A method for controlling the wall thickness uniformity of the combustion channel in additive manufacturing of aerospace propulsion chambers

Abstract

The present invention discloses a method for controlling the wall thickness uniformity of a combustion channel in additive manufacturing of a space propulsion chamber, comprising the following steps: S1, establishing a three-dimensional geometric model and performing layered slicing to generate an initial set of processing control parameters; S2, constructing a thermal echo map before printing the first layer and correcting the control parameters of the first layer; S3, generating a wall thickness error map of the current layer based on the data of the previous layer and constructing a wall thickness error potential energy field map; S4, constructing a target scanning density function, solving the two-dimensional Monge-Ampère equation to generate a path potential function, and generating a pseudo-potential streamline field; S5, inputting the current layer data into a wall thickness prediction neural network model and outputting a predicted value for the wall thickness deviation of the next layer; S6, updating the wall thickness error map of the next layer based on the predicted value for the wall thickness deviation and dynamically adjusting the control parameters; S7, looping through steps S3 to S6 until the printing task is completed. The present invention integrates thermal response modeling with a neural network model to achieve dynamic uniformity control of the combustion channel wall thickness.

Description

Technical Field

[0001] The present invention relates to the field of additive manufacturing control technology, and in particular to a method for controlling the wall thickness uniformity of an additively manufactured combustion channel in an aerospace propulsion chamber. Background Art

[0002] The continuous development of aerospace propulsion technology places higher demands on the structural complexity, precision, and stability of core thermal structural components, such as engine combustion ducts. Selective Laser Melting (SLM), a high-precision, high-degree-of-freedom metal additive manufacturing technology, has become an important means of manufacturing complex internal flow channel structures, such as those in aerospace propulsion chamber combustion ducts. However, due to the limitations of combustion duct wall structures, such as their complex curves, closed rotations, and intensive heat loads, the actual additive manufacturing process is prone to quality issues such as uneven wall thickness, geometric distortion, unstable melt pools, and stress concentration, seriously affecting the subsequent processing and service safety of the components.

[0003] Currently, additive manufacturing wall thickness control typically relies on preset scanning strategies and empirical process parameters, such as using serpentine filling, staggered scanning, and power regulation to statically optimize the forming process. However, most of these control methods are based on global parameter settings and lack the ability to fine-tune the modeling and dynamic response of printing states at different locations, making it difficult to achieve local prediction and closed-loop correction of wall thickness errors. In addition, traditional path planning methods often use straight-line or uniform distribution strategies, failing to fully incorporate the changing characteristics of the spatial error distribution of wall thickness for adaptive path reconstruction, resulting in redundant or insufficient accumulation in local areas, which in turn causes the accumulation of wall thickness deviations.

[0004] In terms of measurement feedback, most existing technologies rely on post-print linear profile measurement or surface point cloud reconstruction methods to obtain structural deviation information, which is unable to achieve dynamic modeling and real-time predictive control during the printing process. Furthermore, even though some studies have introduced infrared thermal imaging or melt pool monitoring methods, the lack of systematic data modeling and physical interpretation paths makes it difficult to establish an effective correlation mechanism between thermal response behavior and subsequent wall thickness errors, making it impossible to effectively transform thermal process characteristics into control feedback.

[0005] In recent years, some research has attempted to incorporate deep learning methods to model and predict additive manufacturing processes. For example, convolutional neural networks are used to extract topographic features, or recurrent neural networks are used to analyze historical data trends. However, most models lack a physical guidance mechanism and have a single input dimension, making it difficult to map prediction results to specific process control parameters. Furthermore, the current lack of dynamic control logic for two-dimensional spatial error fields makes it impossible to establish a precise control strategy allocation mechanism for each scanning unit. This leads to a disconnect between prediction and adjustment, and prevents the formation of a complete closed-loop feedback loop.

[0006] Therefore, how to provide a method for controlling the uniformity of the wall thickness of the combustion channel of the additively manufactured aerospace propulsion chamber is an urgent problem that needs to be solved by those skilled in the art. Summary of the Invention

[0007] One purpose of the present invention is to propose a method for controlling the wall thickness uniformity of the additively manufactured combustion channel of an aerospace propulsion chamber. The present invention integrates thermal response modeling, wall thickness error potential field construction and deep learning prediction network, obtains thermal property differences through non-melting thermal response scanning, dynamically reconstructs the scanning path based on the Monge-Ampère equation, and combines a multi-dimensional wall thickness prediction model to achieve point-by-point adaptive adjustment of control parameters. The method has the advantages of high wall thickness uniformity, fast error response and high path control accuracy, effectively improving the forming quality and structural consistency of the combustion channel of the aerospace propulsion chamber during the additive manufacturing process, and solving the problems of path rigidity, regulation lag and rough wall thickness control in the existing manufacturing process.

[0008] A method for controlling the wall thickness uniformity of a combustion channel in an additively manufactured aerospace propulsion chamber according to an embodiment of the present invention includes the following steps:

[0009] S1. Establish a three-dimensional geometric model of the combustion channel and perform layered slicing to obtain structural information of each layer and generate an initial processing control parameter set;

[0010] S2. Before printing the first layer, use a variable frequency laser to perform a non-melting scan on the combustion channel area to be printed and construct a thermal echo map, and modify the control parameters used for the first layer in the initial processing control parameter set based on the thermal echo map;

[0011] S3. Before printing each layer, a wall thickness error map of the current layer is generated based on the actual forming data of the previous layer, and the wall thickness error value of each printing point is mapped to the wall thickness error potential energy value at the corresponding printing point position to construct a wall thickness error potential energy field map;

[0012] S4. Constructing a target scanning density function according to the wall thickness error potential energy field map, solving the two-dimensional Monge-Ampère equation based on the target scanning density function to generate a path potential function, and generating a pseudo-potential streamline field according to the path potential function as the scanning path of the current printing layer;

[0013] S5. After the current layer is printed, the actual contour image, wall thickness data, pseudo-potential streamline field and control parameters of the current layer are jointly input into the wall thickness prediction neural network model, and the predicted value of the wall thickness deviation of the next layer in each scanning unit is output;

[0014] S6. updating the wall thickness error map of the next layer according to the wall thickness deviation prediction value of the next layer in each scanning unit, and dynamically adjusting the control parameters to complete the adaptive update of the control parameters;

[0015] S7. Loop through steps S3 to S6 until the printing task is completed.

