ENGR 3215

SPRING 2019
PROFESSOR VAN SCOTER
LECTURE 8

1
TOPICS TO REVIEW TO PREPARE FOR MIDTERM

• CONTROL CHARTS & CHAPTER 4 INFERENCES 4.1.1, 4.2, 4.3, 4.4,

4.6
• CHAPTER 6 OF TEXT (WITH OTHER REFERENCES)
• CHAPTER 6.2 & 8.3 (WITH OTHER REFERENCES)
• CHAPTER 8.1 – 8.5 (WITH OTHER REFERENCES)
2
CHAPTER 6 LEARNING OBJECTIVES

• KNOW HOW TO DESIGN VARIABLE CONTROL CHARTS

• KNOW HOW TO SET UP AND USE X BAR AND R CONTROL
CHARTS
• KNOW HOW TO SET UP AND USE X BAR AND S CONTROL
CHARTS
• KNOW HOW TO INTERPRET PATTERNS ON VARIABLE CONTROL
CHARTS
• UNDERSTAND THE STATISTICAL BASIS OF CONTROL CHARTS FOR
VARIABLES 3
 LEARNING OBJECTIVES FROM CHAPTER 8

• INVESTIGATE AND ANALYZE PROCESS CAPABILITY

• UNDERSTAND THE DIFFERENCE BETWEEN PROCESS CAPABILITY
AND PROCESS POTENTIAL

• CALCULATE AND INTERPRET PROCESS CAPABILITY RATIOS

• CALCULATE CONFIDENCE INTERVALS ON PROCESS
CAPABILITY RATIOS
4
CHAPTER 13 LEARNING OBJECTIVES

• EXPLAIN HOW DESIGNED EXPERIMENTS CAN BE USED TO IMPROVE

PRODUCT DESIGN TO IMPROVE PROCESS PERFORMANCE

• EXPLAIN HOW DESIGNED EXPERIMENTS CAN BE USED TO REDUCE THE

CYCLE TIME REQUIRED TO DEVELOP NEW PRODUCTS AND PROCESSES

• UNDERSTAND HOW MAIN EFFECTS AND INTERACTIONS CAN BE

ESTIMATED

• UNDERSTAND THE FACTORIAL DESIGN CONCEPT

5
DATA CLASSIFICATION DECISION TREE

Taken from https://www.hqsc.govt.nz/assets/General-NEMR-files-images-/Scientific-symposium-John-Krueger-Nov-2016.pdf; 2 Feb 2019

OVERVIEW OF CONTROL CHARTS

Taken from https://www.isixsigma.com/tools-templates/control-charts/a-guide-to-control-charts/; 2 Feb 2019

CONTROL CHARTS
• A GRAPH USED TO STUDY HOW A PROCESS CHANGES OVER TIME
• DATA ARE PLOTTED IN TIME ORDER
• A CONTROL CHART ALWAYS HAS A CENTRAL LINE FOR THE AVERAGE, AN
UPPER LINE FOR THE UPPER CONTROL LIMIT AND A LOWER LINE FOR THE LOWER
CONTROL LIMIT, DETERMINED FROM HISTORICAL DATA

• COMPARING CURRENT DATA TO THESE LINES, CONCLUSIONS ABOUT WHETHER

THE PROCESS VARIATION IS CONSISTENT (IN CONTROL) OR IS UNPREDICTABLE
(OUT OF CONTROL, AFFECTED BY SPECIAL CAUSES OF VARIATION)

Taken from https://asq.org/quality-resources/control-chart; 2 Feb 2019

WHEN TO USE A CONTROL CHART
• WHEN CONTROLLING ONGOING PROCESSES BY FINDING AND CORRECTING
PROBLEMS AS THEY OCCUR

• WHEN PREDICTING THE EXPECTED RANGE OF OUTCOMES FOR A PROCESS

• WHEN DETERMINING WHETHER A PROCESS IS STABLE (IN STATISTICAL CONTROL)

• WHEN ANALYZING PATTERNS OF PROCESS VARIATION FROM SPECIAL CAUSES

(NON-ROUTINE EVENTS) OR COMMON CAUSES (BUILT INTO THE PROCESS)
• WHEN DETERMINING WHETHER YOUR QUALITY IMPROVEMENT PROJECT SHOULD
AIM TO PREVENT SPECIFIC PROBLEMS OR TO MAKE FUNDAMENTAL CHANGES TO
THE PROCESS
9

Taken from https://asq.org/quality-resources/control-chart; 2 Feb 2019

CONTROL CHART BASIC PROCEDURE

• CHOOSE THE APPROPRIATE CONTROL CHART FOR YOUR DATA

• DETERMINE THE APPROPRIATE TIME PERIOD FOR COLLECTING AND PLOTTING
DATA

• COLLECT DATA, CONSTRUCT YOUR CHART AND ANALYZE THE DATA

• LOOK FOR “OUT-OF-CONTROL SIGNALS” ON THE CONTROL CHART. WHEN ONE

IS IDENTIFIED, MARK IT ON THE CHART AND INVESTIGATE THE CAUSE

• DOCUMENT HOW YOU INVESTIGATED, WHAT YOU LEARNED, THE CAUSE AND
HOW IT WAS COLLECTED

10

Taken from https://asq.org/quality-resources/control-chart; 2 Feb 2019

CONTROL CHART VARIATION OVER TIME

• CONTROL CHARTS CHARACTERIZE PROCESS CHANGES OVER TIME

• TWO TYPES OF VARIATION OVER TIME:
• COMMON OR INHERENT CAUSE VARIATION – BACKGROUND NOISE DUE TO
INHERENT ELEMENTS OF THE PROCESS – VARIATION THAT IS EXPECTED AND
PREDICTABLE IN A PROCESS (FLUCTUATIONS IN TEMPERATURE, DIFFERENT
OPERATORS, DIFFERENT EQUIPMENT)

• SPECIAL CAUSE VARIATION – USUALLY DUE TO SOME SPECIAL OR ASSIGNABLE

CAUSE, NOT EXPECTED OR PREDICTABLE

11

Taken from nanohub.org/resources/26832/download/SPC_PK02_PG.pdf; 2 Feb 2019

VARIATION IN CONTROL CHARTS

• TWO TYPES OF VARIATION:

• CONTROLLED– PREDICTABLE IN A PROBABILISTIC SENSE AND CAN BE REPRESENTED
BY A CONSISTENT PATTERN THAT CAN BE ATTRIBUTED TO COMMON CAUSE OR
BACKGROUND NOISE (FLUCTUATIONS IN TEMPERATURE, DIFFERENT OPERATORS,
DIFFERENT EQUIPMENT)

• UNCONTROLLED– THE VARIATION IS UNPREDICTABLE, DUE TO A SPECIAL CAUSE.

