WO2021248650A1

WO2021248650A1 - Reverse reconstruction method for high-speed rail sleeper beam process hole

Info

Publication number: WO2021248650A1
Application number: PCT/CN2020/104927
Authority: WO
Inventors: 李鹏一; 程远; 董巍; 唐凯; 徐龙; 胡冬双; 吴晓
Original assignee: 南京英尼格玛工业自动化技术有限公司
Priority date: 2020-06-10
Filing date: 2020-07-27
Publication date: 2021-12-16
Also published as: CN111737796B; CN111737796A

Abstract

The present invention relates to a reverse reconstruction method for a high-speed rail sleeper beam process hole, divided into four stages: collecting point cloud data, and obtaining three-dimensional coordinate information of surface points of a physical model by means of a professional data measurement device; eliminating noise points in point data by means of point cloud noise reduction; reconstructing a curved surface by using the point cloud data, and performing curved surface reconstruction and curved surface splicing on the processed point cloud data or triangular curved surface patch data in reverse engineering software to obtain an STL model of an original product; and repairing a defective surface, performing distance deviation inspection and smoothness detection on a reconstructed surface, performing determination according to design requirements, and performing repeated surface editing on a component which does not meet the requirements until the design requirements of a product are met. The method not only can improve the precision of a mold and the product, but also can greatly shorten a design period and save costs. Especially for some products having complex structures, high curved surface requirements or lack of original geometric data, the method can give full play to the advantages of the products.

Description

A Reverse Reconstruction Method Aiming at Process Hole of High-speed Rail Corbel

Technical field

The invention relates to the field of arc additive materials, in particular to a reverse reconstruction method for high-speed rail corbel technological holes.

Background technique

Reverse Engineering (Revese Engineering, referred to as RE) is to obtain the CAD model of the reversed object through many methods such as shape reversed, process reversed and material reversed. The process of digital modelling. The main purpose of reverse engineering is to reproduce the design or repair the product. By digitizing the physical model of the product, and then restoring the model, clarifying the shape features, functional characteristics, technical specifications, process flow and other technical characteristics, so as to innovate or repair, and realize the function of the product. Redesign of design elements such as, appearance, etc.

Compared with the forward design, the product design based on reverse engineering detects the actual object through measuring equipment to collect point cloud data, and then uses professional software to reconstruct the point, line and surface information of the actual object, and then obtain the three-dimensional shape of the actual object. Reproduce or repair the product through processing equipment such as 3D printing or machine tools to obtain finished parts that meet the requirements.

The common reverse reconstruction method used for process hole processing in the prior art usually has a large distance deviation and poor smoothness, and cannot be applied to process holes with high precision requirements.

Summary of the invention

The purpose of the invention: to provide a reverse reconstruction method for the process hole of the high-speed rail bolster, so as to solve the above-mentioned problems in the prior art.

Technical solution: A reverse reconstruction method for high-speed rail corbel process holes, including the following steps:

Step 1. Obtain point cloud data;

Step 2. Perform point cloud processing and surface reconstruction to obtain a reverse reconstruction model;

Step 3. Perform defect analysis.

In a further embodiment, in step 1, the point cloud data of the unprinted physical model or the semi-finished product in the printing process is photographed by a three-dimensional laser scanner. The three-dimensional laser scanner is preferably lungoCAM, which has the advantages of fast measurement speed, high accuracy, ability to measure complex structures, and good data format compatibility. In order to obtain the product point cloud data completely, multiple images are required to be collected.

In a further embodiment, the step 2 further includes:

Step 2-1: Denoise the point cloud to eliminate the noise points in the point data. The data sampling method is used to simplify the point cloud data and segment the point cloud data by extracting characteristic lines;

Step 2-2. Use point cloud data to reconstruct the surface, and perform surface reconstruction and surface splicing in the reverse engineering software on the processed point cloud data or triangular surface patch data to obtain the original product model;

Step 2-3. Perform distance deviation inspection and smoothness inspection on the reconstructed surface, judge according to the design requirements, and perform repeated surface editing on the parts that do not meet the requirements until the product design requirements are met.

In a further embodiment, the step 3 compares the obtained reverse reconstruction model with the additive layered slicing standard model, and performs defect analysis. If there is no defect, the printing is continued according to the originally planned path; if there is a large deviation , The software can adaptively re-plan the additive path, so that the additive model is as close as possible to the ideal model;

Among them, the defect analysis includes the analysis of size deviation, collapse, and accumulation defects.

The influence of various factors will cause a certain number of noise points and noise points to be generated during the scanning process. These points will seriously interfere with the subsequent surface reconstruction. Therefore, the first step of preprocessing is to use software to remove noise points and noise points. The data is sampled and streamlined, and the amount of data is streamlined without affecting the various characteristics of the product. While it is convenient for subsequent processing, some unnecessary overlapping points are also deleted. When surface reconstruction, it is necessary to specifically analyze the characteristic laws of each surface in the product, and then select the appropriate surface reconstruction method, such as boundary surface, loft surface, sweep surface, free surface, etc.

In a further embodiment, in step 2-2, the coordinates of each point cloud image are further obtained, and each point cloud image is coordinated to form a complete and corrected complete point cloud. The complete point cloud is transformed into the STL model of the product. The scanning of the product by the scanner is divided into frames, and the coordinates of each point cloud are independent, so multiple point cloud data are messy and disorderly, which requires manual splicing of multiple point cloud data. That is, the realignment of the coordinate system. The three-dimensional model of the designed product is optimized according to actual needs. After modeling, the accuracy of the three-dimensional model of the entire product needs to be tested, and the curved surfaces or features with larger errors should be modified according to the test results.

In a further embodiment, the reverse reconstruction method is based on the following system, including a transplanting workstation, and an arc additive workstation arranged on one side of the middle section of the transplanting workstation; the transplanting workstation includes a bidirectional feeder The transplanting track is provided with a corbel fixing seat, and the head and tail of the corbel are positioned and clamped on the fixing seat at a predetermined interval; each corbel is pressed together by a plurality of lower pressing plates, the The lower pressing plate is pressed at the head and tail of the corbel and the middle sections avoiding the corbel craft hole; the transplanting track controls the corbel to advance according to the preset rhythm of the warp. When a complete reverse reconstruction, arc is completed After the three processes of welding and laser cleaning, the processed corbel is sent out of the arc additive workstation, and the next corbel to be processed is sent to the arc additive workstation.

The arc additive workstation includes a safety protection room that encloses a designated work area, the safety protection room is located on both sides of a transplanting track with rolling shutter doors, the transplanting track passes through the rolling door, and the safety protection room is Positions close to the rolling door are respectively provided with a robot tool quick change device, the robot tool quick change device is located on one side of the transplanting track, and a plurality of industrial robots are arranged between the robot tool quick change devices; When the working area is being processed, the rolling shutter door is closed. After the processing is completed, the rolling shutter door is opened, and the bolster is sent out through the transplanting track.

The robot tool quick change device includes a support frame, a quick change plate fixed on one side of the upper part of the support frame, and fixing seats respectively provided on the quick change plate; A cylinder, the output end of the rotary cylinder is fixed with an extended rotating part, and the end of the rotating part is fixed with a contact part that directly contacts the corresponding quick change tool. The robot tool quick change device can make the welding gun, laser cleaning head, and 3D camera cooperate with the additive manufacturing process to realize automatic switching.

In a further embodiment, there are three fixing seats, the welding torch head, laser cleaning head, and 3D camera are respectively arranged on the fixing seats, and one side of the welding torch head, laser cleaning head and 3D camera is fixed with a section When the movable seat adapted to the fixed seat is not exchanged, the welding torch head, the laser cleaning head, and the 3D camera are clamped on the fixed seat by the movable seat, and the rotating part of the rotating cylinder is laterally Compression; the movable seat is fixed with a quick-change lock, and the end of the mechanical arm of the industrial robot is also fixed with a quick-change lock. The above locking structure can protect the welding gun, laser cleaning head, and 3D camera from being taken out at the wrong time, ensuring process reliability.

Beneficial effects: The present invention relates to a reverse reconstruction method for high-speed rail bolster process holes. It uses reverse auxiliary technology to quickly and accurately design a mold for a product with a relatively complex structure. First, use a three-dimensional laser scanner to obtain it Point cloud data, and then use software for point cloud processing and surface reconstruction, and then perform three-dimensional modeling and corresponding optimization design in UG software, and finally obtain a complete mold drawing with the most reasonable design method for subsequent mold processing and Production. This method can not only improve the accuracy of molds and products, but also greatly shorten the design cycle and save costs. Especially for products with complex structures, high surface requirements or missing original geometric data, this method can give more play to its advantages.

Description of the drawings

Figure 1 is a perspective view of a perspective view of the overall system of the present invention.

Fig. 2 is another perspective view of a perspective view of the overall system of the present invention.

Figure 3 is a top view of the overall system of the present invention.

Figure 4 is a three-dimensional view of the arc additive workstation in the present invention.

Fig. 5 is a perspective view of the industrial robot and the robot tool quick change device in the present invention.

Fig. 6 is a partial enlarged view of the robot tool quick change device in the present invention.

Fig. 7 is a schematic diagram of the structure of the processed workpiece corbel in the present invention.

Figure 8 is a working flow chart of the overall system of the present invention.

Fig. 9 is a point cloud diagram of the model reconstructed in the reverse direction in the present invention.

Figure 10 is a technical roadmap of reverse engineering products in the present invention.

Figure 11 is a flow chart of smart printing.

Fig. 12 is a point cloud data diagram of a bearing obtained by photographing with a three-dimensional laser scanner.

FIG. 13 is a schematic diagram of fitting and splicing point clouds in the present invention.

The reference signs in the figure are: transplanting workstation 1, corbel 101, lower pressing plate 102, rolling door 2, safety protection room 3, welding 4, robot tool quick change device 5, support frame 501, quick change plate 502, fixed seat 503, rotating cylinder 504, rotating part 505, master control cabinet 6, robot control cabinet 7, industrial robot 8, quick-change lock 801, welding gun head 9, laser cleaning head 10, and 3D camera 11.

detailed description

In the following description, a lot of specific details are given in order to provide a more thorough understanding of the present invention. However, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features known in the art are not described.

Reverse engineering technology is relative to traditional forward engineering. It mainly measures the existing physical models or parts with a three-dimensional coordinate measuring machine or a three-dimensional laser scanner to obtain point cloud data, and then use the corresponding processing software lungoPNT The process of surface reconstruction with UG and others, and finally obtain the physical three-dimensional model. The reconstructed model can reflect the geometric characteristics and other attributes of the original object, and can be used for various purposes such as analysis, modification, manufacturing, and inspection of the actual object.

The general process of reverse engineering is divided into four stages:

(1) Collecting point cloud data, using professional data measuring equipment and measuring methods to obtain the three-dimensional coordinate information of the surface points of the physical model;

(2) In order to process point cloud data, point cloud noise reduction is needed to eliminate the noise points in the point data. The data sampling method is used to simplify the point cloud data and segment the point cloud data by extracting characteristic lines;

(3) Use point cloud data to reconstruct the surface, and perform surface reconstruction and surface splicing in the reverse engineering software on the processed point cloud data or triangular surface patch data to obtain the original product model;

(4) Due to the existence of noise, directly regenerating the surface through point cloud data, there are often some defects, so it is necessary to further repair the defective surface. Through the distance deviation check and smoothness detection on the reconstructed surface, the judgment is made according to the design requirements, and repeated surface editing is performed on the parts that do not meet the requirements until the product design requirements are met.

The present invention relates to a reverse reconstruction method for high-speed rail corbel technological holes. The method is based on a transferable dual-robot arc 3D printing workstation. The workstation includes a transplanting workstation 1 and an arc additive workstation. The arc additive workstation integrates reverse weight Three processes: structure, arc welding, 4, laser cleaning.

Specifically, the transplanting workstation 1 includes a transplanting track capable of bidirectional feed. The transplanting track is provided with a corbel 101 fixing seat 503, and the head and tail of the corbel 101 are positioned and clamped on the fixing seat 503 at a predetermined interval. superior. Each corbel 101 is pressed together by a plurality of lower pressing plates 102, which are pressed together at the head and tail of the corbel 101, and at multiple sections avoiding the process hole of the corbel 101. The transplanting track controls the corbel 101 to advance according to the preset warp rhythm. After completing a complete reverse reconstruction, arc welding 4, and laser cleaning, the processed corbel 101 will be sent out of the arc additive workstation. The next corbel 101 to be processed is sent to the arc additive workstation.

The arc additive workstation includes a safety protection room 3 that encloses a designated working area. The safety protection room 3 is located on both sides of the transplanting track and is provided with rolling shutter doors 2, and the transplanting track passes through the rolling door 2, and the safety protection A robot tool quick change device 5 is respectively arranged in the interior of the room 3 near the rolling shutter door 2. The robot tool quick change device 5 is located on one side of the transplanting track, and the robot tool quick change device 5 is provided between There are multiple industrial robots8.

The robot tool quick change device 5 includes a support frame 501, a quick change plate 502 fixed on the upper side of the support frame 501, and a fixing seat 503 respectively arranged on the quick change plate 502; two of the fixing seat 503 Rotating cylinders 504 are respectively fixed on the sides. An extended rotating part 505 is fixed at the output end of the rotating cylinder 504, and a contact part that directly contacts the corresponding quick change tool is fixed at the end of the rotating part 505. The robot tool quick change device 5 can make the welding gun, the laser cleaning head 10, and the 3D camera 11 cooperate with the additive manufacturing process to realize automatic switching. There are three fixing bases 503, the welding torch head 9, the laser cleaning head 10, and the 3D camera 11 are respectively arranged on the fixing base 503. One side of the welding torch head 9, the laser cleaning head 10, and the 3D camera 11 is fixed with a section and When the movable seat adapted to the fixed seat 503 is not exchanged, the welding torch head 9, the laser cleaning head 10, and the 3D camera 11 are clamped on the fixed seat 503 by the movable seat, and are fixed on the fixed seat 503 by the movable seat. The rotating part 505 of the rotating cylinder 504 is pressed laterally; a quick-change lock 801 is fixed on the movable seat, and a quick-change lock 801 is also fixed at the end of the mechanical arm of the industrial robot 8. The above locking structure can protect the welding gun, the laser cleaning head 10, and the 3D camera 11 from being taken out at the wrong time, ensuring process reliability.

When the work area is being processed, the rolling shutter door 2 is closed. After the processing is completed, the rolling shutter door 2 is opened, and the bolster 101 is sent out through the transplanting track. A welding machine, a laser cleaning power supply, a master control cabinet 6 and a robot control cabinet 7 connected to the industrial robot 8 are arranged on one side of the arc additive workstation.

The specific working process of the present invention is as follows: firstly, the corbel 101 is manually hoisted to the transplanting workstation 1, the corbel 101 is positioned and clamped on the fixed seat 503, and the corbel 101 is pressed by a plurality of lower pressing plates 102.

After the corbel 101 is clamped, the transplanting workstation 1 is started, and the transplanting track transfers the positioned and fixed corbel 101 to the work area;

Then set the current workpiece coordinates. The industrial robot 8 is first driven to the robot tool quick changer 5, and then accurately driven to the 3D camera 11 when it is in place. When the robot arm of the industrial robot 8 is directly above the 3D camera 11, continue Slowly descend until the two quick-change lock heads 801 are engaged. After the engagement is completed, the rotating cylinder 504 drives the rotating part 505 to rotate away from the movable part. The industrial robot 8 continues to start, so that the 3D camera 11 is separated from its fixed part and continues to drive to the top of the corbel 101 .

Then the 3D camera 11 starts to visually scan the part, analyzes the contour data and compensates for defects, and then reconstructs the model in reverse, and then the computer sets the slice parameters and generates the robot trajectory path. At the same time, the printing and welding 4 process parameters are set.

Then, the industrial robot 8 drives the 3D camera 11 back to the robot tool quick changer 5, places the 3D camera 11 back on the fixed part, and switches to the welding torch head 9, and continues to return to the top of the bolster 101.

After returning to the top of the bolster 101, the laser welding 4 is started. After the welding 4 is completed, the welding torch head 9 is placed back on the fixed part, and the laser cleaning head 10 is switched to continue to return to the top of the bolster 101 for interlayer laser cleaning.

After welding 4, the workpiece is manually hoisted out of the workstation for heat treatment, and the other industrial processing is completed at the same time.

Among them, laser welding uses a welding machine as a heat source and a metal wire as a forming material to plan a continuous spirally ascending slice path for cladding printing. The process is as follows:

1) Select the welding wire and substrate required for forming specific metal structural parts, and determine the process parameters required for forming specific metal structural parts, including welding procedure, wire feeding speed, printing speed, slice layer height, shielding gas type and flow rate, and various parameters The relationship is as follows:

The welding speed is proportional to the wire feeding speed, which can be expressed by the relation (1)

V×F=v×f…………………………(1)

V: welding speed;

F: Cross-sectional area of weld

v: wire feeding speed

f: Cross-sectional area of welding wire

The welding seam section of the workpiece is equivalent to a rectangle, then

F=ld....................................(2)

Among them, l: equivalent rectangular weld width;

d: weld height (namely layer height)

The relationship between wire feeding speed and layer height can be obtained from formulas (1) and (2), as shown in formula (3):

Through the wire feeding speed, the current and voltage values can be read on the control panel, and then the heat input of each 1mm welding wire consumed at the wire feeding speed can be calculated:

Among them, U: arc voltage;

I: welding current;

V: welding speed;

K: relative thermal conductivity;

In the arc additive manufacturing process, the control of heat input is extremely important. If the heat is too low, the weld will not be formed, the workpiece will not be fused, and the heat will cause the workpiece to collapse. Therefore, combining various wire properties with the printing process The relationship of temperature can be inferred suitable for the heat input of the wire, and then the process parameters, such as wire feeding speed, welding speed and layer height can be determined.

2) Wipe the polished flat substrate with absolute ethanol or acetone and fix it on the workbench to ensure its level;

3) The generation of continuous spirally ascending slicing path is as follows:

First, the STL model of the workpiece to be printed is sliced. There are many existing STL model slicing algorithms. We use the STL slicing algorithm based on the geometric characteristics of the triangle to process the STL model, and divide the model into several planes along the Z axis;

Secondly, look for adjacent layers and use the layer with the higher relative position to subtract the layer with the lower relative position to obtain the layer height;

Then randomly select a point on the first layer slice as the starting point (that is, the welding arc point), and then use the following formula to find the offset height in the Z direction between two adjacent points:

Among them, d is the vertical height between the starting point and the end point in the same layer;

X is the number of points per slice;

z is the offset height in the Z direction between each point.

More specifically, the slicing process is as follows:

Divide the model into a number of triangles along the Z axis to obtain the maximum and minimum values of the three-dimensional model in the Z axis, consider the reserved machining allowance, and calculate the total number of layers:

In the formula, Z _max represents the maximum value of the three-dimensional model in the Z-axis direction, Z _min represents the minimum value of the three-dimensional model in the Z-axis direction, Δz represents the layer height, k is the adjustment coefficient, and Δz+k is the preset value The adjustment factor is added on the basis of the layering height to ensure the machining allowance;

Then store each triangle face in each layer of n layers in a dynamic array, and query the value of each triangle face

Value if

Then store the current triangle patch in the j-th group of the dynamic array; if

Then store the current triangle patch in the j-1th group of the dynamic array; if

Then store the current triangle patch in the j+1th group of the dynamic array;

Among them, h _j represents the height of the j-th group, and h _j+1 represents the height of the j+1-th group. The height is taken from the middle value of the minimum and maximum values of the three-dimensional model in the Z-axis direction plus the layer height The product of the number of groups gives:

h _j ＝(Z _min +Z _max )/2+Δz×j

In the formula, Z _min represents the minimum value of the three-dimensional model in the Z-axis direction, Z _max represents the maximum value of the three-dimensional model in the Z-axis direction, Δz represents the layer height, and j represents the number of groups.

Then look for the starting point of the next layer, requiring the point to be the closest to the end point of the previous layer, and connecting the end point of the previous layer to the starting point of the layer, that is, the continuity of the trajectory between the two layers is realized, and the printing process is not Will extinguish the arc.

This method is used in turn to connect all path points of the entire workpiece to generate a continuous spiral path to realize continuous arc additive manufacturing of the workpiece.

4) The welding gun is driven by the robot to move according to the generated continuous spiral path. At the same time, the process parameters are determined according to the method of step 1), and a single weld seam is printed on the substrate. The height of the welding gun from the substrate during the printing process according to the continuous spiral path is gradually Elevated. The combination of the continuous spiral path and the process parameters calculated according to the heat input in 1) can ensure that the dry elongation of the welding wire remains unchanged during the printing process, and the arc will not be extinguished during the entire printing process, and finally a metal structure with good structural performance is formed.

As a preferred solution, the central control unit also optimizes the trajectory of the continuous spiral path:

First set the linear velocity v _{c of the} spiral path:

v _c =ω(Lv ₀ t)

In the formula, ω represents the angular velocity of the torch rotation, L represents the distance between the interpolation starting point and the origin, v ₀ represents the radial velocity, Lv ₀ t is the real-time radius of the workpiece, and t represents the welding time;

Among them, the angular velocity ω of the torch rotation satisfies the following relationship:

In the formula, D represents the distance of the weld bead that the welding gun moves radially when the heat source cooperates with the platform to complete a weld forming process,

Means the radial velocity of the welding torch is averaged;

Then calculate the welding torch's deposition speed v _r :

In the formula, v _c represents the linear velocity of the spiral ascending path, and v ₀ represents the radial velocity;

Then calculate the weld bead spacing, the welding gun moves radially by a weld bead spacing, and the heat source cooperates with the platform to complete a weld formation. The expression for the weld bead spacing D is as follows:

In the formula, n represents the number of welding torches, v ₀ represents the radial velocity, t represents the welding time, ω represents the angular velocity of the torch rotation, and d represents the compensation height;

The compensation height d is determined by the interpolation accuracy and satisfies the following relationship:

Where

Indicates the average value of the radial velocity of the welding torch, and t′ represents the movement time in the interpolation interval;

Then calculate the modified welding speed v _{r repair} :

In the formula, n represents the number of welding torches, v ₀ represents the radial velocity, ω represents the angular velocity of the torch rotation, and d represents the compensation height,

It represents the average value of the radial velocity of the welding torch, and D represents the distance of the welding bead that the welding torch moves in the radial direction when the heat source cooperates with the platform to complete the formation of a weld.

When welding work, it is necessary to calculate the avoidance surface in advance to prevent the welding gun nozzle and the welding gun root from colliding with the side wall of the workpiece. The minimum diameter of the original welding torch nozzle is 22mm. Due to the narrow space at the bottom of the workpiece, the welding torch nozzle is specially made from 22mm to the current 13mm; this measure avoids the problem of inaccessibility of the root torch, because the workpiece is multi-layer and multi-pass welding , When welding to the upper layers, there will be collisions and arc deviations. Therefore, the welding torch needs to monitor the welding torch collision radius where its trajectory goes.

The welding torch trajectory avoidance is avoided by adding eight avoidance surfaces, divided by the outer contour, and the virtual surface is controlled by software. The principle is to detect the collision radius of the welding torch, and the reference is the center line of the gun head. The virtual surface is created, and the virtual surface is created in different positions. The avoidance angle can be set for avoidance, and the automatic avoidance angle is based on the distance to the outer edge of the workpiece. The welding torch change angle is 5-15°.

Reverse engineering technology is relative to traditional forward engineering. It mainly measures the existing physical models or parts with a three-dimensional coordinate measuring machine or a three-dimensional laser scanner to obtain point cloud data, and then use the corresponding processing software to perform The process of surface reconstruction and finally obtaining the physical three-dimensional model. The reconstructed STL model can reflect the geometric characteristics and other attributes of the original object, and can be used for various purposes such as analysis, modification, manufacturing, and inspection of the actual object. The steps are as follows:

Point cloud collection: lungoCAM performs data collection. This scanner has the advantages of fast measurement speed, high accuracy, can measure complex structures, and good data format compatibility. In order to obtain the product point cloud data completely, it is necessary to collect multiple images;

Point cloud data splicing: the scanner scans the product in slices, and the coordinates of each point cloud are independent, so multiple point cloud data are messy and chaotic, which requires multiple point clouds The data is manually spliced, that is, the coordinate system is realigned.

Point cloud data preprocessing: The influence of various factors will cause a certain number of noise points and noise points in the scanning process. These points will seriously interfere with the subsequent surface reconstruction. Therefore, the first step of preprocessing is to use software to remove them. After noise points and miscellaneous points are sampled and streamlined, the amount of data is streamlined without affecting the various features of the product. While it is convenient for subsequent processing, some unnecessary overlapping points are also deleted.

Surface reconstruction: When surface reconstruction, it is necessary to specifically analyze the characteristic laws of each surface in the product, and then select the appropriate surface reconstruction method, such as boundary surface, loft surface, sweep surface, free surface, etc.

Three-dimensional modeling and optimization: optimize the three-dimensional model of the designed product according to actual needs. After modeling, the accuracy of the three-dimensional model of the entire product needs to be tested, and the curved surfaces or features with larger errors should be modified according to the test results.

The method involved in the present invention can control the quality of printed parts based on model optimization, and its intelligence is mainly reflected in the following aspects:

Intelligent recognition: The software automatically recognizes the features that require special treatment contained in the digital model of the workpiece: including overlapping positions, corners, thin walls, small gaps, etc.

Intelligent planning: only need to import the digital model of the workpiece to be printed, without manual work drawings or copy paths, the software automatically plans the printing path and generates the printing program.

Intelligent optimization: The built-in algorithm automatically optimizes the printing sequence, filling strategy, starting and ending arc, path offset, etc., to reduce the occurrence of printing defects.

Intelligent filling algorithm: Provide a variety of filling algorithms suitable for arc additive technology, and continue to improve. (From left to right: grid, straight line, concentric, zigzag line).

The reverse reconstruction technology is mainly after each layer of additive is finished, the unprinted physical model or the semi-finished product in the printing process is photographed by a three-dimensional laser scanner to obtain its point cloud data, and then point cloud processing and surface reconstruction are performed, and then Obtain the reverse reconstruction model, compare it with the standard model of additive layered slices, perform defect analysis (size deviation, collapse, accumulation, etc.), if there is no defect, continue printing according to the originally planned path; if there is a large deviation, the software can Self-adaptively re-plan the additive path, so that the additive model is as close as possible to the ideal model. The entire printing process is shown in Figure 11.

As mentioned above, although the present invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the present invention itself. Various changes in form and details can be made without departing from the spirit and scope of the present invention as defined by the appended claims.

Claims

A method for reverse reconstruction of process holes for high-speed rail corbels, which is characterized by including the following steps:

Step 1. Obtain point cloud data;

Step 2. Perform point cloud processing and surface reconstruction to obtain a reverse reconstruction model;

Step 3. Perform defect analysis.
The method for inverse reconstruction of process holes for high-speed rail bolsters according to claim 1, characterized in that: the step 1 uses a three-dimensional laser scanner to take pictures of unprinted physical models or semi-finished products in the printing process to obtain their points Cloud data.
The method for reverse reconstruction of process holes for high-speed rail corbels according to claim 1, wherein the step 2 further comprises:

Step 2-1: Denoise the point cloud to eliminate the noise points in the point data. The data sampling method is used to simplify the point cloud data and segment the point cloud data by extracting characteristic lines;

Step 2-2. Use point cloud data to reconstruct the surface, and perform surface reconstruction and surface splicing in the reverse engineering software on the processed point cloud data or triangular surface patch data to obtain the original product model;

Step 2-3. Perform distance deviation inspection and smoothness inspection on the reconstructed surface, judge according to the design requirements, and perform repeated surface editing on the parts that do not meet the requirements until the product design requirements are met.
The method for reverse reconstruction of process holes for high-speed rail corbels according to claim 1, characterized in that: in step 3, the obtained reverse reconstruction model is compared with the standard model of additive layered slices to perform defect analysis, If there is no defect, continue printing according to the originally planned path; if there is a large deviation, the software can adaptively re-plan the additive path so that the additive model is as close as possible to the ideal model;

Among them, the defect analysis includes the analysis of size deviation, collapse, and accumulation defects.
The method for inverse reconstruction of process holes for high-speed rail corbels according to claim 3, characterized in that, in step 2-2, the coordinates of each point cloud image are further obtained, and each point cloud image is coordinated In this way, it is spliced into a complete and corrected complete point cloud, and the complete point cloud is transformed into a product model.
The method for inverse reconstruction of process holes for high-speed rail corbels according to claim 1, wherein the method is based on the following system, including a transplanting workstation, and one set in the middle section of the transplanting workstation Side of the arc additive workstation; the transplanting workstation includes a transplanting track capable of bidirectional feed, the transplanting track is provided with a corbel fixed seat, the head and tail of the corbel are positioned and clamped at a predetermined interval in the fixed On the seat; each corbel is pressed by a plurality of lower pressing plates, the lower pressing plates are pressed on the head and tail of the corbel, and the middle sections avoiding the corbel craft hole;

The arc additive workstation includes a safety protection room that encloses a designated work area, the safety protection room is located on both sides of a transplanting track with rolling shutter doors, the transplanting track passes through the rolling door, and the safety protection room is Positions close to the rolling door are respectively provided with a robot tool quick change device, the robot tool quick change device is located on one side of the transplanting track, and a plurality of industrial robots are arranged between the robot tool quick change devices; The robot tool quick change device includes a support frame, a quick change plate fixed on one side of the upper part of the support frame, and a fixing seat respectively provided on the quick change plate; two sides of the fixing seat are respectively fixed with a rotating cylinder The output end of the rotary cylinder is fixed with an extended rotating part, and the end of the rotating part is fixed with a contact part that directly contacts the corresponding quick change tool.
The method for reverse reconstruction of process holes for high-speed rail bolsters according to claim 6, characterized in that there are three fixing seats, and the welding torch head, the laser cleaning head and the 3D camera are respectively arranged on the fixing seats, so One side of the welding torch head, laser cleaning head, and 3D camera is fixed with a movable seat that is adapted to the fixed seat. When not exchanged, the welding torch head, laser cleaning head, and 3D camera are clamped by the movable seat. It is arranged on the fixed seat and pressed laterally by the rotating part of the rotating cylinder; a quick-change lock is fixed on the movable seat, and the end of the mechanical arm of the industrial robot is also fixed with a quick-change lock.