Intelligent closed-loop repair method and system for complex casting defects based on vision and additive manufacturing

Abstract

The application discloses a kind of complex casting defect intelligent closed-loop repair method and system based on vision and additive manufacturing, belong to intelligent manufacturing and special welding technical field.The method first obtains the three-dimensional point cloud data of the repaired area and reconstructs grid model, extracts the depth of defect and minimum opening feature size to execute adaptive classification decision;When triggering deep cavity defect threshold, the system is based on boundary inflation and Boolean difference algorithm, automatically generate the three-dimensional model of multi-chamber conformal gas protection tooling that is consistent with the topology of inner wall of defect, and drive 3D printing equipment to execute manufacturing;While introducing asynchronous double-track scheduling mechanism separates offline tooling preparation and online repair process;Finally, conformal tooling is physically assembled to deep cavity, and under the protection of multi-chamber gas curtain, arc additive filling is executed, and process parameters such as interlayer heat input are collected for closed-loop feedback optimization.The application breaks through the end-to-end unmanned link from unstructured defect data to digital tooling automatic generation to accurate additive repair, significantly improves the repair automation rate and metallurgical quality of complex heterogeneous deep cavity.

Description

Technical Field

[0001] This invention relates to the intersection of intelligent equipment manufacturing, additive manufacturing, and special welding technology. Specifically, it relates to an intelligent closed-loop repair method and system based on vision and additive manufacturing for confined space defects in deep cavities of complex precision castings. Background Technology

[0002] During the casting process of complex thin-walled precision castings (such as high-performance duplex steel pump bodies and valve bodies), shrinkage cavities and cracks inevitably occur in deep cavities or complex flow channels due to the limitations of the alloy's liquid shrinkage and solidification characteristics. Currently, the industry mainly uses tungsten inert gas welding (TIG) or electric arc additive manufacturing (WAAM) technologies to repair these defects through patching and filling.

[0003] In existing conventional repair processes, defect identification and repair path planning primarily rely on manual labor. Defect locations are typically roughly identified using non-destructive testing (NDT), followed by manual grinding and then welding repairs by experienced operators. While some existing technologies have incorporated 3D vision scanning for robotic welding trajectory planning in recent years, for deep-cavity confined defects in complex castings, current vision methods largely focus on extracting the surface macroscopic contours, lacking quantitative assessment and adaptive classification mechanisms for repair strategies based on the underlying topological features of the defect (such as maximum depth and minimum opening size). This leads to the subsequent additive filling often using fixed templates, making it difficult to balance the cladding efficiency of open areas with the accessibility and interference avoidance requirements of confined deep cavities.

[0004] Furthermore, for materials such as duplex stainless steel, which are extremely sensitive to thermal history and protective atmosphere, a strict local inert gas protective environment must be established when performing arc additive repair in deep cavities or narrow slots to suppress the precipitation of harmful phases and prevent weld oxidation. Current technologies typically employ standard argon arc welding nozzles with attached shrouds, or manually fabricated simple baffles (such as high-temperature tape or fixed copper gaskets) for local protection. However, the geometric morphology and spatial location of actual casting defects exhibit high randomness and non-standardization. Standardized protective fixtures cannot achieve a tight fit with the complex curved surfaces surrounding the defects. This geometric mismatch leads to easy jetting or short-circuiting of the protective gas at the gaps, making it difficult to maintain a stable laminar gas curtain at the bottom of the deep cavity. Customizing metal protective covers for each non-standard defect using traditional machining methods results in long processing cycles and high costs, completely failing to meet the timeliness requirements of continuous on-site repair.

[0005] On the other hand, existing arc additive repair systems generally employ open-loop control based on preset static parameters (such as constant welding current and constant interlayer waiting time). When filling metal layer by layer in confined spaces with extremely poor heat dissipation, such as deep cavities, the local interlayer temperature fluctuates drastically as heat accumulates. The open-loop control mode cannot detect the thermal evolution during the forming process and the actual cladding size deviation, which can easily lead to uncontrolled local heat input, resulting in metallurgical defects such as coarse grains and an imbalance in the ratio of ferrite to austenite, severely weakening the final mechanical properties and corrosion resistance of the repaired area.