THE SPECIAL CAUSE NEEDS TO BE IDENTIFIED AND CORRECTED IN ORDER TO
REMOVE THIS UNCONTROLLABLE VARIATION FROM A PROCESS (OUT-OF-CONTROL
– OOC)
12

Taken from nanohub.org/resources/26832/download/SPC_PK02_PG.pdf; 2 Feb 2019

I-MR CONTROL CHART

• I-CHART IS USED TO DETECT TRENDS AND SHIFTS IN THE DATA, AND THUS
IN THE PROCESS (DATA MUST BE ENTERED IN SEQUENCE)

• MR CHART SHOWS SHORT-TERM VARIABILITY IN A PROCESS – AN

ASSESSMENT OF THE STABILITY OF PROCESS VARIATION

• MOVING RANGE IS THE DIFFERENCE BETWEEN CONSECUTIVE

OBSERVATIONS, WHICH IS EXPECTED TO BE PREDICTABLE

• POINTS OUTSIDE THE CONTROL LIMITS INDICATE INSTABILITY, IF THERE ARE

ANY OUT OF CONTROL POINTS, THE SPECIAL CAUSES MUST BE ELIMINATED

13

Taken from https://www.isixsigma.com/tools-templates/control-charts/a-guide-to-control-charts/; 2 Feb 2019

I-MR CONTROL CHART

14

Taken from https://www.isixsigma.com/tools-templates/control-charts/a-guide-to-control-charts/; 2 Feb 2019

I-MR CONTROL CHART – BEST USED WHEN:

1. THE NATURAL SUBGROUP SIZE IS UNKNOWN

2. THE INTEGRITY OF THE DATA PREVENTS A CLEAR PICTURE OF A LOGICAL

SUBGROUP

3. THE DATA IS SCARCE (THEREFORE SUBGROUPING IS NOT YET

PRACTICAL)

4. THE NATURAL SUBGROUP NEEDING TO BE ASSESSED IS NOT YET DEFINED

15

Taken from https://www.isixsigma.com/tools-templates/control-charts/a-guide-to-control-charts/; 2 Feb 2019

XBAR AND RANGE (XBAR-R) CONTROL CHARTS
• ANOTHER COMMON CONTROL CHART FOR CONTINUOUS DATA IS THE
XBAR AND RANGE (XBAR-R) CHART
• THE XBAR-R CHART IS ALSO TWO CHARTS USED IN TANDEM, USED WHEN
YOU CAN RATIONALLY COLLECT MEASUREMENTS IN SUBGROUPS OF
BETWEEN TWO AND 10 OBSERVATIONS
• EACH SUBGROUP IS A SNAPSHOT OF THE PROCESS AT A GIVEN POINT IN
TIME

• THE CHART’S X-AXES ARE TIME BASED, SO THE CHART SHOWS THE
HISTORY OF THE PROCESS, SO IT IS IMPORTANT THAT THE DATA IS IN TIME-
ORDER 16

Taken from https://www.isixsigma.com/tools-templates/control-charts/a-guide-to-control-charts/; 2 Feb 2019

XBAR AND RANGE (XBAR-R) CONTROL CHART

Xbar chart

Xbar-R chart

17

Taken from https://www.isixsigma.com/tools-templates/control-charts/a-guide-to-control-charts/; 2 Feb 2019

XBAR AND RANGE (XBAR-R) CONTROL CHART

• THE XBAR CHART IS USED TO EVALUATE CONSISTENCY

OF PROCESS
AVERAGES BY PLOTTING THE AVERAGE OF EACH SUBGROUP
• IT IS EFFICIENT AT DETECTING RELATIVELY LARGE SHIFTS (TYPICALLY
1.5SIGMA OR LARGER) IN THE PROCESS AVERAGE
• THE R CHART PLOTS THE RANGES OF EACH SUBGROUP
• THE R CHART IS USED TO EVALUATE THE CONSISTENCY OF PROCESS
VARIATION
• LOOK AT THE R CHART FIRST, IF THE R CHART IS OUT OF CONTROL, THEN
THE CONTROL LIMITS ON THE XBAR CHART ARE MEANINGLESS
18

Taken from https://www.isixsigma.com/tools-templates/control-charts/a-guide-to-control-charts/; 2 Feb 2019

XBAR AND RANGE (XBAR-R) CONTROL CHART RESULTS

• AN IN-CONTROL-RANGE, BUT OUT-OF-CONTROL MEAN,

COULD INDICATE A SHIFT IN PROCESS MEAN

• AN OUT-OF-CONTROL (OOC) RANGE AND AN IN

CONTROL MEAN COULD MEAN AN INCREASE IN
VARIABILITY

19

Taken from http://nanohub.org/resources/26832/download/SPC_PK02_PG.pdf/; 2 Feb 2019

NORMAL DISTRIBUTION PROPERTIES
• CURVE IS SYMMETRICAL ABOUT ITS MEAN – IT IS THE APEX OF THE CURVE
• THE WIDTH OF THE BELL-SHAPED CURVE INCREASES AS THE STANDARD DEVIATION
INCREASES AND DECREASES AS THE STANDARD DEVIATION DECREASES

