case study 4

Week10.pdf

Home >Computer Science homework help >case study 4

P erformance measures how well a product functions for its intended uses. If a product has multiple independent functions, its performance can be defined as “performance of the interacting chunks or modules of the product.”1

The performance of the individual chunks or modules, in turn, depends on the properties of the individual modules intended at the time of design. While manufacturing the product according to the design, the intended properties should be embedded, or there can be a deviation of the intended properties.

This deviation of the properties will lead to a fall in the product’s performance level. In turn, the overall product performance will deteriorate as a cascading effect of the falling performances of the individual chunks.

To study this dilemma, we looked at a manufacturer and its casting process of gear boxes used as parts of mechanical power transmission units, and we asked how Six Sigma could be applied to improve product performance.

The design, standards and norms (or intended properties) were set by the customer who needed the product free of defects—such as blowholes, gas poros- ity, cracks, shrinkage, misrun, dimensional accuracy and chilled surface—per the standards in the quality plan.

The objective of this work was to improve product performance by reducing the nonconformities of the manufactured product, thereby improving product performance and ensuring the product adhered to the design and specifications. The deviations in the process, if minimized and controlled, could lead to about 10-15% fewer defective parts shipped to the customer that wouldn’t have been typically caught by inspection.2 The deviations from the standards were found to be the result of the process.

Six Sigma was used to contain the defects or noises in the process, thereby improving performance. The process was mapped through root cause analysis (RCA), and failure mode and effects analysis (FMEA) was carried out using domain knowledge to identify the key process input variables responsible for failures.

Control mechanisms, too, were devised to monitor these inputs and thus reduce defects. After the defects were totally contained, the product manufac- tured more consistently met standards and norms that were set in the design stage by the customer.

At the foundry

In the example, quality consciousness became top of mind with management at the foundry because of increased competition. In any foundry, better quality means more competitive pricing of its product through less rework and rejection. Note that the casting process in a foundry is complex and involves diverse skills, different qualities of ingredients and technology. As variation in any process is inevitable, a high amount of variation makes it challenging to anticipate and manage results.

c a s e s t u d y

Molding a Solution

How Six SigMa

enHanced

product

perforMance

at an indian

foundry

By Prasun Das,

Indian Statistical

Institute, and

Abhranil Gupta,

Auctus Solutions

s i x s i g m a f o r u m m a g a z i n e I a u g u s t 2 0 1 2 I 9

M o l d i n g a S o l u t i o n

Statistical thinking focused on reducing variation found in the casting process, including core making, sand preparation, cupola operation, shot blasting and the measurement system.3 These stages of casting could be viewed as a series of interconnected processes that can result in variation but generate opportuni- ties for improvement if variation is reduced through identification, analysis and adequate control.4

The thinking, methods and tools needed to under- stand the nature of variation must be integrated into the decision-making process. The best results are achieved when principles based on statistical thinking are used as a guideline.

Six Sigma basics

Six Sigma—the systematic method for continuous pro- cess quality improvement and the process of achieving operational excellence—seeks to improve the quality of process outputs by identifying and removing the causes of defects (or errors) and minimizing variability in manufacturing and business processes.

Six Sigma has been on an incredible run for more than 15 years, producing significant bottom-line savings at many small and large organizations.5 Several papers and books have already documented the many benefits to deploying Six Sigma.6-12 Today, organizations have employed Six Sigma strategies in not only manufac- turing processes, but also in many types of business processes aimed to continuously reduce defects.13-17

The design, measure, analyze, improve and control (DMAIC) method is almost universally recognized and should be used with the objective of reducing defects and variation to improve business results and customer satisfaction.

DMAIC first asks project members to define core processes (define phase). Next is quantifying and benchmarking the process using actual data (mea- sure phase). Statistical tools are applied to validate root causes of problems and formulate options for improvement (analyze phase). Ideas and solutions are put to work and then validated (improve phase). Performance tracking mechanisms and measurements are put in place to ensure, at a minimum, the gains made in the project are not lost over a period of time (control phase).

Motivation, objective and scope

Any product is only as good as it is designed. Product design is an ideal scenario—if achieved and no causal relationship is overlooked—that will result in the

desired outputs. This goal should motivate organiza- tions to silence or at least quiet the noise in any process and improve product performance to the highest pos- sible levels.

The foundry’s objective was to use Six Sigma to reduce defects and nonconformity. To define the scope of work (the project’s boundary) and to keep the study’s objective on the right track, the following principles were adopted to yield maximum results:

• No metallurgical innovations would be per- formed.

• No dilution or change in the design and stan- dards, set by the customer, would be done.

• Define, select, measure and improve on the defects would only be done by the selected team.

• Management would periodically review the progress of the DMAIC phases.

Any further changes in scope could only be made through the proper channels and authority.

DMAIC at the foundry

Define phase: The purpose of this phase was to identify the project scope, define its objective and map the core process to quantify the extent of the problem in the next phase. The selection process of this project is out- lined in the project charter shown in Table 1. Figure 1 is a detailed process map of the foundry operations, which shows the process along with identification of critical process variables.

The following are definitions of terms used in the analysis:

• Defect: Any deviation of any characteristic from the standard.

• Defective unit: Any unit (casting item) that doesn’t conform to one or more of the standard

10 I a u g u s t 2 0 1 2 I W W W . a s Q . o r g

table 1. Project charter

Philosophy performance of any product is linked with the attribute and features of the product, which are manifested by the production pro- cess. any decrease in the variability of the process automatically improves the product performance by decreasing the variability in the product features.

Objective reduce the process variability to improve the process, which affects the product performance.

Scope Sand, sand mixing, mold making, cupola operation and pouring.

Present status Quality level of process (sigma level) is around 3.5.

parameters. A defective unit can have more than one defect.

• Defect opportunity: The way one or all types of defects can occur in a unit.

• Defect per million opportunities (DPMO): The number of defects occurring per million defect opportunities.

Measure phase: The purpose of this phase is to quan- tify the problem using available data in terms of the baseline sigma level. In the analysis at the foundry, historical data from the last three years of records were collected, and a Pareto chart was prepared for differ-

ent defect categories to extract the vital few defects. Figure 2 (p. 12) shows that blowholes, gas porosity and shrinkage were causing the majority of the problems (total defects). Table 2 (p. 12) shows the rejection percentage and the subsequent sigma level calculated from the raw data.

Analyze phase: The objective of this phase of our study was threefold:

1. Characterizing the vital few defects. Blowholes/gas porosity. These are defects in the

castings that occur when gas is trapped inside the mold and unable to escape from the vents created to

figure 1. Foundry process map

Takeout pattern

To reclaimed sand store

Store 4

Inspection 2

Operation 25

Transportation 10

Delay 5

To home return

To mold area

To cupola

To store

Assemble

Weigh pig iron

Weigh scrap

Weigh home return

Coal

To cupola

To sand mix

Storage: Storage of in-process, raw material and nished product in the inventory.

Inspection: Points in the system where the inspection is carried out.

Operation: Functions carried out to transform the raw material to nished product.

Transportation: Transporting work in progress inventory from one work station to another.

Delay: Delay in the process for any reason, including no job, no material and absence of workers.

Green

Reclaimed

Betonite

Water

Mold box

Add mold sand

Core box

Issue

S ta

Pattern

C u t

ga te

D ry

C u t

ga te

s P o u ri n g

C o o l

S h a ke

o u t

Fe lt in

In sp

e ct

io n

M e a su

re o u t

p ro

p o rt

io n

C he

ck t em

pe ra

tu re

D ry

m ix

P a n

K e e p r

e a d y

fo r

m o ld

in g

Ram

Mix

K e e p r

e a d y

fo r

p o u ri n g

P u t

in f

u rn

a ce

r e

s i x s i g m a f o r u m m a g a z i n e I a u g u s t 2 0 1 2 I 11

M o l d i n g a S o l u t i o n

table 2. Basic statistics for sigma level

Defects Frequency

Blowhole 98

gas porosity 83

Shrinkage 64

Misrun 50

dimension 20

crack 18

Hard metal 7

number of defects 340

number of castings in charge 1,838

defect opportunity 12,866

defects per million opportunities 26,426.24

Sigma level 3.44

figure 2. Pareto chart for vital few defect analysis

120

180

240

300

360

100

120

Bl ow

ho le

Ga s po

ro si ty

Sh rin

ka ge

M is ru

Di m en

si on

Cr ac

Ha rd

m et

table 3. Process variables and their significance

Process variable Significance

Amount of green sand

amount of green sand, if incorrect, will make incorrect permeability in the mixed sand, which is a determining factor for the formation of the blowholes gas porosity. also, the consistency of the mold will not be satisfac- tory. the grain size of green sand is different than the reclaimed sand. Hence, an incorrect mixture of the two will cause a difference in inter-grain size, which will inhibit the free flow of entrapped gas and cause blowholes.

Amount of reclaimed sand

reducing binders and catalyst is possible in reclaimed sand. also, better dimensional control of castings can be made while reducing large variations in impurity levels and screen analysis. an incorrect mixture of the two will cause a difference in inter-grain size, which will inhibit the free flow of trapped gas and cause blow- holes.

Amount of additives Bentonite is the additive for binding sand, the amount of which is the strength-determining factor for mold/ core. this will affect consistency and strength in the pattern during pouring of hot metal and also affect the dimensionality of the finished products.

Amount of water too much will cause blowholes, and too little will deteriorate mold/core strength significantly after baking. amount of water is directly proportional to forming water vapor inside the mold when hot metal is poured. this water vapor can get trapped and form gas porosity.

Size and age of sieve

Sieve is used to filter out the oversized grains from the sand. this filtration depends on the size of the sieve. Hence, size and age of the sieve becomes important. as the sieve ages, its dimensional features are affect- ed. incorrect size of sand grains results in an incorrect size of mould, which will affect the intergrain distance and cause blowholes.

Pan mixing time the determining factor for the right mix of ingredients. Less time in pan making will result in heterogeneous mixing and, therefore, inferior mold of sand.

Pattern age as the patterns age, the dimensionality of the patterns lose their actual values, and the tapering increases. this causes the internal features to break off while the pattern is taken out after ramming.

Gating cross-section cross-section of the main entry of molten iron that is poured into the mold. while pouring, if the cross-sec- tion is not appropriate, the molten iron may not flow freely, causing defects such as shrinkage or misruns.

Cupola fan speed any decrease or stoppage of the cupola fan will result in low temperature of the molten iron, which will result in defects such as a misrun.

Temperature of molten metal

the accurate control of liquid metal temperature is important in the continuous and large-scale production of a wide variety of castings to reduce casting rejects and scrap.

Pouring time and temperature

pouring time has an indirect relationship with temperature of the molten iron poured into the mold for cast- ing. Lower pouring time will solidify the iron much faster and avoid the shrinkage defect.

Cooling time one of the primary factors for casting. Shorter cooling time will chill the surface, which is not intended by the customer.

12 I a u g u s t 2 0 1 2 I W W W . a s Q . o r g

M o l d i n g a S o l u t i o n

pass the air out of the mold. These have tremendous effect on the casing because it can seriously compro- mise the casting consistency and significantly increas- ing the probability of the casing buckling, which is a serious performance parameter.

Gas porosity is similar to blowholes, but they are very small, smooth bubbles18 that are usually difficult to detect if contained in the subsurface level. This, too, causes the same effect as a blowhole, but sometimes its effects are even more damaging because of its forma- tion in the deep region of castings.

Shrinkage. This is a property of all metals and alloys because they decrease in volume after solidifying and cooling. Two basic types of shrinkage are found, detected and distinguished: linear and volume. The defect mentioned as shrinkage was not the natural property of the metal or alloy, but rather was the shrinkage cavity formed due to expansion, and shrink- ing was the defect.

Three stages of shrinkage and expansion of the gray cast iron due to the phase changes19—coupled with intricate designs of castings in certain areas of a casting—made the casting cavity defective. The defect shrinkage is the cavity formed due to the linear and

volume expansion and contraction. Misrun. This can result from an incomplete union

of two streams of metal that has lost the required fluidity before completely filling the mold cavity. This type of defect occurs when overheated metal is poured into the mold, sand is excessively moist or the molten iron travels through thin cross-sections of castings.

2. Identifying process variables. After studying the process, along with its significance in the foundry oper- ations, the critical process variables that might cause the occurrence of vital few defects were determined. The variables are listed in Table 3.

3. Completing RCA and FMEA. The two key quan- titative techniques used were RCA and FMEA. The RCA for the identified vital few defects was carried out using a cause and effect diagram to identify all possible causes responsible for the generation of these defects (Figure 3). Based on these causes, an FMEA was per- formed to identify the process variables (measurable) responsible, using risk priority number (RPN) from FMEA tables (Table 4, p. 14) so improvement actions could be systematically planned.

The RPN values were based on historical data and domain knowledge. Historical data were used to

figure 3. cause and effect diagram for shrinkage

Shrinkage

Low �uidity

Incorrect phosphorous

ratio

Incorrect amount of phosphorous

Scrap proportion not OK Scrap inspection

defective

Poor casting design

Poor pattern design

Wrong methoding Pattern conditions bad

Pattern old

Improper maintenance

Incorrect gating of mold

Gate cross- section wrong

Inaccurate study of pattern features

Rate of cooling

Temperature gradient

Environment temperature

Sand coverage thickness in mold

s i x s i g m a f o r u m m a g a z i n e I a u g u s t 2 0 1 2 I 13

M o l d i n g a S o l u t i o n

determine the probability of occurrence for each of the parameters. The values of severity were based on the organization’s voice of the customer input.

The following elaborates on the potential causes of the three defects identified from the analyses:

1. Blowhole/gas porosity. Cupola fan not working: Any discrepancy in the

cupola fan operation affects the coal burning, which affects the temperature of the molten iron, which directly affects the castings and can cause blowhole.

Amount of coal: If not enough coal is added, it will not generate the amount of heat (exothermic reac- tion) required. Again, there will be a fall in tempera- ture.

Untrained labor: The ramming of sand is done man- ually, which means hard ramming. This will make the sand inside the mold highly compact, and it will pre- vent air and water vapor from escaping. This trapped air and water vapor will cause blowholes.

Wrong grain size: If the grain size is too small, the intergrain space is too small for the air created to escape.

Wrong sieve: The same reason as wrong grain size. The sieve is used to filter out the small-sized grains out of the sand, which makes it difficult for the air that’s created to escape.

Humid environment: This condition lowers the rate of water evaporation from the mold while it is in the baking stage. Too much water in the mold will cre- ate a high amount of water vapor, which could cause blowholes.

Low drying time: Again, a high amount of water will create water vapor, which may be trapped inside and cause a blowhole.

Improper vent design: Improper design, few vents or blocked vents will prevent the trapped air or water vapor from escaping, causing blowholes.

2. Shrinkage. Inaccurate study of pattern features: When hot mol-

ten iron is poured into the mold, the volume is higher than in its solidified state. So when the hot metal solidi- fies, it goes through a process of contraction. When the liquid metal inside contracts, it needs more liquid metal to fill the gap created between liquid metal and the mold due to contraction. But sometimes this extra metal required may not be available because of the low area of cross-section that is designed. This may lead to shrinkage.

Scrap proportion not OK: Scrap is the main source of other required elements, such as silicon, phospho- rus and sulphur. If the proportion of scrap is not cal- culated correctly, the required percentage of the ele- ments is off, as well as the amount of scrap. This leads to different unwanted properties of the liquid metal, especially the phosphorous, which gives the required fluidity to molten iron. This fluidity is required for the material to flow freely and avoid shrinkage.

Scrap inspection defective: Scrap chemical inspec- tion is done to check for the percentage of differ- ent chemicals. This result, if defective, will create the wrong mix of chemicals in the hot metal.

table 4. FMea for the shrinkage effect

Process Potential

failure effect S e ve

ri ty

Potential causes

O c c u rr

e n c e

Current prevention

Current detection

D e te

c ti

o n

R P

Recommended action

pattern making

incorrect gating

5 inaccurate study of features of pattern

6 n/a n/a 3 90 introduce planning for casting.

charging Low fluidity 7 Scrap proportion not oK

3 n/a n/a 5 105 Measure and pour scrap proportion.

charging Low fluidity 7 Scrap inspection defective

4 n/a n/a 9 252 discuss possible solution with vendor.

pattern making

poor casting design

10 poor methoding 1 n/a n/a 8 80 analyze failures due to method and take corrective action.

pattern making

poor casting design

9 patterns too old 2 n/a n/a 4 72 replace old pattern.

pattern making

poor casting design

3 improper maintenance

5 n/a n/a 2 30 revamp maintenance department.

fMea = failure mode and effects analysis rpn = risk priority number

14 I a u g u s t 2 0 1 2 I W W W . a s Q . o r g

M o l d i n g a S o l u t i o n

s i x s i g m a f o r u m m a g a z i n e I a u g u s t 2 0 1 2 I 15

Poor methoding: Methoding is the planning of the mold and the determination of how the metal will flow inside the mold. If methoding is incorrect, the flow of metal might get constricted, resulting in shrinkage.

3. Misrun. Inaccurate study of pattern features: When hot mol-

ten iron is poured into the mold, the volume is higher than it is in its solidified state. During solidification, more liquid metal is needed to fill the gap created between liquid metal and the mold due to contraction. But sometimes this extra metal required cannot flow to the area where it is required. This creates a misrun in castings.

Improper timing of pouring: If hot metal isn’t poured into the mold in a specified time limit, the metal flowing into the mold may solidify before reach- ing all the places, resulting in a misrun.

Incorrect measurement for centering or a low cross- section thickness away from gate: If the core is not set exactly in the center of the mold, the thickness on either side of the core becomes unequal. This will lead to the constriction of molten metal flowing through the unequal gaps and result in unequal pressure head. Places where it’s less thick will cool faster and not allow molten metal to flow further. This will result in a misrun.

Cupola blower not running properly: The cupola blower distributes the required oxygen for the coal to burn inside the cupola so the hot metal reaches the required temperature. If this blower does not work properly, the temperature of the cupola will fall and cause less fluidity of the material. This will result in a misrun.

From the FMEA, some of the significant processes identified for improvement were better pattern design and maintenance, training, gauge centering, monitor- ing the cupola, and adding the proper proportion of input materials, coal and scrap.

Improve phase: Table 5 shows the action plan for process improvement that was devised in terms of individual activities to be carried out in a systematic manner to prevent the generation of vital defects. After implementing the improvement plans, fresh data were collected.

Of the 525 defect opportunities, two defects were found: one for blowhole and one for shrinkage. From the collected data, DPMO and the improved sigma level were obtained as 3,810 and 4.166, respec- tively.

Control phase: Standard operating practices (SOP) were prepared for each of the six processes: sand

table 5. action plan for improvement

Vital defect

Action plan

Blowhole/ gas

porosity

• check fan speed every 15 minutes during operation.

• check coal amount before adding. • check sieve regularly. • rectify baking time according to atmo-

spheric temperature and humidity. • unload sand mix from chamber after

drying time ended. • check vent before pouring.

Shrinkage • introduce planning for casting. • Measure scrap proportion and pour. • analyze failures due to methoding and

take actions. • replace old patterns based on analysis.

Misrun • discuss with customers about possible changes of riser/runner position.

• train workers in defined skill set and prepare skill sets.

• develop gauge for centering. • check fan speed every 15 minutes

during operation. • Measure scrap proportion and pour. • Make a detailed scrap inspection.

table 6. standard operating procedure for sand mixing in mixing area

Process variables

Activities

Issue

issue green sand, reclaimed sand and bentonite (additive) from stock.

Instruments/equipment

Sieve size Sieve sand. discard oversized grains. take proportionate volumes of sand and additive.

Process

Mix by volume Dry mixture Core sand Mold sand

green sand 5 parts 4 parts

reclaimed sand 5 parts 6 parts

yellow clay

Bentonite ¾ part 1¼ part

Final mixture

Mix by volume water 4 parts 4 parts

Machine/ equipment

pan-making machine.

time Start pan-making machine. note time. end at stipulated time.

Visually inspect the mixed sand. record, quantify and date.

M o l d i n g a S o l u t i o n

mixing, mold making, core making, cupola operation, pouring, and shake out and fettling operation. An SOP is shown in Table 6 (p. 15) as an example of the control system implementation. The feedback system of the six processes was designed to control the input materials and process.

Lessons learned

This quality improvement journey was launched to improve product performance, and Six Sigma accom- plished that with castings at the foundry. The monthly vendor evaluation report received by management from the foundry’s customers showed a marked improvement in quality. In other areas such as delivery and cost, the customers didn’t seem to notice much difference.

As expected, certain agility also was observed in the manufacturing system as a result of reduced rework. Other lessons learned from the analysis included:

• To produce a good quality product and retain its performance over time, the minimum require- ment is to reduce the variation in the input to the process and the process itself.

• Correlation between output characteristics and input parameters must be statistically derived and based on actual data.

• Close control on input rather than output vari- ables is required. More time and effort must be spent in this area rather than controlling output.

• Significant improvement in productivity is pos- sible by statistically controlling the process.