[0006] In summary, existing casting defect repair technologies suffer from significant process gaps and technical limitations in areas such as adaptive classification of unstructured defect features, rapid response and precise forming of conformal protective tooling, and thermodynamic dynamic closed-loop feedback in additive manufacturing processes. Therefore, there is an urgent need in this field for an intelligent closed-loop method and system that deeply integrates 3D visual perception, additive manufacturing processes, and automatic control to address the digital gaps and quality consistency issues in the repair of defects in complex and confined spaces. Summary of the Invention

[0007] To address the shortcomings of existing technologies in repairing defects in confined spaces of complex castings, such as insufficient quantification of topological features, geometric mismatch of protective tooling, and unstable metallurgical quality caused by open-loop heat input control, the present invention aims to provide an intelligent closed-loop repair method and system for complex casting defects based on vision and additive manufacturing.

[0008] To achieve the above objectives, the present invention provides the following technical solution: On the one hand, this invention provides an intelligent closed-loop repair method for defects in complex castings based on vision and additive manufacturing, comprising the following steps: The three-dimensional point cloud data of the area to be repaired is acquired, and after denoising and registration, an STL three-dimensional mesh model representing the morphology of the defect is reconstructed. The maximum depth feature value and minimum opening feature size of the defect are extracted from the STL three-dimensional mesh model, and an adaptive classification judgment is made by calling a dynamic threshold model that is limited by the mechanical shape interference limit and the upper limit of material process heat input. When a deep cavity confined defect is identified, based on the inner wall topology of the STL three-dimensional mesh model, the boundary expansion and Boolean difference algorithm are applied to automatically generate an inner shell profile with inner and outer surfaces fitting together. Virtual baffles and fluid uniformity lattice mesh with a porosity of 30% to 50% are automatically inserted into this profile, and the slice file of the multi-chamber conformal gas protection tooling is output. The dual-track asynchronous scheduling mechanism is triggered to send the sliced ​​file to the 3D printing equipment for offline manufacturing; The physical assembly of the solidified multi-chamber conformal gas protection fixture is placed onto the area to be repaired. Under the isolation and protection of the multi-chamber gas curtain, the arc additive manufacturing is carried out layer by layer filling repair. Simultaneously, the interlayer heat input and cladding size deviation data are collected to perform closed-loop feedback and dynamically adjust the additive manufacturing process parameters of subsequent layers.

[0009] Furthermore, the adaptive classification determination includes: when the maximum depth feature value is not greater than the depth threshold and the minimum opening feature size is not less than the opening threshold, it is determined to be an open defect and the conventional spatial path is directly called for additive repair; otherwise, it is determined to be a deep cavity confined defect and the conformal tooling automatic generation process is triggered.

[0010] Furthermore, the dual-track asynchronous scheduling mechanism includes: during the offline preparation track for conformal tooling printing and manufacturing, the robot on the online execution track continuously performs routine repair tasks on castings that have not triggered deep cavity confined defects; after the conformal tooling is materialized, the corresponding casting is guided back to the online execution track to perform arc additive repair and closed-loop feedback steps.

[0011] On the other hand, the present invention provides an intelligent closed-loop repair system for complex casting defects that implements the above-mentioned repair method, comprising: The 3D vision acquisition module is configured to scan the casting to be repaired and output three-dimensional point cloud data; The edge computing and control hub has a built-in 3D modeling engine and process decision algorithm library, and is configured to perform defect feature extraction, adaptive classification and judgment and automatic generation of conformal tooling. A heterogeneous additive manufacturing cluster, comprising at least one metal selective laser melting device and at least one photopolymer ceramic 3D printing device, is configured to receive slice files and allocate printing paths based on estimated peak temperatures to perform solidification manufacturing; The robotic additive manufacturing execution terminal has a built-in repair control closed loop and is configured to perform arc additive filling with the assistance of a multi-chamber conformal gas protection tooling. It can also dynamically switch between multiple cladding modes based on instructions from the edge computing and control center.

[0012] Compared with the prior art, the present invention has the following beneficial effects: (1) An adaptive classification of defect repair strategies based on multidimensional geometric features was achieved. This invention abandons the traditional experience-based judgment that relies on manual visual inspection. By reconstructing a three-dimensional mesh model and extracting the "depth-opening" two-dimensional features, a dynamic threshold model is constructed by combining mechanical interference and thermal input. This mechanism transforms non-standardized casting defects into calculable quantitative indicators, automatically separating open defects from deep cavity confined defects. While ensuring the accessibility of the additive manufacturing execution terminal space and interference avoidance, it effectively avoids over-design and waste of computational resources.

