A metal workpiece ultrafast laser processing system based on surface quality feedback

Abstract

The application belongs to the technical field of additive manufacturing equipment post-processing, laser rapid forming equipment and laser beam precision machining, and particularly relates to a metal workpiece ultrafast laser processing system based on surface quality feedback. The system integrates an ultrafast laser processing unit, an online detection unit, a motion control unit and a central processing unit on the same rack assembly. The central processing unit performs collaborative control on initial topography detection, working condition identification, defocusing spot and power calculation, processing path generation, round feedback control and termination determination of the surface of the additive manufacturing workpiece through a data acquisition module, a state recognition module, a parameter calculation module, a scanning strategy module, a feedback decision module and a termination output module. According to the initial maximum height difference, the cumulative removal depth and the surface quality feedback, the application adaptively adjusts the scanning strategy and the laser processing parameters, realizes the surface ultrafast laser closed-loop finishing machining, and improves the machining stability and the quality consistency.

Description

Technical Field

[0001] This application belongs to the fields of additive manufacturing equipment post-processing, laser rapid prototyping equipment and laser beam precision machining technology. Specifically, it relates to an ultrafast laser processing system for metal workpieces based on surface quality feedback. Background Technology

[0002] Ultrafast laser processing technology, characterized by short pulse widths, high peak power density, and small heat-affected zones, enables high-precision removal of material surfaces and is now widely used in surface finishing, polishing, texturing, and microstructure fabrication of metal workpieces. In additive manufacturing equipment and laser rapid prototyping equipment manufacturing, laser rapid prototyping technologies such as SLM can fabricate complex metal components. However, the resulting metal workpieces often exhibit problems such as large peak-valley undulations, significant particle adhesion, pronounced layering, uneven local melting, and high roughness, making it difficult to directly meet the surface quality requirements of applications such as high-precision assembly, fatigue service, fluid contact, or friction and wear. Therefore, using ultrafast laser processing systems as surface finishing and post-processing equipment for additively manufactured parts has become an important technical means to improve the surface quality, dimensional consistency, and service performance of laser-produced parts.

[0003] Existing ultrafast laser finishing equipment for post-processing of additively manufactured parts typically includes a laser, scanning galvanometer, worktable, and control system. During operation, operators usually pre-set parameters such as laser power, scanning speed, scanning spacing, and number of processing passes, and the equipment then performs laser processing on the workpiece surface according to these parameters. Therefore, it suffers from the following shortcomings in terms of equipment structure and unit configuration:

[0004] 1. Existing systems lack adaptability to the initial surface condition of the formed parts. Most existing control devices only form conventional connections with the laser, galvanometer scanning assembly, and motion platform. They typically only have basic parameter setting functions such as laser power, scanning speed, scanning spacing, and number of scans, lacking configurations for hierarchical identification and differentiated control based on the initial surface undulations of the additive manufacturing formed parts. For metal workpieces prepared by different laser rapid prototyping equipment, or different areas on the same formed part with significantly different surface conditions, existing systems often still use a uniform or fixed scanning execution method, making it difficult to adjust the processing strategy according to the actual surface condition, resulting in poor equipment adaptability.

[0005] 2. Existing systems lack sufficient energy input matching capability under defocus conditions. The existing power output section, defocus adjustment section, and focusing optics section typically lack linkage. The laser output power setting is independent of the defocus state of the workpiece surface relative to the focal plane, making it difficult to synchronously reflect changes in the actual spot size under defocus conditions in the power output control. Consequently, the system struggles to maintain stable and reasonable energy input per unit area under different defocus conditions, easily leading to localized excessively high or low energy levels, thus affecting the removal uniformity and surface consistency of the processed area.

[0006] 3. The existing system lacks a close structural connection between the scanning execution component and the actual processing state of the workpiece surface. Although the galvanometer scanning assembly, motion platform, laser output component, and control device can work together to complete the processing action, the components mostly coordinate in a conventional manner, without forming a system configuration that dynamically adjusts the operation based on changes in the surface state of the additively manufactured part. Especially when dealing with workpieces with significant surface undulations and those with nearly flat surfaces, the existing system struggles to reflect different processing strategies and execution intensities through its internal structural configuration, thus failing to balance processing efficiency and accuracy.

[0007] 4. Although the surface detection unit, scanning execution unit, power output unit, defocus adjustment unit and motion control unit in the existing equipment can realize detection, processing or displacement actions respectively, the functional units are mostly conventionally coordinated and have not yet formed a system structure for linkage control around the initial surface state of the workpiece, the change of the defocus spot, the feedback of the processing process and the termination judgment. As a result, the equipment has insufficient adaptability to different initial surface states, and the processing stability and surface quality consistency still need to be improved.

[0008] Therefore, it is necessary to provide an ultrafast laser processing system for metal workpieces based on surface quality feedback, so that it can be applied to the surface finishing post-processing scenarios of additive manufacturing parts, laser rapid prototyping parts and other metal workpieces, in order to improve the shortcomings of existing systems in terms of initial surface state adaptation, energy control matching under defocus conditions and detection and processing collaborative configuration, thereby improving the stability, consistency and adaptability of the surface finishing process of metal workpieces. Summary of the Invention

[0009] To address the shortcomings of existing ultrafast laser processing systems for metal workpieces, such as insufficient adaptability to initial surface conditions, inadequate matching between power control and defocusing conditions, and insufficient coordination between detection and processing, this application provides an ultrafast laser processing system for metal workpieces based on surface quality feedback. By setting up a working condition identification and strategy generation module corresponding to the initial surface condition of the workpiece, a power control module coupled with defocusing adjustment and actual spot state, and combining the coordinated configuration of the online detection unit and the ultrafast laser processing unit, adaptive scanning control for the same processing area, energy matching control under defocusing conditions, and surface quality feedback adjustment during processing are achieved. This improves the processing stability, surface quality consistency, and adaptability to different initial surface conditions in the surface finishing process of metal workpieces.

[0010] This application provides an ultrafast laser processing system for metal workpieces based on surface quality feedback, including a frame assembly, an ultrafast laser processing unit, an online inspection unit, a motion control unit, and a central processing unit. The frame assembly includes a base, a column, and a crossbeam. The motion control unit is disposed on the base, and a fixture for fixing the metal workpiece is mounted on it. The ultrafast laser processing unit and the online inspection unit are disposed on the crossbeam and located above the fixture. The processing optical path of the ultrafast laser processing unit faces the surface of the metal workpiece, and the detection end of the online inspection unit faces the surface of the metal workpiece to be processed. There is a pre-calibrated fixed offset between the detection position and the processing position. The central processing unit is disposed on the side of the frame assembly and is communicatively connected to the ultrafast laser processing unit, the online inspection unit, and the motion control unit, respectively. It is used to control the motion control unit to switch the metal workpiece between the detection position and the processing position according to the fixed offset, so that the detection area corresponds to the processing area.

[0011] The central processing unit includes at least a data acquisition module, a status recognition module, a parameter calculation module, a scanning strategy module, a feedback decision module, and a termination output module. The data acquisition module is connected to the online detection unit and is configured to receive measurement data output by the online detection unit, reconstruct the measurement data into a three-dimensional height matrix z(i,j) within the measurement area, calculate the initial maximum height difference and initial arithmetic mean roughness of the surface to be processed based on the three-dimensional height matrix, and calculate the current cumulative removal depth and current arithmetic mean roughness based on the detection results of the same area registered one-to-one with the current processing area after each round of processing.

[0012] The status recognition module is connected to the data acquisition module and is configured to receive the initial maximum height difference and compare the maximum height difference with a preset working condition recognition threshold to determine whether the current surface to be processed is in a large amount removal working condition or a fine homogenization working condition; the working condition recognition result is used to determine the target cumulative removal depth setting, the basic scan group number calculation, the initial planned total scan volume setting, the configuration of the number of new scans per round, the average power setting, and the termination logic judgment.