• APPROXIMATELY 68% OF THE DATA IS WITHIN ONE STANDARD DEVIATION OF THE MEAN,
95% WITHIN TWO, AND 99% WITHIN THREE STANDARD DEVIATIONS OF THE MEAN
• THE CENTRAL LIMIT THEOREM IMPLIES THAT THE SUM OF Ƞ INDEPENDENTLY DISTRIBUTED
RANDOM VARIABLES IS APPROXIMATELY NORMAL, REGARDLESS OF THE DISTRIBUTION OF
THE INDEPENDENT VARIABLES. THE APPROXIMATION IMPROVES AS Ƞ INCREASES

• IN GENERAL, IF XI ARE IDENTICALLY DISTRIBUTED, AND THE DISTRIBUTION DOES NOT DEPART
RADICALLY FROM THE NORMAL, THEN THE CENTRAL LIMIT THEOREM WORKS QUITE WELL,
AND THESE CONDITIONS ARE FREQUENTLY MET FOR QUALITY ENGINEERING PROBLEMS
20

Taken from http://nanohub.org/resources/26832/download/SPC_PK02_PG.pdf/; 2 Feb 2019

XBAR CONTROL CHARTS

• CONTROL LIMITS SHOW WHAT THE PROCESS IS ACTUALLY DOING (WHILE

SPECIFICATION LIMITS SAY WHAT THE PROCESS SHOULD BE DOING)

• EACH POINT ON A VARIABLES CONTROL CHART IS USUALLY MADE UP OF A SET OF

MEASUREMENTS – FOR TWO REASONS

1. IT RESULTS IN A PREDICTABLE NORMAL (BELL-SHAPED_ DISTRIBUTION FOR THE

OVERALL CHART, DUE TO THE CENTRAL LIMIT THEOREM

2. THIS WILL RESULT IN TIGHTER CONTROL LIMITS BECAUSE THE AVERAGING EFFECT IN
EACH GROUP SMOOTHS OUT INDIVIDUAL HIGH AND LOW MEASUREMENTS, RESULTING
IN A CONTROL CHART THAT CAN DETECT SMALLER CHANGES IN THE PROCESS THAN
ONE WHICH PLOTS ONE POINT FOR EACH MEASUREMENT

21

Taken from http://www.syque.com/quality_tools/toolbook/Control/how.htm/; 2 Feb 2019

XBAR AND XBAR-R CONTROL CHARTS
• TWO CONTROL CHARTS MUST BE DRAWN WHEN TRACKING
VARIABLES

• JUST MEASURING THE AVERAGE OF SUBGROUPS COULD RESULT IN

SIGNIFICANT VARIATION WITHIN THE SUBGROUPS BEING MISSED

• THE CONTROL LIMITS FOR THE RANGE CHART ARE NOT SYMMETRICAL
ABOUT THE CENTER AVERAGE LINE

• THE DISTRIBUTION FOR THE RANGE CHART IS SKEWED, AND IT IS ALSO

COMMON FOR THE LOWER CONTROL LIMIT OF A RANGE CHART TO BE
ON THE ZERO LINE
22

Taken from http://www.syque.com/quality_tools/toolbook/Control/how.htm/; 2 Feb 2019

UNDERSTANDING VARIATION IN SUBGROUPS

23

Taken from http://www.syque.com/quality_tools/toolbook/Control/how.htm/; 2 Feb 2019

UNDERSTANDING VARIATION IN CONTROL CHARTS

• WHEN A PROCESS IS STABLE AND IN CONTROL,

IT DISPLAYS COMMON CAUSE VARIATION THAT
IS INHERENT IN THE PROCESS, I.E., BASED ON
PAST EXPERIENCE IT CAN PREDICT HOW THE
PROCESS WILL VARY IN THE FUTURE.

24

Taken from: https://www.isixsigma.com/tools-templates/control-charts/a-guide-to-control-charts/; 2/12/19

STATISTICAL BASIS OF CHARTS

• SUPPOSE A QUALITY CHARACTERISTIC IS NORMALLY DISTRIBUTED

WITH MEAN  (MU) AND STANDARD DEVIATION  (SIGMA),
WHERE BOTH SIGMA AND MU ARE KNOWN.

• IF X1 + X2 + …. +X𝒎 IS A SAMPLE OF SIZE 𝒎, THEN THE

𝚺𝒙𝒊
SAMPLE AVERAGE, XBAR =
𝒎
• XBAR IS NORMALLY DISTRIBUTED WITH MEAN  AND STANDARD
DEVIATION OF XBAR = /𝒏1/2
25
STATISTICAL INFERENCE