[0013] (2) An end-to-end automated generation link for conformal protective tooling was constructed, eliminating geometric topological mismatches in the protective flow field. For the irregular inner walls of deep cavities in complex castings, this invention achieves 100% holographic topological fit between the tooling shell and the defective inner wall through Boolean subtraction, thereby eliminating the local gas leakage gaps caused by standardized protective covers from the root. At the same time, by automatically dividing multiple chambers and embedding a flow-equalizing lattice grid through an algorithm, the traditional time-consuming and labor-intensive machining molding process is transformed into a digital flow of "data input equals output slices", which greatly improves the response speed and flow field uniformity of the special gas protective tooling.

[0014] (3) A dual-track asynchronous scheduling and heterogeneous machine group collaboration mechanism was established, taking into account both customized gas supply requirements and overall production cycle time. Addressing the issue of process waiting caused by the long manufacturing cycle of 3D printed tooling, this invention introduces an asynchronous scheduling logic that separates "offline generation preparation" and "online additive manufacturing execution," ensuring the continuous operation rate of high-value robot workstations. Simultaneously, process allocation between metal printing and ceramic printing is performed based on the estimated peak temperature, achieving an optimal balance between manufacturing economy and reliability in ultra-high temperature operations.

[0015] (4) Closed-loop control of the thermodynamic state in the additive manufacturing process was achieved, ensuring the metallurgical consistency of special alloy materials. Addressing the issues of heat accumulation and phase imbalance that easily occur when repairing materials such as duplex stainless steel in confined spaces, the robot execution terminal of this invention synchronously incorporates feedback monitoring of interlayer heat input and dimensional deviations. Through dynamic adaptive switching between modes such as pulsed cold metal transition, the interlayer temperature and heat input are strictly constrained within the metallurgical safety range, thereby effectively suppressing the precipitation of harmful phases and grain coarsening, and improving the comprehensive mechanical properties and corrosion resistance service life of the repaired components. Attached Figure Description

[0016] Figure 1 is an interactive diagram of the overall physical and data architecture of the intelligent closed-loop repair system for complex casting defects provided in an embodiment of the present invention; Figure 2 is a flowchart of the decision tree algorithm for three-level defect feature extraction and adaptive classification judgment provided in an embodiment of the present invention; Figure 3 A topological operation geometric logic evolution diagram of the automatic CAD generation process of the multi-chamber conformal gas protection tooling provided in the embodiments of the present invention; Figure 4 is a schematic diagram of the process flow and timing control state machine of the dual-track asynchronous scheduling mechanism provided in the embodiment of the present invention; Figure 5 is a schematic diagram of the closed-loop repair section and process feedback data flow of arc additive manufacturing under the assembly state of the solidified conformal tooling. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the following embodiments are only for explaining the invention and are not intended to limit the scope of protection of the invention. Where there is no conflict, the embodiments and features in this application can be combined with each other.

[0018] Example 1: Physical and Computational Architecture of Intelligent Closed-Loop Repair System As shown in Figure 1, this embodiment provides an intelligent closed-loop repair system for defects in complex castings. Its system architecture comprises three main layers: data acquisition, edge computing, and manufacturing execution. In the data perception layer, a 3D vision acquisition module (preferably a combination of a blue light laser scanner with strong anti-reflective properties and a binocular stereo vision sensor) is deployed, configured to acquire high-precision three-dimensional point cloud data of the casting to be repaired.

[0019] At the computational decision-making layer, an edge computing and control hub is deployed. This hub serves as the system's "computational brain," housing a 3D modeling engine (such as a geometry processing engine based on the Open CASCADE kernel) and a process decision algorithm library. It is configured to perform parallel processing of point cloud registration, defect feature extraction, Boolean topology operations, and scheduling instruction distribution.

[0020] At the physical execution layer, a heterogeneous additive manufacturing cluster and a robotic additive execution terminal are deployed. The heterogeneous additive manufacturing cluster includes a selective laser melting (SLM) machine for printing stainless steel tooling and a photopolymerization ceramic 3D printing (SLA / DLP) machine for printing ultra-high temperature resistant ceramic tooling; the robotic additive execution terminal integrates a pulsed TIG welder, a CMT (cold metal transfer) welder, and an infrared thermal imaging thermometer.

[0021] Example 2: Defect Multidimensional Feature Extraction and Adaptive Classification Algorithm To eliminate the blind spots of traditional manual evaluation, this system runs an adaptive classification and judgment algorithm within the edge computing and control center (as shown in Figure 2).

[0022] Sub-step S11: Perform statistical filtering and ICP registration on the acquired 3D point cloud data to reconstruct and generate a high-fidelity STL 3D mesh model.

[0023] Sub-step S12: Call the feature recognition operator to automatically extract the maximum depth feature value H and the minimum opening feature size Dmin of the defect in the STL 3D mesh model.

[0024] Sub-step S13: The system invokes a preset dynamic threshold model. The depth threshold Hth and opening threshold Dth in the threshold model are not fixed constants, but are dynamically calculated based on the outer diameter size dgun of the additive gun body of the robot additive execution terminal, the effective extension length Lext and the upper limit of process heat input of the given material (for example, usually set Dth = dgun + Δδ, where Δδ is the safety interference redundancy; Hth ≤ Lext).

[0025] Sub-step S14 (Classification Routing): Execution logic judgment: IF(H ≤ Hth AND Dmin ≥ Dth), then this area is determined as an open defect. The system directly issues slicing instructions, invokes conventional spatial free paths to perform additive filling, and does not trigger tooling generation, so as to avoid waste of computing power and materials.

[0026] Execution logic judgment: IF(H>Hth OR Dmin<Dth), then this area is determined as a deep-cavity restricted defect. Since the gun body cannot probe in or is very prone to interference and collision, the system forcibly triggers the "conformal tooling automatic generation process" and the "asynchronous dual-track scheduling mechanism".

[0027] Example 3: Boolean algorithm for automatic CAD generation of conformal protective tooling When the processing flow for deep-cavity restricted defects is triggered, to ensure that the gas protection flow field and the inner wall of the non-standard defect achieve 100% topological fitting, the edge calculation and control center perform the geometric topological operation shown in Figure 3: Topology extraction and expansion: The system automatically extracts the healthy inner wall profile data of the casting within a preset distance outside the defect boundary (e.g., 20mm expansion outward), and uses it as the protection reference surface.

[0028] Boolean offset and difference operation: Offset the protection reference surface along the normal direction into the inner cavity space by a specific distance (e.g., 3mm-5mm) to form an outer shell curved surface; then, perform a Boolean subtraction operation between the three-dimensional solid enveloped by the outer shell curved surface and the STL model of the inner wall of the original casting. The result of this operation is: all overlapping space portions are removed, and an inner shell profile with an outer surface being a regular geometry and an inner surface 100% mirror-fitting the rough topology of the inner wall of the defect is automatically generated.

[0029] Implantation of multi-chamber and lattice structure: Along the direction of the bottom-layer additive repair path planned by the system, a half-height virtual partition is automatically inserted inside the generated inner shell profile, cutting the protection space into a main protection chamber and a tail pre-blowing chamber; at the same time, at the air inlet section of each chamber, a BCC (Body-Centered Cubic) or Gyroid three-dimensional lattice grid with a porosity of 30% to 50% is automatically generated by invoking the lattice algorithm, which serves as the fluid uniform flow lattice grid.

[0030] Finally, the system outputs the merged multi-chamber conformal gas protection tooling as a standard slice file (such as CLI or SLC format). This data-driven design process reduces the traditional machining and molding time of several days to seconds.

[0031] Example 4: Cluster Allocation and Asynchronous Dual-Track Scheduling Based on Dynamic Prediction To resolve the timeliness conflict between the 3D printing manufacturing cycle and the production line cycle time, this embodiment introduces a dual-track asynchronous scheduling mechanism controlled by a state machine (as shown in Figure 4).

