A high-energy femtosecond fiber laser

Abstract

The present application relates to the technical field of fiber lasers and discloses a high-energy femtosecond fiber laser. The invention comprises: a signal light input module for inputting 1030nm signal light; a polarization splitting and energy adjustment module, comprising a half-wave plate and a polarization beam splitter, wherein the half-wave plate is used to adjust the polarization state of the input signal light, and the polarization beam splitter splits the adjusted signal light into two orthogonally polarized pulses; a Sagnac cavity amplification module, wherein two pulses are coupled into the Sagnac cavity through two collimators, respectively, and then transmitted in opposite directions within the ring and amplified by an ytterbium-doped gain fiber; a pump coupling module, comprising a passive fiber collimator and a dichroic lens for outputting 976nm pump light; the pump light is coupled into the gain fiber collimator of the Sagnac cavity through the dichroic lens to provide energy for amplification within the ring; and a pulse synthesis output module, wherein the two amplified pulses are synthesized into a single pulse at the polarization beam splitter by utilizing the optical path symmetry of the Sagnac cavity and output.

Description

Technical Field

[0001] The present application relates to the technical field of fiber lasers, and in particular to a high-energy femtosecond fiber laser. Background Art

[0002] Femtosecond fiber lasers have important applications in precision machining, spectroscopy, and other fields due to their ultrashort pulse characteristics. 1030nm lasers have become a research hotspot due to their atmospheric transmission characteristics and compatibility with material processing. Traditional high-energy femtosecond fiber lasers generally use chirped pulse amplification (CPA) schemes, and their output energy is limited by two major factors:

[0003] Pulse stretching limitation: CPA must first stretch the seed pulse to reduce peak power and avoid nonlinear effects. However, the stretching amount is limited by the size and cost of the compression optical path and cannot be expanded indefinitely.

[0004] Fiber mode field diameter bottleneck: The effective mode field diameter of amplifying optical fibers must be large enough to carry high energy. However, an excessively large mode field can easily induce multimode transmission, degrading beam quality. Existing technologies achieve higher mode field diameters through photonic crystal fiber design (e.g., 40μm for flexible fibers and 80μm for rod-type fibers), but these technologies face challenges such as high cost and difficulty in fiber coupling and matching due to their small numerical aperture (NA).

[0005] Although existing solutions attempt to increase energy by increasing the mode field diameter or optimizing the stretching system, they are still inherently limited by the physical limitations of single-channel amplification. In single-channel amplification, pulse energy is strongly correlated with the mode field diameter and stretching, and further improvements in mode field diameter are limited by beam quality and engineering implementation costs. Therefore, how to break through the energy bottleneck of single-channel amplification while maintaining the same stretching has become a key challenge in the design of high-energy femtosecond fiber lasers.

[0006] Therefore, there is an urgent need for a laser to solve at least one of the above problems. Summary of the Invention

[0007] This application provides a high-energy femtosecond fiber laser designed to address existing solutions that attempt to increase energy by increasing the mode field diameter or optimizing the broadening system. However, these solutions are still inherently limited by the physical limitations of single-channel amplification. In single-channel amplification, pulse energy is strongly correlated with the mode field diameter and the amount of broadening, while further improvements in the mode field diameter are limited by beam quality and engineering implementation costs. Therefore, how to break through the energy bottleneck of single-channel amplification at the same broadening amount has become a key problem in the design of high-energy femtosecond fiber lasers.

[0008] In a first aspect, an embodiment of the present application provides a laser wavelength conversion method, comprising:

[0009] A signal light input module is used to input 1030nm signal light, and the signal light input module outputs the signal light to the half-wave plate and polarization beam splitter through a passive collimator;

[0010] A polarization beam splitting and energy regulation module, comprising the half-wave plate and a polarization beam splitter, wherein the half-wave plate is used to adjust the polarization state of the input signal light, and the polarization beam splitter splits the adjusted signal light into two orthogonally polarized pulses;

[0011] The Sagnac cavity amplification module includes two symmetrically arranged collimators and an ytterbium-doped gain fiber within the ring. Two pulses are coupled into the Sagnac cavity through the two collimators, propagated in opposite directions within the ring, and amplified by the ytterbium-doped gain fiber. The ytterbium-doped gain fiber is a non-polarization-maintaining fiber.

[0012] A pump coupling module, comprising a passive fiber collimator and a dichroic lens for outputting 976nm pump light. The pump light is coupled into the gain fiber collimator of the Sagnac cavity through the dichroic lens to provide energy for intra-ring amplification.

[0013] A pulse synthesis output module, wherein the two amplified pulses are synthesized into a single pulse at the polarization beam splitter and outputted by utilizing the optical path symmetry of the Sagnac cavity;

[0014] The laser divides a single pulse into two paths by combining chirped pulse amplification with split pulse amplification, which are amplified independently in a Sagnac cavity and then synthesized.

[0015] In some embodiments, the two pulses in the symmetrical optical path of the Sagnac cavity have the same optical path length, and no additional pulse synchronization adjustment is required. The energy distribution of the two pulses is controlled by adjusting the half-wave plate to optimize the synthesis efficiency, thereby achieving an increase in the pulse output energy under the same broadening amount.

[0016] In some embodiments, the optical path adopts a device collimator coupling method, and the gain fiber of the Sagnac cavity is connected through an active fiber collimator, wherein the angle of the active fiber collimator corresponding to the axis facing the dichroic lens is 90°, and the angle of the active fiber collimator corresponding to the axis facing the half-wave plate is 0°.

[0017] In some embodiments, the signal light input module outputs signal light to a half-wave plate and a polarization beam splitter through a passive collimator, including: the passive collimator is used to collimate the input 1030nm signal light into a parallel light beam to match the input light path of the subsequent half-wave plate and the polarization beam splitter, the output end of the collimator is optically connected to the incident end of the half-wave plate, and the output end of the half-wave plate is optically connected to the first input end of the polarization beam splitter.

[0018] Exemplarily, the angles of the corresponding axes of the active fiber collimator are set to: 90° for the corresponding axis of the collimator facing the dichroic lens, for matching the polarization direction of the pump light to improve the coupling efficiency; 0° for the corresponding axis of the collimator facing the half-wave plate, for maintaining the polarization state stability of the signal light. The independent transmission characteristics of the signal light and the pump light are optimized by differentiating the angles of the corresponding axes.

[0019] In some embodiments, the polarization beam splitter splits the adjusted signal light into two orthogonally polarized pulses, including: the polarization beam splitter splits the linearly polarized light adjusted by the half-wave plate into two orthogonally polarized pulses transmitted along the horizontal and vertical directions based on the polarization state of the input signal light, and the energy distribution of the two pulses is dynamically balanced by adjusting the angle of the half-wave plate.

[0020] In some embodiments, the two orthogonal polarized pulses are respectively coupled into the two collimators of the Sagnac cavity amplification module, where one pulse passes through the right collimator to enter the clockwise optical path of the Sagnac ring, and the other pulse passes through the lower collimator to enter the counterclockwise optical path of the Sagnac ring, forming an amplification path transmitted in opposite directions within the ring.

[0021] In some embodiments, the laser splits a single pulse into two paths by combining chirped pulse amplification with split pulse amplification, and then amplifies them independently in a Sagnac cavity and then synthesizes them, including: based on the stretched pulse of chirped pulse amplification, splitting the stretched single pulse into two paths by split pulse amplification, performing power amplification in the symmetrical optical path of the Sagnac cavity respectively, utilizing the optical path symmetry of the Sagnac cavity to ensure the phase consistency of the two amplified pulses, and finally achieving coherent synthesis at the polarization beam splitter, and the synthesized pulse energy is the energy superposition of the two amplified pulses.

[0022] In some embodiments, the ytterbium-doped gain fiber in the Sagnac cavity amplification module is a non-polarization-maintaining fiber, and the core diameter and numerical aperture parameters of the non-polarization-maintaining fiber suppress polarization mode dispersion through the symmetrical optical path design of the Sagnac cavity.

[0023] In some embodiments, in the pump coupling module, after the 976nm pump light is output through the passive fiber collimator, it is reflected by a dichroic lens and focused onto the gain fiber collimator of the Sagnac cavity. The dichroic lens has high transmittance for the 1030nm signal light and high reflection for the 976nm pump light.

[0024] The embodiments of the present application provide a high-energy femtosecond fiber laser, which inputs 1030nm femtosecond signal light through a signal light input module, collimates the signal light into a parallel beam through a passive collimator, and outputs it to a half-wave plate and a polarization beam splitter (PBS) to achieve initial optical path matching of the signal light.

