High flatness supercontinuum fiber laser based on nanosecond / sub-nanosecond pulse and fiber length matching

Abstract

The application relates to a high-flatness nanosecond / sub-nanosecond pulse fiber laser based on nonlinear length matching, and belongs to the technical field of fiber lasers. The application aims to solve the problems of large spectrum fluctuation of an existing supercontinuum light source and excessively high residual pump peak in the 1-micron wave band. The technical scheme is as follows: a "nonlinear spectrum homogenization length" design criterion is proposed, a nanosecond / sub-nanosecond pulse seed source, a power amplification unit and a specific-length graded-index multimode optical fiber are cooperatively designed, the fiber length is preferably 30-150 meters, the length is matched with a pulse long walk-off length, the synergistic effect of cascade Raman scattering and geometric parameter instability is enhanced, pump light energy is efficiently transferred to a wider spectrum range, the pump peak is suppressed to be comparable to a continuous background, and a high-flatness supercontinuum spectrum is directly generated. Without complex multi-wavelength pumping or feedback, the application is full-fiber, simple and reliable, simultaneously utilizes a beam self-cleaning effect to maintain a near-Gaussian distributed beam quality, and is suitable for the fields of precision spectroscopy and imaging.

Description

Technical Field

[0001] This invention relates to the field of fiber laser technology, and specifically to a high-flatness supercontinuum fiber laser based on matching nanosecond / sub-nanosecond pulses with fiber length. Background Technology

[0002] Supercontinuum fiber lasers, due to their wide spectral coverage and high spatial coherence, have significant application value in precision optics fields such as optical coherence tomography, high-resolution spectroscopy, and biomedical imaging. These applications not only require a wide spectral coverage of the light source but also impose stringent requirements on the flatness of its output spectrum (i.e., the uniformity of the spectral power density distribution). Spectral fluctuations directly lead to problems such as uneven signal-to-noise ratio within the imaging depth and decreased accuracy of quantitative analysis. Therefore, achieving high-flatness supercontinuum output has become a key bottleneck driving this technology towards high-end applications.

[0003] Currently, technologies for generating supercontinuum mainly follow the following technical paths, but all of them have significant limitations in achieving high flatness:

[0004] The first type of approach is based on photonic crystal fibers (PCFs). This type of approach utilizes the extremely high nonlinear coefficients of PCFs to achieve efficient spectral broadening. For example, CN105759531B discloses a system using a multi-stage large-mode-field fiber amplifier combined with PCFs, aiming to improve average power; CN105762626B uses germanium oxide PCFs to extend the spectrum to the mid-infrared. However, this type of approach has inherent drawbacks: First, PCFs have small cores, making splicing with ordinary fibers difficult and the end faces easily damaged, making it difficult to adapt to stable injection of nanosecond / sub-nanosecond pulses, resulting in low system reliability and engineering sophistication (CN113659412A also points out this problem). More importantly, the resulting spectra are usually dominated by soliton dynamics, prone to strong modulation and discrete Raman peaks, generally exhibiting poor flatness, and lacking effective active modulation mechanisms.

[0005] The second type of approach is based on amplifying conventional optical fibers and then connecting them to nonlinear optical fibers. To overcome power limitations, solutions such as CN105759531B employ large-mode-area rare-earth-doped optical fibers to construct a master oscillator power amplifier (MOPA) system. Laser pulses are first obtained and then coupled to the nonlinear fiber for broadening. While this improves power handling capacity, new problems arise: the multimode beam output from the amplification chain excites complex multimode interactions in the nonlinear fiber, leading to disordered output spectral structure and uncontrollable flatness; if polarization-maintaining devices are used to suppress mode noise, as pointed out in CN103199415B, the system complexity and cost will increase significantly.

[0006] The third category of approaches aims to improve spectral characteristics, but employs complex control schemes. A few works focus on output stability or spectral width, but the methods used are typically costly or have limited applicability. For example, CN103199415B stabilizes output spectral power through real-time monitoring and feedback control of the polarization state; however, the system is complex and offers limited improvement to the flatness of nanosecond / sub-nanosecond pulses. CN104201545B uses two lasers of different wavelengths to combine and pump a cascaded fiber to splice a broad spectrum, but this system is structurally complex, costly, and does not address the spectral fluctuation problem within a single wavelength band.

