In recent years, more and more companies have begun to focus on improving the reliability of their products.Reliability has become one of the important indicators for measuring product quality.The reliability refers to the probability that the product will perform the specified function within the specified time under the specified conditions of use.One problem encountered by many enterprises in implementing reliability management is that when the reliability management of products develops to a certain stage, statistical modeling methods are bound to be used, and since the reliability statistical modeling method is relatively more complicated, many are called six standards. The consultants of the poor black belt masters dare not touch it easily, which invisibly casts a mysterious veil on the reliability analysis.
When studying the quality elements of products, Zhu Lan, a quality management guru, summed up various quality characteristics and divided them into three categories: first, technical characteristics such as strength, hardness, voltage, etc. Second, psychological characteristics, such as Perception, vision, taste, etc.; third, time-related characteristics such as reliability, maintainability, etc.This shows that reliability is an indispensable part of quality control.
Many well-known auto manufacturers have been working on reliability for many years, and GM is one of the first companies to introduce reliability concepts for product design and production.After a long period of reliability work, GM's reliability level has increased significantly.In recent years, the North American Automotive Vehicle Reliability (VDS) survey report released by JDPower, the authoritative automotive research organization in the United States, shows that GM's Buick brand has been in the top three for a long time and its performance is very stable.
In addition to the automotive industry, reliability engineering is also widely used in other manufacturing industries.Keep an eye on the valuable experience of companies such as Intel and Dow Chemical in integrating reliability tools into the company's original R&D, production and customer site use value chain. It is not difficult to find that one of the key factors for success is to make the reliability not It has become a special favorite of a few R&D elites. Instead, it visualizes and simplifies reliability analysis, making it an ordinary tool that many technicians can use, and can be used to obtain decision-making improvements.
In all fairness, it is quite difficult to achieve this more than a decade ago. The feeling of ordinary companies is that reliability is “unexpected”.However, the rapid development and maturity of modern computer technology in recent years has enabled enterprises to meet the preconditions of traditional technical obstacles. Many reliable tool carriers represented by professional quality control statistical software JMP have created an interactive and visual advanced analysis. The mode has the advantages of simple operation and high efficiency of analysis, which makes the reliability go down the altar and move toward “popularization”.
The following sections will focus on several major application directions for reliability analysis and the help of JMP interactive visualization reliability analysis.
l Life data distribution
In reliability engineering, if the product life data can be fitted to a certain probability distribution, the characteristics of this distribution can be used to make product design and manufacturing evaluation decisions.However, the difficulty is that the reliability data has a limit characteristic, and its probability distribution is much more complicated and more difficult to understand than the distribution of general data.With the help of JMP software, engineers can easily identify the distribution of the most matching life data in a few seconds by simply clicking the mouse to find the distribution type of the most matching life data in a few seconds. Predict the product's service life, failure probability, failure probability density and failure rate.
壽命資料分佈壽命資料分佈壽命資料分佈Life data distribution
l Reliability prediction
After understanding the distribution of life data, we can further make reliability predictions.JMP can help us predict the number of failures in the future.Future failures can be estimated using interactive graphs that adjust factors such as future production and contract duration.
可靠度預測可靠度預測可靠度預測Reliability prediction
l accelerated life test
If you want to further quantify the impact that different load conditions may have on your product, you need to use a statistical method that is much more complicated than traditional one-way regression analysis.JMP software simplifies this analysis process, transforming complex statistical principles into a series of graphic combinations that vividly show the impact of accelerated life factors (such as temperature in this example) on product life at different levels. .
加速壽命試驗加速壽命試驗加速壽命試驗Accelerated life test
l Parameter survival model
In many cases, it is not only the temperature and one load condition that affects the life of the product, but also the humidity, vibration, radiation, pressure, chemicals, dust, voltage, current and human factors.How can we comprehensively and intuitively express the impact of all factors on product life?In JMP, the solution to this problem is to accurately represent the relationship between various influencing factors (such as coating, Celsius temperature, etc. in this example) and product reliability in an easy-to-understand "model descriptor" form. Synchronously predict failure probability and product life under different conditions.
參數生存模型參數生存模型參數生存模型Parameter survival model
l Product aging analysis
If we want to have a deeper understanding of the quality characteristics of the product, we can not only use "failure", "no failure" to indicate the quality status of the product, but also use a continuous variable to describe the changes in the characteristics of the product.As long as relevant data is recorded, engineers can use JMP to perform more accurate model fitting on the aging process of product characteristics (such as Current in this example) in linear or nonlinear form, and predict product failure based on this. The time of occurrence, as well as the law of product characteristics over time.
產品老化分析產品老化分析產品老化分析Product aging analysis
l Reliability experiment design
The reliability analysis data is mainly derived from the reliability test. The reliability test is generally long and destructive to the product.In order to reduce costs, we need to use the fewest experimental samples and the shortest experimental time to get the most information.JMP is a software that enjoys a reputation in the field of experimental design. Engineers can flexibly and conveniently calculate the number and length of samples required to carry out reliability testing and reliability verification experiments, and can also set a single factor to consider the interaction between factors. (such as the torque in this example, etc.) or the combination of experimental conditions in the two-factor accelerated life test, and the number of samples under each experimental condition.
可靠度實驗設計可靠度實驗設計可靠度實驗設計Reliability experiment design
