An online self-adaptive control method for arc additive manufacturing topography

Abstract

The present application belongs to the technical field of electric arc additive manufacturing, and discloses an online self-adaptive regulation method for the appearance of electric arc additive manufacturing. According to the deviation between the actual formed part appearance and the theoretical appearance of the part, the forming trajectory and forming parameters of the unformed part are dynamically adjusted. In this way, a layer closed-loop control mechanism is formed, which can reduce the forming deviation and improve the flatness of the deposited layer surface. The method can be widely applied in the field of electric arc additive manufacturing, effectively solving the problem of poor cumulative forming precision of electric arc additive manufacturing, improving the forming quality and surface flatness during additive manufacturing, ensuring the stability of the process, and realizing a more accurate and reliable manufacturing process.

Description

Technical Field

[0001] This invention belongs to the field of electric arc additive manufacturing technology, and relates to an online adaptive control method for the morphology of electric arc additive manufacturing. Background Technology

[0002] Additive manufacturing, commonly known as 3D printing, is a layer-by-layer discrete manufacturing technology based on digital 3D CAD models. It rapidly achieves the overall shaping of parts by adding materials in the form of powder, filaments, or sheets layer by layer. Direct energy deposition (DED) is a major additive manufacturing technology widely used in the rapid prototyping of medium to large-sized metal components. This technology utilizes electron beams, lasers, plasma arcs / electric arcs, etc., as heat sources, melting metal filaments at a certain speed and injecting them into a molten pool to achieve rapid additive manufacturing.

[0003] However, due to the large heat input and molten pool size of fused wire deposition technology, the height of a single layer is relatively high, resulting in lower precision, poor surface quality, and a higher likelihood of defects in the formed components (Shi X, Yan C, Feng W, et al. Effect of high layer thickness on surface quality and defect behavior of Ti-6Al-4V fabricated by selective laser melting[J]. Optics & Laser Technology, 2020, 132: 106471.). Furthermore, before plasma arc fused deposition forming, slicing and layering are required, with the layer thickness preset to a fixed value. However, during the deposition process, many influencing factors cause deviations between the deposition width and height of each layer and the theoretical design. As the deposition process progresses, these deviations gradually accumulate and amplify, eventually leading to the inability to perform the deposition process, or even the scrapping of the formed parts (Xiong J, Zhang Y, Pi Y. Control of deposition height in WAAM using visual inspection of previous and current layers[J]. Journal of Intelligent Manufacturing, 2021, 32(8):2209-2217.). Therefore, it is urgent to design an apparatus and method to ensure the stability, reliability and smooth progress of the deposition process.

[0004] Therefore, the present invention aims to provide a method for solving the above-mentioned problems, so as to achieve closed-loop control of surface morphology in the arc additive manufacturing process. By dynamically planning and controlling the forming parameters of the lower layer based on the upper layer morphology, the device and method of the present invention can reduce forming deviations, improve the surface smoothness of the deposited layer, and ultimately achieve a more accurate and reliable manufacturing process. Summary of the Invention

[0005] In the actual forming process of arc additive manufacturing, fluctuations in process parameters and changes in forming boundary conditions can lead to differences between the actual deposition morphology and dimensions and the theoretical preset values. Furthermore, in the layer-by-layer discrete deposition process, the morphology of the upper deposition layer directly affects the forming of the lower layer, causing these differences to accumulate and propagate continuously during the open-loop forming process, affecting the subsequent surface smoothness and forming quality.

[0006] To address the issue of poor forming accuracy in arc additive manufacturing technology, this invention aims to provide an online adaptive control method for arc additive manufacturing morphology. This method dynamically plans and controls the forming parameters of the lower layer based on the upper-layer morphology, effectively improving the surface quality and forming accuracy of arc additive formed parts. This effectively solves the problem of poor forming accuracy in existing arc additive manufacturing technologies.