[0032] Heterogeneous cluster dynamic allocation: Before issuing slice files, the edge computing and control center allocates resources based on the volume of the defect 3D model and... The characteristics of the material to be repaired are used to estimate the peak heat accumulation temperature during the additive filling process. When the estimated peak temperature is lower than the softening critical threshold of stainless steel, the scheduling instruction is issued to the metal selective laser melting equipment; if there is an extremely severe heat dissipation bottleneck that causes the estimated peak temperature to exceed the limit, the task is assigned to the photopolymer ceramic 3D printing equipment to manufacture ultra-high temperature resistant ceramic conformal tooling.

[0033] Dual-track asynchronous scheduling: The system establishes an offline preparation track and an online execution track, operating on two threads. Castings currently triggering deep-cavity confined defects are logically marked as "suspended" by the system and transferred to a physical buffer station to await printing completion (offline preparation track). During this period, the robot additive manufacturing terminal is not idle but is scheduled to continuously perform routine filling on "open-cavity defect castings" in the buffer queue that have not triggered deep-cavity alarms (online execution track). Once the multi-chamber conformal gas protection fixture has been printed and inspected by the heterogeneous machine group, the system changes the status label of the originally suspended casting to "ready," guiding it back to the online execution track. This asynchronous scheduling mechanism maximizes the OEE (Overall Equipment Effectiveness) of high-value robot workstations.

[0034] Example 5: Additive Execution with Assembly Feedback and Thermodynamic Closed-Loop Control As shown in Figure 5, the physical assembly of the multi-chamber conformal gas protection fixture is clamped onto the confined cavity above the defect. At this point, the inner surface of the fixture fits perfectly with the bottom wall of the casting, forming a sealed local micro-protective chamber.

[0035] Multimodal additive manufacturing with closed-loop heat input: An Ar-He-N2 ternary mixed protective gas is supplied internally via a tooling system. For duplex steel, the robotic additive manufacturing terminal 400 executes a trajectory of "contour offset + zone filling". During the additive stacking process, an infrared thermal imaging thermometer collects the interlayer temperature (Tlayer) in real time.

[0036] The control center operates a thermodynamic closed-loop feedback algorithm, strictly constraining the interlayer temperature within a metallurgically safe range of 60°C to 100°C. When the system detects that the Tlayer is approaching the 100°C upper limit, the edge computing center dynamically issues intervention commands, automatically switching the robot's welding mode from the high-heat-input CMT mode to the low-heat-input pulse TIG precision cladding mode, or forcibly inserting an adaptive interlayer cooling waiting time. This closed-loop control strategy fundamentally suppresses the austenite phase loss and ferrite grain coarsening problems that are prone to occur in duplex steel under deep cavity confined heat dissipation conditions, ensuring a high degree of consistency in the mechanical properties of the repaired components.

[0037] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent closed-loop repair of defects in complex castings based on vision and additive manufacturing, characterized in that, The process includes the following steps: Defect scanning and mesh reconstruction step: acquiring 3D point cloud data of the area to be repaired, and denoising and registering the 3D point cloud data to reconstruct an STL 3D mesh model representing the defect morphology; Feature extraction and adaptive classification step: extracting the maximum depth feature value and minimum opening feature size of the defect from the STL 3D mesh model; The system calls the preset dynamic threshold model to perform adaptive classification and judgment. When the judgment result is a deep cavity confined defect, the automatic generation process of conformal tooling is triggered. Automatic generation steps for conformal tooling: Based on the inner wall topology of the STL 3D mesh model, the boundary expansion and Boolean difference algorithms are applied to automatically generate a 3D CAD model of a multi-chamber conformal gas protection tooling that matches the inner wall topology, and output slice files; Asynchronous manufacturing and physical assembly steps: The sliced ​​file is sent to the 3D printing equipment to perform offline manufacturing; After manufacturing is completed, the solidified multi-chamber conformal gas protection fixture is physically assembled onto the area to be repaired; Arc additive repair and closed-loop feedback steps: Inert protective gas is delivered through the assembled multi-chamber conformal gas protection fixture; Under the isolation and protection of the multi-chamber gas curtain, the robot is driven to perform arc additive layer-by-layer filling repair; Simultaneously collect interlayer heat input and cladding size deviation data, and perform closed-loop feedback optimization to dynamically adjust the additive manufacturing process parameters of subsequent layers.

2. The intelligent closed-loop repair method for complex casting defects based on vision and additive manufacturing according to claim 1, characterized in that, In the feature extraction and adaptive classification steps, the dynamic threshold model is jointly constrained and defined by the mechanical shape interference limit of the additive manufacturing terminal and the upper limit of the process thermal input of the repair material. The adaptive classification decision includes: When the maximum depth feature value is not greater than the depth threshold and the minimum opening feature size is not less than the opening threshold, it is determined to be an open defect, and the conventional spatial path is directly called for additive repair. When the maximum depth feature value is greater than the depth threshold, or the minimum opening feature size is less than the opening threshold, it is determined to be a deep cavity confined defect.

3. The intelligent closed-loop repair method for complex casting defects based on vision and additive manufacturing according to claim 1, characterized in that, The specific execution algorithm for the automatic generation step of the conformal tooling includes: Extract the surface data of the area around the defect within a preset range in the STL three-dimensional mesh model to construct a protection reference surface; The protective reference plane is offset along the normal direction to form the outer shell surface, and the outer shell surface and the inner wall model of the area to be repaired are subjected to Boolean subtraction operation to generate an inner shell surface with 100% fit between the inner and outer surfaces; Along the additive repair path planned by the system, virtual partitions are automatically inserted into the inner shell profile to divide a multi-chamber structure that includes at least a pre-blowing chamber and a main protective chamber. A three-dimensional lattice mesh with a porosity of 30% to 50% is automatically generated at the fluid inlet of each chamber as a fluid flow equalization structure, and the slice file is output after being merged with the multi-chamber structure.

4. The intelligent closed-loop repair method for complex casting defects based on vision and additive manufacturing according to claim 1, characterized in that, The method also includes a dual-track asynchronous scheduling mechanism, specifically comprising: When the automatic tooling generation process is triggered, the system will assign the repair task corresponding to the current casting to be repaired to the offline preparation track and transfer the casting to be repaired to the buffer station to wait. During the manufacturing tasks of the 3D printing equipment performed on the offline preparation track, the robot on the online execution track continuously performs routine repair tasks on other castings that have not triggered deep cavity confinement defects; Once the materialized multi-chamber conformal gas protection tooling has been manufactured and verified, the system guides the corresponding casting to be repaired back to the online execution track, triggering the arc additive repair and closed-loop feedback steps.

5. A complex casting defect intelligent closed-loop repair system implementing the method of any one of claims 1 to 4, characterized in that, include: The 3D vision acquisition module is configured to scan the casting to be repaired and output the three-dimensional point cloud data; The edge computing and control center is communicatively connected to the 3D vision acquisition module, and has a built-in 3D modeling engine and process decision algorithm library. It is configured to execute the feature extraction and adaptive classification steps as well as the automatic generation step of the conformal tooling. The heterogeneous additive manufacturing cluster is communicatively connected to the edge computing and control center, and receives the slice file and executes the physical printing manufacturing of the multi-chamber conformal gas protection tooling. The robotic additive manufacturing terminal is configured to perform arc additive layer-by-layer filling and repair with the assistance of the solidified multi-chamber conformal gas protection tooling.

6. The intelligent closed-loop repair system for complex casting defects according to claim 5, characterized in that, The 3D vision acquisition module includes at least one of a line laser scanner, a structured light camera, or a binocular stereo vision sensor.

7. The intelligent closed-loop repair system for complex casting defects according to claim 5, characterized in that, The heterogeneous additive manufacturing fleet includes at least one metal selective laser melting device and at least one photopolymer ceramic 3D printing device; The edge computing and control hub is configured to automatically allocate printing paths based on the estimated peak temperature of the area to be repaired: when the estimated peak temperature is lower than the material critical threshold, the metal selective laser melting equipment is scheduled to print stainless steel conformal tooling; when the estimated peak temperature is higher than the material critical threshold, the photopolymer ceramic 3D printing equipment is scheduled to print ultra-high temperature resistant ceramic conformal tooling.

8. The intelligent closed-loop repair system for complex casting defects according to claim 5, characterized in that, The robotic additive manufacturing execution terminal has a built-in repair control closed loop for duplex steel materials, configured to strictly constrain the interlayer temperature of additive manufacturing within the range of 60°C to 100°C, and automatically switch between pulsed cold metal transition mode and fine tungsten inert gas welding cladding mode according to the instructions issued by the edge computing and control center.

Images