• THE NORMAL DISTRIBUTION HAS A SPECIAL NOTATION: X –

N(, 2)
• IT IMPLIES THAT X IS NORMALLY DISTRIBUTED WITH MEAN 
AND VARIANCE 2

• IF X1, X2, . . . ,𝑿𝒏 IS A RANDOM SAMPLE OF SIZE 𝒏 FROM

THIS PROCESS, THEN THE DISTRIBUTION OF THE MEAN XBAR
IS N(, 2/𝒏)
26
STATISTICAL BASIS OF CHARTS

• FURTHERMORE, THE PROBABILITY IS 1-  THAT ANY SAMPLE MEAN

WILL FALL BETWEEN

+Z/2 X = +Z/2 /𝒏1/2 AND -Z/2 X = -Z/2 /𝒏1/2

(6.1)
IT IS CUSTOMARY TO REPLACE Z/2 BY 3, SO THAT THREE-SIGMA
LIMITS ARE EMPLOYED

IF A SAMPLE MEAN FALLS OUTSIDE OF THESE LIMITS, IT IS AN

INDICATION THAT THE PROCESS MEAN IS NO LONGER EQUAL TO  27
GOING OUT OF CONTROL

28
IN PRACTICE, WE USUALLY DON’T KNOW  AND 

• XBAR = 𝚺𝒎𝒙𝒊 /𝒎
• R = XMAX – XMIN
• RBAR = 𝚺𝒎𝑹𝒊 /𝒎
29
CONTROL LIMITS FOR X CHART

30
ESTIMATORS FOR SIGMA AND MU

31
CONTROL LIMITS FOR THE X CHART – PROCESS
MEAN CHART

32
CONTROL LIMITS FOR R CHART

33
ESTIMATORS FOR SIGMA AND MU

34
THE R CHART – PROCESS VARIABILITY CHART

• CENTER LINE WILL BE RBAR

• WE ASSUME THE STANDARD DEVIATION OF R IS NORMALLY
DISTRIBUTED,

• WE NEED AN ESTIMATE OF R

35
IN PRACTICE, WE USUALLY DON’T KNOW  AND 

• THERE IS A WELL KNOWN RELATIONSHIP BETWEEN THE

RANGE OF A SAMPLE FROM A NORMAL DISTRIBUTION
AND THE STANDARD DEVIATION OF THAT DISTRIBUTION.
• RANDOM VARIABLE W = R /  = RELATIVE RANGE
• THE PARAMETERS OF THE DISTRIBUTION OF W ARE A
FUNCTION OF THE SAMPLE SIZE N
• AN ESTIMATOR OF  IS
36
CONTROL LIMITS FOR R CHART

37
ESTIMATING PROCESS CAPABILITY (CP)

• THE X VALUE GIVES THE MEAN OF OUR INDEPENDENT

VARIABLE

• THE STANDARD DEVIATION IS GIVEN BY

38
EXAMPLE 6.1 FROM TEXT – USED FOR HOMEWORK

39
ESTIMATING PROCESS CAPABILITY (CP)

Xbar+/-variability

40
WHEN TO USE XBAR R VERSUS XBAR S CHARTS

• 50 SUBGROUPS WERE GENERATED WITH 15 SAMPLES IN EACH SUBGROUP

USING A MEAN OF 50 AND STANDARD DEVIATION OF 5. IN TOTAL,
THERE WERE 750 VALUES.

• SIMULATED VALUES USING MEAN = 50 AND STANDARD DEVIATION = 5

• FOR EACH SUBGROUP, OF N=5 TO
15 VALUES, THE RANGE WAS
COMPUTED AND THEN THE AVERAGE RANGE ACROSS ALL 50
SUBGROUPS WAS CALCULATED. THE AVERAGE RANGE WAS THEN
DIVIDED BY THE APPROPRIATE D2 CONSTANT FOR EACH SUBGROUP MADE
UP OF N = 5 TO N = 15 VALUES.
41

Taken from: https://andrewmilivojevich.com/xbar-r-chart-versus-xbar-s-chart/; 2/13/19

STANDARD DEVIATION ESTIMATES
Standard Deviation Estimates Table of Average Standard
Based on the Range Deviation Estimates

42

Taken from: https://andrewmilivojevich.com/xbar-r-chart-versus-xbar-s-chart/; 2/13/19

STANDARD DEVIATION ESTIMATES

43

Taken from: https://andrewmilivojevich.com/xbar-r-chart-versus-xbar-s-chart/; 2/13/19

 WHEN TO USE XBAR R VERSUS XBAR S CHARTS

• THE POPULATION STANDARD DEVIATION IS SHOWN AS A HORIZONTAL LINE

THROUGH A STANDARD DEVIATION AT 5. NOTICE THAT THE STANDARD
DEVIATION, BASED ON THE AVERAGE RANGE, FOR N = 5 TO 11
ESTIMATES THE POPULATION STANDARD DEVIATION WELL. BUT, FOR THE
SAME RANGE OF SUBGROUPS (N = 5 TO 11), THE AVERAGE STANDARD
DEVIATION UNDER-ESTIMATES THE POPULATION STANDARD DEVIATION.

• BASED ON THIS OBSERVATION, WE’RE INCLINED TO BELIEVE THAT THE

AVERAGE RANGE ESTIMATES THE POPULATION STANDARD DEVIATION
WELL BETWEEN N = 5 THROUGH N = 11 VALUES PER SUBGROUP.
44

Taken from: https://andrewmilivojevich.com/xbar-r-chart-versus-xbar-s-chart/; 2/13/19

 WHEN TO USE XBAR R VERSUS XBAR S CHARTS
• BUT WHAT HAPPENS WHEN YOU HAVE N = 12 OR MORE VALUES IN EACH
SUBGROUP? WE CAN SEE THAT THE STANDARD DEVIATION BASED ON THE AVERAGE
RANGE OVER ESTIMATES THE POPULATION STANDARD DEVIATION. IN THIS CASE, THE
CONTROL LIMITS FOR THE XBAR CHART WOULD BE WIDER.

• BUT NOTICE THAT THE AVERAGE STANDARD DEVIATION FOR N = 12 OR MORE

VALUES PER SUBGROUP ESTIMATES THE POPULATION WELL. IN THE CASE, THE
CONTROL LIMITS FOR THE XBAR CHART WOULD BE JUST RIGHT.

• BASED ON THIS SIMULATION, WE WOULD SUGGEST THAT WE USE THE XBAR R

CHART FOR N = 11 OR LESS VALUES PER SUBGROUP. OR, IF WE HAD N = 12 OR
MORE VALUES PER SUBGROUP WE WOULD SUGGEST THE USE THE XBAR S CHART.
45

Taken from: https://andrewmilivojevich.com/xbar-r-chart-versus-xbar-s-chart/; 2/13/19

PROCESS CAPABILITY REFERS TO UNIFORMITY

TWO WAYS TO THINK ABOUT THIS VARIABILITY:

• THE NATURAL OR INHERENT VARIABILITY IN A CRITICAL-TO-
QUALITY CHARACTERISTIC AT A SPECIFIED TIME – THAT IS,
“INSTANTANEOUS” VARIABILITY
• THE VARIABILITY IN A CRITICAL-TO-QUALITY
CHARACTERISTICS OVER TIME
46
PROCESS CAPABILITY ANALYSIS IS USED IN
MULTIPLE PRODUCT CYCLE PHASES
• ASSISTING DEVELOPMENT ACTIVITIES PRIOR TO
MANUFACTURING; SELECTING A PROCESS, SPECIFYING
PERFORMANCE REQUIREMENTS FOR EQUIPMENT; SELECTING
BETWEEN COMPETING SUPPLIERS AND OTHER ASPECTS OF
SUPPLY CHAIN MANAGEMENT
• PREDICTING HOW WELL THE PROCESS WILL HOLD THE
TOLERANCES
• IN ANALYZING THIS VARIABILITY RELATIVE TO PRODUCT
REQUIREMENTS OR SPECIFICATIONS
47
PROCESS CAPABILITY ANALYSIS IS USED IN
MULTIPLE PRODUCT CYCLE PHASES (2)

