CN106202727B - Design method and system of special-shaped cable drawing die - Google Patents
Design method and system of special-shaped cable drawing die Download PDFInfo
- Publication number
- CN106202727B CN106202727B CN201610545822.7A CN201610545822A CN106202727B CN 106202727 B CN106202727 B CN 106202727B CN 201610545822 A CN201610545822 A CN 201610545822A CN 106202727 B CN106202727 B CN 106202727B
- Authority
- CN
- China
- Prior art keywords
- die
- elongation
- pass
- inner hole
- section
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
Landscapes
- Physics & Mathematics (AREA)
- Geometry (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computational Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Metal Extraction Processes (AREA)
Abstract
The invention provides a design method and a system of a special-shaped cable drawing die, wherein the method is a method for designing multi-pass drawing die matching by combining finite element simulation, an artificial neural network, a genetic algorithm and software modeling on the basis of simulation and artificial intelligence, and the design method can ensure the local and global load balance of the die. According to the invention, by means of simulation, optimization and software design, the hole pattern of the drawing die is optimally designed, so that the performance of the drawing die and the drawing equipment can be fully exerted, the service life of the die is greatly prolonged, the time for replacing and maintaining the die is reduced, the production efficiency is greatly improved, and the method is very important for improving the economic benefit of the whole metal wire and pipe drawing industry.
Description
Technical Field
The invention relates to a design method in the technical field of dies, in particular to a design method and a system of a special-shaped cable drawing die based on artificial intelligence and finite element analysis and load balancing.
Background
The profiled cable is a non-circular cross-section wire of various geometries other than a circular cross-section, as opposed to a circular cross-section wire. The special-shaped wire is widely applied to the fields of aerospace, machinery, automobiles, electronic industry, communication and the like. The smart grid is one of seven new industries in the world war, and can promote the rapid development of the wire and cable industry in China. Meanwhile, the novel application field of the cable is greatly expanded due to the construction of a new energy power station, the development of equipment manufacturing, the popularization of rail transit and the like. The special-shaped wire is widely applied to the concentric stranded overhead conductor and the electromagnetic wire, compared with the circular concentric stranded overhead conductor, the molded line stranded overhead conductor has larger conductor section utilization rate, and the diameter can be reduced by about 10 percent when the sections are the same; when the diameter is equal, the effective section can be increased by 20-25%. In order to improve the transmission capability of the extra-high voltage transmission line, a cable conductor with a larger section must adopt a split conductor so as to reduce the increase of conductor resistance and the reduction of transmission capacity caused by skin effect and proximity effect during the transmission of alternating current. The dividing conductor generally adopts special-shaped lines (such as S-shaped lines, Z-shaped lines and trapezoidal lines), and the special-shaped lines have the advantages of increasing the utilization rate of the section of the conductor, reducing the sag rate of the conductor, having good corrosion resistance, small damage of broken lines, strong self-damping performance, reducing wind load, reducing the possibility of galloping of the overhead conductor and the like.
The special-shaped cable is widely adopted abroad, particularly developed countries, but the production and application in China are still in the starting stage at present, and the indexes of the drawing die such as the processing precision, the durability and the like are all required to be further improved. The quality of the special-shaped cable mold is guaranteed in the production of high-quality special-shaped wires and cables, so that the research of the special-shaped drawing mold has important significance in the development of the wire and cable industry in China and in energy conservation and consumption reduction.
At present, the production of domestic special-shaped wire rods mainly adopts a method of roller die drawing and fixed die drawing. The roller die drawing is a drawing and rolling process combining drawing and rolling, and rollers are driven to rotate by friction force between a drawing piece and the drawing piece, so that the drawing piece is deformed in a hole pattern. Compared with the drawing of a fixed die, the drawing force of the roller die drawing is small, the energy consumption is low, the compression ratio is large, the processing limit is high, the requirement on lubrication is low, the service life of the die is long, the requirement on blanks is low (such as the properties of the blanks, such as section deformation, size change, welding, and the like, surface defects and the like), the elastic limit and the yield strength of finished products are improved, the section hardness distribution is uniform, the residual stress is small, and the like. The production of the fixed die generally employs a method of wire drawing and die matching. The method for wire drawing and die matching is that the size of a round billet or a flat billet is determined according to the size, the shape and the material of a finished product during the wire drawing, the size and the shape of each drawing pass are determined, and the intermediate annealing time is determined. However, when the wire drawing and die matching method is adopted to draw the molded line, the deformation nonuniformity of the workpiece is more serious than that of the round wire drawing, the formed residual stress is larger, and the possibilities of shape bending, inaccurate size and local cracks are also higher. Therefore, when the special-shaped wire drawing die is designed, the defects of the traditional design method are overcome, and the problems of size and stress caused by the traditional design are solved by means of big data.
As is well known, artificial intelligence is a new technical science for researching and developing theories, methods, techniques and application systems for simulating, extending and expanding human intelligence, and aims to research how to simulate the intelligent behavior of human beings by using modern tool systems such as computers. The development of artificial intelligence technology provides an effective method for analyzing and processing production data and information, and intelligent wings are added to the manufacturing technology. The artificial intelligence technique is particularly suitable for solving the problems of particular complexity and uncertainty, and can be widely applied to all links of the manufacturing process. The neural network is used as a cross research field integrating multiple subjects such as brain science, neuropsychology, information science and the like, is applied to multiple aspects such as pattern recognition, machine learning, expert systems and the like, and becomes an active field in artificial intelligence research. Aiming at the characteristic of strong learning capacity of the neural network, the invention combines the neural network and the finite element technology based on the software modeling of the continuous linear variation hole pattern design method, and is applied to the design of the special-shaped wire drawing die, thereby not only simplifying the design flow of the die and ensuring the accuracy of the design size of the die, but also greatly improving the aspects of the wear resistance, the service life and the like of the die.
Through the search of the prior art documents, mussoti 2011 proposes a method for designing a transition hole-shaped cross section in the middle of a wire drawing die by using a CAXA electronic drawing board. And under the condition that the drawing pass and the final forming section shape are determined, designing the hole pattern of the die for transferring the circular section to the section required by the molded line. Finally, the cross section of each transition hole is approximately obtained in the form of a sample line, and the CAXA software is used for machining on a wire cut electric discharge machine, so that a relatively accurate hole pattern is obtained. However, the design method needs to study the flow characteristics of metal processing, the modeling process is complex, and scaling adjustment is used when the area of the hole-type interface of each pass is determined, so that certain errors exist in the design parameters.
The application discloses a method for designing and optimizing copper alloy pipe casting and rolling process parameters, which is a Chinese invention application with the publication number of 1979496A and the application number of 200510047903.6, and the application discloses a method for designing and optimizing copper alloy pipe casting and rolling process parameters, wherein a database is used as a design basis, a neural network is used as a design method of process parameters and process indexes, a genetic algorithm is used as a process parameter optimization means, and the process parameters of the copper alloy pipe casting and rolling are designed and optimized by comprehensively integrating the neural network, the genetic algorithm, finite element simulation, test design, CAD parametric design and database technologies in process design and parameter optimization. The invention has high automation degree, can be suitable for the processing deformation of the copper alloy pipe, and leads the personnel lacking rich professional knowledge to be capable of making an accurate and standard processing technology. But the method cannot be applied to the design of the special-shaped cable drawing die.