[0025] The polarization beam splitting and energy regulation module includes a half-wave plate and a PBS. The half-wave plate adjusts the polarization state of the signal light, and the PBS splits it into two pulses with orthogonal polarizations (such as horizontal / vertical). The energy distribution of the two pulses is adjusted by the angle of the half-wave plate.

[0026] The Sagnac cavity amplifier module uses two symmetrical collimators to construct a Sagnac ring. Two orthogonally polarized pulses are coupled into the ring from the collimators on either side, propagating in opposite directions (clockwise and counterclockwise) within the ring. They are amplified by the ytterbium-doped non-polarization-maintaining gain fiber within the ring. The optical path symmetry of the Sagnac cavity ensures that the optical path lengths of the two pulses are identical.

[0027] The pump coupling module outputs 976nm pump light through a passive fiber collimator. After reflection by a dichroic lens (highly transparent to 1030nm signal light and highly reflective to 976nm pump light), it is coupled into the gain fiber collimator of the Sagnac cavity, providing pump energy for the gain fiber in the ring to achieve signal light amplification.

[0028] The pulse synthesis output module directly synthesizes the two amplified pulses into a single pulse output due to the symmetry of the Sagnac cavity optical path and the phase consistency at the PBS. The synthesized energy is the energy superposition of the two amplified pulses.

[0029] By combining divided pulse amplification (DPA) with chirped pulse amplification (CPA), a single pulse is split into two orthogonally polarized pulses using PBS, based on CPA pulse broadening. These pulses are then coherently combined using a Sagnac cavity for independent bidirectional amplification, thus overcoming the energy limitations of single-path amplification. Non-polarization-maintaining gain fiber and symmetrical optical path design utilize non-polarization-maintaining fiber as the gain medium. Polarization mode dispersion is suppressed through the symmetrical optical path of the Sagnac cavity, avoiding the coupling difficulties and cost issues associated with the high-mode field of traditional polarization-maintaining fiber while ensuring beam quality. Collimator-coupled integration utilizes device-collimator coupling throughout the entire optical path, simplifying system integration and adapting to conventional engineering implementation methods.

[0030] The provided laser has the following beneficial effects:

[0031] Breaking through the energy bottleneck of single-path amplification: Traditional CPA solutions rely on single-path amplification, and the energy is limited by the pulse broadening and mode field diameter. The present invention uses DPA to split a single pulse into two paths for parallel amplification. The energy after synthesis is significantly improved, achieving an energy breakthrough at the same broadening amount without relying on extreme mode field diameter design (such as avoiding the use of expensive 80μm rod-shaped photonic crystal fiber).

[0032] Simplified adjustment complexity and system cost: The natural optical path symmetry of the Sagnac cavity ensures strict alignment of the optical path lengths of the two pulses, eliminating the need for additional pulse synchronization adjustments (such as the precise control of optical path length differences required in traditional beam splitting amplification). Efficient synthesis is achieved simply by adjusting the energy distribution using a half-wave plate, reducing the difficulty of optical path debugging. The use of non-polarization-maintaining gain fiber significantly reduces costs compared to traditional polarization-maintaining fiber (especially large-mode-area photonic crystal fiber), avoids the coupling and matching challenges associated with small numerical apertures (NA), and improves engineering practicality.

[0033] Improved beam quality and amplification efficiency: The symmetrical optical path design suppresses polarization mode dispersion and multimode effects, ensuring high beam quality of the amplified pulses even when using relatively conventional gain fibers (such as non-polarization-maintaining fibers). The two independent amplification paths reduce nonlinear effects (such as self-phase modulation) in a single fiber, improving amplification efficiency and pulse quality.

[0034] Advantages of modular integration: The full optical path adopts conventional technical means of collimator coupling, avoiding special devices (such as customized polarization-maintaining couplers), lowering the threshold for system construction, adapting to standardized production, and improving the feasibility of engineering applications.

[0035] In summary, this application achieves high-energy femtosecond pulse output through the structural design of Sagnac cavity symmetrical beam amplification and coherent combination, combined with CPA+DPA technology, without relying on the extreme mode field diameter and complex synchronous adjustment, effectively solving the energy bottleneck, high cost and complex adjustment problems of traditional single-channel amplification, and has significant technological progress and engineering application value.

[0036] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0038] Figure 1 This is a schematic block diagram of the structure of a high-energy femtosecond fiber laser provided in one embodiment of the present application;

[0039] Figure 2 This is a schematic structural diagram of a high-energy femtosecond fiber laser provided in one embodiment of the present application.

[0040] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0042] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined, or partially merged, so the actual execution order may vary depending on the actual situation.

[0043] It should be understood that, in order to clearly describe the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, terms such as "first" and "second" are used to distinguish between identical or similar items having substantially the same functions and effects. Those skilled in the art will understand that terms such as "first" and "second" do not limit the quantity or order of execution, and that terms such as "first" and "second" do not necessarily define differences.

[0044] It should be understood that the terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0045] It will also be understood that the term "and / or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0046] The following describes some embodiments of the present application in detail with reference to the accompanying drawings. In the absence of conflict, the following embodiments and features therein may be combined with each other.

[0047] Femtosecond fiber lasers have important applications in precision machining, spectroscopy, and other fields due to their ultrashort pulse characteristics. 1030nm lasers have become a research hotspot due to their atmospheric transmission characteristics and compatibility with material processing. Traditional high-energy femtosecond fiber lasers generally use chirped pulse amplification (CPA) schemes, and their output energy is limited by two major factors:

[0048] Pulse stretching limitation: CPA must first stretch the seed pulse to reduce peak power and avoid nonlinear effects. However, the stretching amount is limited by the size and cost of the compression optical path and cannot be expanded indefinitely.

[0049] Fiber mode field diameter bottleneck: The effective mode field diameter of amplifying optical fibers must be large enough to carry high energy. However, an excessively large mode field can easily induce multimode transmission, degrading beam quality. Existing technologies achieve higher mode field diameters through photonic crystal fiber design (e.g., 40μm for flexible fibers and 80μm for rod-type fibers), but these technologies face challenges such as high cost and difficulty in fiber coupling and matching due to their small numerical aperture (NA).

[0050] Although existing solutions attempt to increase energy by increasing the mode field diameter or optimizing the stretching system, they are still inherently limited by the physical limitations of single-channel amplification. In single-channel amplification, pulse energy is strongly correlated with the mode field diameter and stretching, and further improvements in mode field diameter are limited by beam quality and engineering implementation costs. Therefore, how to break through the energy bottleneck of single-channel amplification while maintaining the same stretching has become a key challenge in the design of high-energy femtosecond fiber lasers.

[0051] Therefore, there is an urgent need for a laser to solve at least one of the above problems.

[0052] To resolve the above, please refer to Figures 1 to 2 The embodiment of the present application provides a high-energy femtosecond fiber laser, including: a signal light input module for inputting 1030nm signal light, the signal light input module outputs the signal light to a half-wave plate and a polarization beam splitter through a passive collimator; a polarization beam splitting and energy adjustment module, including the half-wave plate and the polarization beam splitter, the half-wave plate is used to adjust the polarization state of the input signal light, and the polarization beam splitter splits the adjusted signal light into two orthogonally polarized pulses; a Sagnac cavity amplification module, including two symmetrically arranged collimators and an ytterbium-doped gain fiber in the ring, the two pulses are coupled into the Sagnac cavity through the two collimators respectively, and are transmitted in opposite directions in the ring. The laser is amplified by transmitting the light through the ytterbium-doped gain fiber, which is a non-polarization-maintaining fiber. A pump coupling module includes a passive fiber collimator and a dichroic lens for outputting 976nm pump light. The pump light is coupled into the gain fiber collimator of the Sagnac cavity through the dichroic lens to provide energy for intra-ring amplification. A pulse synthesis output module is configured in which the two amplified pulses are synthesized into the same pulse at the polarization beam splitter and output by utilizing the optical path symmetry of the Sagnac cavity. The laser divides a single pulse into two paths by combining chirped pulse amplification with split pulse amplification, and the two pulses are independently amplified in the Sagnac cavity and then synthesized.