• IN PLANNING THE SEQUENCE OF PRODUCTION PROCESSES

WHEN THERE IS AN INTERACTIVE EFFECT OF PROCESSES ON
TOLERANCES; ASSISTING MANUFACTURING IN ELIMINATING
OR REDUCING VARIABILITY

• ASSISTING IN ESTABLISHING AN INTERVAL BETWEEN

SAMPLING FOR PROCESS MONITORING
48
PROCESS CAPABILITY ANALYSIS USING A
PROBABILITY PLOT
AT LEAST 100 OR MORE OBSERVATIONS SHOULD BE AVAILABLE FOR THE PLOT TO BE MODERATELY
STABLE SO A REASONABLY RELIABLE ESTIMATE OF PROCESS CAPABILITY MAY BE MAINTAINED.

- THE FOLLOWING STEPS SHOULD BE FOLLOWED PRIOR TO DATA COLLECTION:

1. CHOOSE MACHINE (S) TO BE USED, REPRESENTATIVE OF THE POPULATION. DESIGNED

EXPERIMENTS SHOULD BE USED TO ISOLATE HEAD-TO-HEAD VARIABILITY FOR MULTI
WORKSTATIONS.

2. SELECT THE PROCESS OPERATING CONDITIONS. CAREFULLY DEFINE CONDITIONS – THESE MAY
BE SOURCES OF VARIATION IN CAPABILITY TO UNDERGO FURTHER STUDIES.

3. SELECT A REPRESENTATIVE OPERATOR. IF ESTIMATING OPERATOR VARIABILITY, RANDOMLY

SELECT OPERATOR.

4. CAREFULLY MONITOR DATA-COLLECTION PROCESS AND RECORD TIME ORDER WITH DATA.
49
50
PROCESS CAPABILITY RATIOS

• IT IS FREQUENTLY CONVENIENT TO HAVE A SIMPLE, QUANTITATIVE WAY TO EXPRESS

PROCESS CAPABILITY. ONE WAY TO DO SO IS THROUGH THE PROCESS CAPABILITY
RATIO (PCP) CP, INTRODUCED IN CHAPTER 6.
• PROCESS CAPABILITY RATIOS ARE WIDELY MIS-USED. SIGMA AND MU ARE ALMOST
ALWAYS UNKNOWN AND MUST BE REPLACED BY AN ESTIMATE. ASSUMPTIONS ON THE
NORMALITY OF A DISTRIBUTION, THE SKEWNESS OF A DISTRIBUTION, OR
AUTOCORRELATED DATA ARE INCORRECT.

51
PERCENTAGE OF THE SPECIFICATION BAND USED
UP BY THE PROCESS

From Example 6.1, the hard

bake process uses 83.89 percent
of the specification band.

52
DOES A PROCESS CAPABILITY RATIO ALWAYS
NEED TO BE TWO-SIDED?

• WHAT IF WE WANT A MINIMUM THICKNESS, OR A MINIMUM

STRENGTH? (LOWER SPECIFICATION LIMIT)
• WHAT IF WE WANT A MAXIMUM ELONGATION OR A
MAXIMUM ABSORPTION? (UPPER SPECIFICATION LIMIT) 53
54
55
A MEASURE OF ACTUAL CAPABILITY

56
CONFIDENCE INTERVALS AND PROCESS CAPABILITY
RATIOS

• PROCESS CAPABILITY RATIOS ARE ‘POINT ESTIMATES’ AND

AS SUCH, ARE SUBJECT TO STATISTICAL FLUCTUATIONS

• AN ALTERNATIVE THAT HAS BECOME STANDARD PRACTICE

IS TO REPORT CONFIDENCE LIMITS FOR PROCESS CAPABILITY
RATIOS

57
CONFIDENCE INTERVALS ON PROCESS CAPABILITY
RATIOS

58
59
 Process
Capability
Analysis
using
Control
Charts

60
Since LSL = 200

61
PROCESS CAPABILITY ANALYSIS USING DESIGNED
EXPERIMENTS

• A DESIGNED EXPERIMENT IS A SYSTEMATIC APPROACH TO

VARYING THE INPUT CONTROLLABLE VARIABLES IN THE
PROCESS AND ANALYZING THE EFFECTS OF THESE PROCESS
VARIABLES ON THE OUTPUT.

• A DESIGNED EXPERIMENT IS USED TO ISOLATE AND ESTIMATE

THE SOURCES OF VARIABILITY IN A PROCESS.

62
WHAT IS EXPERIMENTAL DESIGN?

• A DESIGNED EXPERIMENT IS A TEST

OR SERIES OF TESTS IN WHICH
PURPOSEFUL CHANGES ARE MADE TO
THE INPUT VARIABLES (MACHINES,
METHODS OR PEOPLE) OF A PROCESS
SO THAT WE MAY OBSERVE AND
IDENTIFY CORRESPONDING CHANGES
IN THE OUTPUT RESPONSE

63
 OBJECTIVES OF THE DESIGNED EXPERIMENT
1. DETERMINING WHICH VARIABLES ARE MOST
INFLUENTIAL ON THE RESPONSE, Y
2. DETERMINING WHERE TO SET THE INFLUENTIAL
X’S SO THAT Y IS NEAR THE NOMINAL
REQUIREMENT
3. DETERMINING WHERE TO SET THE INFLUENTIAL
X’S SO THAT VARIABILITY IN Y IS SMALL
4. DETERMINING WHERE TO SET THE INFLUENTIAL
X’S SO THAT THE EFFECTS OF THE
UNCONTROLLABLE VARIABLES Z ARE
MINIMIZED
64
 WHEN TO USE DESIGNED EXPERIMENTS