Therefore, how to design a design method for rapidly obtaining optimized parameters and more accurately and reliably drawing a special-shaped cable has become the key for solving the problems.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a design method and a system of a special-shaped cable drawing die, which can design a hole type section of the drawing die, which not only accords with a hole type change rule of gradual transition, but also meets speed ratio parameters of a drawing machine.
According to the first aspect of the invention, the method for designing the special-shaped cable drawing die is provided, wherein on the basis of simulation and artificial intelligence, finite element simulation, an artificial neural network, a genetic algorithm and software modeling are combined, and the method can ensure that the designed die has the characteristics of local load balance and global load balance;
the finite element simulation is used for calculating two parts of stress:
one part is as follows: the section area and the elongation of the inlet wire are the same, but the geometric parameters are different, the equivalent stress of each section of the surface of the inner hole of the drawing die and the average stress value of the surface of the inner hole; comparing to obtain the geometric parameters of the die with the minimum equivalent stress variation range of each section of the surface of the inner hole, wherein the obtained geometric parameters are the optimal die geometric parameters meeting the local load balance under the cross section area and the elongation of the inlet wire;
the other part is that: after the cross section area and the elongation of the inlet wire are changed for many times, the equivalent stress of each section of the surface of the inner hole of the drawing die corresponding to different geometric parameters and the average stress value of the surface of the inner hole are obtained; obtaining optimal die geometric parameters under different inlet wire cross-sectional areas and elongation rates through simulation optimization calculation;
the neural network is used for constructing a mapping relation between the average stress value of the inner hole surface of the drawing die and the initial section area and the elongation of the parent metal;
the genetic algorithm is used for determining the value range of drawing passes of the die and carrying out optimization solution on the elongation rate under each pass;
the software modeling is based on a continuous 'linear' deformation hole pattern design method, the mold structure of the whole drawing process is designed, and the mold structure under each pass is divided according to the design value of the elongation.
Preferably, the neural network is trained by taking the incoming line cross-sectional area and the elongation as inputs and taking the optimal geometric parameters of the die and the corresponding average stress value of the surface of the inner hole of the die as outputs, so as to construct a BP neural network topological structure of the mapping relation between the average stress value of the surface of the inner hole of the drawing die and the initial cross-sectional area and the elongation of the parent metal.
Preferably, the local load balancing means: the variation range of the equivalent stress of each section of the inner hole surface of the drawing die with the same inlet wire cross-sectional area and elongation but different geometric parameters is the minimum value range.
Preferably, the global load balancing means: in the whole drawing process of multi-pass drawing, the change range of the average stress value borne by the surface of the inner hole of the drawing die in each pass is the minimum value range.
Specifically, the design method of the special-shaped cable drawing die comprises the following steps:
the method comprises the following steps of firstly, selecting an orthogonal test design method, and under the condition that the elongation percentage and the initial sectional area of a drawing die are the same, carrying out finite element simulation by using the geometric parameters of the drawing die, including a compression area half angle, a compression area length and a sizing area length, as design variables, and calculating the equivalent stress of each section of the surface of an inner hole of the drawing die and the average stress value of the surface of the inner hole under different geometric parameters;
secondly, comparing the value ranges of equivalent stress of all sections of the surface of the inner hole of the drawing die under different geometric parameters in the first step in order to meet the requirement of die local load balance, wherein the geometric parameter corresponding to the drawing die with the minimum value range is the optimal die design geometric parameter under the elongation and the initial sectional area;
changing the elongation and the initial sectional area of the drawing die, and repeating the simulation calculation of the first step to obtain equivalent stress and average stress values under different elongations and initial sectional areas;
fourthly, constructing a neural network system, taking the cross section area and the elongation of the incoming line as input, taking the optimal geometric parameters of the die and the average stress value of the corresponding inner hole surface of the die as output, training, and constructing a BP neural network topological structure of the mapping relation between the average stress value of the inner hole surface of the drawing die and the initial cross section area and the elongation of the parent metal;
fifthly, taking the safety coefficient of each sub-drawing in the multi-pass drawing process as an optimization target, taking the elongation of each sub-drawing as an optimization parameter, adopting a genetic algorithm as an optimizer to determine the value range of the drawing pass, and solving the elongation of each sub-drawing under different drawing passes;
sixthly, under different drawing passes, according to the elongation design value of each sub-pass, obtaining the optimal die design geometric parameter value and the corresponding average stress value of the inner hole surface in the neural network;
step seven, comparing the variation ranges of the average stress values among the sub-passes under different drawing passes obtained in the step six to meet the overall load balance of the die, wherein the die drawing pass with the minimum variation range of the average stress is the optimal drawing pass of the die;
and eighthly, designing the die structure in the whole drawing process based on a continuous linear deformation hole pattern design method, and dividing the die structure in each pass according to the design value of the elongation.
More preferably, in the fifth step, the safety factor K is between 1.40 and 2.00.
More preferably, the eighth step specifically includes the following steps:
(1) drawing a finished product size graph and a blank size graph of a rectangular line;
(2) drawing a ruled surface;
(3) extracting the cross section of each pass of the die;
(4) and extracting a section design drawing, and converting the designed sample line into a section curve consisting of line segments and circular arcs.
According to a second aspect of the present invention, there is provided a system for designing a profiled cable drawing die for implementing the above method, comprising:
a finite element simulation module for calculating two part stresses, one part being: the section area and the elongation of the inlet wire are the same, but the geometric parameters are different, the equivalent stress of each section of the surface of the inner hole of the drawing die and the average stress value of the surface of the inner hole; comparing to obtain the geometric parameters of the die with the minimum equivalent stress variation range of each section of the surface of the inner hole, wherein the obtained geometric parameters are the optimal die geometric parameters meeting the local load balance under the cross section area and the elongation of the inlet wire; the other part is that: after the cross section area and the elongation of the inlet wire are changed for many times, the equivalent stress of each section of the surface of the inner hole of the drawing die corresponding to different geometric parameters and the average stress value of the surface of the inner hole are obtained; obtaining optimal die geometric parameters under different inlet wire cross-sectional areas and elongation rates through simulation optimization calculation;
the neural network module takes the incoming line cross-sectional area and the elongation as input, takes the optimal geometric parameters of the die and the average stress value of the corresponding inner hole surface of the die as output for training, and constructs a BP neural network topological structure of the mapping relation between the average stress value of the inner hole surface of the drawing die and the initial cross-sectional area and the elongation of the parent metal;
the genetic algorithm module determines the value range of drawing passes by using a genetic algorithm and optimizes and solves the elongation value of each pass under different drawing passes;
and the software modeling module is used for designing the die structure in the whole drawing process based on a continuous 'linear' deformation hole pattern design method, and dividing the die structure in each pass according to the design value of the elongation.
Preferably, the genetic algorithm module finds the optimal die design geometric parameter value and the corresponding average stress value of the inner bore surface in the neural network according to the elongation design value of each sub-pass under different drawing passes.
Preferably, in order to satisfy the global load balance of the die, the genetic algorithm module calculates the variation range of the average stress value between each sub-pass under different drawing passes, and compares the variation ranges to obtain the die drawing pass with the minimum variation range of the average stress, wherein the drawing pass is the optimal design pass of the die.
