The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1741-0401.htm

IJPPM
58,3 REFLECTIVE PRACTICE
Using DMAIC Six Sigma to
254 systematically improve shopfloor
Received June 2008
production quality and costs
Accepted August 2008 Sameer Kumar and Michael Sosnoski
Opus College of Business, University of St Thomas, Minneapolis,
Minnesota, USA
Abstract
Purpose – This study sets out to highlight how a leading manufacturer of tooling, with its
continuous improvement drive, has leveraged DMAIC Six Sigma potential to realize cost savings and
improved quality on their shopfloor. The study examines one of the shopfloor chronic quality issues
which deals with finding a way to reduce the amount of warp incurred in Amada A-Station punches
during the heat-treat process.
Design/methodology//approach – The goals of this specific study were achieved by utilizing Six
Sigma tools such as “brainstorming, process mapping, fish-bone diagrams, histograms, and control
charts”.
Findings – The analysis resulted in a number of findings and recommendations. A major contributor
to the warp was the method used to fix the parts. It was found that, by using a new fixture to hang the
parts, the amount of warp could be greatly reduced. The major recommendation was that the new
heat-treat fixture design be implemented as soon as a reliable fixture design has been proven. The
value of implementing this new fixture design equates to roughly $10,000 per year in savings, which is
due to the elimination of a secondary grinding operation and elimination of scrap parts. The company
expects millions of dollars in annual savings as DMAIC Six Sigma process is rolled out to areas on the
shopfloor.
Originality/value – The case study shows DMAIC Six Sigma process is an effective and novel
approach for the machining and fabrication industries to improve the quality of their processes and
products and profitability through driving down manufacturing costs.
Keywords Six sigma, Control charts, Production economics, Tool manufacturers, Quality control
Paper type Case study

