Weather and Mood
Today, the weather is sunny. The sunlight pours down unreservedly, covering everything in the world with a golden glow. In the deep - blue sky, several white clouds are floating leisurely, as if they were a wonderful picture casually sketched by nature. Such wonderful weather really makes people feel happy. Sunshine symbolizes light and hope, dispelling the gloom and dullness and brightening people's hearts. When one is in a good mood, everything around seems extraordinarily beautiful. It's as if even the flowers and plants by the roadside are dancing merrily. This kind of pleasant mood acts like a catalyst, giving people more energy and enthusiasm to immerse themselves in learning about the knowledge of the Seven QC Tools.
An overview of the classification of the seven QC tools
The seven QC tools are important tools in the field of quality management and are mainly divided into three categories.
Simple seven-step handwashing method
It includes Gantt charts, flowcharts, the 5W2H method, the foolproof method, the radar method, statistical charts, and trend charts. These techniques are simple to understand and easy to operate, and can play an important role in daily work management and problem - solving. For example, the Gantt chart is an intuitive and effective self - management tool for progress; statistical charts can make abnormal data clearly visible at a glance, facilitating comparison and drawing conclusions.
The old seven QC toolsQCQC The old seven QC tools
There are cause-and-effect analysis diagrams, Pareto charts, check sheets, stratification methods, scatter diagrams, histograms, and control charts. The old seven management tools are widely used in data collection, analysis, and problem-solving at the production site. They can help enterprises identify the key factors affecting product quality and thus take targeted improvement measures.
The New Seven Tools of QCQCQuality Control
It covers the relationship diagram, systematic diagram method, KJ method, arrow diagram method, matrix diagram method, PAPC method, and matrix data analysis method. The New Seven Management and Planning Tools place more emphasis on the thinking process and management methods. They are suitable for solving complex quality management problems and can assist enterprises in grasping problems holistically and formulating systematic solutions.
Count data and measurement data- count data - 、 measurement data
In the data statistics of quality management, there is a distinction between count data and measurement data. Count data refers to the data obtained by counting, such as the number of qualified products or the number of defects. Simply put, it is the data obtained by "counting". For example, in a batch of products, the number of qualified products or the number of defects on the products is counted.Measurement data, on the other hand, is the data obtained through measurement, such as weight, time, content, and length. That is to say, it is the data obtained by "measuring". Examples include the length, weight, and concentration of products. For data with decimal points, the rounding - off method is usually used for processing.
Composition of the Seven QC ToolsQCSeven QC Tools
The seven QC tools consist of five charts, one table, and one method.
Five pictures
They are Pareto chart, scatter diagram, histogram, control chart, and cause-and-effect diagram (fishbone diagram) respectively. The Pareto chart can help us identify the key factors affecting quality; the scatter diagram is used to analyze the relationship between two variables; the histogram can visually display the distribution of data; the control chart is used to monitor whether the production process is stable; and the fishbone diagram can clearly present the causes of problems.
- a form a table - a person of striking appearance a handsome young man
Checklist (Gantt Chart). The checklist is mainly used for daily management, data collection, and management improvement. It can help us complete the necessary data collection work in the shortest possible time.
One method / One approach
Stratification method. The stratification method stratifies the data to identify the factors causing data differences, so as to take targeted measures and solve problems in quality management.
Introduction to the Seven Simple Techniques
Gantt chart
Gantt charts are widely used. They can be used for work schedule arrangement, enabling us to clearly know the start time, end time and duration of each task. They can also be used to check work progress and promptly identify whether the work is proceeding as planned. Moreover, they help us keep track of the current situation and understand the progress of the ongoing work. Additionally, they can be applied to daily plan management.It is the simplest and most effective self - management tool for work progress. Through a simple chart format, we can have a clear overview of the work progress at a glance.
Statistical chart (bar chart)
Statistical charts can make abnormal data clearly visible at a glance. Through the comparison of the lengths of bars, we can quickly identify abnormal situations in the data. It is easy to compare the differences between different data and draw conclusions. It is the most commonly used type of chart, which can be seen everywhere in newspapers and magazines. Moreover, the stratification method is also applied in the statistical process.
Run chart (Trend chart)
The trend chart is mainly used for managing data changes over time. Through it, we can understand the current situation, identify problem areas, and compare the data effects and differences in different time periods. It is the simplest way to understand data differences and has a wide range of applications. For example, a trend chart of the defective product rate can intuitively reflect the changing trend of the defective product rate over time.
Flow chart; Flow diagramA flow chart will be helpful in making a program. flow chart flow diagram 、
Flowcharts are used to represent work content, enabling us to clearly understand each link of the work. They make it easy to grasp workstations and clarify the position and function of each work step. Flowcharts are also commonly used in education and explanations, serving as a simple way to illustrate work descriptions and content.
Circle diagramCircular chart Round graph diagram chart graph
A pie chart is used to compare the composition ratios of various parts. It arranges the data from largest to smallest in a clockwise direction and divides the circle into several sectors, directly depicting the proportion of each item. The stratification method is also used in the analysis process.
Introduction to the Old Seven Tools
Checklist (CHECK LIST)ChecklistCHECK LIST
Checklists can be used for daily management, helping us to supervise and inspect work in an orderly manner. They are also used for data collection to ensure the integrity and accuracy of data. Moreover, they can be used to improve management. By analyzing the collected data, we can identify the problems in management and make improvements. A checklist can help everyone complete the necessary data collection in the shortest possible time.
Stratification method
The main purpose of the stratification method is to use it to identify the factors causing data differences and then take appropriate measures. Usually, stratification is carried out according to the 4M factors (man, machine, material, and method). The stratification method itself doesn't have a specific graph. Generally, it borrows other graphs to present data, and the data are arranged in descending order.
Plato (Count value statistics)Pareto chart Pareto chart (Count value statistics)
The Pareto chart is a commonly used tool in the data management of the production site. The data collected at the production site must be effectively analyzed and utilized to reflect its value. The Pareto chart classifies and organizes this data and presents it in a chart, enabling us to fully understand the problem points and important causes. Its definition is as follows: Based on the collected data, it is sorted and classified according to different criteria such as defective causes, defective conditions, defective items, and the locations after the occurrence of defects, to find the cause, condition, or location with the largest proportion. The data is then arranged in descending order, and a graph with cumulative values is created. From the Pareto chart, we can see which item has a problem and the degree of its influence, so as to identify the problem and take improvement measures for the problem points. Therefore, it is also called the ABC chart, which mainly analyzes the control of the first 2 - 3 important items. Since the chart is arranged in descending order, it can also be called a ranking chart.The process of creating a Pareto chart is relatively complex, including steps such as determining the classification items of the data (which can be classified from two aspects: results and causes), deciding the period for data collection and collecting the data, organizing the data according to the classification items and creating a statistical table, entering the data on the graph paper and drawing a bar chart, plotting the cumulative curve and cumulative ratio, and recording necessary information. There are also some points to note when drawing the chart. For example, the items on the horizontal axis should be arranged in descending order, and the "other" item should be placed at the end; the width of the bar chart should be consistent, and the ratio of the vertical axis to the horizontal axis should be 3:2.
Control chart
What is a control chart
In order to ensure that the on - site quality situation meets the requirements of "management" operations, we usually detect the quality characteristics of products to judge whether the "management" operations are normal. Since the quality characteristics will fluctuate over time and under various conditions, it is necessary to set a reasonable upper and lower limit to detect whether the on - site production process is in the "management" state. This is the basic origin of the control chart.The control chart was invented by Dr. Walter A. Shewhart, an American quality control master, in 1924. Its main concept is to compare the actual quality characteristics of products with the control limits of the process capability judged based on past experience, and present them graphically in chronological order.
Basic characteristics
Generally, the vertical axis of a control chart is set for the quality characteristics of products, with the graduation based on the data of process changes. The horizontal axis represents the group codes, numbers, or dates (such as year, month, and day) of the inspected products. The points are plotted on the chart in sequence according to the time or the manufacturing order.There are three straight horizontal lines on the chart. The middle one is the Center Line (CL), usually drawn as a solid blue line. The one at the upper - left is the Upper Control Limit (UCL), and the one at the bottom is the Lower Control Limit (LCL). The upper and lower control limits are generally represented by red dashed lines, indicating the acceptable range of variation. The lines connecting the points representing the actual quality characteristics of products are mostly drawn as solid black lines.
Principle of control chart
During the manufacturing process, regardless of how precise the equipment and environment are, there will always be variations in the quality characteristics of products. The causes of these variations can be divided into chance (random) causes and assignable (non - random) causes.The control chart is based on the theory of three standard deviations in the normal distribution. The center line represents the mean value, and the upper and lower control limits are determined by adding and subtracting three standard deviations (±3σ) from the mean. This method is used to determine whether there are any problems in the manufacturing process, and it was invented by Dr. Walter A. Shewhart.
Types of control charts
According to the nature of data, control charts can be divided into variable control charts and attribute control charts. The data of variable control charts are obtained through actual measurement with measuring tools, such as continuous characteristics like length, weight, and concentration. Commonly used ones include the mean - range control chart and the mean - standard deviation control chart.The data of attribute control charts are obtained by counting units, such as discontinuous data like the number of non - conformities and the number of defects. Commonly used ones include the proportion of non - conformities control chart and the number of non - conformities control chart.Variable control charts are sensitive, making it easy to investigate the root cause. They can promptly reflect non - conformities and stabilize product quality. However, they require a relatively high sampling frequency, are time - consuming and cumbersome, and need trained personnel to measure and calculate the data.The data required for attribute control charts can be obtained by simple methods, which makes it more convenient to understand the overall quality situation. But they cannot find the root cause of non - conformities, lack timeliness, and are prone to delaying opportunities.
Drawing of control charts
Taking the commonly used measurement value control chart (X - R) as an example, the drawing steps are as follows:1. First, collect more than 100 data and arrange them in the order of measurement.2. Group the data into about 20 - 25 groups, with 2 - 5 data in each group (usually 4 - 5 data are used).3. Record the data of each group in the corresponding columns of the data table.4. Calculate the average value X and the range R of each group.5. Calculate the overall average X and the average range R.6. Calculate the control limits.7. Draw the center line and the control limits, and plot each data point on the chart.8. Record the history of each data and special reasons for future reference, analysis, and judgment.
Key points for plotting control points
Any change information such as the names of various projects, control characteristics, measurement units, equipment types, operators (measurers), sample sizes, material types, and environmental changes should be clearly filled in for the analysis and organization of the data.
Relevant settings of the control chart for metrological value change
Method for determining the width of control limits
In the control charts for variable data (such as X - R, X - s, etc.), the determination of the control limit widths of the X - control chart and the R - control chart follows specific principles. Usually, the sample size (n) of each group is used as a reference. Specifically, the unit scale of the X - control chart is approximately 1/n times that of the R - control chart.In practical operation, the width of the vertical control limits is generally controlled within 20 - 30 mm. This width can clearly show the vertical distribution range of the data. The interval between each group on the horizontal axis is about 2 - 5 mm. A reasonable interval helps to distinguish the data of different groups, facilitating subsequent observation and analysis.
Line drawing and marking
When drawing a control chart, the center line (CL) should be recorded with a solid line. This is because the center line represents the average level of the data and serves as a stable reference benchmark. Using a solid line can highlight its importance and stability. The control limits, on the other hand, are recorded with dashed lines. The dashed lines can form a sharp contrast with the center line and give an implication of a boundary. Meanwhile, symbols such as CL, UCL (Upper Control Limit), and LCL (Lower Control Limit) need to be marked on each line according to the line type. This enables users to quickly identify the meaning represented by each line, facilitating the interpretation and analysis of data.
Numerical digit calculation
There are clear requirements for calculating the number of digits of CL, UCL, and LCL. It is sufficient to have two more digits than the measured values. This is to ensure a certain level of accuracy while avoiding the difficulties in subsequent calculations and analyses caused by overly complex numerical values. For the average calculation of each group of data, one more digit than the measured values should be taken. In this way, it can reflect the average situation of the data without increasing the computational burden due to an excessive number of digits.
Rules for point plotting
There are various forms for plotting points, such as [·], [○], [△], [×], etc. To ensure the standardization and consistency of control charts, it is best to uniformly specify within the factory which form of point plotting should be used. In this way, when different personnel observe and analyze the control charts, misunderstandings caused by different point - plotting forms can be avoided, which improves the universality and comprehensibility of the control charts.
Control chart drawing interval
For variable control charts, the minimum distance between the two control charts when plotting should be more than 20 mm. If conditions permit, it is best to keep an interval of about 30 mm. An appropriate interval can prevent the two control charts from interfering with each other, make the data presentation clearer, and facilitate independent observation and comparative analysis of the data on the two control charts.
Interpretation of control charts
Judgment of the control state
When the manufacturing process is in a stable state, the control chart will exhibit specific characteristics.
Most of the data points are concentrated near the center line: This indicates that the fluctuations in the data revolve around the average level. Most of the data do not show significant deviations, which means that the manufacturing process is relatively stable and is less affected by random factors.
A few data points fall near the control limits: Since the manufacturing process will inevitably be affected by some random factors, there will be a few data points close to the control limits. However, this does not mean that the process is abnormal. As long as the data points do not exceed the control limits, they all fall within the normal fluctuation range.
The distribution and fluctuation of data points are in a random state without any discernible rules: A random distribution indicates that the manufacturing process is not affected by systematic factors. The changes in each data point are independent and conform to the normal probability distribution law.
No data points go beyond the control limits: The control limits are determined based on the statistical characteristics of the process. If no data points exceed the limits, it indicates that the fluctuations of the process are within the acceptable range and the process is in a stable state.
Can the control limits be extended to serve as the judgment criteria for subsequent process control
There are the following conditions for determining whether the control limits can be extended for subsequent process control:
When there are more than 25 consecutive points within the control limits (with a probability of 93.46%): When 25 consecutive points fall within the control limits, it indicates that the process has a relatively high level of stability during this period. There is a high probability that the subsequent process can also remain stable. At this time, extending the control limits can be considered.
When among 35 consecutive points, no more than 1 point falls outside the control limits: In the observation of 35 consecutive points, if no more than 1 point exceeds the control limits, it indicates that although the process may have occasional minor fluctuations, the overall state remains relatively stable. This can also serve as a reference for extending the control limits.
When no more than 2 points fall outside the control limits among 100 consecutive points: When the number of observed points increases to 100, the judgment on the process stability will be more accurate. If no more than 2 points exceed the control limits, it indicates that the process has good stability. The process can be considered to be in a controlled state, and to some extent, extending the control limits can be considered.
It should be noted that even if the manufacturing process meets the above - mentioned conditions and can be considered to be in a controlled state without changing the control limits, once there are data points exceeding the control limits, there must be abnormal causes for these out - of - limit points. The causes must be investigated and eliminated to ensure the quality and stability of the manufacturing process.
Principles for verification and interpretation
Each data point should be regarded as a distribution rather than a simple point. Each data point does not merely represent a specific data value but a state distribution of the process at that moment. Through the analysis of each data point, we can understand the fluctuation of the process at different times and the potential changing trends.
The movement of data points represents the changes in the manufacturing process. Even without an abnormal cause, there will still be differences among the data points within the control limits. The direction and trend of the data points' movement reflect the dynamic changes in the manufacturing process. Even when there is no obvious abnormal cause, due to various random factors in the manufacturing process, there will be certain differences among the data points within the control limits, which is a normal fluctuation phenomenon. By observing the movement of the data points, it is possible to detect potential changes in the manufacturing process in a timely manner and take measures for adjustment in advance.
General principles for detecting abnormalities: Although the original text mentions "as shown in the figure", here we need to clarify that in practical applications, according to some common statistical laws and experiences, when there are situations such as consecutive rising or falling points, or multiple points concentrated on one side, it may indicate that there are abnormalities in the process, which requires further in - depth analysis.
Precautions for using control charts
Operation standardizationoperation workoperationHomework standardization
Before using control charts, on - site operations must complete standardized operations. Standardized operations can ensure the consistency and stability of the production process, reduce data fluctuations caused by differences in human factors and operation methods, and enable control charts to accurately reflect the real situation of the production process.
Determine the control items
Before using the control chart, it is necessary to determine the control items first, including the selection of quality characteristics and the determination of the sampling quantity. The selection of quality characteristics should be related to the key quality indicators of the product and be able to directly reflect the quality status of the product. The determination of the sampling quantity should be carried out according to the characteristics of the process and actual needs. It should not only ensure that the changes in the process can be accurately reflected, but also avoid increasing costs and workload due to excessive sampling.
Avoid using specification values to replace control limits
The control limits are calculated based on the statistical data of the production process, which reflect the actual fluctuation range of the process. The specification values, on the other hand, are the design requirements of the product. The meanings and uses of the two are different. Under no circumstances should the specification values be used to replace the control limits. Otherwise, it will lead to misjudgment of the process state and the inability to detect abnormal situations in the process in a timely manner.
Selection of control chart types
The selection of control chart types should be coordinated with the determination of control items. Different control items may require different types of control charts for monitoring. For example, the X - R control chart can be selected for measurement data, while the P control chart, C control chart, etc. can be chosen for count data. Selecting an appropriate control chart can more effectively monitor the quality status of the process.
Sampling method
The sampling method should be based on the principle of obtaining a reasonable sample group. A reasonable sample group should be representative and accurately reflect the overall situation of the process. Methods such as random sampling and stratified sampling can be adopted to ensure that the samples drawn cover various states of the process and avoid incorrect judgments about the process due to sampling bias.
Exception handling
When ideas go beyond the boundaries or show abnormal states, various measures must be used for research and improvement, or statistical methods should be applied to identify and eliminate the causes of the abnormalities. Tools such as fishbone diagrams and Pareto charts can be used to analyze the causes, find out the root causes of the abnormalities, and take corresponding measures for improvement to ensure the stability of the manufacturing process and product quality.
Selection of group size
In the X - R control chart, the most suitable group size (n) is generally in the range of n = 4 - 5. A group size within this range can achieve a good balance between ensuring the representativeness of the data and the accuracy of statistical analysis. If the group is too small, it may not accurately reflect the process fluctuations. If the group is too large, it will increase the difficulty and cost of data processing.
Problem of the lower limit of the R control chart
The R control chart has no lower limit. This is because the R value is obtained by subtracting the minimum value from the maximum value in the same group of data. Since the minimum value cannot be greater than the maximum value, the R value must be non - negative. If a negative lower limit is set for the R control chart, it has no practical meaning. Therefore, the lower limit is usually not set.
Process capability requirements
To make the control chart effective, the Cp value (process precision) in the product process capability should be greater than 1. The Cp value reflects the matching degree between the actual processing ability of the process and the product specification requirements. When the Cp value is greater than 1, it indicates that the process has sufficient ability to meet the product specification requirements, and the control chart can effectively monitor the quality status of the process. If the Cp value is less than 1, it means that the process capability is insufficient. Even if the control chart shows that the process is in a stable state, the product may still have a relatively high non - conformity rate. In this case, the process needs to be improved to enhance the process capability.