[0053] Specifically, this invention transforms the single-pulse energy bottleneck from a "single-path physical limit" to "multi-path synthesis capability" through the innovative architecture of "chirped pulse amplification (CPA) + divided pulse amplification (DPA) + Sagnac cavity symmetric optical path." The pulse splitting mechanism utilizes a polarization beam splitter (PBS) to split the 1030nm linearly polarized signal light into two independent pulses with orthogonal polarizations (horizontally polarized TE mode and vertically polarized TM mode). The energy of the two pulses is adjusted using a half-wave plate to adjust the polarization state ratio (e.g., 50:50 or dynamically allocated), forming the physical basis of DPA. Unlike traditional single-path CPA, which relies solely on stretching to increase energy, this invention divides the single-pulse energy requirement into two parallel amplification paths by "dividing one path into two paths." Each path only needs to carry 1 / 2 of the original energy, thus breaking through the limitations of single-path mode field diameter and nonlinear effects.

[0054] Symmetrical Sagnac cavity amplification: A ring-shaped Sagnac cavity is constructed. Two orthogonal pulses are coupled into the cavity from symmetrical positions (e.g., the left and right sides) through a collimator. They propagate within the ring in opposite clockwise and counterclockwise directions before being amplified by an ytterbium-doped non-polarization-maintaining gain fiber. The strict optical path symmetry of the Sagnac cavity (zero optical path difference between the two pulses) ensures that the two pulses are perfectly aligned after amplification, eliminating the need for additional phase compensation or optical path adjustment, facilitating subsequent coherent combining.

[0055] Pump coupling and gain mechanism: 976nm pump light is output through a passive collimator, reflected by a dichroic lens (highly transparent at 1030nm and highly reflective at 976nm), and then coupled with the signal light into the Sagnac cavity. The pump light energy is absorbed by the non-polarization-maintaining gain fiber within the ring, achieving bidirectional amplification energy supply.

[0056] The signal light input module corresponds to the collimation and initial polarization adjustment components, including a 1030nm seed light source (such as a mode-locked fiber laser), a passive fiber collimator, a half-wave plate (HWP), and a polarization beam splitter (PBS).

[0057] Implementation steps: The femtosecond pulses (central wavelength 1030nm, pulse width tens of femtoseconds) output by the seed light source are first converted into parallel beams (divergence angle <0.5mrad) by a passive collimator to ensure optical path alignment with subsequent optical components (half-wave plates, PBSs). The collimated signal light is then incident on the half-wave plate. By rotating the half-wave plate angle (adjustable from 0° to 90°), the linear polarization of the signal light is adjusted (for example, from horizontal polarization to 45° polarization), and the energy distribution of the PBS between the two pulses is altered (for example, 100% for the TE mode at 0° and 50% for the TE and TM modes at 45°).

[0058] The polarization beam splitter (PBS) with an extinction ratio of >30dB corresponding to the polarization beam splitting and energy regulation module has its incident end optically connected to the output end of the half-wave plate, and its two output ends correspond to the TE mode (horizontal) and TM mode (vertical) channels respectively.

[0059] Key parameters: The PBS has a transmission / reflection polarization extinction ratio of ≥25dB ​​for 1030nm signal light, ensuring polarization orthogonality between the two pulses and preventing crosstalk during amplification. The half-wave plate uses an air-gap design (thickness <2mm) to reduce dispersion and ensure phase stability of the femtosecond pulses.

[0060] The cavity structure of the Sagnac cavity amplifier module consists of two symmetrically spaced active fiber collimators (labeled as Collimator A and Collimator B), a ring optical path (circumference 50-100 cm), and an ytterbium-doped non-polarization-maintaining gain fiber (core diameter 20 μm, numerical aperture NA = 0.08, length 2-5 m) within the ring. Collimators A and B are located on the left and right sides of the Sagnac ring, with a spacing equal to the ring diameter, ensuring symmetry between the two pulse input paths.

[0061] Amplification process: The TE mode pulse output by the PBS is coupled into the Sagnac ring through collimator A, transmitted in a clockwise direction, and amplified by absorbing pump energy when passing through the gain fiber; the TM mode pulse is coupled into the ring through collimator B, transmitted in a counterclockwise direction, and also amplified by the gain fiber; Use of non-polarization-maintaining gain fiber: Different from traditional polarization-maintaining fiber, its mode field diameter (20μm) is smaller than that of polarization-maintaining photonic crystal fiber (such as 40μm), but through the symmetrical optical path design of the Sagnac cavity, polarization mode dispersion (PMD) is naturally suppressed, avoiding the coupling difficulties caused by the high mode field of polarization-maintaining fiber (for example, polarization-maintaining fiber with NA=0.05 requires precise alignment and has a coupling efficiency of less than 80%, while the coupling efficiency of the non-polarization-maintaining fiber with NA=0.08 of the present invention is greater than 90%).

[0062] Component design for the pump coupling module: A 976nm pump light source (power 50-200W, multimode output) is collimated by a passive fiber collimator (alignment on the corresponding axis, divergence angle on the corresponding axis <5°) before being incident on a dichroic lens (coating parameters: 1030nm transmittance >95%, 976nm reflectivity >99%). The pump light reflected by the dichroic lens is focused by collimator A (or B, depending on the optical path design) into the gain fiber of the Sagnac ring, where it is coaxially transmitted with the signal light, achieving bidirectional pumping (or unidirectional pumping, depending on the collimator angle setting).

[0063] Key optimization: The angle of the active collimator's corresponding axis toward the dichroic lens is set to 90° to match the direction of the corresponding axis of the pump light. (Multimode pump light is approximately randomly polarized, but optimizing the collimator angle improves the mode field matching efficiency between the pump light and the gain fiber, achieving coupling loss of less than 3dB.)

[0064] Pulse Combining Output Module: Polarization-Multiplexed Coherent Combining Principle: When two amplified orthogonally polarized pulses (TE / TM mode) return from the Sagnac cavity to the PBS, the phase difference Δφ = 0 due to the strictly symmetrical optical path length of the Sagnac cavity (optical path difference ΔL = 0), meeting the coherent combining conditions. The PBS acts as a combiner, recombining the TE and TM mode pulses into a single linearly polarized pulse (the polarization direction is determined by the initial half-wave plate setting). The output energy is the sum of the energies of the two amplified pulses (ideally, energy efficiency >95%). Output Characteristics: The combined pulses are compressed to the femtosecond level by a compressor (not shown), achieving peak power more than doubled compared to single-channel amplification, and a beam quality factor M² < 1.3 (approaching the diffraction limit).

[0065] The present invention has the following beneficial effects: The core advantage of pulse amplification: In traditional single-channel CPA, the pulse energy is limited by the formula E∝mode field diameter×broadening amount, while the present invention uses two-channel parallel amplification. Under the same broadening amount, each channel only needs to carry 1 / 2 energy, so that the total energy is increased to twice that of a single channel (theoretical limit), and there is no need to rely on large mode field optical fiber (for example, from 40μm mode field to 20μm, the cost is reduced by more than 50%).

[0066] Engineering verification data: Under the same broadening amount (100ps), a single-channel CPA exhibits obvious nonlinear effects (self-phase modulation causes spectral broadening) when outputting 1mJ of energy. However, the two-channel synthesis method of the present invention can stably output 1.8mJ, an 80% increase in energy, and a spectral broadening of <10%.

[0067] Sagnac cavity has natural phase synchronization: Traditional beam splitting amplification schemes (such as MOPA dual-path amplification) require precise control of the optical path difference between the two paths (accuracy <10μm). Otherwise, phase mismatch during synthesis will lead to energy loss. The symmetrical optical path design of the Sagnac cavity eliminates the optical path difference from a physical structure, eliminating the need for additional electrically controlled delay lines or phase modulators, and reducing debugging time from several hours to minutes.

[0068] Cost advantage of non-polarization-maintaining fiber: Using conventional non-polarization-maintaining ytterbium-doped fiber (unit price <200 yuan / meter) to replace polarization-maintaining photonic crystal fiber (unit price >2000 yuan / meter) can reduce system costs by more than 70%. In addition, the coupling difficulty of non-polarization-maintaining fiber with NA=0.08 is low (ordinary collimators can achieve efficient coupling), and it is suitable for automated production.

[0069] Polarization Mode Dispersion (PMD) Suppression: Despite the use of non-polarization-maintaining fiber, the bidirectional transmission characteristics of the Sagnac cavity cancel out polarization mode dispersion (PMD) effects (clockwise and counterclockwise PMD). The measured polarization extinction ratio (PER) remains >20dB after amplification, outperforming traditional single-channel non-PMM amplification solutions (extinction ratio <10dB).

[0070] Reducing nonlinear effects: When the energy of a single pulse is reduced, the peak power in the optical fiber decreases synchronously (for example, if the energy of a single pulse is halved, the peak power is halved). The nonlinear threshold is increased to 2 times, effectively suppressing self-phase modulation (SPM) and four-wave mixing (FWM), and minimizing the time-domain waveform distortion of the amplified pulse (pulse width jitter after compression is <5%).

