Reliability is a very important index for servo system. Reliability prediction is one of the important contents of reliability design.
In different stages of product design and development, reliability prediction in different depths is required. Focusing on the development of professional foundation, this paper describes the first time to use WQS10.0 reliability comprehensive analysis software to model and predict the reliability of servo controller products, to model all electronic components, standard parts and printed circuit boards of the products, and to get rid of the original and rough design mode of traditional component counting method. Professional technology is developing in depth. With the development of digital servo control technology, digital servo controller is widely used in servo system.
Although digital servo controller has been developed for 12 years, it still uses traditional rough component counting method or approximation method based on previous development experience to predict reliability. The accuracy of prediction is limited, and it has limited guiding significance for subsequent reliability growth. In view of the reliability prediction mode of component counting method used in the past model development, a reliability working platform for servo controller is established.
Based on Windchill Quality Solutions 10.0 (hereinafter referred to as WQS10.
0) reliability platform, various kinds of information such as components, mechanical parts and structures are counted. Prediction and fault analysis. The software specializes in the analysis of servo controller products, promotes the promotion of electronic products and the promotion of servo controller specialty. WQS10.0 software is the leader of comprehensive reliability analysis software in the world. It is adopted by many aerospace and national defense research institutes in China. Combining with the development of existing models, a model of motherboard servo controller is selected as a pilot project, and reliability prediction and fault analysis are carried out synchronously based on WQS10.0 software.
According to the work plan, five aspects of work have been carried out, including software installation of reliability platform, product tree creation, reliability model establishment, fault platform construction and reliability prediction database creation. The platform has already been configured in the institute, but it has not been fully applied in product development. In the process of development, we make full use of the platform in our institute, complete the construction and application training and practice of WQS10.0 platform, and have the ability to carry out formal work. Different from the previous reliability prediction of pure electronic components, WQS10.0 software platform is used for reliability work. A detailed product tree is needed. The product tree is composed of components.
Each component contains structure, mechanical parts and components. Using the reliability platform, we can not only see the failure rate of the whole product, but also the failure rate of each component. Different digital servo controllers have different specific requirements, but their main functions are similar. Servo controllers products are gradually developing to platform, modularization and serialization. Layered product design of controllers is carried out from the basic layer, component layer and product layer. Therefore, the product tree is built based on the functional composition of a certain type of product, and has the ability to expand to other models in the follow-up work. Servo controller consists of five PCB: power board, code 1-0; CPU board, code 2-0; A/D, D/A conversion circuit, code 3-0; power amplifier circuit and LVDT conversion circuit (power amplifier board), code 4-0; motherboard, code 5-0.
The product structure of servo controller is shown in Figure 1.
Using Windchill reliability prediction module of WQS10.0 reliability platform software, the product tree of servo controller is established, as shown in Figure 2. WQS10.0 predictive function interface is divided into two forms, the upper form is the system tree, and the lower form is the parts table, FMEA table, predicted data, event tree graph and RBD corresponding to the system tree. When building the system tree, it is necessary to establish the relevant information base of components, determine the environmental parameters, select the prediction model, and establish the prediction parameter base. WQS10.0 software can select corresponding prediction models for system-level, component-level and part-level products respectively. The software’s predictive standard and setting of environment have “inheritance” characteristics, that is, after selecting a predictive standard at the product component level (for the top level), thermostatic element the software will default to the predictive standard and working temperature of the component if the lower unit does not select the predictive standard. If the predicted model and environment of a subsystem or component are different from that of a top-level system, we need to specify the model and ambient temperature of the subsystem or component.
The conventional and predicted parameters of more than 1000 components are set in the WQS10.0 system tree as the basic parameters of reliability prediction. Quantity and adjustment factors in conventional data are important parameters for reliability prediction, as shown in Figure 3. Predicted data include import, quality level, number of transistors and learning factors. Predicted data directly affect the predicted results, as shown in Figure 4. By analyzing the design of the servo controller, the reliability model of the non-redundant servo controller is a series model, as shown in Fig. 5. The failure rate of each part is calculated by using the [Pi Factor] button of the predicted data in the WQS10.0 product tree, and various parameters in the failure rate calculation formula of the part are given, as shown in the Pi Factor dialog box in Figure 6. More than 1000 failure rates of 84 kinds of parts including structure parts, components, standard parts and printed circuit boards have been calculated, and the basic database of reliability prediction of servo controller parts has been formed. On the basis of establishing the database of component reliability prediction, the database of component reliability prediction including component failure rate, MTBF, reliability and availability is established by using WQS10.0 software. By selecting the system from the main menu, calculating, predicting and calculating tasks, the calculation results of the components shown in Figure 7 are obtained. The reliability prediction results can be exported to Excel tables and printed. After the establishment of reliability prediction database for parts and components, the reliability of servo controller products is predicted according to task profile. WQS10.
0 reliability prediction module can comprehensively evaluate the reliability index of various task profiles of products, establish related task profiles, determine the task environment, working temperature and time ratio of different task stages. WQS10.0 gives the reliability prediction results of failure rate, MTBF, reliability and availability of task profiles, as shown in the figure. 8 and 9 are shown. The product tree is constructed by components, each of which contains structure, standard parts and components.
Using the reliability platform, we can not only look at the failure rate of the whole product, but also look at the indicators of each component. The product tree established in this way is different from the previous single calculation of the failure rate of the whole product. Servo controller specialty carries out reliability work based on software tools for the first time, establishes the reliability prediction process in the form of product tree and grasps the reliability prediction method based on WQS software tools. Through this platform, designers can be freed from the previous manual calculation process, and promote the automated reliability prediction process for information. Foundation is laid for the construction of chemical research and development process. Through detailed analysis, extensive investigation, comprehensive collection and sorting out, the database resources of general components and mechanical parts for servo controller specialty have been formed, the product tree structure and reliability prediction database have been established, and the reliability prediction database for more than 1000 components such as components, mechanical parts and printed circuit boards has been established.
Previously, it has covered the main components in the research products and provided the guarantee for the integrity of the basic data. Because the components and mechanical components are added to the reliability model, a complete set of reliability prediction schemes including mechanical parts and electronic components of the product is realized, which is compared with the traditional components. The reliability prediction method of device counting greatly improves the coverage and authenticity of reliability prediction.
The efficiency and accuracy of reliability prediction are greatly improved. According to the traditional process, in the process of reliability calculation, designers need to find out the failure parameters of each component from GJB299C standard, consult a large number of component reliability data manually and repeatedly, and calculate the failure rate by combining environmental factors and working time. Because the inquiry of component failure parameters in this standard needs to be done manually. It takes time and effort to finish. Compared with manual calculation, WQS software tool greatly improves the scientificity and accuracy of reliability prediction. It can accumulate resource databases and further improve the efficiency of reliability prediction of new products. Servo controller is an electronic product with high reliability requirement and complexity. On the basis of using WQS10 software to complete reliability prediction, the following technical work will be carried out to establish FMEA database of servo controller.
The professional reliability data of servo controller will be accumulated continuously, and the professional foundation will be consolidated so as to continuously improve the reliability of products. And the software platform is extended to all products to improve the level of professional fault prediction design.