[0013] The parameter calculation module is connected to the state recognition module, motion control unit, and ultrafast laser processing unit, and is configured to calculate parameters based on the material's single-pulse ablation threshold, laser wavelength λ, and focal plane beam waist radius. 1. Preset the defocusing amount Δz and the laser repetition frequency, and calculate the actual effective spot radius ω(Δz), the actual processed spot diameter D, and the average output power of the laser at the defocusing position;

[0014] The scanning strategy module is connected to the state recognition module and the parameter calculation module, and is configured to generate the actual processing path based on the working condition recognition result, power calculation result, and overlap rate matching result, and control the ultrafast laser processing unit to perform scanning processing; the feedback decision module is connected to the data acquisition module, the state recognition module, and the scanning strategy module, and is configured to perform closed-loop control based on the initial morphology parameters, the working condition recognition result, and the feedback measurement result after each processing cycle; the termination output module is connected to the feedback decision module, and is also connected to the ultrafast laser processing unit and the motion control unit, and is configured to output a control command to stop processing or continue processing based on the deterioration counter, the relative change rate of surface roughness β, and the comparison result between the current cumulative removal depth and the target cumulative removal depth.

[0015] In a preferred implementation, the scanning strategy module further includes an overlap rate matching submodule and a scanning path generation submodule. The overlap rate matching submodule is configured to determine the pulse overlap rate and the scanning line overlap rate based on the actual processing spot diameter D, the laser repetition frequency, the galvanometer scanning speed v, and the scanning line spacing L. The scanning path generation submodule is configured to generate the actual processing path based on the working condition identification result, the power calculation result, and the overlap rate matching result.

[0016] In a preferred implementation, the parameter calculation module is further configured to determine the actual effective spot size and the average output power of the laser at the defocus position according to the following relationship:

[0017] ;

[0018] D = 2ω(Δz);

[0019] And make the peak energy density at the defocus position satisfy the product of the material's single-pulse ablation threshold and the safety factor ξ, wherein the safety factor ξ ranges from 1.05 to 1.35.

[0020] In a preferred implementation, the overlap rate matching submodule is further configured to determine the pulse overlap rate and the scan line overlap rate according to the following relationship:

[0021] Pulse overlap rate ;

[0022] Scan line overlap ;

[0023] and control and Equal, or satisfy | - |≤ε, where ε is 0.01-0.05; when = At that time, the scan line spacing L satisfies L=v / f.

[0024] In a preferred implementation, the scanning path generation submodule is further configured to generate a vertical cross-scanning path. The vertical cross-scanning path includes first completing a layer of parallel line scanning along the X direction, and then completing a layer of parallel line scanning along the Y direction, with the scanning directions of the two layers being perpendicular to each other. Odd-numbered layers are X-direction scanning layers, and even-numbered layers are Y-direction scanning layers. For areas outside the coverage range of the high-speed galvanometer scanning head, the central processing unit controls the motion control unit to perform step stitching.

[0025] In a preferred implementation, the feedback decision module further includes a target cumulative removal depth generation unit, a basic scan group number calculation unit, and a round scan increment configuration unit. The target cumulative removal depth generation unit is configured to calculate the target cumulative removal depth based on the initial maximum height difference. Generate target cumulative removal depth And satisfying =k· Where the value of k ranges from 1.51 to 2.0; the basic scan group number calculation unit is configured to calculate the removal depth d of a single vertical cross-scan and the cumulative removal depth of the target. Calculate the number of basic scan groups And satisfy The incremental scanning configuration unit is configured to configure the number of scans ΔN to be performed in each round based on the working condition identification result. In the case of large margin removal, ΔN is 6-20 times; in the case of fine homogenization, ΔN is 1-5 times.

[0026] In a preferred implementation, the termination output module is further configured to output a stop processing command when any of the following conditions are met: the deterioration counter is greater than or equal to 2 and the current cumulative removal depth is greater than or equal to 0.9 times the target cumulative removal depth; 0 ≤ β < δ and the current cumulative removal depth is greater than or equal to 0.9 times the target cumulative removal depth, provided that the current round is in a non-deterioration branch; or 0 ≤ β < δ is met in two consecutive rounds in a non-deterioration branch.

[0027] In a preferred implementation, each round of feedback detection corresponds to the same measurement area A. The central processing unit registers the corresponding area between the detection position and the processing position based on the fixed bias and the calibration mapping relationship. The cumulative removal depth and the current arithmetic mean roughness are both obtained based on the registration results of the same measurement area A.

[0028] In a preferred implementation, the preset working condition identification threshold is further defined as follows: When the initial maximum height difference is greater than When the state recognition module determines that the current surface to be processed is in a large excess material removal condition; when the initial maximum height difference is less than or equal to At that time, the state recognition module determines that the surface to be processed is in a fine homogenization state.

[0029] In a preferred embodiment, the ultrafast laser processing unit further includes an ultrafast laser, a beam expander, a first high-reflectivity mirror, a second high-reflectivity mirror, a high-speed galvanometer scanning head, and a large-aperture flat-field focusing mirror. The pulsed laser beam output by the ultrafast laser passes sequentially through the beam expander, the second high-reflectivity mirror, the first high-reflectivity mirror, the high-speed galvanometer scanning head, and the large-aperture flat-field focusing mirror before irradiating the surface of the metal workpiece.

[0030] The beneficial effects of this application are:

[0031] First, the ultrafast laser processing system for metal workpieces based on surface quality feedback of this application integrates the ultrafast laser processing unit, online detection unit, motion control unit, and central processing unit onto the same frame assembly. The central processing unit coordinates the data acquisition module, state recognition module, parameter calculation module, scanning strategy module, feedback decision module, and termination output module. This enables the system to classify and identify the initial surface state of the workpiece based on the initial maximum height difference and initial arithmetic mean roughness obtained from online detection. Furthermore, it links the target cumulative removal depth, the number of basic scan groups, the initial planned total scan volume, and the number of additional scans per round, thereby overcoming the shortcomings of existing systems in terms of adaptability to different initial surface states and the simplistic scanning strategy. The problem is that, by establishing a correlation between the defocusing amount, the actual processing spot size, and the average output power of the laser through the parameter calculation module, a more reasonable energy input can be achieved under different defocusing conditions. This improves the problems of excessively high or low local energy, uneven removal, and poor surface consistency caused by the independent power output and defocusing state in the existing system. In addition, by using an online detection unit to perform round-by-round feedback detection on the same area, and combining the feedback decision module and the termination output module to implement closed-loop control, the processing process can be adjusted or a stop command can be output in a timely manner based on the cumulative removal depth and the current arithmetic mean roughness after processing. This improves the synergy between detection and processing, the stability of the processing process, and the consistency and adaptability of the surface finishing of the metal workpiece.

[0032] Secondly, in the preferred implementation, this application further establishes an overlap rate matching submodule and a scan path generation submodule in the scanning strategy module to create a coordinated matching relationship between the actual processing spot diameter D, laser repetition frequency, galvanometer scanning speed v, and scan line spacing L. This ensures that the pulse overlap rate and scan line overlap rate remain consistent or approximately consistent, thereby improving the uniformity of energy coverage along the scanning direction and transverse direction within a unit area and reducing strip residue, directional texture, and local under-ablation phenomena. Simultaneously, the scan path generation submodule employs a single vertical cross-scan method, making material removal in different directions more balanced. This facilitates the repeated energy reception of local peaks in different directions, thereby enhancing peak reduction and flattening effects, and improving surface removal uniformity and morphology convergence.