[0016] Optionally, the initial processing control parameter set includes the scanning path, laser power, scanning speed, scanning spacing and spot size corresponding to each layer, and is generated by matching the cross-sectional geometric features of each layer of channels in the three-dimensional geometric model.

[0017] Optionally, the S2 specifically includes:

[0018] S21. Without applying melting laser power, apply variable frequency laser scanning in a frequency range of 10 Hz to 10 kHz to the combustion channel area to be printed, controlling the laser power to be no higher than 20% of the material melting threshold to stimulate thermal response changes in the area to be printed;

[0019] S22, using an infrared thermal imager to collect thermal response time series data of each scanning point during the laser scanning process, and record the temperature change during the temperature rise and cooling process of each scanning point;

[0020] S23, processing the thermal response time series data of each scanning point to extract thermal response characteristic parameters, wherein the thermal response characteristic parameters include peak temperature, heating time, half-life and temperature decay slope, and forming a thermal response characteristic matrix;

[0021] S24, comparing the thermal response characteristic matrix with a preset thermal behavior model point by point, wherein the preset thermal behavior model uses the response of aluminum alloy powder at a standard bulk density as a reference curve library, and adopts the Euclidean distance minimum principle to determine the thermal behavior deviation of each actual scanning point;

[0022] S25, constructing a two-dimensional difference image using the thermal behavior deviation value of each scanning point as the pixel grayscale value, and constructing a thermal echo map by projecting the normalized thermal behavior deviation into the thermal map, wherein the thermal echo map reflects the change trend of the thermal absorption coefficient of the local area;

[0023] S26. According to the variation range of the thermal absorption coefficient of each scanning point in the thermal echo spectrum, adjust the first layer laser power, scanning speed and spot diameter of the corresponding area, generate a first layer corrected processing control parameter set and use it for the first layer printing.

[0024] Optionally, the S3 specifically includes:

[0025] S31, obtaining actual forming data after the previous layer is printed, the actual forming data including a three-dimensional contour image and actual wall thickness measurement data, wherein the three-dimensional contour image is acquired by an optical coherence tomography scanner, and the actual wall thickness measurement data is obtained by image reconstruction and calibration point spacing calculation;

[0026] S32, one-to-one correspondence between the target design wall thickness of the current layer and the actual wall thickness measurement data at each printing point position, calculating the wall thickness error value of each printing point, and forming a wall thickness error distribution matrix;

[0027] S33, performing two-dimensional interpolation and Gaussian smoothing on the wall thickness error distribution matrix to obtain a wall thickness error map of the current layer, wherein the wall thickness error map records the error size and spatial position of each printing point in the form of an image;

[0028] S34, setting an error mapping function to map the wall thickness error value of each printing point to the wall thickness error potential energy value at the corresponding printing point position;

[0029] S35. Construct a wall thickness error potential energy field diagram based on the wall thickness error potential energy value distribution obtained by the error mapping function. The wall thickness error potential energy field diagram is represented in the form of a two-dimensional matrix. Each matrix unit corresponds to a printing point position, and the wall thickness error potential energy value of the printing point position is recorded to form a continuous spatial distribution data structure for describing the size and directional change trend of the wall thickness deviation in each area.

[0030] Optionally, the error mapping function is specifically expressed as:

[0031] ;

[0032] in, Indicates the current layer printing point The potential energy value of the wall thickness error at Indicates the current layer printing point The wall thickness error value at 、 、 、 and Indicates the preset control coefficient.

[0033] Optionally, the S4 specifically includes:

[0034] S41. Normalize the wall thickness error potential energy field diagram to obtain the normalized wall thickness error potential energy value , the value range is between;

[0035] S42, constructing a target scanning density function based on the normalized wall thickness error potential energy value ,make and There is a positive correlation and the integral conservation constraint is satisfied:

[0036] ;

[0037] in, Indicates the current layer printing area. Indicates the total amount of scan path resources in the current layer;

[0038] S43, construct the Monge-Ampère equation with directional coupling and scan the density function with the target Solve the path potential function for the target term , the Monge-Ampère equation is:

[0039] ;

[0040] in, represents the determinant value, represents the Hessian matrix of the path potential function, represents the coupling coefficient, represents the directional coupling matrix, which is defined as:

[0041] ;

[0042] in, Indicates the direction of the wall thickness error gradient;

[0043] S44, setting the path potential function boundary condition to the Neumann boundary condition, stipulating that the path potential function normal derivative is zero on the printing area boundary to maintain the boundary continuity of the path field;

[0044] S45. Solve the Monge-Ampère equation to obtain the path potential function The two-dimensional distribution results of ;

[0045] S46. Calculate the weighted gradient field vector for scanning path guidance:

[0046] ;

[0047] in, represents the weighted gradient field vector, and represents the control weight coefficient, represents the gradient of the path potential function, represents the gradient of the normalized wall thickness error potential field;

[0048] S47 , constructing a pseudo-potential streamline field according to the directional distribution of the weighted gradient field vector as a scanning path for the current printing layer.

