Commonly used tools for total quality management
The so-called seven commonly used tools for total quality management are the seven commonly used methods for collecting and analyzing quality data, analyzing and identifying quality problems, and controlling and improving the level of quality in carrying out total quality management activities. These methods are not only scientific, but also practical, as the team leader should be the first to learn and master them, and lead the workers to apply to the actual production.
A checklist
Checklist is also known as a survey form, statistical analysis form. Checklist is the QC seven practices in the simplest and most used practices. But perhaps because of its simplicity and not pay attention to, so the checklist used in the process of a number of problems.
The purpose of using the checklist:
Systematically collect data, accumulate information, confirm the facts and data can be roughly organized and analyzed. That is, to confirm what is there and what is not there or whether what should be done is done (checking for omissions).
Second, the method of ranking chart
The method of ranking chart is an effective way to find out the main factors affecting product quality.
Steps to make a ranked graph:
1, collect data, that is, in a certain period of time to collect data on product quality issues. For example, you can collect 1 month or 3 months or half a year and so on in the period of scrap or nonconforming product data.
2, layered, listed in the data table, the data to be collected, according to the different issues of layered processing, each layer can also be called a project; and then statistics on the various types of problems (or each project) the number of times the recurrence of the (i.e., frequency); according to the size of the frequency of the order of the order, from large to small sequentially listed in the data table as a basic basis for calculations and graphing.
3, the calculation, that is, according to the data in column (3), accordingly calculate the percentage of each type of problem in the total problem, counted in column (4), and then calculate the cumulative percentage, counted in column (5).
4, make a ranking chart. That is, according to the data in the table above for graphing. It should be noted that the cumulative percentage should be labeled on the right side of each item, and then from the origin, point to point to connect with a straight line, so as to make the Pareto curve.
Three, cause and effect diagram method
Cause and effect diagram is also known as the characteristics of the main cause of the diagram or fishbone diagram. According to its shape, some people also call it a dendrogram or fishbone diagram. It is an effective tool for finding the causes of quality problems.
Drawing cause and effect analysis of the notes:
1, affect the quality of the product's big reason, usually from five major aspects to analyze, namely, people, machines, raw materials, processing methods and working environment. Each big reason and then materialized into a number of reasons, the reasons and then materialized into small reasons, the more detailed the better, until measures can be taken.
2, the discussion should give full play to technical democracy, brainstorming. When others speak, are not allowed to interrupt, do not carry out the argument. All kinds of opinions should be recorded.
Fourth, layered method
Layered method, also known as classification, is to analyze the impact of quality (or other problems) causes. We know that it is difficult to make sense of `if many causes of different nature are stirred together. The way to do this is to categorize the data collected for different purposes and to group together data of the same nature, collected under the same production conditions. In this way, the facts reflected in the data can be made more obvious, more prominent, easy to identify the problem, the right medicine.
Enterprise data processing is often classified according to the following principles:
1) according to different time points: such as different shifts, different dates for classification;
2) according to the operator points: such as new and old workers, men, women, different age classification;
3) according to the use of equipment: such as by different types of machine tools, different jigs and fixtures, and so on. Classification
4) according to the operating method: such as different cutting dosage, temperature, pressure and other working conditions for classification;
5) according to the raw material: such as different feed units according to different feeding time, different material composition, etc. for classification.
6) Classified by different means of detection.
7) other classifications: such as different factories, use units, use conditions, climate conditions and other classifications.
In short, because our purpose is to separate the different qualitative problems. Convenient to analyze the problem to find the cause. Therefore, there are various methods of classification, and there are no hard and fast rules.
Fifth, the histogram method
Histogram (Histogram) is the abbreviation of the frequency histogram. It is a series of rectangles of equal width and height to represent the data graph. The width of the rectangles represents the interval of the data range, and the height of the rectangles represents the number of data in a given interval.
The role of the histogram
(1) show the state of quality fluctuations;
(2) more intuitively convey information about the quality status of the process;
(3) through the study of quality fluctuations in the state of the process after the state of the process can be grasped, so as to determine where to focus efforts to improve the quality of work.
Six, control chart method
Control chart method is in the form of control charts, to determine and forecast the quality of the production process fluctuations in a commonly used statistical methods of quality control. It can directly monitor the process quality dynamics of the production process, with the stabilization of production, quality assurance, active prevention role.
1, the types of control charts:
Control charts in practice, according to the quality of data can usually be divided into two categories of seven.
Control charts for metrological data
Xbar-R chart (mean value - extreme deviation chart)
Xbar-S chart (mean value - standard deviation chart)
X-MR chart (singular value - moving extreme deviation chart)
X-R control chart (median chart)
Control charts for counting data
P chart ( Failure rate graph)
np chart (the number of failed products graph)
c chart (the number of failed graph)
u chart (the number of failed graphs per unit of product)
2, the observation of the control chart
If the points fall outside the control limits, it should be judged that an abnormal change in the process.
If the dots do not jump out of the control limits, but the arrangement of the following conditions, but also to determine the process has an abnormal change;
1) dots on the side of the center line for more than 7 consecutive occurrences;
2) more than 7 consecutive dots up or down
3) dots on the side of the center line for many times, such as 11 consecutive points, at least 10 points (can not be consecutive). dots (can be discontinuous) on the same side of the centerline
4) at least 2 dots (can be discontinuous) out of 3 consecutive dots appear beyond the 2 horizontal lines above or below (i.e., very close to the control boundaries)
5) dots show periodic shifts
In the X-R charts, the X-R charts, and the X-Rs charts, the control of the extreme deviation, R, and the moving extreme deviation, Rs, is observed. This is generally normal as long as the dots are not outside the control boundaries.
Scatterplot method
Scatterplot method, refers to the analysis of the relationship between the data of the two factors, to control the product quality of the relevant factors of an effective method.
In production practice, there are often a number of variables *** in a unity, they are interconnected, mutual constraints, and under certain conditions and mutual transformation. Some variables have a deterministic relationship between them, the relationship between them, can be expressed in a functional relationship, such as the area of the garden and its radius relationship: S = ?r2 ;Some variables but there is a correlation between them, that is, there is a relationship between these variables, but can not be accurately derived from the value of a variable to the value of another variable. These two related data are listed, punched with dots on a coordinate chart, and then the relationship between the two factors is observed. Such a graph is called a scatterplot or correlation plot.
The scatterplot method is often used in factory production, for example, the relationship between the moisture content of cotton yarn and elongation, the relationship between the room temperature and paint viscosity when spraying paint; heat treatment of steel quenching temperature and hardness of the relationship between; parts machining cutting volume and machining quality, etc., will be used in this way. The figure below is a scatter diagram reflecting the relationship between the quenching temperature and hardness of steel.
;