More improvement opportunities

Enhancing the performance of your product is possible through the use of Six Sigma. This foundry looking to improve the castings of a gear box benefited from:

1. Defining product performance and subfunctions linked to casting defects.

2. Securing effective management involvement and commitment.

3. Linking project selection to business goals and product performance.

4. Implementing employee training and fostering teamwork.

5. Tracking and monitoring progress. Further improvement potential was found in areas

such as changing in the foundry’s layout for better material movement, computerizing process control to eliminate human errors, and controlling the chemical composition of melt through a humidity-controlled

plant environment and numerical-controlled chemical analyzer.

REFERENCES

1. Christoph H. Loch and Stylianos Kavadias, Handbook of New Product Develop- ment Management, Butterworth-Heinemann, 2008.

2. Jack ReVelle, Manufacturing Handbook of Best Practices: An Innovation, Produc- tivity and Quality Focus, CRC Press, 2002.

3. Prasun Das, “Focus on Improvement of Casting Quality Using Statistical Principles,” Indian Foundry Journal, Vol. 55, No. 10, 2009, pp. 38-50.

4. Ronald D. Snee, “Statistical Thinking and Its Contribution to Total Quality, American Statistician, Vol. 44, No. 2, 1990, pp. 116-121.

5. Roger Hoerl, “One Perspective on the Future of Six Sigma,” International Journal of Six Sigma and Competitive Advantage, Vol. 1, No. 1, 2004, pp. 112- 119.

6. Forrest W. Breyfogle III, Implementing Six Sigma: Smarter Solutions Using Sta- tistical Methods, John Wiley and Sons, 1999.

7. Mikel Harry and Richard Schroeder, Six Sigma: The Breakthrough Strategy Revolutionizing the World’s Top Corporation, Doubleday, 2000.

8. Peter S. Pande, Robert P. Neuman and Roland R. Cavanagh, The Six Sigma Way: How GE and Other Top Companies Are Honing Their Performance, McGraw- Hill, 2000.

9. Ronald D. Snee, “Impact of Six Sigma on Quality Engineering,” Quality Engineering, Vol. 12, No. 3, 2000, pp. 9-14.

10. Jiju Antony and Ricardo Banuelas, “Key Ingredients for the Effective Implementation of Six Sigma Program,” Measuring Business Excellence, Vol. 6, No. 4, pp. 20-27.

11. Ronald D. Snee and William F. Rodebaugh, “Project Selection Process,” Quality Progress, September 2002, pp. 78-90

12. Ronald D. Snee and Roger Hoerl, Leading Six Sigma Companies, FT Pren- tice-Hall, 2003.

13. Prasun Das, “Reduction in Delay in Procurement of Materials Using Six Sigma Philosophy,” Total Quality Management and Business Excellence, Vol. 16, No. 5, 2005, pp. 645-656.

14. Prasun Das, “A Statistical Thinking Perspective Through Six Sigma Phi- losophy for Improving Healthcare Services,” Vision-Journal of Management and Allied Sciences, Vol. 2, No. 4, 2006, pp. 38-47.

15. Nandini Das, Susanta Gauri and Prasun Das, “Six Sigma Principles in Mar- keting: An Application,” International Journal of Six Sigma and Competitive Advantage, Vol. 2, No. 3, 2006, pp. 243-262.

16. Prasun Das, Shirshendu Roy and Jiju Antony, “An Application of Six Sig- ma Methodology to Reduce Lot-to-Lot Shade Variation of Linen Fabrics,” Journal of Industrial Textiles, Vol. 36, No. 3, 2007, pp. 227-251.

17. Prasun Das and Shirshendu Roy, “Reducing Defect Generations in E- learning Business: A Six Sigma Approach,” International Journal of Six Sigma and Competitive Advantage, Vol. 5, No. 2, 2009, pp. 113-126.

18. Jonathan S. Colton, “Casting Defects and Design Issues,” presentation, 2009, Georgia Institute of Technology, www.me.gatech.edu/jonathan. colton/me4210/castdefect.pdf.

19. Ibid.

BIBLIOGRAPHY

Monroe Charles A., and Christopher Beckermann, “Development of a Hot Tear Indicator for Steel Castings,” Materials Science and Engineering A 413- 414, 2005, pp. 30-36.

Montgomery, Douglas C., Introduction to Statistical Quality Control, sixth edi- tion, John Wiley & Sons Inc., 2008.