• THEY MAY BE USED EITHER IN

PROCESS DEVELOPMENT PROCESS
OR PROCESS TROUBLESHOOTING
TO IMPROVE PERFORMANCE OR TO
OBTAIN A PROCESS THAT IS ROBUST
OR INSENSITIVE TO EXTERNAL
SOURCES OF VARIABILITY
65
WHY USE DESIGNED EXPERIMENTS

EXPERIMENTAL DESIGN METHODS CAN BE VERY USEFUL IN

ESTABLISHING STATISTICAL CONTROL OF A PROCESS AS WELL AS:
• IMPROVE YIELD
• REDUCE VARIABILITY AND CLOSER CONFORMANCE TO THE
NOMINAL
• REDUCE DEVELOPMENT TIME
• REDUCE OVERALL COSTS
66
WHY USE DESIGNED EXPERIMENTS (2)

EXPERIMENTAL DESIGN METHODS CAN ALSO PLAY A MAJOR ROLE IN

ENGINEERING DESIGN ACTIVITIES, WHERE NEW PRODUCTS ARE
DEVELOPED AND EXISTING ONES IMPROVED –WIDELY USED IN SIX SIGMA
(DFSS) ACTIVITIES. APPLICATIONS OF STATISTICAL EXPERIMENTAL
DESIGN IN ENGINEERING DESIGN INCLUDE:
• EVALUATION AND COMPARISON OF BASIC DESIGN CONFIGURATION
• EVALUATION OF MATERIAL ALTERNATIVES
• DETERMINATION OF KEY PRODUCT DESIGN PARAMETERS THAT IMPACT
PERFORMANCE
67
WHY USE DESIGNED EXPERIMENTS (3)

USE OF EXPERIMENTAL DESIGN IN THESE AREAS CAN RESULT

IN IMPROVED MANUFACTURABILITY OF THE PRODUCT,
ENHANCED FIELD PERFORMANCE AND RELIABILITY, LOWER
PRODUCT COST, AND SHORTER PRODUCT DEVELOPMENT TIME

IN RECENT YEARS, DESIGNED EXPERIMENTS HAVE FOUND

EXTENSIVE APPLICATION IN TRANSACTIONAL AND SERVICE
APPLICATIONS, INCLUDING E-COMMERCE.
68
GUIDELINES FOR DESIGNING AN EXPERIMENT

69
FACTORIAL DESIGNED EXPERIMENTS

• THERE IS CONSIDERABLE SCOPE FOR REDUCING THE NUMBER OF

RESOURCES USED IN RESEARCH BY DESIGNING BETTER EXPERIMENTS

• TO OPTIMIZE THE EXPERIMENTS TO USE THE SMALLEST NUMBER OF

TESTS, ONE METHOD IS TO USE FACTORIAL EXPERIMENT DESIGNS

• THE AIM IS TO MAXIMIZE THE SIGNAL/NOISE RATIO SO THAT THE

NUMBERS OF EXPERIMENTAL SUBJECTS REQUIRED TO DETECT A GIVEN
TREATMENT RESPONSE IS MINIMIZED USING POWER AND SAMPLE SIZE
CALCULATIONS
70

Taken from: https://academic.oup.com/ilarjournal/article/43/4/223/981833; 2/18/19

FACTORIAL DESIGNED EXPERIMENTS (FEDS)

• THE STRATEGY IS TO IDENTIFY THE MOST IMPORTANT FACTORS AND

LEVELS OF THE FACTORS THAT DETERMINE OUTPUT AND THEN TO USE
THESE FACTORS IN NORMAL PRODUCTION

• IN SOME CASES, FEDS HAVE BEEN USED TO GAIN AN

UNDERSTANDING OF THE FACTORS INFLUENCING THE OBSERVED
RESPONSE, RATHER THAN TO OPTIMIZE FUTURE EXPERIMENTS

• IN FEDS EACH TEST WILL CONTRIBUTE TO UNDERSTANDING THE

EFFECT OF ALL OF THE FACTORS UNDER EXPLORATION
71

Taken from: https://academic.oup.com/ilarjournal/article/43/4/223/981833; 2/18/19

REVIEW OF HOMEWORK PROBLEMS

72
HOMEWORK #1
• SOLUTION:
WE ARE GIVEN K = 24 SUBSETS FOR THIS SITUATION AS THERE ARE IN TOTAL OF
24 DATA SETS. SINCE, THERE ARE TOTAL OF FOUR PH MEASUREMENTS BEEN
TAKEN EVERY HOUR, SO FOR EACH SUBSET N = 4.

THE FIRST THING THAT IS REQUIRED TO BE DONE IS TO CALCULATE THE MEAN

AND RANGE FOR EXPERIMENTAL SUBSET. WE USE AVERAGE() EXCEL
FUNCTION TO CALCULATE THE MEAN AND THE FORMULA MAX() – MIN() TO
CALCULATE THE RANGES. ONCE WE ARE DONE CALCULATING THESE VALUES,
WE NEED TO CALCULATE THE AVERAGE, XBARBAR AND AVERAGE RANGE, RA.
THESE ARE SIMPLY THE MEANS OF THE MEAN AND RANGE OF THIS SUBSET.
73
HOMEWORK #1EXPERIMENTAL DATA ONLY

• WE NOW KNOW THAT THE VALUE OF XGA = 6.997. WITH THIS WE CAN
CALCULATE THE UPPER CONTROL LIMIT THAT IS UCL AND LOWER CONTROL
LCL FOR THE XBAR CONTROL TYPE CHART.
LIMIT THAT IS
UCL = XGA + A2 RA UCL=6.99+0.729*0.1217 = 6.99+.0887 = 7.0787
LCL = XGA – A2 RA LCL = 6.99-0.729*0.1217 = 6.99-.0887 = 6.9013

WHEN N = 4, A2 = 0.729. SO, FROM THE FORMULA AND VALUES ABOVE WE

GET,
UCL = 7.0787 AND LCL = 6.9013.
• GREATEST XGA IS 7.08 SMALLEST IS 6.94 => WITHIN CONTROL
74
HOMEWORK #1EXPERIMENTAL DATA ONLY

• WE NOW KNOW THAT THE VALUE OF RA = 0.1217. WITH THIS WE CAN ALSO
CALCULATE THE UPPER CONTROL LIMIT THAT IS UCL AND LOWER CONTROL
LIMIT THAT IS LCL FOR THE RANGE CONTROL TYPE CHART.
UCL = D4 RA AND LCL = D3 RA

FOR N = 4 HERE, D3 = 0, D4 = 2.282. WE NOW GET,

LCL = 0 AND UCL = 2.282*0.1217 = 0.2777
GREATEST RA IS 0.25, SMALLEST IS 0.03 => WITHIN CONTROL

75
HOMEWORK #1- SECOND PART- USED HISTORICAL
DATA TO CREATE CHART PARAMETERS
• SOLUTION:
WE ARE GIVEN K = 24 SUBSETS FOR THIS SITUATION AS THERE ARE IN TOTAL OF
24 DATA SETS. SINCE, THERE ARE TOTAL OF FOUR PH MEASUREMENTS BEEN
TAKEN EVERY HOUR, SO FOR EACH SUBSET N = 4.

THE FIRST THING THAT IS REQUIRED TO BE DONE IS TO CALCULATE THE MEAN

AND STANDARD DEVIATION FOR HISTORICAL SUBSET. WE USE AVERAGE()
EXCEL FUNCTION TO CALCULATE THE MEAN AND THE FORMULA STDEV() TO
CALCULATE THE STANDARD DEVIATIONS. ONCE WE ARE DONE CALCULATING
THESE VALUES, WE NEED TO CALCULATE THE AVERAGE, XBARBAR AND
AVERAGE STANDARD DEVIATION, SA. THESE ARE SIMPLY THE MEANS OF THE
MEAN AND STANDARD DEVIATION OF THE HISTORICAL SUBSET.

76
HOMEWORK #1 HISTORICAL DATA ONLY

• MY THOUGHT WAS THAT YOU WOULD USE THE HISTORICAL DATA TO

CALCULATE VALUES OF XBARBAR AND SBAR THAT WOULD BE USED TO SPECIFY
VALUES FOR THE PROCESS.

• UCL = XBARBAR + A2 RA UCL=7.011+0.729*0.1217 = 7.011+.0887 =

7.099
LCL = XBARBAR – A2 RA LCL = 7.011-0.729*0.1217 = 7.011-.0887 =
6.9223
• CENTERLINE = 7.011
GREATEST XBARBAR IS 7.08 SMALLEST IS 6.94 => WITHIN CONTROL
77
HOMEWORK #1EXPERIMENTAL DATA ONLY

• WE NOW KNOW THAT THE VALUE OF SBAR = .06. WITH THIS WE CAN
CALCULATE THE UPPER CONTROL LIMIT THAT IS UCL AND LOWER CONTROL
LCL FOR THE SBAR CONTROL TYPE CHART.
LIMIT THAT IS
UCL = B4 *SBAR=UCL=2.266*0.06 = 0.1359
LCL = B3 * SBAR =UCL=0*0.06 = 0
CENTER LINE = SBAR = .06
SO, FROM THE FORMULA AND VALUES ABOVE WE GET: UCL = 0.1359 AND
LCL = 0. CENTERLINE = 0.06
• EXPERIMENTAL DATA: GREATEST SBAR IS 0.15 SMALLEST IS 0 => WITHIN
CONTROL
78
HOMEWORK PROBLEMS FOR 2/17

CHAPTER 6
• 6.6
• 6.8
• 6.10
• 6.18
• 6.42
79
• 6.6. THE NET WEIGHT (IN OZ) OF A DRY BLEACH PRODUCT IS TO
• BE MONITORED BY X AND R CONTROL CHARTS USING A SAMPLE SIZE OF N = 5. DATA FOR 20
PRELIMINARY SAMPLES ARE SHOWN IN TABLE 6E.6.

• (A) SET UP X AND R CONTROL CHARTS USING THESE DATA.

• XBARBAR = 16.268; RBAR = 0.475
• DOES THE PROCESS EXHIBIT STATISTICAL CONTROL? YES
• (B) ESTIMATE THE PROCESS MEAN AND STANDARD DEVIATION. 16.268;
SIGMAHAT=RBAR/D2=0.204

• (C) DOES FILL WEIGHT SEEM TO FOLLOW A NORMAL DISTRIBUTION? YES

• (D) IF THE SPECIFICATIONS ARE AT 16.2 ± 0.5, WHAT CONCLUSIONS WOULD YOU DRAW ABOUT
PROCESS CAPABILITY? 16.7 – 15.8 - PROCESS CAN HANDLE ACCORDING TO CONTROL CHART,
BUT CALCULATION SHOWS .

• (E) WHAT FRACTION OF CONTAINERS PRODUCED BY THIS PROCESS IS LIKELY TO BE BELOW THE
LOWER SPECIFICATION LIMIT OF 15.7 OZ?

80
• 6.8 REWORK EXERCISE 6.3 USING THE S CHART.

• THE DATA SHOWN IN TABLE 6E.3 ARE THE DEVIATIONS FROM NOMINAL DIAMETER FOR
HOLES DRILLED IN A CARBON-FIBER COMPOSITE MATERIAL USED IN AEROSPACE
MANUFACTURING.
• THE VALUES REPORTED ARE DEVIATIONS FROM NOMINAL IN TEN-THOUSANDTHS OF AN
INCH.
• (A) SET UP X AND R CHARTS ON THE PROCESS. IS THE PROCESS IN STATISTICAL
CONTROL?
• (B) ESTIMATE THE PROCESS STANDARD DEVIATION USING THE RANGE METHOD. SIGMA
=RBAR/D2 D2=4.918
• (C) IF SPECIFICATIONS ARE AT NOMINAL ±100, WHAT CAN YOU SAY ABOUT THE
CAPABILITY OF THIS PROCESS? CALCULATE THE PCR CP. (USL-LSL)/6SIGMA >1

81
• 6.10 TABLE 6E.7 SHOWS 15 ADDITIONAL SAMPLES FOR THE PIS-TON RING PROCESS (TABLE
6.3), TAKEN AFTER THE INITIAL CONTROL CHARTS WERE ESTABLISHED. PLOT THESE DATA ON THE
X AND R CHART DEVELOPED IN EXERCISE 6.9. IS THE PROCESS IN CONTROL?

• XBAR CHART IS OUT OF CONTROL.

82
• 6.18 CONSIDER THE X AND R CHART THAT YOU ESTABLISHED IN EXERCISE 6.9 FOR THE PISTON
RING PROCESS. SUPPOSE THAT YOU WANT TO CONTINUE CONTROL CHARTING PISTON RING
DIAMETER USING N = 3. WHAT LIMITS WOULD BE USED ON THE X AND R CHART?

• XBARBAR = (SUM(I=1-M)NIXBARI)/SUM(I=1-M)NI (6-30); SHAT (6.31)

• LIMITS - (6.27) & (6.28)
• CONSTANTS A3, B3 AND B4 WILL DEPEND ON THE SAMPLE SIZE USED IN EACH INDIVIDUAL
SUBGROUP

83
• 6.42 CONTROL CHARTS FOR X AND S WITH N = 4 ARE MAINTAINED ON A QUALITY
CHARACTERISTIC. THE PARAMETERS OF THESE CHARTS ARE AS FOLLOWS:

x Chart s Chart
UCL = 201.88 UCL = 2.266
Center line = 200.00 Center line = 1.000
LCL=198.12 LCL = 0

• Both charts exhibit control. Specifications on the quality characteristic are 197.50 and
202.50. What can be said about the ability of the process to produce product that conforms
to specifications?

84
HOMEWORK PROBLEMS FOR 2/21

CHAPTER 6
• 6.4
• 6.15
• 6.16
• 6.33
• 6.55
85
• PROBLEM 6.4
• THE THICKNESS OF A PRINTED CIRCUIT BOARD IS AN IMPORTANT QUALITY
PARAMETER. DATA ON BOARD THICKNESS (IN INCHES) ARE GIVEN IN TABLE
6E.4 FOR 25 SAMPLES OF THREE BOARDS EACH.
• (A) SET UP X AND R CONTROL CHARTS. IS THE PROCESS IN STATISTICAL
CONTROL?

• (B) ESTIMATE THE PROCESS STANDARD DEVIATION.

• (C) WHAT ARE THE LIMITS THAT YOU WOULD EXPECT TO CONTAIN NEARLY
ALL THE PROCESS MEASUREMENTS?

• (D) IF THE SPECIFICATIONS ARE AT 0.0630 IN. ± 0.0015 IN., WHAT IS THE
VALUE OF THE PCR CP?

86
• PROBLEM 6.15
• PARTS MANUFACTURED BY AN INJECTION MOLDING PROCESS ARE SUBJECTED TO A
COMPRESSIVE STRENGTH TEST. TWENTY SAMPLES OF FIVE PARTS EACH ARE COLLECTED, AND
THE COMPRESSIVE STRENGTHS (IN PSI) ARE SHOWN IN TABLE 6E.11.
• (A) ESTABLISH X AND R CONTROL CHARTS FOR COMPRESSIVE STRENGTH USING THESE DATA. IS THE
PROCESS IN STATISTICAL CONTROL?

• (B) AFTER ESTABLISHING THE CONTROL CHARTS IN PART (A), 15 NEW SUBGROUPS WERE COLLECTED
AND THE COMPRESSIVE STRENGTHS ARE SHOWN IN TABLE 6E.12. PLOT THE X AND R VALUES
AGAINST THE CONTROL UNITS FROM PART (A) AND DRAW CONCLUSIONS.

87
• PROBLEM 6.16
• RECONSIDER THE DATA PRESENTED IN EXERCISE 6.15.
(A) REWORK BOTH PARTS (A) AND (B) OF EXERCISE 6.15 USING
THE X AND S CHARTS.
(B) DOES THE S CHART DETECT THE SHIFT IN PROCESS VARIABILITY
MORE QUICKLY THAN THE R CHART DID ORIGINALLY IN PART (B) OF
EXERCISE 6.15?
•

88
• PROBLEM 6.33
• SPECIFICATIONS ON A CIGAR LIGHTER DETENT ARE 0.3220 AND
0.3200 IN. SAMPLES OF SIZE 5 ARE TAKEN EVERY 45 MIN WITH THE
RESULTS SHOWN IN TABLE 6E.17 (MEASURED AS DEVIATIONS FROM
0.3210 IN 0.0001 IN.).
• (A) SET UP AN R CHART AND EXAMINE THE PROCESS FOR
STATISTICAL CONTROL.

• (B) WHAT PARAMETERS WOULD YOU RECOMMEND FOR AN R

CHART FOR ON-LINE CONTROL?

• (C) ESTIMATE THE STANDARD DEVIATION OF THE PROCESS.

• (D) WHAT IS THE PROCESS CAPABILITY? 89
 • PROBLEM 6.55
• THE VISCOSITY OF A POLYMER IS MEASURED HOURLY.
MEASUREMENTS FOR THE LAST TWENTY HOURS ARE SHOWN IN TABLE
6E.21.
(A) DOES VISCOSITY FOLLOW A NORMAL DISTRIBUTION?
• (B) SET UP A CONTROL CHART ON VISCOSITY AND A MOVING
RANGE CHART. DOES THE PROCESS EXHIBIT STATISTICAL CONTROL?

• (C) ESTIMATE THE PROCESS MEAN AND STANDARD DEVIATION.

70 9

60 8
40 45 5
7
50 45 50 8
6
40 50 55 6
5
55 60 0
30 4
60 65 1
20 3 90
2
10
1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 0