[0049] Optionally, the wall thickness prediction neural network model includes:

[0050] The spatial feature encoding module is used to receive and process the actual contour image, wall thickness data, pseudo-potential streamline field and control parameters of the current layer, and extract multi-scale spatial features through a two-dimensional convolutional encoder;

[0051] A temporal feature extraction module is used to load historical wall thickness error maps and control parameter sequences of multiple consecutive printing layers, and use a gated recurrent unit to extract the temporal evolution characteristics of the error changes;

[0052] A feature fusion module is used to splice and fuse the multi-scale spatial features and the time evolution features in the feature dimension, and enhance the joint expression ability of local patterns and global trends through residual connections to generate fused joint features;

[0053] The prediction generation module is used to generate a two-dimensional wall thickness deviation prediction matrix based on the fused joint features through a two-dimensional convolution decoder, wherein the two-dimensional wall thickness deviation prediction matrix represents the wall thickness deviation prediction value of the next layer in each scanning unit.

[0054] Optionally, the S6 specifically includes:

[0055] Based on the predicted value of the wall thickness deviation of the next layer in each scanning unit, define the first threshold , the second threshold and the third threshold , dynamically update the various control parameters in the initial processing control parameter set of the next layer according to the set rules:

[0056] If the wall thickness deviation prediction value is less than or equal to the first threshold , the change rate of the wall thickness deviation prediction value is less than or equal to the second threshold And the deviation fluctuation of the scanning unit of two consecutive layers is less than or equal to the third threshold , the control parameters remain unchanged and it is determined to be a stable area;

[0057] If any of the conditions is met: the wall thickness deviation prediction value is greater than the first threshold , the change rate of the wall thickness deviation prediction value is greater than the second threshold , but the deviation fluctuation of the scanning unit in two consecutive layers is less than or equal to the third threshold , it is determined to be a slight correction area, and the control parameters are adjusted based on the wall thickness deviation prediction value and the preset adjustment factor:

[0058] ;

[0059] in, represents the adjusted laser power, Indicates the laser power before adjustment, represents the laser power adjustment factor, Indicates the predicted value of wall thickness deviation;

[0060] ;

[0061] in, Indicates the adjusted scanning speed, Indicates the scanning speed before adjustment. Indicates the scan speed adjustment factor;

[0062] ;

[0063] in, Indicates the adjusted scanning distance, Indicates the scanning distance before adjustment. Indicates the scan spacing adjustment factor;

[0064] ;

[0065] in, Indicates the adjusted spot size, Indicates the spot size before adjustment, represents the spot size adjustment factor;

[0066] The scan path remains unchanged over the slightly corrected area;

[0067] If the wall thickness deviation prediction value is greater than the first threshold , the change rate of the wall thickness deviation prediction value is greater than the second threshold And the deviation fluctuation of the scanning unit of two consecutive layers is greater than the third threshold , it is determined to be the key correction area, and the laser power, scanning speed, scanning spacing and spot size are adjusted according to the maximum correction amplitude upper limit ±15%, and the pseudo-potential streamline field is regenerated for the key correction area, and the scanning path is updated.

[0068] The beneficial effects of the present invention are:

[0069] (1) By introducing non-melting scanning and thermal echo mapping of variable frequency lasers before printing the first layer, it is possible to accurately identify the local thermal physical response differences of the powder bed before printing, effectively correct the first layer control parameters, and improve the consistency and stability of the initial printing conditions;

[0070] (2) A wall thickness error potential energy field map constructed based on actual forming data is used. By solving the Monge-Ampère equation for coupled directionality, the scanning path and the local error spatial distribution are controlled in a linked manner, so that the scanning trajectory dynamically tends to the low-potential error region, thereby suppressing the accumulation and diffusion of local wall thickness deviations.

[0071] (3) Construct a wall thickness prediction neural network model to simultaneously extract wall thickness data, path field images, and historical error evolution sequences, achieving cross-level and multi-dimensional accurate prediction of wall thickness deviations and improving the spatiotemporal perception capability of the prediction model;

[0072] (4) Based on the wall thickness deviation prediction results, an error level area map is constructed to achieve point-by-point dynamic adaptive adjustment of the laser power, scanning speed, scanning spacing, and spot size control parameters, significantly enhancing the manufacturing process's rapid response to local errors and closed-loop control capabilities;

[0073] (5) A full-process dynamic control mechanism from thermal response modeling, error potential field analysis, path optimization, prediction feedback to control adjustment was constructed, which effectively solved the problems of fixed path, static parameters and feedback delay in the existing technology, and realized the system optimization of wall thickness control under highly complex curved surface structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0075] Figure 1 This is an overall flow chart of a method for controlling the wall thickness uniformity of a combustion channel in additive manufacturing of a space propulsion chamber proposed by the present invention;

[0076] Figure 2 A flowchart of the Monge-Ampère equation for generating path potential functions and pseudo-potential streamline fields for a method for controlling the wall thickness uniformity of a combustion channel in additive manufacturing of aerospace propulsion chambers proposed in the present invention;

[0077] Figure 3 This is a schematic diagram of the wall thickness prediction neural network model structure of a method for controlling the wall thickness uniformity of a combustion channel in additive manufacturing of an aerospace propulsion chamber proposed in the present invention. DETAILED DESCRIPTION

[0078] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0079] refer to Figure 1-Figure 3 A method for controlling the wall thickness uniformity of a combustion channel in an additively manufactured aerospace propulsion chamber comprises the following steps:

[0080] S1. Establish a three-dimensional geometric model of the combustion channel and perform layered slicing to obtain structural information of each layer and generate an initial processing control parameter set;

[0081] S2. Before printing the first layer, use a variable frequency laser to perform a non-melting scan on the combustion channel area to be printed and construct a thermal echo map, and modify the control parameters used for the first layer in the initial processing control parameter set based on the thermal echo map;

[0082] S3. Before printing each layer, a wall thickness error map of the current layer is generated based on the actual forming data of the previous layer, and the wall thickness error value of each printing point is mapped to the wall thickness error potential energy value at the corresponding printing point position to construct a wall thickness error potential energy field map;

[0083] S4. Constructing a target scanning density function according to the wall thickness error potential energy field map, solving the two-dimensional Monge-Ampère equation based on the target scanning density function to generate a path potential function, and generating a pseudo-potential streamline field according to the path potential function as the scanning path of the current printing layer;

[0084] S5. After the current layer is printed, the actual contour image, wall thickness data, pseudo-potential streamline field and control parameters of the current layer are jointly input into the wall thickness prediction neural network model, and the predicted value of the wall thickness deviation of the next layer in each scanning unit is output;

[0085] S6. updating the wall thickness error map of the next layer according to the wall thickness deviation prediction value of the next layer in each scanning unit, and dynamically adjusting the control parameters to complete the adaptive update of the control parameters;

[0086] S7. Loop through steps S3 to S6 until the printing task is completed.

[0087] The present invention establishes a closed-loop additive manufacturing control process for aerospace propulsion chamber combustion channel structures, combining thermal scanning pre-sensing, wall thickness error potential field construction, path function optimization and depth prediction model linkage control to form a dynamic iterative mechanism from forming prediction to path planning and then to parameter adjustment. This method can accurately identify local error trends before each printing layer, and generate a dynamic adaptive scanning path based on a mathematical and physical model to achieve on-demand adjustment of scanning density and energy input, thereby effectively suppressing the cross-layer accumulation of wall thickness deviations. This method not only significantly improves the uniformity and control granularity of curved wall thickness, but also enhances the local sensitivity of path response, solving the problems of control hysteresis and parameter rigidity in complex wall structures.

[0088] In this embodiment, the initial processing control parameter set includes the scanning path, laser power, scanning speed, scanning spacing and spot size corresponding to each layer, and is generated by matching the cross-sectional geometric features of each layer channel in the three-dimensional geometric model.

[0089] In this embodiment, S2 specifically includes:

[0090] S21. Without applying melting laser power, apply variable frequency laser scanning in a frequency range of 10 Hz to 10 kHz to the combustion channel area to be printed, controlling the laser power to be no higher than 20% of the material melting threshold to stimulate thermal response changes in the area to be printed;

[0091] S22, using an infrared thermal imager to collect thermal response time series data of each scanning point during the laser scanning process, and record the temperature change during the temperature rise and cooling process of each scanning point;

[0092] S23, processing the thermal response time series data of each scanning point to extract thermal response characteristic parameters, wherein the thermal response characteristic parameters include peak temperature, heating time, half-life and temperature decay slope, and forming a thermal response characteristic matrix;

[0093] S24, comparing the thermal response characteristic matrix with a preset thermal behavior model point by point, wherein the preset thermal behavior model uses the response of aluminum alloy powder at a standard bulk density as a reference curve library, and adopts the Euclidean distance minimum principle to determine the thermal behavior deviation of each actual scanning point;

[0094] S25, constructing a two-dimensional difference image using the thermal behavior deviation value of each scanning point as the pixel grayscale value, and constructing a thermal echo map by projecting the normalized thermal behavior deviation into the thermal map, wherein the thermal echo map reflects the change trend of the thermal absorption coefficient of the local area;

[0095] S26. According to the variation range of the thermal absorption coefficient of each scanning point in the thermal echo spectrum, adjust the first layer laser power, scanning speed and spot diameter of the corresponding area, generate a first layer corrected processing control parameter set and use it for the first layer printing.

[0096] By introducing non-melting variable-frequency laser scanning and constructing thermal echo maps, a spatial tomographic perception mechanism for the local thermal physical state before printing was established, enabling the acquisition of high-resolution thermal response characteristics without affecting the powder bed state. By using a standardized thermal behavior model to perform point-by-point matching of the actual thermal response data, a differential distribution map of the real-space thermal absorption characteristics was obtained, enabling precise correction of the first-layer processing parameters. This step significantly enhances the adaptability of the initial printing stage to the local powder thermal stability, reduces the risk of energy input imbalance in the initial forming phase, and helps control the source of thermal distortion and error diffusion trends.

[0097] In this embodiment, S3 specifically includes:

[0098] S31, obtaining actual forming data after the previous layer is printed, the actual forming data including a three-dimensional contour image and actual wall thickness measurement data, wherein the three-dimensional contour image is acquired by an optical coherence tomography scanner, and the actual wall thickness measurement data is obtained by image reconstruction and calibration point spacing calculation;

[0099] S32, one-to-one correspondence between the target design wall thickness of the current layer and the actual wall thickness measurement data at each printing point position, calculating the wall thickness error value of each printing point, and forming a wall thickness error distribution matrix;

[0100] S33, performing two-dimensional interpolation and Gaussian smoothing on the wall thickness error distribution matrix to obtain a wall thickness error map of the current layer, wherein the wall thickness error map records the error size and spatial position of each printing point in the form of an image;

[0101] S34, setting an error mapping function to map the wall thickness error value of each printing point to the wall thickness error potential energy value at the corresponding printing point position;

[0102] S35. Construct a wall thickness error potential energy field diagram based on the wall thickness error potential energy value distribution obtained by the error mapping function. The wall thickness error potential energy field diagram is represented in the form of a two-dimensional matrix. Each matrix unit corresponds to a printing point position, and the wall thickness error potential energy value of the printing point position is recorded to form a continuous spatial distribution data structure for describing the size and directional change trend of the wall thickness deviation in each area.

[0103] The process of constructing a wall thickness error potential energy field provides a continuous spatial field foundation for path optimization, transforming discrete printing error data into a continuous control target with physical guidance. By mapping wall thickness errors into potential energy values ​​and generating a spatial error field, a quantitative representation and directional description of localized areas of wall thickness anomalies are achieved. This effectively addresses the problem that traditional error control models based on contour lines or profiles cannot reflect spatial gradient changes, providing a more refined and differentiable input representation for path function solution and control parameter scheduling.

[0104] In this embodiment, the error mapping function is specifically expressed as:

[0105] ;

[0106] in, Indicates the current layer printing point The potential energy value of the wall thickness error at Indicates the current layer printing point The wall thickness error value at 、 、 、 and Indicates the preset control coefficient.

[0107] The error mapping function employed incorporates high-order control factors during its design, enabling flexible adjustment of the nonlinearity of the error value's response to the potential energy, thereby endowing the error potential energy field with enhanced resolution and dynamic response range. By combining exponential terms, inverse hyperbolic tangent terms, and multiple slope control factors, the error mapping function possesses controllable characteristics in terms of boundary smoothing, enhanced center response, and gradient transition. This effectively suppresses the risk of distortion in the path guidance results caused by outliers, enhancing the convergence of the path field and practical machining feasibility.

[0108] In this embodiment, the S4 specifically includes:

[0109] S41. Normalize the wall thickness error potential energy field diagram to obtain the normalized wall thickness error potential energy value , the value range is between;

[0110] S42, constructing a target scanning density function based on the normalized wall thickness error potential energy value ,make and There is a positive correlation and the integral conservation constraint is satisfied:

[0111] ;

[0112] in, Indicates the current layer printing area. Indicates the total amount of scan path resources in the current layer;

[0113] S43, construct the Monge-Ampère equation with directional coupling and scan the density function with the target Solve the path potential function for the target term , the Monge-Ampère equation is:

[0114] ;

[0115] in, represents the determinant value, represents the Hessian matrix of the path potential function, represents the coupling coefficient, represents the directional coupling matrix, which is defined as:

[0116] ;

[0117] in, Indicates the direction of the wall thickness error gradient;

[0118] S44, setting the path potential function boundary condition to the Neumann boundary condition, stipulating that the path potential function normal derivative is zero on the printing area boundary to maintain the boundary continuity of the path field;

[0119] S45. Solve the Monge-Ampère equation to obtain the path potential function The two-dimensional distribution results of ;

[0120] S46. Calculate the weighted gradient field vector for scanning path guidance:

[0121] ;

[0122] in, represents the weighted gradient field vector, and represents the control weight coefficient, represents the gradient of the path potential function, represents the gradient of the normalized wall thickness error potential field;

[0123] S47 , constructing a pseudo-potential streamline field according to the directional distribution of the weighted gradient field vector as a scanning path for the current printing layer.

[0124] By incorporating the Monge-Ampère equation into the scanning path construction process and setting a target scan density function coupled with a directional gradient, a high degree of coupling between the potential function generation of the scanning path and the error field distribution can be achieved. The calculation of the path potential function not only satisfies local density constraints but also guides the path trend to converge to a low-error region through boundary conditions and weighted gradient fields, achieving adaptive control of the scanning trajectory in space. This path generation strategy is mathematically closed and physically reasonable, significantly outperforming traditional path grid methods or contour extension strategies, and exhibits greater deformation adaptability and continuity stability in controlling the internal wall thickness of curved surfaces.

[0125] In this embodiment, the wall thickness prediction neural network model includes:

[0126] The spatial feature encoding module is used to receive and process the actual contour image, wall thickness data, pseudo-potential streamline field and control parameters of the current layer, and extract multi-scale spatial features through a two-dimensional convolutional encoder;

[0127] A temporal feature extraction module is used to load historical wall thickness error maps and control parameter sequences of multiple consecutive printing layers, and use a gated recurrent unit to extract the temporal evolution characteristics of the error changes;

[0128] A feature fusion module is used to splice and fuse the multi-scale spatial features and the time evolution features in the feature dimension, and enhance the joint expression ability of local patterns and global trends through residual connections to generate fused joint features;

[0129] The prediction generation module is used to generate a two-dimensional wall thickness deviation prediction matrix based on the fused joint features through a two-dimensional convolution decoder, wherein the two-dimensional wall thickness deviation prediction matrix represents the wall thickness deviation prediction value of the next layer in each scanning unit.

[0130] The wall thickness prediction neural network model fully integrates the spatial structural characteristics of the printing process, control parameter information, and historical error evolution sequences, demonstrating a high degree of contextual modeling capabilities. By extracting local geometric patterns through convolution and capturing temporal dynamic features through gated structures, and retaining residual connection paths during the fusion phase, the model effectively enhances its ability to learn non-stationary and nonlinear error trends. The resulting two-dimensional prediction matrix corresponds one-to-one with the spatial scanning units, providing a high-resolution prediction basis for precise dynamic adjustment of control parameters and enabling efficient integration of the prediction model into the control logic.