[0033] Third, in the preferred implementation, this application further optimizes the parameter calculation module according to the defocusing amount Δz, laser wavelength λ, and focal plane beam waist radius. The correspondence between the actual spot size and the actual spot size determines the spot size at the defocus position. Furthermore, the peak energy density at the center of the defocus position is made to satisfy the product of the material's single-pulse ablation threshold and the safety factor ξ. This allows for the synchronous correction of the laser's average output power when the defocus conditions change, ensuring that the energy input remains within a reasonable range that is slightly higher than the ablation threshold and required for stable ablation. This not only helps to avoid local deep pits, heat accumulation, and expansion of the heat-affected zone caused by excessive energy density, but also helps to avoid insufficient removal and unstable processing caused by insufficient energy density.

[0034] Fourth, in the preferred implementation, this application further enhances the system by setting a target cumulative removal depth generation unit, a basic scan group number calculation unit, and a round scan increment configuration unit in the feedback decision module, enabling the system to determine the initial maximum height difference based on the target cumulative removal depth generation unit, a basic scan group number calculation unit, and a round scan increment configuration unit. Predetermine the target cumulative removal depth to match the surface allowance. Furthermore, the number of basic scan groups is obtained by combining the depth d removed by a single vertical cross-scan. This allows the overall processing scale to have a clear depth basis and planning boundary, avoiding the underprocessing or processing redundancy caused by existing systems relying solely on experience to set the number of processing times. At the same time, the system differentiates the number of additional scans ΔN in each round based on the working condition identification results. Under the condition of large excess removal, a larger round scan increment is used to improve peak reduction efficiency, while under the condition of fine homogenization, a smaller round scan increment is used to shorten the detection feedback cycle. This allows the processing process to take into account both the early removal efficiency and the later surface convergence accuracy, reducing invalid scans and overprocessing.

[0035] Fifth, in the preferred implementation, this application further distinguishes between "continuous quality degradation after reaching the target depth" and "insufficient improvement or continuous stagnation in the non-deterioration state" by having the termination output module determine termination based on the deterioration counter, the relative change rate of surface roughness β, and the relationship between the current cumulative removal depth and the target cumulative removal depth. This allows processing to be stopped in time when the surface quality begins to deteriorate in the opposite direction or the marginal benefit of continuing processing decreases significantly, avoiding the problems of invalid scanning, over-removal, and surface re-roughening caused by relying solely on a fixed number of processing times or a single termination condition in existing systems. At the same time, the central processing unit retains the processing state corresponding to the historical best roughness as the final processing result when processing is terminated, so that the final output of the system no longer simply depends on the last round of detection value, but can stably select the processing state with the best surface quality throughout the entire processing process.

[0036] Sixth, in the preferred implementation, this application further ensures that each round of feedback detection corresponds to the same measurement area A, and the central processing unit registers the corresponding areas between the detection position and the processing position based on a fixed bias and calibration mapping relationship. This ensures that the cumulative removal depth and the current arithmetic mean roughness are always based on the comparison of the same surface area, effectively avoiding feedback distortion and judgment bias caused by detection area drift, area misalignment, or differences in the morphology of different areas.

[0037] Seventh, in a preferred implementation, this application further identifies a threshold by setting a preset operating condition. The status recognition module identifies the initial maximum height difference and the preset working condition threshold. The comparison results distinguish the surfaces to be processed into two categories: large-scale removal and fine homogenization. This allows the system to classify different initial surface states before processing begins, avoiding the inadequacy of existing systems that use a uniform processing mode for workpieces with significant surface undulations.

[0038] Eighth, in a preferred embodiment, the ultrafast laser processing unit of this application further includes an ultrafast laser, a beam expander, a first high-reflectivity mirror, a second high-reflectivity mirror, a high-speed galvanometer scanning head, and a large-aperture flat-field focusing mirror. The pulsed laser beam passes through the aforementioned optical components in sequence and then irradiates the surface of the metal workpiece, thereby forming a complete and stable laser transmission, reflection, scanning, and focusing optical path within the system. This not only helps to modulate the laser beam output from the ultrafast laser into an effective beam suitable for high-speed scanning of the galvanometer and surface finishing, improving the stability of laser energy transmission and the controllability of the processing area coverage, but also facilitates the precise focusing and rapid two-dimensional scanning of the laser beam on the workpiece surface. Attached Figure Description

[0039] Figure 1 This is a simplified structural diagram of the ultrafast laser processing system for metal workpieces based on surface quality feedback according to the present invention;

[0040] Figure 2 This is a functional block diagram of the ultrafast laser processing system for metal workpieces based on surface quality feedback according to the present invention. Detailed Implementation

[0041] To enable those skilled in the art to better understand the technical solutions of this application, the following will provide a more detailed description of this application in conjunction with the accompanying drawings and embodiments.

[0042] The directional terms such as above, below, left, right, front, and back used in this application are based on the positional relationships shown in the attached drawings. Different attached drawings may result in different positional relationships, therefore they should not be interpreted as limitations on the scope of protection.

[0043] In this application, the terms "installation," "connection," "interlocking," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, an integral connection, a mechanical connection, an electrical connection, or a connection that allows communication between components. They can also refer to a direct connection or an indirect connection through an intermediate medium. They can refer to the internal connection of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in this application can be understood according to the specific circumstances.

[0044] Example

[0045] As per the instruction manual Figure 1 and Figure 2 As shown, this invention provides an ultrafast laser processing system for metal workpieces based on surface quality feedback, used for ultrafast laser finishing of the surface of metal workpiece 11, particularly suitable for reducing surface roughness, smoothing peaks and valleys, and improving surface quality of additively manufactured metal workpieces such as SLM. This ultrafast laser processing system for metal workpieces uses the initial maximum height difference of the surface to be processed as the basis for processing. and initial arithmetic mean roughness As the initial state input, the material single-pulse ablation threshold Actual processing spot size at the defocus position, pulse overlap rate overlap with scan lines As a process control parameter, and based on the cumulative removal depth during the processing. Current arithmetic mean roughness Historical best roughness Deterioration counter The system incorporates the relative change rate of surface roughness β as a feedback variable, thus forming an adaptive closed-loop machining system oriented towards surface quality convergence. This system not only adaptively switches machining strategies and parameters based on different initial surface states, but also achieves station reuse, coordinate reuse, and optical path reuse for the same workpiece surface area through the coordinated arrangement of the online detection unit and the ultrafast laser processing unit on the same frame, the fixed offset switching between the detection and processing positions, and the calibration mapping between the detection coordinate system and the processing coordinate system. This reduces reference drift caused by repeated workpiece disassembly and relocation, and improves the consistency between machining and inspection.

[0046] In this embodiment, the ultrafast laser processing system for metal workpieces based on surface quality feedback includes a frame assembly, an ultrafast laser processing unit, an online detection unit, a motion control unit, a central processing unit, a human-machine interaction unit, and a dust removal and fume extraction unit.

[0047] The frame assembly includes a base, columns, and crossbeams. A workpiece clamping area and a processing area are located above the base. The columns are fixed to both sides of the top surface of the base, and the crossbeams connect to the upper ends of the two columns. The ultrafast laser processing unit, online inspection unit, and dust removal and fume extraction unit are mounted on the crossbeams or the mounting plates connected to the crossbeams. The motion control unit is mounted on the top surface of the base. An electrical control cabinet is located on the side of the frame assembly, and the central processing unit and human-machine interface unit are located within the electrical control cabinet on the side of the frame assembly. Preferably, the online inspection unit and the ultrafast laser processing unit are mounted on the same frame, sharing the same mechanical coordinate reference and workpiece clamping reference, but the inspection position and the processing position have a pre-calibrated fixed offset distance along the X / Y direction. The central processing unit controls the motion control unit to switch between the inspection position and the processing position according to this fixed offset, so that the same metal workpiece 11 can sequentially complete initial inspection, laser processing, and cycle feedback inspection without being unclamped. With the above arrangement, the processing optical path, the detection optical path and the workpiece motion trajectory are coordinated around the same workstation system, thereby realizing optical path reuse and adaptive processing control for the same processing object.

[0048] The ultrafast laser processing unit includes an ultrafast laser 1, a beam expander 2, a first high-reflectivity mirror 3, a second high-reflectivity mirror 4, a high-speed galvanometer scanning head 5, and a large-aperture flat-field focusing mirror 6. The ultrafast laser 1 is fixedly mounted on the upper part of the crossbeam of the frame assembly. The beam expander 2 is connected to the output end of the ultrafast laser 1. The second high-reflectivity mirror 4 and the first high-reflectivity mirror 3 are arranged sequentially along the output direction. The high-speed galvanometer scanning head 5 is located downstream of the reflected light path of the first high-reflectivity mirror 3. The large-aperture flat-field focusing mirror 6 is mounted on the output side of the high-speed galvanometer scanning head 5. The pulsed laser beam output from the ultrafast laser 1 passes sequentially through the beam expander 2, the second high-reflectivity mirror 4, the first high-reflectivity mirror 3, the high-speed galvanometer scanning head 5, and the large-aperture flat-field focusing mirror 6 before irradiating the surface of the metal workpiece 11 fixed in the fixture 7. The beam expander 2 is configured to expand and collimate the laser beam until it matches the entrance aperture of the high-speed galvanometer scanning head 5. The high-speed galvanometer scanning head 5 is configured to form a two-dimensional scanning trajectory under the control of the central processing unit. The large-aperture flat-field focusing lens 6 is configured to focus the scanned laser beam onto the workpiece machining surface. The ultrafast laser 1 is preferably configured as a femtosecond or picosecond laser with an output wavelength of 1030nm, 515nm, or 343nm and a repetition frequency that is adjustable from 100kHz to 10MHz. Through the above optical path arrangement, the system can adaptively adjust the average output power, repetition frequency, scanning speed, and defocusing amount according to different working conditions under the control of the central processing unit, achieving stable finishing under different surface conditions.

[0049] The motion control unit is a precision linear displacement module 8, which includes an X-axis linear module, a Y-axis linear module, a Z-axis linear module, and a fixture 7. The Y-axis linear module is mounted on the moving part of the X-axis linear module, and the Z-axis linear module is mounted on the moving part of the Y-axis linear module. The fixture 7 is fixedly connected to the output end of the Z-axis linear module, and the metal workpiece 11 is mounted on the fixture 7. The X-axis and Y-axis linear modules are configured to perform in-plane translation of the workpiece to switch between the detection position and the processing position, and between different processing areas. The Z-axis linear module is configured to adjust the axial distance between the workpiece surface and the theoretical focal plane of the large-diameter flat-field focusing lens 6 to set the processing defocus amount Δz. To improve positioning accuracy, the X-axis, Y-axis, and Z-axis linear modules are all equipped with position feedback encoders, and the position feedback signals are input to the central processing unit.

[0050] The online inspection unit is configured to share the same frame and workpiece clamping reference with the ultrafast laser processing unit, and completes initial surface morphology acquisition and feedback inspection after each round of processing while the metal workpiece 11 is not unclamped. The online inspection unit preferably employs a three-dimensional surface profilometer 9, which can be a white light interferometric surface profilometer, a laser confocal surface profilometer, or a surface roughness measuring instrument with three-dimensional surface morphology acquisition capabilities. The central processing unit pre-establishes the calibration mapping relationship between the detection coordinate system of the three-dimensional surface profilometer 9 and the scanning coordinate system of the high-speed galvanometer scanning head 5, and, combined with the fixed offset between the detection position and the processing position, achieves accurate switching and area correspondence between the metal workpiece 11 and the detection position. Preferably, each round of inspection corresponds to the aforementioned measurement area A, and the cumulative removal depth... The comparison with the roughness parameter is based on the registration results of the same region. Therefore, although the online detection unit and the ultrafast laser processing unit have fixed offsets between the detection position and the processing position in space, they establish coordinate multiplexing and optical path multiplexing relationships around the same processing area in the control logic, thereby ensuring the consistency between the feedback quantity and the processing area.

[0051] In this embodiment, the 3D surface profilometer 9 performs two functions: First, before formal processing, it performs a surface scan on the measurement area A of the surface to be processed to obtain the height data of each sampling point relative to the reference surface and inputs it into the central processing unit, which then calculates the initial maximum height difference of the surface to be processed. and initial arithmetic mean roughness Secondly, after each processing cycle ends and is paused, the current processing area is re-inspected to obtain the current cumulative removal depth. and the current arithmetic mean roughness Compared with existing split-type inspection solutions, this embodiment unifies the initial morphology measurement and processing feedback measurement into the same online inspection link. At the same time, it completes the correspondence between the processing position and the inspection position through fixed offset switching and coordinate mapping, avoiding repeated disassembly and relocation of the workpiece between the inspection position and the processing position, thereby reducing reference drift and improving the reliability of closed-loop control.

[0052] Before entering the formal processing flow, it is preferable to pre-treat the surface of the metal workpiece 11. Pre-treatment may include: placing the metal workpiece 11 in an ultrasonic cleaner and sequentially ultrasonically cleaning it with acetone, anhydrous ethanol, and deionized water to remove oil, powder residue, cutting fluid residue, oxide deposits, and other loose contaminants from the workpiece surface; the ultrasonic cleaning time for each cleaning solution is preferably 3-10 minutes. After cleaning, the workpiece 11 is dried using clean compressed air, nitrogen, or constant-temperature hot air to keep the workpiece surface clean and dry. After pre-treatment, the metal workpiece 11 is fixed on the fixture 7 and moved to the detection position of the three-dimensional surface profilometer 9 by the precision linear displacement module 8. The three-dimensional surface profilometer 9 performs a surface scan of the measurement area A of the surface to be processed, acquiring the height data of each sampling point within the measurement area relative to the reference average surface. The central processing unit receives the sampling data output by the three-dimensional surface profilometer 9 and reconstructs the sampling data to obtain the three-dimensional height matrix z(i,j) of each sampling point within the measurement area A. Based on this height matrix, the central processing unit calculates the initial maximum height difference of the surface to be processed. and initial arithmetic mean roughness .in, The vertical height difference between the highest peak and the lowest valley point among all sampling points within the same measurement area A is used to reflect the peak-valley extreme value span within the measurement area. It is the area average of the absolute height of each sampling point in measurement area A relative to the reference average surface, used to reflect the average amplitude of the overall microscopic undulations of the surface.

[0053] The central processing unit preferably adopts a main control computer 10, which is equipped with at least a data acquisition module, a status recognition module, a parameter calculation module, a scanning strategy module, a feedback decision module, and a termination output module.

[0054] The data acquisition module is connected to the 3D surface profilometer 9 and is used to receive the measurement data output by the 3D surface profilometer 9 and reconstruct the measurement data into a 3D height matrix z(i,j) within the measurement area. The data acquisition module calculates the initial maximum height difference based on the height matrix. and initial arithmetic mean roughness Simultaneously, after each processing cycle, based on the detection results of the same area that is registered one-to-one with the current processing area, the current cumulative removal depth is calculated. and the current arithmetic mean roughness Therefore, the data acquisition module constitutes the raw data entry point, regional registration entry point, and state quantity update entry point in the entire closed-loop control link.

