (excluding self-focusing) is to enlarge the effective mode area by optimizing geometry designs and material choices of fiber structure. Unlike the passive power delivery fiber, this task is more challenge for active fiber, as doping-induced refractive index changing, and thermal stability will be issues. A straightforward design approach to maintain pure single-mode operation is to increase the core diameter, with the NA reducing correspondingly. However, the downside is the waveguide effect gets weaker, and consequently light is easier lost from the core when the fiber is bent. More works on fiber designs for addressing these challenges are related to photonic crystal fibers techniques. It is possible to make single mode operation in a multi-mode supported fiber, by building up preferentially amplify, or attenuate for specific mode, while the mixing or coupling between modes should be controlled to minimum. There have been works focused on using differential gain by selective RE doping across the core , and differential-bend-loss by controlled bending of the fiber. The leakage channel and chirally coupled core fibers are designed to selectively couple propagation mode to high loss mode. The high order mode but not the fundamental mode is coupled to leaky mode, which will substantially be attenuated. The multifilament core and multicore fiber arrange filaments or cores in a two-dimensional array. There are evanescent-field coupling among cores, and the overall structure can exhibit single-mode guidance with large mode area.
From the aspect of power generation, Investigation of advanced fiber for mitigating nonlinearity will be still the most critical issue in increasing the output power for cw fiber lasers. It has been estimated that the maximum single core output powers of the ytterbium doped fiber laser should be at several tens of kilowatts level based on present technique. However, single-mode operation is not indispensable for lots of high power lasers applications. The single or near single mode operation in the MM fiber which is developed by balancing the mode quality, the achievable power, and the damage threshold of the fiber, can offer possibilities to archive higher output power. In addition to that, as the emission wavelength of well developed thin disk laser is still covered by the Yb ion's absorption band and is more close to the emission band of the Yb ion, the research on novel architecture using thin disk laser to tandem pump the special designed Yb doped fiber laser also offers the potential to increase output power of fiber laser and develop novel fiber laser with useful function.
The proposed research will focus on advanced fiber, especially for the evanescent-field coupled waveguides, also called multi-core fibers (MCF). The main aim is to design and realize novel types of active MCF for increasing the output power of fiber laser with good beam quality, and for suppressing the SBS effect. Besides that, Based on the advanced thin-disk laser, and the novel MCF fiber, the investigation on the novel laser architectures will also be performed.
The Outline of the Project
According to the above proposed objectives, the research work can be divided to two main phases. The first phase will be focused on fiber design and fabrication, and the expected deliverable is the novel active fiber with improved performance in nonlinearity mitigation and bending resistant. The other phase is about the novel fiber laser architecture, and investigation of the novel tandem pump configuration based on thin-disk laser will be performed.
Mathematical Model and Design Strategy
The main nonlinearities for cw operating fiber laser is SRS and SBS. Although both of them can be mitigated by the enlarged mode area, the SBS is still too strong for increasing the power of single frequency laser in the LMA fiber. The proposed research aimed to suppress the SBS in the LMA fiber for mitigating both SRS and SBS. According to the previous research, the SBS threshold can be expressed by :
The Î±u is acoustic attenuation coefficient for the acoustic mode of order u, Aeff is the optical effective mode area, G(Ñ´max) is the SBS effective gain coefficient at the peak frequency, K is the polarization factor. We can see form the equation. Beside the mode area, the SBS can be suppressed by increasing the acoustic loss, reducing the overlap integral, and the SBS effective gain coefficient. The is the normalized overlap integral of the electric and acoustic fields and it can be expressed as :
The E0 is the optical field associated with the fundamental mode, and Ï?u the field of a longitudinal acoustic eigen-mode of order u. The overlap integral can be changed by modifying the fiber refractive index profile and acoustic velocity profile. The acoustic loss can be changed by glass composition design. As different dopants have different effects on optical and acoustic properties, it is possible to create suitable dopants profile in the core and cladding to reducing the overlap integral or increasing the acoustic loss. Table 1 is some common dopants used for making silica glass based fibers. The profile of the optical and acoustic field can be indicated by optical and acoustic refractive indices. Similar to the optical refractive index, acoustic refractive index is defined as na(r) = VL Silica /VL (r) , where VL(r) is the longitudinal acoustic velocity in the core, and VL Silica is the longitudinal acoustic velocity of pure silica glass.
Table 1. Trend of optical and acoustic refractive index change of different dopants in silica
Optical refractive index
Acoustic refractive index
One straightforward approach to modify the loss for optical and acoustic field of a fiber structure is created a type of optical guiding and acoustic anti-guides with a dopant material(Fig 4 (a)), such as Al2O3, and it has been demonstrated in . The other approach is to reduce optical and acoustic field overlap, with different dopants in the core (Fig 4 (b)). The resultant optical and acoustic refractive index profiles of above approaches are shown below.
Fig. 4. Dopant designs for reducing the overlap of the optical and acoustic fields
The strategies shown in fig 4 are based on single core fiber. There are quite a few research works on improving the effective mode field in single core fiber, and it is little room to enlarge the effective mode field areas further without detrimental effect in single core fiber LMA. Recently, multicore fiber based LMA has been investigated as passive delivery fiber , and as active fiber in the novel laser architecture. The supported optical mode field of MCF can be designed by core size and core interval; the profile of the acoustic and optical field can be modified by the distribution of dopants and doping area size; and the loss of the optical and acoustic can be controlled by doping material. Thus, it is worth investigating a novel active MCF supporting a few modes or only single supermode with the reduced overlap between the acoustic and optical field.
Yb Ge F/B
Fig 5 the schematic of the proposed 19 core double cladding fiber
A fiber design strategy to suppress the SBS is shown in Fig 5. An optical guide while acts as an acoustic anti-guide in the effective optical field areas of MCF will be fabricated by manipulating dopants in core and cladding, for example as shown in fig 5, by choosing Al2O3 in core and GeO2 in cladding. Because the fields of optical and acoustic are separated, the interaction between the optical and acoustic waves is weaker. Furthermore, the MCF will be designed to support a few modes or only one supermode, that benefits for manipulating refraction index to increase threshold of the SBS in the single mode MCF, as the âˆ†n=neffâˆ'nclad of single mode MCF is larger than single mode single core fiber(SMF), for example, the index difference âˆ†n=neffâˆ'nclad is 3.69Ã—10âˆ'4 in the case of the 19-core fiber reported in and only 1.60Ã—10âˆ'4 for the SMF, providing more room of âˆ†n for manipulating dopants. Finally, the reference value of parameters such as the diameter of each core, the core interval, and the doped areas of each core, will be archived by numerical calculation.
The optical field in the waveguide can be solved by numerical calculation the Maxwell equations. Like the optical field, by numerical calculating the nonlinear acoustic equation, the acoustic field can be obtained. After that the SBS threshold can be calculated with equation (1), (2).
From the nonlinear acoustic equation, we can obtain the equation that determines the longitudinal acoustic eigen-modes. The acoustic modes that contribute to the SBS associated with the optical fundamental mode have constant azimuth. Neglecting the damping factor, the radial distribution of such a mode can be expressed as:
The Î©u is the acoustic frequency and the Î²u is the propagation constant of the acoustic mode, VL(r) is the longitudinal acoustic velocity profile across the fiber. The wave equation for optical field in waveguide is derived from the general Maxwell and can be written as:
The EO is the optical field, ko=2p/l is the wave number of the optical field, and no(r) is the refraction index profile across the fiber. The optical mode is efficiently backscattered by the acoustic mode when the phase-match condition, Î² = 2Î²u, is fulfilled, where Î² is the propagation constant of the optical field. The Î² is determined by the optical wavelength Î», the effective refractive index no,eff, and it can be expressed as: Î² =kono,eff=2pno,eff/l.
Determined by the structure, the acoustic field in the proposed 19 core fiber is confined in the inter cladding, and the acoustic index can affect the confining effect. As the position of inter cladding is fixed, once the doping concentration is chosen, the acoustic field will be determined. The optical field in the MCF is determined by both the doping concentration and the geometry structure of the MCF. It is the geometry structure of the MCF provides the extra room to design the optical field with desired mode.
The field of a longitudinal acoustic (Ï?u) can be numerical calculated with finite-element method. beside the finite-element method, previous research has indicated the Ï?u can be solved by utilizing the solver for optical scalar wave equation after defining a few new terms for acoustic wave. to numerical solve the equations (4), as numerical calculation by the finite element method is still valid when strong coupling exists between the different cores, the mode structure of the optical filed in the MCF is also calculated by finite-element method based on commercial available software such as Fimmwave or Comsol-Multiphysics.
After knowing the Ï?u and the EO, the can be calculated by taking the Ï?u and E0 into equation (2). Finally, taking the into equation (1), the Pth of the designed fiber can be obtained. The theoretical M2 propagation factor can be computed with the method in. For a doping state, different Pth value and mode structures can be achieved for different geometry parameter, such as single core diameter and core interval. Optimizing the geometry parameter is necessary to get the high Pth value with good mode structures. Finally, repeatedly implementing above step for different doping condition, a series of optimized reference parameters can be obtained.
The home institute - IFSW has equipped the fiber manufacturing facility ' consisting of a modified chemical vapor deposition (MCVD) preform production lathe and the new commissioned drawing tower. The prototype of the proposed double-clad MFC fiber will be produced in the IFSW by the stack and draw technique.
Investigation on the Laser Architectures
The mode field mismatch and high operation power set the obstacles on employing the state of art fiber communication components in the high power laser architecture. Beside that some of critical components for high laser, such Bragg gratings in the LMA cores, large mode area pump coupler or combiner for high power diode are still in the initial stages. Above aspects cause the architecture for the exited high power laser is limited comparing with well developed communication band fiber laser. Most of the previous research on increasing the output power focused on developing the LMA fiber. As the difficulty of increasing the output power by enlarging the mode field is increasing continuously, it is time to consider improve the laser output from other aspect.
The high brightness power scalable thin disk laser acting as the in-band pumping source can generate less quantum defect heat than 9Ã—Ã—nm laser diode, providing the potential to developing novel or improving existed laser structure by using components which are thermal damage or degradation sensitive. Thus there are rooms to increase the output power or improve the efficiency by developing the laser architecture with thin disk laser and special designed high power components.
One of proposed architecture improvements is to replace the butt-coupled HR-mirror in the existed laser with the Bragg grating in the core of the double cladding fiber.
Fig 6 (a) the butt-coupling mirror based laser architecture
Fig 6 (b) the FBG based laser architecture
FBG1,2 reflectors for the laser radiation; FBG3 reflectors for the pump light;
The reflection rate of the FBG1 is around 99%, that of FBG2is around 50%.
For the butt-coupled mirror based laser architecture (Fig 6 (a)), as the butt-coupled mirror will reflect both the pump light and the laser radiation on a very small area, the energy densities will be extremely high in core and cladding near the conjunct point. To withstand such high power densities, special material substrate such as the sapphire is needed to remove the heat very quickly. Even though, the core power density is still close to the damage threshold of the mirror coating. Furthermore, as the pump wavelength is closed to the laser radiation wavelength, to fabricate the dichroic mirror will be rather difficult. Although the Bragg grating inside the fiber core still has the problem of thermal damage, the damage threshold of FBG will be higher than mirror face. Considerable power increasing is expected for replaying mirror with the Bragg grating. From the aspect of fabrication, the wavelength of the Bragg grating is determined by the mask period and the refraction index of the fiber core, it will be easy to fabricate two FBG with spectrum interval larger than 4nm, which is enough to separate the in band pump light and laser radiation. Finally the FBG also can provide the facility to control the laser wavelength, and the laser output wavelength will be determined by the corresponded reflector. In fig 6 they are FBG1,2.
Fig 7 the proposed hybrid laser architecture
DMCF: doped multicore double cladding fiber, SMF: single-mode fiber, PMSF: polarization maintained single mode fiber ISO1: polarization dependent optical isolator, Amp: amplifier, DM: dichroic mirror, BS: beam splitter,Li (i = [1; 5]): plus lenses, FBG1: fiber Bragg grating for laser radiation, FBG2: fiber Bragg grating for pump light, PZT: Piezoelectric Ceramics
Limited by components, many well developed communication fiber laser techniques such as wavelength tuning and polarization stabilizing cannot be projected to high power fiber laser area directly. A promising method to solve the problem is to develop hybrid architecture which employs a low power single mode fiber feedback loop to control the high power laser. By applying the advanced communication laser techniques in the single mode feedback loop, the high power fiber laser with wavelength tuning and polarization stabilizing can be realized. The proposed hybrid laser architecture with wavelength tuning and polarization stabilizing is shown in fig 7.
The thin disk emitted pump light is coupled to the double cladding fiber with a dichroic mirror and a plus lens. The FBG in pump wavelength is employed in the far end of the active fiber. A small part of the emitted beam is reflected and coupled to the feedback loop whereas the most of power is coupled out from the laser cavity. The lens set (L4 to L6) constitutes a free-space imaging system for projecting the far field of the DMCF onto the input face of the SMF. Polarization independent optical isolators and circulator in the feedback loop determined the light traveling direction and eliminated the unwanted reflected light. Furthermore, the polarization independent optical isolator shapes the light to single polarization state, which will be persevered in the feedback loop by polarization maintained fiber. The FBG1 is fixed in a stretch preloaded piezoelectric ceramics, and the reflected wavelength can be tuned slightly by driving the PZT with Bi-directional signal. After passing the feed pack fiber, the DMCF will amplify the seed laser radiation, and consequently power will be scaled and the polarization can be preserved.
Cite This Dissertation
To export a reference to this article please select a referencing stye below: