A method and system for controlling the size of the side surface plating of a special-shaped edge of a printed circuit board

Abstract

The application provides a method and system for controlling the size of the irregular edge side of a printed circuit board. The method includes: analyzing the design file to extract the geometric element data of the target edge that needs side plating processing; partitioning the target edge at the geometric feature mutation, calculating the geometric feature parameters of each partition, and classifying them into pre-set edge type categories; predicting the edge current density distribution characteristics based on the geometric feature parameters and the type of each partition, and determining the compensation correction amount; calculating the differential compensation amount to obtain the compensation vector according to the type and geometric feature parameters of each partition combined with the compensation correction amount; and converting the compensation vector into the process parameters of the downstream processing procedure and establishing a cascade constraint relationship. The method applies differential compensation to irregular edges of different geometric types that match their current density distribution characteristics, eliminating the problem of local over-plating or under-plating under the unified compensation parameter mode, and ensuring consistency when the process is executed through cascade constraints.

Description

Technical Field

[0001] This application relates to the field of printed circuit board manufacturing technology, and in particular to a method and system for controlling the dimensionality of the side plating on irregularly shaped edges of printed circuit boards. Background Technology

[0002] In the manufacturing process of printed circuit boards, side plating is used to form a conductive plating layer on the sides of the board edges to meet the electrical connection requirements for gold finger soldering or inter-board splicing. When the board shape includes irregular edge structures such as arcs, right angles, and acute-angled notches, controlling the coverage size of the side plating faces technical challenges.

[0003] In existing methods for controlling the dimensions of irregular edge side plating, engineers manually interpret edge information in design documents to determine the plating area and uniformly set compensation parameters for all edge types. However, different geometric edge types exhibit varying plating distribution characteristics during electroplating, and the uniform compensation parameters cannot accommodate these differences, leading to insufficient or excessive plating coverage in areas such as the outer side of convex arcs and sharp corners. Furthermore, there is a lack of parameter transfer mechanisms between the edge analysis results at the design stage and subsequent milling and electroplating processes. Each process makes independent decisions, and the analysis results from preceding stages cannot effectively constrain subsequent stages. Summary of the Invention

[0004] To address the shortcomings of existing technologies where uniform compensation parameters for irregular edge side plating cannot adapt to the differences in plating distribution among edges with different geometries, and where there is a lack of parameter cascade constraints between processes, this application provides a method and system for controlling the dimensions of irregular edge side plating on printed circuit boards. This method automatically parses design files to extract target edges and classifies them according to geometric features. Based on the current density distribution characteristics of each zone, it determines differentiated compensation amounts and cascades the compensation parameters to downstream processing steps to form constraints. This enables precise control of side plating dimensions for irregular edges with different geometries. Specifically, this application provides the following technical solutions: This application provides a method for controlling the side plating dimensions of irregular edges on printed circuit boards, comprising: parsing design documents and extracting geometric element data of the target edge requiring side plating treatment; dividing the target edge into regions at abrupt changes in geometric features, calculating geometric feature parameters for each region and classifying them into a preset edge type category; predicting the edge current density distribution characteristics of each region based on the geometric feature parameters and the edge type category, and determining the compensation correction amount for each region; calculating the differentiated compensation amount for each region based on the edge type category and the geometric feature parameters, combined with the compensation correction amount, to obtain a compensation vector; converting the compensation vector into process parameters for downstream processing steps, and establishing a cascade constraint relationship between the compensation vector and the process parameters.

[0005] Optionally, the geometric feature abrupt change includes curvature abrupt change and angle inflection point; wherein, the curvature abrupt change is the position where the curvature difference between adjacent geometric elements exceeds the curvature abrupt change threshold, and the angle inflection point is the position where the angle between the direction vectors of adjacent straight line elements exceeds the angle inflection threshold.

[0006] Optionally, the edge type categories include six types: straight line segment, convex arc, concave arc, right angle, acute angle, and obtuse angle; wherein, based on the curvature value and interior angle value of each partition, each partition is assigned to the corresponding edge type category through a fractal decision tree.

[0007] Optionally, the calculation of the differential compensation amount for each partition includes: determining a baseline compensation amount from the calibrated baseline compensation coefficients according to the edge type category of the partition; selecting a corresponding correction function according to the edge type category, wherein for arc-type partitions, the curvature correction amount is determined according to the curvature parameter, and for corner-type partitions, the angle correction amount is determined according to the inner angle parameter; and combining the curvature correction amount or the angle correction amount with the baseline compensation amount to obtain the empirical compensation amount for each partition.

[0008] Optionally, the prediction of the edge current density distribution characteristics of each partition includes: mapping the edge geometry of each partition to a standard electric field configuration based on conformal transformation, solving the current density distribution function on the standard electric field configuration, determining the mean of excess current density of each partition based on the current density distribution function, and determining the compensation correction amount based on the mean of excess current density.

[0009] Optionally, the calculation of the differential compensation amount for each partition further includes: determining the fusion weight based on the geometric complexity of each partition, and weighting and fusing the compensation correction amount and the empirical compensation amount according to the fusion weight to obtain the final differential compensation amount for each partition.

[0010] Optionally, the process parameters of the machining process include the path offset of the contour milling process and the partition masking parameters of the electroplating process; the cascade constraint relationship includes the constraint that the combination of the path offset and the partition masking parameters does not exceed the equipment capability boundary.

[0011] Optionally, the method further includes: applying a length attenuation factor to partitions with a length less than a critical length to suppress overcompensation; and performing a smooth transition process at the boundary where the differential compensation amount of adjacent partitions jumps beyond a preset threshold.

[0012] Optionally, the method further includes: measuring the actual side coating coverage size during the production process, comparing the measured value with the design target value section by section to obtain the deviation value; when multiple consecutive batches of the same edge type show deviation in the same direction, correcting the reference compensation coefficient corresponding to that type according to the deviation value.

[0013] This application also provides a system for controlling the side plating dimensions of irregular edges on printed circuit boards, comprising: an edge extraction module for parsing design documents and extracting geometric element data of the target edge requiring side plating; a partition identification module for partitioning the target edge at abrupt changes in geometric features, calculating geometric feature parameters for each partition, and classifying each partition into a preset edge type category; a compensation prediction module for predicting the edge current density distribution characteristics of each partition based on the geometric feature parameters and edge type category of each partition, and determining the compensation correction amount for each partition; a compensation calculation module for calculating the differentiated compensation amount for each partition based on the edge type category and geometric feature parameters of each partition, combined with the compensation correction amount, to obtain a compensation vector; and a parameter binding module for converting the compensation vector into process parameters for downstream processing steps, and establishing a cascade constraint relationship between the compensation vector and the process parameters.

[0014] The beneficial effects of this application are as follows: By automatically parsing geometric elements in the design file and classifying target edges according to geometric features, the reliance on subjective experience in manual interpretation is eliminated, making the edge recognition process deterministic and repeatable. By applying compensation amounts matching the geometric feature parameters to different edge types, replacing the uniform compensation parameter mode in existing technologies, various irregular edges can receive compensation adapted to their electroplating layer distribution characteristics, eliminating local over-plating or under-plating. By using the edge current density distribution characteristics to predict the compensation correction amount, the current density concentration effect in the electroplating process is transformed from a phenomenon that needs to be suppressed into a basis for assisting in the calculation of compensation amounts, improving the compensation accuracy of complex geometric areas. By establishing a cascade constraint relationship between the compensation vector and the process parameters of downstream machining processes, the analysis results at the design end can directly constrain the parameter settings of subsequent milling and electroplating processes. The parameters of each process are interconnected rather than independently decided, ensuring consistency during process execution. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the method for controlling the plating dimensions of irregularly shaped edges of printed circuit boards according to embodiments of this application.

[0016] Figure 2 A sub-flowchart for design document parsing and target edge extraction provided for embodiments of this application.

[0017] Figure 3 This is a sub-flowchart for edge partition recognition and geometric feature parameterization provided in the embodiments of this application.

[0018] Figure 4 A sub-flowchart for current density distribution prediction and fusion decision-making provided in the embodiments of this application.

[0019] Figure 5 The following is a sub-flowchart for the compensation calculation and smoothing verification provided in the embodiments of this application.

[0020] Figure 6 This is a sub-flowchart for the cascading binding and verification of process parameters provided in the embodiments of this application.

[0021] Figure 7 This is a schematic diagram of a six-category edge type classification decision tree provided in an embodiment of this application.

[0022] Figure 8 This is a schematic diagram of the structure of the printed circuit board irregular edge side plating size control system provided in the embodiments of this application. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0025] To make the technical solutions and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and embodiments.

[0026] The method for controlling the side plating dimensions of irregularly shaped edges on printed circuit boards provided in this embodiment is applicable to boards with irregular edge structures such as arcs, right angles, acute angle notches, protrusions, and slopes. It enables differentiated control of side plating coverage dimensions throughout the entire process from design documents to production. This method operates within the engineering system and electroplating production line control system of printed circuit board manufacturing, using copper layer data and outline data from the design documents as input sources and board stack-up structure information as auxiliary parameters.

[0027] The execution environment of this method comprises two collaborative units: an engineering workstation and a production line control system. The engineering workstation is responsible for all calculations from S100 to S400, performing partition identification and compensation calculations for individual boards. The production line control system receives process parameter files and distributes these parameters to the CNC units of the milling and electroplating equipment. The two systems can communicate via a local area network to transfer data files.

[0028] This method applies to sheet metal specifications where the maximum external dimensions do not exceed [specific dimensions]. The thickness of the sheet metal is to Between, the minimum feature size of the irregular edge is not less than .

[0029] like Figure 1 As shown, this method includes the following steps: S100, parse the design file and extract the geometric element data of the target edge that needs side plating treatment.

[0030] like Figure 2 As shown, this step automatically identifies the edge regions that need side plating from the provided Gerber RS-274X format file set and extracts the corresponding geometric element sequence for each edge region.

[0031] S110: Parse the header parameters of the design file to determine the coordinate format and drawing tool definition.

[0032] The Gerber RS-274X format uses ASCII text encoding to display the graphic information of each layer of the printed circuit board. The parsing process reads the coordinate format parameter from the file header, which defines the integer number of bits used for the coordinate values. and decimal places The coordinate resolution is from Confirmed. For example, when the coordinate format is a 3-digit integer plus 6-digit decimal with units in millimeters, the coordinate resolution is [value missing]. Original integer value divided by This yields the actual coordinate values. Simultaneously, the D-code definition table is parsed to obtain the aperture parameters for each numbered drawing tool, where D01 is the exposure drawing mode, D02 is the movement-without-drawing mode, and D03 is the blinking mode.

[0033] The parsing of coordinate format parameters uses regular expressions to match the format declaration lines in the file header. There are two coordinate modes: absolute and incremental. This method uniformly uses absolute coordinate mode, and when incremental coordinate mode is detected, it is converted to absolute coordinates during the parsing stage. The D-code definition table establishes the correspondence between tool numbers and aperture geometry parameters. Aperture types include circular, rectangular, and elliptical. Side marking layers typically use a circular aperture, the diameter of which is the value used in the coverage determination. parameter.

[0034] The instruction stream processor needs to pay attention to the difference between multi-quadrant and single-quadrant modes when handling circular interpolation instructions. In multi-quadrant mode, the arc can span multiple quadrants, and the center offset points directly to the center; in single-quadrant mode, the arc is confined to a single quadrant. This method defaults to using multi-quadrant mode for parsing. For the special case where the start and end points coincide, it is interpreted as a complete circle in multi-quadrant mode and as a zero-length segment in single-quadrant mode.

[0035] The calculation of the length of geometric elements varies depending on the type. (Length of line segment elements...) The length of the arc element ,in The value is the central angle in radians, calculated by the angle between the starting vector and the ending vector, and determined according to the direction of rotation.

[0036] S120 parses the drawing instructions line by line, converts them into structured geometric elements, and forms an ordered chain of geometric elements that define the outline of the plate.

[0037] The instruction stream processor maintains a drawing state machine containing three state variables: current coordinate position, current interpolation mode, and current D-code. It scans the drawing instructions in the outline layer file line by line. When a G01 instruction is detected, the interpolation mode is set to linear interpolation, and the start and end coordinates are recorded to form a straight line segment. When a G02 instruction is detected, the interpolation mode is set to clockwise circular interpolation, and the start, end, and center offsets are recorded to form a clockwise circular arc element. When a G03 instruction is detected, the interpolation mode is set to counterclockwise circular interpolation, and counterclockwise circular arc elements are formed in the same way. After all instructions are parsed, an ordered chain of geometric elements is obtained. The elements in the chain are connected end-to-end to form the complete closed outline of the sheet metal.

[0038] For example, for a sheet metal part with irregular edges, its outline layer file contains four drawing instructions. The first instruction is the G01 linear interpolation instruction, with the starting coordinates being... The endpoint coordinates are , forming a length of The second element is the G02 clockwise circular interpolation command, starting from... The destination is The center offset is Therefore, the absolute coordinates of the circle's center are determined as the starting point plus an offset. radius of arc Articles 3 and 4 are both G01 linear interpolation instructions, which respectively form from... to The vertical line segment and from to The horizontal line segment. The ordered geometric element chain obtained after analysis contains 4 elements: line segment, clockwise arc, line segment, and line segment, with the first and last coordinates of the elements strictly connected.

[0039] S130, spatially overlay and match the side marking layer graphic with the outline, and extract the overlapping segment as the target edge of the side marking layer.

[0040] The side marking layer file is parsed in the same way to obtain the marking graphic data. The layer alignment engine performs an overlay determination for each geometric element in the outline: it calculates the coordinates of the midpoint of the element and determines the distance from the midpoint to the nearest graphic element in the side marking layer. ,when When the element is determined to belong to the side plating area, The aperture diameter of the D code used for the side marking layer. To match tolerance All the outline elements identified as side-plating areas are organized into an ordered set of geometric elements in their original order to obtain the target edge set for side plating. .

[0041] For example, continuing the previous example, suppose the side-plated identification layer has a D code with a diameter of... The drawing line is drawn along the first three shape elements. The midpoint of the first element is used as the starting point. The distance from the midpoint to the marker line is The judgment condition is This is correct; the element belongs to the side plating area. The midpoint of the second element is approximately... The distance from the marking line is approximately This also satisfies the judgment condition. For the midpoint of the third element... The distance from the marking line is The condition is met. The midpoint of the 4th element... The closest distance to the marking line is more than The condition is not met. Final output: Contains 1 target edge Its ordered geometric element chain is a sequence of the first 3 elements.

[0042] When the design file does not contain a side-plated marking layer, the target edge set is empty, the system outputs a prompt message and terminates the subsequent process.

[0043] The layer-based matching engine employs different distance calculation strategies for different geometric element types. For line segment markers, the distance from the midpoint to the marker line is calculated using the formula for the perpendicular distance from a point to a line segment. For arc markers, the distance from the midpoint to the circle containing the arc is combined with the arc's angle range for determination.

[0044] For complex sheet metal parts with multiple independent side-plated edges, when multiple consecutive adjacent elements in the outline belong to the side-plated area, they are organized into a single target edge. When there are interruptions due to elements not belonging to the side-plated area, they are segmented into independent edges at the interruptions. Matching tolerance. The settings take into account the safety margins of both cumulative coordinate accuracy error and interlayer alignment deviation.

[0045] Error handling is performed on abnormal data during the parsing process. When the relative difference between the distance from the start point to the center of the arc segment and the distance from the end point to the center exceeds... If the error is deemed to be due to data precision error, the arithmetic mean of the two radii is taken as the effective radius of the arc. When there are self-intersections in the outline, the outline is split into two independent closed outlines based on the first intersection point and processed separately. The geometric elements of each edge in the target edge set are organized into a doubly linked list structure according to the parsing order. Each node of the linked list stores the element's type identifier, start coordinates, end coordinates, center coordinates or direction vector, element length, and curvature value.

[0046] For sheet metal parts containing multiple independent side-plated edges, each edge is in the aggregate... The components are sorted counter-clockwise from their starting coordinates to ensure consistency in subsequent partitioning operations. When the same outline is covered by multiple side markings, they are merged into a single target edge for processing, avoiding duplicate calculations. (Information on sheet metal layer thickness is also included.) Obtain the corner partition radius from the borehole layer file or user parameter input for subsequent steps. and equivalent penetration depth All depend on this parameter.

[0047] S200, the target edge is partitioned at the point of geometric feature abrupt change, and the geometric feature parameters of each partition are calculated and classified into a preset edge type category.

[0048] like Figure 3 As shown, this step partitions each side plated target edge extracted by S100 according to geometric feature differences, and calculates quantified geometric feature parameters for each partition as input for subsequent compensation calculations. The partitioning is based on two types of geometric feature abrupt changes: curvature abrupt changes and angle transitions.

[0049] S210: Traverse the ordered geometric element chain of the target edge, calculate the curvature value of each element, and set partition boundaries at the curvature abrupt change points.

[0050] Traverse the target edge Ordered geometric element chain Calculate the curvature value for each element: the curvature of the line segment element. Curvature of circular arc elements ,in The radius of the arc is determined by the Euclidean distance between the coordinates of the starting point and the center of the circle, i.e. The unit is The curvature sign is defined as follows: when the center of the circle is located on one side of the solid part. This indicates a convex arc, where the center of the arc is located on the outer side of the plate. It represents a concave arc.

[0051] Calculate the curvature difference between adjacent elements pairwise. .when At that time, and The boundary of the partition is set at the intersection. Curvature abrupt change threshold. The range of values ​​is The determination was made through standard test piece calibration: standard test pieces containing straight segments and circular arc transitions of different radii were prepared. After performing the standard side plating process on each transition point, the differences in coating coverage and distribution were observed. The smallest adjacent element curvature difference corresponding to the visually distinguishable coverage difference was used as the standard. The calibration value.

[0052] The method for determining the curvature sign is based on the cross product of the solid orientation vector of the plate and the direction of the center of the circle. The solid orientation of the plate is defined as the left side along the direction of travel of the ordered geometric element chain. When the dot product of the direction vector from a point on the arc to the center of the circle and the direction inside the plate is positive, the curvature is positive, representing a convex arc; when it is negative, the curvature is negative, representing a concave arc.

[0053] For transition points where the direction changes abruptly but the curvature remains continuous, the curvature difference between two circular arcs of different radii in the same direction may be less than a threshold when they meet end to end. In this case, no partition boundary is set, and the two circular arcs are classified into the same partition. The equivalent curvature is taken as the arc-length weighted average. For example, and Convex circular arcs in the same direction, curvature difference They are grouped into the same partition.

[0054] For example, when When, the line segment abruptly changes to a radius. The arc triggers partition cutting; the lower limit of this threshold Corresponding radius Sensitivity for transition detection between arcs and straight lines.

[0055] S220 detects the angular transitions between adjacent line segments and marks the angular regions at the transition points.

[0056] For two adjacent line segments in an ordered geometric element chain, calculate their direction vectors. and The exterior angles of two vectors The corresponding interior angle . When the outside corner At this point, the boundary is marked as a corner region. Angular turning threshold. The range of values ​​is .

[0057] The corner division is defined by extending the radius of the corner division to both sides from the corner point. The area, in which , The thickness of the sheet metal is expressed in units of 1. The physical basis for this value is that the main area of ​​influence of electric field distortion at the corner is approximately half the thickness of the plate, and the divergence range of the electric field lines at the corner is limited by the cross-sectional dimensions of the plate. For example, when the plate thickness... hour, That is, each side of the corner point The edges of the area are designated as the jurisdiction of the corner district.

[0058] S230 calculates the three-dimensional geometric feature parameters of curvature, internal angle and segment length for each partition, and classifies each partition into six edge type categories based on the three-dimensional features.

[0059] like Figure 7 As shown, the fractal decision tree determines the type of each partition according to priority hierarchy. For the corner partition marked by S220, the type is determined based on the included angle. Specific value determination: Classified as acute angle type Classified as a right angle type Classified as obtuse angle type. For non-angular regions, based on curvature value... determination: Classified as a convex arc type, Classified as a concave arc type, It is classified as a straight line segment.

[0060] The precise coordinates of the partition boundary points are determined using the following method. For boundaries with abrupt changes in curvature, the boundary point is the connection point between two adjacent geometric elements. The area of ​​the corner partition extends to both sides. The extension measurement is performed along the arc length direction of the edge curve. For circular arc elements, the arc length extension... The corresponding central angle increment is When the corner partition extends beyond the current element and enters an adjacent element, the extension is truncated to the end of the current element, and the remaining extension continues to be applied to the adjacent element.

[0061] 3D geometric feature parameters of each partition The required accuracy for calculation is as follows: curvature should be retained to 4 significant digits, and angles should be retained to 4 significant digits. Precision, length retained Accuracy. These accuracy requirements are matched to the sensitivity of the correction function in subsequent compensation calculations.

[0062] For a circular arc with a very small radius, when When the angle is an arc, its behavior is close to that of a corner, so it is classified as an acute angle and the equivalent interior angle is assigned a value of 1. ,in This is the central angle corresponding to the arc.

[0063] For lengths less than the minimum partition length The partitions are merged as follows: if adjacent partitions on both sides of a shorter partition are of the same type, they are merged into either side; otherwise, they are merged into the longer side. After merging, the weighted average curvature of the target partition is recalculated. and merge length .

[0064] For example, continuing the target edge from the previous example It contains 3 geometric elements. The first element is a line segment. The second element is a clockwise arc with a radius of... , A positive curvature indicates a convex circular arc; the third element is a straight line segment. The curvature difference between adjacent elements is and All of these trigger partition cutting. Simultaneously, there is a gap between the 3rd and 4th elements (which belong to the non-side-plated area and are not considered here). Right-angle turns are marked as corner partitions. This ultimately yields the partition feature set. It contains 4 partitions: It is a line segment type. , , ; It is a convex arc type. , , ; It is a line segment type. , , ; It is a right-angle type. not applicable, , .

[0065] The complete decision rules of the fractal decision tree are executed in the following priority order. The highest priority is corner type detection: if a partition is marked as a corner partition by S220, the decision is based on the internal angle... Directly categorized The angle is acute. The degree is a right angle type. The angle is obtuse. The next priority is curvature type detection: check the curvature value of non-angular regions. , Classified as a convex arc type, Classified as a concave arc type, Classified as a line segment. Threshold. Corresponding radius For arcs with a radius greater than this, the current concentration effect is negligible.

[0066] For a short straight line segment or short circular arc segment located between two corner regions, when its length When there is overlap in jurisdiction, the entire segment is classified as a corner region, and the type is determined by the smaller of the two inner angles of the two corners. Consecutive regions of the same type are merged, and the curvature of the merged region is calculated as a length-weighted average, with the length being the sum of the lengths of the segments. For example, two adjacent convex arc segments... and The curvatures are respectively and The lengths are respectively and Curvature after merging Length after merging .

[0067] Plate thickness The impact on partition recognition is reflected in the corner partition radius. The value that can be taken. When hour The corners have a smaller impact range; when hour The influence range of the corner expands, and the short straight line segments between adjacent corner points may be completely absorbed into the corner region.

[0068] S250, based on the geometric feature parameters and edge type category of each partition, predict the edge current density distribution characteristics of each partition, and determine the compensation correction amount of each partition.

[0069] like Figure 4 As shown, this step uses equipotential line theory to predict the edge current density distribution of each zone during the electroplating process, transforming the current density concentration effect from a harmful phenomenon that needs to be suppressed into a physical basis for accurately calculating the plating layer expansion width.

[0070] S251, the edge geometry of each partition is equivalent to a two-dimensional electric field boundary condition, and the corresponding conformal transformation mapping function is selected according to the edge type.

[0071] The edge geometry configuration mentioned in this step refers to the two-dimensional conductor boundary shape presented by the edge contour lines of each partition on a cross-section perpendicular to the plate surface. Specifically, the edge geometry configuration of the straight segment partition is a semi-infinite planar conductor boundary, the edge geometry configuration of the convex arc partition is an outwardly convex cylindrical conductor boundary, the edge geometry configuration of the concave arc partition is an inwardly concave cylindrical conductor boundary, and the edge geometry configuration of the corner partition is the included angle between two straight conductor boundaries. The wedge-shaped conductor boundary formed by the intersection.

[0072] The process of equivalencing the above edge geometry to two-dimensional electric field boundary conditions is based on the following assumptions to establish an equivalent electric field model. Assumption 1: Far-field uniformity condition: The distance from the anode to the cathode edge in the electroplating tank is much larger than the partition feature size. The initial electric field generated by the anode can be considered a uniform electric field near the cathode edge, and the direction of the electric field intensity points towards the cathode surface along the macroscopic normal direction of the cathode surface. Assumption 2: Cathode equipotential condition: The conductor surface on the side of the plate edge is an equipotential surface, serving as the first type of boundary condition (Dirichlet boundary condition) for the Laplace equation, with the potential value set to zero. Assumption 3: Insulating boundary condition: The normal current density in the resist coating region covering the upper and lower surfaces of the plate and the exposed non-conductor surfaces is zero, serving as the second type of boundary condition (Neumann boundary condition) for the Laplace equation.

[0073] Based on the above assumptions, the standard electric field configurations corresponding to the edge geometry of each partition are as follows. For the straight-segment type partition, the standard electric field configuration is a semi-infinite planar conductor in a uniform electric field, with the cathode surface being a straight equipotential boundary, electric field lines perpendicularly incident on the conductor surface, and the current density being equal everywhere. For the convex circular arc type partition, the standard electric field configuration is a cylindrical conductor cathode in a uniform far field, with the cathode surface being a radius... The convex arc equipotential boundary causes electric field lines to converge at the edge region, resulting in a higher current density on the outer side of the conductor surface compared to the far-field reference value. For the concave arc type partition, the standard electric field configuration is an electric field distribution concave inside a cylindrical surface. The cathode surface is an inwardly curved arc equipotential boundary, and the concave surface causes electric field lines to diverge at the edge region, resulting in a lower current density on the conductor surface compared to the far-field reference value. For the diagonal type partition, the standard electric field configuration is an electric field distribution near the tip of a wedge-shaped conductor cathode, with two intersecting conductor surfaces forming an outer angle. The wedge-shaped equipotential boundary causes the electric field lines to converge at a high degree at the wedge tip, resulting in a singular current concentration.

[0074] The current distribution during electroplating follows the Laplace equation. For the aforementioned standard electric field configurations, conformal transformations are used to map complex boundaries to simple boundaries in a uniform electric field, allowing direct solution of the current density distribution on the transformed standard plane. The Laplace equation exhibits form invariance under conformal transformations, and the magnitude of the transformed derivative... The amplification factor equal to the current density, i.e. ,in This represents the uniform current density on the transformed standard plane. The above two-dimensional conformal transformation model is applicable to scenarios where the plate thickness is much smaller than the edge feature size. When the plate thickness is comparable to the partition length, the three-dimensional edge effect is compensated by tilting the fusion weights towards the empirical model.

[0075] For the straight line segment type partitioning, semi-infinite planar conductors in a uniform electric field do not require geometric transformations and can be transformed using identity transformations. The magnitude of the transformed derivative is Uniform current density distribution There is no concentrated effect. The convex and concave circular arc types are partitioned, and the cylindrical conductor boundary is mapped to a straight boundary using a logarithmic transformation. ,in Using complex plane coordinates with the center of the arc as the origin, this transformation maps the arc's equipotential lines to straight equipotential lines, and the electric field problem after the transformation degenerates into the electric field of a parallel plate. Diagonal type partitioning maps the wedge-shaped conductor boundary to a half-plane boundary using a power transformation. ,in Representing exterior angles in radians, this transformation will change the angle to... The wedge-shaped region unfolds at an angle of The electric field problem in the half-plane degenerates into a uniform electric field problem in a semi-infinite plane conductor.

[0076] S252, solve the current density distribution function on the transformed standard configuration to determine the mean excess current density of each partition.

[0077] For convex arc type partitions, the normal depth along the thickness direction of the plate. from To equivalent penetration depth Within the range, the current density distribution is as follows The engineering approximation of the mean excess current density obtained by performing area-average integration on this distribution is: ,in For the curvature of the partition. This linear approximation is in The relative deviation between the time and the exact solution is less than Equivalent penetration depth The range of values ​​is The determination was made through back calibration using standard test pieces.

[0078] For concave arc type partitions, the mean excess current density is A negative value indicates that the current density in the edge region is lower than the reference value.

[0079] Diagonal partitioning, with current density near the corners exhibiting a power-law distribution. ,in The distance from the corner point The radius of the corner section, Let be the current concentration index. Under typical cutoff conditions in printed circuit board manufacturing, the closed-form solution for the area-weighted mean of the corner excess current density is: For example, right-angled partitions. , , , ; partitioning acute angles , , , .

[0080] Partitioning by line segment type The current density does not exhibit excessive distribution.

[0081] S253 determines the compensation correction amount for each partition based on the average excess current density.

[0082] Substituting the average excess current density into the calculation formula for the compensation correction, the compensation correction for each zone is obtained. (Hereinafter also referred to as physical prediction compensation amount) ): in This is the baseline compensation coefficient for the type to which this partition belongs. The excess current density conversion efficiency factor has a range of values. , For the target coating thickness, in units Baseline compensation coefficient Obtained through calibration using standard test pieces, its physical meaning is the proportional coefficient of the coating width expansion per unit of target plating thickness. Excess current density conversion efficiency factor. The conversion ratio of excess current density to the lateral expansion width of the coating is determined through a two-stage sequential calibration protocol: the first stage is determined independently. The second phase is based on the established... Reverse push .

[0083] Partitioning by line segment type The physical prediction compensation amount degenerates into pure baseline compensation. Set non-negative baseline constraints for concave arc type partitions. This is to prevent excessively small predicted values ​​under extreme conditions.

[0084] For example, partitioning a convex circular arc , , , .Pick , , Substituting into the formula, we get For acute angle partitioning, ,Pick Substituting into .

[0085] The derivation process for the above approximation of excess current density on a convex circular arc is as follows. With a radius of... Cylindrical cathode surface, normal depth The current density at that point is obtained from Gauss's law. The denominator For depth The radius of curvature of the equipotential surface. Excess current density is defined as... For the normal depth from arrive The average integral of the area of ​​the interval execution The exact solution to this integral is ,in .when Performing a Taylor expansion on the logarithmic terms and substituting the values, we obtain the engineering approximation. .

[0086] The execution method of the two-phase sequential calibration protocol is as follows: the first phase is determined independently. Three standard test pieces of convex circular arcs with different radii were prepared, and the actual coating spread width was measured. As a benchmark, the portion exceeding the benchmark is attributed to curvature effects, and the fit is... This minimizes the mean squared error between the model's predicted and measured values. The second stage is based on the determined... Reverse push Calculate the average theoretical excess current density of each test piece, and then... Fitting the data with the only free parameter ensures that the predicted values ​​match the measured values.

[0087] For example, take The convex arc standard test piece, target plating thickness Actual measured extended width Baseline forecast Excess portion .Pick , ,but The single set of back-calculated values ​​is outside the valid range. The reason is that the measurement error and parameter coupling effect of a single test piece make single-point backtesting unstable, requiring joint fitting of multiple sets of test pieces with different radii to eliminate individual biases. A stable combination is obtained through joint fitting of multiple sets of test pieces. and .

[0088] The derivation of the excess current density for diagonal-type partitions is as follows. The current density near the corners exhibits a power-law distribution. This originates from the exact solution of the two-dimensional Laplace equation under wedge-shaped geometric boundary conditions. The area-weighted average integral in... And the cutoff radius ratio The engineering approximate closed solution at time is The relative deviation from the exact numerical integral is less than ,in The lower limit is truncated for the value at the corner point, indicating that the distance from the corner point is less than the minimum value. Singular regions are not included in the area integral.

[0089] The applicability boundary of the above physical model is constrained by three conditions. Condition one is the effective domain of the linear approximation: the engineering approximation requirements for convex circular arcs. When the product exceeds this threshold, the relative deviation of the linear approximation increases rapidly, at which point the system switches to the exact logarithmic integral formula for calculation. Condition two is a geometric scale constraint: partition length. Must be greater than This ensures that the current density distribution within the partition has fully developed to a steady state, and that the edge transition zone does not occupy the main body of the partition. Condition three is the linear response assumption of the electroplating process: the model assumes that the lateral expansion width of the plating layer is linearly proportional to the excess current density. This assumption applies to the target plating thickness. This holds true under normal process conditions. However, when the plating thickness exceeds this range, nonlinear effects emerge, necessitating the introduction of a second-order correction term. For concave arc-type partitions, the physical mechanism by which the current density is lower than the baseline value lies in the fact that the concave geometry causes the equipotential lines to be sparsely distributed in the edge region, reducing the current received per unit area on the cathode surface and suppressing the lateral expansion of the coating. The strength of the concave current shielding effect is inversely proportional to the radius of the concave arc; the smaller the radius, the stronger the shielding, and the higher the average excess current density. The negative value deepens as the absolute value of curvature increases. When the physical prediction compensation drops to an extremely low value due to the shielding effect, the non-negative floor constraint... Effective to prevent underplating.

[0090] S300, based on the edge type category and geometric feature parameters of each partition, and combined with the compensation correction amount, calculate the differentiated compensation amount for each partition to obtain the compensation vector.

[0091] like Figure 5 As shown, this step selects the corresponding empirical compensation calculation formula for each partition based on its type label, and then weights and fuses the empirical compensation amount with the physical prediction compensation amount of S250 according to the fusion weight to obtain the final differentiated compensation amount. All compensation amounts All are defined as the reserved width of the copper layer extending from the design outline to the outside of the board in the board direction, in units of It is always a positive value.

[0092] Before performing compensation calculations, the system retrieves various types of benchmark compensation coefficients from the benchmark parameter library. and equipment capability boundary matrix The baseline compensation coefficient is obtained through calibration using standard test pieces: at least three standard test pieces of each of the six categories—straight line segments, convex arcs, concave arcs, right angles, acute angles, and obtuse angles—are prepared. After performing the standard side plating process on the standard test pieces, the actual plating coverage width is measured using metallographic sectioning. The baseline compensation coefficient for each type is then determined. ,in This represents the average measured coating width of this type of test piece. The target coating thickness.

[0093] S310 calculates the baseline compensation amount for each partition and applies the corresponding correction function according to its type to obtain the empirical compensation amount.

[0094] The benchmark compensation coefficient mentioned in this step As described in S253 All parameters are obtained through standard test piece calibration and represent the coating width expansion ratio coefficient corresponding to each unit of target coating thickness.

[0095] For each partition According to its type, refer to the benchmark compensation coefficient table Retrieve corresponding coefficients Calculate the benchmark compensation amount .

[0096] Apply curvature correction function to arc-type partitions The correction function for convex circular arc type is: The concave arc type is Non-circular arc type .in As a reference radius, Sensitivity is corrected for convex circular arcs. Sensitivity is corrected for concave circular arcs. The range of values ​​for the curvature correction factor is limited to [value missing]. When the value exceeds the boundary, the boundary value is used.

[0097] Apply angle correction function to corner type partitions ,in This refers to angle correction sensitivity. The range of values ​​for the angle correction factor is limited to... For non-corner type partitions, take... .

[0098] When partition length Less than the critical length When applying a length attenuation factor To prevent overcompensation in short partitions. partition take .

[0099] The experience compensation amount for each partition is .

[0100] For example, partitioning a convex circular arc , , Baseline compensation amount Curvature correction: , This section is an arc-shaped, non-corner part. Partition length , Experience compensation amount .

[0101] Curvature correction factor value range The engineering basis is as follows. Lower limit Corresponding concave arc radius For concave circular arcs with radii smaller than this, the current shielding effect is extremely strong, and further reducing the compensation amount may lead to insufficient coverage. Upper limit Corresponding convex arc radius Exceeding this curvature and further increasing the compensation amount will introduce the risk of overcompensation. The range of angle correction factor values... The upper and middle limits correspond to the truncation of the extremely acute angle region.

[0102] The four-factor concatenation multiplication of empirical compensation ensures that each correction effect acts independently: curvature correction only applies to arc types, angle correction only applies to corner types, and length decay only applies to short partitions. The priority mechanism of the fractal decision tree ensures that each partition belongs to only one type, and there are no composite correction scenarios.

[0103] S320 determines the fusion weights based on the geometric complexity of each partition and performs weighted fusion of physical prediction compensation and empirical compensation.

[0104] The fusion formula is Among them, the fusion weight Geometric complexity index of partitions Decide: Geometric complexity index ,in This is the upper limit of saturation for the product of curvature and length. For reference complexity constants. In the angle terms of non-corner type partitions. Unified The angle term is .

[0105] The more complex the geometry of the partition, the more... The larger the value, the higher the fusion weight. The closer The more a region relies on the predictions of a physical model, the simpler its geometry. The closer The more it relies on the stability of empirical models, the more it depends on the type of line segment. and , , It uses an empirical model entirely.

[0106] Geometric complexity index The design logic lies in comprehensively considering the curvature and angle effects of the partitions. (Curvature-length product) Reflects the cumulative intensity of the curvature effect, saturation upper limit To prevent extreme parameter combinations from producing unreasonably high complexity values. Angle term. This reflects the enhancing effect of corner sharpness on current concentration.

[0107] Reference complexity constant The meaning is: when Time-fusion weights The physical model and the empirical model each contribute half. For high-complexity partitions, the empirical model exceeds the calibration range and its accuracy decreases, making the physical model more reliable; for low-complexity partitions, the current concentration effect of the physical model is negligible, and the empirical model is accurate enough.

[0108] For example, partitioning a convex circular arc , , Fusion results Divide the straight line segment into sections. , , .Pick , .

[0109] S330 performs a smooth constraint check on the jump in compensation amount between adjacent partitions and outputs the final compensation vector.

[0110] Check the jump variables of compensation amounts in adjacent partitions When the jump exceeds Linear transition interpolation is performed at partition boundaries. Boundary checks are also performed on the final compensation values ​​for all partitions. Must Within the range, take the value below the lower limit. Exceeding the upper limit triggers a process feasibility warning.

[0111] Final output compensation vector For example, the compensation amount for each partition after fusion calculation in this case is: This corresponds to four partitions: line segments, convex circular arcs, line segments, and right angles. Perform smooth constraint verification: the maximum value of adjacent jumps is... No smoothing is required. Perform boundary checks: all partitions are within... Within the specified range, there is no truncation. Output the final compensation vector. .

[0112] Benchmark compensation coefficient The typical value ranges for each type are as follows. Line segment type The effect is mainly influenced by the composition of the electroplating solution and the anode distance. Convex arc type The reason it's higher than the straight section is because the current concentration effect of the convex curved surface enhances the lateral expansion of the coating. (Concave arc type) The lower angle than the straight line segment is due to the current shielding effect of the concave surface suppressing lateral expansion. Right-angle type. acute angle type obtuse angle type .

[0113] Curvature Correction Sensitivity The basis for the value: Convex arc, correction factor This indicates that the compensation amount has been amplified. Measured data shows that the coating extension width of this curvature arc is approximately [a fraction] of that of the straight line segment. to times, This ensures the predicted value falls within the center of the measured range. Concave arc correction sensitivity. Slightly lower This reflects the physical asymmetry that the current shielding effect of a concave surface is weaker than the current concentration effect of a convex surface.

[0114] Angle correction sensitivity Make the right-angle partition correction factor acute angle Degree partitioning obtuse angle Degree partitioning These values ​​match the measured data of the standard test piece: the coating extension width of the right-angled test piece is approximately [missing value] of the straight line segment. to The acute angle test piece is approximately times, and the acute angle test piece is approximately to The obtuse-angled specimen is approximately [number] times ... to times.

[0115] The smooth transition process is implemented by establishing a width of [missing information] at the boundary of adjacent partitions. The compensation amount is linearly interpolated within the transition interval to ensure the continuity of the milling path offset at the partition boundary.

[0116] S400, the compensation vector is transformed into process parameters for downstream processing steps, and a cascade constraint relationship is established between the compensation vector and the process parameters.

[0117] like Figure 6 As shown, this step transforms the partition compensation vector calculated by S300 into specific process parameters for the profile milling and electroplating processes, and establishes cascaded constraint relationships between parameters to ensure execution consistency.

[0118] S410, generate the path offset of the profile milling operation based on the compensation vector.

[0119] For each partition Milling path offset ,in For the selected milling cutter diameter, in units This offset is defined as the normal distance from the center trajectory of the milling cutter to the designed outline. The accuracy of the milling path offset is limited by the milling positioning accuracy. When the offset difference between adjacent partitions is less than At this time, smaller offsets are uniformly increased to larger values ​​to eliminate the risk of insufficient accuracy.

[0120] The milling path is generated segment by segment in the partitioning order. For straight segment types, the offset path is the original straight segment translated along the normal direction. Parallel line segments afterwards. For convex arc type partitions, the offset path is an arc concentric with the original arc, with an increased radius. That is, the offset radius For concave arc types, partition by offset radius. ,when The error was initially identified as milling cutter interference, triggering a failed V2 check. For diagonal type partitioning, the offset path extends along the angle bisector at the corner point. This forms a transition arc or polygonal line connection. The connection point of adjacent offset paths is determined by geometric intersection calculation. When two offset paths have no intersection, a connecting arc is inserted to achieve a smooth transition. The milling feed rate is set differently according to the partition type: the standard feed rate is used for straight segment partitions. The circular arc partition and corner partition reduce the feed rate to To ensure uniform cutting.

[0121] For example, a milling cutter is selected. For the compensation vector The milling path offset for each partition is Partition and The offset difference between them is Greater than milling positioning accuracy Each maintains its independent value. (Partitioning) and The offset difference between them is Also greater than Each retains its independent value.

[0122] S420 generates partitioned masking parameters for the electroplating process based on the compensation vector.

[0123] Based on the boundary coordinates and compensation values ​​of each partition in the compensation vector, the contour data of the electroplating masking mold is generated. The opening width of the masking mold corresponds to the compensation amount of each zone, and the opening position corresponds to the coordinate range of the zone on the edge of the sheet metal. The alignment deviation between the masking mold contour and the zone boundary is controlled within... Within.

[0124] The geometric design of the masking mold follows these rules: each partition corresponds to an opening window on the mold, and the width of the window in the plate direction is equal to... ,in For the thickness of the sheet metal, allowance is provided on both sides. The coating extension area. The length of the window along the edge direction is equal to the edge projection length of the partition. The width of the isolation strip between adjacent windows shall not be less than [amount missing]. This ensures the mechanical strength of the mold structure. When the difference in compensation between adjacent zones is less than... At this time, the corresponding windows are merged into a continuous opening to simplify mold processing. The positioning datum hole of the mold is aligned with the process positioning hole of the sheet metal, using a two-pin, one-face positioning method. The tolerance fits of the main positioning pin and the secondary positioning pin are as follows: and Ensure that the alignment deviation between the mold window and the partition boundary is within the specified range after clamping. Within.

[0125] S430 establishes a constraint relationship where the combination of milling path offset and partition occlusion parameters does not exceed the equipment capacity boundary.

[0126] Perform a process consistency check matrix to verify the following constraints: the milling cutter center trajectory corresponding to the milling path offset does not interfere; the electroplating current density is within the equipment's allowable range. The total width of the compensated edges does not exceed the design safety clearance of the plate. ,in This represents the spacing between the nearest neighboring graphics. All parameter combinations must not exceed the device capability boundary matrix. When the output is within the specified range, it passes; when any combination of parameters exceeds the limit, it outputs a warning and provides correction suggestions.

[0127] Equipment capability boundary matrix Record the physical limit parameters of the milling and electroplating equipment. Milling equipment constraints include: minimum path offset not less than... The maximum path offset does not exceed The offset jump between adjacent path segments does not exceed the milling positioning accuracy. The electroplating equipment constraints include: the minimum width of the shielding mold opening is not less than [number missing]. (Machining accuracy limitations) Maximum width not exceeding (Due to space limitations in the electroplating tank), the total opening area for simultaneous electroplating of a single board shall not exceed the effective area of ​​the cathode. (Current distribution uniformity requirement). The verification matrix is ​​executed in priority order from V1 to V5. If any verification fails, the subsequent verification will be terminated and a detailed report of the first failed item will be output. The report includes the specific parameter value of the failure, the over-limit amplitude, and the suggested parameter adjustment direction.

[0128] When verification fails, the system performs parameter rollback correction. For V2 concave region interference failure, the correction strategy is to reduce the compensation amount of that concave arc section to... The corresponding maximum allowable value is displayed, and a warning is output to the user indicating that the actual compensation amount for the partition is lower than the calculated optimal value. For V3 safety clearance failures, the correction strategy is to truncate the compensation amount for the out-of-limit partition to... The feasibility of applying the V4 current density failed. The corrected strategy is to adjust the opening area of ​​the shielding mold to reduce the total electroplating area back to within the allowable range of the equipment. The corrected parameters need to be re-executed with a complete verification process to confirm that there are no new conflicts.

[0129] After all verifications pass, the system encapsulates the milling path offset parameters and electroplating masking parameters into a process parameter file for output. The milling parameter file contains the start coordinates, end coordinates, arc parameters, and normal offset of each path segment, in a format conforming to the path description specifications of CNC machining centers. The electroplating parameter file contains the outline coordinate sequence of the masking mold and the dimension annotations of each window, in a format conforming to the data exchange standards for mold machining drawings. The two files establish a correspondence through a common partition number index, ensuring that the partition boundaries of the milling path are strictly aligned with the window boundaries of the masking mold.

[0130] S500 measures the actual side plating coverage size during the production process, compares the measured value with the design target value section by section to obtain the deviation value, and when multiple batches of the same edge type show deviation in the same direction, the reference compensation coefficient corresponding to that type is corrected according to the deviation value.

[0131] For each zone edge of the first piece and the sampled pieces, the actual coating coverage width is measured using optical measuring equipment. Calculate the deviation for each zone. Positive values ​​indicate overplating, and negative values ​​indicate underplating. The tolerance is... Anything exceeding this limit is considered out of tolerance.

[0132] Perform attribution analysis on the out-of-tolerance zones: if only the corner type zones are over-plated, the angle correction factor is too large; if only the arc type zones are over-plated, the curvature correction sensitivity is too large; if all types are over-plated, the electroplating process parameter deviation is the cause.

[0133] When multiple consecutive batches of the same edge type category exhibit deviations in the same direction, online correction of the baseline compensation coefficient is triggered. The threshold for determining consecutive batches is set based on the statistical characteristics of the production line, exemplarily set at 3 batches; for high-precision production lines, this can be tightened to 2 batches, and for production lines with large batch fluctuations, it can be relaxed to 5 batches. The correction formula is as follows: ,in The range of values ​​for the step size factor is as follows: Recommended value The single correction amplitude shall not exceed Current value ,Right now If three consecutive batches exceed the tolerance and remain outside the tolerance after correction, a forced shutdown for re-inspection will be triggered.

[0134] The attribution decision tree covers both over-plating and under-plating directions. Over-plating direction: Over-plating only at corners is identified as an excessively large angle correction factor; over-plating only at convex arcs is identified as an excessively large curvature correction sensitivity; over-plating of all types is identified as a deviation in electroplating process parameters. Under-plating direction: Under-plating only at corners is identified as an excessively small angle correction factor; under-plating only at concave arcs is identified as excessive correction; under-plating of all types is identified as a deviation in process parameters.

[0135] The shutdown conditions follow a two-level progressive principle. Level 1: Forced shutdown. A shutdown command is output when three consecutive batches exceed the tolerance and three consecutive batches still exceed the tolerance after correction. Level 2: Early warning shutdown. [The last part, "certain type," appears to be an unrelated fragment and is omitted from the translation.] The cumulative correction amount exceeds the initial value A prompt to recalibrate will appear. After recalibration, [the following will occur]. Reset to the new value, and clear the cumulative correction amount to zero.

[0136] In another embodiment, the process monitoring employs statistical process control methods to perform trend analysis on the deviation data. Control charts are plotted for deviation values ​​of various partitions. Preventive correction is triggered when the deviation data shows a continuous seven-point increasing or decreasing trend, with the correction step size factor set to... Suppress the trend in a gentle manner and intervene early before the deviation reaches the tolerance limit.

[0137] For example, suppose a batch of convex arc type partitions is over-plated for 3 consecutive batches, with an average deviation Correction calculation Exceeding the limit value After the amplitude limit is implemented .

[0138] The output format of the milling path offset data is an offset command recognizable by the CNC machining center. A path segment definition is generated for each partition, including the start coordinates, end coordinates, and normal offset of the segment. Path segments are connected end-to-end through coordinate continuity. When the offsets of adjacent partitions are identical after precision standardization, the path segments are merged into a single continuous segment to simplify the program.

[0139] The design rules for the electroplating partition masking parameters are as follows. For each partition, the opening width of the masking mold is equal to... The opening length is equal to the projected length of the partition on the plate surface. The mold material is an insulating material resistant to electroplating solution corrosion, and the machining accuracy requirement for the opening edge is [specific value missing]. The positioning of the mold and the sheet metal is achieved using a combination of locating pins and fixtures, with alignment accuracy controlled within [specific parameters]. Within.

[0140] The consistency verification matrix contains 5 verification rules. V1 Milling Accuracy Verification: The difference between adjacent offsets must not be less than the milling positioning accuracy; otherwise, it will be processed uniformly. V2 Concave Area Interference Verification: The center trajectory of the milling cutter in the concave arc segment must not break through the inner side. V3 Compensation Width Safety Verification: The compensated edge does not exceed the safety distance. V4 Current Density Feasibility Verification: The standard current density is within the allowable range of the equipment. V5 Safety Distance Verification: The maximum compensation amount is less than half the distance between adjacent graphics.

[0141] For example, for milling cutter Offset of each partition V1 check: Adjacent differences and All greater than No uniformity is required. V3 verification: Passed. Output the process parameter file after all 5 items pass.

[0142] like Figure 8 As shown, this application also provides a printed circuit board irregular edge side plating size control system, including an edge extraction module, a partition recognition module, a compensation prediction module, a compensation calculation module and a parameter binding module.

[0143] The edge extraction module is used to parse the design file and extract the geometric element data of the target edges that require side plating. This module runs on an engineering workstation and includes a file header parser, an instruction stream processor, a geometric element constructor, and a layer alignment engine. The file header parser extracts the coordinate format and D-code definition table; the instruction stream processor maintains a drawing state machine to parse drawing instructions line by line; the geometric element constructor constructs line segment or arc segment objects according to the instruction type and parameters; and the layer alignment engine performs spatial overlay matching between the outline layer and the side plating mark layer.

[0144] The partitioning identification module is used to partition the target edge at abrupt changes in geometric features, calculate geometric feature parameters for each partition, and classify it into a preset edge type category. This module internally includes a curvature calculator, a partition boundary detector, a partition cutter, a feature extractor, a fractal engine, and a short partition merger, transforming the original chain of geometric elements into a set of partitions with type labels and quantization features.

[0145] The compensation prediction module is used to predict the edge current density distribution characteristics of each partition based on the geometric feature parameters and edge type category of each partition, and to determine the compensation correction amount for each partition. This module internally includes a geometric equivalent modeler, a transformation function selector, an analytical solution calculator, a coating width integrator, and a fusion decision unit, and predicts the coating distribution based on the first principles of electroplating physics.

[0146] The compensation calculation module is used to calculate the differentiated compensation amount for each partition based on the edge type category and geometric feature parameters of each partition, combined with the compensation correction amount, to obtain a compensation vector. This module selects a correction function according to the partition type and performs empirical compensation amount calculation, then fuses it with the physical prediction compensation amount.

[0147] The parameter binding module is used to transform the compensation vector into process parameters for downstream processing operations and establish a cascaded constraint relationship between the compensation vector and the process parameters. This module internally includes a milling offset calculator, a masking mold designer, a consistency verification engine, and a data formatter, which converts the compensation parameters into a format acceptable to the production line system and performs equipment capability boundary checks.

[0148] Each of the above modules can be implemented by the processor executing computer program instructions in memory.

[0149] The edge extraction module and the partition recognition module transmit the set of side-coated target edges. Each edge records its start coordinates, end coordinates, and an ordered chain of geometric elements. The partition feature set is transferred between the partition identification module and the compensation prediction module. This includes the partition number, type label, curvature value, interior angle, segment length, and boundary coordinates. The compensation prediction module passes the physical prediction compensation amount and fusion weights to the compensation calculation module. The compensation calculation module passes the final compensation vector and the type and boundary coordinate information of each partition to the parameter binding module.

[0150] The exception handling mechanisms for each module are as follows: The edge extraction module outputs an exception code and terminates when a file format error is detected. The partition recognition module treats the entire edge as a single partition when the number of partitions is zero. The compensation prediction module reverts to a purely empirical model when a division-by-zero exception occurs in the physical model calculation. The compensation calculation module truncates the compensation vector and outputs a warning when the compensation vector exceeds its range. The parameter binding module outputs a failure report when validation fails.

[0151] In another embodiment, the system is deployed in a distributed architecture environment. The edge extraction module and partition identification module run on the engineering design workstation, the compensation prediction module and compensation calculation module run on the engineering calculation server, and the parameter binding module runs on the production line control terminal. Structured data packets are transmitted between the modules via the factory's internal network, and data transmission uses a request-response mode to ensure synchronization and consistency. When a network transmission timeout or data verification failure occurs, the sending end automatically retransmits until the receiving end confirms successful reception. The maximum number of retransmissions is three; if more than three retransmissions are performed, a communication error alarm is output, and the process is paused pending manual intervention.

[0152] In another embodiment, regarding the plate thickness Side plating dimensions are controlled for thin sheet metal parts. The equivalent penetration depth of thin sheet metal parts... The value is reduced accordingly to Cornering radius Because corners have a smaller impact range, many partitions marked as corner types in thick plates may be merged into adjacent partitions in thin plates.

[0153] In yet another embodiment, regarding the plate thickness Thick plate components undergo dimensional control of side plating. The equivalent penetration depth of thick plate components... The value increases to Cornering radius The current concentration effect is more prominent in thick plates. The physical model has a greater advantage in prediction accuracy in highly complex partitions compared to the empirical model, and the fusion weights are generally biased towards the physical model.

[0154] Optionally, the method also supports side plating control for multi-layer stacked boards. For stacked boards containing multiple copper layers, the complete process from S100 to S400 is executed independently for each layer, and the compensation vector for each layer is calculated independently. When generating the milling path offset, the parameter binding module takes the maximum value of the compensation amount of each layer as the unified offset.

[0155] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware.

[0156] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for controlling the dimensional plating of irregularly shaped edges on the side of a printed circuit board, characterized in that, include: Parse the design file and extract the geometric element data of the target edge that needs side plating treatment; The target edge is divided into partitions at abrupt changes in geometric features, and geometric feature parameters are calculated for each partition and assigned to a preset edge type category; The geometric feature abrupt change includes curvature abrupt change and angle inflection point; wherein, the curvature abrupt change is the position where the curvature difference between adjacent geometric elements exceeds the curvature abrupt change threshold, and the angle inflection point is the position where the angle between the direction vectors of adjacent straight line elements exceeds the angle inflection threshold. Based on the geometric feature parameters and edge type categories of each partition, the edge current density distribution characteristics of each partition are predicted, and the compensation correction amount of each partition is determined. Based on the edge type category and geometric feature parameters of each partition, and combined with the compensation correction amount, the differentiated compensation amount of each partition is calculated to obtain the compensation vector; The compensation vector is transformed into process parameters for downstream processing steps, and a cascade constraint relationship is established between the compensation vector and the process parameters.

2. The method according to claim 1, characterized in that, The edge type categories include six types: straight line segment, convex arc, concave arc, right angle, acute angle, and obtuse angle; among them, based on the curvature value and the included angle value of each partition, each partition is assigned to the corresponding edge type category through a fractal decision tree.

3. The method according to claim 2, characterized in that, The calculation of the differential compensation amount for each partition includes: The baseline compensation amount is determined from the calibrated baseline compensation coefficients based on the edge type category described in the partition. The corresponding correction function is selected according to the edge type category, wherein the curvature correction amount is determined based on the curvature parameter for the arc type partition, and the angle correction amount is determined based on the inner angle parameter for the corner type partition. The curvature correction amount or angle correction amount is combined with the reference compensation amount to obtain the empirical compensation amount for each zone.

4. The method according to claim 3, characterized in that, The predicted edge current density distribution characteristics of each partition include: Based on conformal transformation, the edge geometry of each partition is mapped to the standard electric field configuration, and the current density distribution function is solved on the standard electric field configuration. The mean excess current density of each partition is determined based on the current density distribution function, and the compensation correction amount is determined based on the mean excess current density.

5. The method according to claim 3 or 4, characterized in that, The calculation of the differential compensation amount for each partition also includes: The fusion weight is determined based on the geometric complexity of each partition. The compensation correction amount and the empirical compensation amount are then weighted and fused according to the fusion weight to obtain the final differentiated compensation amount for each partition.

6. The method according to claim 1, characterized in that, The process parameters of the machining process include the path offset of the contour milling process and the partition masking parameters of the electroplating process; the cascade constraint relationship includes the constraint that the combination of the path offset and the partition masking parameters does not exceed the equipment capability boundary.

7. The method according to claim 1, characterized in that, Also includes: Apply a length attenuation factor to partitions with lengths less than the critical length to suppress overcompensation; Smooth transition processing is performed at the boundary where the differential compensation amount of adjacent partitions jumps beyond the preset threshold.

8. The method according to claim 1, characterized in that, Also includes: During the production process, the actual side coating coverage size is measured, and the measured value is compared with the design target value section by section to obtain the deviation value. When multiple batches of partitions of the same edge type exhibit deviations in the same direction, the benchmark compensation coefficient corresponding to that type is corrected based on the deviation value.

9. A system for controlling the dimensional plating of irregularly shaped edges on the side of a printed circuit board, characterized in that, include: The edge extraction module is used to parse the design file and extract the geometric element data of the target edges that need to be side-plated. The partitioning identification module is used to partition the target edge at the point of geometric feature abrupt change, calculate the geometric feature parameters of each partition and classify it into a preset edge type category; The compensation prediction module is used to predict the edge current density distribution characteristics of each partition based on the geometric feature parameters and edge type category of each partition, and to determine the compensation correction amount of each partition. The compensation calculation module is used to calculate the differentiated compensation amount for each partition based on the edge type category and geometric feature parameters of each partition, combined with the compensation correction amount, to obtain the compensation vector; The parameter binding module is used to convert the compensation vector into process parameters for downstream processing steps and establish a cascade constraint relationship between the compensation vector and the process parameters. The geometric feature abrupt change includes curvature abrupt change and angle inflection point; wherein, the curvature abrupt change is the position where the curvature difference between adjacent geometric elements exceeds the curvature abrupt change threshold, and the angle inflection point is the position where the angle between the direction vectors of adjacent straight line elements exceeds the angle inflection threshold.

Images