[0007] It is particularly important to note that when the pump source is a nanosecond / subnanosecond pulsed laser (with a common center wavelength of 1 micrometer) widely used in industry and scientific research, the above-mentioned technologies face unique challenges in achieving spectral flattening. Commercially available 1-micrometer-band nanosecond / subnanosecond pulsed fiber lasers have a wide single-pulse energy range, typically from tens of microjoules to several millijoules, with repetition frequencies in the kHz to MHz range, corresponding to average powers ranging from tens to hundreds of watts. These lasers themselves possess good power stability and industrial reliability. However, due to their long time-domain width (typically 1-100 ns) and correspondingly long walk-off length, nanosecond / subnanosecond pulses significantly enhance stimulated Raman scattering in nonlinear fibers. Under traditional design approaches, this typically results in one or more strong, isolated Raman-Stokes peaks in the output spectrum, while a large amount of energy at the pump wavelength is consumed, forming deep valleys, severely disrupting the continuity and flatness of the spectrum. Existing schemes often suppress the Raman effect as a harmful factor or lack effective guidance for it, failing to utilize its energy transfer characteristics to actively shape a flat spectrum.

[0008] The fourth approach simplifies the process by using graded-index multimode fiber. Recent advancements, such as CN113659412A and prior user-provided patents, have shown that replacing photonic crystal fiber with graded-index multimode fiber with a larger core diameter can effectively reduce splicing difficulty and cost, and improve output beam quality through its "beam self-cleaning" effect. However, these works primarily focus on addressing power enhancement, beam quality, or spectral width issues, without considering "spectral flatness" as a core optimization indicator or the invention's objective. The resulting supercontinuum often exhibits significant intensity fluctuations over a wide spectral range, and its flatness has not been systematically designed and controlled.

[0009] In summary, existing technologies face a contradiction: solutions capable of achieving broadband output suffer from poor spectral flatness; while a few solutions focusing on spectral characteristics are impractical due to system complexity, high cost, or incompatibility with nanosecond / sub-nanosecond pulses. Therefore, how to leverage the characteristics of nanosecond / sub-nanosecond pulses and graded-index multimode fibers to actively guide nonlinear dynamic processes through ingenious rather than complex system design, thereby directly generating a high-flatness supercontinuum while ensuring broadband and high beam quality, is a pressing technical challenge in this field. Summary of the Invention

[0010] The technical problem to be solved by this invention is to address the prominent issues of poor output spectral flatness of nanosecond / sub-nanosecond pulsed supercontinuum light sources in the prior art, as well as the complexity and difficulty in practical application of existing flattening schemes. This invention provides a nanosecond / sub-nanosecond pulsed fiber laser with a relatively simple structure, capable of actively controlling the nonlinear process, thereby directly outputting a wide-spectrum, high-flatness spectrum while maintaining high beam quality.

[0011] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0012] A high-flatness spectral nanosecond / subnanosecond pulsed fiber laser based on nonlinear length matching is proposed. The core of this laser is to take "nonlinear spectral averaging length" as the core design criterion. By actively matching the temporal characteristics of the nanosecond / subnanosecond pulse with a section of graded refractive index multimode fiber of a specific length, a transmission system is constructed that can enhance Raman scattering and guide its synergistic effect with nonlinear effects such as geometric parametric instability, thereby achieving efficient spectral flattening.

[0013] The laser includes a nanosecond / sub-nanosecond pulse seed source, a power amplification unit, a mode matching unit, and a nonlinear spectral flattening unit arranged sequentially along the optical path.

[0014] Furthermore, the nanosecond / sub-nanosecond pulse seed source is used to generate an initial laser pulse with a center wavelength of 1064 nm; its pulse width is 0.11 nanoseconds to 100 nanoseconds, and its repetition frequency is 1 kHz to 1 MHz.

[0015] Furthermore, the power amplification unit adopts a master oscillation power amplification structure to amplify the power of the seed pulse; the single pulse energy of its output laser is in the range of microjoules to millijoules, and the average power is in the range of watts to tens of watts.

[0016] Furthermore, the mode matching unit is used to couple the amplified laser to the subsequent nonlinear spectral flattening unit in a low-loss, fundamental mode-dominated manner; it can be implemented through a transition fiber or a mode field adapter.

[0017] Furthermore, the nonlinear spectral flattening unit, which is the core component of the present invention, is composed of a section of graded-index multimode fiber of a specific length; the length of the fiber is specially designed to match the "nonlinear spectral flattening length" of the nanosecond / sub-nanosecond pulse in this fiber, and its preferred range is 30 meters to 150 meters.

[0018] The highly nonlinear multimode fiber exhibits a parabolic or approximately parabolic graded refractive index (GRIN) distribution. Its radial refractive index n(r) is formed by the distribution of refractive index modulating elements in the fiber core and can be characterized as follows:

[0019]

[0020] In the formula, n0 is the maximum refractive index at the center of the fiber core. R is the relative refractive index difference, r is the core radius, r is the radial coordinate, and α is the refractive index distribution coefficient. α is preferably between 1.8 and 2.2 to form a parabolic or approximately parabolic structure.