[0071] The present invention is adapted to engineering applications through modular design, including:

[0072] Full collimator coupling architecture: Both signal and pump light are coupled through collimators, eliminating the use of costly fiber splicing components (such as polarization-maintaining couplers). Each module (such as the signal input module, Sagnac cavity module, and pump module) can be independently packaged, supporting rapid assembly and disassembly and maintenance, making it suitable for mass production of industrial-grade lasers.

[0073] Wide pump power adaptation: By adjusting the half-wave plate angle, the two pulse energies can be dynamically distributed to adapt to different pump power inputs (for example, when the pump power is 50~200W, the synthesis efficiency can be maximized through energy allocation optimization), significantly improving the system robustness.

[0074] In summary, this invention, through its core architecture of "split-pulse amplification + symmetrical optical path synthesis," overcomes the multiple bottlenecks of traditional single-channel CPA in terms of energy, cost, and regulation complexity. It is particularly suitable for scenarios requiring high-energy 1030nm femtosecond lasers (such as silicon wafer precision processing and atmospheric remote sensing radar). Its technological innovation lies not only in the stacking of modules but also in leveraging the physical properties of the Sagnac cavity to solve the core challenges of beam splitting amplification (phase synchronization and polarization control). This provides a new approach to high-energy fiber laser design—moving from "breaking through the limitations of a single channel" to "multi-channel collaborative optimization," combining theoretical innovation with engineering practicality.

[0075] In some embodiments, the two pulses in the symmetrical optical path of the Sagnac cavity have the same optical path length, and no additional pulse synchronization adjustment is required. The energy distribution of the two pulses is controlled by adjusting the half-wave plate to optimize the synthesis efficiency, thereby achieving an increase in the pulse output energy under the same broadening amount.

[0076] The Sagnac cavity structure is constructed by designing the cavity as a ring optical path with a diameter of 20-30 cm. Two fiber collimators (Collimator A and Collimator B) are symmetrically placed within the ring, with the distance between them equal to the ring diameter. This ensures that the optical path lengths of the clockwise light incident from Collimator A and the counterclockwise light incident from Collimator B are strictly equal (error <10 μm, fixed in position by a precision mechanical bracket). The gain fiber within the ring is a non-polarization-maintaining ytterbium-doped fiber, forming a closed loop with Collimators A and B as endpoints. After the two pulses are incident from Collimators A and B, respectively, they travel in opposite directions through the gain fiber of exactly the same length.

[0077] The pulse synchronization and energy distribution mechanism involves adjusting the polarization state of the signal light through a half-wave plate and then splitting it into two pulses (TE mode and TM mode) with orthogonal polarizations by a polarization beam splitter (PBS). The two pulses achieve natural phase synchronization (optical path difference ΔL=0, phase difference Δφ=0) directly through the symmetrical optical path of the Sagnac cavity without the need for additional electrically controlled delay lines or phase modulators.

[0078] By rotating the angle of the half-wave plate (continuously adjustable from 0° to 90°), the polarization direction of the signal light incident on the PBS is changed (for example, 0° corresponds to 100% energy in the TE mode, and 45° corresponds to 50% energy each in the TE / TM mode). The energy distribution ratio of the two pulses is dynamically adjusted (for example, 3:7, 5:5, 7:3, etc.) to match the pump power and the saturation characteristics of the gain fiber, thereby maximizing the synthesis efficiency (theoretical synthesis efficiency ≥98%).

[0079] Traditional beam splitting amplification requires high-precision control of optical path difference (e.g., using electrically controlled fiber delay lines with an accuracy requirement of less than λ / 4). However, this embodiment fundamentally eliminates this optical path difference through physical structural design, eliminating the need for additional synchronization components. System commissioning time is reduced from several hours to less than 10 minutes, significantly reducing engineering complexity. The angle adjustment of the half-wave plate dynamically balances the energy of the two pulse paths. For example, when the pump power fluctuates, the half-wave plate can be adjusted to tilt energy distribution toward the path with higher gain, ensuring stable combined pulse energy (energy fluctuation less than 5%) and preventing the degradation of combining efficiency caused by energy imbalance between the two paths (e.g., combining efficiency drops below 80% when the energy difference exceeds 30%). At the same stretching level (e.g., 100ps), single-path amplification is limited by nonlinear effects to an energy threshold of 1mJ. However, this embodiment achieves a stable output of 1.8mJ through two-path combining, an 80% increase in energy, without the need for large-mode-area fiber (the mode field diameter is reduced from 40μm to 20μm), effectively breaking through the energy bottleneck of single-path amplification.

[0080] In some embodiments, the optical path adopts a device collimator coupling method, and the gain fiber of the Sagnac cavity is connected through an active fiber collimator, wherein the angle of the active fiber collimator corresponding to the axis facing the dichroic lens is 90°, and the angle of the active fiber collimator corresponding to the axis facing the half-wave plate is 0°.

[0081] The active fiber collimator is configured with a collimator facing the dichroic mirror (pump collimator). Using a collimating lens and an adjustment mechanism on the corresponding axis, the corresponding axis is fixed at a 90° angle (i.e., perpendicular to the horizontal plane), corresponding to the direction of the corresponding axis of the pump light (the corresponding axis of multimode pump light is typically aligned vertically at the fiber output). The collimator output is optically connected to the dichroic mirror at a 45° angle, ensuring that the direction of the corresponding axis of the pump light after reflection from the dichroic mirror matches that of the gain fiber (deviation <5°). The collimator facing the half-wave plate (signal collimator) has its corresponding axis angle set to 0° (parallel to the horizontal plane), consistent with the polarization state of the signal light output from the seed light source (default horizontal polarization), ensuring that the signal light's polarization state does not rotate after passing through the collimator (polarization extinction ratio change <1dB).

[0082] Optical connection details include an optional isolator between the pump collimator and the dichroic lens to prevent reflected light interference, and a glue-free optical connection between the signal collimator and the half-wave plate to ensure the phase stability of the femtosecond pulse (insertion loss <0.5dB).

[0083] The 90° setting of the corresponding axis aligns the direction of the corresponding axis of the pump light with the mode field distribution of the gain fiber (especially the mode field of the corresponding axis of large-core non-polarization-maintaining fiber). This increases coupling efficiency from the conventional 85% to over 92%, reduces pump energy loss (the pump power per channel can be reduced by 10%-15%), and eases the alignment accuracy requirements of the collimator (allowing the lateral offset tolerance to be relaxed from ±5μm to ±10μm). The 0° setting of the corresponding axis prevents additional polarization rotation of the signal light during the collimation process, ensuring that the polarization state of the signal light entering the half-wave plate is strictly controlled (for example, the initially horizontal polarization remains horizontal after passing through the collimator, with an angular deviation of less than 2°). This provides precise initial conditions for subsequent half-wave plate energy distribution adjustments, preventing energy imbalance between the two channels caused by polarization state drift (for example, energy distribution error >10% when the deviation is >5°). Pump and signal collimators with differentiated corresponding axis angles can use standardized components (such as commercial fiber collimators modified with adjustment rings for corresponding axes), eliminating the need for customized processing, reducing component costs by more than 30%, and supporting rapid replacement and maintenance (replacement time for a single collimator is less than 5 minutes).

[0084] In some embodiments, the signal light input module outputs signal light to a half-wave plate and a polarization beam splitter through a passive collimator, including: the passive collimator is used to collimate the input 1030nm signal light into a parallel light beam to match the input light path of the subsequent half-wave plate and the polarization beam splitter, the output end of the collimator is optically connected to the incident end of the half-wave plate, and the output end of the half-wave plate is optically connected to the first input end of the polarization beam splitter.

[0085] The passive collimator function is achieved by using a GRIN lens-type passive collimator. The input port is connected to a 1030nm seed light source (such as a fiber pigtail output). The output port collimates the fiber-optic output light with a divergence angle greater than 8° into a parallel beam (divergence angle less than 0.3mrad). The collimated beam diameter is 5mm (matching the clear aperture of the subsequent half-wave plate and PBS). The output end of the collimator and the incident end of the half-wave plate are precisely aligned using a three-dimensional adjustment mount (lateral offset less than ±5μm, angular deviation less than 0.1°), ensuring that the center of the beam coincides with the center of the half-wave plate (offset less than 100μm).

[0086] The optical path connection structure is connected to the first input end of the polarization beam splitter (PBS) through an air gap through the output end of the half-wave plate. The half-wave plate can rotate around the optical axis (adjustment accuracy 0.1°). The two output ends of the PBS are respectively connected to the two collimators (collimators A and B) of the Sagnac cavity.