[0055] The status recognition module is used to receive the output of the data acquisition module. and will Identification threshold for preset working conditions Comparison. When > When the current surface to be processed is determined to be in a condition of large-scale material removal; when ≤ At that time, it is determined that the surface to be processed is in a fine and homogeneous state. For typical laser additive manufacturing of metal parts, it can be taken as follows: =80μm was used as the statistical reference threshold; at this time, when When >80μm, it usually corresponds to >8μm indicates a large peak-to-valley difference and numerous sharp peaks and deep valleys on the surface. The system identifies this as a large margin removal condition, and the subsequent laser removal strategy focuses on improving single-round removal efficiency and rapid peak reduction; when When ≤80μm, it usually corresponds to A depth of ≤8μm indicates that the surface is in a state of low peak-to-valley difference. The system identifies this as a fine homogenization condition, and subsequent processing strategies focus on suppressing overcutting and improving surface convergence and uniformity. The result of this condition identification will be used simultaneously to guide the setting of the cumulative removal depth of the target, the calculation of the number of basic scan groups, the setting of the initial planned total number of scans, the configuration of the number of new scans per round, the setting of average power, and the termination logic judgment, thereby reflecting the adaptive processing characteristics of this system for different initial morphological states.

[0056] The parameter calculation module is used to calculate the single-pulse ablation threshold of the material. Laser wavelength λ, focal plane beam waist radius Preset defocus amount Δz and laser repetition frequency Calculate the actual effective spot radius ω(Δz), the actual processed spot diameter D, and the average output power of the laser at the defocus position. Single-pulse ablation threshold This refers to the minimum single-pulse energy density at which a material surface just begins to undergo stable ablation under given laser wavelength, pulse width, and incident conditions. It can be obtained through single-pulse ablation experiments combined with the Gaussian beam threshold fitting method. Then, the main control computer 10 calculates the actual spot size at the defocus position based on the relationship between the defocus amount Δz and the Gaussian beam transmission, and then introduces a single-pulse energy density safety factor ξ to ensure that the center peak energy density at the defocus position meets the requirements. =ξ· Thus, the average output power is obtained. Preferably, the value of ξ ranges from 1.05 to 1.35, and the value of the defocusing amount Δz preferably ranges from 0.1 mm to 6.0 mm. Through the above calculations, the peak energy density at the defocusing position can be maintained at a level slightly higher than the single-pulse ablation threshold of the material, thereby ensuring stable ablation while avoiding excessively high focal energy that could lead to local deep pits, heat accumulation, and expansion of the heat-affected zone.

[0057] Preferably, the parameter calculation module has the following built-in calculation relationship:

[0058] ;

[0059] D = 2ω(Δz);

[0060] .

[0061] The main control computer 10 calculates the target average output power based on the above relationship. Afterwards, The parameters are sent to the ultrafast laser 1 for setting, and the surface of the metal workpiece 11 is moved to the corresponding defocus position through the Z-axis displacement of the precision linear displacement module 8. In this way, the system links and controls the material threshold, defocus amount, spot size and laser power, instead of using a fixed power for rough processing, thereby further demonstrating the adaptive processing capability for different workpiece surface conditions and different processing stages.

[0062] The scanning strategy module includes an overlap rate matching submodule and a scanning path generation submodule. The overlap rate matching submodule receives the actual processing spot diameter D and the laser repetition frequency. The scanning speed v and the line spacing L of the galvanometer are calculated, and the pulse overlap rate is calculated respectively. overlap with scan lines Among them, pulse overlap rate This refers to the degree of overlap along the scanning direction of adjacent working spots formed by two consecutive laser pulses on the workpiece surface within the same scan line; scan line overlap rate. This refers to the degree of overlap between adjacent active areas of two adjacent scan lines perpendicular to the scanning direction. They can be expressed as follows:

[0063] Pulse overlap rate ;

[0064] Scan line overlap .

[0065] In a preferred embodiment, the main control computer 10 controls the pulse overlap rate. overlap with scan line They are equal or approximately equal, that is, they satisfy the condition that they are equal or approximately equal. = or | - |≤ε, where ε is preferably 0.01-0.05. When further controlled to... = When the scan line spacing L is preferably satisfied, L = v / Through the above matching control, a pulse spatial coverage density that is approximately consistent along the scanning direction and the lateral direction can be obtained within a unit area, thereby reducing stripe residue, directional texture, and local under-ablation phenomena.

[0066] The scanning path generation submodule generates the actual processing path based on the working condition identification results, power calculation results, and overlap rate matching results. Preferably, the basic processing unit generated by this submodule is a single vertical cross-scan. A single vertical cross-scan refers to completing one layer of parallel line scanning along the X direction, followed by another layer of parallel line scanning along the Y direction, with the scanning directions of the two layers perpendicular to each other. Preferably, odd-numbered layers are X-direction scanning layers, and even-numbered layers are Y-direction scanning layers; after each unidirectional scan, the workpiece coordinate system or galvanometer scanning coordinate system is rotated 90° to ensure that the scanning direction of the subsequent layer is strictly perpendicular to the previous layer. Compared with single-direction parallel scanning, alternating vertical cross-scanning in the X and Y directions can make material removal in different directions more balanced, reduce the directional texture formed on the surface due to repeated ablation in a single direction, and allow local peaks to repeatedly receive energy in different directions, thereby improving peak reduction and flattening effects; in controlling... and Under equal or approximately equal conditions, the uniformity of removal and the convergence of surface morphology can be further improved. For areas outside the coverage of the galvanometer, the main control computer 10 further controls the precision linear displacement module 8 to complete step splicing to achieve continuous and uniform coverage of the surface to be processed.

[0067] The feedback decision module is used to perform closed-loop control based on initial morphology parameters, working condition identification results, and feedback measurement results after each processing cycle. This feedback decision module includes at least a target cumulative removal depth generation unit, a basic scan group number calculation unit, a cycle scan increment configuration unit, a state initialization unit, a state variable update unit, a convergence judgment unit, and a stall judgment unit.

[0068] Among them, the target cumulative removal depth generation unit is based on the initial maximum height difference. Generate target cumulative removal depth Preferably, satisfy =k· The safety factor k ranges from 1.51 to 2.0. This setting is because if the subsequent cumulative removal depth is less than or equal to... However, there may still be areas where deep valleys are not completely covered, uneven material removal space, or measurement errors leading to residual surface undulations; therefore, this invention does not simply set the target cumulative removal depth as equal to... Equal, but in A safety margin is reserved on the basis to improve the stability and adequacy of macro-level smoothing. However, when k < 1.5, the safety margin is insufficient, making it difficult to completely eliminate deep valley defects; when k > 2.0, it significantly increases processing time and the risk of heat accumulation. Therefore, By limiting the removal to between 1.5 and 2.0 times the initial maximum height difference, a balance can be struck between sufficient removal and processing economy.

[0069] The basic scan group calculation unit calculates the number of scan groups based on the removal depth d of a single vertical cross-scan and the cumulative removal depth of the target. Calculate the number of basic scan groups Preferred to satisfy =ceil( / d). Here, the removal depth d in a single vertical cross-scan refers to the depth removed within a given actual processing spot diameter D and laser repetition frequency. Scanning speed v, scan line spacing L, pulse overlap rate Scan line overlap rate and average output power Under the given conditions, the average material removal depth is the result of one vertical cross-scan of the surface to be processed.

[0070] It should be noted that 'd' is neither the depth of a single pulse pit nor the depth of a single-layer scan in a single direction. Instead, it refers to the average normal removal amount corresponding to an orthogonal double-layer scan consisting of single-layer scans in the X and Y directions, which serves as the basic processing unit. This removal depth 'd' can be obtained through preliminary experiments. Specifically, a sample with the same material or similar surface condition as the final workpiece is selected, and a single vertical cross-scan is performed under the same processing conditions. Then, a 3D surface profilometer 9 is used to compare the height of the same measurement area before and after processing, and the average removal depth is calculated and averaged at multiple locations. (Number of basic scan groups) The target cumulative removal depth is achieved under ideal average removal conditions. The theoretical minimum number of vertical cross-scans required. The up-rounding function ceil(·) is used to ensure that the theoretical cumulative removal amount is not less than the target cumulative removal depth, thereby avoiding insufficient theoretical processing amount due to simple rounding or down-rounding.

[0071] Furthermore, this embodiment will As the basis for setting the initial planned total scan volume, that is, according to the central processing unit... Establish an initial planned total scan reference for the entire processing process. This reference is used to define the initial overall processing scale, assist in forming a cumulative plan for each round, and provide an initial boundary reference for subsequent closed-loop adjustments. In other words, It not only serves as a theoretical scan scale parameter, but also acts as an input for the initial planned scan volume at the system level.

[0072] The round-scan incremental configuration unit is used to dynamically configure the number of scans ΔN to be performed in each round based on the working condition identification result. If the working condition identification result is a large allowance removal working condition, it indicates that the initial maximum height difference of the surface to be processed is... When the surface has significant peaks and valleys and a large amount of material remaining, it is preferable to set ΔN to a larger value to improve single-round processing efficiency, accelerate the removal speed of surface peaks, and approach the target cumulative removal depth as quickly as possible. In a preferred embodiment, ΔN can be set to 6-20 times, more preferably 10 times, under the condition of large material removal. If the condition identification result is a fine homogenization condition, it indicates that the initial maximum height difference of the surface to be processed is large. When the surface is relatively small and close to being smooth, it is preferable to set ΔN to a smaller value so that the system can pause detection after a short processing step, promptly identifying the improvement, stagnation, or deterioration trend of surface roughness, and avoiding excessive local removal caused by too many scans in a single round. In a preferred embodiment, ΔN can be set to 1-5 times under fine homogenization conditions, more preferably 2 times. Therefore, the present invention does not uniformly apply a fixed processing step to all surfaces to be processed, but rather adaptively switches the processing strategy based on the degree of surface undulation represented by the initial maximum height difference.

[0073] The state initialization unit is used to establish an initial baseline for subsequent round-by-round comparisons and feedback updates before formally entering the first round of processing. Specifically, the main control computer 10 performs the following initialization operation: [Instructions for setting the base scan group number are missing from the original text]. As a reference value for the initial planned total scan count; the total number of planned scans in the first round is set to [value]. (1) = ΔN; the initial arithmetic mean roughness is... Set as the initial reference roughness = Initialize the degradation counter =0; initialize the historical best roughness to 0. = A roughness convergence threshold δ is set, preferably 3%-5%. It should be noted that in this embodiment, (n) is defined as the total number of planned scans at the end of round n, while the number of newly executed vertical cross scans in the current round n is ΔN. During round 1, the system executes new scans according to ΔN, bringing the total number of planned scans to [value missing]. (1) = ΔN; If entering the next round, the total number of planned scans will be calculated as follows: (n+1)= (n)+ΔN update. Based on the above definition, (n) reflects the cumulative planned progress, and ΔN reflects the new step size in a single round. The two have a clear division of labor.

[0074] After initialization, the system enters the nth round of processing, where n ≥ 1. In the nth round of processing, the main control computer 10 controls the ultrafast laser 1, the high-speed galvanometer scanning head 5, and the precision linear displacement module 8 to perform ΔN additional vertical cross-scans according to the aforementioned scanning path, so that the cumulative planned number of scans reaches [number missing]. (n). After completing this round of scanning, the system pauses processing and enters a phased inspection. The 3D surface profilometer 9 measures the surface morphology of the current processing area, and the main control computer 10 obtains the current cumulative removal depth. and the current arithmetic mean roughness . This represents the total normal removal amount of the workpiece surface relative to the initial reference plane from the start of processing to the end of the nth round; This indicates the average surface roughness within measurement area A under the current processing condition.

[0075] The state variable update unit obtains the state variable update value in the nth round. and Then, the current state is evaluated in two dimensions: firstly, by comparison. Compared with the previous effective reference roughness saved before entering this round of judgment The size relationship is used to determine whether the surface quality is improving, remaining the same, or deteriorating; secondly, comparison... Cumulative removal depth with target The magnitude of these values ​​is used to determine whether the current processing has reached or is close to the target depth stage. Here, Preferably, the state variable update unit updates the state variable before entering the nth round of judgment. The data is temporarily stored in the middle, that is... = . This indicates the previous valid comparison benchmark retained before entering this round of judgment: in round 1, = In the nth round (n≥2), This represents the baseline roughness retained after the (n-1)th round of judgment. When the previous round deteriorates and the system adopts a conservative update strategy, It can maintain the baseline value before the deterioration, therefore It is not mechanically equivalent to the roughness measured in the (n-1)th round, but rather equivalent to the roughness of the previous effective reference. Specifically, when the roughness is detected... > This indicates that the surface roughness after the current wheel machining is greater than the previous effective reference value, meaning that the surface quality is showing a deteriorating trend. At this point, the system further adjusts the current cumulative removal depth... Has the target cumulative removal depth been reached or is close to being achieved? Perform a branch check. If the condition is met... ≥0.9 This indicates that the current removal depth has reached or is close to the predetermined target, but the roughness has begun to deteriorate. In this case, the deterioration is considered a true quality degradation after reaching the target depth, and the deterioration counter is incremented and updated accordingly. = +1; if <0.9 If the roughness increases in this round, it indicates that the processing depth has not yet reached the predetermined target. Even if an increase in roughness is detected in this round, it is considered a process fluctuation in the stage where the depth has not reached the target, and the deterioration counter remains unchanged. Furthermore, in this deterioration branch, the system does not use β as a convergence criterion for termination judgment. Preferably, β in this round can be recorded as invalid or not updated, thereby avoiding cross-conflicts caused by "roughness deterioration" and "insufficient improvement" sharing the same criterion. At this time, to prevent abnormal fluctuations in a single round from amplifying the error in the trend judgment of the next round, Ideally, it should remain unchanged, or it should be updated to max(in a conservative manner). , ).

[0076] When detected ≤ When the current surface roughness is not worsened compared to the previous effective reference value, it indicates that the surface roughness is in an improved or stable state. In this case, the system determines that the current machining state is healthy and performs a positive update on the state variables: first, the deterioration counter is reset to zero. =0; secondly, set the current roughness to 0. Compared with historical best roughness If a comparison is made, < Then update = Next, in this non-deterioration branch, the relative change rate β of surface roughness is calculated, preferably satisfying:

[0077] β=| - | / .

[0078] Finally, the current measured roughness is used as the benchmark value for the next round of judgment, i.e., the roughness is updated. = Through the above design, the system can continuously record the historical best roughness during the processing, while ensuring that the calculation of β is always based on the previous value. Complete, then execute. Update, thereby eliminating the conflict between the calculation order of β and... Conflicts between update orders.

[0079] The convergence judgment unit is used to determine whether the surface quality has entered the convergence stage when the current processing wheel is in a non-deteriorating branch, based on the relationship between β and the preset roughness convergence threshold δ. When the value of β is large, it indicates that the surface roughness has decreased significantly after this processing wheel, and the surface quality is still effectively improving; when the value of β is small, it indicates that the roughness improvement brought about by this processing wheel is limited, and the surface quality tends to converge or stagnate. In this embodiment, the convergence threshold for surface roughness change is denoted as δ, preferably with a value range of 0.03-0.05. When 0≤β<δ, it indicates that the surface roughness improvement brought about by the current processing wheel is less than the preset minimum effective improvement threshold, and the processing effect can be considered to be tending to converge or entering a stagnant stage; when β≥δ, it indicates that the current processing wheel can still bring about effective improvement.

[0080] The stagnation detection unit is used to identify whether the processing has entered a continuous ineffective improvement stage. Preferably, the stagnation detection unit can determine whether the system has entered a stagnation period by recording two consecutive rounds in a non-deteriorating branch where both satisfy 0 ≤ β < δ. That is, the system only uses β for convergence and stagnation judgment when the roughness has not deteriorated; while when the roughness deteriorates, β is used for convergence and stagnation judgment. Degradation protection is performed. By separating "degradation" and "stagnation" into two parallel but non-overlapping judgment paths, the system's termination logic becomes clearer.

[0081] The termination output module is used to determine the deterioration counter. Surface roughness relative change rate β and current cumulative removal depth Cumulative removal depth with target Based on the comparison results, a control command is output to stop or continue processing. Preferably, the termination output module controls the ultrafast laser 1 to stop output and end the current processing when any of the following conditions are met: First, ≥2 and ≥0.9 This indicates that the current processing has entered a stage of continuous quality degradation after reaching or approaching the target depth; secondly, assuming the current wheel is in a non-deteriorating branch, 0≤β<δ and ≥0.9 This indicates that the current cumulative removal depth has reached or is close to the target depth, and the improvement in surface roughness is less than the preset convergence threshold; thirdly, 0 ≤ β < δ is satisfied in two consecutive rounds under the non-deterioration branch, indicating that although the theoretical depth target may not have been fully reached, the processing has entered an ineffective stagnation period, and the marginal benefit of continuing processing is extremely low. If none of the above stopping conditions are met, the main control computer 10 updates the total number of cumulative planned scans for the next round according to the cumulative plan logic, i.e. (n+1)= (n)+ΔN, and return to the scan path generation submodule to continue generating the next round of vertical cross scan trajectory, thus entering the next round of closed loop iteration.

[0082] In this embodiment, the main control computer 10 can also be connected to a human-machine interface unit to allow operators to input material type, laser wavelength, pulse width, defocus range, safety factor ξ, convergence threshold δ, and process template recall information, and display the initial maximum height difference. Initial arithmetic mean roughness Current cumulative removal depth Current arithmetic mean roughness Target cumulative removal depth Current round number, current cumulative planned total number of scans (n), number of basic scan groups ΔN, number of new scans per round, and degradation counter Information such as alarm status.

[0083] The dust removal and smoke extraction unit includes a dust collection hood, a suction pipe, a filter assembly, and a fan assembly. The dust collection hood is positioned above or to the side of the processing area and faces the processing point. One end of the suction pipe is connected to the dust collection hood, and the other end is connected to the filter assembly and the fan assembly in sequence.

[0084] Preferably, the dust removal and fume extraction unit is communicatively connected to the central processing unit, receiving start / stop control commands from the central processing unit and being able to synchronize with the output status of the ultrafast laser 1. When the system enters the laser processing stage, the central processing unit synchronously controls the dust removal and fume extraction unit to start; when the system pauses detection, terminates processing, or an alarm occurs, the central processing unit controls the dust removal and fume extraction unit to shut down after a delay or switch its working state according to a preset program. The dust removal and fume extraction unit operates continuously during laser processing, allowing processing fumes to enter the suction pipe along the dust collection hood and be filtered by the filter assembly before being discharged, thereby maintaining the cleanliness and stability of the detection optical path and the processing area.

[0085] The working process of the above system is as follows: First, after cleaning and drying the surface of the metal workpiece 11, it is fixed on the fixture 7; then, the precision linear displacement module 8 moves the metal workpiece 11 to the detection position of the three-dimensional surface profiler 9, and the three-dimensional surface profiler 9 acquires the height data of all sampling points in the measurement area A; the main control computer 10 reconstructs the detection data and calculates... and The process involves several steps: identifying the working condition, generating the cumulative target removal depth, calculating the number of basic scan groups, setting the initial planned total number of scans, configuring the number of new scans per round, calculating the spot size at the defocus position, calculating the average output power, matching the overlap rate, and generating the scan path. Next, the main control computer 10 controls the precision linear displacement module 8 to move the metal workpiece 11 from the detection position to the processing position based on the fixed offset between the detection position and the processing position. It also controls the ultrafast laser 1, the high-speed galvanometer scanning head 5, and the precision linear displacement module 8 to perform the first round of ΔN vertical cross-scans, bringing the cumulative planned total number of scans to [number missing]. (1); After reaching the set scanning increment for this round, the system pauses laser output, and then the precision linear displacement module 8 returns the workpiece to the detection position. The three-dimensional surface profiler 9 re-acquires the current surface morphology, and the main control computer 10 calculates based on coordinate calibration mapping and registration in the same area. and And update accordingly. , , and β; subsequently, the output module terminates according to , β and Determine whether the termination conditions are met. If the termination conditions are met, then retrieve... The corresponding optimal state is taken as the final result, and the system is stopped; if the stopping condition is not met, then proceed as follows: (n+1)= (n)+ΔN updates the total planned scan count for the next round and enters the next processing round. This cycle continues until the system issues a stop command. Therefore, it is evident that the system, through condition identification, parameter linkage, round-incremental control, and feedback detection within the same area, forms a truly adaptive processing closed loop, rather than an open-loop processing with fixed parameters and a fixed total scan count.

[0086] In one specific embodiment, when the object to be processed is an additively manufactured stainless steel workpiece with large surface peaks and valleys, the three-dimensional surface profilometer 9 detects... For thicknesses >80μm, the main control computer 10 identifies it as a large allowance removal condition and configures a larger ΔN to improve single-round removal efficiency and peak reduction speed. During this stage, the system reduces intermediate pauses, improving processing efficiency in the large allowance stage; when... Approaching step by step and As the improvement rate decreases, the main control computer 10 begins to increase its sensitivity to the deterioration count and convergence threshold, thereby preventing further processing from causing reverse surface deterioration. Conversely, when the initial surface of the object to be processed is already relatively smooth, i.e. When the surface area is no larger than 80μm, the system enters a fine homogenization process. The main control computer 10 is configured with a smaller ΔN, enabling the system to repeat detection and judgment at shorter intervals, thereby suppressing overcutting and improving the final surface convergence accuracy. Thus, workpieces with different initial surface states can undergo adaptive processing on the same hardware system through different control parameter sets, different scan increments, and switching between detection and processing at the same workstation.

[0087] In summary, this invention is not a simple conventional assembly of an ultrafast laser, galvanometer, displacement platform, and surface profilometer. Instead, it constructs a comprehensive ultrafast laser processing system for metal workpieces, centered around surface quality feedback. This system integrates initial morphology detection, working condition identification, threshold power calculation, overlap rate matching, vertical cross-scanning, fixed offset switching between detection and processing positions, calibration and mapping of the detection and scanning coordinate systems, dual feedback of depth and roughness, continuous degradation protection, convergence stagnation identification, and optimal state output. The key lies in using the initial maximum height difference... As the core input for condition identification and target cumulative removal depth setting, the material single-pulse ablation threshold is used. The actual spot size at the out-of-focus position and the safety factor ξ are linked to set the average output power. With pulse overlap rate overlap with scan lines Matching control improves energy distribution uniformity, and alternating vertical cross-scanning in the X and Y directions enhances peak clipping and flattening effects. Establish an initial plan to scan the total volume as a reference, and use ΔN to establish a round-by-round incremental execution strategy. (n) represents the cumulative planned progress and is expressed as the cumulative removal depth. Current arithmetic mean roughness Historical best roughness Deterioration counter A surface quality convergence control model is established by combining the relative roughness change rate β calculated only under the non-deterioration branch; at the same time, by reusing the workstation, coordinates and optical paths of the online detection unit and the ultrafast laser processing unit under the same frame, stable, adaptive and highly consistent control of the surface quality convergence process of metal workpieces is achieved.

[0088] The above descriptions are merely embodiments of this application, and common knowledge regarding specific structures and characteristics in the solutions is not described in detail here. It will be apparent to those skilled in the art that this application is not limited to the details of the above exemplary embodiments, and that this application can be implemented in other specific forms without departing from the spirit or essential characteristics of this application. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of this application is defined by the appended claims rather than the foregoing description. Therefore, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within this application. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A system for ultrafast laser processing of metal workpieces based on surface quality feedback, characterized in that, The system includes a frame assembly, an ultrafast laser processing unit, an online inspection unit, a motion control unit, and a central processing unit. The frame assembly includes a base, a column, and a crossbeam. The motion control unit is mounted on the base and has a fixture for fixing the metal workpiece. The ultrafast laser processing unit and the online inspection unit are mounted on the crossbeam and located above the fixture. The processing optical path of the ultrafast laser processing unit faces the surface of the metal workpiece, and the detection end of the online inspection unit faces the surface of the metal workpiece to be processed. There is a pre-calibrated fixed offset between the detection position and the processing position. The central processing unit is located on the side of the frame assembly and is communicatively connected to the ultrafast laser processing unit, the online inspection unit, and the motion control unit. It is used to control the motion control unit to switch the metal workpiece between the detection position and the processing position according to the fixed offset, so that the detection area corresponds to the processing area. The central processing unit includes at least a data acquisition module, a status recognition module, a parameter calculation module, a scanning strategy module, a feedback decision module, and a termination output module. The data acquisition module is connected to the online detection unit and is configured to receive measurement data output by the online detection unit, reconstruct the measurement data into a three-dimensional height matrix z(i,j) within the measurement area, calculate the initial maximum height difference and initial arithmetic mean roughness of the surface to be processed based on the three-dimensional height matrix, and calculate the current cumulative removal depth and current arithmetic mean roughness based on the detection results of the same area registered one-to-one with the current processing area after each round of processing. The state recognition module is connected to the data acquisition module and is configured to receive the initial maximum height difference and compare the maximum height difference with a preset working condition recognition threshold to determine whether the current surface to be processed is in a large amount removal working condition or a fine homogenization working condition. The working condition identification results are used to determine the target cumulative removal depth setting, the basic scan group number calculation, the initial planned total scan volume setting, the configuration of the number of new scans per round, the average power setting, and the termination logic judgment; The parameter calculation module is connected to the state recognition module, motion control unit, and ultrafast laser processing unit, and is configured to calculate parameters based on the material's single-pulse ablation threshold, laser wavelength λ, and focal plane beam waist radius.

1. Preset the defocusing amount Δz and the laser repetition frequency, and calculate the actual effective spot radius ω(Δz), the actual processed spot diameter D, and the average output power of the laser at the defocusing position; The scanning strategy module is connected to the state recognition module and the parameter calculation module, and is configured to generate the actual processing path based on the working condition recognition result, power calculation result, and overlap rate matching result, and control the ultrafast laser processing unit to perform scanning processing; the feedback decision module is connected to the data acquisition module, the state recognition module, and the scanning strategy module, and is configured to perform closed-loop control based on the initial morphology parameters, the working condition recognition result, and the feedback measurement result after each processing cycle; the termination output module is connected to the feedback decision module, and is also connected to the ultrafast laser processing unit and the motion control unit, and is configured to output a control command to stop processing or continue processing based on the deterioration counter, the relative change rate of surface roughness β, and the comparison result between the current cumulative removal depth and the target cumulative removal depth.

2. The ultrafast laser processing system for metal workpieces based on surface quality feedback according to claim 1, characterized in that, The scanning strategy module includes an overlap rate matching submodule and a scanning path generation submodule. The overlap rate matching submodule is configured to determine the pulse overlap rate and the scanning line overlap rate based on the actual processing spot diameter D, the laser repetition frequency, the galvanometer scanning speed v, and the scanning line spacing L. The scanning path generation submodule is configured to generate the actual processing path based on the working condition identification result, the power calculation result, and the overlap rate matching result.

3. The ultrafast laser processing system for metal workpieces based on surface quality feedback according to claim 1, characterized in that, The parameter calculation module is configured to determine the actual effective spot size and the average output power of the laser at the defocus position according to the following relationship: ； D = 2ω(Δz); And make the peak energy density at the defocus position satisfy the product of the material's single-pulse ablation threshold and the safety factor ξ, wherein the safety factor ξ ranges from 1.05 to 1.

35.

4. The ultrafast laser processing system for metal workpieces based on surface quality feedback according to claim 1, characterized in that, The overlap rate matching submodule is configured to determine the pulse overlap rate and scan line overlap rate according to the following relationship: Pulse overlap rate ; Scan line overlap ; and control and Equal, or satisfy | - |≤ε, where ε is 0.01-0.05; when = At that time, the scan line spacing L satisfies L=v / f.

5. The ultrafast laser processing system for metal workpieces based on surface quality feedback according to claim 2, characterized in that, The scanning path generation submodule is configured to generate a vertical cross-scanning path. The vertical cross-scanning path includes first completing a layer of parallel line scanning along the X direction, and then completing a layer of parallel line scanning along the Y direction, with the scanning directions of the two layers being perpendicular to each other. Odd-numbered layers are X-direction scanning layers, and even-numbered layers are Y-direction scanning layers. For areas outside the coverage range of the high-speed galvanometer scanning head, the central processing unit controls the motion control unit to perform step stitching.

6. The ultrafast laser processing system for metal workpieces based on surface quality feedback according to claim 2, characterized in that, The feedback decision module includes a target cumulative removal depth generation unit, a basic scan group number calculation unit, and a round scan increment configuration unit. The target cumulative removal depth generation unit is configured to calculate the target cumulative removal depth based on the initial maximum height difference. Generate target cumulative removal depth And satisfy =k· Where the value of k ranges from 1.51 to 2.0; the basic scan group number calculation unit is configured to calculate the removal depth d of a single vertical cross-scan and the cumulative removal depth of the target. Calculate the number of basic scan groups And satisfy The incremental scanning configuration unit is configured to configure the number of scans ΔN to be performed in each round based on the working condition identification result. In the case of large margin removal, ΔN is 6-20 times; in the case of fine homogenization, ΔN is 1-5 times.

7. The ultrafast laser processing system for metal workpieces based on surface quality feedback according to claim 2, characterized in that, The termination output module is configured to output a stop processing command when any of the following conditions are met: the deterioration counter is ≥2 and the current cumulative removal depth is ≥0.9 times the target cumulative removal depth; 0≤β<δ and the current cumulative removal depth is ≥0.9 times the target cumulative removal depth when the current round is in a non-deterioration branch; or 0≤β<δ is met for two consecutive rounds in a non-deterioration branch.

8. The ultrafast laser processing system for metal workpieces based on surface quality feedback according to claim 1, characterized in that, Each round of feedback detection corresponds to the same measurement area A. The central processing unit registers the corresponding area between the detection position and the processing position based on the fixed bias and the calibration mapping relationship. The cumulative removal depth and the current arithmetic mean roughness are both obtained based on the registration results of the same measurement area A.

9. The ultrafast laser processing system for metal workpieces based on surface quality feedback according to claim 1, characterized in that, The preset working condition identification threshold is: When the initial maximum height difference is greater than When the state recognition module determines that the surface to be processed is in a large amount of material removal condition; When the initial maximum height difference is less than or equal to At that time, the state recognition module determines that the surface to be processed is in a fine homogenization state.

10. The ultrafast laser processing system for metal workpieces based on surface quality feedback according to claim 1, characterized in that, The ultrafast laser processing unit includes an ultrafast laser, a beam expander, a first high-reflectivity mirror, a second high-reflectivity mirror, a high-speed galvanometer scanning head, and a large-aperture flat-field focusing mirror. The pulsed laser beam output by the ultrafast laser passes sequentially through the beam expander, the second high-reflectivity mirror, the first high-reflectivity mirror, the high-speed galvanometer scanning head, and the large-aperture flat-field focusing mirror before irradiating the surface of the metal workpiece.

Images