Introduction
The primary purpose of this case study is to underscore how a leading manufacturer of
tooling with its continuous improvement initiative is utilizing DMAIC Six Sigma
capability. Results achieved are reflected in manufacturing cost savings and quality of
product and process improvements on their shopfloor. The study specifically examines
one of the shopfloor chronic quality issues to reduce the amount of warp incurred in
Amada A-Station punches during the heat-treat process. Using DMAIC Six Sigma
International Journal of Productivity methodology, this study leverages various Six Sigma tools such as “brainstorming,
and Performance Management
Vol. 58 No. 3, 2009
process mapping, fish-bone diagrams, histograms, and control charts” to provide
pp. 254-273 decision support to a Punch Value Stream improvement team.
q Emerald Group Publishing Limited
1741-0401
The remainder of the paper is laid out as follows: The first section reviews the
DOI 10.1108/17410400910938850 existing work reported in literature that relates to the study described in this paper.
Second, it describes the Six Sigma framework and the integration of various quality Using DMAIC
tools to the chosen quality improvement project by re-engineering an improved
heat-treat process to mitigate warping. The benefit-cost analysis of Six Sigma
Six Sigma
implementation as part of continuous improvement is presented next. This is followed
by elaborating on the managerial implications of systematic DMAIC Six Sigma
implementation in the entire organization. Finally, conclusions and recommendations
for future work are presented. 255
Literature review
The literature review reported in this section can be summarized in three groups –
waste reduction in inventory (supply chains), Six Sigma companies, and material
warping.
A significant challenge in today’s manufacturing is to be both efficient and highly
effective in terms of customer satisfaction. The latter results in an emphasis on both
relations with customers and the service provided to such customers (Hitt et al., 1999).
Kotler (1997, p. 109) argued, “Customers are scarce; without them, the company ceases
to exist. Plans must be laid to acquire and keep customers”. The level of competition to
capture customers in both domestic and international markets demands that
organizations be quick, agile, and flexible to compete effectively (LaLonde, 1997;
Fliedner and Vokurka, 1997; Wang et al., 2005; Lapide, 2000).
Time and quality based competition focuses on eliminating waste in the form of time,
effort, defective units, and inventory in manufacturing-distribution systems (Larson and
Lusch, 1990; Schönberger and El-Ansary, 1984; Schultz, 1985). In addition, there has been
a significant trend to emphasize quality, not only in the production of products or
services, but also, throughout all areas in a company (Coyle et al., 1996; Scalise, 2003).
Ted Farris of the University of North Texas reports 66 percent of the cash-to-cash
improvements since 1986 in the majority of industry segments have come from
reductions in days of inventory (Trunick, 2005). According to Farris, the cash-to-cash
formula adds accounts receivable to inventory and then subtracts accounts payable. If
any improvement to operations results in a reduction in the amount of inventory held
within the company, days of inventory are reduced. But, if the company is merely
shifting inventories intra-company to hold the goods somewhere else, there is no change
in the cash-to-cash cycle. This inventory is still carrying an associated accounts payable
tag – it has not shifted to a receivable because it has not been sold.
Many companies have developed and implemented Six Sigma approaches. Among
the noted ones are: General Electric (GE), DuPont, Honeywell, and Samsung. A brief
perspective is provided in the following paragraphs on how these companies have
applied Six Sigma to similar needs.
We begin with General Electric. A key element of GE’s approach to Six Sigma is
tailoring underlying methodologies to specific needs and characteristics of its business
units. GE has taken the generic Six Sigma methodology for process innovation, and
has tailored their specific needs of system design and implementation as well as
product development activities.
DuPont combines Six Sigma principles with the SCOR (Supply Chain Operations
Reference) model, which scopes five core management processes – including “plan,
source, make, deliver, and return”. DuPont’s Six Sigma approach utilizes a quality
function deployment (QFD) tool – a method for converting customers’ requirements to
products, processes or services.
IJPPM Honeywell is known for extensive application of lean methodologies, which has
become a major tool in their Six Sigma implementation. Honeywell developed a
58,3 proprietary Six Sigma approach called Six Sigma Plus which links lean manufacturing
concepts and tools, such as “value stream maps and thought process maps”, into a
general Six Sigma framework.
At Samsung, Six Sigma projects usually focus on either redesigning processes and
256 systems or improving performance levels of existing systems. In Six Sigma parlance,
the former is addressed most commonly through DMADOV (define, measure, analyze,
design, optimize, verify), while the latter is addressed using DMAIC (define, measure,
analyze, improve, control). Samsung’s Business Team estimated that among the
Supply Chain Management (SCM) projects at Samsung, about 75 percent would
involve redesigning processes, while the remaining 25 percent would focus on process
improvement. The DMADOV methodology, while useful, could not provide the
necessary support to execute the entire range of SCM projects at Samsung. In most
cases, SCM Six Sigma projects result in system development. Accordingly, a new
approach was needed. DMAEV (define, measure, analyze, enable, and verify) is the
resulting approach. Additionally, Samsung incorporated the concept of five design
parameters (process, operation rule and policy, organization role and responsibility,
performance measure, and system), process modeling and value chain map techniques,
and SCM related investment value analysis methods. Samsung’s endeavor for global
optimum is continuing and SCM Six Sigma is expected to play an enabling role (Yang
et al., 2007; Knowles et al., 2005).
Several metallurgical books were reviewed in order to find information related to
materials warping during heat-treat. Most books did not discuss this topic although two
books were found that did reference the topic. The first was an out-of-print book
published by Crucible Materials Corporation titled Tool Steel for the Non-Metallurgist. In
a few short paragraphs, this book summarized two types of distortion during heat-treat
size-change and shape-change. In this project, we are concerned about shape-change. The
book went on to discuss how stresses during machining, heating too fast during
heat-treat, and insufficient support during heat-treat can lead to shape distortion
(Crucible Materials Corporation, 1988, p. 38). The second book, titled Tool Materials,
explained in more detail the causes of shape-change distortion. This book explained that
rapid heating causes distortion, but, this is typically found in large complex shapes. The
book also explained that when tool steels are heated over 1,100-degrees Fahrenheit, the
yield strength becomes so low that deformation is often caused by gravity. Thus, long
parts must be properly supported to prevent distortion (1995, p. 348). This information
will be used to brainstorm factors that cause the warped parts.

Six Sigma framework

The project team followed the DMAIC process within a Six Sigma framework. The five
DMAIC phases and key tools used in each phase are listed below:
(1) D ¼ Define. Pareto Analysis; Project Charter.
(2) M ¼ Measure. Descriptive Statistics; Process Capability Analysis.
(3) A ¼ Analyze. Detailed Process Map; Fish-Bone Diagram.
(4) I ¼ Improve. Experimentation; New Process.
(5) C ¼ Control. Statistical Process Control.
The five Six Sigma phases (Define, Measure, Analyze, Improve and Control) are Using DMAIC
discussed in detail in this section.
Six Sigma
Define phase
The first phase in a DMAIC Six Sigma project is the Define phase that uses Pareto
analysis to identify and prioritize projects that will be worked on as part of continuous
improvement process in the company. This is followed by using a project charter to 257
organize the major project categories.
Pareto analysis. Wilson Tool is one of the world’s leaders in supplying tooling to the
fabrication industry. Wilson Tool manufactures tooling for punch presses, press
brakes, and stamping equipment. Wilson Tool has an award winning continuous
improvement program, mainly due to its efforts in lean manufacturing. Wilson Tool is
now starting to look at the tools available in the DMAIC Six Sigma process to further
its continuous improvement efforts.
Over the last year, the shopfloor employees at Wilson Tool have been tracking quality
issues and errors. The data is logged in a database containing over 15,000 entries and is
used to track quality issues. Each department within the company displays a Pareto
Chart of the top errors. Figure 1, displayed below, is a Pareto Chart for the Punch Value
Stream. Pareto Charts are an easy to read tool for identifying top quality issues.
The Pareto Chart (Figure 1) shows the top ten errors of the Punch Value Stream.
The area of focus for this project is the issue of warped parts. Although the problem of
warped parts is not the greatest source of quality problems, it is a problem that has
existed for many years and should be significantly decreased, if not eliminated.
It is the goal of Wilson Tool to eliminate the cause of the top two or three quality
issues in each area on an annual basis. Over multiple years, the majority of the quality
issues should be solved. The quality issue selected for this project was Warped, which

