Quality control tools

Flow Chart, Check sheet, Histogram, Pareto Analysis, Scatter Diagram, Control Chart, and Cause and Effect Diagram are the basic seven control tools considered for achieving quality. (See Figure 6.2 for Quality control tools)

Figure 6.2: Quality control tools

a) Flow Chart: Flow chart is a visual representation of process showing the various steps. It helps in locating the points at which a problem exists or an improvement is possible. Detailed data can be collected, analysed, and methods for correction can be developed using flow charts. The various steps include:

 Listing out the various steps or activities in a particular job

 Classifying them as a procedure or a decision

Each decision point generates alternatives. Criteria and consequences that go with decisions are amenable to evaluation for purposes of assessing quality. The flow chart helps in pin-pointing the exact points at which errors have crept in. (See Figure 6.3 for Sample flow chart)

Figure 6.3: Sample flow chart

b) Check Sheet: Check sheets are used to record the number of defects, types of defects, locations at which they are occurring, times at which they are occurring, and workmen by whom they are occurring. The sheet keeps a record of the frequencies of occurrence with reference to possible defect causing parameters. It helps to implement a corrective procedure at the point where the frequencies are more. (See Table 61. for Sample check sheet)

Table 6.1” Sample check sheet No. of

Defects Day

1 2 3 4 5

1 / / ^{/// /// ///// //}

2 / //// /// //// ///

3 // ////// //// // //

4 // //// /// // //

5 /// ////// /// / ///

6 // //// /// /// //

The table shows that the number of defects 1 and 5 are not many as compared to defect no 2 which increased over the days and appears to be stabilising at the higher side and therefore needs to be attended

Reject Reject

Open Inspect Inspect

Open Pack

immediately. The column which shows days can be changed to observed by the hour, if need be.

c) Histogram – Histograms are graphical representations of distribution of data (See Figure 6.4 for Sample histogram chart). They are generally used to record huge volumes of data about a process. They reveal whether the pattern of distribution has a single peak, or many peaks and also the extent of variation around the peak value. This helps in identifying whether the problem is serious. When used in conjunction with comparable parameters, the visual patterns help us to identify the problem which should be attended to.

14 parameter. Sometimes, the percentages are shown to demonstrate the relative contribution of each of the parameters.

d) Pareto Analysis: Pareto analysis is a tool for classifying problem areas according to the degree of importance and attending to the most important. Pareto principle, also called 80-20 rule, states that 80 percent of the problems that we encounter arise out of 20 percent of items. If we find that, in a day, we have 184 assemblies having problems and there are 11 possible causes, it is observed that 80 percent of them, that is, 147 of them have been caused by just 2 or 3 of them. It will be easy to focus on these 2 or 3 and reduce the number of defects to a great

extent. When the cause of these defects has been attended, we will observe that some other defect becomes predominantly observed and if the process is continued, we are marching toward zero defects.

e) Scatter Diagram: Scatter diagram is used when we have two variables and want to know the degree of relationship between them (See Figure 6.5 for Sample scatter diagram). We can determine if there is cause and effect relationship between the variables and the degree of extent over a range of values of the variables. Sometimes, we can observe that there is no relationship, in which we can change one parameter being sure that it has no effect on the other parameter.

VARIABLE 2

Figure 6.5: Sample scatter diagram

We can see that the change in variable 2 does not have much effect on variable 1. The other interpretation can be that for a small change in variable 1, the effect on variable 2 is more.

f) Control Charts: Control charts are used to verify whether a process is under control. Variables, when they remain within a range, will render the product and maintain the specifications. This is called the quality of conformance. The range of permitted deviations is determined by design parameters. Samples are taken and the mean and range of the variable of each sample (subgroup) is recorded. The mean of the means of the samples gives the control lines. Assuming normal distribution, we expect 99.97 percent of all values to lie within the Upper Control Limit (UCL) and Lower Control Limit (LCL) – corresponding to + 3. The graphical

representation of data helps in changing settings to bring back the process closer to the target.

Example 1

A shaft is to be made with a diameter of 25mm. The area required to be ground is between +0.01 and -0.02mm by a process of centre less grinding. A sample of 5 numbers is taken every hour and the observations are recorded as under. (See Table 6.2 for Sample check sheet)

Table 6.2: Sample check sheet Samples

Time 1 2 3 4 5

9 AM 24.98 24.99 25.00 25.04 25.01 10 AM 25.01 25.02 25.00 25.01 25.00 11 AM 24.99 24.98 25.02 25.02 24.97 12 Noon 24.97 24.99 25.01 25.04 25.03 2 PM 25.01 25.02 25.00 25.03 25.01 3 PM 24.99 24.98 25.02 24.97 25.00 4 PM 24.97 24.99 25.01 25.04 25.03 5 PM 25.01 25.02 25.00 25.03 25.01 The method to be followed is as under.

1. Draw a line diagram taking the means of every hour

2. Draw the R chart and X charts and determine whether the process is under control.

The procedure to find the range for the sample readings of each hour is:

1. Find the mean of the readings of each hour, that is, x

2. Add all the means calculated above and take the mean of the means.

Then you will get the mean for all samples. (See Table 6.3 for Mean of all samples)

(Cont. in next page)

g) Cause and Effect Diagram: Cause and effect diagram represents all the possible causes which lead to a defect on quality characteristics.

These are arranged in such a way that different branches representing (Cont. from previous page) Mean for all samples 25.00

On a graph sheet y-axis represents the dimension. The mean value is drawn as a horizontal line near the middle of the y-axis while the horizontal axis represents the serial number of the samples. Variations of the dimensions get marked on both sides of the mean line.

n

causes connect the stem in the direction of the discovery of the problem (See Figure 6.6 for Sample cause and effect diagram). When each of them is investigated thoroughly we will be able to pin-point some factors which cause the problem. We will also observe that a few of them can have cumulative effect or even a cascading effect.

Figure 6.6: Sample cause and effect diagram

When we observe that we have excessive defects from a machine, we try to identify all possible sources of the causes of defects. We make a study of each of them and try to correct it.