[0007] The technical solution of the present invention:

[0008] A method for online adaptive control of morphology in arc additive manufacturing, comprising the following steps:

[0009] Step 1: Slice the 3D model of the part to be formed into layers, plan the forming path, and obtain the spatial trajectory and process parameters of the part deposition forming; import the planned spatial trajectory and process parameters into the CNC system and the arc forming system, and the CNC system drives the arc forming system to perform layer-by-layer additive forming on the substrate according to the planned forming path.

[0010] Step 2: During the arc additive manufacturing process, a morphology sensor is used to acquire the surface morphology information of the deposited layer online, and the results are transmitted to the computer control system;

[0011] Step 3: The computer control system analyzes the actual morphology information of the sedimentary layer and obtains the morphology deviation by comparing it with the expected morphology information;

[0012] Step 4: Based on the morphology difference results, dynamically plan and update the subsequent forming trajectory and forming process parameters to suppress surface morphology differences and improve forming quality and accuracy.

[0013] Furthermore, this online adaptive control method for the morphology of arc additive manufacturing is based on an online detection and adaptive control system for the morphology of arc additive manufacturing, which includes a CNC system, a motion platform, a computer control system, an arc additive manufacturing system, and a morphology detection system.

[0014] Furthermore, the CNC system drives the arc forming system or motion platform to form relative motion according to a preset forming trajectory. During the motion, the wire feeding mechanism of the arc forming system melts the metal wire and deposits it on the surface of the forming substrate, thereby obtaining a complete metal part of the desired shape.

[0015] Furthermore, the topography sensor is a structured light sensor, used for online monitoring of the deposition layer topography and feeding back the topography results to the computer control system in real time in the form of a three-dimensional point cloud. The three-dimensional point cloud needs to be preprocessed to remove noise points and segment the effective point cloud data of the part surface. The actual forming surface height corresponding to each point is calculated as the average height of all topography point clouds within a spatial range with a radius of R centered on the discrete point.

[0016] Furthermore, using only topographic deviation as the control input is insufficient to accurately reflect the dynamic characteristics of the control process, and the control process is prone to over-adjustment with fluctuating values. It is best to consider the topography of the first two layers, i.e., the variation in interlayer topographic deviation, to enable faster convergence of the control curve. Assuming the part to be formed consists of n deposition layers... This represents the expected height of the surface of the nth deposition layer at (x, y).

[0017] The actual deposition height obtained by measurement:

[0018] ce n (x,y)=e n (x,y)-e n-1 (x,y)

[0019] Among them, e n (x,y) represents the topographic height deviation at the current layer position (x,y), e n-1 (x,y) represents the topographic height deviation at position (x,y) in the previous layer of the current layer, ce n (x,y) represents the variation in topographic height between the two layers.

[0020] Furthermore, in step three, based on the morphological differences, the subsequent forming trajectory and forming process parameters are dynamically planned and updated using a fuzzy control method. The fuzzy control method employs a dual-input Mamdani-type two-dimensional fuzzy logic controller, which includes a fuzzy generator and a fuzzy canceller. The morphological height deviation e at the current layer (x,y) position is then calculated. n (x,y) and the change in topographic height deviation cen (x,y) is input into the fuzzy logic inference engine, and the trajectory adjustment amount and the corresponding process parameter adjustment amount at the (x,y) position are inferred according to the fuzzy rules.

[0021] The trajectory adjustment amount for the corresponding (x,y) position in the next layer is:

[0022] H infer (x,y)=H set (x,y)+e n (x,y)

[0023] Among them, H set (x,y) is the preset height of the next layer trajectory at (x,y), H infer (x,y) represents the inference height of the next layer trajectory at (x,y);

[0024] The process parameter is the travel speed, specifically the travel speed V at the (x,y) position in the next layer. infer (x,y) is:

[0025] V infer (x,y)=V set (x,y)+ΔV(x,y)

[0026] Among them, V set (x,y) is the preset deposition rate of the next deposition layer at (x,y); ΔV(x,y) is the walking speed adjustment amount obtained by logical reasoning.