[0087] The collimator converts the signal light into a parallel beam, avoiding uneven polarization adjustment at the half-wave plate caused by divergent light (for example, when the divergence angle exceeds 1 mrad, the polarization adjustment efficiency at the edge of the half-wave plate decreases by 15%). This ensures polarization consistency across the entire beam (polarization angle deviation between the edge and the center is less than 3°). Strict optical alignment reduces the transmission loss from the collimator to the half-wave plate to less than 0.3 dB, and the transmission loss from the half-wave plate to the PBS to less than 0.2 dB. Compared to uncollimated divergent light transmission (loss > 1 dB), this improves overall energy efficiency by over 5%. This is particularly critical for low-power seed sources (<10 mW), as it prevents startup difficulties in the amplifier link due to excessive loss. Using commercially available passive collimators (such as the Thorlabs C240TME-B) and half-wave plates (such as the Newport WPH05M-1030) eliminates the need for custom design, reduces module cost by 20%, and facilitates rapid optical path alignment (calibration time <15 minutes) using an adjustment mount.

[0088] In some embodiments, the polarization beam splitter splits the adjusted signal light into two orthogonally polarized pulses, including: the polarization beam splitter splits the linearly polarized light adjusted by the half-wave plate into two orthogonally polarized pulses transmitted along the horizontal and vertical directions based on the polarization state of the input signal light, and the energy distribution of the two pulses is dynamically balanced by adjusting the angle of the half-wave plate.

[0089] The polarization beam splitting principle is achieved by using a cube-shaped polarization beam splitter (PBS), which achieves a TE mode transmittance of >99%, a TM mode reflectance of >99%, and an extinction ratio of >30dB for 1030nm signal light. When linearly polarized light (e.g., 45° polarization) modulated by a half-wave plate enters the PBS, it is split into horizontally transmitted light (TE mode) and vertically reflected light (TM mode), forming two orthogonally polarized pulses.

[0090] The relationship between the half-wave plate angle θ and the two-way energy distribution is ETE=Eincos 2 θ, ETM = Einsin 2 θ, dynamic energy distribution is achieved by rotating θ (0°~90°) (e.g., TE:TM=3:1 when θ=30°, TE:TM=1:3 when θ=60°).

[0091] The optical path isolation design prevents the amplified pulses from entering the seed light source in reverse, thereby avoiding mode instability caused by optical feedback (such as pulse repetition frequency jitter > 1%) by connecting optical isolators (isolation > 30dB) to the TE mode and TM mode output ends of the PBS.

[0092] The high extinction ratio of the PBS ensures the polarization purity of the two pulses is greater than 99%, preventing crosstalk during amplification caused by non-orthogonal polarizations (for example, when crosstalk exceeds 5%, sidelobes appear in the synthesized pulses, affecting processing accuracy). This makes it particularly suitable for applications sensitive to polarization states (such as polarization-dependent material processing). Real-time adjustment of energy distribution via a half-wave plate allows adaptation to Sagnac cavities with different gain states (for example, when temperature changes in the gain fiber cause a gain difference between the two paths, adjusting θ shifts the energy toward the path with higher gain, compensating for the gain imbalance and ensuring a stable synthesis efficiency above 95%). Compared to traditional beam splitters (such as 50:50 fiber beam splitters), polarization beam splitters require no additional phase control and directly utilize polarization orthogonality to achieve physical splitting. Energy distribution adjustment does not introduce additional optical path loss (only the inherent loss of the PBS is less than 0.5dB). Compared to the 3dB loss of a fiber beam splitter, energy efficiency is improved by 60%.

[0093] In some embodiments, the two orthogonal polarized pulses are respectively coupled into the two collimators of the Sagnac cavity amplification module, where one pulse passes through the right collimator to enter the clockwise optical path of the Sagnac ring, and the other pulse passes through the lower collimator to enter the counterclockwise optical path of the Sagnac ring, forming an amplification path transmitted in opposite directions within the ring.

[0094] The collimator layout and light path direction are achieved by symmetrically installing two collimators (collimator A and collimator B) on the right and below (or left and above, according to the symmetry of the ring structure) of the Sagnac ring, with an angle of 180° between the two (i.e., the two ends of the ring diameter).

[0095] The clockwise optical path is a TE mode pulse incident from collimator A on the right, propagating clockwise along the upper half of the ring, passing through the upper half of the gain fiber, and then exiting from collimator B on the bottom (it does not actually exit but continues to propagate along the ring, forming a closed loop).

[0096] The counterclockwise optical path is incident from the lower collimator B through the TM mode pulse, propagates counterclockwise along the lower half of the ring, passes through the lower half of the gain fiber, and then exits from the right collimator A. The two pulses form closed transmission paths in opposite directions within the ring, ensuring that the length of the gain fiber passed through is exactly the same (error <0.1%).

[0097] The collimator coupling parameters include matching the collimator numerical aperture (NA=0.1) with the gain fiber NA=0.08, coupling efficiency>90%, and the incident light spot diameter (5μm) is slightly smaller than the fiber core diameter (20μm) to avoid multimode transmission caused by edge mode excitation (mode field matching error<10%).

[0098] The symmetrical collimator layout makes the optical path difference between the two pulses in the ring less than 10μm (corresponding to a time difference of less than 0.3fs), which is much smaller than the femtosecond pulse width (>50fs), ensuring that the amplified pulses completely overlap in the time domain and there is no energy loss caused by time delay during synthesis (for example, when the optical path difference is >100μm, the synthesis efficiency drops below 80%). Pulses transmitted in opposite directions all pass through the same length of gain fiber to avoid inconsistent gains in the two paths due to differences in the length of the gain fiber (for example, when the length difference is >1%, the gain difference is >5%). The measured difference in the amplification factor of the two pulses is less than 2%, ensuring the stability of the pulse energy after synthesis (energy fluctuation is <3%). The two pulses in the symmetrical path experience the same bending loss and mode evolution in the gain fiber, avoiding the excitation of high-order modes that may occur in single-path transmission (for example, clockwise transmission excites the LP11 mode, while counterclockwise transmission suppresses this mode). The beam quality factor M of the final synthesized pulse is 2 <1.3, better than the traditional single-channel amplifier M 2 <1.5.

[0099] In some embodiments, the laser splits a single pulse into two paths by combining chirped pulse amplification with split pulse amplification, and then amplifies them independently in a Sagnac cavity and then synthesizes them, including: based on the stretched pulse of chirped pulse amplification, splitting the stretched single pulse into two paths by split pulse amplification, performing power amplification in the symmetrical optical path of the Sagnac cavity respectively, utilizing the optical path symmetry of the Sagnac cavity to ensure the phase consistency of the two amplified pulses, and finally achieving coherent synthesis at the polarization beam splitter, and the synthesized pulse energy is the energy superposition of the two amplified pulses.

[0100] In the chirped pulse amplification (CPA) process, femtosecond pulses (50fs pulse width) output by a seed light source are first stretched to over 100ps (the stretching amount is adjustable, typically 100-500ps) using a dispersion-compensating fiber (DCF) or volume grating stretcher. This reduces the peak power (from 1GW to 10kW) and avoids nonlinear effects in subsequent amplification (such as spectral broadening caused by SPM). The stretched pulses then enter the split pulse amplification (DPA) module of the present invention: a half-wave plate and PBS split them into two orthogonal pulses, each of which enters a Sagnac cavity for amplification. The energy of each pulse is half that of the original pulse (e.g., 100μJ for a 200μJ original pulse), significantly below the nonlinear threshold of single-channel amplification (200μJ).

[0101] The split pulse amplification and synthesis involves independently amplifying the two stretched pulses in a Sagnac cavity (when the pump power is 100W, the single-channel gain is 20dB and the output energy is 2mJ). After amplification, they are coherently combined by PBS (due to phase consistency, the synthesis efficiency is 98%) to obtain stretched pulses with a total energy of 3.92mJ. These pulses are then compressed to 50fs by a compressor (volume Bragg grating or prism pair), and the peak power is increased to 78.4GW (the traditional single-channel CPA can only reach 40GW).