Figure 1.
Pareto chart
IJPPM refers to bent parts. The issue of warped parts only accounted for 3 percent of the
58,3 quality issues although it was decided that it would be a good test for the Six Sigma
process because it was narrowly defined, measurable, and had been a known problem
within Wilson Tool for many years.
The “A-station Amada Punch” is one of Wilson Tools oldest products. This punch
is 5/8-inches in diameter and 8-inches long; a picture of the assembly that uses this
258 punch (which is the inner most component of this assembly) is displayed in Figure 2.
This punch is made from tool steel, which requires turning, heat-treat, and grinding
processes. This assembly is used in a punch press to cut or form sheet metal. This
product has a known problem of warping during the heat-treat process. To overcome
this problem, additional material had been added to the diameter to allow for the
additional grinding needed to straighten this punch. This extra material requires the
diameter of the punch to be ground twice due to depth-of-cut limitations of the grinding
process. The warping problem was escalated early this fall when an exceptionally bad
batch of punches was produced. These punches were so warped that they crashed the
machine, which caused machine downtime and required pre-sorting of parts to prevent
further crashes. The goal of this Six-Sigma project is to determine how to heat-treat
this product without causing it to warp.

Figure 2.
Amada punch assembly
Project charter Using DMAIC
The problem statement, project objective statement, project scope/limitations, project Six Sigma
goals and targets, expected benefits and general information on the project are
specified as part of the Project Charter.
Problem statement. Amada A-Station Punches have a diameter of 0.6245-inches, are
8.16900 long, and are made of tool steel, which requires a heat-treat operation. Because of
the long narrow shape of these punches, they have always had a problem with warping 259
during heat-treat, which causes an additional manufacturing process and waste in the
form of scrap.
Project objective statement. The objective of this project is to reduce the amount that
these punches warp during heat-treat to less than 0.00500 over the length of the punch.
This will eliminate any scrap related to warping and it will allow the Punch Value
Stream to grind the outside diameter of the punches in one operation.
Project scope/limitations. Limitations associated with this project include:
.
the tool steel grade must remain the same;
. the hardness of the punches must not change;
.
the toughness of the punches must not change, which means the rate of quench
during heat-treat must remain the same; and
.
the cost to manufacture these parts must not increase.
Project goals and targets. The goal of this project is to produce punches with less than
six out of 18,000 punches that warp greater than 0.00500 without adding cost to the
process. This acceptable defect rate of 334 DPM was established based on accepting
one defective punch every other month by the area supervisor.
This goal seemed reasonable considering the large amount of known variability in
the current process. Once this goal has been established, the goal can be reset and the
process refined further, as time permits.
Expected benefits. Hard dollar savings of $12,500 per year are expected with the
elimination of the second grind operation. Savings due to a reduction in scrapped parts
is expected to be $1,646.40 per year. It could be theorized that straighter punches would
perform better and create less side loading in the assembly, which would increase
product quality as an added benefit.
General information. The customer of this project is the Punch Value Stream, in
particular the outside diameter grinder. The current business objective of the division is
to decrease manufacturing cost in order to increase divisional profits. The project team
included – Project Lead; Tool Room Support; Project Sponsor and Business Unit Leader
of the Punch Value Stream. Assistance was provided by one of Wilson Tools internal
Heat-Treat Operator, a Metallurgist for Wilson Tool and also a Heat-Treat Vendor.

Measure phase
The amount of after-heat-treat warp was measured in each part using a 0.0001-inch
Dial Indicator. The parts were supported on each end using V-blocks. The parts were
rotated while the total indicator run-out was recorded. Plate 1 shows the fixture used
during the inspection process to measure the warp contained in the punches. The
fixture utilized V-blocks to support each end of the punch. The Dial-Indicator measured
the amount of warp contained in each punch as the punches were rotated.
IJPPM
58,3

260

Plate 1.
Inspection fixture

Descriptive statistics. Basic descriptive statistics were calculated using Mini-Tab for the
current process and are listed in Table I. These statistics were based on a sample of 100
measurements. The mean warp of the process is 0.004300 and the standard deviation is
0.001900 . The sample had a range of 0.000500 to 0.011700 . The current process produces a Cp
of 1.00 (which was just coincidental) and a DPM of 1,350. (Note: a Cp was used in place of
a Cpk since an error can only deviate from zero in the positive direction.). These statistics
showed that the current process produces a large amount of warp.
The histogram displayed in Figure 3 shows that the population of the current
process is bi-modal and has some outliers with as much as 0.01200 of warp. The current
process produced a mean warp of 0.0043” with a standard deviation of 0.001900 . The
sample found that the current process produced parts with as much as 0.011700 of warp.
Process capability analysis. The Range Control Chart (based on sample size of 5 for each
data point), displayed in Figure 4, shows that the current process is out of control since
data point 9 is outside the upper control limit (UCL). The other problem with the process is
the gap of 0.00700 between control limits, which is unacceptable for the desired process.
The Mean Control Chart (based on sample size of 5 for each data point), displayed in
Figure 5, also shows that the process is out of control since data points 2, 3, 5, 6, and 9
are outside the control limits. Again, the average warp, and the spread between the
control limits, is too large.

Analyze phase
The analysis of the process will use two Six Sigma tools: a detailed process map and a
fish-bone diagram (otherwise known as a cause-and-effect diagram).

Table I. Variable n Mean SE mean SD Minimum Q1 Median Q3 Maximum

Basic descriptive statistics
for current process Current value 100 0.0043 0.0002 0.0019 0.0005 0.0028 0.0041 0.0057 0.0117
 Using DMAIC
Six Sigma

261

Figure 3.
Histogram of current
process

Figure 4.
Range control chart for the
current process

Figure 5.
Mean control chart for the
current process (s
unknown)
IJPPM Detailed process map. The value stream map, displayed in Figure 6, illustrates the
manufacturing process for Amada Punches. The map starts with the truck that
58,3 delivers the raw material to the plant which is then stored in the raw material storage
location until it is issued to a production order. The raw material is then transported to
the turning center via handcart. The turning-center machines all the features into the
punch but leaves 0.01000 extra material on the outside diameter and 0.00500 on each end
262 of the punch. This extra material will be ground off after the heat-treat operations.
After the punch is turned, it is then transported to an outside heat-treat vendor where it
is heat-treated. After heat-treat, the punch is transported back to Wilson Tool where
the ends and outside diameters are ground. The warped punches are causing problems
at the outside diameter grinder, which means that one of the steps prior to this machine
is the cause of the warp. The parts were measured before heat-treat and found to be
very straight. When the parts were measured after heat-treat, they were found to be
quite warped. This means that the heat-treat process is causing the punches to warp.
Fish bone diagram. In order to narrow the focus of the project, parts were inspected
before heat-treat and the total warp before heat-treat was 0.000000 to 0.000400 . This
narrowed the project focus to the heat-treat process. The fish-bone diagram contained
in Figure 7 was used to identify possible causes to the warped punches during the
heat-treat process. This project focused on the heat-treat fixtures along with the
process used to load the punches in the fixtures. These two factors were selected
because they were easy to observe and control, whereas many of the other factors
required observations during the heat-treat process, which is very difficult since the
temperatures inside the heat-treat oven are over 1,000-degrees Fahrenheit.

Improve phase
Of the two types of variation, common cause and special cause, the team decided to
focus on the special cause variation first because the assignable causes make it easier
to address. Of the special cause variation, the loading of parts into the heat-treat
fixtures and the heat-treat fixtures moving during heat-treat seemed to be the easiest to
address.
Experimentation. The group brainstormed ideas and decided to focus on two
options, hanging the parts in a new fixture and laying the parts flat on the bottom of
the heat-treat fixture. It was theorized that by hanging the parts, gravity would
keep the parts straight. The other theory was, that by laying the parts flat on the
bottom of the fixture they will be fully and evenly supported. It was decided to test
both methods and compare the results to a baseline using the current fixture. All three
methods would be heat-treated in the same batch, and made from the same batch of
raw material, in order to reduce outside variations.
The Tool Room fabricated a simple fixture to test the hanging of the parts. A slot
was added to the parts to allow the parts to hang from the fixture. Figure 8 shows the
heat-treat fixture loaded with parts to be tested.
The picture displayed in Figure 8a shows the prototype fixture used to test heat
treating the punches hanging from a groove. Figure 8b shows the groove that was added
to allow the punches to hang. The theory behind this fixture is that hanging punches will
not collapse under their own weight during the high heat of the heat-treat process.
After heat-treat, results from the first test were compiled and summarized in the
histogram displayed in Figure 9. This test compared the current process, with fixtures
parts standing on end, to the process of hanging parts, and the process of laying parts
 Using DMAIC
Six Sigma

263

Figure 6.
Process map for Amada
punch production
IJPPM
58,3

264

Figure 7.
Fish bone diagram

Figure 8.
Prototype heat-treat
fixture
 Using DMAIC
Six Sigma

265

Figure 9.
Histogram of test one
results – notched, flat,
normal

flat on the bottom of the heat-treat basket. The current method of fixturing the parts
showed that the average warp was 0.00700 with a standard deviation of 0.001700 . Laying
the parts flat decreased the average warp to 0.00600 but the standard deviation
increased to 0.002400 , which means we added variability to the process (which is
undesirable). Hanging the parts showed great results; the average warp was 0.00200
with a standard deviation of 0.000700 . If this process can be repeated we will have
reached our goal with a defect rate of 9-DPM. In summary, the test showed that
hanging the parts produced the least amount of warp and least amount of variation.
Laying parts flat decreased the warp but increased the amount of variation.
New process. Since the test of hanging the parts worked so well, the group decided to
continue to refine the hanging concept. As the first test was conducted, the group
received feedback related to the design of the heat-treat fixture. Comments included:
.
The fixture is very heavy which makes it difficult to move and increases the heat
up time in the ovens. The additional weight associated with the fixture will increase
the cost to heat-treat because the company is charged on a per pound basis.
.
The total weight of the fixture, with punches loaded, should be light enough for
an operator to lift by hand. This might require more fixtures with fewer parts.
.
The keyhole slots seemed to work well for loading and for holding the parts in
the ovens.
. A locking mechanism is needed to keep the punches in the keyhole slots because
the punches tended to fall out of the fixture when the fixture is moved.
.
Additional holes are needed in the top and bottom of the fixture to allow air to
flow from the top of the fixture through the bottom during the quench.

One of the barriers to the new process was the notch that was added to the punch. The
R&D department did not have any concerns about the notch decreasing the performance
of the punch. Although concerns existed about how the customer would perceive the
notch. It was determined that notch should not be added. In order to get around the notch
IJPPM problem, another test was performed, which compared hanging the parts from the notch
58,3 to hanging the parts from the top oil groove. Figure 10 illustrates the difference between
the top oil groove and the groove that was added for test 1. Before conducting this second
test, it was theorized that the results from hanging the parts from the oil groove would be
less impressive than hanging from the notch since the rounded oil grove allowed the
punches to hang at an angle. In contrast to the group’s theory, hanging the parts from the
266 oil grove produced straighter parts. The histogram displayed in Figure 11 shows that the
un-notched parts produced less warp, on average they were warped 0.001300 , and they
were more consistent with a standard deviation of 0.000800 . This meant that the parts
could hang without changing the design of the punch.
As mentioned earlier, Figure 10 shows another option to hanging the parts by using
the already existing oil grooves. Originally it was suspected that the rounded oil
grooves would not work as well as the square shoulders of the added groove. Testing
proved that the oil groove actually worked better than the groove that was added.
The histogram displayed in Figure 11 above shows the results from test two, which
compared hanging parts from the notch to hanging parts from the first oil groove. The
test proved that hanging parts from the rounded oil groove produced straighter and
more consistent parts than by hanging parts form the square notch that was added for
test purposes.
After it was determined that the parts could be hung from the oil groves, the Tool
Room designed a new fixture that would be used for production purposes as displayed
in Plate 2. This new fixture incorporated all the things learned from the previous tests.
The new fixture was made of stainless steel sheet metal, which is lightweight. It held
72 punches so that when the fixture was fully loaded it would weigh less than
50-pounds, which allows the operators to safely lift the fixture by hand. The fixture
had many holes to allow air to flow during the quench process. To lock the punches in
place after loading, steel rods were place between the rows of punches. The picture

Figure 10.
Oil groove illustration
 Using DMAIC
Six Sigma

267

Figure 11.
Histogram of test two
results: hanging parts
from the oil groove

Plate 2.
New fixture design

(Plate 2) shows the new heat-treat fixture fully loaded and ready to be shipped to the
heat-treat vendor. This fixture design is much less expensive to manufacture and
incorporated many of the improvements identified during the first two tests.
When the new heat-treat fixtures returned from heat-treat, it was immediately
obvious that the fixtures were not able to withstand the head-treat process. Under the
high temperatures of heat-treat, both the sides and the top sagged until some of the
punches hit the bottom of the fixture. The failed heat treat fixture is displayed in
Plate 3.
IJPPM
58,3

268

Plate 3.
Failed heat-treat fixture

Control phase
The results of the new fixture were less impressive than the results from the test
fixture. The decreased performance can be attributed to the fact that the fixture warped
and some of the punches hit the bottom of the fixture, which caused those punches to
bend. The histogram displayed in Figure 12 shows the results from the new fixture.
The parts had a mean warp of 0.002800 with a standard deviation of 0.0013. The process
using the production fixtures produced a cp of 1.842 if the upper tolerance were to
remain at 0.01000 . Since our goal is to eliminate the second grind operation, our upper
tolerance limit will be 0.00500 , which produces a cp of 0.559 or 45,294 DPM. New fixtures
will need to be re-designed and made of thicker material to be more rigid. This should

Figure 12.
Histogram for the process
of hanging parts during
heat-treat
prevent the fixtures from sagging and allow the fixtures to be used many times before Using DMAIC
they need to be replaced. Six Sigma
Table II contains basic descriptive statistics for the process of hanging parts during
heat-treat. These statistics were based on a sample of 46 measurements, which
produced a mean warp of 0.002800 with a standard deviation of 0.001300 . The sample had
a range of 0.000800 to 0.005700 . These statistics showed that the process of hanging parts
during heat treat produced much less warp than the current process. 269
The histogram illustrated in Figure 12 shows the results from hanging parts during
heat-treat.
Statistical process control. The Range Control Chart and the Mean Control Chart
displayed in Figures 13 and 14 show the results from the process of hanging parts

Table II.
Variable n Mean SE mean SD Minimum Q1 Median Q3 Maximum Basic descriptive statistics
for the process of hanging
Current value 46 0.0028 0.0002 0.0013 0.0008 0.0018 0.0028 0.0038 0.0057 parts during heat-treat

Figure 13.
Range control chart for
hanging parts during heat
treat

Figure 14.
Mean control chart for
hanging parts during heat
treat (s unknown)
IJPPM during heat-treat. The process appears to be better, but still out of control, which can be
assigned to the fixtures sagging during heat-treat. Once the new heat-treat fixtures are
58,3 fabricated, the test will need to be repeated. The data from the first set of fixtures will
be discarded and an additional 30-data points will need to be gathered before it is
determined that the process is stable.
The Range Control Chart (Figure 13) shows that the new process is in control. The
270 chart displays only five points because testing was stopped due to the sagging of the
heat-treat fixtures.
The Mean Control Chart (Figure 14) shows the new process is out of control since
data point 2 is below the lower control limit. The out of control state of the new process
is most likely due to the heat treat fixtures that sagged.

Return on quality
In order to change the process, it must be determined that the benefits from the new
process outweigh the costs. Assuming the new fixture achieves the intended goals, the
grind stock will be reduced to 0.00500 and the second grind operation eliminated. The
labor savings due to eliminating the second grind operation equals 15-seconds per part
(0.25-minutes) at a shop rate of $2.50 per minute or $0.63 per part. This savings will be
recognized on approximately 20,000 parts per year which equates to an annual savings
of $12,500 (0:25 £ $2:50 £ 20; 000 ¼ $12; 500).
The improved process should reduce the scrap associated with bent parts. The
current method produced approximately 8 scrapped parts per month, valued at $17.15
each, which equated to $1,646.40 per year (8 parts per month £ 12 months £ $17.15
per punch ¼ $1; 646:40 per year). The process would also provide some soft costs such
as better punch performance due to straighter punches and less handling costs since
the parts could be transported, cleaned, and heat-treated in the same fixture. (The
current process involves putting each part in a bin for transporting, placing each part
in a basket for cleaning, placing parts in the heat-treat fixture, and finally placing parts
back in the bin for transportation.)
The cost of the new process includes the cost of the stainless steel used to make the
new fixtures and the cost to stamp, bend, and weld the fixtures. The fixtures were
stamped from a 2400 £ 4800 sheet metal. Each sheet cost $46.45 and produced two
fixtures for a total raw material cost of $23.23 per fixture ($46:45=2 ¼ $23:23). Using a
punch press allowed the operator to quickly cut the sheet metal. The operator then bent
the sides of the fixture and spot-welded it together. These three steps consumed
seven-minutes of labor per fixture at a shop rate of $2.50. Thus, the total labor cost to
produce one fixture was $17.50. The cost per fixture is summarized in Table III.
The total cost of the project depends on the fixture life. At this time, the fixture life is
unknown. Assuming that a fixture design is created that allows each fixture to be used
three times before it is discarded and each fixture is used once per month, the fixtures
will need to be replaced four times per year. Based on these assumptions, 88-fixtures
will need to be fabricated (22 fixtures replaced 4 times ¼ 88 fixtures). The 88-fixtures
required for a year’s production will cost $3,583.80 (88 fixtures £ $40.73 per
fixture ¼ $3; 583:80).
Table III shows the cost per fixture, including materials and labor, was $40.73 each.
In order to convert the entire production process to the new fixture, 22 fixtures will be
needed. Since the fixture life has not yet been established, an estimated fixture life of
 Using DMAIC
Item Cost Units
Six Sigma
Sheet metal (2400 £ 4800 ) $46.45
Fixtures per sheet 2
Raw material cost $23.23 Per fixture
Punch press labor 3 Min.
Press brake labor 2 Min. 271
Spot weld labor 2 Min.
Total labor 7 Min.
Shop rate $2.50 Per min.
Total labor cost $17.50
Total fixture cost $40.73
Number of fixtures needed 22
Fixture life 3 Heat treat cycles Table III.
Total cost $3,583.80 Fixture cost

three heat treat cycle was used. Thus, the annual fixture cost for the new process would
be $3,583.80.
The worst-case fixture life scenario must be examined since the life expectancy of
the fixtures is unknown at this point. Assuming that the fixtures were discarded after
each use, a total of 22-fixtures would need to be replaced each month for an annual
fixture consumption of 264 fixtures (22 fixtures replaced 12 times ¼ 264 fixtures). The
total cost of producing 264 fixtures is $10,751.40 per year (264 fixtures £ $40.73 per
fixture ¼ $10; 751:40). Based on this information, even if the fixtures have a life of only
one heat-treat cycle, the total cost ($10,751.40) would not exceed the benefits of this
project ($14,146.40).
The total savings per year due to this project is directly affected by the fixture life,
which is unknown at this point. Assuming the fixtures are used three times before
being discarded, then the total project savings per year will be $10,562.60 (based on a
fixture life of three heat treat cycles) as illustrated in Table IV. If the fixtures are
discarded after each use, then the total project savings per year will be $3,395 as
illustrated in Table IV. In either scenario, the project will save the company money.

Managerial implications
Conducting this project had a large impact on areas outside of the heat-treat and
grinding operations. Other areas of the shopfloor benefited from this project, the
product quality improved in general, Wilson Tool as a whole benefited, and Wilson
Tools customers benefited.
Besides reducing scrap and eliminating operations, the fixtures also reduced
non-value added activities at the turning center before heat-treat where operators were

Fixture life 3-loads $ Fixture life 1-load $

Decreased scrap savings 1,646.40 Decreased scrap savings 1,646.40

Decreased grinding operations savings 12,500.00 Decreased grinding operations savings 12,500.00
Fixture cost 3,583.80 Fixture cost 10,751.40 Table IV.
Total savings 10,562.60 Total savings 3,395.00 Total project savings
IJPPM no longer required to clean parts. It also reduced the amount of handling of parts by
allowing the parts to be transported, cleaned, and heat-treated in the same container.
58,3 Both of these improvements made the job easier, and better, for the employees.
This project increased the product quality. When punches are kept from warping, the
punch shanks will be better aligned with the threads. This will allow the punch head to
fasten properly on the punch. With proper alignment of parts that make up the assembly,
272 the assembly as a whole will function better. This leads to increased product life.
With improved quality comes a decrease in cost. By decreasing costs, Wilson Tool
can better meet their goals for profits and avoid price increases. This will allow Wilson
Tool to increase market share, which secures jobs for employees at Wilson Tool.
Finally, the customers of Wilson Tool will benefit from increased quality and
sustained or decreased prices. High levels of quality will also insure Wilson Tool’s
existence, thus Wilson can continue to serve its valued customers.
Recall the Pareto Chart in Figure 1 that showed warped parts as a relatively small
percentage (3 percent) of the total errors in the Punch Value Stream yet it will save the
company $10,000 per year. As the DMAIC Six Sigma process is rolled out to other projects
in the Punch Value Stream area, the savings should increase to 100,000s of dollars
($10,000/3 percent ¼ $333; 333). Assuming similar savings can be recognized throughout
the five other shopfloor value streams in the company, the savings could exceed millions
of dollars ($333,333 £ 6 value streams ¼ $2,000,000) which amount to increase in profit
by 2 percent of company’s annual revenue. The projected cost savings are very significant
when the low overall implementation cost is considered.