[0027] Fuzzification is the process of converting precisely acquired physical quantities into fuzzy quantities that can be recognized by a fuzzy inference engine. The conversion mapping function between precise physical quantities and fuzzy quantities is called the membership function. The range of values ​​for precise physical quantities is the basic universe of discourse. According to process experiments, the commonly used range of walking speed change TS can be set to [0.4, 0.8] m / min, the range of interlayer weld height deviation e is [-3, 3] mm, and the range of height deviation change ce is [-2, 2] mm. That is, the universes of discourse for input and output quantities are TS∈[-0.2, 0.4], e∈[-3, 3], and ce∈[-2, 2], respectively.

[0028] In the fuzzy control method, the deviation 'e' of the forming height at various locations on the deposition layer surface is divided into 7 levels: negative large, negative medium, negative small, zero, positive small, positive medium, and positive large, with corresponding fuzzy subsets set as NB, NM, NS, ZO, PS, PM, and PB. The change in morphological height error 'ce' is divided into 5 levels: negative large, negative small, zero, positive small, and positive large, with corresponding fuzzy subsets set as NB, NS, ZO, PS, and PB. The change in output walking speed 'TS' is also divided into 5 levels: negative large, negative small, zero, positive small, and positive large, with corresponding fuzzy subsets set as NB, NS, ZO, PS, and PB. The specific value range of each fuzzy subset is based on the aforementioned basic universe of discourse. A membership function is used to map precise physical quantities to fuzzy quantities, thereby ensuring that the fuzzy control method can accurately adjust the walking speed at different deviation levels to optimize the forming quality of the deposition layer.

[0029] Furthermore, in the fuzzy control method, the fuzzy rules are designed based on experience and prior shaping and adjustment experience. The basic principle is to control the step size of fuzzy adjustment according to the deviation and the change of deviation. Specifically, when the deviation and the change of deviation deviate from the zero point in the same direction, a large adjustment step size in the opposite direction is selected to reduce the error as soon as possible; when the deviation and the change of deviation deviate from the zero point in opposite directions, the step size needs to be reduced for fine adjustment.

[0030] Furthermore, in step two, the acquisition of the surface morphology information of the deposition layer requires the calibration of the measurement coordinate system and the forming coordinate system before forming, and the determination of the transformation matrix between the two. Then, according to the calibration parameters, the measurement point cloud in the measurement coordinate system is transformed to the printing coordinate system and jointly analyzed and calculated with the pre-planned lower layer forming trajectory.

[0031] The beneficial effects of this invention are:

[0032] 1. This invention effectively reduces interlayer and cumulative deviations by real-time detection and analysis of the deposition morphology of each layer and timely adjustment of process parameters such as travel speed, thereby significantly improving the geometric accuracy and dimensional consistency of the parts.

[0033] 2. This invention utilizes fuzzy control methods, combined with multi-layer morphological deviation changes, to finely regulate the forming process, significantly improve surface smoothness, reduce surface defects, and enhance overall surface quality.

[0034] 3. This invention employs an online adaptive shape control method for arc additive manufacturing based on fuzzy control to control the shape of parts. By comparing with parts that have not undergone shape control, it is proven that this method can significantly improve the shape quality. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the process of the present invention.

[0036] Figure 2 This is a schematic diagram of the device structure of the present invention.

[0037] Figure 3 The single-channel straight-wall feature under interlayer control is shown in (a) for macroscopic morphology and (b) for layer-by-layer height distribution curve.

[0038] Figure 4 The first part is a single-channel straight-wall feature without interlayer control, where (a) is the macroscopic morphology and (b) is the layer-by-layer height distribution curve.

[0039] In the figure: 1-substrate; 2-soldering torch; 3-line structured light sensor; 4-deposited layer. Detailed Implementation

[0040] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings and technical solutions.

[0041] Please see Figure 1 and Figure 2 As shown, the online adaptive control method for the morphology of arc additive manufacturing of the present invention includes the following steps:

[0042] A method for online adaptive control of morphology applied in arc additive manufacturing can be achieved through the following steps:

[0043] Step 1: Import the 3D model of the part to be formed into the trajectory planning software to obtain the forming CNC code, and then import it into the CNC system.