[0131] In this embodiment, S6 specifically includes:

[0132] Based on the predicted value of the wall thickness deviation of the next layer in each scanning unit, define the first threshold , the second threshold and the third threshold , dynamically update the various control parameters in the initial processing control parameter set of the next layer according to the set rules:

[0133] If the wall thickness deviation prediction value is less than or equal to the first threshold , the change rate of the wall thickness deviation prediction value is less than or equal to the second threshold And the deviation fluctuation of the scanning unit of two consecutive layers is less than or equal to the third threshold , the control parameters remain unchanged and it is determined to be a stable area;

[0134] If any of the conditions is met: the wall thickness deviation prediction value is greater than the first threshold , the change rate of the wall thickness deviation prediction value is greater than the second threshold , but the deviation fluctuation of the scanning unit in two consecutive layers is less than or equal to the third threshold , it is determined to be a slight correction area, and the control parameters are adjusted based on the wall thickness deviation prediction value and the preset adjustment factor:

[0135] ;

[0136] in, represents the adjusted laser power, Indicates the laser power before adjustment. represents the laser power adjustment factor, Indicates the predicted value of wall thickness deviation;

[0137] ;

[0138] in, Indicates the adjusted scanning speed, Indicates the scanning speed before adjustment. Indicates the scan speed adjustment factor;

[0139] ;

[0140] in, Indicates the adjusted scanning distance, Indicates the scanning distance before adjustment. Indicates the scan spacing adjustment factor;

[0141] ;

[0142] in, Indicates the adjusted spot size, Indicates the spot size before adjustment, represents the spot size adjustment factor;

[0143] The scan path remains unchanged over the slightly corrected area;

[0144] If the wall thickness deviation prediction value is greater than the first threshold , the change rate of the wall thickness deviation prediction value is greater than the second threshold And the deviation fluctuation of the scanning unit of two consecutive layers is greater than the third threshold , it is determined to be the key correction area, and the laser power, scanning speed, scanning spacing and spot size are adjusted according to the maximum correction amplitude upper limit ±15%, and the pseudo-potential streamline field is regenerated for the key correction area, and the scanning path is updated.

[0145] A control parameter adjustment mechanism based on predicted wall thickness deviations enables point-by-point dynamic adaptive adjustment of laser power, scanning speed, scanning spacing, and spot size. By setting multi-level response thresholds and rate-of-change criteria, a three-level control state is established: stable area, minor correction area, and key correction area, giving the system multi-level responsiveness. Furthermore, the path reconstruction strategy is used to update the scanning strategy for error-sensitive areas, effectively achieving local compensation and global balanced adjustment in error accumulation areas, significantly improving the accuracy and real-time performance of local quality control in complex curved surface printing.

[0146] Example 1:

[0147] To verify the feasibility of this invention, it was applied to a laser additive manufacturing (LAM) project for the hot-end structure of a short-burn liquid oxygen-kerosene propulsion chamber at an aerospace engine development unit. A typical vortex combustion channel section was selected as the test object. This channel structure, with a target wall thickness of 1.2mm to 1.5mm, was formed using selective laser melting of aluminum alloy powder. It features a typical variable-curvature inner wall and non-uniform energy flux density distribution, making it one of the most challenging hot-end structures to precisely control.

[0148] Traditional processes for this type of structure use a strategy of constant scanning spacing, constant laser power, and a fixed speed. While this method can achieve the desired configuration, local thickness deviations exceeding ±80μm often occur in arc transition zones and areas of sudden wall thickness changes, severely impacting channel cooling capacity and combustion stability, leading to rework rates exceeding 22%. However, the application of the wall thickness uniformity control method provided by this invention enables a comprehensive, intelligent upgrade to the original process.

[0149] First, during the printing preparation phase, this method introduces a variable-frequency laser to perform non-melting scanning of the powder layup layer. The laser frequency is set to vary from 200Hz to 2000Hz, and the laser power is controlled at 15% of the original powder melting threshold. Combined with the infrared thermal imager's recording of the thermal response process, a complete thermal echo map is constructed. This map reveals the spatial distribution of delayed thermal absorption response in local areas such as guide corners and overhanging edges, effectively guiding the differentiated correction of first-layer parameters. The laser power in areas with uneven thermal response was reduced by 8%, and the scanning speed was increased by 5%, effectively improving the stability of the first-layer forming.

[0150] Secondly, during the additive manufacturing process, before printing each layer, the system automatically extracts the three-dimensional contour map and actual wall thickness data of the previous layer after forming, and constructs a wall thickness error map for the current layer. The error map is processed by Gaussian filtering and spatial interpolation to generate a wall thickness potential field map. By setting the potential energy mapping function, the error range is mapped to a continuous field from 0 to 1. Based on this field, the directional Monge-Ampère path potential function is solved, and a pseudo-potential streamline field is generated as the actual scanning path, forming a trajectory distribution that adaptively matches the error trend. During the printing process, the scanning trajectory in the high-error potential area is significantly denser. At the same time, the power is reduced, and the path tends to the area with smaller errors, achieving local harmonic control.

[0151] After each printed layer, the system collects the actual control parameters, contour image, and generated path field for that layer. These are combined with the error sequences from multiple historical layers and fed into a residual prediction network model to generate a prediction matrix for the next layer's wall thickness deviation. The prediction accuracy averages 93.7% across different regions, and at local outliers, the prediction deviation is kept within ±15μm, far exceeding traditional empirical models.

[0152] During the control parameter adaptation phase, three-stage control thresholds were set based on the predicted results. This resulted in a dynamic adjustment range of scanning speed of ±12%, spot size control accuracy of 20μm, and laser power step control of ±0.5W. The total printing process involved data collection and feedback for 1,185 layers, with a path adjustment rate of 83.5%, and an average of 3.6 control parameter adjustments per layer.

[0153] During the final inspection, a high-precision three-dimensional coordinate measurement system and a multi-layer CT scanner were used to obtain wall thickness distribution data for the entire channel. Statistics showed that within the 130mm total length, 92.4% of the area had a wall thickness error within ±20μm, and the maximum error did not exceed ±38μm. Compared with the original process solution, the overall wall thickness control accuracy increased by 71.6%, and the rework rate decreased to 4.3%. The total printing time was 27 hours and 48 minutes, and the consumables rate was controlled within an error range of ±3%.

[0154] This embodiment verifies the effectiveness and engineering adaptability of the method of the present invention in high-precision wall thickness control by applying it to the actual printing task of the complex combustion channel structure of the aerospace propulsion chamber. Compared with the traditional printing strategy of fixed paths and static parameters, this solution realizes a full-process closed-loop control mechanism of thermal response prediction before printing, dynamic path regulation during printing, and intelligent prediction of errors after printing. Experimental data show that this method significantly reduces local wall thickness errors, improves the spatial resolution of path control and the accuracy of parameter adjustment, successfully improves the wall thickness control accuracy to within ±20μm, and reduces the rework rate by more than 80%, effectively solving the problems of path rigidity, error accumulation and control lag in the existing technology.

[0155] In addition, this embodiment demonstrates the excellent adaptability of the method of the present invention in dealing with complex thermal load structures and high-gradient wall thickness change areas. By constructing a wall thickness error potential energy field and introducing the Monge-Ampère path optimization mechanism, the scanning path can intelligently tend to the area with smaller errors according to the real-time error distribution, effectively avoiding the regional aggregation of wall thickness deviations. At the same time, the wall thickness prediction neural network model that integrates the deep residual network and historical layer data not only improves the accuracy of the prediction, but also provides a reliable basis for the feedforward adjustment of the control parameters. The dynamic adjustment of the control parameters in each layer ensures the optimal matching of the energy distribution during the processing process, and significantly improves the delayed response problem of traditional printing methods in wall thickness control. This embodiment is not only replicable and generalizable, but also provides practical technical support for the quality control of additive manufacturing of high-performance propulsion chambers, and has important engineering application value and promotion significance.

[0156] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for controlling the wall thickness uniformity of a combustion channel in an aerospace propulsion chamber additive manufacturing process, characterized in that: The steps include: S1. Establish a three-dimensional geometric model of the combustion channel and perform layered slicing to obtain structural information of each layer and generate an initial processing control parameter set; S2. Before printing the first layer, use a variable frequency laser to perform a non-melting scan on the combustion channel area to be printed and construct a thermal echo map, and modify the control parameters used for the first layer in the initial processing control parameter set based on the thermal echo map; S3. Before printing each layer, a wall thickness error map of the current layer is generated based on the actual forming data of the previous layer, and the wall thickness error value of each printing point is mapped to the wall thickness error potential energy value at the corresponding printing point position to construct a wall thickness error potential energy field map; S4. Constructing a target scanning density function according to the wall thickness error potential energy field map, solving the two-dimensional Monge-Ampère equation based on the target scanning density function to generate a path potential function, and generating a pseudo-potential streamline field according to the path potential function as the scanning path of the current printing layer; S5. After the current layer is printed, the actual contour image, wall thickness data, pseudo-potential streamline field and control parameters of the current layer are jointly input into the wall thickness prediction neural network model, and the predicted value of the wall thickness deviation of the next layer in each scanning unit is output; S6. updating the wall thickness error map of the next layer according to the wall thickness deviation prediction value of the next layer in each scanning unit, and dynamically adjusting the control parameters to complete the adaptive update of the control parameters; S7. Loop through steps S3 to S6 until the printing task is completed.

2. The method for controlling the wall thickness uniformity of the combustion channel of the aerospace propulsion chamber additive manufacturing according to claim 1 is characterized in that: The initial processing control parameter set includes the scanning path, laser power, scanning speed, scanning spacing and spot size corresponding to each layer, and is generated by matching the cross-sectional geometric features of each layer channel in the three-dimensional geometric model.

3. The method for controlling the wall thickness uniformity of the combustion channel of the aerospace propulsion chamber additively manufactured according to claim 1 is characterized in that: The S2 specifically includes: S21. Without applying melting laser power, apply variable frequency laser scanning in a frequency range of 10 Hz to 10 kHz to the combustion channel area to be printed, controlling the laser power to be no higher than 20% of the material melting threshold to stimulate thermal response changes in the area to be printed; S22, using an infrared thermal imager to collect thermal response time series data of each scanning point during the laser scanning process, and record the temperature change during the temperature rise and cooling process of each scanning point; S23, processing the thermal response time series data of each scanning point to extract thermal response characteristic parameters, wherein the thermal response characteristic parameters include peak temperature, heating time, half-life and temperature decay slope, and forming a thermal response characteristic matrix; S24, comparing the thermal response characteristic matrix with a preset thermal behavior model point by point, wherein the preset thermal behavior model uses the response of aluminum alloy powder at a standard bulk density as a reference curve library, and adopts the Euclidean distance minimum principle to determine the thermal behavior deviation of each actual scanning point; S25, constructing a two-dimensional difference image using the thermal behavior deviation value of each scanning point as the pixel grayscale value, and constructing a thermal echo map by projecting the normalized thermal behavior deviation into the thermal map, wherein the thermal echo map reflects the change trend of the thermal absorption coefficient of the local area; S26. According to the variation range of the thermal absorption coefficient of each scanning point in the thermal echo spectrum, adjust the first layer laser power, scanning speed and spot diameter of the corresponding area, generate a first layer corrected processing control parameter set and use it for the first layer printing.

4. The method for controlling the wall thickness uniformity of the combustion channel of the aerospace propulsion chamber additively manufactured according to claim 1 is characterized in that: The S3 specifically includes: S31, obtaining actual forming data after the previous layer is printed, the actual forming data including a three-dimensional contour image and actual wall thickness measurement data, wherein the three-dimensional contour image is acquired by an optical coherence tomography scanner, and the actual wall thickness measurement data is obtained by image reconstruction and calibration point spacing calculation; S32, one-to-one correspondence between the target design wall thickness of the current layer and the actual wall thickness measurement data at each printing point position, calculating the wall thickness error value of each printing point, and forming a wall thickness error distribution matrix; S33, performing two-dimensional interpolation and Gaussian smoothing on the wall thickness error distribution matrix to obtain a wall thickness error map of the current layer, wherein the wall thickness error map records the error size and spatial position of each printing point in the form of an image; S34, setting an error mapping function to map the wall thickness error value of each printing point to the wall thickness error potential energy value at the corresponding printing point position; S35. Construct a wall thickness error potential energy field diagram based on the wall thickness error potential energy value distribution obtained by the error mapping function. The wall thickness error potential energy field diagram is represented in the form of a two-dimensional matrix. Each matrix unit corresponds to a printing point position, and the wall thickness error potential energy value of the printing point position is recorded to form a continuous spatial distribution data structure for describing the size and directional change trend of the wall thickness deviation in each area.

5. The method for controlling the wall thickness uniformity of the combustion channel of the aerospace propulsion chamber additively manufactured according to claim 4 is characterized in that: The error mapping function is specifically expressed as: ； in, Indicates the current layer printing point The potential energy value of the wall thickness error at Indicates the current layer printing point The wall thickness error value at 、 、 、 and Indicates the preset control coefficient.

6. The method for controlling the wall thickness uniformity of the combustion channel of the aerospace propulsion chamber additive manufacturing according to claim 1 is characterized in that: The S4 specifically includes: S41. Normalize the wall thickness error potential energy field diagram to obtain the normalized wall thickness error potential energy value , the value range is between; S42, constructing a target scanning density function based on the normalized wall thickness error potential energy value ,make and There is a positive correlation and the integral conservation constraint is satisfied: ； in, Indicates the current layer printing area. Indicates the total amount of scan path resources in the current layer; S43, construct the Monge-Ampère equation with directional coupling and scan the density function with the target Solve the path potential function for the target term , the Monge-Ampère equation is: ； in, represents the determinant value, represents the Hessian matrix of the path potential function, represents the coupling coefficient, represents the directional coupling matrix, which is defined as: ； in, Indicates the direction of the wall thickness error gradient; S44, setting the path potential function boundary condition to the Neumann boundary condition, stipulating that the path potential function normal derivative is zero on the printing area boundary to maintain the boundary continuity of the path field; S45. Solve the Monge-Ampère equation to obtain the path potential function The two-dimensional distribution results of ; S46. Calculate the weighted gradient field vector for scanning path guidance: ； in, represents the weighted gradient field vector, and represents the control weight coefficient, represents the gradient of the path potential function, represents the gradient of the normalized wall thickness error potential field; S47 , constructing a pseudo-potential streamline field according to the directional distribution of the weighted gradient field vector as a scanning path for the current printing layer.

7. The method for controlling the wall thickness uniformity of the combustion channel of the aerospace propulsion chamber additively manufactured according to claim 1, characterized in that: The wall thickness prediction neural network model includes: The spatial feature encoding module is used to receive and process the actual contour image, wall thickness data, pseudo-potential streamline field and control parameters of the current layer, and extract multi-scale spatial features through a two-dimensional convolutional encoder; A temporal feature extraction module is used to load historical wall thickness error maps and control parameter sequences of multiple consecutive printing layers, and use a gated recurrent unit to extract the temporal evolution characteristics of the error changes; A feature fusion module is used to splice and fuse the multi-scale spatial features and the time evolution features in the feature dimension, and enhance the joint expression ability of local patterns and global trends through residual connections to generate fused joint features; The prediction generation module is used to generate a two-dimensional wall thickness deviation prediction matrix based on the fused joint features through a two-dimensional convolution decoder, wherein the two-dimensional wall thickness deviation prediction matrix represents the wall thickness deviation prediction value of the next layer in each scanning unit.

8. The method for controlling the wall thickness uniformity of a combustion channel in an aerospace propulsion chamber additively manufactured according to claim 1, characterized in that: The S6 specifically includes: Based on the predicted value of the wall thickness deviation of the next layer in each scanning unit, define the first threshold , the second threshold and the third threshold , dynamically update the various control parameters in the initial processing control parameter set of the next layer according to the set rules: If the wall thickness deviation prediction value is less than or equal to the first threshold , the change rate of the wall thickness deviation prediction value is less than or equal to the second threshold And the deviation fluctuation of the scanning unit of two consecutive layers is less than or equal to the third threshold , the control parameters remain unchanged and it is determined to be a stable area; If any of the conditions is met: the wall thickness deviation prediction value is greater than the first threshold , the change rate of the wall thickness deviation prediction value is greater than the second threshold , but the deviation fluctuation of the scanning unit in two consecutive layers is less than or equal to the third threshold , it is determined to be a slight correction area, and the control parameters are adjusted based on the wall thickness deviation prediction value and the preset adjustment factor, and the scanning path remains unchanged in the slight correction area; If the wall thickness deviation prediction value is greater than the first threshold , the change rate of the wall thickness deviation prediction value is greater than the second threshold And the deviation fluctuation of the scanning unit of two consecutive layers is greater than the third threshold , it is determined to be the key correction area, and the laser power, scanning speed, scanning spacing and spot size are adjusted according to the maximum correction amplitude upper limit ±15%, and the pseudo-potential streamline field is regenerated for the key correction area, and the scanning path is updated.

Images