Conclusion and recommendations for future work

It is recommended that the company move forward with the new heat-treat fixture
design. This recommendation is based on the assumption that the desired quality level
can be achieved by hanging the punches, and given the cost benefit analysis. It is also
recommended that the company implement an ongoing control process using control
charts to monitor the amount of warp in the punches returning from heat-treat.
This project has demonstrated the power of the DMAIC Six Sigma process to solve
problems that appeared unsolvable. Wilson Tool had been suffering from warped parts
for nearly a half-century. The solution was identified, tested, and ready for
implementation within a few months. This project also demonstrated the power of
teamwork. Using cross-functional teams and involving employees from many areas of
the company allowed for the idea generation needed to solve this problem. Finally,
using the tools associated with the DMAIC Six Sigma process allowed this project to
confidently prove the outcome of the process.
The DMAIC Six Sigma process is an excellent fit for the machining and fabrication
industries. Any company that needs solutions for quality related problems should
benefit from this process. Companies will be most likely to succeed if their top-level
management is supportive of a continuous improvement culture. Companies should also
choose projects that can be measured and that potentially have a good return on quality.

References
Coyle, J.J., Bardi, E.J. and Langley, C.J. (1996), The Management of Business Logistics, 6th ed.,
West Publishing, St Paul, MN.
Crucible Materials Corporation (1988), Tool Steel for the Non-Metallurgist, Crucible Materials
Corporation, Syracuse, NY, p. 36.
Fliedner, G. and Vokurka, R.J. (1997), “Agility: competitive weapon of the 1990s and beyond”, Using DMAIC
Production and Inventory Management Journal, Vol. 38 No. 3, pp. 19-24.
Hitt, M.A., Clifford, P.G., Nixon, R.D. and Coyne, K. (1999), Dynamic Strategic Resources:
Six Sigma
Development Diffusion and Integration, Wiley, New York, NY.
Knowles, G., Whicker, L., Femat, J.H. and Canales, F.D.C. (2005), “A conceptual model for the
application of six sigma methodologies to supply chain improvement”, International
Journal of Logistics: Research and Applications, Vol. 8 No. 1, pp. 51-65. 273
Kotler, P. (1997), Marketing Management, 9th ed., Prentice-Hall, Englewood Cliffs, NJ.
LaLonde, B.J. (1997), “Supply chain management: myth or reality?”, Supply Chain Management
Review, Vol. 1, Spring, pp. 6-7.
Lapide, L. (2000), “True measures of supply chain performance”, Supply Chain Management
Review, July/August, pp. 25-8.
Larson, P.D. and Lusch, R.F. (1990), “Quick response retail technology: integration and
performance measurement”, International Review of Retail, Distribution and Consumer
Research, Vol. 1 No. 1, pp. 17-35.
Scalise, D. (2003), “Six Sigma in action”, Hospitals & Health Networks, Vol. 77 No. 5, p. 57.
Schönberger, R.J. and El-Ansary, A. (1984), “Just-in-time purchasing can improve quality”,
Journal of Purchasing and Materials Management, Vol. 20, pp. 1-7.
Schultz, D.P. (1985), “Just-in-time systems”, Stores, Vol. 67, April, pp. 28-31.
Trunick, P.A. (2005), “Time is inventory”, Logistics Today, Vol. 46 No. 4, pp. 26-7.
Wang, G., Huang, S.H. and Dismukes, J.P. (2005), “Manufacturing supply chain design and
evaluation”, International Journal of Advanced Manufacturing Technology, Vol. 25,
pp. 93-100.
Yang, H.M., Choi, B.S., Park, H.J., Suh, M.S. and Chae, B. (2007), “Supply chain management Six
Sigma: a management innovation methodology at the Samsung Group”, Supply Chain
Management: An International Journal, Vol. 12 No. 2, pp. 88-95.

Further reading
Davis, J. (1995), Tool Materials, ASM International, Materials Park, OH, p. 348.

About the authors

Sameer Kumar is currently a Professor of Decision Sciences and Qwest Endowed Chair in Global
Communications and Technology Management in the Opus College of Business, University of St
Thomas. Major areas of research interests include optimization concepts applied to various
aspects of global supply chain management, information systems, technology management,
product and process innovation, quality engineering and capital investment justifications.
Sameer Kumar is the corresponding author and can be contacted at: [email protected]
Michael Sosnoski has an Industrial Engineering degree from the University of Minnesota and
is currently employed at Wilson Tool International. Michael has had several positions at Wilson
Tool including manufacturing engineering, the project manager responsible for establishing
Wilson Tool Shanghai, and he is currently the Lead Manufacturing Engineer. Michael has
played a key role in Wilson Tool’s continuous improvement program and is an SAP Super-user.

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints