_{1}

^{*}

In this article, we studied the bearings made by one company in Shanghai. Through statistical process controlling the quality characteristic of bearings’ diameters and multi-vary analysis is applied to find the key variation factors which have an influence on the quality characteristic of the bearings, the quality level of the bearings of this company is improved.

The manufacturing in China is developing fast through 30 years reform and opening up. China has firmly established itself as a manufacturing power in the world. At the same time, we should see that Chinese manufacturing is large but not strong. The improvement of product quality lags behind the growth of economic scale. The whole level of product quality has a gap with the developed countries. With the bearing for instance, China is a power of bearings, but the most is belonging to medium or lower end product. It is the lag of R&D capability, equipment and handcrafts that result in the precision of bearing unsure. If the precision of bearing can not meet the quailfication, the steel plate is unqualified. Therefore, it is essential subject to study through technology improvement which can advance bearing quality and increase enterprise benefit.

SPC (statistical process control) is a tool of process control by means of mathematical statistics method. It analyses and controls the process by using the statistical law nature of figure fluctuation. Control chart becomes the one of most important tool of management after birth of the first control chart.

Zhen He, Ershi Qi, Shenghu Zhang [

As to how to control the problem of multiple variation sources, product quality problems caused by the variation source can be known if we can make sure of sources of variation and proportions in all variation factors. Then we can take relevant measures according to the relative size of variation source and proportion.

In this article, the CHTD5/7 model bearings made by one company in Shanghai were taken for research objects. The quality remand of the diameter is considered as the key factor because it has a directly effect on the final assembly of pumps.

CHTD5/7-type self-bearing diameter quality requirements: Φ (60 ± 0.02) mm. When the products are processing, five samples per half hour are taken. And the data of its inner diameters are shown in the

a. Unit: mm So we can get the control Chart (see

From the

From ^{2}.

Following the above equation, we can obtain: C_{pk} ≈ 0.77. It shows that the process capability is deficient because 0.77 less than 1.

In multi-vary analysis (MVA), the variation sources of process quality characteristics are divided into time to time variation, piece to piece variation and within piece variation. After on-site analysis, the main factor affecting the diameter of bearings is the taper and the non-concentricity. The different tapers of two ends of bearings

can not make the bearings keep in parallel. It maybe has an influence on the contact area when bearings are used. And the different non-concentricity of bearings can make the circle centers of bearing two ends unsymmetrical and cause these bearings can not be assembly.

First, the systematic analysis chart of quality variation is drawn (see

The sample data at 8:00, 9:00, 10:00, 11:00, 12:00 are collected to analyze the quality characteristics of diameters of bearings.

• This can not only ensure the continuity of time, but also collect sufficient data.

• Considering that the bearing itself has a certain errors and there are some errors existing in measure, we take three bearing samples in every time span.

• We twirl each bearing to read the data of the maximum and the minimum of the left and the right. Then we can get four data of every bearing.

There are data of monitoring as follows. (Tables 2-7)

The multi-vary data analysis at 8:00 is as follows:

1) Within piece variation:

Different tapers variation = |the value of the left average diameter of samples – the value of the right average diameter of samples| = 60.014 − 60.010 = 0.004.

Different non-concentricity variation = the maximum of average diameter of samples – the minimum of average diameter of samples = 60.012 − 60.011 = 0.001.

8:00 the average diameter value of three samples = 60.012.

2) Piece to piece variation:

Sample 1 − 2 = |the average of sample 1 – the average of sample 2| = 60.013 − 60.012 = 0.001.

Sample 2 − 3 = |the average of sample 2 – the average of sample 3| = 60.013 − 60.011 = 0.002.

3) Time to time variation:

By parity of reasoning, five time span variation values are obtained. (

By analogy, we can obtain five time variation values.

Through field analysis on the above sample data, we can offer a proposal on improving process quality of bearings. (

We take the inner diameter data of bearings of one process through the above adjustment. (

From ^{2}.

By analogy, we can obtain five time variation values.

a. Unit: mm.

According to the Equation (1), we can obtain C_{pk} = 1.11. For 1 ≤ C_{pk} ≤ 1.33, the process capability is normal. The goal has now been finally attained through the multivary analysis of bearings.

In brief, the process capability is improved through the multi-vary analysis to the CHTD5/7 model bearings made by one company in Shanghai, so the quality of bearings is made better.