[0021] Through the specific parameter combination design described above, when nanosecond / subnanosecond pulses propagate in a selected graded-index multimode fiber, the cascaded stimulated Raman scattering (SPR) effect is significantly enhanced because its walk-off length is much greater than the physical length of the fiber. This enhanced Raman process not only continuously consumes pump light energy and generates significant redshift spectral components, but more importantly, it works in synergy with the naturally excited geometric parametric instabilities in the fiber. The multi-band discrete sidebands generated by GPI in the early stages of transmission provide a wide range of seeds for spectral broadening; while the enhanced Raman effect acts as an efficient "energy transporter," working in synergy with GPI and cascaded four-wave mixing and other nonlinear processes to drive the energy to be efficiently and uniformly redistributed across a wide spectral range on both sides of the pump wavelength, thereby effectively filling the spectral valley structure and ultimately achieving significant spectral broadening and high flatness output at the fiber output end. At the same time, the inherent beam self-cleaning effect of this graded-index fiber ensures that the output beam maintains excellent spatial quality with a near-Gaussian distribution throughout the entire spectral range during the aforementioned complex nonlinear spectral shaping process.

[0022] The core of this invention lies in achieving precise matching between the time-domain characteristics of nanosecond / subnanosecond pulses and the length of graded-index multimode fiber, so as to actively excite and utilize the flattened dynamic path of "enhanced Raman-geometric parametric instability synergy". To achieve this goal, we provide the following quantitative design method and derivation.

[0023] 1. Establishment of matching criteria: Analysis based on Raman walk-off length

[0024] The primary principle of matching design is to ensure that the propagation length of the nanosecond / subnanosecond pulse in the fiber is less than its first-order Stokes Raman walk-off length. This allows the Raman scattering process to occur almost "collinearly" throughout the fiber, enabling sufficient gain accumulation and thus significant enhancement.

[0025] For the center wavelength is Pump light and the first-order Stokes light it generates The walk-off length caused by its group velocity mismatch It can be approximated as:

[0026]

[0027] in, The pulse time domain width (full width at half maximum). For group velocity, It is the difference between the first-order dispersion (reciprocal of the group velocity) of the pump light and the Stokes light.

[0028] For a typical 1064 nm nanosecond / subnanosecond pulse (e.g. In quartz-based graded-index optical fibers, the group velocity difference of the corresponding first-order Stokes light (~1120 nm) is very small, and the calculated... Typically, the length of optical fiber ranges from several hundred meters to several kilometers. However, the optical fiber length selected in this invention... The distance is between 30 and 150 meters, therefore it always meets the following requirements:

[0029]

[0030] This inequality is the fundamental difference between nanosecond / sub-nanosecond pulse schemes and picosecond / femtosecond pulse schemes. For picosecond pulses ( ),That Only a few tens of meters away, and At comparable scales, the Raman process is suppressed by a severe walk-off effect.

[0031] 2. Quantitative Design of "Nonlinear Spectral Homogenization Length"

[0032] Based on satisfying the above-mentioned enhanced Raman conditions, the final length of the optical fiber... It needs to approach a "nonlinear spectral averaging length". This length is the scale at which Raman energy transfer, GPI seed generation, and multimode nonlinear coupling effects reach a dynamic equilibrium, and its design requires a trade-off between nonlinear effects and transmission loss.

[0033] One engineering design approach is to consider the total length as the sum of two characteristic lengths:

[0034]

[0035] (Nonlinear mode coupling length): Characterizes the length of energy transfer from higher-order modes to the fundamental mode in a multimode system through nonlinear processes (such as cross-phase modulation and four-wave mixing), initially forming a broadband seed. It is inversely proportional to the nonlinear coefficient and the peak power.

[0036] (Spectral flattening length): Characterizes the additional length required to achieve high flatness by uniformly redistributing energy across wavelengths through enhanced Raman scattering and cascade processes, ultimately filling spectral valleys, based on an initial broadband foundation. It is strongly dependent on Raman gain and the dispersion characteristics of the optical fiber.

[0037] Given nanosecond / subnanosecond pulse parameters (energy, pulse width) and a selected graded-index fiber, Relatively short (usually in the range of several meters to ten meters), while It is a decision The main part. Through numerical solution of the multimode generalized nonlinear Schrödinger equation (MM-GNLSE) or systematic experimental scanning, it was found that, at a pump power of hundreds of watts, the high flatness output described in this patent can be achieved. The optimal range is 30 meters to 150 meters. When the rice is dry, the planarization process is insufficient (as in Example 1); when While flatness may be better at a certain time, transmission loss will significantly reduce total output power and long-wavelength sideband strength (as shown in Example 3). A trade-off between flatness and efficiency must be made depending on the application.

[0038] Compared with existing technologies, the advantages of this invention lie in achieving high-flatness spectral output. Through the collaborative parameter design of "nanosecond / sub-nanosecond pulses - graded-index fiber of specific length," it creatively guides the synergistic evolution mechanism of enhanced Raman scattering and geometric parametric instability, effectively solving the problem of isolated and drastically fluctuating Raman peaks in the spectrum pumped by nanosecond / sub-nanosecond pulses. This enables the direct output of a supercontinuum spectrum with high flatness. Excellent beam spatial quality is maintained by fully utilizing the inherent nonlinear spatial self-cleaning effect in graded-index multimode fiber. This allows the output beam to spontaneously purify and stabilize in a near-Gaussian distribution during the complex process of achieving spectral broadening and flattening, ensuring high beam quality with a low M² factor. The system structure is simplified and highly reliable. This invention abandons traditional flattening schemes that rely on multi-wavelength pump combining or complex real-time feedback control. It achieves performance goals only by precisely matching the pump pulse temporal characteristics and fiber length—two key design degrees of freedom. The system has high integration through all-fiber technology, a simple structure, significantly improved stability and reliability, and reduced manufacturing costs. It provides clear and quantifiable design guidance, innovatively proposing the physical concept of "nonlinear spectral averaging length" and transforming it into a core design parameter. This provides a clear theoretical basis and convenient engineering approach for optimizing laser systems for different nanosecond / subnanosecond pulse parameters (such as energy and pulse width) to obtain optimal spectral flatness. Attached Figure Description

[0039] Figure 1 This is a schematic diagram illustrating the general principle and structure of the supercontinuum fiber laser system proposed in this invention.

[0040] Figure 2 This is a schematic diagram of the specific implementation structure of Embodiment 1 of the present invention.

[0041] The diagram is labeled as follows: 1. Laser seed source; 2. First semiconductor pumped laser; 3. First fiber isolator; 4. First combiner; 5. Small-sized active fiber; 6. Second fiber isolator; 7. Second semiconductor pumped laser; 8. Second combiner; 9. Large-sized active fiber; 10. 10m long high nonlinear multimode fiber; 11. Laser output head.

[0042] Figure 3 This is a schematic diagram of the specific implementation structure of Embodiment 2 of the present invention.

[0043] The diagram is labeled as follows: 1. Laser seed source; 2. First semiconductor pumped laser; 3. First fiber isolator; 4. First combiner; 5. Small-sized active fiber; 6. Second fiber isolator; 7. Second semiconductor pumped laser; 8. Second combiner; 9. Large-sized active fiber; 12. 30m long high nonlinear multimode fiber; 11. Laser output head.

[0044] Figure 4This is a schematic diagram of the specific implementation structure of Embodiment 4 of the present invention.

[0045] The diagram is labeled as follows: 1. Laser seed source; 2. First semiconductor pumped laser; 3. First fiber isolator; 4. First combiner; 5. Small-sized active fiber; 6. Second fiber isolator; 7. Second semiconductor pumped laser; 8. Second combiner; 9. Large-sized active fiber; 13. 100m long high nonlinear multimode fiber; 11. Laser output head.

[0046] Figure 5 This is a diagram showing the output light spot at wavelengths of 808nm, 1064nm, 1310nm, and 1550nm in Embodiment 3 of the present invention. Figure 5 (a) Output spot pattern at 808nm Figure 5 (b) Output spot pattern at 1064nm Figure 5 (c) Output spot pattern at 1310nm Figure 5 (d) Output spot pattern at 1550nm;

[0047] Figure 6 This is the output spectrum of Embodiment 1 of the present invention (10m long highly nonlinear multimode optical fiber);

[0048] Figure 7 This is the output spectrum of Embodiment 2 of the present invention (30m long highly nonlinear multimode optical fiber);

[0049] Figure 8 This is the output spectrum of Embodiment 3 of the present invention (100m long high nonlinear multimode optical fiber). Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0051] This invention provides a high-flatness spectral nanosecond / subnanosecond pulsed fiber laser based on nonlinear length matching. By actively matching the temporal characteristics (pulse width, repetition frequency) of the nanosecond / subnanosecond pulse with a section of graded-index multimode fiber of a specific length, a transmission system is constructed that can enhance Raman scattering and guide it to deeply synergize with nonlinear effects such as geometric parametric instability (GPI), thereby achieving efficient spectral flattening.

[0052] Please see Figure 1The laser comprises, in sequence along the optical path, a nanosecond / sub-nanosecond pulse seed source, a power amplification unit, and a nonlinear spectral flattening unit. Optionally, a mode matching unit can be provided between the power amplification unit and the nonlinear spectral flattening unit to optimize mode field matching and reduce coupling loss.

[0053] Nanosecond / sub-nanosecond pulse seed source: Used to generate nanosecond or sub-nanosecond pulsed lasers with a center wavelength of 1 micrometer (e.g., 1064 nm).

[0054] Power Amplification Unit: Employs a master oscillator power amplifier (MOPA) structure, including at least one amplification stage, to amplify the energy of the seed pulse to a level sufficient to excite significant nonlinear effects.

[0055] Nonlinear spectral flattening unit: Consists of a section of graded-index multimode fiber with a length ranging from 30 meters to 150 meters. This length range is configured to enhance the cascaded Raman scattering effect of nanosecond / subnanosecond pulses during transmission and to work in conjunction with geometric parametric instabilities, thereby outputting a broadband and highly flat supercontinuum laser.

[0056] Example 1

[0057] Please see Figure 2 This embodiment provides a nanosecond pulsed supercontinuum laser based on a 10-meter-long highly nonlinear multimode fiber, used to elucidate the physical phenomenon that the spectral flattening process is limited when the nonlinear interaction length is insufficient. Figure 2 As shown, the laser includes, in sequence along the laser transmission direction: a laser seed source 1, a two-stage master oscillator power amplifier (MOPA) structure, a 10-meter-long highly nonlinear multimode fiber 10, and a laser output head 11.

[0058] The laser seed source 1 is a nanosecond pulsed fiber laser with a center wavelength of 1064 nm. The time-domain width (full width at half maximum) of the output pulse of this seed source is approximately 4 nanoseconds, and the repetition frequency is fixed at 63 kHz. This parameter provides specific time-domain initial conditions for the evolution of subsequent nonlinear effects.

[0059] To boost the energy of the seed pulse to a level sufficient to excite a significant nonlinear process, the system employs a two-stage MOPA amplification structure. The primary amplification unit consists of a first semiconductor pump laser 2, a first fiber isolator 3, a first beam combiner 4, and a section of small-sized active fiber 5. Specifically, the weak nanosecond pulse from the seed source 1 first passes through the first fiber isolator 3 to ensure unidirectional light transmission and prevent back-reflected light from damaging the source; subsequently, the pulse is introduced into the signal input of the first beam combiner 4. Simultaneously, the pump light output from the first semiconductor pump laser 2 is injected into the pump input of the first beam combiner 4. After the pulsed light and pump light are combined within the first beam combiner 4, they are coupled together into the small-sized active fiber 5 (e.g., a ytterbium-doped fiber with a core diameter of approximately 10 micrometers) for initial energy amplification. The small fiber core helps maintain good beam quality during the initial amplification stage.

[0060] The optical signal, after primary amplification, enters the final amplification unit. This unit consists of a second fiber isolator 6, a second semiconductor pump laser 7, a second combiner 8, and a section of large-size active fiber 9. After passing through the second fiber isolator 6, the optical signal enters the signal input end of the second combiner 8. The higher-power pump light output from the second semiconductor pump laser 7 is injected into the second combiner 8 and combined with the signal light again. The combined light is coupled into the large-size active fiber 9 (e.g., a double-clad ytterbium-doped fiber with a core diameter of 30 micrometers or larger) for power amplification. The large core size effectively reduces the power density within the fiber, increases the system's damage threshold, and allows for the injection of higher pump power, ultimately outputting a high-power 1064 nm nanosecond pulse with an average power in the range of several watts.

[0061] The amplified high-power nanosecond pulses need to be efficiently injected into the nonlinear medium. In this embodiment, the output end of the large-size active optical fiber 9 is connected to the input end of the 10-meter-long highly nonlinear multimode optical fiber 10 using a standard arc fusion splicing process. To ensure low loss and long-term reliability at the splice point, the fiber end face can be precisely cut and cleaned before splicing. The 10-meter-long highly nonlinear multimode optical fiber 10 is one of the core components of this invention. It is a graded-index multimode optical fiber with specific parameters such as a core diameter of 50 micrometers and a numerical aperture of 0.2. This type of fiber has a parabolic refractive index distribution, which not only supports multimode transmission to carry high power, but its special waveguide structure also provides a basis for the self-cleaning effect in nonlinear space.

[0062] The laser output head 11 is connected to the output end of a 10-meter-long highly nonlinear multimode fiber 10. The laser output head 11 typically includes a precision focusing lens group or a simple end cap structure. Its function is to safely export the generated supercontinuum laser from the system with a low divergence angle and high beam quality, and to facilitate coupling with subsequent application equipment.

[0063] The working process and physical phenomena revealed in this embodiment are as follows: A high-power 1064nm nanosecond pulse, amplified by the MOPA structure, is injected into a graded-index multimode fiber 10 that is only 10 meters long. At this relatively short length, the interaction time and distance between the optical pulse and the fiber medium are limited. The experimentally observed evolution of the output spectrum is as follows: Figure 5 As shown, at lower pump power, the spectrum broadens symmetrically around the pump wavelength through effects such as self-phase modulation. With increasing injection power, the spectrum gradually extends towards longer wavelengths, as nonlinear processes such as stimulated Raman scattering begin to take effect. However, the overall broadening range of the spectrum is limited across the entire power variation range, and obvious modulation and undulation structures are visible in the spectral profile, failing to form a smooth, continuous spectral band. This experimental result clearly demonstrates that a transmission length of 10 meters is insufficient to allow various nonlinear effects, such as geometric parametric instability (GPI), cascaded Raman scattering, and four-wave mixing, to fully develop and co-evolve. The pulse has not yet undergone a complete "spectral homogenization" process in the fiber, and the transfer and redistribution of energy between different frequency bands are insufficient, resulting in poor flatness of the final output spectrum. This example serves as a contrast, demonstrating from the opposite perspective that there is a critical fiber length; shorter lengths make it difficult to achieve high-flatness spectral output, providing a benchmark for optimizing the length in subsequent examples.

[0064] Example 2

[0065] Please see Figure 3 This embodiment provides a nanosecond pulsed supercontinuum laser based on a 30-meter-long highly nonlinear multimode fiber, aiming to demonstrate that spectral flatness is significantly improved when the fiber length increases to a critical range, thereby verifying the optimization range of the "nonlinear spectral averaging length". Figure 3 As shown, the overall architecture of this laser is highly consistent with that of Embodiment 1, and its optical path also includes, in sequence: laser seed source 1, second-stage master oscillator power amplifier (MOPA) structure, and laser output head 11. The core change in this embodiment is that the nonlinear spectral flattening unit is replaced with a 30-meter-long highly nonlinear multimode fiber 12 to study the regulatory effect of the length parameter on the spectral evolution dynamics.

[0066] The laser seed source 1, the first semiconductor pump laser 2, the first fiber isolator 3, the first beam combiner 4, the small-sized active fiber 5, the second fiber isolator 6, the second semiconductor pump laser 7, the second beam combiner 8, the large-sized active fiber 9, and the laser output head 11 are all the same in model 1 in terms of type, function, and connection relationship. This means that the pumping conditions injected into the nonlinear fiber—i.e., a wavelength of 1064 nm, a pulse width of approximately 4 ns, a specific repetition frequency, and a nanosecond pulse with an average power of several watts after amplification—are basically consistent in the two experiments, thus ensuring that the only variable in the comparative experiment is the length of the nonlinear fiber.

[0067] The input end of the 30-meter-long highly nonlinear multimode fiber 12 is connected to the output end of the large-size active fiber 9 via fusion splicing to achieve an all-fiber connection. This fiber 12 is the same type of product as the fiber 10 in Example 1, meaning it has the same graded refractive index distribution, core diameter (50 micrometers), and numerical aperture (0.2) waveguide parameters. The only difference is that its physical length is precisely controlled to 30 meters. This length selection is based on a potential "spectral homogenization" range suggested by previous theoretical research and numerical simulations for current nanosecond pulse parameters.

[0068] The working process and observed physical phenomena of this embodiment are as follows: A high-power nanosecond pulse, under the same conditions as in Embodiment 1, is injected into this 30-meter-long graded-index multimode fiber 12. The longer transmission distance provides more time and space for the nonlinear interaction between the pulse and the fiber medium. Figure 6 The experimental spectral results shown reveal an evolutionary behavior that is distinctly different from that of Example 1. As the pump power increases from a lower level, the output spectrum not only extends more rapidly toward longer wavelengths (above 1800 nm), but also extends more effectively toward shorter wavelengths (below 900 nm), exhibiting a more balanced bidirectional broadening characteristic.

[0069] Crucially, at medium power levels (e.g., approximately 3W), the morphology of the output spectrum undergoes a fundamental transformation. The sharp fluctuations and valley structures that might have previously existed in the spectral profile due to discrete sidebands or Raman peaks essentially disappear, replaced by a continuous, broad, and smoothly contoured band. This "smoothing" of the spectral morphology indicates that, during the 30-meter transmission process, a sufficient and efficient co-evolution has occurred between the initial multi-band discrete seeds generated by geometric parametric instability (GPI) and various nonlinear mechanisms such as self-phase modulation, cross-phase modulation, and enhanced cascaded Raman scattering. Energy is able to undergo complex and continuous transfer and redistribution among different spectral components, effectively filling spectral gaps and smoothing out intensity fluctuations.

[0070] This experimental result has clear indicative significance: for the specific nanosecond pulse pump source of this invention, a fiber length of 30 meters approaches an efficient range for achieving global spectral flattening, i.e., close to the so-called "nonlinear spectral averaging length." At this length, the nonlinear process is fully developed while avoiding excessive transmission loss that may be introduced by excessively long fibers (as will be shown in subsequent embodiments). Therefore, Embodiment 2 not only demonstrates a specific implementation method that can output a high-flatness supercontinuum, but more importantly, it experimentally confirms that by precisely controlling the single variable of fiber length, the nonlinear dynamic path can be actively guided, thereby effectively optimizing the core quality of the output spectrum—flatness. This constitutes a key innovation of this invention.

[0071] Example 3

[0072] Please see Figure 4 This embodiment provides a nanosecond pulse supercontinuum laser based on a 100-meter-long highly nonlinear multimode fiber, used to explore the unique nonlinear dynamic path of nanosecond pulses at longer transmission scales and their crucial role in achieving optimal spectral flatness. Figure 4 As shown in the schematic diagram, the core architecture of this laser is consistent with that of Embodiments 1 and 2, including the same laser seed source 1, two-stage MOPA amplification structure (composed of components 2 to 9), and laser output head 11. The fundamental difference in this embodiment is that the nonlinear spectral flattening unit is replaced with a 100-meter-long highly nonlinear multimode fiber 13, thereby constructing a synergistic system of "nanosecond pulse - 100-meter fiber".

[0073] The input end of the 100-meter-long highly nonlinear multimode fiber 13 is fused to the output end of the large-size active fiber 9. Its fiber type, core diameter (50 micrometers), and numerical aperture (0.2) are exactly the same as those of fibers 10 and 12 in the aforementioned embodiments, ensuring consistent waveguide characteristics. The only variable is that the length is significantly increased to 100 meters. This length selection aims to take advantage of the fact that the first-order Stokes walk-off length (approximately 1000 meters) corresponding to a nanosecond pulse (pulse width ~4 ns) is much larger than the actual length of the fiber.

[0074] The physical processes and output characteristics revealed in this embodiment are of landmark significance. When the same high-power 1064nm nanosecond pulse is injected into this 100-meter-long optical fiber, its nonlinear evolution exhibits characteristics drastically different from those of shorter optical fibers. For example... Figure 7The output spectrum clearly shows that at relatively lower pump power, the spectrum begins to broaden significantly, accompanied by the formation and redshift of a very prominent Raman-Stokes peak. This phenomenon is the primary characteristic of the nonlinear dynamics in this embodiment: due to the extremely weak walk-off effect of nanosecond pulses in a hundred-meter fiber, stimulated Raman scattering obtains unprecedented sustained gain and energy accumulation conditions, thus being greatly enhanced.

[0075] This enhanced Raman process is the core mechanism for achieving ultra-high flatness output. It continuously and efficiently transfers energy from the pump wavelength (1064 nm) to longer wavelengths, causing a large amount of energy from the original pump peak to be consumed, becoming extremely weak or even almost disappearing in the final output spectrum. This fundamentally avoids the destruction of the overall spectral flatness by the strong pump peak in traditional schemes. More importantly, this powerful Raman energy transfer process has a profound synergy with the naturally excited geometric parametric instabilities (GPIs) in the optical fiber. The broadband seed generated by GPIs at the transmission front end, especially the mid-infrared components, is effectively "relayed" by the enhanced Raman effect and pushed to longer wavelengths; at the same time, the Raman process combined with cascaded four-wave mixing establishes dense energy exchange channels throughout the broadened spectrum (especially in spectral valley regions that are difficult to fill in traditional schemes).

[0076] Ultimately, these synergistic nonlinear effects drive a thorough and uniform redistribution of energy across a wide spectral range of 820 nm to 2280 nm. Experimental results show that, at an output power of 1.5 W, an ultraflat continuous spectrum with a 10-dB bandwidth exceeding 1400 nm and an extremely smooth spectral profile can be obtained, with overall fluctuations suppressed to an extremely low level. Figure 7 This signifies that, through the specific parameter combination of "nanosecond pulse-hundred-meter fiber," Raman scattering, often regarded as a source of harmful spectral fluctuations, has been successfully transformed into a core constructive force driving global spectral flattening.

[0077] This embodiment also systematically evaluates the spatial quality of the output beam during such extreme spectral broadening and shaping. For example... Figure 5 As shown in (a) to (d), the near-field beams collected at the four characteristic wavelengths of 808 nm, 1064 nm, 1310 nm, and 1550 nm all exhibit a stable and regular near-Gaussian distribution. This result conclusively proves that even after undergoing strong and complex nonlinear interactions in a multimode fiber of hundreds of meters in length, the inherent beam self-cleaning effect of graded-index fiber remains effective. This effect ensures that energy continuously concentrates on lower-order modes throughout the entire supercontinuum generation process, ultimately maintaining a high spatial quality close to the diffraction limit across the entire spectrum, achieving synergistic optimization of spectral flatness and beam spatial quality.

[0078] In summary, the three embodiments demonstrate that 10 meters illustrates the incomplete spectral homogenization when the length is insufficient; 2 meters reveals that high flatness output can be obtained when the length approaches the optimized "nonlinear spectral homogenization"; and 3 meters, through the matching of extreme length with nanosecond pulse characteristics, explores and utilizes the unique dynamic path of enhanced Raman scattering and GPI synergy, achieving not only better spectral flatness but also maintaining excellent beam quality. These three embodiments together constitute a complete chain of experimental evidence, systematically verifying that by actively designing and matching the "nanosecond pulse time-domain characteristics" and the "gradient refractive index multimode fiber length," nonlinear dynamics can be guided towards the desired flattening direction. This is precisely the core innovation and technological advancement of this invention.

[0079] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A high-flatness spectral nanosecond / subnanosecond pulsed fiber laser based on nonlinear length matching, characterized in that, Including those connected sequentially along the optical path: Nanosecond / sub-nanosecond pulse seed source, used to generate nanosecond / sub-nanosecond pulsed lasers with a center wavelength of 1 micrometer; A power amplification unit, the input of which is connected to the output optical path of the nanosecond / sub-nanosecond pulse seed source, is used to amplify the power of the nanosecond / sub-nanosecond pulse laser. as well as The nonlinear spectral flattening unit, whose input end is optically connected to the output end of the power amplification unit, is composed of a section of graded refractive index multimode fiber. The length of the graded-index multimode fiber should ensure that when the nanosecond / sub-nanosecond pulsed laser is transmitted in the graded-index multimode fiber, its walk-off length is much greater than the physical length of the graded-index multimode fiber. This enhances the cascaded Raman scattering effect and works in conjunction with the geometric parametric instability to output a broadband and highly flat supercontinuum laser.

2. The high flatness spectral nanosecond / subnanosecond pulsed fiber laser according to claim 1, characterized in that, The length of the graded-index multimode fiber is 30 meters to 150 meters.

3. The high-flatness spectral nanosecond / subnanosecond pulsed fiber laser according to claim 1, characterized in that, The graded-index multimode fiber has a core diameter of ≥20 micrometers, a numerical aperture of 0.1 to 0.3, and a parabolic or approximately parabolic graded-index distribution.

4. The high flatness spectral nanosecond / subnanosecond pulsed fiber laser according to claim 1, characterized in that, The nanosecond pulse seed source generates pulses with a width of 0.1 nanoseconds to 100 nanoseconds and a repetition frequency of 1 kHz to 1 MHz.

5. The high flatness spectral nanosecond / subnanosecond pulsed fiber laser according to claim 1, characterized in that, The power amplification unit is a main oscillation power amplification structure, including at least one amplification stage. Each amplification stage includes a pump source, a combiner, and an active optical fiber.

6. The high-flatness spectral nanosecond / subnanosecond pulsed fiber laser according to claim 1, characterized in that, The output end of the power amplifier unit and the input end of the graded-index multimode fiber are connected by fusion splicing to achieve a fully fiber optic connection.

7. The high-flatness spectral nanosecond / subnanosecond pulsed fiber laser according to claim 1, characterized in that, It also includes a mode matching unit, which is disposed between the output end of the power amplification unit and the input end of the graded refractive index multimode fiber; the mode matching unit is a transition fiber or mode field adapter, used to achieve mode field diameter matching, and to couple the amplified nanosecond / subnanosecond pulsed laser to the graded refractive index multimode fiber in a fundamental mode-dominated manner.

Images