Compared with the prior art, the invention has the following beneficial effects:
1. the method takes the optimization result of a finite element as a sample, trains an artificial neural network, and establishes a mapping relation between the average stress value of the inner hole surface of the drawing die, the optimal die size parameter and the initial section area and elongation of the parent metal;
2. the invention adopts a genetic optimization algorithm to optimize the distribution of the elongation rate in the multi-pass drawing;
3. the invention has balanced local and global loads borne by the die in the multi-pass drawing process, and the forming size of the former pass is the initial size of the compression area of the latter pass, so as to ensure that the compression area of the latter pass is simultaneously contacted with the wire rod, and reduce the nonuniformity of internal stress strain of the wire rod, thereby designing the hole type section of the wire drawing die which not only accords with the hole type change rule of gradual transition, but also meets the speed ratio parameter of a wire drawing machine.
In conclusion, the artificial neural network is used for calculating the optimized value of the drawing force by virtue of simulation optimization, so that the finite element calculation amount is greatly reduced; the problem of the distribution of the elongation rate in the multi-pass drawing is solved by combining the strong global optimization capability of the genetic algorithm, so that the workload of the design process is remarkably reduced; based on the principle of local and global load balance of the die, the stress balance in the drawing process of each drawing pass is ensured, the abrasion of the die is effectively slowed down, and the service life of the die is prolonged; by means of software modeling and optimized design of the hole pattern of the drawing die, the performance of the drawing die and the drawing equipment can be fully exerted, the service life of the die is greatly prolonged, the time for replacing and maintaining the die is reduced, the production efficiency is greatly improved, and the method is very important for improving the economic benefit of the whole metal wire and pipe drawing industry.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:
FIGS. 1a to 1d are schematic block diagrams of the method of the present invention;
FIG. 2 is a schematic diagram of a pull process according to an embodiment of the present invention;
FIG. 3 shows (a), (b), (c), and (d) are schematic diagrams of a rectangular drawing die according to an embodiment of the present invention;
FIG. 4(a), (b) and (c) are design diagrams of the tile-shaped pull-out mold according to an embodiment of the present invention;
FIG. 5 is a schematic drawing of the tile-shaped wire drawing process according to one embodiment of the present invention;
FIG. 6 is a block diagram of the system of the present invention.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that variations and modifications can be made by persons skilled in the art without departing from the spirit of the invention. All falling within the scope of the present invention.
As shown in fig. 1, a method for designing a special-shaped cable drawing die is a method for designing a multi-pass drawing die matching by combining finite element simulation, an artificial neural network, a genetic algorithm and software modeling on the basis of simulation and artificial intelligence, and the design method can ensure local and global load balance of the die. The method comprises the following steps:
as shown in fig. 1a, the finite element simulation is used to calculate two-part stresses:
one part is as follows: the section area and the elongation of the inlet wire are the same, but the geometric parameters are different, the equivalent stress of each section of the surface of the inner hole of the drawing die and the average stress value of the surface of the inner hole; comparing to obtain the geometric parameters of the die with the minimum equivalent stress variation range of each section of the surface of the inner hole, wherein the obtained geometric parameters are the optimal die geometric parameters meeting the local load balance under the cross section area and the elongation of the inlet wire;
the other part is that: after the cross section area and the elongation of the inlet wire are changed for many times, the equivalent stress of each section of the surface of the inner hole of the drawing die corresponding to different geometric parameters and the average stress value of the surface of the inner hole are obtained; obtaining optimal die geometric parameters under different inlet wire cross-sectional areas and elongation rates through simulation optimization calculation;
the neural network is shown in figure 1b, and is a neural network for constructing a mapping relation between an average stress value of the surface of an inner hole of the drawing die and the initial section area and the elongation of the parent metal;
the genetic algorithm is shown in fig. 1c and is used for determining the value range of the drawing pass of the die and carrying out optimization solution on the elongation rate under each pass;
the software modeling, as shown in fig. 1d, is based on a hole pattern design method of continuous "linear" deformation, and designs the die structure of the whole drawing process, and divides the die structure of each pass according to the design value of elongation.
The method is based on a continuous 'linear' deformation hole pattern design method, namely the forming size of the previous pass is the initial size of the compression area of the next pass in the multi-pass drawing process, so as to ensure that the compression area of the die of the next pass is simultaneously contacted with the wire rod, and reduce the nonuniformity of the internal stress strain of the drawn wire rod as much as possible, thereby designing a wire drawing die hole pattern section model which not only accords with the gradually-transitional hole pattern change rule, but also meets the speed ratio parameter of a wire drawing machine.
Example 1
As shown in fig. 2, the present embodiment provides a method for designing a profiled cable drawing die, which is performed through the following steps:
first step, simulation optimization
Firstly, establishing a geometric model of a die and a rectangular wire in a pretreatment module of an ANSYS Workbench; then loading friction coefficients obtained through experiments and stress strain data of the metal materials into the model, and dividing grids to obtain a finite element model for simulating the drawing process; in the simulation process, the bottom of the drawing die is applied with fixed constraint, the axial displacement load of-60 mm is applied to the front end of the wire rod and is used for simulating the applied drawing force, the drawing speed is set to be 3m/s, and the simulation time is 0.02 s; selecting a copper material as a rectangular wire material, determining the initial cross section area and the elongation value of the wire, then selecting an orthogonal test design method, taking the geometric parameters (the half angle range of a compression zone is 10-18 degrees, the length of the compression zone is 1-5mm, and the length of a sizing zone is 2-6mm) of the drawing die as design variables, carrying out finite element simulation, and calculating the equivalent stress of each section of the surface of the inner hole of the drawing die and the average stress value of the surface of the inner hole under different geometric parameters.
Second step, local load balancing of the mold
And comparing to obtain the geometric parameters of the die with the minimum equivalent stress variation range of each section of the surface of the inner hole, wherein the obtained geometric parameters are the optimal die geometric parameters meeting the local load balance under the inlet wire section area and the elongation.
Third, expanding the design sample of the mold
And changing the initial cross-sectional area and the elongation percentage of the rectangular wire for multiple times, repeating the modeling and post-treatment processes in the first step, then selecting an orthogonal test design method, carrying out finite element simulation, calculating the equivalent stress of each section of the surface of the inner hole of the drawing die under different geometric parameters and the average stress value of the surface of the inner hole, screening out the corresponding optimal die geometric parameters under different initial cross-sectional areas and elongation percentages, and obtaining more die design parameter samples.
Step four, establishing a BP artificial neural network
Taking the incoming line cross-sectional area and the elongation as input, taking the optimal geometric parameters of the die and the average stress value of the corresponding inner hole surface of the die as output for training the established neural network system, and constructing a BP neural network topological structure of the mapping relation between the average stress value of the inner hole surface of the drawing die and the initial cross-sectional area and the elongation of the parent metal;
the neural network structure is as follows:
(1) inputting: 2 input parameters, namely the initial sectional area and the elongation of the parent metal;
(2) hidden layer: two layers with 10 neurons and 5 neurons, respectively;
(3) an output layer: 6 neurons are the objective function value; the hidden layer adopts logsig as a transfer function, and the output layer adopts prulin as a transfer function;
carrying out neural network training on the established network model, and setting a minimum expected error value err _ coarse to be 0.001; setting the maximum cycle number max _ epoch to 35000; setting the learning rate lr of the correction weight to be 0.01; after 1000 times of training, the training is finished.
Fifthly, solving the value range of the total drawing pass and the elongation of each pass by a genetic algorithm
Taking the safety coefficient of each sub-drawing in the multi-pass drawing process as an optimization target, taking the elongation of each sub-drawing as an optimization parameter, adopting a genetic algorithm as an optimizer to determine the value range of the drawing pass, and solving the elongation of each sub-drawing under different drawing passes;
in the present example, the cross-sectional area of the base material was 70.89, and the elongation was 7.8; selecting 5-10 passes as the optimized pass range of the drawing according to the existing drawing experience, respectively calculating the optimal result of the multi-pass drawing of the copper wire rod from 5 passes to 10 passes by adopting a genetic algorithm, and as can be seen from the calculated result, the safety coefficient K is increased along with the increase of the pass number N, and the wire breakage phenomenon is more easily caused in the drawing process with the less pass number; observing the optimization result can find that:
when the number N of the drawing passes is 5, the safety coefficient K is about 1.02;
when the number N of the drawing passes is 6, the safety coefficient K is about 1.14;
when the number N of the drawing passes is 7, the safety coefficient K is about 1.27;
when the number N of the drawing passes is 8, the safety coefficient K is about 1.4;
when the number N of the drawing passes is 9, the safety coefficient K is between 1.48 and 1.58;
when the number N of the drawing passes is 10, the safety coefficient K is between 1.7 and 2.1;
according to the calculation result, when the number N of the drawing passes is 8 and 9, the safety coefficient K is between 1.4 and 2.0, and the condition of safe drawing is met. Therefore, the optimization result of the final drawing pass is 8 or 9 passes.
Wherein the parameters of the genetic algorithm are set as follows:
taking 200 of population scale; the crossover operator adopts two-point arithmetic crossover, and the crossover probability is 0.8; the mutation operator adopts Gaussian mutation, and the mutation probability is 0.2; the maximum number of genetic generations was 500.
Sixthly, global load balancing of the die
When the drawing pass is 8, the optimum geometric parameter values of the die and the corresponding average stress values of the inner hole surface (511.65MPa, 454.1MPa, 467.33MPa, 471.59MPa, 539.94MPa, 556.57MPa, 589.7MPa and 598.97MPa) of each pass are determined in a neural network by using the initial cross-sectional area 70.89 of the base material and the elongation values (1.31, 1.3, 1.3, 1.25, 1.28, 1.34 and 1.26) of each pass as known conditions, and the variation range of the average stress of the inner hole surface of the die at the pass is found as follows: 454.1MPa to 598.97 MPa;
when the drawing pass is 9, the initial cross-sectional area 70.89 of the parent metal and the elongation values (1.28, 1.26, 1.27, 1.26, 1.23, 1.23, 1.26, 1.35, 1.18) of each sub-pass are used as known conditions, the optimal geometric parameter value of the die and the corresponding average stress value of the inner bore surface (509.09MPa, 301.28MPa, 396.2MPa, 450.71MPa, 538.31MPa, 599.3MPa, 632.1MPa, 574.33MPa, 611.5MPa) of each sub-pass are obtained in the neural network, and the variation range of the average stress of the inner bore surface of the die under the sub-pass is known as follows: 301.28MPa to 632.1 MPa;
comparing the variation range of the average stress of the inner hole surface of the die in the two drawing passes, the variation range of the stress in the 9 passes is smaller, so that the 9 passes are selected as the drawing passes of the rectangular wire drawing die.
Seventh step, AutoCAD modeling
According to the optimized design parameters of each pass, using AutoCAD software to perform modeling:
(1) drawing a rectangular line finished product size diagram and a blank size diagram: drawing a finished product size diagram of the rectangular electromagnetic wire by using CAD software, establishing a surface area, finding out the barycentric coordinate of the surface area, and drawing the size of an incoming wire blank at the position 10mm higher than the barycentric position, as shown in figure 3 (a).
(2) Drawing a ruled surface: and (3) selecting an upper section and a lower section (a finished product size diagram and a blank size diagram of a rectangular wire) for lofting to obtain a ruled surface gradually transiting from a circular section to a molded wire, as shown in fig. 3 (b).
(3) Extracting the section of each pass of the die: and according to the selected speed ratio, firstly calculating the target area of the intermediate transition section as the basis for determining the section height in the subsequent step. The middle transition section is cut with slice command to make its area meet certain speed ratio requirement (if the cut area is larger than the target area, the cut height is reduced; vice versa), as shown in fig. 3 (c).
(4) And newly building a layer, selecting all curved surfaces under the new layer, and extracting edges by using an xedges command to obtain a section curve of a sample line, namely the section design drawing of each required pass. Since the spline line obtained here cannot be recognized by the wire-cut electrical discharge machining software when the wire-cut electrical discharge machining is performed using the slow wire, the designed spline line should be converted into a cross-sectional curve composed of a line segment and a circular arc (for example, conversion by a dove-tail kit software can be adopted), as shown in fig. 3 (d).
(5) The half angle of a compression area of the special-shaped drawing die is controlled by changing the height of the special-shaped drawing die, and machining production is realized on a slow wire-moving wire cut electrical discharge machine.
Example 2
As shown in fig. 4 and 5, the present embodiment provides a method for designing a special-shaped cable drawing die, which comprises the following steps:
first step, simulation optimization
Firstly, a geometric model of a die and a tile-shaped wire rod is established in a pretreatment module of an ANSYS Workbench. And then loading the friction coefficient obtained through experiments and the stress-strain data of the metal material into the model, and dividing the grids to obtain the finite element model for simulating the drawing process. In the simulation process, the bottom of the drawing die is applied with fixed constraint, an axial displacement load of-60 mm is applied to the front end of the wire rod and used for simulating the applied drawing force, the drawing speed is set to be 3m/s, and the simulation time is 0.02 s. The method comprises the steps of selecting a rectangular wire material as a low-carbon steel, determining an initial cross-sectional area and an elongation value of the wire, selecting an orthogonal test design method, taking geometric parameters (a compression area half-angle range is 4-10 degrees, a compression area length is 2-8mm, and a sizing area length value is 1-6mm) of a drawing die as design variables, carrying out finite element simulation, and calculating equivalent stress of each section of the surface of an inner hole of the drawing die and an average stress value of the surface of the inner hole under different geometric parameters.
Second step, local load balancing of the mold
And comparing to obtain the geometric parameters of the die with the minimum equivalent stress variation range of each section of the surface of the inner hole, wherein the obtained geometric parameters are the optimal die geometric parameters meeting the local load balance under the inlet wire section area and the elongation.
Third, expanding the design sample of the mold
And changing the initial cross section area and the elongation percentage of the tile-shaped wire for multiple times, repeating the modeling and post-treatment processes in the first step, then selecting an orthogonal test design method, carrying out finite element simulation, calculating the equivalent stress of each section of the surface of the inner hole of the drawing die under different geometric parameters and the average stress value of the surface of the inner hole, screening out the corresponding optimal geometric parameters of the die under different initial cross section areas and elongation percentages, and obtaining more die design parameter samples.
Step four, establishing a BP artificial neural network
Taking the incoming line cross-sectional area and the elongation as input, taking the optimal geometric parameters of the die and the average stress value of the corresponding inner hole surface of the die as output for training the established neural network system, and constructing a BP neural network topological structure of the mapping relation between the average stress value of the inner hole surface of the drawing die and the initial cross-sectional area and the elongation of the parent metal;
the neural network structure is as follows:
(1) inputting: 2 input parameters, namely the initial sectional area and the elongation of the parent metal;
(2) hidden layer: two layers with 10 neurons and 5 neurons, respectively;
(3) an output layer: the objective function values are 6 neurons. The hidden layer adopts logsig as a transfer function, and the output layer adopts prulin as a transfer function.
Carrying out neural network training on the established network model, and setting a minimum expected error value err _ coarse to be 0.001; setting the maximum cycle number max _ epoch to 35000; the learning rate lr of the correction weight is set to 0.01. After 1500 training passes, the training is finished.
Fifthly, solving the value range of the total drawing pass and the elongation of each pass by a genetic algorithm
Taking the safety coefficient of each sub-drawing in the multi-pass drawing process as an optimization target, taking the elongation of each sub-drawing as an optimization parameter, adopting a genetic algorithm as an optimizer to determine the value range of the drawing pass, and solving the elongation of each sub-drawing under different drawing passes;
in this example, the cross-sectional area of the base material was 152.84, and the elongation was 6.7. According to the existing drawing experience, 5-10 passes are selected as the optimized pass range of the drawing in the embodiment, the genetic algorithm is adopted to respectively calculate the optimal result of the multi-pass drawing of the low-carbon steel wire rod from 5 passes to 10 passes, and the calculated result shows that the safety coefficient K is increased along with the increase of the pass number N, and the wire breakage phenomenon is more likely to occur in the drawing process with the smaller pass number. Observing the optimization result can find that:
when the number N of the drawing passes is 5, the safety coefficient K is about 1.01;
when the number N of the drawing passes is 6, the safety coefficient K is about 1.09;
when the number N of the drawing passes is 7, the safety coefficient K is about 1.14;
when the number N of the drawing passes is 8, the safety coefficient K is about 1.32;
when the number N of the drawing passes is 9, the safety coefficient K is 1.42;
when the number N of the drawing passes is 10, the safety factor K is between 1.5 and 1.89.
The calculation result shows that when the number N of the drawing passes is 9 and 10, the safety coefficient K is between 1.4 and 2.0, and the condition of safe drawing is met. Therefore, the optimization result of the final drawing pass is 9 or 10 passes.
Wherein the parameters of the genetic algorithm are set as follows:
taking 300 of population scale; the crossover operator adopts two-point arithmetic crossover, and the crossover probability is 0.8; the mutation operator adopts Gaussian mutation, and the mutation probability is 0.2; the maximum genetic passage number is 800.
Sixthly, global load balancing of the die
When the drawing pass is 9, the initial cross-sectional area of the parent metal 152.84 and the elongation values (1.15, 1.18, 1.22, 1.23, 1.22, 1.24, 1.30, 1.31, 1.28) of each sub-pass are used as known conditions, the optimal geometric parameter values of the die and the corresponding average stress values (650.126MPa, 557.048MPa, 671.503MPa, 732.618MPa, 733.409MPa, 748.226MPa, 694.598MPa, 701.97MPa, 711.55MPa) of the inner hole surface of the die in each sub-pass are obtained in a neural network, and the change range of the average stress of the inner hole surface of the die in the sub-pass is known as follows: 557.048MPa to 748.226 MPa;
when the drawing pass is 10, the initial cross-sectional area of the parent material 152.84 and the elongation values (1.09, 1.13, 1.18, 1.23, 1.24, 1.23, 1.25, 1.24, 1.26, 1.27) of the respective sub-passes are used as known conditions, the optimal geometric parameter values of the die and the corresponding average stress values (562.22MPa, 536.18MPa, 668.21MPa, 640.71MPa, 690.32MPa, 710.55MPa, 632.16MPa, 707.35MPa, 699.5MPa, 720.64MPa) of the inner hole surface of the die in each sub-pass are obtained in a neural network, and the variation range of the average stress of the inner hole surface of the die in the sub-pass is found to be: 536.18MPa to 720.64 MPa;
comparing the variation range of the average stress of the inner hole surface of the die in the two drawing passes, the variation range of the stress in 10 passes is smaller, so 10 passes are selected as the drawing passes of the rectangular wire drawing die.
Seventh step, AutoCAD modeling
According to the optimized design parameters of each pass, using AutoCAD software to perform modeling:
(1) drawing a rectangular line finished product size diagram and a blank size diagram: drawing a finished product size diagram of the rectangular electromagnetic wire by using CAD software, establishing a face area, finding out the barycentric coordinate of the face area, and drawing the size of an incoming wire blank at the position 25mm higher than the barycentric position, as shown in FIG. 4 (a).
(2) Drawing a ruled surface: and (3) selecting an upper section and a lower section (a finished product size diagram and a blank size diagram of the tile-shaped line) for lofting to obtain a ruled surface gradually transiting from the circular section to the molded line, as shown in fig. 4 (b).
(3) Extracting the section of each pass of the die: and according to the selected speed ratio, firstly calculating the target area of the intermediate transition section as the basis for determining the section height in the subsequent step. And (5) intercepting the intermediate transition section by using a slice command, so that the area of the intermediate transition section meets a certain speed ratio requirement (if the intercepted area is larger than the target area, the intercepting height is reduced, and vice versa).
(4) And newly building a layer, selecting all curved surfaces under the new layer, and extracting edges by using an xedges command to obtain a section curve of a sample line, namely the section design drawing of each required pass. Since the spline line obtained here cannot be recognized by wire-cut electrical discharge machining software when wire-cut electrical discharge machining is performed using a slow wire, the designed spline line should be converted into a cross-sectional curve composed of a line segment and a circular arc, as shown in fig. 4 (c).
(5) The half angle of a compression area of the special-shaped drawing die is controlled by changing the height of the special-shaped drawing die, and machining production is realized on a slow wire-moving wire cut electrical discharge machine.
The method for drawing the special-shaped wire rod can design the hole type section of the drawing die which not only accords with the gradually-transitional hole type change rule, but also meets the speed ratio parameter of the drawing machine. In the actual production and processing process of the die, the performance of the drawing die and the drawing equipment can be fully exerted, the service life of the die is greatly prolonged in the actual die application, the time for replacing and maintaining the die is also reduced, the production efficiency is greatly improved, and the method is very important for improving the economic benefit of the whole metal wire and pipe drawing industry.
Example 3
The embodiment provides a design method of a special-shaped cable drawing die, which is carried out through the following steps:
first step, simulation optimization
Firstly, a geometric model of a die and an S-shaped wire rod is established in a pretreatment module of an ANSYS Workbench. And then loading the friction coefficient obtained through experiments and the stress-strain data of the metal material into the model, and dividing the grids to obtain the finite element model for simulating the drawing process. In the simulation process, the bottom of the drawing die is applied with fixed constraint, the axial displacement load of-60 mm is applied to the front end of the wire rod and used for simulating the applied drawing force, the drawing speed is set to be 1m/s, and the simulation time is 0.05 s. Selecting a copper material as an S-shaped wire material, determining the initial cross section area and the elongation value of the wire, then selecting an orthogonal test design method, taking the geometric parameters (the half angle range of a compression zone is 6-13 degrees, the length of the compression zone is 4-10mm, and the length of a sizing zone is 3-8mm) of the drawing die as design variables, carrying out finite element simulation, and calculating the equivalent stress of each section of the surface of the inner hole of the drawing die and the average stress value of the surface of the inner hole under different geometric parameters.
Second step, local load balancing of the mold
And comparing to obtain the geometric parameters of the die with the minimum equivalent stress variation range of each section of the surface of the inner hole, wherein the obtained geometric parameters are the optimal die geometric parameters meeting the local load balance under the inlet wire section area and the elongation.
Third, expanding the design sample of the mold
And changing the initial cross section area and the elongation percentage value of the S-shaped wire for multiple times, repeating the modeling and post-processing processes in the first step, then selecting an orthogonal test design method, carrying out finite element simulation, calculating the equivalent stress of each section of the surface of the inner hole of the drawing die under different geometric parameters and the average stress value of the surface of the inner hole, screening out the corresponding optimal geometric parameters of the die under different initial cross section areas and elongation percentages, and obtaining more die design parameter samples.
Step four, establishing a BP artificial neural network
Taking the incoming line cross-sectional area and the elongation as input, taking the optimal geometric parameters of the die and the average stress value of the corresponding inner hole surface of the die as output for training the established neural network system, and constructing a BP neural network topological structure of the mapping relation between the average stress value of the inner hole surface of the drawing die and the initial cross-sectional area and the elongation of the parent metal;
the neural network structure is as follows:
(1) inputting: 2 input parameters, namely the initial sectional area and the elongation of the parent metal;
(2) hidden layer: two layers with 10 neurons and 5 neurons, respectively;
(3) an output layer: the objective function values are 6 neurons. The hidden layer adopts logsig as a transfer function, and the output layer adopts prulin as a transfer function.
Carrying out neural network training on the established network model, and setting a minimum expected error value err _ coarse to be 0.001; setting the maximum cycle number max _ epoch to 35000; the learning rate lr of the correction weight is set to 0.01. After 1500 training passes, the training is finished.
Fifthly, solving the value range of the total drawing pass and the elongation of each pass by a genetic algorithm
Taking the safety coefficient of each sub-drawing in the multi-pass drawing process as an optimization target, taking the elongation of each sub-drawing as an optimization parameter, adopting a genetic algorithm as an optimizer to determine the value range of the drawing pass, and solving the elongation of each sub-drawing under different drawing passes;
in this example, the cross-sectional area of the base material was 85.10, and the elongation was 7.4. According to the existing drawing experience, 5-9 passes are selected as the optimized pass range of the drawing in the embodiment, the genetic algorithm is adopted to respectively calculate the optimal result of the multi-pass drawing of the S-shaped copper wire from 5 passes to 9 passes, and the calculated result shows that the safety coefficient K is increased along with the increase of the pass number N, and the wire breakage phenomenon is more likely to occur in the drawing process with the smaller pass number. Observing the optimization result can find that:
when the number N of the drawing passes is 5, the safety coefficient K is about 1.03-1.08;
when the number N of the drawing passes is 6, the safety coefficient K is about 1.18;
when the number N of the drawing passes is 7, the safety coefficient K is about 1.29;
when the number N of the drawing passes is 8, the safety coefficient K is about 1.42-1.48;
when the number N of the drawing passes is 9, the safety coefficient K is between 1.63;
when the number N of the drawing passes is 10, the safety factor K is between 1.81 and 2.25.
According to the calculation result, when the number N of the drawing passes is 8 and 9, the safety coefficient K is between 1.4 and 2.0, and the condition of safe drawing is met. Therefore, the optimization result of the final drawing pass is 8 or 9 passes.
Wherein the parameters of the genetic algorithm are set as follows:
taking 200 of population scale; the crossover operator adopts two-point arithmetic crossover, and the crossover probability is 0.8; the mutation operator adopts Gaussian mutation, and the mutation probability is 0.2; the maximum number of genetic generations was 500.
Sixthly, global load balancing of the die
When the drawing pass is 8, the initial cross-sectional area 85.10 of the parent metal and the elongation values (1.30, 1.28, 1.29, 1.30, 1.27, 1.25, 1.32, 1.27) of the respective sub-passes are used as known conditions, the optimal geometric parameter value of the die and the corresponding average stress value of the inner hole surface (612.11MPa, 553.20MPa, 637.28MPa, 590.19MPa, 648.70MPa, 699.37MPa, 710.71MPa, 640.33MPa) of each sub-pass are obtained in the neural network, and the result shows that the variation range of the average stress of the inner hole surface of the die under the pass is: 553.20MPa to 710.71 MPa;
when the drawing pass is 9, the initial cross-sectional area 85.10 of the parent metal and the elongation values (1.28, 1.25, 1.23, 1.24, 1.24, 1.26, 1.25, 1.30, 1.20) of each sub-pass are used as known conditions, the optimal geometric parameter value of the die and the corresponding average stress value of the inner bore surface (521.16MPa, 501.33MPa, 579.30MPa, 611.52MPa, 629.09MPa, 575.32MPa, 671.69MPa, 570.21MPa, 567.18MPa) of each sub-pass are obtained in a neural network, and the change range of the average stress of the inner bore surface of the die under the sub-pass is known as follows: 501.33MPa to 671.69 MPa;
comparing the variation range of the average stress of the inner hole surface of the die in two drawing passes, the variation range of the stress in 8 passes is smaller, so 8 passes are selected as the drawing passes of the rectangular wire drawing die.
Seventh step, AutoCAD modeling
According to the optimized design parameters of each pass, using AutoCAD software to perform modeling:
(1) drawing a rectangular line finished product size diagram and a blank size diagram: drawing a finished product size graph of the rectangular electromagnetic wire by using CAD software, establishing a face area, finding out the gravity center coordinate of the face area, and drawing the size of an incoming wire blank at the position 13mm higher than the gravity center position.
(2) Drawing a ruled surface: and (3) selecting an upper section and a lower section (a finished product size diagram and a blank size diagram of the S-shaped line) for lofting to obtain a ruled surface gradually transiting from the circular section to the molded line.
(3) Extracting the section of each pass of the die: and according to the selected speed ratio, firstly calculating the target area of the intermediate transition section as the basis for determining the section height in the subsequent step. And (5) intercepting the intermediate transition section by using a slice command, so that the area of the intermediate transition section meets a certain speed ratio requirement (if the intercepted area is larger than the target area, the intercepting height is reduced, and vice versa).
(4) And newly building a layer, selecting all curved surfaces under the new layer, and extracting edges by using an xedges command to obtain a section curve of a sample line, namely the section design drawing of each required pass. Since the spline line obtained here cannot be recognized by the wire-cut electrical discharge machining software when the wire-cut electrical discharge machining is performed using a slow-moving wire, the designed spline line should be converted into a cross-sectional curve composed of a line segment and an arc by means of the dove-show toolbox software.
(5) The half angle of a compression area of the special-shaped drawing die is controlled by changing the height of the special-shaped drawing die, and machining production is realized on a slow wire-moving wire cut electrical discharge machine.
Example 4
As shown in fig. 6, corresponding to the methods described in embodiments 1 to 3, the present embodiment provides a design system of a profiled cable drawing die for implementing the above methods, including:
a finite element simulation module for calculating two part stresses, one part being: the section area and the elongation of the inlet wire are the same, but the geometric parameters are different, the equivalent stress of each section of the surface of the inner hole of the drawing die and the average stress value of the surface of the inner hole; comparing to obtain the geometric parameters of the die with the minimum equivalent stress variation range of each section of the surface of the inner hole, wherein the obtained geometric parameters are the optimal die geometric parameters meeting the local load balance under the cross section area and the elongation of the inlet wire; the other part is that: after the cross section area and the elongation of the inlet wire are changed for many times, the equivalent stress of each section of the surface of the inner hole of the drawing die corresponding to different geometric parameters and the average stress value of the surface of the inner hole are obtained; obtaining optimal die geometric parameters under different inlet wire cross-sectional areas and elongation rates through simulation optimization calculation;
the neural network module takes the incoming line cross-sectional area and the elongation of the finite element simulation module as input, takes the optimal geometric parameters of the die and the average stress value of the corresponding inner hole surface of the die as output for training, and constructs a BP neural network topological structure of the mapping relation between the average stress value of the inner hole surface of the drawing die and the initial cross-sectional area and the elongation of the parent metal;
the genetic algorithm module is used for determining the value range of drawing passes by using a genetic algorithm, optimizing and solving the elongation value of each pass under different drawing passes, then taking the obtained drawing passes and the initial sectional area of the parent metal as the input of the trained neural network module, and solving the optimal drawing pass which meets the minimum change range of the average stress of the surface of the inner hole of the die;
and the software modeling module is used for designing the mold structure in the whole drawing process according to a continuous 'linear' deformed hole pattern design method, and dividing the mold structure under each pass according to the design value of the elongation rate of each pass calculated by the genetic algorithm module.
The corresponding implementation techniques of each module in the system are similar to the corresponding steps in the method, and are not described herein again.
In conclusion, the invention optimizes the hole pattern of the drawing die by means of simulation, optimization and software design, not only can fully play the performances of the drawing die and the drawing equipment, but also can greatly prolong the service life of the die and reduce the time for replacing and maintaining the die for a high-speed automatic drawing production line, thereby greatly improving the production efficiency and being very important for improving the economic benefit of the whole metal wire and pipe drawing industry.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes and modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention.
Claims (9)
1. A design method of a special-shaped cable drawing die is characterized by comprising the following steps: on the basis of simulation and artificial intelligence, finite element simulation, an artificial neural network, a genetic algorithm and software modeling are combined, and the method can ensure that the designed die has the characteristics of local load balance and global load balance;
the finite element simulation is used for calculating two parts of stress:
one part is as follows: the section area and the elongation of the inlet wire are the same, but the geometric parameters are different, the equivalent stress of each section of the surface of the inner hole of the drawing die and the average stress value of the surface of the inner hole; comparing to obtain the geometric parameters of the die with the minimum equivalent stress variation range of each section of the surface of the inner hole, wherein the obtained geometric parameters are the optimal die geometric parameters meeting the local load balance under the cross section area and the elongation of the inlet wire;
the other part is that: after the cross section area and the elongation of the inlet wire are changed for many times, the equivalent stress of each section of the surface of the inner hole of the drawing die corresponding to different geometric parameters and the average stress value of the surface of the inner hole are obtained; obtaining optimal die geometric parameters under different inlet wire cross-sectional areas and elongation rates through simulation optimization calculation;
the neural network is used for constructing a mapping relation between the average stress value of the inner hole surface of the drawing die and the initial section area and the elongation of the parent metal;
the genetic algorithm is used for determining the value range of drawing passes of the die and carrying out optimization solution on the elongation rate under each pass;
the software modeling is based on a continuous 'linear' deformation hole pattern design method, a die structure of the whole drawing process is designed, and the die structure under each pass is divided according to a design value of the elongation;
the neural network is trained by taking the cross-sectional area and the elongation of the incoming line as input and taking the optimal geometric parameters of the die and the average stress value of the corresponding inner hole surface of the die as output, so that the BP neural network topological structure of the mapping relation between the average stress value of the inner hole surface of the drawing die and the initial cross-sectional area and the elongation of the parent metal is constructed.
2. The method for designing a profiled cable drawing die according to claim 1, wherein the local load balancing is: the variation range of the equivalent stress of each section of the inner hole surface of the drawing die with the same inlet wire cross-sectional area and elongation but different geometric parameters is the minimum value range.
3. The method for designing the special-shaped cable drawing die as claimed in claim 1, wherein the global load balancing is that: in the whole drawing process of multi-pass drawing, the change range of the average stress value borne by the surface of the inner hole of the drawing die in each pass is the minimum value range.
4. The method for designing a profiled cable drawing die as claimed in any one of claims 1 to 3, wherein the method comprises the steps of:
the method comprises the following steps of firstly, selecting an orthogonal test design method, carrying out finite element simulation by using geometric parameters of a drawing die, including a compression area half angle, a compression area length and a sizing area length, as design variables under the condition that the elongation and the initial sectional area of the drawing die are the same, and calculating equivalent stress of each section of the surface of an inner hole of the drawing die and an average stress value of the surface of the inner hole under different geometric parameters;
secondly, comparing the value ranges of equivalent stress of all sections of the surface of the inner hole of the drawing die under different geometric parameters in the first step in order to meet the requirement of die local load balance, wherein the geometric parameter corresponding to the drawing die with the minimum value range is the optimal die design geometric parameter under the elongation and the initial sectional area;
changing the elongation and the initial sectional area of the drawing die, and repeating the simulation calculation of the first step to obtain equivalent stress and average stress values under different elongations and initial sectional areas;
fourthly, constructing a neural network system, taking the cross section area and the elongation of the incoming line as input, taking the optimal geometric parameters of the die and the average stress value of the corresponding inner hole surface of the die as output, training, and constructing a BP neural network topological structure of the mapping relation between the average stress value of the inner hole surface of the drawing die and the initial cross section area and the elongation of the parent metal;
fifthly, taking the safety coefficient of each sub-drawing in the multi-pass drawing process as an optimization target, taking the elongation of each sub-drawing as an optimization parameter, adopting a genetic algorithm as an optimizer to determine the value range of the drawing pass, and solving the elongation of each sub-drawing under different drawing passes; then, the obtained drawing pass and the initial section area of the parent metal are used as the input of the trained neural network system;
sixthly, under different drawing passes, according to the elongation design value of each sub-pass, obtaining the optimal die design geometric parameter value and the corresponding average stress value of the inner hole surface in the neural network system;
step seven, comparing the variation ranges of the average stress values among the sub-passes under different drawing passes obtained in the step six to meet the overall load balance of the die, wherein the die drawing pass with the minimum variation range of the average stress is the optimal drawing pass of the die;
and eighthly, designing the die structure in the whole drawing process by using AutoCAD software based on a continuous 'linear' deformed hole pattern design method, and segmenting the die structure under each pass according to the design value of the elongation rate of each pass obtained by a genetic algorithm.
5. The design method of the special-shaped cable drawing die as claimed in claim 4, wherein in the fifth step, the safety coefficient K is between 1.40 and 2.00.
6. The design method of the special-shaped cable drawing die as claimed in claim 4, wherein in the eighth step, the method specifically comprises the following steps:
(1) drawing a finished product size diagram and a blank size diagram;
(2) drawing a ruled surface;
(3) extracting the cross section of each pass of the die;
(4) and extracting a section design drawing, and converting the designed sample line into a section curve consisting of line segments and circular arcs.
7. A design system of a profiled cable drawing die for implementing the method of any one of the preceding claims 1 to 6, characterized in that: the method comprises the following steps:
a finite element simulation module for calculating two part stresses, one part being: the section area and the elongation of the inlet wire are the same, but the geometric parameters are different, the equivalent stress of each section of the surface of the inner hole of the drawing die and the average stress value of the surface of the inner hole; comparing to obtain the geometric parameters of the die with the minimum equivalent stress variation range of each section of the surface of the inner hole, wherein the obtained geometric parameters are the optimal die geometric parameters meeting the local load balance under the cross section area and the elongation of the inlet wire; the other part is that: after the cross section area and the elongation of the inlet wire are changed for many times, the equivalent stress of each section of the surface of the inner hole of the drawing die corresponding to different geometric parameters and the average stress value of the surface of the inner hole are obtained; obtaining optimal die geometric parameters under different inlet wire cross-sectional areas and elongation rates through simulation optimization calculation;
the neural network module takes the incoming line cross-sectional area and the elongation of the finite element simulation module as input, takes the optimal geometric parameters of the die and the average stress value of the corresponding inner hole surface of the die as output for training, and constructs a BP neural network topological structure of the mapping relation between the average stress value of the inner hole surface of the drawing die and the initial cross-sectional area and the elongation of the parent metal;
the genetic algorithm module is used for determining the value range of drawing passes by using a genetic algorithm, optimizing and solving the elongation value of each sub-pass under different drawing passes, then taking the obtained drawing passes and the initial section area of the parent metal as the input of the trained neural network module, and solving the optimal drawing pass which meets the minimum change range of the average stress of the surface of the inner hole of the die;
and the software modeling module is used for designing the mold structure in the whole drawing process according to a continuous 'linear' deformed hole pattern design method, and dividing the mold structure under each pass according to the design value of the elongation rate of each pass calculated by the genetic algorithm module.
8. The system for designing the special-shaped cable drawing die according to claim 7, wherein the genetic algorithm module is used for solving an optimal die design geometric parameter value and a corresponding average stress value of the inner hole surface in a neural network according to the elongation design value of each sub-pass under different drawing passes.
9. The system for designing the special-shaped cable drawing die according to claim 7, wherein in order to meet the global load balance of the die, the genetic algorithm module calculates the variation range of the average stress value among the sub-passes in different drawing passes, and compares the variation ranges to obtain the die drawing pass with the minimum variation range of the average stress, wherein the drawing pass is the optimal design pass of the die.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610545822.7A CN106202727B (en) | 2016-07-12 | 2016-07-12 | Design method and system of special-shaped cable drawing die |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610545822.7A CN106202727B (en) | 2016-07-12 | 2016-07-12 | Design method and system of special-shaped cable drawing die |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106202727A CN106202727A (en) | 2016-12-07 |
| CN106202727B true CN106202727B (en) | 2020-05-22 |
Family
ID=57477767
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610545822.7A Active CN106202727B (en) | 2016-07-12 | 2016-07-12 | Design method and system of special-shaped cable drawing die |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106202727B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108509666B (en) * | 2017-02-27 | 2021-05-25 | 中国科学院金属研究所 | Pure aluminum wire multi-pass cold drawing plastic deformation mode prediction method based on ANSYS/LS-dyna |
| CN108090282B (en) * | 2017-12-18 | 2021-03-05 | 上海理工大学 | Multi-physical-field coupling simulation analysis method for machining gap of slow wire-cut electrical discharge machining |
| CN110298070B (en) * | 2019-05-15 | 2023-12-08 | 成都数模码科技有限公司 | Standard punch fixing seat generation and distribution method |
| CN112026132B (en) * | 2020-08-11 | 2022-03-29 | 江苏亨通高压海缆有限公司 | Special-shaped single line die for direct-current submarine cable and design method thereof |
| CN115705443A (en) * | 2021-08-04 | 2023-02-17 | 徐州徐工液压件有限公司 | Method for manufacturing steel pipe cold-drawing die |
| CN114798779B (en) * | 2022-03-23 | 2023-06-02 | 中南大学 | Intelligent metal drawing processing device and method based on neural network |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1624396A1 (en) * | 2004-07-09 | 2006-02-08 | Yazaki Corporation | Method of calculating the shape of a wiring structure, calculating apparatus, and computer -readable recording medium |
| CN105200391A (en) * | 2014-05-29 | 2015-12-30 | 上海交友钻石涂层有限公司 | Preparation method of diamond-coated drawing mold for water lubrication |
| CN105444935A (en) * | 2015-12-18 | 2016-03-30 | 深圳供电局有限公司 | Portable knob insulator stress detection device |
-
2016
- 2016-07-12 CN CN201610545822.7A patent/CN106202727B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1624396A1 (en) * | 2004-07-09 | 2006-02-08 | Yazaki Corporation | Method of calculating the shape of a wiring structure, calculating apparatus, and computer -readable recording medium |
| CN105200391A (en) * | 2014-05-29 | 2015-12-30 | 上海交友钻石涂层有限公司 | Preparation method of diamond-coated drawing mold for water lubrication |
| CN105444935A (en) * | 2015-12-18 | 2016-03-30 | 深圳供电局有限公司 | Portable knob insulator stress detection device |
Non-Patent Citations (1)
| Title |
|---|
| Numerical and experimental investigation of trapezoidal wire cold drawing through a series of shaped dies;Lin Z 等;《The International Journal of Advanced Manufacturing Technology》;20150228;第76卷(第5-8期);第1383-1391页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106202727A (en) | 2016-12-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106202727B (en) | Design method and system of special-shaped cable drawing die | |
| Ko et al. | Application of artificial neural network and Taguchi method to preform design in metal forming considering workability | |
| Kukhar et al. | Shape indexes for dieless forming of the elongated forgings with sharpened end by tensile drawing with rupture | |
| Knust et al. | Preform optimization for hot forging processes using genetic algorithms | |
| CN1979496A (en) | Copper-alloy pipe-material casting-milling technology parameter designing and optimizing method | |
| CN108062427A (en) | The method that gradient rate controlling based on numerical computations reduces turbine disk forging residual stress | |
| Shao et al. | Evolutionary forging preform design optimization using strain-based criterion | |
| Lambiase | Optimization of shape rolling sequences by integrated artificial intelligent techniques | |
| CN104077439A (en) | Numerical simulation method of novel high-strength steel spoke drawing punching combined process | |
| Alimirzaloo et al. | A novel method for preform die design in forging process of an airfoil blade based on Lagrange interpolation and meta-heuristic algorithm | |
| CN103514325A (en) | Finite element numerical simulation method of spoke three-spinning-roller dip-separation powerful spinning technology | |
| Ge et al. | Optimized design of tube hydroforming loading path using multi-objective differential evolution | |
| Wu et al. | An upper bound solution of axial metal flow in cold radial forging process of rods | |
| Deng et al. | A blank optimization design method for three-roll cross rolling of complex-groove and small-hole ring | |
| Heng et al. | Towards intelligent design optimization: Progress and challenge of design optimization theories and technologies for plastic forming | |
| CN115146529A (en) | Method for predicting cold shot blasting strengthening residual stress of automotive suspension spring | |
| Qin et al. | Rivet die design and optimization for electromagnetic riveting of aluminium alloy joints | |
| Zhang et al. | Selective element fission approach for fast FEM simulation of incremental sheet forming based on dual-mesh system | |
| CN107025354A (en) | A kind of window lifting plate forming technology optimization method based on range analysis | |
| Feng et al. | Research on the optimization mechanism of loading path in hydroforming process | |
| Huang et al. | Development of energy-saving optimization for the oval-edging oval pass design using genetic algorithm | |
| Povorov et al. | Method for calculating of cross-sectional dimensions of sheet blank at intermediate stages of roller forming process | |
| CN115221622A (en) | Method for optimizing assembly, positioning, clamping and layout of large-size composite material fuselage wall panel | |
| CN106734797B (en) | A kind of variable-section variable wall thickness axis of swaging is swaged the determination method of mold finishing segment length | |
| Park et al. | Preform optimization for bevel gear of warm forging process |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |