Application of statistical technology in steel bar

2022-08-02
  • Detail

On the application of statistical technology in the production of steel drums after China's entry into world trade, in order to facilitate international trade, countries need to follow more international norms and practices in their economic and trade activities in the future. Steel drums are used as product packaging and exported with their products. Therefore, the implementation of international standards will become the focus of quality management in the steel barrel industry and a powerful competitive weapon for steel barrels to enter the international market. At present, most steel barrel industries have established quality management systems and passed ISO9000 standard system certification. As a quality management worker in a steel barrel enterprise, through years of experience in implementing ISO9000 standards in the enterprise, I found in the conversation with peers that at present, in the steel barrel industry, many enterprises still have deficiencies in the selection of statistical technology in the actual operation of the quality management system, which can not comprehensively solve the practical problems existing in the production of steel barrel enterprises. To this end, I would like to talk about some of my own views in this regard, for peer reference only

I. importance of adopting statistical technology in steel barrel enterprises

in the process of establishing the quality management system, the selection of statistical technology is very important, which directly affects the operation effect of the system and the continuous improvement mechanism. At present, the statistical techniques used in the actual production of steel drums are mostly embodied by data analysis

the quality management system established and maintained according to the ISO9000 family standards of 2000 edition has obtained a lot of data. Especially through monitoring and measurement, a large amount of data from customers and interested parties, quality management system, processes and products can be obtained. It would be a waste if these data were not used and allowed to survive. As a kind of information, data is a basic resource for enterprise management, especially for quality management. The control of steel barrel quality cannot be separated from data. Collect these data to master and analyze the production status and the factors causing problems, so as to take corrective and preventive measures to achieve the purpose of continuous improvement of product quality. Therefore, the correct application of data and data analysis is not only an important link in the quality management of steel barrels, but also a scientific basis for quality management

II. Classification of data in steel barrel production

various data will be encountered in steel barrel production. Some can be measured (such as steel barrel diameter, curling width, curling thickness, etc.); Some can be read out (for example, the steel barrel wave is accompanied by the increasingly mature striation number and ring reinforcement number of domestic aircraft R & D technologies such as C919, Yun 20, J 20, etc.); Some cannot be measured or directly counted (such as the appearance quality after spraying). Below, we classify these measurable and unmeasurable data:

1. Measured value data (measurable data)

measured value data can be taken continuously, which can usually be measured with measuring tools and instruments. Such as the height of the steel barrel, the curling width, the material thickness, the air pressure, the leak test time, etc

2. Count value data (readable data)

count value data is data that cannot be taken continuously and can only be calculated in numbers. This kind of data can be "counted" without measurement. Since flexible packaging materials are mainly high molecular polymers or their related materials, they are discrete. Such as number of nonconforming products, unit defects, number of corrugations, number of ring reinforcement, number of pores, etc

note: Although the discreteness of the count value is expressed as an integer, the integer value is not necessarily the measured value data. For example, the first pass rate of the product = = 98%, which is a non integer. However, it is not measured by instruments, and it does not have continuity. Instead, it is obtained by calculation. It is the count data of relative numbers. The nature of such data is the same as that of its molecules

III. application of control chart in steel barrel production

1. Purpose of control chart

control chart is a kind of record graph of steel barrel production process quality. It can judge whether the steel barrel production process is in a stable state, so as to find out the causes and take measures

2. Classification of control charts

the control chart of steel barrel production process can be divided into count value control chart and measure value control chart

the control chart of measurement value is divided into two types: average value range control chart (-r) and median range control chart (-r). The process control of steel barrel production generally adopts the average value range control chart (- R), which is not only convenient to operate, but also has a strong ability to detect the instability of the production process, and the control effect is ideal

3. Method of average range control chart (-r)

① data collection

generally, it is better to collect 100 data. (recent production process data under the same conditions)

② data grouping and filling in

group the collected data according to the measurement time sequence. Group k=20~25, group size (sample size) n=2~6

③ calculate the average number and range of samples in each group r

== i

n - sample size (number of data in the group)

x - the ith data value

- indicates from Xi to xn

r=xmax xmin

xmax - maximum value of data within the group

xmin - minimum data within the group

④ calculate the total average and average range

= i

= i

k - the number of groups in the data group

⑤ calculation of graph

center line cl=

upper control line ucl= +a2

lower control line lcl= - a2

note: A2 is the mean control limit coefficient related to n

⑥ calculation of R graph

center line cl=

upper control line ucl= d4

lower control line lcl= d3

note: D3 and D4 are range control limit coefficients related to n(check from the coefficient table of the measurement control chart)

⑦ work - r

draw the control limit. Draw the R diagram first, and then draw. The R chart is arranged at the top of the chart. The center line is represented by a solid line, and the control limit is represented by a dotted line

⑧ trace points

point the Xi value and RI value of each group on the graph and R graph respectively. When tracing, the abscissa of the two figures shall be aligned, and the super handicap shall be expressed with different symbols

⑨ mark the drawing name, sample size, sampling time, drafter and other reference items

4, - R chart observation and analysis

quality control of steel barrel production process, only drawing the control chart will lose the important role of the control chart. It is important to observe and analyze the control chart, find out the key of the problem, and take measures to restore the process to a controlled state. In this way, the control chart can play its effective role and achieve the purpose of prevention and control

two principles for judging process stability

① 25 consecutive points are within the control limit

there are 35 consecutive points, only one of which exceeds the control limit

there are 100 consecutive points, and only 2 points exceed the control limit

② there is no defect in the point arrangement within the control limit

judgment principle of process instability

one of the following phenomena can determine that the process is unstable

① the point exceeds the control limit (including the point on the control limit)

point ② is in the warning area. The process can be judged to be unstable under one of the following conditions

2 of the 3 consecutive points are in the warning area

three of the seven consecutive points are in the warning area

4 of the 10 consecutive points are in the warning area

③ if the point is within the control limit, but the point arrangement has one of the defects such as chain, tendency or periodicity, the judgment process is unstable

continuous chain: the phenomenon of continuous points on one side of the centerline. Chain length ≥ 7 points is abnormal; Pay attention to the development of the process in case of 5 points; In case of 6 points, the cause investigation shall be started

discontinuous chain: most points are on the side of the centerline. Continuous 11 points, with 10 points on one side of the centerline; There are 14 consecutive points, and 12 points are on the side of the centerline; 14 of the 17 consecutive points are on the side of the center line; There are 16 points on one side of the centerline for 20 consecutive points

tendency: the point rises or falls continuously

periodicity: the change of points (such as rise or fall) occurs with a certain interval

this is the refinement of the statistical technical analysis method we used in the actual production of steel barrels. In order to make the statistical technology widely used in steel barrel production, experts in the same industry are kindly invited to put forward their own opinions and carry out discussion here

Author: thus creating a new sustainable material Zhang Rong

Copyright © 2011 JIN SHI