[0044] The trajectory planning software slices the 3D model into layers to obtain the contours of each slice layer of the 3D model of the metal part. Then, it fills the contours of each layer with trajectories to obtain the scanning path of each layer and plans them in sequence.

[0045] The thickness of the slice layer and the forming process parameters are determined by basic process experiments or experience. The basic process parameters include key factors such as forming voltage, current, and travel speed.

[0046] Step 2: During the forming process, the actual formed shape of the part is measured online.

[0047] The motion platform drives the structured light sensor to scan the upper surface of the part, obtaining a three-dimensional point cloud of the part surface in the measurement coordinate system.

[0048] The transformation matrix from the measurement coordinate system to the forming coordinate system is determined through calibration experiments before forming. After the measurement is completed, the point cloud of the part surface is transformed into the forming coordinate system and saved.

[0049] Step 3: Compare and analyze the differences between the actual formed morphology and the expected formed morphology.

[0050] First, the raw point cloud data of the line laser is preprocessed to obtain the smoothed point cloud of the deposition layer. The preprocessing steps include segmentation and denoising.

[0051] Discretize the spatial trajectory to be formed in the next layer into a set of equally spaced discrete points P, P = P ij ∈R 3 |i=1,2,…M;j=1,2,…N, where M is the number of trajectories and N is the number of discrete points on a trajectory.

[0052] In the process of arc additive manufacturing, the height of the weld bead corresponding to the trajectory will fluctuate slightly and frequently. If the adjacent points are too close, it will lead to excessively frequent parameter changes, affecting the stability of the forming process and the welding quality. If the adjacent points are too far apart, the parameter adjustment effect will not meet expectations. Taking all factors into consideration, the spacing between the discrete points of the trajectory is set to 1mm.

[0053] The deviation calculation takes the discrete trajectory position point as the center and the width of the melt pool as the diameter. The height of the topographic point cloud in this area is taken as the actual height of the trajectory position. The difference between the actual height and the expected height is used to obtain the height deviation of the corresponding discrete trajectory point.

[0054] Step 4: Adjust the trajectory and shaping parameters of the next layer based on fuzzy inference.

[0055] A Mamdani-type two-dimensional fuzzy logic controller was established, using height deviation and height deviation change as inputs and walking speed change as outputs.

[0056] The deviation *e* of the morphological height at various locations on the deposition layer surface is classified into 7 levels: {negative large, negative medium, negative small, zero, positive small, positive medium, positive large}, with the corresponding fuzzy subsets set as {NB, NM, NS, ZO, PS, PM, PB}, and the corresponding values ​​as {-3mm, -2mm, -1mm, 0mm, 1mm, 2mm, 3mm}. Similarly, the variation *ce* of the morphological height error is classified into 5 levels: {negative large, negative small, zero, positive small, positive large}, with the corresponding fuzzy subsets set as {NB, NS, ZO, PS, PB}, and the corresponding values ​​as {-2mm, -1mm, 0mm, 1mm, 2mm}. The output walking speed change is divided into 5 levels {negative large, negative small, zero, positive small, positive large}, and the corresponding fuzzy subsets are set as {NB, NS, ZO, PS, PB}, with corresponding values ​​of {0.4m / min, 0.5m / min, 0.6m / min, 0.7m / min, 0.8m / min}.

[0057] The membership function is a triangular function near the equilibrium point and a Gaussian function at the edges. The membership degree at the intersection of two adjacent membership functions is set to 0.5. The centroid method is used for the defuzzification process.

[0058] Once the membership function is determined, the fuzzy control rule table can be designed based on the rule manual, expert experience, and prior experimental results. The basic idea is to control the step size of the fuzzy adjustment based on the deviation and its change. For example, when the deviation and its change deviate from zero in the same direction, a larger adjustment step size in the opposite direction should be selected to quickly reduce the error; when the deviation and its change deviate from zero in opposite directions, the step size should be reduced for fine adjustment.

[0059]

[0060] like Figure 2 As shown, Figure 4 As shown, the hardware system of this invention consists of a Gocator 2430 optical structure sensor and a MIG welding system.

[0061] Figure 3 This paper demonstrates the layered slicing and path trajectory of a 160 mm long, 20-layer straight-wall feature structure formed layer by layer using two reciprocating deposition methods. During the forming process, all process parameters remained consistent except for the travel speed. The process parameters for the forming process are set as follows:

[0062]

[0063] The online adaptive control method for the topography of arc additive manufacturing based on interlayer fuzzy control proposed in this invention is used to deposit and form parts, such as... Figure 3 As shown in (a). During the forming process, all process parameters except for the travel speed remained consistent. Starting from the third layer, the travel speed was adaptively determined by the fuzzy controller based on the actual measured surface morphology. The system's process parameter control resolution was set to 1 mm. During the forming process, the height distribution curve of each layer was as follows: Figure 3 As shown in (b), the results show that the feature part is well formed overall, with uniform height distribution in each layer and minimal fluctuations. The increasing trend of the height difference between the two ends and the middle area during the forming process is effectively blocked and tends to stabilize, ensuring the consistency and stability of the forming quality.

[0064] In contrast, when forming without using the method of this invention, the formed straight-walled structural member is as follows: Figure 4 As shown in (a). After each layer is formed, the surface morphology is measured online, and the height distribution curve of the feature is obtained as shown in (a). Figure 4As shown in (b), the results indicate that even with a reciprocating stacking strategy, the height difference between the two ends and the middle end of the straight-walled feature gradually increases from the third layer onwards. By the 20th layer, the height difference between the two ends of the weld bead and the middle area is close to 10.2 mm. This "collapse" phenomenon at the arc end leads to poor gas protection during the forming process, which in turn affects the arc morphology and forming stability, resulting in a decrease in forming quality. Simultaneously, the effective forming area decreases, and when the height difference further increases to a certain extent, it can even cause molten pool flow, making it impossible to continue the forming process.

[0065] The comparative experiments above demonstrate that the online adaptive control method for the morphology of arc additive manufacturing based on interlayer fuzzy control of the present invention can significantly improve the surface morphology stability and forming accuracy during the forming process, effectively suppress the accumulation of forming deviations, and improve forming quality and reliability.

Claims

1. An online adaptive control method for arc additive manufacturing topography, characterized in that, The steps are as follows: Step 1: Slice the 3D model of the part to be formed into layers, plan the forming path, and obtain the spatial trajectory and process parameters of the part deposition forming; import the planned spatial trajectory and process parameters into the CNC system and the arc forming system, and the CNC system drives the arc forming system to perform layer-by-layer additive forming on the substrate according to the planned forming path. Step 2: During the arc additive manufacturing process, a morphology sensor is used to acquire the surface morphology information of the deposited layer online, and the results are transmitted to the computer control system; To obtain information on the surface morphology of the deposition layer, it is necessary to calibrate the measurement coordinate system and the forming coordinate system in advance before forming, determine the transformation matrix between the two, and then transform the measurement point cloud in the measurement coordinate system to the printing coordinate system according to the calibration parameters, and perform joint analysis and calculation with the pre-planned lower layer forming trajectory. Step 3: The computer control system analyzes the actual morphology information of the sedimentary layer and obtains the morphology deviation by comparing it with the expected morphology information; It is assumed that the part to be formed consists of n a layer deposition layer, a first n a layer deposition layer surface at a x,y ) expected height, a measured actual deposition height: wherein, is a topography height deviation at a position of a current layer ( x,y ), is a topography height deviation at a position of a previous layer of the current layer ( x,y ), is a change in topography height deviation between the two layers; Step 4: Based on the morphology difference results, dynamically plan and update the subsequent forming trajectory and forming process parameters to suppress surface morphology differences and improve forming quality and accuracy; Based on the morphological differences, the subsequent forming trajectory and forming process parameters are dynamically planned and updated using a fuzzy control method. The fuzzy control method employs a dual-input Mamdani-type two-dimensional fuzzy logic controller, which includes a fuzzy generator and a fuzzy canceller. The current layer ( x,y ) Topographic height deviation at location and changes in morphological height deviation The fuzzy logic inference engine is used as input to infer the next layer trajectory adjustment amount and its corresponding value based on fuzzy rules. x,y Adjustment amount of positional process parameters; The next layer corresponds to ( x,y The position trajectory adjustment amount is: in, For the next layer trajectory in ( x,y The preset height at ) For the next layer trajectory in ( x,y The reasoning height at ) The process parameter is the travel speed, and the next layer corresponds to ( x,y Walking speed at position for: in, For the next deposition layer in ( x,y The preset stacking speed at point ); The walking speed adjustment amount obtained through logical reasoning; The morphology sensor is a structured light sensor, used for online monitoring of the morphology of the deposition layer, and feeding back the morphology results to the computer control system in real time in the form of a three-dimensional point cloud.

2. The online adaptive control method for morphology in arc additive manufacturing according to claim 1, characterized in that, The online adaptive control method for the morphology of electric arc additive manufacturing is based on an online detection and adaptive control system for the morphology of electric arc additive manufacturing, which includes a CNC system, a motion platform, a computer control system, an electric arc additive manufacturing system, and a morphology detection system.

3. The online adaptive control method for morphology in arc additive manufacturing according to claim 2, characterized in that, The CNC system drives the arc forming system to form relative motion according to the preset forming trajectory. During the motion, the wire feeding mechanism of the arc forming system melts the metal wire and deposits it on the surface of the forming substrate, thereby obtaining a complete metal part of the desired shape.

4. The online adaptive control method for morphology in arc additive manufacturing according to claim 1, characterized in that, In the fuzzy control method, the deviation of the forming height at each position on the surface of the deposition layer is... e The judgment criteria deviation is divided into 7 levels: negative large, negative medium, negative small, zero, positive small, positive medium, and positive large, with corresponding fuzzy subsets set as NB, NM, NS, ZO, PS, PM, and PB; the deviation of forming height e The range is [-3, 3] mm; for the variation of topographic height error ce Divided into 5 levels: negative large, negative small, zero, positive small, and positive large, corresponding to fuzzy subsets set as NB, NS, ZO, PS, and PB; topography height error variation ce The range is [-2,2] mm; the output walking speed change TS is divided into 5 levels: negative large, negative small, zero, positive small, positive large, and the corresponding fuzzy subsets are set as NB, NS, ZO, PS, PB; the range of walking speed change TS is [0.4,0.8] m / min.

5. The online adaptive control method for morphology in arc additive manufacturing according to claim 4, characterized in that, The deviation of the forming height at various locations on the surface of the deposition layer e The judgment criteria deviation is divided into 7 levels: {negative large, negative medium, negative small, zero, positive small, positive medium, positive large}, with corresponding values ​​of {-3 mm, -2 mm, -1 mm, 0 mm, 1 mm, 2 mm, 3 mm}; for the variation of morphological height error... ce The walking speed is divided into 5 levels {negative large, negative small, zero, positive small, positive large}, with corresponding values ​​of {-2 mm, -1 mm, 0 mm, 1 mm, 2 mm}; the output walking speed change is divided into 5 levels {negative large, negative small, zero, positive small, positive large}, with corresponding values ​​of {0.4 m / min, 0.5 m / min, 0.6 m / min, 0.7 m / min, 0.8 m / min}.

6. The online adaptive control method for morphology in arc additive manufacturing according to claim 5, characterized in that, In the fuzzy control method, the fuzzy rules are designed based on experience and prior shaping and adjustment experience. The basic principle is to control the step size of fuzzy adjustment according to the deviation and the change of deviation. Specifically, when the deviation and the change of deviation deviate from the zero point in the same direction, a large adjustment step size in the opposite direction is selected to reduce the error as soon as possible; when the deviation and the change of deviation deviate from the zero point in opposite directions, the step size needs to be reduced for fine adjustment.

Images