MECE5208 Reliability Engineering Dr.

Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)

Chapter 7- 7 QC Tools

7.1 Introduction
Ishikawa, creator of the "fishbone diagram" for root cause analysis and colleague of Deming,
assembled a set of seven tools for quality control. Ishikawa’s seven basic tools of quality: cause-
and-effect diagrams, flowcharting, check sheets, Pareto diagrams, control charts, histograms and
scatter diagrams. Ishikawa’s seven basic tools are also referred to as the 7QC tools. Kaoru
Ishikawa once said “As much as 95% of quality problems can be solved with seven
fundamental tools”.

1. Flowcharts /stratification

A flow chart is a visual tool that depicts the flow or sequence of a process (flow of
information, tasks, people, material or decision).
 It makes the complex become simple
 It Promotes a common understanding
 It show the logical steps in a process and how various elements within a process, system
(organization) are related
 They can be used to determine and analyze potential problems in quality planning
and quality control
 By documenting these logical steps, the team can identify where quality problems
might occur and then develop approaches to proactively manage them.
 Flowcharting also helps create a process that is repeatable.

Figure 7.1
If you can’t describe what you’re doing as a process, you don’t know what you’re doing!
Edwards Deming

Basic symbols of the flow chart

MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)

Figure 7.2

 WE WHANT TO MAKE IMPROVEMENT ON A TOASTER (reduce defects by 25%).

 Most defect happens during final testing

Figure 7.3

Figure 7.4

2. Check Sheets

A check sheet is a simple tool for collecting, organizing and analyzing data.
MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)
 Check sheets are particularly effective for doing inspections, enabling focus on the
particular attributes that may be contributing to potential or identified quality problems.
 It makes data collection easy and accurate
 Very effective for quick analysis
 It is also known as Tally sheet.
7.1 Example: what is wrong with cheek sheet below?

Table 7.2

 Data collected using Check sheet need to be meaningfully classified.

 A technic that separates data gathered from a variety of sources so that patterns can be
seen is stratification or process fl

 Table 7.5: Toaster

How do we know what to focus on? The pareto chart?

3. Pareto Diagrams

A Pareto chart or diagram, is a specific type of histogram that allows for analysis of data in
search of the Pareto principle or the 80/20 rule, which states that a large number of defects or
problems are caused by a small number of causes or for many outcomes, roughly 80% of
consequences (consequences) come from 20% of causes (the "vital few").

 A Pareto diagram is an ordered bar graph showing the number of defects and their causes
ranked by frequency. The bars on the diagram graphically show the number and
percentage of causes individually and the line shows the cumulative value. Pareto charts
MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)
help focus attention on the most critical issues to get the most benefit.

Figure 7.5: Pareto Diagram

 80% of the land was own by 20% of the people.

 Help us separate the vital few from the trivial many.

Fig. 7.6 Toaster example

Why these issues are happening?

4. Cause and Effect Diagrams

The cause and effect diagram or or Ishikawa diagram is a visual too to explore all the potential
factors that may be causing or contributing to a particular problem (effect).
 Cause-and-effect diagrams, , were developed by Kaoru Ishikawa to illustrate and help
determine how various factors relate to potential problems.
 The head of the fish is the effect and each bone of the fish is a cause that leads to that
effect.
 When all the bones are filled in, the diagram lets you look at all the possible causes
(individual or combinations) of the effect (or problem) so that you can develop a solution
to mitigate that effect.
 The diagram allows organized thought and encourages orderly consideration of the
MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)
factors that result in a certain outcome.

Figure 7.7: Cause and Effect Diagram

Figure 7.8: Cause and Effect Diagram for PCB failure

What is the relationship between the defect rate and the humidity level?

5. Scatter Diagrams

It is a visual analysis tool that shows the possible relationship between two variables.

 Scatter diagrams plot two variables, the independent variable and the dependent variable,
to graphically show the relationship between them.

 The X -axis in the diagram represents one characteristic (usually the independent
variable), and the Y -axis measures the other.

 To interpret the diagram, look at two characteristics of the clustering:

o Tightness: The closer the cluster is to a diagonal line drawn through the graph,
the more the two variables are likely to be linearly correlated. High correlation
between the characteristics means that a change in one characteristic will be
accompanied by a change in the other.
MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)
o Direction: If the correlation is positive, then as one variable increases so does the
other, and the line will have a positive slope (from lower left to upper right). On
the other hand, if the correlation is negative, it implies that as one characteristic
increases, the other decreases, and the line will have a negative slope (from lower
right to upper left).

Figure 7.9: Scatter Diagram

 Two variable can correlate without having a cause and effect relationship
 The strength of the relationship between two variables can be expressed using Pearson
correlation coefficient.

Figure 7.10: Pearson correlation

 How is relative humidity changing over time within the facility?

6. Histograms

A histogram is a type that graphs the frequency of occurrence of continuous data and is a
useful toot for displaying, summarizing and analyzing data.
 Vertical bar graph represents the frequency of each measured category (known as bins) of
variable.
 The histogram is particularly useful for identifying common causes.
MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)
 The histogram can be ordered, similar to a Pareto chart, or unordered.
 Every process or product has some level of variation that will occur in a pattern.
The best way to see that pattern is to graph your data using a histogram.

7. Control Charts

Statistical Process Control (SPC) is a collection of statistical tools (control charts) that
allow us to ensure that a process is in control (stable/normal variation) or out of statistical
control (unstable/special cause variation).

 Control charts are used to determine if processes are in most processes experience a
degree of random, normal process variation (or common cause variation); that is to
say, most processes do not achieve target performance all the time.

 Control charts provide a mechanism for establishing a statistically objective range of

acceptable variation around the target performance, thereby enabling attention to be
focused on special cause variations (those that fall outside of the established performance
limits).

 Dr. Walter A.Shewhart of Bell Telephone Laboratories developed a theory of

control charts in 1920s.

 Variation due to assignable causes can be economically discovered using control charts.

 Variation due to random or chance causes cannot be economically discovered (using

control charts) and cannot be removed without making basic changes in process.
MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)
a) Types of Control Charts

 Control Charts for Variables:

1. Average (X-bar) and range (R) charts

2. Average (X-bar) and standard deviation (s)charts
3. Individual (X) and moving average (MR) charts

 Control Charts for Attributes

1. Proportion Nonconforming (p) charts

2. Number Defective (np) Charts
3. Number of defects (c) charts
4. Nonconformities per unit (u) charts

b) Rules to detect special causes: Criteria for Out of Control

While investigating consider:

a. Tool/die wear/breakage
b. Change of material
c. Change of operator
d. Change of measuring equipment
e. Equipment condition
f. External factors such as voltage fluctuations, temperature/humidity changes,
vibrations etc.
g. Take action to eliminate the assignable cause so that it will not recur
h. Recalculate the control limits ignoring subgroups
i. corresponding to the assignable causes

c) Key elements of a control chart

 Control limits create the boundaries between common cause variation and special cause
variation.
MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)

 Rational subgrouping: it is a collection of units that are all produced under the same
conditions.
o Use a good rational subgrouping by collecting data all produced under the same
conditions.
o By including only the normal, inherent process variation in our rational subgroup,
we ensure that our control limits are appropriately sensitive to special cause of
variation.

 X-Bar and R chart equations (we monitor the charts to detect special cause variation)

o X-bar chart monitors the mean of a process.

MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)

o R-(Range) chart monitors the variation within a process.

Table 7.xx X-Bar and R chart constants

MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)

Example 1: Is the following process in control?

MECE5208 Reliability Engineering Dr. Tsapi Kevin
NAHPI, DEPARTMENT OF MECHANICAL AND INDUSTRIAL ENGINEERING (MIE)

Improving a process goes far beyond root cause analysis. You must be resolving issues by making
changes to your product or process. this can be done using the 7 management and planning tools.

Example 2

A company wants to monitor run-out of valve seat inserts on assembly used in cylinder heads.
• Data is collected every 30 minutes with subgroup size of 3
• Part has upper tolerance of 0.080 (80 microns)
• Data is shown in the worksheet.
• Plot X-bar and Range Control Charts and identify out of control points
• Calculate Process Capability from the dat
• Discuss your results