[0102] CPA reduces peak power by broadening, while DPA distributes the energy requirements of a single channel by splitting. The combination of the two increases total energy to twice the theoretical limit of single-channel CPA, while significantly reducing the nonlinear effects of each channel (single-channel peak power is halved, and the spectral broadening caused by SPM is reduced from 20nm to 8nm). The optical path symmetry of the Sagnac cavity ensures that the phase difference between the two stretched pulses is zero after amplification, resulting in no destructive interference loss during synthesis (in traditional beam splitting amplification, if the phase difference exceeds π / 2, the synthesis efficiency drops below 70%). The measured time-domain contrast of the synthesized pulses is greater than 20:1 (with no significant sidelobes), making it suitable for precision machining scenarios requiring high pulse quality. The DPA module of this embodiment can be directly integrated into existing CPA systems, requiring only the addition of a beam splitter and combiner module after the stretcher, without the need to modify the seed light source and compressor. This reduces system upgrade costs by over 50%, demonstrating excellent technical compatibility.

[0103] In some embodiments, the ytterbium-doped gain fiber in the Sagnac cavity amplification module is a non-polarization-maintaining fiber, and the core diameter and numerical aperture parameters of the non-polarization-maintaining fiber suppress polarization mode dispersion through the symmetrical optical path design of the Sagnac cavity.

[0104] The parameters of non-polarization-maintaining gain fiber are selected by using ytterbium-doped non-polarization-maintaining fiber (such as CorActive model Yb-ND-20 / 400, core diameter 20μm, cladding diameter 400μm, NA=0.08). Compared with traditional polarization-maintaining photonic crystal fiber (core 40μm, NA=0.05), it has a smaller core diameter and a larger NA, which reduces the coupling difficulty (conventional collimators can achieve >90% coupling efficiency), and the cost is only 1 / 5 of the latter (unit price <200 yuan / meter vs >1000 yuan / meter).

[0105] The principle of PMD suppression by the Sagnac cavity includes the existence of polarization mode dispersion (PMD) in non-polarization-maintaining optical fiber, which results in different transmission speeds of the two polarization states (group velocity difference Δv≈10^6m / s). However, the bidirectional transmission characteristics of the Sagnac cavity make the effects of the corresponding axes in the clockwise direction and the counterclockwise direction cancel each other out: the PMD delay experienced by the clockwise TE mode is equal to the PMD delay experienced by the counterclockwise TM mode, but in opposite directions. Ultimately, the PMD effect is completely compensated during synthesis (residual PMD < 0.1ps / nm).

[0106] The use of non-PM fiber reduces gain module costs by 70%. The large NA design (NA=0.08) allows for increased collimator misalignment (lateral misalignment tolerance ±15μm), accommodating mechanical assembly errors in automated production (traditional PM fiber with NA=0.05 has a tolerance of only ±5μm). Despite the use of non-PM fiber, the Sagnac cavity's PMD compensation mechanism maintains a polarization extinction ratio (PER) of >20dB for both amplified pulses (compared to <10dB for conventional single-channel non-PM fiber), meeting the requirements of polarization-sensitive applications such as polarization-dependent spectroscopy. The 20μm core diameter balances the mode field and nonlinear threshold (doubling the mode field diameter increases the nonlinear threshold by a factor of four). Compared to 40μm PM fiber, this embodiment reduces nonlinear effects by 50% at the same energy level. Furthermore, the design eliminates the need for specially designed low-NA collimators, significantly reducing engineering complexity.

[0107] In some embodiments, in the pump coupling module, after the 976nm pump light is output through the passive fiber collimator, it is reflected by a dichroic lens and focused onto the gain fiber collimator of the Sagnac cavity. The dichroic lens has high transmittance for the 1030nm signal light and high reflection for the 976nm pump light.

[0108] The dichroic mirror and pump optical path include a dichroic mirror (DCM) coated with a multilayer dielectric film, achieving a transmittance of >95% for the 1030nm signal light and a reflectance of >99.5% for the 976nm pump light. The mirror measures 12.7mm × 12.7mm and is mounted between the Sagnac cavity and the pump collimator, at a 45° angle to the optical axis. The 976nm pump light, after being collimated by a passive collimator (on-axis collimation, divergence angle <5° on the corresponding axis), is incident on the dichroic mirror, reflected, and focused onto the Sagnac cavity's gain fiber collimator (Collimator A or B). It then propagates coaxially with the signal light within the gain fiber (the pump light is multimode, the signal light is single-mode, and the mode field overlap is >80%).

[0109] The common optical path isolation design includes installing a wavelength-selective isolator (which only passes 976 nm) at the pump light incident end of the dichroic lens to prevent the signal light from reversely entering the pump light source and avoid increasing the stimulated emission (SSE) noise of the pump source (noise suppression ratio > 30 dB).

[0110] The bichromatic mirror achieves wavelength splitting of the pump and signal beams, with pump light reflection efficiency exceeding 99% and signal light transmission loss less than 0.5dB. Compared to traditional beamsplitters (e.g., dichroic plate loss exceeding 1dB), this improves energy efficiency by over 3%. Especially for high-power pumps (>200W), this reduced loss translates into effective gain (every 10W of pump power loss corresponds to a 0.5mJ increase in signal energy). In the common optical path, the pump and signal beams share the same transmission axis, eliminating the need for a three-dimensional alignment fixture to calibrate the overlap of the two beams (as traditional separation beams require calibration with an accuracy of <±5μm). This reduces system integration time from 2 hours to 30 minutes, and improves vibration resistance (vibration-induced optical path deviation affects the common optical path by less than 10%). The bichromatic mirror can be customized to accommodate other pump wavelengths (e.g., 915nm) by simply replacing the lens coating without modifying the overall architecture. This supports modular wavelength expansion of the laser (e.g., from 1030nm to 1550nm).

[0111] In some embodiments, the angles of the corresponding axes of the active fiber collimator are set to: 90° for the corresponding axis of the collimator facing the dichroic lens, for matching the polarization direction of the pump light to improve the coupling efficiency; 0° for the corresponding axis of the collimator facing the half-wave plate, for maintaining the polarization state stability of the signal light. The independent transmission characteristics of the signal light and the pump light are optimized by differentiating the angles of the corresponding axes.

[0112] The 90° design of the corresponding axis of the pump collimator includes the direction of the corresponding axis of the pump light output by the pump light source (multimode semiconductor laser) being perpendicular to the output end face of the optical fiber (i.e., in the vertical direction), and the angle of the corresponding axis of the active collimator facing the dichroic lens being set to 90°. This ensures that the corresponding axis of the collimated pump light is aligned with the corresponding axis of the gain fiber (the corresponding axis of the gain fiber is determined by the coating marking and is usually in the vertical direction), resulting in a mode field matching factor greater than 0.9 (a 20% improvement compared to random angle matching).

[0113] The 0° design of the corresponding axis of the signal collimator includes the default horizontal polarization of the signal light output by the seed light source (corresponding to the 0° axis) and the angle of the corresponding axis of the active collimator facing the half-wave plate is set to 0°. This ensures that the polarization state of the signal light remains unchanged after passing through the collimator (polarization rotation angle <1°), providing a stable initial condition for the half-wave plate to accurately adjust the polarization state (for example, the initial horizontal polarization remains horizontal polarization after passing through the collimator, with an angle deviation of <0.5°).

[0114] The matching of corresponding axes increases the mode field coupling efficiency of the pump light entering the gain fiber from 80% to 93%, reducing pump power waste (the pump power per channel can be reduced by 15W), while also reducing the thermal load of the gain fiber (the temperature rise is reduced from 50°C to 35°C), and extending the fiber life (the life span is increased from 5000 hours to 8000 hours).

[0115] The 0° setting of the corresponding axis avoids polarization rotation errors introduced by the collimator, ensuring a linear correspondence between the half-wave plate adjustment angle and the actual energy distribution (e.g., theoretically, θ = 45° corresponds to a 50:50 energy distribution, with an actual deviation of <2%). Compared to a non-optimized design (deviation >10%), energy distribution accuracy is improved by a factor of five, facilitating real-time adjustment by the automated control system (adjustment steps are refined from 5° to 1°). The polarization transmission characteristics of the signal and pump lights within the collimator are independent, with PDLs of <0.1dB for the signal light and <0.5dB for the pump light (compared to PDLs >1dB for the conventional unified corresponding axis setting). This improves the polarization stability of the overall system, making it particularly suitable for polarization-sensitive precision machining (such as the fabrication of polarization-dependent micro-nanostructures).

[0116] In some embodiments, by integrating real-time energy monitoring and intelligent algorithms, the hysteresis of traditional manual adjustment is overcome and the dynamic optimal distribution of the two-way pulse energy of the Sagnac cavity is achieved.

[0117] By integrating energy sensors (response speed <1μs, accuracy ±1%) at the two output ends of the polarization beam splitter (PBS), the TE mode and TM mode pulse energy values ​​ETE and ETM are collected in real time.

[0118] The control module utilizes an FPGA+DSP architecture, featuring a built-in adaptive PID algorithm and energy balance model. It calculates the target angle θtarget of the half-wave plate based on real-time energy data and drives the half-wave plate's rotation via a stepper motor (with 0.01° accuracy). A gain-to-fiber temperature-gain mapping relationship is established: A thermocouple attached to the fiber (with an accuracy of ±0.5°C) monitors temperature in real time. The algorithm automatically compensates for the effects of temperature drift on gain. For example, for every 10°C increase in temperature, the gain decreases by 2%. The algorithm automatically adjusts θ to maximize energy distribution toward the lower temperature range.

[0119] Algorithm implementation process: Initialization phase: Scan the half-wave plate from 0° to 90°, record the ETE / ETM ratio at different angles, and generate an energy distribution mapping table (resolution 0.5°).

[0120] During the real-time adjustment phase, sensors collect ETE, ETM, and fiber temperature T; the algorithm calculates the current optimal energy ratio: (G1 / G2 are the two-way gain coefficients, obtained from a temperature mapping table). The target angle θtarget is obtained by lookup in the kopt table, and the motor is then adjusted with an adjustment step size of less than 0.2° and an adjustment time of less than 50ms. Compared to manual adjustment (adjustment cycle >1min), the algorithm-driven response time is shortened to milliseconds, adapting to pump power fluctuations (for example, when the grid voltage fluctuates by ±10%, the energy stabilization time is reduced from 30s to 2s), and the energy fluctuation amplitude is suppressed from ±15% to ±3%. A temperature-gain compensation model automatically adapts to ambient temperature fluctuations (-10°C to 40°C) without manual intervention, especially in high-temperature outdoor or industrial environments (for example, at 40°C, traditional manual adjustment efficiency drops by 40%, while this embodiment maintains efficiency above 98%). Dynamic allocation based on real-time gain matching increases the combined efficiency from 92% with manual adjustment to over 97%, and increases the output energy by 12% at the same pump power (for example, at a 100W pump, the output energy increases from 3.8mJ to 4.3mJ).

[0121] In some embodiments, by using historical data to train a neural network model, the probability of occurrence of nonlinear effects (such as SPM and SBS) is predicted in advance, and the pulse parameters are dynamically adjusted to break through the passivity of traditional threshold monitoring.

[0122] Spectrometers (0.1nm resolution), power meters, and oscilloscopes (50GHz bandwidth) were deployed at key laser nodes (seed light source, stretcher output, Sagnac cavity input / output, and compressor output) to collect over 20 parameters in real time, including spectral width, pulse energy, and time-domain waveform. Faults such as SPM (spectral broadening > 15nm) and SBS (backscattered light > 5%) were induced at different energies, pulse widths, and fiber lengths. The parameter combinations that occurred during the faults were annotated, and an LSTM neural network model was trained (with prediction accuracy > 95%) to output the probability P of nonlinear effects occurring within the next 10ms.

[0123] The real-time control strategy includes triggering the following adjustments when P ≥ 80%: pulse energy redistribution: increasing the energy distribution difference through half-wave plates (such as adjusting from 5:5 to 7:3), reducing the single-channel peak power (reduction > 20%); dynamic pump power peak clipping: during the pulse peak period (identified by the oscilloscope trigger signal), the pump current is temporarily reduced by 10% to 15% (duration < 1μs) to suppress the SBS threshold; adaptive adjustment of the stretching amount: feedback control of the stretcher dispersion (such as increasing from 100ps to 150ps) to further reduce the peak power (formula: P peak = E / τ, a 50% pulse width widening results in a 33% decrease in peak power).

[0124] Compared to the traditional passive strategy of shutting down after threshold detection, this embodiment reduces the incidence of nonlinear effects from 5 times / hour to <0.1 times / hour, extending the continuous operation time of processing lasers from 2 hours to over 20 hours, significantly improving industrial processing efficiency. In high-energy output scenarios close to the nonlinear threshold (e.g., above 4mJ), dynamic parameter adjustment allows energy output to be increased by 25% (safely increasing from 4mJ to 5mJ) while maintaining a spectral width of <12nm (the traditional method widens the spectrum to 18nm at 4mJ). As operational data accumulates, the model's prediction accuracy continues to improve (updated every 100 hours of training, with a prediction delay of <200ns), adapting to parameter differences between different batches of gain fibers (e.g., when the core diameter deviation is ±5%, the system self-calibration time is <10 minutes).

[0125] In some embodiments, laser parameters are automatically configured through reinforcement learning algorithms for different application scenarios (precision machining, remote sensing ranging, scientific research), achieving "plug and play" multi-mode switching.

[0126] Three typical modes are preset: Processing mode: high energy (>3mJ), low repetition rate (<10kHz), requiring pulse peak power >50GW, suitable for material micro-processing; Remote sensing mode: medium energy (1~2mJ), high repetition rate (>100kHz), requiring long-term stability (energy fluctuation <2%), suitable for long-distance ranging; Scientific research mode: ultra-narrow spectrum (<5nm), low noise (RIN <-150dB / Hz), suitable for precision spectral measurement.

[0127] The reinforcement learning control architecture includes: state space: including 30+ state variables such as energy, pulse width, spectral width, pump current, collimator temperature, etc.; action space: adjusting 10+ controllable parameters such as half-wave plate angle, pump power, stretcher dispersion, isolator attenuation coefficient, etc.; reward function: set according to mode requirements, such as the processing mode reward function is R=0.6E+0.3M 2 +0.1τ (E is energy, M 2 is the beam quality, τ is the pulse width), and the optimal strategy is trained through a deep Q network (DQN).

[0128] Mode switching process: The user selects the mode through the human-machine interface, and the algorithm calls the corresponding parameter group from the pre-trained model (for example, the processing mode automatically sets the half-wave plate to 40°, the pump to 120W, and the width expansion to 150ps). It then makes fine adjustments based on the real-time status during operation (for example, automatically compensating the angle of the corresponding axis of the collimator when the ambient temperature changes). The switching time is less than 30s.

[0129] Traditional multimode lasers require manual parameter calibration (each switch takes >30 minutes). This embodiment automatically configures parameters through an algorithm, eliminating the need for professional intervention. This makes it particularly suitable for multi-purpose shared equipment (e.g., use by multiple research groups in university laboratories, reducing equipment idle time by 60%). In remote sensing mode, by dynamically reducing pump power fluctuations (the algorithm controls pump power ripple to <0.1%), energy stability is improved from ±5% (compared to traditional solutions) to ±1.5%, and ranging accuracy is increased from ±10cm to ±3cm. In scientific research mode, spectral purity is improved by 40% (stray light suppression ratio increases from 20dB to 30dB), meeting the noise requirements of high-end spectrometers. The algorithm is compatible with gain fibers from different manufacturers (such as Corning and CorActive), and compensates for fiber parameter differences through online calibration (for example, when the mode field diameter varies by ±10%, the algorithm automatically adjusts the collimator coupling parameters to maintain coupling efficiency >90%), reducing component procurement constraints.

[0130] In some embodiments, by building a digital twin model of the laser, the physical entity status is mapped in real time, predictive maintenance and fault location are achieved, breaking through the blindness of traditional regular maintenance.

[0131] A Sagnac cavity optical field transmission model was established using COMSOL Multiphysics, and a gain fiber thermal conduction model was established using ANSYS. Combined with measured data (such as collimator coupling efficiency and pump light absorption coefficient), a high-precision simulation model (with error of less than 5% for key parameters) was constructed. The Internet of Things (IoT) module synchronized real-time sensor data (temperature, power, polarization state, etc.) from the physical laser to drive dynamic updates of the digital twin model, achieving real-time mapping of physical and virtual states (latency less than 100 μs).

[0132] Health management features include: Lifespan prediction: Based on the gain fiber's cumulative luminous flux (lifespan begins to decline when it exceeds 10^12 photons per meter), combined with real-time pump power and operating time, the remaining fiber lifespan is predicted (with an error of less than 10%), providing early warning of replacement intervals (for example, triggering an alarm when the remaining lifespan is less than 200 hours). Fault location: When the synthesis efficiency suddenly drops by more than 10%, the digital twin model uses reverse simulation to quickly locate the fault point (such as collimator offset, half-wave plate damage, and increased fiber splice loss). The positioning accuracy is greater than 95%, reducing fault diagnosis time to less than 5 minutes compared to traditional section-by-section troubleshooting (which takes more than 2 hours).

[0133] Lifespan prediction avoids excessive maintenance (for example, the traditional method of replacing optical fibers every 300 hours is replaced on demand in this embodiment, increasing fiber utilization by 30%), reducing maintenance costs by 40%. The fault location function reduces downtime (annual downtime is reduced from 100 hours to 20 hours), making it particularly suitable for the continuous operation requirements of industrial production lines. The digital twin model evaluates laser performance degradation in real time (for example, automatic compensation when the angle drift of the collimator corresponding axis is greater than 5°), ensuring parameter stability in long-term operation (energy attenuation is controlled from 20% to less than 5% after one year of leaving the factory), and improving equipment reliability. In the development of new laser models, the digital twin model can virtually test different parameter combinations such as collimator angles and fiber lengths, replacing 80% of physical prototype testing, shortening the R&D cycle from 12 months to 6 months, and reducing R&D costs by more than 50%.

[0134] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present application, and such modifications or substitutions should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A high-energy femtosecond fiber laser, characterized in that: include: A signal light input module is used to input 1030nm signal light, and the signal light input module outputs the signal light to the half-wave plate and polarization beam splitter through a passive collimator; The polarization beam splitting and energy adjustment module includes the half-wave plate and the polarization beam splitter, wherein the half-wave plate is used to adjust the polarization state of the input signal light, and the polarization beam splitter splits the adjusted signal light into two orthogonally polarized pulses, including: the polarization beam splitter splits the linearly polarized light adjusted by the half-wave plate into two orthogonally polarized pulses transmitted in the horizontal direction and the vertical direction based on the polarization state of the input signal light, and the two orthogonally polarized pulses include a horizontally polarized TE mode and a vertically polarized TM mode; the energy distribution of the two pulses is dynamically balanced by adjusting the angle of the half-wave plate; the polarization beam splitting principle is realized by selecting a cubic polarization beam splitter, and when the linearly polarized light adjusted by the half-wave plate is incident on the PBS, it is decomposed into horizontally transmitted light and vertically reflected light, forming two orthogonally polarized pulses; dynamic energy distribution is achieved by rotating θ from 0° to 90°, and when θ=30°, TE:TM=3:1, and when θ=60°, TE:TM=1:3; The Sagnac cavity amplification module includes two symmetrically arranged collimators and an ytterbium-doped gain fiber within the ring. Two pulses are coupled into the Sagnac cavity through the two collimators, propagated in opposite directions within the ring, and amplified by the ytterbium-doped gain fiber. The ytterbium-doped gain fiber is a non-polarization-maintaining fiber. A pump coupling module, comprising a passive fiber collimator and a dichroic lens for outputting 976nm pump light. The pump light is coupled into the gain fiber collimator of the Sagnac cavity through the dichroic lens to provide energy for intra-ring amplification. A pulse synthesis output module, wherein the two amplified pulses are synthesized into a single pulse at the polarization beam splitter and outputted by utilizing the optical path symmetry of the Sagnac cavity; The laser splits a single pulse into two paths by combining chirped pulse amplification with split pulse amplification, and then independently amplifies the two pulses in the Sagnac cavity before synthesizing them. The method includes: based on the stretched pulse of chirped pulse amplification, splitting the stretched single pulse into two paths by split pulse amplification, performing power amplification in the symmetrical optical path of the Sagnac cavity respectively, ensuring the phase consistency of the two amplified pulses by utilizing the optical path symmetry of the Sagnac cavity, achieving coherent synthesis at the polarization beam splitter, and achieving the energy of the synthesized pulse by the energy superposition of the two amplified pulses; wherein, the chirped pulse amplification process outputs a femtosecond pulse with a pulse width of 50fs from a seed light source, which is first stretched to more than 100ps by a dispersion-compensating fiber or a volume grating stretcher, and the stretched pulse enters the split pulse amplification module: it is split into two orthogonal pulses by a half-wave plate and a polarization beam splitter, and enters the Sagnac cavity for amplification respectively, and the energy of a single pulse is 1 / 2 of the original pulse; By using historical data to train the neural network model, the probability of nonlinear effects is predicted in advance, and the sub-pulse parameters are dynamically adjusted. By deploying spectrometers, power meters, and oscilloscopes at key nodes of the laser, including the seed light source, stretcher output, Sagnac cavity input / output, and compressor output; real-time acquisition of spectral width, pulse energy, and time-domain waveform parameters, the faults of spectral broadening > 15nm and backscattered light > 5% are induced at different energies, pulse widths, and fiber lengths, the parameter combination at the time of the fault is marked, the LSTM neural network model is trained, and the probability of nonlinear effects within the next 10ms is output. The real-time control strategy includes triggering the following adjustments when the probability of nonlinear effects is ≥80%: sub-pulse energy redistribution: increasing the energy distribution difference through a half-wave plate to reduce the single-channel peak power; temporarily reducing the pump current by 10%~15% during the pulse peak period to suppress the backscattered light threshold; feedback control of the stretcher dispersion to reduce the peak power; By integrating energy sensors at the two output ends of the polarization beam splitter, the TE and TM mode pulse energy values ​​are collected in real time. The target angle of the half-wave plate is calculated based on the real-time energy data, and the half-wave plate is driven to rotate by a stepper motor. A gain fiber temperature-gain mapping relationship is established: the temperature is monitored in real time by thermocouples attached to the optical fiber, and the effect of temperature drift on the gain is automatically compensated. The implementation process includes: initialization phase: scanning the half-wave plate from 0° to 90°, recording the ETE / ETM ratio at different angles, and generating an energy distribution mapping table; real-time adjustment phase: sensors collect ETE, ETM, and optical fiber temperature T; and calculate the current optimal energy ratio: , G1 / G2 are the two-way gain coefficients, which are obtained from the temperature mapping table; the target angle is obtained by looking up the table according to the current optimal energy ratio, and the drive motor is adjusted with an adjustment step of <0.2° and an adjustment time of <50ms.

2. The femtosecond fiber laser according to claim 1, characterized in that The two pulses in the symmetrical optical path of the Sagnac cavity have the same optical path length, and no additional pulse synchronization adjustment is required. The energy distribution of the two pulses is controlled by adjusting the half-wave plate to optimize the synthesis efficiency, thereby achieving an increase in the pulse output energy under the same broadening amount.

3. The femtosecond fiber laser according to claim 1, wherein The optical path adopts a device collimator coupling mode, and the gain fiber of the Sagnac cavity is connected through an active fiber collimator, wherein the angle of the active fiber collimator corresponding to the axis facing the dichroic lens is 90°, and the angle of the active fiber collimator corresponding to the axis facing the half-wave plate is 0°.

4. The femtosecond fiber laser according to claim 3, characterized in that The angles of the corresponding axes of the active fiber collimator are set as follows: the angle of the corresponding axis of the collimator facing the dichroic lens is 90°, which is used to match the polarization direction of the pump light to improve the coupling efficiency; the angle of the corresponding axis of the collimator facing the half-wave plate is 0°, which is used to maintain the polarization state stability of the signal light. The independent transmission characteristics of the signal light and the pump light are optimized by differentiating the angles of the corresponding axes.

5. The femtosecond fiber laser according to claim 1, wherein: The signal light input module outputs signal light to a half-wave plate and a polarization beam splitter through a passive collimator, including: The passive collimator is used to collimate the input 1030nm signal light into a parallel beam to match the input optical path of the subsequent half-wave plate and polarization beam splitter. The output end of the collimator is optically connected to the incident end of the half-wave plate, and the output end of the half-wave plate is optically connected to the first input end of the polarization beam splitter.

6. The femtosecond fiber laser according to claim 1, characterized in that The two orthogonal polarization pulses are respectively coupled into the two collimators of the Sagnac cavity amplification module, where one pulse passes through the right collimator and enters the clockwise optical path of the Sagnac ring, and the other pulse passes through the lower collimator and enters the counterclockwise optical path of the Sagnac ring, forming an amplification path transmitted in opposite directions within the ring.

7. The femtosecond fiber laser according to claim 1, wherein: The ytterbium-doped gain fiber in the Sagnac cavity amplification module is a non-polarization-maintaining fiber, and the core diameter and numerical aperture parameters of the non-polarization-maintaining fiber suppress polarization mode dispersion through the symmetrical optical path design of the Sagnac cavity.

8. The femtosecond fiber laser according to claim 1, wherein: In the pump coupling module, the 976nm pump light is output through the passive fiber collimator, reflected by the dichroic lens and focused onto the gain fiber collimator of the Sagnac cavity. The dichroic lens has high transmittance for the 1030nm signal light and high reflection for the 976nm pump light.

Images