CN101199994A - Intelligent laser cladding metal parts - Google Patents

Abstract

An intelligent closed-loop control method that uses an expert system to detect and control the process of laser cladding forming metal parts in real time. According to the real-time detected characteristic parameters of the front and rear cladding zones of the molten pool and the target cladding zone characteristic parameters determined by the CAD model, the expert system uses the existing knowledge in the knowledge base and the calculation model in the comprehensive database to perform diagnosis and reasoning , adjust the process parameters in real time, control the cladding molding process, and update the knowledge base and comprehensive database in real time. This intelligent closed-loop control system can reduce errors caused by inaccurate calculation models, and can adjust process parameters in real time according to the characteristics of the parts to be clad, reducing or even avoiding control lag and achieving good control effects; It is necessary to install milling (grinding) equipment, which improves production efficiency, product quality and the ability to form complex parts.

Description

Intelligent laser cladding metal parts

1. Technical field

The invention relates to a rapid prototyping process in the field of metal material forming, in particular to a closed-loop control method for laser cladding forming metal parts.

2. Background technology

Laser cladding molding technology does not need molds, and can quickly manufacture full-density, high-performance metal parts with complex shapes and structures through a single process based on a computer three-dimensional model. During the forming process, the laser beam is used to heat and melt the metal powder and form a molten pool on the substrate or workpiece. According to the three-dimensional model, the metal is fused and piled up layer by layer, and finally a part is formed. This technology can greatly shorten the time from new product development to market launch, and greatly reduce product processing costs. It is especially suitable for the characteristics of fast, flexible, diversified and personalized development of modern enterprises. It is used in new automobile manufacturing, aerospace, instrumentation, The fields of manufacturing high-performance special parts for medical and health, national defense and military industry, and civil high-precision parts, especially the rapid manufacturing of functionally graded materials, superhard materials and intermetallic compound parts that are difficult to process by conventional methods, will have extremely broad applications. Application prospects.

However, the quality of parts formed by laser cladding is not stable at present, and the parts must be subjected to subsequent mechanical processing to obtain the accuracy and roughness required by users, so that the superiority of this technology is weakened, so that it cannot be promoted. application.

For this problem, two measures are mainly adopted at present, that is, the laser cladding closed-loop control system or the laser cladding forming and milling (grinding) combined system. The former has the problem of control lag and is limited by the accuracy of the calculation model, while the latter needs to add milling (grinding) equipment, which is inefficient and has limited molding capabilities for complex parts. In general, the effect of laser cladding molding parts quality control is not very obvious.

The most commonly used in the intelligent control system is the expert system. Expert system includes knowledge base, inference engine, comprehensive database, man-machine interface, interpretation program and knowledge acquisition program, which can control some complex processes very well.

3. Contents of the invention

The reason why the quality of laser cladding molded parts is unstable is that during the molding process, the process parameters are easy to fluctuate, which makes the size and shape of the cladding trace (cladding band) formed somewhere undesired change; moreover, in In the subsequent cladding, the existing defects will expand, making the convex place more convex, the concave place more concave, the thick part becomes thicker, and the thin part becomes thinner, resulting in poor precision and roughness of the part. Meet the requirements, even difficult to finally mold a complete part. However, the relationship between the process parameters of laser cladding molding and the characteristic parameters of the cladding tape is complicated, and the analytical model is difficult to establish. In the existing control system, there is a problem of hysteresis in the control of some process parameters, etc. Difficulty in quality control. This patent solves these problems and realizes rapid and short-process molding of materials, that is, one-step molding of dense metal parts.

This patent integrates intelligent technology into laser cladding molding technology, and mainly uses an expert system to conduct online closed-loop detection and control of the laser cladding molding process. According to the real-time detected characteristic parameters of the front and rear cladding zones of the molten pool and the target cladding zone characteristic parameters determined by the CAD model, the expert system uses the existing knowledge in the knowledge base and the calculation model in the comprehensive database to perform diagnosis and reasoning , adjust the process parameters in real time, control the cladding molding process, and update the knowledge base and comprehensive database in real time. Each record (knowledge) in the knowledge base of the expert system used must include: each process parameter, part material thermophysical parameter, part structural size parameter, substrate material thermophysical parameter and structural size, cladding zone characteristic parameter and its in-workpiece The parts in the laser beam should be classified according to the material of the part, the structural size of the part, the type of laser, etc. In addition to data such as material thermophysical parameters and part structure dimensions determined by CAD models, the knowledge comes from the data detected during the molding process. The system uses more than two sensors next to the laser working head, and the sensor field of view covers the molten pool. In a large area around, real-time monitoring is carried out on the characteristics of the cladding zone (parameters such as bumps and convex points on the workpiece surface, cladding zone width, etc.) at the front and rear of the molten pool (there are more than two detection modules); real-time detection at the rear of the molten pool The obtained results together with the corresponding process parameters and the cladding strip characteristic parameters of the corresponding parts before cladding the layer provide new data for the knowledge base; at the same time, the original data in the knowledge base may be updated. This is the "knowledge acquisition" of "self-study". Combined with the laser scanning direction during the forming process, the program distinguishes and selects the characteristic signals of the cladding zone representing the front and rear of the molten pool respectively, and compares the real-time detection results in front of the molten pool with the target characteristic parameters of the subsequent cladding zone at this place, Combine the current process parameters and the records in the knowledge base to conduct "diagnosis" and decide how to adjust the process parameters, that is, to decide whether to use the existing data in the knowledge base or use the model to calculate new process parameters, which is equivalent to the "reasoning machine" effect. The comprehensive database mainly stores data related to the calculation model (such as coefficients, applicable condition data, etc.), and the system can correct the control model in time according to the comparison and analysis between the expected cladding target and the actual cladding results. It is "adaptive". The man-machine interface and interpreting program here are related to the input of initial data, CAD model, and problem display.

After the implementation of this patent, compared with the existing control technology, the error caused by the inaccurate calculation model can be reduced, and the process parameters can be adjusted in real time according to the characteristics of the part to be clad, so as to reduce or even avoid the control lag and achieve good Compared with the combination of laser cladding molding and milling (grinding) cutting, the expert system can effectively control the cladding process, so there is no need to install milling (grinding) cutting equipment, which improves production efficiency and product quality and Molding capability for complex parts.

4. Description of drawings

Figure 1 is a flow chart of intelligent process detection and control.

5. Specific implementation

Combined with Figure 1, the specific implementation process is as follows:

1) Before the start of laser cladding molding, (1) install the detection and control devices, input data related to the calculation model and detection and control programs into the system, so that each module shown in Figure 1 can work normally; (2) ) The user inputs the CAD model of the part, the thermal physical parameters of the material and the environmental parameters, etc.

2) During the laser cladding molding process, the system automatically performs the following tasks: (1) According to the cladding target determined by the CAD model and the detection module 1, the characteristics of the cladding belt in front of the current molten pool (concave and convex points on the workpiece surface, melting According to the results of real-time detection, reasoning and judging whether the data in the knowledge base can be used to adjust the process parameters of the upcoming cladding work; (2) If the data in the knowledge base is available, directly press the data in the knowledge base Data adjustment process parameters, if the data in the knowledge base is not available, call the calculation model in the comprehensive database, refer to similar process conditions, calculate new process parameters, and send control instructions to the execution module; (3) detection module 2 The characteristics of the cladding zone behind the molten pool (parameters such as concave and convex points on the surface of the workpiece, the width of the cladding zone, etc.) are detected, and the results are used together with the corresponding process parameters and the characteristic parameters of the cladding zone at the corresponding part before cladding the layer as the database A new data record is stored; (4) At the same time, compare the characteristics of the cladding zone behind the molten pool detected by the detection module 2 with the cladding target determined by the CAD model, if it is within the error range, there is no need to correct the calculation model , if it is not within the error range, the calculation model must be corrected, the corrected results are stored in the comprehensive database, and the problem is displayed on the display device; (5) Enter the next cycle until a complete dense metal is formed Component.

Claims (5)

1. A closed-loop control method in the process of forming parts, which uses an intelligent expert system to control the process of laser cladding and forming metal parts in real time. The characteristic parameters and the characteristic parameters of the target cladding strip determined by the CAD model use the existing knowledge in the knowledge base and the calculation model in the comprehensive database to diagnose and reason, adjust the process parameters in real time, control the cladding molding process, and update the knowledge in real time libraries and comprehensive databases. 2. The closed-loop control method in the process of forming parts according to claim 1, characterized in that: the model calculation is combined with the retrieval of knowledge (experience data of actual molding) in the knowledge base to jointly provide the required data for process control. 3. The closed-loop control method in the molding part process according to claim 1, characterized in that: the supplement and update of data (knowledge) in the knowledge base is mainly carried out in the molding process, and the cladding strip characteristic parameters detected in real time, The corresponding process parameters are the data sources of the knowledge base. 4. The closed-loop control method in the forming part process according to claim 1, characterized in that: the data such as coefficients and application conditions relevant to the calculation model are stored in the comprehensive database, the same as the knowledge in the knowledge base, and the comprehensive database The data is also updated and supplemented in real time. 5. The closed-loop control method in the process of forming parts according to claim 1, characterized in that: during the forming process, the characteristic information of the cladding belt before and after the molten pool is detected at the same time, and after data conversion and analysis, the systems are respectively controlled in time Pre-state and control effects.

Images