0% found this document useful (0 votes)

331 views

CSP Mathematics

This document presents some acceptance quality limit (AOQ) and average outgoing quality limit (AFI) values for the Dodge and Torrey continuous sampling plan 3 (CSP-3). It first provides a mathematical proof of the AOQ formula for CSP-3. It then applies this formula to generate tables of AOQ and AFI values for different sample sizes and acceptance numbers for CSP-3. The purpose is to help quality control inspectors apply CSP-3 by providing key statistical values for it.

Uploaded by

Jairo Castro Maestre

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

331 views

CSP Mathematics

Uploaded by

Jairo Castro Maestre

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 178

' 0

\
AEC RESEARCH AND DEVELOPMENT REPORT
MASTER Y-1585
Engineeri ng and
Equipment

SOME AOQ AND AFI VALUES FOR THE DODGE AND

- v
TORRY CSP-3 CONTINUOUS SAMPLI NG PLAN

R. H. Hull

UNION CARBIDE CORPORATION

NUCLEAR DIVISION
OAK RIDGE Y-12 PLANT

operated for the ATOMIC ENERGY COMMISSION unrler U. S. GOVERNMENT Contract W-7405 eng 26

•
OAK RIDGE Y-12 PLANT
P . O.Bo x Y
OAK RIDGE , TENNESSEE 37830

OISTRIBUTION Of THIS OOCUMfNT IS UNUM~

DISCLAIMER

This report was prepared as an account of work sponsored by an

agency of the United States Government. Neither the United States
Government nor any agency Thereof, nor any of their employees,
makes any warranty, express or implied, or assumes any legal
liability or responsibility for the accuracy, completeness, or
usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately
owned rights. Reference herein to any specific commercial product,
process, or service by trade name, trademark, manufacturer, or
otherwise does not necessarily constitute or imply its endorsement,
recommendation, or favoring by the United States Government or any
agency thereof. The views and opinions of authors expressed herein
do not necessarily state or reflect those of the United States
Government or any agency thereof.
DISCLAIMER

Portions of this document may be illegible in

electronic image products. Images are produced
from the best available original document.
Printed in the United States of America , Available from Clearinghouse for Federal
Scientific and Technical Information, Notional Bureau of Standards,
U.S. Deportment of Commerce, Springfield, Virginia 22151
Price: Printed Copy $3.00; Microfiche $0.65

LEGAL NOTICE - - - - - - - - - - - - - - - -

This report was prepared as an account of Government sponsored work. Neither the United States,
nor the Commission, nor any person acting on behalf of the Commission:
A, Makes any warranty or representation, expressed or imp I ied, with respect to the occuracy,
completeness, or usefulness of the information contained in this report, or that the use of
any information, apparatus, method, or process disclosed in this report may not infringe
privately owned rights; or
B. Assumes any I iobil ities with respect to the use of, or for damages resulting from the use of
any information, apparatus, method, or process disclosed in this report .
As used in the above, "person acting on behalf of the Commission" includes any employee or
contractor of the Commission, or employee of such contractor, to the extent that such employee
or contractor of the Commission, or employee of such contractor prepares, disseminates, or
provides access to, any information pursuant to his employment or contract with the Commission,
or his employment with such contractor.
Date Issued: July 7, 1967 Document Y-1585
~----------

Engineering and Equipment

TID-4500

RC:: $ 3 .C'Jc>; MN
----
UNION CARBIDE CORPORATION
Nuclear Division

Y-12 PLANT

Contract W-7405-eng-26
With the US Atomic.Energy Commission

SOME AOQ AND AFI VALUES FOR THE DODGE AND

TORREY CSP-3 CONTINUOUS SAMPLING PLAN

R. H. Hull, Jr

This report is based on a study by the author as

partial fulfillment of requirements for the degree
of Master of Science in Statistics from The Uni-
versity of Tennessee. L EG A L N 0TlcE
This report was prepared as an account of Gov~r111Jc1il -'Ponoorod work. l'l1;>1thP.r the United.
States, nor the Commission, nor any person acting oD behalf of the Comm.lesion:
A. Makes any warranty or representation, expre.Ssed or implied, with respect to the accu-
racy, completeness, or "usefulness of the'finformallori contained in this report, or that the use
of any Information, apparatus, method, or prQcess disclosed In this report may not infringe ·
prlvatel~ owned rlgblo; u1· - 1
_ B. Assumes any llabiltUes with respect to the u~e of, or for damages resulting from the
use of any Information, apparatus, method, or proce~s disclosed in this report.
As used in the above, "person acting on behal~ of the Commission" Includes any em-
ployee or contractor of the Commission, or employee of such contractor, to the extent that
such employee or contractor of the Conlmlsslon, oi employee of such contractor prepares,
disseminates, or provides access to, any Information pursuant to his employment or contract
with the Commission, or his employment with such contractor.

Oak Ridge, Tennessee

~1,STRIBUTION Of THIS DOCUMENT. IS UNllMlltD. I

. IJ
ii

Document _ Y-1585
__.;;___;,...;;....;;...;;_.. __
Engineering and Equipment
TID-4500

Distribution:

Ackerson, R. D. (AFSC) 0 'Hara, M. E.

Bailey, E. W. Oliphant, G. W.
Ballenger, H.F. (2) Perry, A. E.
Bernander, N. K. Phillips, L. R.
Burkhart, L. E. Rader, D. H.
Center, C. E. (0 RGDP) Reavis, J.P.
Christ man, A. M. Schwenn, M. F.
Douglass, T. E. Smith, H. F.
Ebert, J. W. Smith, R. D.
Ellingson, R. D. Sommerfeld, K. W.
Evans, G. W. Stoner, H. H.
Gambill, E. F. Thomas, F. W.
Gritzner, V. B. Tilson, F. V.
Harwell, W. L. (0 RGDP)(5) Trotter, T. C.
Hemphill, L. F. Waters, J. L.
Hensley, C. E. Wesley, R. L.
Huber, A. P. (ORGDP) Whitson, W. K.
Hull, R. H., Jr (15) Winkel, R. A. (Paducah)
Keller, C. A. (AEC-0 RO )(4) Wood, D. C.
Little, J. C. Yaggi, W. J.
Mclendon, J. D. Zurcher, E.
Meeks, E. J. (AFSC) Y-12 Central Files (5)
Mitchel, G. W. Y-12 Central Files (Y-12 RC)

In addition, this report is distributed in accordanc~ with the category Engineer-

ing and Equipment, as given in the 11 USAEC Standard Distribution Lists for Un-
classified Scientific and TechniCal Reports", TID-4500.
iii

ABSTRACT

The main purposes of this study were (1) to present a mathematical

proof of the AOQ formula for CSP-3, and (2) to apply this formula in deriving

AOQ and AFI values for CSP-3.

Before AOQ and AFI values were generated, it was necessary to verify

the CSP-3 AOQ formula. The proof revealed AOQ to be

[
qi (1-q k+4 + q4)(1-f) J
This above formula is different from the formula given in error by

Dodge and Torrey in the March, 1951 issue of Industrial Quality Control.

Because of the complexity of the AOQ formula and time consuming cal-

culations involved, a computer program was written in Fortran IV language

for the IBM 7090 computer in order to determine AOQ and AFI values. For

each set of values generated, the per cent defective, p, ranged from 0.5 per

cent to 50.0 per cent at increments of 0.5 per cent. The values of i used in this

study ranged from 5 to 200 units, and the sampling rate, f, began with 5 per

cent and terminated with 50 per cent at increments of 5 per cent. An assumption

which was made throughout this stl.~dy is that i = k. The computer program was

written in such a manner that AOQ and AFI values could be generated for a
iv

given i value at various levels of f and p with only one data card. Therefore,

AOQ and AFI values other than those appearing in this study may be generated

with this computer program.

In addition, curves for obtaining values of f and i for a given value of

AOQL were determined. When compared to the corresponding CSP-2 values,

an examination revealed that the i values for CSP-3 were reduced by about

4 units when AOQL is 2 per cent; at 3 per cent AO QL, the CSP-3 i values were

about 3 units less than the CSP-2 values; and for AOQL values greater than 3

per cent, the i values for CSP-3 were about 2 units less than the CSP-2 values.
v

TABLE OF CONTENTS

CHAPTER PAGE

I. INTRODUCTION •• . . .. . .. . .. .. .. . .. ..
The Prob Iem • .. . . .. ..... ... 3

Statement of the Problem • . . . . . . . . .. .. . 3

Importance of the Study ••• .... ... 4

Method of Procedure ••••• . . . . . . . . ... . . . . 6

Organization of the Thesis. . ... . . .. .. . .. 6

11. HISTORY OF CONTINUOUS SAMPLING INSPECTION ••• . . . . . . 8

Dodge-Type Plans •• . . . . ...... 8

Single-Level Plans . .... .... 9

Multi-Level Plans ••• . ....... 11

Other Plans •••••• .... 16

Ill. PROCEDURE FOR CSP-3 AND MATHEMATICAL

VERIFICATION OF AOQ ••• .. . .. 20

Procedure for CSP-3 •••• a a e • • a e .. . . .

I a 20

Mathematical Verification of AOQ •• . . . . . . ... 21

IV. DISCUSSION OF THE STUDY ••••• ..... 35

Techniques Used for Obtaining the Results . . . . . . . . . . . . . 35

CHAPTER PAGE

Choice of i, f, and p • • • • • • • • • • • • • • • • • • ~ •. • • • • 35

The Program .......... .. ... . . .. . . ... .. . 37

Results of the Study.. . .. . . . ... .. . .. .. . .. . . .. 38

AO Q Results • • • • • • • • • • • • • • • • • • • • , • • • • • • • 38

AFI Results .. .. . . .. . . .. . . . . . . . . . . .. . . .. 41

Curves for Determining Values off and for a Given

Value of AOQL . . . . . . . . . . . . . . . . . .. . . . . . . . . 44

v. SUMMARY AND RECOMMENDATIONS • • • • • • • • • • • • • • • · •• 47

Summary • . . . . . . • . . . . . . • . . . • • . . • . • . . . . . . 47

Recommendations .... ... ....... ..... ....... 49

BIBLIOGRAPHY .. .. .. . .. ...... .. .. . .. .. ... . .. .

APPENDIXES . . • • • • • . • • • . • • • . • . • • • • • • • . • • • . • . • 54

APPENDIX A • • • • • • . • • • . • • • . • . . • • . • • • • • • • . • . • 55

APPENDIX B. • • • • • • • • • • • • . • • • • • • • • . • • • • • • • . • 56
vii

LIST OF TABLES

·TABLE PAGE

I. AOQ and AFI Values for CSP-3

f = 0.050 and i = 5 • • • • • • • • • • . . . . . . . . . . ... 56

11. AOQ and AFI Values for CSP-3

f=0.100andi=5 ••••••••••• • • • • • 0 • ~ • • • 57

111. AOQ and AFI Values for CSP-3

f = O. 150 and i = 5 • • • • • • . . . . . . . . . .. . . . . . . 58

IV. AOQ and AFI Values for CSP-3

f = 0.200 and i = 5 • • • • • • • ... ............ 59

V. AOQ and AFI Values for CSP-3

f = 0.250 and i ,.. 5 • • • • • • . . . . . . . . . . . . . . ... 60

VI. AOQ and AFI Values for CSP-3

...
f = 0.300 and i = 5 • • • • • • • • • • • • • ....... 61

VII. AOQ and AFI Values for CSP-3

f = 0.350 and i = 5 • • • • • • ... . . . .... 62

VIII. AOQ and AFI Values for CSP-3

f = 0.400 and i =5 • • • • • • •. • • • ~ . . . . . . . . . . •. 63

IX. AOQ and AFI Values for CSP-3

f = 0.450 and i = 5 . . . . · · · .· · · • • · · • · • · · • • · 64

. '.?
viii

TABLE PAGE

X. AOQ and AFI Values for CSP-3

f = 0.500 and i :z: 5 •••••• ..... . . ... .. ... . . 65

XI. AOQ and AFI Values for CSP-3

f = 0.050 and i ,.. l0 • • • • • • . . . .. . . . . . 66

XII. AOQ and AFI Values for CSP-3

f = O. 100 and i = 10 • • • • • • • . . . . . .. . . . . . '• 67

XIII. AOQ and AFI Values for CSP-3

f = O. 150 and i = l 0 • • • • • • • . .. . . . . . . . . . . . . . . 68

XIV. AOQ a.nd AFI Values for CSP-3

f = 0.200 and i = 10 •••••• .... ............ 69

XV. AOQ and AFI Values for CSP-3

f = 0.250 and i = 10 •••••• . . . . . . . . . . . . . . . ... 70

XVI. AOQ and AFI Values for CSP-3

f :z: 0.300 and i = 10 • • • • • • • . .. . ... .. .... . . . ' 71

XVII. AOQ and AFI Values for CSP-3

f = 0.350 and i = l 0 • • • • • • ... ... ... . . ... . .. 72

XVIII. AOQ and AFI Values for CSP-3

f = 0.400 and i ,.. 10 • • • • • • . . .. . . . . . . . . . . .

~ ' 73
...

TABLE PAGE

XIX. AOQ and AFI Values for CSP-3

f == 0.450 and i == 10 • • • • • • • • ~ -. •• 74

XX. AOQ and AFI Values for CSP-3

f = 0.500 and i == l0 • • • • • • .' . ........ 75

XXI. AOQ and AFI Values for CSP-3

f == 0.050 and i == 20 • • • • • • • • .... . ... .. .. . . 76

XXll. AOQ and AFI Values for CSP-3

f == O. l 00 and i ::: 20 • • • • • • • • • • . . . . . . . . . . .... 77

XXlll. AOQ and AFI Values for CSP-3

f == O. 150 and i == 20 • • • • • • . . . . . . . . ... . . .

. . 78

XXIV. AOQ and AFI Values for CSP-3

f == 0.200 and i == 20 • • • • • • ... .. .. ........ . 79

XXV. AOQ and AFI Values for CSP-3

f == 0.250 and i == 20 • • • • • • • • • • • . .. . ... .... 80

XXVI. AOQ and AFI Values for CSP-3

f == 0.300 and i ,,.. 20 • • • • • • • • • • • • Cl • • • • .• • • • .• 81

XXVll. AOQ and AFI Values for CSP-3

f ::: 0.350 and i == 20 • • • • • • . . . . . . . . . . . . ·. . 82'

TABLE PAGE

XXVlll. AOQ and AFI Values for CSP-3

f = 0.400 and i = 20 • • • • • • • • • • .. .. ... . . . . .

. 83

XXIX. AOQ and AFI Values for CSP-3

f = 0.450 and i = 20 • • • • • • • . ... . . . . 84

XXX. AOQ and AFI Values for CSP-3 ·

f = 0.500 and i = 20 • • • • • • • 85

XXXI. AOQ and AFI Values for CSP-3

f = 0.050 and i = 30 • • • • • • ........... ..... 86

XXXll. AOQ and AFI Values for CSP-3

f = O. l 00. and i = 30 ••••••• 87

XXXlll. AOQ and AFI Values for CSP-3

f = O. 150 and i = 30 • • • • • • . .... .. ... ...... 88

XXXIV. AOQ and AFI Values for CSP-3

f = 0.200 and i = 30 • • • • • • • • • • .. .. .... .. .. 89

XXXV. AOQ and AFI Values for CSP-3

f = 0.250 and i = 30 • • • • • • • • • 90

XXXVI. AO Q and AF I Values for CSP-3

f = 0.300 and i = 30 • • • • • • • • • , .. .. ...... .. 91

TABLE PAGE

XXXVll. AOQ and AFI Values for CSP-3

f = 0.350 and i = 30 • • • • • • • .. . . . . ... .... . .

. 92

XXXVlll. AOQ and AFI Values for CSP-3

f = 0.400 and i = 30 • • • • • • • • ... 93

XXXIX. AOQ and AF I Values for CSP-3

f = 0.450 and i = 30 • • • • • • • .... .. .... ... . .

. 94

XL. AOQ and AFI Values for CSP-3

f = 0.500 and i = 30 • • • • • • • • .... ...... .... 95

XLI. AOQ and AFI Values for CSP-3

f = 0.050 and i = 40 • • • • • • ...... .. ...... .. 96

XLll. AOQ and AFI Values for CSP-3

f = O. l 00 and i = 40 •••••• ... . . . . . .. .. . . . . 97

XLlll. AOQ and AFI Values for CSP-3

f = O. 150 and i = 40 • • • • • • • • • .. . . .. .. ... .. . 98

XLIV. AOQ and AFI Values for CSP-3

f = 0.200 and i = 40 • • • • • • . . . . ... . . . . . . . . .. 99

XLV. AO Q and AF I Values for CSP-3

f = 0.250 and i = 40 • • • • •• ... . .... ..... . .. 100

xii

TABLE PAGE

XLVI. AOQ and AFI Values for CSP-3

f = 0.300 and i = 40 • • • • • • • • • • • • .. .. . . .. . . 101

XLVll. AOQ and AFI Values for CSP-3

f = 0.350 and i = 40 • • • • • • • • • • • 102

XLVlll. AOQ and AFI Values for CSP-3

f = 0.400 and i = 40 • • • • • • • . .. . . . . .. .... . . 103

XLIX. AOQ and AFI Values for CSP-3

f = 0.450 and i = 40 • • • • • • • • • • • • . .. .. .. ... 104

L. AOQ and AFI Values for CSP-3

f = 0.500 and i = 40 • • • • • • ..... .• 105

LI. AOQ and AFI Values for CSP-3

f = 0.050 and i = 50 • • • • • • • • • • • . . . . . . .. . . . 106

Lii. AOQ and AFI Values for CSP-3

f = 0.100 and i = 50 •••••• . . . . . . . . . . . . . . .... 107

Liii. AOQ and AFI Values for CSP-3

f = 0. 150 and i =50 • • • • • • ..' . . . . ...... . 108

LIV. AOQ and AFI Values for CSP-3

f = 0.200 and i =50 • • • • • • .. .. .... .. . 109

xiii

TABLE PAGE

LV. AOQ and AFI Values for CSP-3

f = 0.250 and i = 50 • • • • • • • • • .......... . . .

. 110

LVI. AOQ and AFI Values for CSP-3

f = 0.300 and i =50 • • • • • • • • • • . . . . . . . .. . . . . . 111

LVll •. AO Q and AFI Values for CSP-3

f = 0.350 and i = 50 • • • • • • • • • ............. 112

LVlll. AOQ and AFI Values for CSP-3

f = 0.400 and i = 50 • • • • • • .... . . . . .... .. . . 113

LIX. AOQ and AFI Values for CSP-3

f = 0.450 and i = 50 • • • • • • • • • . . . . . . . . . . . ... .114

LX. AOQ and AFI Values for CSP-3

f ,,,, 0.500 and i = 50 • • • • • • ...... ....... . .. 115

LXI. AOQ and AFI Values for CSP-3

f = 0.050 and i = 60 • • • • • • • .... . ........ .. 116

LXll. AOQ and AFI Values for CSP-3

f = O. 100 and i = 60 •••• · •• .. . . . . . ... . . .. . . 117

LXlll. AOQ and AFI Values for CSP-3

f = O. 150 and i = 60 • • • • • • • • ... . . . ........ 118

xiv

TABLE PAGE

LXIV. AOQ and AFI Values for CSP-3

f = 0.200 and i = 60 • • • • • • • • .. .. .. .... .... 119

LXV. AO Q and AF I Va I ues for CS P-3

f = 0.250 and i = 60 • • • • • • • • • • .. .. ....... . 120

LXVI. AOQ and AFI Values for CSP-3

f = 0.300 and i = 60 • • • • • • • .. .. . .. . . . . .. . . 121

LXVll. AOQ and AFI Values for CSP-3

f = 0.350 and i = 60 • • • • • • • • .............. .122

LXVI 11. AOQ and AF I Values for CSP-3

f = 0.400 and i = 60 • • • • • • • • • • • . . . . . .. .. . . 123

LXIX. AOQ and AFI Values for CSP-3

f = 0.450 and i = 60 • • • • • • • • .... ....... 124

LXX. AOQ and AFI Values for CSP-3

f = 0.500 and i = 60 ••••••••• ·• • • • • • • • • • • • • . 125

LXXI. AOQ and AFI Values for CSP-3

f = 0.050 and i =80 • • • • • • . . .. . . . . . . . . . . . . . 126

LXXll. AOQ and AFI Values for CSP-3

f = 0. l 00 and i = 80 • • • • • • ...... ...... .. .. 127

TABLE PAGE

LXXlll. AOQ and AFI Values for CSP-3

f = O. 150 and i = 80 ••••••••• .. ......... .. 128

LXXIV. AOQ and AFI Values for CSP-3

f = 0,200 and i = 80 • • • • • • ...... . ....... .. 129

LXXV. AOQ and AFI Values for CSP-3

f == 0.250 and i == 80 • • •••• ......... ....... 130

LXXVI. AOQ and AFI Values for CSP-3

f = 0.300 and i = 80 •••••• .. . .. .. .... . ... . 131

LXXVll. AOQ and AFI Values for CSP-3

f = 0.350 and i = 80 • • • • • • • • • ........ ..... 132

LXXVlll. AOQ and AFI Values for CSP-3

f = 0.400 and i = 80 • • • • • • .. .. . . . . . . .. .. . . 133

LXXIX. AOQ and AFI Values for CSP-3

f = 0.450 and i = 80 • • • • • • .. ...... ........ 134

LXXX. AOQ and AFI Values for CSP-3

f = 0.500 and i = 80 • , • • • • • • ......... ..... 135

LXXXI. AO Q and AF I Values for CSP-3

f = 0.050 and i = 100 • • • • • . . . .. . . . . . . .... . 136

xvi

TABLE PAGE

LXXXll. AO Q and AF I Values for CSP-3

f = O. 100 and i = 100 • • • • • • . . .... .. .. . .. . . 137

LXXXlll. AOQ and AFI Values for CSP-3

f = 0.150 and·i = 100 •••••• 138

LXXXIV. AOQ and AFI Values for CSP-3

f = 0.200 and i = 100 ••••• ................ 139

LXXXV. AOQ and AF I Values for CSP-3

f = 0.250 and i = 100 • • • • • • • .. .. .... .. . . . . 140

LXXXVI. AOQ and AFI Values for CSP-3

f = 0.300 and i = 100 • • • • • • .. ... ....... 141

LXXXVll. AOQ and AFI Values for CSP-3

f = 0.350 and i = 100 • • • • • • • • ... ........' . 142

LXXXVlll. AOQ and AFI Values for CSP-3

f = 0.400 and i = 100 • • • • • • • • . . . .......... 143

LXXXIX. AOQ and AFI Values for CSP-3

f = 0.450 and i = 100 • • • • • • • ..............' 144

XC. AOQ and AFI Values for CSP-3

f = 0.500 and i = 100 • • • • • 145

xvii

TABLE PAGE

XCI. AOQ and AFI Values for CSP-3

f = 0.050 and i = 200 • • • • • ........ ... . .. .. 146

XCll. AOQ and AFI Values for CSP-3

f = O. l 00 and i = 200 • • ••• ... . . . . . . .. .. . . . 147

XClll. AOQ and AFI Values for CSP-3

f = O. 150 and i = 200 • • • • • • • . . . . . . .. . . . . . . 148

XCIV. AOQ and AFI Values for CSP-3

f = 0.200 and i = 200 • • • • • • • • . . . . ...... . . . 149

XCV. AOQ and AFI Values for CSP-3

f = 0.250 and i =200 ••••••• .. .. ....... ... 150

XCVI. AOQ and AFI Values for CSP-3

f = 0.300 and i = 200 ••••••••••••••••••••• 151

XCVll. AOQ and AFI Values for CSP-3

f = 0.350 and i = 200 • • • • • • • • • • • • • • • • • • • • • 152

XCVlll. AOQ and AFI Values for CSP-3

f = 0.400 and i = 200 • • • • • • • • • • • • • • • • • • • • • 153

XCIX. AOQ and AFI Values for CSP-3

f =0.450 and i = 200 • • • • • • • • • • • • • • • • • • • • • 154

xviii

TABLE. PAGE

C. AOQ and Ari Values for CSP-3

f = 0.500 and i = 200 • • • • • . .. . . . . . ...... .. 155

xix

LIST OF FIGURES

FIGURE PAGE

1. Procedure for H-106 Continuous Sampling Inspection Plan

for f = 1/2 • • • • • • • • • ... 15

2. Procedure for CSP-3 . . . . . . . . . . .. . . . . . . .. 22

3. AOQ Curves, i Varying, f = 0.30 . . . . . . . . . . . . . . .. 39

4. AOQ Curves, f Varying, i = 20. ........ .... ...... 40

5. AF I Curves, i Varying, f = 0.30 ••• .... 42

6. AF I Curves, f Varying, i = 20 • • • • • • 0 • • • • 0 • 43

7. Curves for Determining Values off and i for a Given Value

of AO QL • • • , • • • • • • • • • • • • • • • • • • • • • • ...... -45
CHAPTER I

INTRODUCTION

A continuous sampling inspection plan is one which is designed to

be applied continuously as the items are completed and presented for acceptance.

As each item is presented, the decision is made to inspect or not to inspect

it according to: (l) previous inspection results, and (2) a random selection
1
procedure that gives the proper probability of selecting each item. One

of the basic and most important advantages of continuous sampling plans over

lot-by-lot acceptance plans is that in applying the continuous plan, it is not

necessary to accumulate a lot before making a decision. This advantage is

especially vital when producing costly items because: (1) defective items

wil I be. spotted sooner after they occur than if the items were accumulated

for lot-by-lot acceptance and,. therefore, (2) the production or fabrication

group can be kept currently informed as to the status of production items.

1
inspection and Quality Control Handbook (interim) H-106, Multi-Level
Continuous Sampling Procedures and Tables for Inspection by Attributes,
A Report Prepared by the Department of the Navy (Washington: Office of the
Assistant Secretary of Defense, 1958), p. l.
2

A frequently used continuous sampling plan is designated as CSP-3,

2
which was originated by Harold F. Dodge and Mary N. Torrey. Before further

discussion of CSP-3, it would be advantageous to examine briefly the original

plans upon which CSP-3 is based. Dodge's initial continuous sampling plan
3
is referred to as CSP-1. At the beginning of inspection, 100 per cent of

the product is examined. When i consecutive· units are free of defects, 100

per cent inspection is discontinued and a fraction, f, of product is inspected.

While sampling, if a defective unit is found, reversion to 100 per cent in-

spection is made. In an effort to satisfy complaints that immediate return

to 100 per cent inspection after only one defect was too severe a restriction,
4
Dodge and Torrey devised CSP-2. The only additional stipulation under

CSP-2 is that sampling inspection will be discontinued and a return to 100 ·

per cent inspection will occur when two defective items are found within

a sample of k units. It is assumed that k = i in almost al I instances although

this is not a requirement. For purposes of providing greater protection against

2
Harold F. Dodge and Mary N. Torrey, "Additional Continuous Sampling
Inspection Plans," Industrial Quality Control, 7:7-12, March, 1951.
3
Harold F. Dodge, "A Sampling Inspection Plan for Continuous Pro-
duction," The Annals of Mathemat'ical Statistics, 14:264-79, September, 1943.
4
Dodge and Torrey, loc. cit.
3

unexpected surges of defective items, an additional restriction was added

to CSP-2. This new plan is referred to as CSP-3. When applying CSP-3 under

sampling inspection, after a defective item has been found, the next four

items are inspected. If a defect is found within the four items, 100 per cent

inspection is invoked. A more complete discussion and diagram of the opera-

tion of CSP-3 appears in Chapter Ill.

I. THE PROBLEM

Statement of the Problem

A frequently encountered quality which is used .to describe a given

sampling plan is called the. Average Outgoing Quality,. or AOQ. The curve

of AOQ is a graphic device which depicts the relationship between incoming

and final outgoing quality. Final outgoing quality includes those items accepted

on the basis of the sampling plan and excludes other defective items which

Were rejected on the basis of the sampling plan. In other words, if the frac-

tion defective of incoming items is small, most of the production will be

accepted and therefore, the outgoing quality will have a low fraction defective.

However, if the other extreme exists (a high fraction defective of incoming

items), 100 per cent inspection would. be invoked resulting in a low fraction

defective of outgoing quality;. In summary, an AOQ value is usuall°y expressed

as per cent defective which, in the long run, will not be exceeded, regardless
5
of the quality submitted to the inspector.

Another important graphic device, which has proved to be helpful in

estimating inspection costs, is the average fraction inspected, or AF I curve.

The average fraction of total items inspected is o. direc;:t function of the in-

coming quality. As the fraction defective of incoming production increases,

the average fraction of total items inspected will also increase.

The purposes of this study are: (1) to present a mathematical proof

of the AOQ formula for CSP-3, (2) to apply this formula in deriving AOQ

curves for CSP-3, (3) to generate AF I curves for CS P-3, and (4) to offer

suggestions for possible areas of further study.

Importance of the Study

In the article discussing CSP-3 in Industrial Qua Iity Control, Dodge

and Torrey stated that AOQ values for CSP-2 may be used to approximate
6
the CSP-3 AOQ values. In some instances, however, it would be advantageous

to use exact AOQ vo.lues for CSP-3 instead of CSP-2 approximations. One

such instance exists at the Union Carbide Corporation-Nuclear Division,

5
Dodge, op. cit., p. 264.
6
Dodge and Torrey, op. cit., p. 10.
- --
5

Y-12 Plant, located at Oak Ridge, Tennessee. A continuous inspection sampling

plan which is based upon CSP-3 is in use at the Y-12 Plant. Since very little

information about CSP-3 is available, especially AOQ cur:ves, it is most

desirable for management and the administrators of the sampling plan to

hav~ n-1ore exact AOQ values for CSP-3 in lieu of CSP-2 approximations.

The information contained in this study can supply the Y-12 Plant with these

useful AOQ values.

The Average Outgoing Quality Limit (AOQL) describes the worst quality

of material that a given sampling plan is capable of producing. Mathematically,

the AOQL is the maximum AOQ value for any given plan. Therefore, with

the use of the AOQ curves derived in this study, AOQL values for CSP-3 may

be determined approximately.

Another use for the AOQ curves presented in this study may be for

purposes of determining the average fraction of total units inspected. With

the use of the formula:

AFI =l - AOQ,
p

AF I = average fraction of total units inspected, and

p = fraction defective of units produced,

average fraction inspected curves for CSP-3 may easily be determined.7

II. METHOD OF PROCEDURE

The AO Q and AF I values for CSP-3 were derived with the use of the

formula presented in Chapter Ill. ~n assumption which was made throughout

this study is that i = k. Therefore, with the exception of the general proof

of the AOQ formula appearing in Chapter 111, i will be used to refer to both

i and k for CSP-3. Because of the complexity and time-consuming calculations

involved, a computer program was written for purposes of determining the

AOQ and AFI values. For each curve generated, the per cent defective, p,

ranged from 0.5 to 50.0 per cent in increments of 0.5 per cent. A large number

of AOQ and AFI values were then derived for various levels off and i.

Ill. ORGANIZATION OF THE THESIS

Chapter I includes an introduction, a statement of the prob.lam, its

importance, and the method of procedure.

Chapter 11 presents a review of the literature concerning continuous

sampling procedures.

7
1bid., p. 11.
7

Chapter 111 contains a full discussion and diagram for CSP-3 and gives

mathematical proof of the formula for determining AOQ values.

Chapter IV describes the manner in which the AOQ and AFI curves

were generated and contains a summary and discussion of the results.

Chapte1· V is a summary of the. preceding chapters and, in addition,

presents conclusions and suggestions for areas of future study.

CHAPTER II

HISTORY OF CONTINUOUS SAMPLING INSPECTION

Since the introduction of continuous sampling inspection plans in 1943

by Dodge, several modifications of the original plan in addition to other types

of continuous inspection plans have been devised.·

I. DODGE-TYPE PLANS

Although previously mentioned, because of its importance and con-

tribution to other plans, the procedures for Dodge's first plan, CSP-1, shall

be discussed further. CSP-1 is a simple and easy-to-administer plan whereby

i consecutive units must be found free of defects before beginning the sampling

of a fraction, 1/k, of.units produced. If a defective unit is found while sampling,

100 per cent inspection is initiated immediately, completing the cycle.

Using the original Dodge plan as a basis, several other continuous

1 2 3
sampling plans were proposed. These plans include CSP-2, CSP-3, CSP-4,

1
H. F. Dodge and M. N. Torrey, "Additional Continuous Sampling In-
spection Plans," Industrial Quality Control, 7:7-12, March, 1951.
2
1bid., pp. 9-12.
3
c. Derman, M. V. Johns, Jr.,and G. J. Lieberman, "Continuous Sampling
Procedures Without Control," The Annals of Mathematical Statistics, 30:
1175-1191, December, 1959.
9
4 5 6 7 8 9
CSP-5, H-106, MLP, MLP-r x 1, MLP-T, and MLP-r x s.

Single-Level Plans

In 1951, Dodge and Torrey expanded CSP-1 into two additional con-

tinuous sampling plans, CSP-2 and CSP-3. Under these three continuous

sampling plans, Dodge and Torrey assumed that the process was in a state

of statistical control. The process is defined as being in a state of statistical

control if for every item produced the probability that it is defective is a

positive constant p ~ 1 where p is independent of the condition or defective-

ness of all the other items produced. In the original sampling plans, Dodge

and Torrey guaranteed an AOQL with the assumption that the process was

5
1nspection and Quality Control Handbook (interim) H- l06, Multi-Level
Continuous Sampling Procedures and Tables for Inspection by Attributes, a
Report Prepared by the Department of the Navy (Washington: Office of the
Assistant Secretary of Defense, 1958) pp. 1-16.
6
Gerald J. Lieberman and Herbert Solomon, "Multi-Level Continuous
Sampling Plans," The Annals of Mathematical Statistics, 26:686-704, December,
1955.
7
C. Derman, S. Littauer, and H. Solomon, "Tightened Multi-Level Con-
tinuous Sampling Plans," The Annals of Mathematical Statistics, 28:395-404,
March, 1957.
8
1bid.
9
,,, 1bid.
10

in a state of statistical control. In 1953 Lieberman showed that this procedure,

without the assumption of statistical control, guaranteed an AOQL, although

10
·different from that specified by Dodge.

The plans CSP-4 and CSP-5 were devised in an effort to assure an

AOQL without assuming the process to be in statistical control. Under the

procedures for CSP-4 and CSP-5, changes from the original Dodge plan were

made in the rule of action when a defective item was found while sampling

was in effect. Following are the steps outlining the procedures for CSP-4:

1. Begin by inspecting 100 per cent of the units consecutively as pro-

duced until i successive units are found free of defects.

2. After examining i successive defect-free units, discontinue 100

per cent inspection, and inspect only a fraction 1/k of the units,

choosing the unit to be inspected at random from a segment

of size k.

3. If, while sampling, a defect is found, revert immediately to 100

per cent inspection, eliminating from the production process

the remaining k-1 items of the segment, and returning to 100

per cent inspection with the next item following the eliminated

10
Gerald J. Lieberman, "A Note on Dodge's Continuous Inspection
Plan," The Annals of Mathematical Statistics, 24:480--84, September, 1953.
11

segment. This completes the cycle. As in Dodge's original plan,

all defective units are corrected or replaced with good units.

The second continuous sampling plan, CSP-5, which was derived without

the assumption of the existence of a state of statistical control is the same as

CSP-4 except for Step 3. Under Step 3, the procedure for CSP-5 is as follows:

If, while sampling, a defective unit is found, screen the remaining k-1 items

in the segment. After screening, begin 100 per cent inspection with the next

item produced. Continue 100 per cent inspection until i successive defect-free

units are found. These i units do not include the k-1 screened units. It is

significant to note that for a given k and i, the AOQL values for CSP-4 and

CSP-5 are always slightly higher than Dodge's CSP-1 AO QL values.

Multi-Level Plans

In an effort to satisfy complaints that the original Dodge plans required

. too abrupt a change between 100 per cent inspection and sampling inspection,

Lieberman and Solomon devised a multi-level continuous sampling plan,

referred to as MLP. In addition to allowing a smoother transition between

100 per cent inspection and sampling inspection, the authors contended that

MLP required 100 per cent inspection only when the quality was very poor

and required a minimum amount of inspection.when quality was good. The MLP

inspection scheme, which is an extension of CSP-1, consists of a number of

sampling levels subject to the provision that transitions can only occur between

adjacent levels. These sampling levels or rates decrease in a geometric pro-

gression as the product qualityimproves.Followingarethe procedures for MLP:

l. Begin by inspecting 100 per cent of the units consecutively as pro-

duced until i successive units are found free of defects.

2. After examining i successive defect-free units, discontinue 100

per cent inspection and inspect only a fraction, f, of the units.

If the next i units are free of defects, proceed to the next level.

If a defect occurs in the above i units, revert immediately to

100 per cent inspection.

3. If, while sampling at rate f, i units are found to be free of defects,

discontinue sampling at rate f and proceed at a sampling rate

of r. If the next i units are free of defects, proceed to the next

level. If a defect occurs in the above i units, !"evert immediately

to sampling at rate f.

4. If, while sampling at rate r, units are found to be free of defects,

discontinue sampling at rate ~and proceed at a sampling rate

3
of f • If the next i units are free of defects, proceed to the next

level. If a defect occurs in the above i units, revert immediately to

sampling at rater.
13

The above procedure may continue until a sampling rate off is reached.

As in CSP-1, al I known defective items should either be corrected or replaced

with good items. It is significant to observe that for the special case where

k = l, MLP reduces to the original Dodge plan, CSP-1.

Under MLP, after finding a defect while sampling at rote f, a reversion

to sampling rate (- l is mode. In an apparent effort to increase the penalty

for finding a defective unit, three tightened multi-level continuous sampling

plans, based upon MLP, were originated. These plans are referred to as

MLP-r x l, MLP-r x s, and MLP-T.

When operating under the MLP-r x l plan, if i successive defect-free

units . ore found while sampling at the j-th level, proceed to sample at the

(j + l )-th level. If a defect is found before i successive acceptable units occur,

revert immediately to the (j - r)-th level, if j > r, or to 100 per cent inspection

if j ~ r. In .the special case where r = l, MLP-r x l reduces to the original MLP

plan.

The MLP-r x s pion is basically the same as MLP-r x l, except that

when i successive units are found free of defects while sampling at the j-th

level, sampling begins at level (j + s). If a defect is found, the some action

as required under MLP-r x l should be token. In the spe,iol situation where

r = s, MLP-r x s reduces to the original MLP plan. Since MLP-r x s is a

tightened plan, with respect to the MLP plan, only cases where r > s should

be considered.

The third tightened multi-level plan, MLP-T, is the same as MLP-r x 1

except upon finding a defective item a reversion to 100 per cent inspection

is made immediately. Of the three tightened multi-level plans, MLP-T is

the most strict in the sense that a reversion to 100 per cent inspection is

accomplished sooner after finding a defect.

In 1958, the Department of Defense issued a handbook, referred to as

H-106, which presented a multi-level continuous sampling scheme. The first

level of this plan very closely resembles CSP-3. The only difference between

Level of the H-106 plan and CSP-3 is that after inspecting the four units

following a sampling defect, and no defects are found within these four units,

the H-106 plan requires sampling to continue until the next i - 4 units are

inspected; whereas, under the same conditions, CSP-3 requires sampling

to continue until i acceptable units are inspected. As in the MLP plans, the

H-106 plan provides for a reduction of the sampling rate when the process

exhibits a quality level much better than the AOQL. The H-106 plan calls

for three sampling levels with the rate of change of sampling from level to
3
level being in a geometric progression, f, f, and f • Figure 1 shows a flow

diagram of the H-106 inspection scheme for f = 1/2.

100% Level
lnspect-100%
of the units in
the order
i...11---------------------------------------.....,
produced.

If i consecutive units If a defective is

are found to he free found among the
of defectives,shift to 4 in~p~cted units,
Level No. 1. shift to 100%
inspection.

'
Sampling Rate
~evel No. 1
Inspect at rate
f = 1/2 -State 1°
Resume in-
spection at
rate f-1/2
If a defective is lound while
sampling at rate f = 1/2, shift
to State lR~.-----------------iunits following
- - - - - - - If the 4 inspected units are Found
to be free of defectives, shift to
State R
1
Inspect next 4

the defe~tive
found.
State 1° . - - - - - - - - - - - - " '
- - - - - - - - - - I f a defective is found
If i inspected units ofter resuming inspection If a defective is
If i-4 inspected units at rate f= 1/2, shift to
are found to be free found among the
are found to be free 100% inspection-.---"'
of defectives, shift 4 inspected units,
of defectives, shift
to Level No. 2. shift to State 1R.
to Level No. 2.

Somplin 9 Rate 1---------lf a defective is found while State 2 R

Level No. 2 sampling at rate fL=l/4,shift lnpscct noxt 4
Inspect at rate
State 2°
ta State 2R-.----------------o-t units Following
f2 = 1/4 ~:.::.::_.::.__-1-' _________ lf the 4 inspected units ore the defective
found.
Resume in- found to be free of defectives,
spect~on at shift to State 2° . - - - - - - - - -..
4
Lr:.::a::.te:.f;..-=__:lf__/::.- - - - - - - - - - - - If a defective is found
If i inspected units If i- 4 inspected units afterres~ming inspection If a defective is
are found to be free are found to be free of at rate f = 1/4, shift to found among the
of defectives, shift defectiv~s, shift to State 1R. 4 inspected units,
to Level No. 3. LevP.I No. 3 , shift to State 2R.

Sampling Rate
I
State 3R
Level No. 3 1---------- If a d~fective is (ound whi.le
Inspect next 4
Inspect atrate sampling at rate rJ= 1/8, shift
units following
f3 = 1/8 State 3 o to State 3R. the defective
i-::.=_::_::....._-f"I_________ If the 4 inspected units are
Resume in-
found.
found to be free of defectives,
spection at shift to State 3° . - - - - - - - - - - - - " '
ral~ r3~ ljs 1------ - - - - - - - I f a defective is found
If i-4 inspected units are ofter resuming inspection
found to be free of defec- at rate f3= 1/8, shift to
tive" shift tn LP.vel No, 3. State 2R . - - - - - - - - - - - - - - - - - - "

FIGURE

PROCEDURE FOR H-106 CONTINUOUS SAMPLING INSPECTION

PLAN FOR f = 1/2
16

As in most Dodge-type plans, with the exception CSP-4 and CSP-5,

the H-106 plan assumes the production process to be in a state of statistical

control. In 1966, White provided several AOQL values for the H-106 plan
11
under the assumption that the process was not in a stQte of statistical control.

These AOQL values were referred to as unrestricted, and designated UAOQL,

Thus, the UAO QL may be defined as the long-run per cent of defective items

in the outgoing product after inspection as a result of the plan's performance

against the worst possible quality sequence of produced items. The author

points out that in some instances the UAOQL has been found to be as much as

four times greater than the corresponding AOQL calculated under. the as-

sumption of the process' being in statistical control.

II. OTHER PLANS

In 1945, Wald and Wolfowitz proposed a continuous inspection scheme

12
referred to as SPA. Unlike the Dodge-type plans, the SPA plan begins with

sampling inspection. A predetermined number of units are sampled until

l1 Leon S. White, "The Evaluation of H-106 Continuous SamplingPlans

Under the Assumption of Worst Conditions," Journal of the American Statistical
Association, 61:833-841, September, 1966.
12
A. Wald and J. Wolfowitz, "Sampling Inspection Plans for Continuous
Production Which Insure a Prescribed Limit on the Outgoing Quality," The
Annals of Mathematical Statistics, 16:30-49, March, 1945•

.. .
17

more than a fixed number of defective items are found while sampling. Whenever

this fixed number of defective units is exceeded, 100 per cent inspection is

required. Sampling inspection is reinstated when the cumulative proportion

defective of the inspected units is equal to or less than the prescribed limit.

The authors state that SPA will guarantee a given AOQL whether or not the

production process is in a state of statistical control. However, when the process

is under statistical control, the SPA plan requires a minimum amount of

inspection as well as guaranteeing a specified AOQL.

The authors reported that although the SPA plan will guarantee the

desired AOQL with inspection at a minimum when the production process

is in statistical control, the plan is rather unfavorable as far as prot!!ction

against accepting a short run of unusually bad production. In order to achieve

better local stability, Wald and Wolfowitz proposed a slight modification of

13
SPA, referred to as SPc. The SPC plan requires the beginning of 100. per cent

inspection more quickly in situations where a number of defects occur within

a short production run, regardless of the cumulative proportion defective.

The authors emphasize that the use of the SPC plan does not increase the

AOQL over the corresponding SPA values. However, the amount of inspection

13
1bid., pp. 43-46.
18

will be increased, especially under very high local stability require-

ments.

In 1954, Girshick proposed a continuous inspection scheme similar

14
to the plans of Wald and Wolfowitz. The Girshick plan begins by sampling

1/1< units and continues at this rate until a predetermined number of defective

units, m, have been observed. At this time, if the cumulative sample size, n,

is less than another predetermined . number, N, one begins anew. However,

if the cumulative sample size is greater than or equal to N, the plan requires

100 per cent inspection of (N - n) k units before beginning the cycle anew.

The values of m and N are mainly dependent upon the desired AOQL. As in

the plans of Wald and Wolfowitz, the big advantage of the Girshick plan is

that it holds the AOQL whether or not the process is in a state of statistical

control.

In summary, the basic difference between the Dodge-type plans and the

Wald and Wolfowitz-type plans is the concept of sampling versus initial 100

per cent inspection. Al I Dodge-type plans begin with 100 per cent inspection

and initiate sampling inspection when the production reaches an acceptable

quality level. The Wald and Wolfowitz-type plans initially call for sampling

14
M. A. Girshick, Technical Report No. 16, Applied Mathematics and
Statistics Laboratory, Stanford University, July, 1954.
19

inspection, with 100 per cent inspection being required only when the production

reaches an unacceptable level •

.•
20

CHAPTER Ill

PROCEDURE FOR CSP-3 AND MATHEMATICAL

VERIFICATION OF AOQ

The following section of the study consists of a discussion and diagram

of procedures for CSP-3. In addition, a mathematical verification of the AOQ

formuia for CSP-3 is presented.

I. PROCEDURE FOR CSP-3

Following is a detailed description of the procedure for CSP-3: l

1. In the beginning, inspect 100 per cent of the units consecutively

as produced until i successive units are found free of defects.

2. When i successive units are found free of defects, discontinue 100

per cent inspection and inspect only a fraction, f, of the units,

selecting these units in a random manner.

3. If a defective unit is found while sampling, inspect the next four

units. If a defect is found in the four units, revert to 100 per cent

inspection and follow the procedure outlined in Step 1. If no defect

1
Harold F. Dodge and Mary N. Torrey, "Additional Continuous Sampling
Inspection Plans," Industrial Quality Control, 7:9-10, March, 1951.
21

is found in the four units, continue sampling, but keep count

of the number of units inspected after a defect is found.

4. If a defect is found in the next i or less sample units; revert to

100 per cent inspection and follow the procedure outlined in·

Step l. If no defect is found in the next i or less sample units,

continue sampling until the next defect is found. When the next

defect is found, proceed as outlined in Step 3.

Figure 2 presents the above procedure in the form of a flow diagram

for easy reference.

II. MATHEMATICAL VERIFICATION OF AOQ

Before generating AOQ and AFI values, it was necessary tq verify,

mathematically, the AOQ formula for CSP-3. The formulae to follow, with

.the exception of 7 and 8, were presented in earlier CSP-1 and CSP-2 works
2
of Dodge and Torrey. The proofs and derivations of some of these formulae,

however, were incomplete. In 1957, C. C. Thigpen fully developed the basic

2
1bid., pp. 10-11; and H. F. Dodge, "A Sampling Inspection Plan for
Continuous Production," The Annals of Mathematical Statistics, 14:264-79,
September, 1943.
22

Inspect 100% of units consecutively until i

ST ART-----.-t
units in succession arc found clear of defects,

When i units in succession

are found clear of defects:

Discontinue 100% inspection, and inspect only a

fraction, f, of lhe units, selecting individual s1;1mple
unit~ one at a time f1u111 Ifie flow of product in such a
manner as to assure an unbiased sample.

When a defect is found:

Inspect the next 4 units.

If no defect is found If a defect is found

in 4 •.•nits: in 4 units:

Continue sampling, but keep count Revert to 100%

of the number of units inspected af- inspection.
ter finding the defect.

If a defect is found in the

If no defect is found in the next k or less sample units
next k sample units inspected: inspected:

Continue sampling until Revert to 100%

the next defect is found. inspection.

FIGURE 2

PROCEDURE FOR CSP-3

23
. 3
formulae for CSP-2. Since the basic CSP-2 formulae also are applicable

to CSP-3, the proofs of these formulae were not duplicated in this study.

Following are the definitions and formulae for 100 per cent inspection:

l. The probability of a failure sequence occurring, denoted byP ,

1
or a series of less than i acceptable units followed by a defective is:

2 3 i-1 l i i
(l + q + q + q + ••• + q ) p = -1 (l-q ) p = 1-q •
-q
i
2. Since the probability of getting a failure sequence is 1-q , the probe-

bility of getting a passing sequence, denoted by Q , or a series of

1
i or more acceptable units before finding a defective, is qi. There-

3. This section defines the expression for the average number of

failure sequences before finding i successive defect-free units as:

Denote this average value by G.

3
C. C. Thigpen, "Extension of the Dodge CSP-2 Continuous Sampling
Plan to the Case i Not Equal to k" (unpublished Master's thesis, The University
of Tennessee, Knoxville, 1957), pp. 11-17.
24

4. Having defined the average number of failure sequences before finding

i successive defect-free units, the average number of units in

a failure sequence next is defined as:

i
1. .!.9_i-- h 1 •
p
1-q
a:n

s. From Definitions 3 and 4, the average number of units inspected

in a 100 per cent inspection is defined to be Gh + i

1
= u.
6. This section defines the average number of units inspected in a

sequence of i or more good units followed by a defective unit.

The expression will define the average number of units in a

sequence of good units followed by a defective unit, given that

the sequence includes i good units initially. This expression is

defined to be ~ + i = g1•
7. The probability of a failure sequence of less than four units occurring
4
is 1-q , and is denoted by R.
8. From Definition 7, it follows that the probability of getting at least
4
four good units is q , and is denoted by R.

Similar definitions and expressions for sampling inspection are given

as follows:
25

9. The probability of a sequence of less than k good units and then a

defective is 1-qk = P2 • Combined with Q and R, a complete

1
cycle is made. Then, this cycle of a1RP 2 is the probability of

beginning with 100 per cent inspection, inspecting i consecutive

good unit, beginning sampling inspection, finding a defective

unit, inspecting the next four units, finding a defective unit

within these four units, and then a reversion to 100 per cent

inspection. Note that all cycles are terminated either with P

2
(the probability of a. failure sequence of k or less units) or

with R (the probability of a failure sequence of less than four

units).

10. The probability of finding k or more good sample units before finding

the first defective is qk, and is denoted by a2 •

11. The average number of un'its inspected in a failure sequence of

less than k good sample units followed by a defective unit is

k
defined to be h
2
=..!_ - ; .
p 1-q
12. The average number of units in a sequence of k or more good units

followed by a defective unit is defined to be g

2
= ~ + k.
13. The expressions defining AFI and AOQ now can be derived. From

Definition 5, u is the average number of units inspected in a

given 100 per cent inspection. Similarly, d is defined to be

the average number of times that a group of four consecutive

units is examined. During sampling inspection, a fraction, f,

of units are inspected. If the average number of units accepted

during one sampling inspection is denoted by v, the average

number of units inspected during one sampling inspection is fv.

Therefore, the average fraction of total units inspected in the

long run, AF I, is:

AF I = u + d + fv •
u+d+v

As mentioned before, one inspection cycle extends from the beginning

of 100 per cent inspection until it is reverted to again. The expected value

of the total number of units inspected for any given plan may be found by

considering the number of units that would be inspected in all possible cycles.

Probability of Possible Average Number Inspected in Such

Cycles Cycles

Q1R Q1R(g1+4)

Q1RP2 + Q1RP2(gl+h2+4)

Q1RQ2R + Q1RQ2R(gl+g2+2·4)

Q1RQ2RP2 + Q1RQ2RP2(gl+g2+h2+2•4)

Q1RQ2RQ2R + Q RQ Ra R(g +2g +3·4)

1 2 2 1 2
27

QlRQ2RQ2RP2 + Q Ra Ra RP (g +2g +h +3·4)

1 2 2 2 1 2 2

a 1 R~R~RQ R 2 + Q Ra Ra Ra R(g +3g +4•4)

1 2 2 2 1 2

+ ••• +

PlQlR P la lR(h 1+9 1+4)

pl Ql RP2 + PlQlRP2(hl+gl+h2+4)

Pl al ROiR + p 1Q 1RQ2R (h t 9 1+g2+2 •4)

pl Ql RQ2RP2 + p 1Q1RQ2RP2(h l+gl+g2+h2+2 •4 )

Pl Ql RQ2RQ2R + P a Ra RQ R(htg +2g +3•4)

1 1 2 2 1 2

pl Ql RQ2RQ2RP2 + P a Ra Ra RP (h +g 2g +h +3·4)
1 1 2 2 2 1 1 2 2

Pl Ql RQ2RQ2RQ2R + P a RQ RQ RQ R(h +g +3g +4·4)

1 1 2 2 2 1 1 2

+ ••• +
2 - 2 -
pl Ql R p 1 Q1R(2hl +gl +4)
2 2
pl Ql RP2 + p 1 Q lRP 2(2h tg1 +h2 +4)
2 - 2 -
pl Ql RQ2R + p 1 Q lRQ2R(2hl +gl +92 +2·4)
2 2
pl Q1RQ2RP2 +pl QlRQ2RP2(2hl+gl+g2+h2+2•4)
2 - 2 -
pl QlRQ2RQ2R + Pl Q l RQ Ra R(2h l +g l +2g +3·4)
2 2 2
2 2 .
pl Ql RQ2RQ2RP 2 +Pl Q Ra Ra RP (2h +g +2g +h +3·4)
1 2 2 2 1 1 2 2
2 - 2 -
pl QlRQ2RQ2RQ2R +Pl Ql RQ Ra RQ R(2h +g +3g +4•4)
2 2 2 1 1 2

+ ••• +

+ ••• +
28

Sum of all terms for 9 :

1
(Q1R91 + Q1RP291 + Q1RQ2Rgl + Q1RQ2RP291 + ••• )

+ (PlQlRgl + P1Q1RP291 + P1Q1RQ2R91 + P1Q1RQ2RP291 + ••. )

2 - 2 2 - 2 )
+ (Pl Ql R91 + pl Ql RP 291 + pl Ql RQ2 R91 + pl Ql RQ2 RP 291 + •••
3- 3 3 - 3 )
+ (Pl Q1R91 + pl Q1RPi91 +pl QlRQ2R91 +pl a,Ra2RPi91 + •••

+ ••• +
- 2 3 2 3
= Q1R9 1(1 + p1 + p1 + p 1 + •") + Q1Rp 29 1 ( l + p 1 + p 1 + p 1 + ••• ) +
- 2 3
Q1 RQ2 R9 1 ( l + p 1 + p 1 + p 1 + •'•) + ••. •

Since,

2 3 1
l + p 1 + p 1 + p 1 + ... = 1-P = Q'
1 1

therefore,

a1'R91
+ + •••
a,
=R91 + RP291 + RQ2R91 + RQ2RP29l + •••
2 2
= (R + R p 2 )9 1 + RQ2 (R + Rp 2 )9 1 + R Q2 (R + Rp 2 )9 1 + •••
2 2
=CR+ RP2)(9l + RQ291 + R Q2 91 + ••• )

=(R + RP2) (i-~Q2)91

=~:RR~2jgl = 91'
29

Sum of all terms for h :

(p 1Q 1Rh 1 + p 1Q1Rp 2h1 + p 1Q 1RQ2 Rh l + p l Q l RQ2Rp2h1 + ••• )

2 - 2 2 - 2 )
+ (P a R2h + P a RP 2h +Pl a Ra R2h +Pl a Ra RP 2h + •••
1 1 1 1 1 2 1 1 2 1 1 2 2 1
+ •••
- 2 2
=h a R(P +2P + ••• )+h a RP (P +2P + ••• )
1 1 1 1 1 1 2 1 1
- 2
+h 0 Ra R(P +2P + ••• )+ ••••
1 1 2 1 1
Since,
2 2
Pl + 2P + ···= P (1 + 2P + 3P + ••• )
1 1 1 1
pl pl
= (1-P )2 = ;2 ,
1 1
therefore,

Sum of al I terms for g :

(QlRQ2R92 + QlRQ2RP292 + QlRQ2RQ2R29 2 + QlRQ2RQ2RP2 29 2 + ••• )

30
2 2 2 -
+ ••• )+(Pl Ql RQ2Rg2 +pl Ql RQ2RP2g2 +pl Ql RQ2RQ2R292
2
+pl Ql RQ2RQ2RP2 292 + ••• ) + •.• +
. - 2 2
= 92Q 1RQ2R(l +pl +pl + ••• ) + 92Q 1RQ2RP2(l +pl +pl
- 2
+ ••• ) + 29 2Q1RQ2RQ2R(l +pl+ pl + ••• ) + •••.

= g Ra R+ g Ra RP + 2g2RQ Ra R+ 2g Ra Ra RP + •••
2 2 2 2 2 2 2 2 2 2 2
,. - 2 2 2- 3 2
=g (RQ R+R a P +2R a R+2R a P + ••• )
2 2 2 2 2 2 2
- - 2
= g RQ~(R + RP +2RQ R + 2R a P + ••• )
2 2 2 2 2
- 2 2 -
= g Ra (R + RP ) + 2g R a (R + RP ) + •••
2 2 2 2 2 2
- 2 2
= (R + RP )(g RQ + 292 R a + ••• )
2 2 2 2

. = (R + RP 2) ( RQ2 2\ 92
\Cl-RQ ) ~;
2

- 2
RRQ R P a
2 2 2
- g + g
- (1-RQ )2 2 (1-RQ )2 2
2 2

= g (RQ2 (1-RQ2))
2 (1-RQ2)2 .

Sum of all terms for h :

2
31

+(Pl Ql RP2h2 + PlQlRQ2RP2h2 + PlQlRQ2RQ2RP2h2 + ••• )

2 2 2
+(Pl QlRP2h2+Pl QlRQ2RP2h2+Pl QlRQ2RQ2RP2h2+ ••• )

+ ••.
:-
2 2
= h2QlRP2(l +pl+ pl + ••• ) + h2QlRQ2RP2(l +pl+ pl
2
+ ••• ) + h2QlRQ2RQ2RP2(l +pl+ pl + ••. ) + •••

= h2 RP2 + ~RQ 2 RP 2 + h2Ra 2Ra RP + •••

2 2
2 3 2
= h2(RP 2 + R Q2P 2 + R Q2 p 2 + ••• )
- 2 2
= h2RP 2(1 + RQ + R a + ••• )
2 2

= h2 c~=~J.
Sum of all terms containing four consecutive inspections:

[4(Q R) + 4 (QlRP ) +4 (2Q Ra R) + 4 (2Q Ra RP ) + ••• ]

1 2 1 2 1 2 2
+ [ 4(P a R) + 4(P Q RP ) + 4(2P a Ra R) + 4(2P a Ra RP )+ •••
1 1 1 1 2 1 1 2 1 1 2 2
J
r 2 - 2 2 - 2
+ L4(P 1 Q lR) + 4(P l Q lRP 2) + 4(2P l Q 1RQ2R) +4(2Pl Ql RQ2 RP 2)+ •••
]

+ •.•
- 2 2
= 4Q lR(l + pl + pl + ••• ) + 4Q lRP 2(1 + pl+ pl + ••• )
- . ·2
+ {4)(2)Q R.a R(l + Pl + P + ."..) + •.•
1 2 1
- . . - 2 2 2-
= 4R + 4RP 2 + (4)(2)RQ 2R + (4)(2)R a 2P 2 + (4)(3)R a 2 R
3 2
+ (3)(4)R a P + •••
2 2
-- 2 2 . 2 2
= 4R(l + 2R~ + 3R a 2 + ••• ) + 4RP 2(1 + 2RQ 2 + 3R a 2 + ••• ) + •••
·~,
32

4R 4RP2
= +---
(l-RQ2)2 (l-RQ2)2

4 (l-R) + 4R (l-Q )
2
=
(l-RQ2)2

4
= l•RQ = d =average number of times that a group of four consecutive
2 units is examined.

Therefore,

u + fv

Since,

AFI = u + d + fv
u+d+v'

and
AOQ = p (1-A_F I),
33

then,

AO Q = (l _ u + fv + d)
p u+v+d

= ( v (1-f) )·
p \u+d+v

Thus, by substitution, ~~)(~-~k+.4)(1-qk+4 + q4)(1-f)

AOQ = p
1-ii + 4~( 1k+)<1-qk+4 + q4) + t+4
pq ~ ~ 1-q y 1-q

qi ( 1k+4.) (1-q k+4 + q 4)(1-f)

1-q
=P 1

j j ( 1 ) k+4 4 4f I
f (l-q ) + q k+4 (l-q +q )+ 114
1-q 1-q

q i (1-q k+4 + q4 )(1-f) ]

[

This above AOQ formula is different from the formula of:

(1-f) qi (2-qk+4) ]
AO Q = P [ i - k+4 i k+4 i k '
f (1-q )(1-q ) + q (2-q ) + 4fpq q
4
given in error by Dodge and Torrey. In the Dodge and Torrey formula, the

apparent assumption was that in the part of the cycle requiring four units

4
Dodge and Torrey, op. cit., p. 11.
34

to be inspected, 100 per cent inspection of i units was initiated if a defective

unit was found in fewer than four units.

CHAPTER IV

DISCUSSION OF THE STUDY

This chapter will describe the techniques used for obtaining the results,

a verification of their accuracy, and, finally, a summary of the results.

I. TECHNIQUES USED FOR OBTAINING THE RESULTS

A program was written for the I BM 7090 computer to generate AO Q

and AFI values. This program is outlined in Appendix A. As shown in Appendix B,

tables representing 100 different AOQ and AFI curves were generated for

various levels off and i.

Choice of i, f, and p

In a study of this type, realistic and practieal values should be chosen

for the variables under consideration. For example, the AOQ and AFI ~clues

generated when i exceeds 200 units would be of little use in most instances.

Also, the general shape of both the AOQ and AFI curves for values of i greater

than 200 units becomes obvious by examining AOQ and AFI curves for values of

i slightly less than 200 units. Therefore, AOQ and AFI tables were generated

using i values of 5, 10, 20, 30, 40, 50, 60, 80, 100, and 200 units.
36

Another variable for which the range was limited in this study was

f, the fraction of units sampled. The sampling i:ates used began with 5 per cent

and terminated with 50 per cent, in increments of 5 per cent. In earlier works

of Dodge and Torrey concerning CSP-1 and CSP-2, the upper limit off was
1
established at 50 per cent. In many instances a sampling rate greater than

50 per cent would be of little practical value because the cost of administering

the plan would be more than the savings realized by use of the plan.

The choice of the levels of p, the per cent defective, was dependent

upon the level of p at which the AOQL occurred. Initially, a program was

written with p varying from 0.5 to 20.0 per cent in increments of 0.5 per cent.

However, for small values of f and i, the AOQL approximations were not

reached. In order to obtain these needed AOQL approximations, the program

was modified with p varying from 0.5 to 50.0 per cent in increments of 0.5

per cent. By using increments of p of 0.5 per cent, AOQL values could be

approximated closely enough for practical purposes.

1
Harold F. Dodge, "A Sampling Inspection Plan for Continuous Pro-
duction," The Annals of Mathematical Statistics, 14:272, September, 1943;
and Harold F. Dodge and Mary N. Torrey, "Additional Continuous Sampling
Inspection Plans," Industrial Quality Control, 7:8, March, 1951.
37

The Program

The computer was programmed in Fortran IV language to generote

AOQ ond AFI tables for various levels off, p, and i. The-program was written

in such a manner that AOQ ond AFI volues were generated for a given i at

various levels of f and p with only one data card. By writing the program in

this manner, AOQ and AFI values could be generated for any values off, p,

and by merely changing the data cards without altering the program itself.

Also,. by restricting the increments of p to smaller values, closer AOQL

approximations could be obtained.

In a study of this nature, it is important to estoblish the accuracy of

the results. The only errors likely to be encountered would be programming

errors. In order to verify the program, several AOQ and AFI values were

determined with the use of a desk calculator. The maximum difference between

the programmed values and the desk calculator values was found to be 0.000013.

This close agreement between programmed and calculated values was inter-

preted as prima .facie evidence that the AOQ and AF I values generated were

accurate.
38

II. RESULTS OF THE STUDY

The following section of the study consists of a discussion of the results

and examples of AOQ and AFI curves. Also, a chart for determining f and i

values for a given AOQL is presented.

AOQ Results

When studying the AOQ results, it is important to remember that this

relationship between the incoming and outgoing. quality is based on two funda-

mental principles: (1) the AOQ values are not to be applied to an isolated

finite quantity of production, but instead, should be applied to an overall average

of many production runs, and (2) ar.iy defective units found are to be repaired

or replaced with good units. So, the AOQ may be thought of as the long-run

percentage of defective items finally accepted.

Figures 3 and 4 are graphic examples of AOQ curves. AOQ values

used to derive these curves as well as other AOQ values are presented in

Appendix B in tabular form. The AOQ curves displayed in Figure 3 show

the effect of varying i while holding f constant at 30 per cent. The important

fact concerning the AOQ curves pictured in Figure ·3 is that for a constant f,

the AOQL values decrease as i increases. Also, as i increases, the level of

p at which the AOQL occurs decreases. For example, when i= 5, the AOQL

is about 9 per cent and occurs at p = 24 per cent; however, when i is increased
39

10.0-

,; = 5
Y.0

..,. 8.0
c
"'
u
~ 7.0

~6.0
c
0 5.0
Cl
c
·c;
"3"' 4.0
0

"'"'
1:! 3.0
"'~
2.0

1.0

10 15 30 35
Per Cent DefectivP., p

FIGURE 3

AOQ CURVES, VARYING, f = 0.30

10.0

9.0•--

...
c
8,0

" 7.0
u
~
a.
c
~6.0
0
:>
"Dl 5.0
c
·c;
-5 4.0
0
"w3.o
~"
2.0

1.0

0
5 10 25 30 35
Per Cent [)efP.r.tive, p

FIGURE 4

AOQ CURVES, f VARYING, i = 20

to 20, the AOQL decreases to about 3.2 per cent with p also decreasing to

about 8 per cent.

Presented in Figure 4 are examples of AOQ curves with i= 20 and

f varying. It is significant to note that the AOQL values are reduced as f

increases. The level of p at which the AOQL occurs, however, changes only

slightly when f varies, compared· to the changes shown in Figure 3 with i

varying and f constant. For example, when f = 25 per cent, the AOQL is about

3.6 per cent and occurs at p = 8 per cent, but when f is increased to 50 per

cent, the AO QL is reduced to about 1.8 per cent with p decreasing to only

about 7 per cent. In summary, an immediate conclusion from Figures 3 and 4

is that AOQL values are reduced as values off and i increase.

AFI Results

Another graphic device often used in conjunction with continuous sampling

is the AFI curve. These curves show the fraction of the total items inspected

as a function of the inccming per cent defective. Fraction-inspected curve.s

are often very useful to management as a guide in estimating inspection costs

and scheduling inspection workloads.

Figure 5 displays a group of AFI curves with i varying and f held constant

at 30 per cent. As expected, for a constant p the total fraction of units inspected

increas·es as- i increases. Figure 6 shows a group of AFI curves with f varying
42

100

..,.c 80
"u
:;;
a. 70
c
-0
2 60
u
"~
a.
50
.e
·c:
::::>
~ 40
0
c
-~u
30
E!
u..
] 20
,_0
10

0
5 10 15 20 25 30 35
Per Cent Defective, p

FIGURE 5

AFI CURVES, VARYING, f = 0.30

100
·~

-;::-
c:
Q)
u
Q;
a. 70
~
-0
2
u
Q)
a.
~
50
.t!
·c:
::>
0 40
c:
.!!
u 30
e
u..
0
0 20
I-

Per Cent Defective, p

FIGURE 6

AFI CURVES, f VARYING, i"" 20

and i held constant at 20 units. Obviously, for a given level of p, as f increases

the total fraction of units inspected also incr-eases.

Curves for Deter-mining Values off and i for a Given Value of AOQL

l·n order- to obtain curves for determining values off and i for a given

value of AOQL for CSP-3, it was necessary to expand the computer program

which was previously used for- purposes of deriving AO Q values appearing

in Appendix B. Since the curves in Figure 7 represent only integral AOQL

values ranging from 2 to 10 per cent, it was necessary to generate a great

number of AOQ curves in order to extract the AOQL values needed to plot

accurately the curves shown. After an AO QL is es tab Ii shed, the curves in

Figure 7 may be used to determine the most desirable combination off and

i for a given situation. For example, an AOQL of 5 per cent may be achieved

with f = 20 per cent and i = 17 units, or with f = 3 per cent and i = 41 units.

Although previously discussed, the curves in Figure 7 clearly reemphasize

the fact that fora constantvalueof f, the AOQL values are reduced as i increases.

In addition to the curves for determining values off and i for a given

value of AOQL for CSP-3, Figure 7 also shows the corresponding curves
2
for CSP-2. An examination of the curves shows that for a given value of
45

i, Number of Units

FIGURE 7

CURVES FOR DETERMINING VALUES OF f AND i

FOR A GIVEN VALUE OF AOQL
46

AOQL and f, the values of i for CSP-3 are decre~ed from their corresponding

CSP-2 values. For example, if. AOQL =2 per cent and f = 10 per cent, the

value of i for CSP-3 is 68 units, whereas i = 72 units for CSP-2. Also, from

Figure 7, it is shown that the difference between the i values of CSP-2 and

CSP-3 at 2 per cent is about 4 units; the difference at 3 per cent AOQL is

about 3 units; and for AOQL values greater than 3 per cent the difference

is about 2 units.

Dodge and Torrey stated that the CSP-:3 values of i for. a given AOQL

arid f are equal to those for CSP-2 when AOQL is less than 2 per cent and are

less than those for CSP-2 by no more than 2 units when AOQL is 10 per
3 .
cent. A difference of about 2 units at 2 percentAOQLwasnotedbetween

the Dodge and Torrey study and the results of this study; otherwise, there

is close agreement between the two studies. The slight differences between

the two studies mentioned above could probably be attributed to the use of

different AOQ formulae, as discussed in Chapter Ill.

In summary, these differences between CSP-2 and CSP-3 may or may

not be significant depending upon the size of i, the cost of inspection, and the

avail ab ii ity of inspection manpower.

3
1bid., p. 10.
47

CHAPTER V

SUMMARY AND RECOMMENDATIONS

This study was prompted by the need to know AOQ and AFI values for

CSP-3 instead of CSP-2 approximotions.

I. SUMMARY

Before AO Q and AF I values were generated, it was necessary to verify

the CSP-3 AOQ formula. ·The proof revealed AOQ to be

qi (1-qk+4 + q4)(1-f) ]"

[

1
This formula is different from the formula given in error by Dodge and Torrey.

Because of the complexity of the AOQ formula and time-consuming

calculations involved, a computer program was written in Fortran IV lanuage

for the IBM 7090 computer in order to determine AOQ and AFlvalues. For

each set of values generated, the per cent defective, p, ranged from 0.5 to

50.0 per cent in increments of 0.5 per cent. The values of i used in this study

ranged from 5 to 200 units, and the sampling rate, f, began with 5 per cent

1
Harold F. Dodge and Mary N. Torrey, "Additional Continuous Sampling
Inspection Plans," Industrial Quality Control, 7: 10, March, 1951.
48

and terminated with 50 per cent at increments of 5 per cent. An assumption

which was made throughout this study is that i = k. The computer program

was written in such a manner that AOQ and AFI values could be generated

for a given i value at various lev·elsof f and p with only one data card. Therefore,

AOQ and AFI values other than those appearing in this study may be generated

with this computer program.

In addition, curves for obtaining values of f and i for a given value of

AOQL were determined. When compared to the corresponding CSP-2 values,

an examination revealed that the i values for CSP-3 were reduced by about

4 units when AOQL is 2 per cent; at 3 per cent AOQL the CSP-3 i values

were about 3 units less than the CSP-2 values; and for AOQL values greater

than 3 per cent, the i values for CSP-3 were about 2 units less than the CSP-2

values. Whether or not these differences are significant depends upon the

individual situations. For example, if i is small, a difference of 2 units may

be significant, especially if inspection costs are high. Also, in situations

,_.

where i is small, a difference of 2 units may or may not be significant depending

upon the availability and flexibility of inspection manpower.

II. RECOMMENDATIONS

Although there is an abundance of literature pertaining to continuous

sampling inspection plans, there are several areas which could warrant further

study and investigation. Some of these unexplored areas concerning CSP._3

include the fol lowing suggestions;

1. Since the first level of the H-106 multi-level plan is very similar

to CSP-3, a d_irect comparison of the two plans' AOQ, AFI, and

curves for determining f and i would be beneficial. From a study

of this nature, one could determine the values off and i for which

H-106 and CSP-3 were most similar and most different.

2. As discussed in Chapter 11, theMLP inspection scheme is a multi-level

extension of CSP-1. A similar extension of CSP-3 could be a

beneficial investigation.

·3. The CSP-3 plan states that upon finding a defective item, inspect

the next four items. An investigation of the effect of inspecting

n items, where n .j 4, after finding the first defective item could

be of some practical value. The AOQ formula for CSP-3 could

be generalized for any value, n, by replacing the number 4 with n.

Some preliminary work with values other than 4 has been done
2
by Dodge and Torrey, but these works are unpublished.

4. In 1957, C. C. Thigpen investigated the effect of i not equal to k

3
for CSP-2. A similar study for CSP-3 and a comparison of these

results to the results of the CSP-2 investigation ~ould be bene-

ficial.

2
statement by Harold F. Dodge, personal interview.
3
c. C. Thigpen, "Extension of the Dodge CSP-2 Continuous Sampling Plan
to the Case i Not Equal to k" (unpublished Master's thesis, The University of
Tennessee, Knoxville, 1957).
>-
:::c
a..
~
<.!>
0
_J
co
co
...
52

BIBLIOGRAPHY.

Derman, C., M. V. Johns, and G. J. Lieberman. ;'Continuous Sampling Pro-

cedures Without Control," The Annals of Mathematical Statistics, 30: 1175-
1191, December, 1959.
Derman, C., S. Littauer, and H. Solomon. "Tightened Multi-Level Continuous
Sampling Plans," The Annals of Mathematical Statistics, 28:395-404, March,
1957,

Dodge, Harold F. "A Sampling Inspection Plan for Continuous Production,"

The Annals of Mathematical Statistics, 14:264-79, September, 1943.

Dodge, Harold F. and Mary N. Torrey. "Additional Continuous Sampling

Inspection Plans," Industrial Quality Control, 7:7-12, March, 1951.

Girshick, M. A., Technical Report No. 16, Applied Mathematics and Statistics
Laboratory, Stanford University, July, 1954.

Inspection and Quality Control Handbook (interim) H-106. Multi-Level Con-

tinuous Sa mp I ing Procedures and Tables for Inspection by Attributes,
A Report Prepared by the Department of the Navy. Washington: Office
of the Assistant Secretary of Defense, 1958.

Lieberman, Gerald J. "A Note on Dodge's Continuous Inspection Plan." The

Annals of Mathematical Statistics, 24:480-84, September, 1953.

Lieberman, Gerald J. and Herbert Solomon. "Multi-Level Continuous Sampling

Plans," The Annals of Mathematical Statistics, 26:686-704, December, 1955.

Thigpen, C. C. "Extension of the Dodge CSP-2 Continuous Sampling Plan to

The Case i Not Equal to k. 11 Unpublished Master's Thesis, The University
of Tennessee, Knoxville, 1957. .

Wald, A. and J. Wolfowitz, "Sampling Inspection Plans for Continuous Pro-

duction Which Insure a Prescribed Limit on the Outgoing Quality," The
Annals of Mathematical Statistics, 16:30-49, March, 1945.
53

White, Leon S. "The Evaluation of H-106 Continuous Sampling Plans Under

the Assumption of Worst Conditions," Journal of the American Statisticc;:il
Association, 61 :833-41, September, 1966.
V>
w
x
0
z
w
a..
a..
<(
' '
55

APPENDIX A

THE COMPUTER PROGRAM

DIMENSION PPllOCl,AAllOOl,AFillDOl
IT AB#C
2 READ( 5, I IPS,PE,PD,FS,FE,FO.,K
I FORMATl6FI0.3,I51
Flff S
5 P#P S
L #I
3 Q #I .-P
QK #QtHK
QP #QtH( K+4 I
QF#QiH4
A#P * ( I .- FI •QK• ( I • +QF-QP I I ( F * (I • -QK I * 11 • -QP I+ QK • ( I.• +QF-QP I +4. •F•P•Q
IK I
PP ( L I #P
AAILl#A+l.E-E
AF I IL I# I .- I A /P I
L #L +I
P #P +P 0
IF( P-PE 13, 4, 4
4 ITABllITAB+I
100 FORMAT<IHI/ IH0,42X,SHTABLE,75X,I4/IH0,34X,28HAOQ ANC AFI VALUE
IS FOR CSP-3/IH ,40X,IH#,F6.3,8H AND #,14/IH0,19X,IHP,8X,3HAOQ,8X
2,3HAFI,12X,IHP,ex,3HAOQ,8X,3HAFI/ IH)
200 FORMAT< IHI/ IH0,42X,5HTABLE,75X,I4/IH0,3~X,29HAOQ AND AFI VALUE
IS FOR CSP-3/IH ,40X,IH#,F6.3,8H AND #,I3/IH0,19X,IHP,8X,3HAOQ,8X
2,3HAFI,!2X,IHP,EX,3HAOQ,8X,3HAFI/ IHI
IFIK.GE.ICdl wRITE (6,1001 ITAB,F,K
IFIK.LT.ICOI WRITE (6,2001 ITAB,F,K
00 20L #I , 50
WR ITE ( 6, I CI I PP ( LI , AA IL I , AF I ( Ll , PP ( L+ 50 l , AA ( L +SO I , AF I I L +5 JI
I 0 I FORM AT I I 7 X, F 6. 3 , F I 1 • 6, F I I • 6 , F I .3 • 3 , F I I • 6 , FI I • 6 I
20 CONTINUE
WRITEl6, 1021
102 FORMATllH9,15XI
F#F+t=D
IF I F- FE I 5, 2, 2
E\I D
56
APPENDIX B

p AOQ AF! p AOQ AFI

C.C05 O.Q04745 'J. n51 063 n.2ss 0.140489 0 ,449063
C. CIC fJ.009477 i].052338 0.26D 0.138715 0. 46 64 79
C. [I 5 o. 0141 93 n.o53S08 8.265 0.136712 0.4E4108
c.c2c 0.018891 Cl. 055461 ;J.27:J o. 134486 0.5[1905
(.[25 [,023568 !],057292 u. 2"f5 0.132047 o. s 19829
[.[3[ 0.023221 !). 059297 0.280 0.129407 o.537832
[ .(35 0.032848 n.061475 0.285 0.126578 0.555866
[,[4[ 0.037447 o. 06 ·rn26 iJ.290 o. 123574 0. 5 7 38 84
C.C45 0.042014 0.[)66354 0.295 0.120408 0.591836
c.csc C:.046547 a. 069061 0. 30Q 0. 11 7098 0.609673
C.C55 C.051043 J. rl7 I 954 o. 305 0.113659 0.627347
[.[6[ 0.055498 o. n75037 0.310 0.110109 0.644811
[.(65 Q.059909 D. 078.319 0.315 0. 106464 0. 6~20 I 9
[,[7[ 0.064274 o.081so5 o. 32'.J 0. 102743 0 .678927
[.[75 0.063587 0.085505 0.325 0.098964 0.695496
C.C3C 0.072846 ~). 089427 0.330 0.095143 0.711688
C.C85 C.077046 n.!193582 0.335 0.091299 '). 7 2 74 66
C.[9[ O. 081 I 82 0.097979 D.340 0.087447 0. 7 4 28 02
[.(95 0.085250 !] • 02629 0.345 0.083605 0. 757667
[.10[ o. 089246 o. 07544 0.350 fJ .079787 0. 7 7 20 38
C.lll5 O.C9~"3163 o. 12735 0.355 IJ.076007 0.785896
C.I IC 0.096996 !] • 18214 o. 360 0.072279 0.7<19225
C. II 5 c. 00741 fJ. 23994 0.365 o.n6s61s 0. 8 120 I 4
o.12c o. 04389 o. 30088 o. 370 o.065026 n.s24254
C. I 2 5 o. 07936 f]. 36508 fl.375 0.061522 IJ.8~5942
C. I 3C c. I I 37:, D. 43269 0. 38 fJ 0.058111 0. 84 70 77
[.135 o. 14698 t1. 50382 0.385 0.0548(]0 0.857662
0.14[ lj. I 7899 o. 57862 D.390 0.051597 0.8677[)1
c .145 o. 20970 o. 65722 0.395 0.()48505 0.877203
[.15( o. 23904 n. 73975 0.40Q 0.045529 0.886177
C.155 o. 26692 1. 82632 0.405 0.042672 0.8Cf4637
C.16[ o. 29327 o. 91 707 0.410 D.039936 0.9[2596
C.165 c. 31 800 n.201211 0.415 0.031321 0. 91OD71
C. I 7C [. 34104 f).211154 0.420 n.034828 0.91 7077
[.175 Q. 36230 !J. 221 545 0.425 D.032456 0. 92 ~6 .32
0.18[ o. 38169 n. 232 392 0.430 0.030205 0.929756
C. I 0 5 c. 3?915 Cl. 211 3 703 0.435 n.02ao12 0. 9 3 54 6 7
C. I 9[ o. 41458 0.255482 0.44J 0.026055 0. 94Cl7 85
[.195 [. 42 792 0.267733 0.445 0.024151 0.945728
c.2oc o. 1~3909 !1.280456 "J.45iJ 0.022357 0. 95 03 I 7
c.2os (.I Ii I~ 802 n.293A'19 0.455 n.D2f1671 0.954570
c.21c 0.145465 0.307309 fJ.460 0.019087 n. 9585 01
c.21s [,145893 :J.32lli29 0.465 0.017602 0.962145
o.22c 0.146080 :J. 335998 0.470 0.016213 f).9t5504
C.225 Q. I l16024 D. 351 004 0.475 IJ. 011~9 I 5 0. 968600
C.23C C.145721 o. 366431 0.490 0.013704 0.971450
[.235 C.145169 [).382259 0.485 0.012576 D.974070
C.24( 0.144369 o. 39·3464 J.490 n.011527 0. 9764 76
C.245 C.143320 fl. 415021 0.495 0.010552 [). 97 86 8 3
[.25[ C. I 42 02 5 !1.431898 n.500 o.no9648 0.9E0705
67

TABLE XII

AOQ AND AFI VALUES FOR CSP-3

f = 0. 100 AND i = 10

p AOQ AFI p AOQ AFI

[.cos o. 004490 !J.102012 0.255 0.093724 0 .632454
C.CIC 0.008956 0.104419 0.260 0.091362 0.61111609
C. CI 5 C.013392 0.107186 0.265 0.088895 0.664547
c.02c o. 01 7794 ri.110209 o. 270 0.086337 0 .6·E02 32
[.[25 C.022157 0.113712 0.275 0.083702 0 .6'75629
0.(3( o.n26477 0.117445 0.280 0.081001 0. 710709
C.C35 D.C3fJ748 0.121492 0.285 0.078249 0. 7 2 54 42
[.(4[ 0.034967 n. 25821 0.290 0.075458 0. 7 3 9800
[.[45 0.039129 o. 3Q462 0.295 0.072640 0. 75 .?:7 61
[.[5[ 0.043230 f]. 35406 0.300 0.069809 0.767304
c.C55 C.047264 o. 40657 0.305 0.066974 0. 7 8 04 I I
[.(6[ 0.051227 iJ. 46221 0.310 0.064149 0. 793068
C.C65 C.055113 o. 52103 0.315 0.061342 0 .8C5263
C.C7C 0.058918 J. 58310 0.320 0.058564 D.816986
C.C75 C.062636 !J. 64849 0.325 0.055824 0. 82 82 33
(.(8( [.066262 o. 71727 o. 330 0. 05 31 30 0. 83 9000
c.C85 C.069789 1. 78954 o. 335 0.050489 CJ.849287
0.(9( 0.073212 o. 86537 0.340 0.047908 0. 85 90 95
C.C95 0.076524 fJ. 194484 0.345 0.045392 0. 8684 30
c. DC C.079720 n.202so4 iJ.350 0.042946 o.e77296
r.. 05 o.082792 n. 21 I 503 0.355 0.040576 0 .8'E5702
o. IC o. 085.735 0.220589 D.360 0.038283 0. 3'73659
(. I 5 c. 088542 f).~30069 a. 365 0.036071 0.901176
c. 2[ 0.091206 0.239947 o. 37[] 0.033941 0. 908267
c. 25 C.093721 n.250230 0.375 0.031896 0 • 9 I 49 4 4
o. 3( 0.096080 o. 260921 0.380 D.029935 0.921223
( . 35 C.098277 o. 272 021 0.385 0.028060 0.927117
c. 4( o. 00305 0.283532 o.390 D.026270 0. 93 26 42
c. '• 5 c:. 02159 :1. 295454 0.395 0.024564 0. 9 3 78 I 4
o. 5( o. 03832 o. 307784 0.400 O.fJ22941 0.911264A
c. 55 o. 05320 0.320517 0.405 0.021400 0.947161
[. 6( o. 06617 0.333646 0.410 0.019939 0.951368
c. 65 c. 07718 CJ.3471.64 0.415 0.018557 0.9~5285
c. 7[ o. 00620 o. 361 059 n.42CJ 0.017250 o.9~e928
c. 75 o. 09319 o.375317 :J.425 0.016018 0 • 96 23 I I
(. 8[ Q. 09814 o. 389923 0.430 D.014857 0. 9{; 54 49
"'
L 85
• c. I lJ I UI Q. 404857 0.435 0.013765 n.968357
c. 9( o. I 0181 :1. 42 0099 0.440 0.012739 0.971048
r. 9': Q. I 0053 J.435624 0.445 0.011776 0.973536
[ .20( c. 09719 n.451?+07 1).450 0.010875 0. 97 58 .14
C.205 [. 09179 0.467418 0.455 D.010031 0.977954
o.21c o. 08438 iJ.483627 0.460 0. 009 243 0. 97 99 07
[. 21 : c. Q7500 0.500000 0.465 0. 008507 0 .981704
o.22c Q. 06369 0.516503 o. 47J 0 .007822 0.9E.!358
[.225 c. 05053 fJ.533098 0.475 n. no11 a4 0.9E4376
C.23( o. 03558 o. 549749 J.480 0. 006590 o.9e6270
C.235 c. 01 892 0.566416 0.485 0.006039 0.987548
C.24C c. 00066 !J.583060 0.490 n.no5528 0.9E8718
[.245 C.098088 0.599640 0.495 O.D051J55 D.9E9788
C.25[ 0.095971 n.616110 0.500 D.OCJ4617 0.9<;Q766
68

TABLE XIII

. AOQ AND AFI VALUES FOR CSP-3

c.cos o. 004 737 o. 052571 IJ. 255 0.000049 0.999809
C.CIC 0.009422 I]. 05 7751 0.260 0.000038 0. 9'798 55
c.c1s 0.014022 0.065177 0.265 0.000029 0.999890
c.c2c C.018503 o. on866 0.270 o. 011n1J 22 n. 99 99 11
c.C25 0.022825 [1.1)87020 0.275 0.000017 0. 9<; 99 37
C.C3C C.026942 0.101946 0.280 0.001013 0.99995~
[.[35 0.030799 0.120021 0.285 0.000010 0. 9999 64
C.C4C 0.034334 0.141656 0.290 0.000009 0. 9999 73
C.C45 0.037473 n.167267 0.295 0.000006 D.999980
c.C5[ o.040139 0.197226 o. 300 D.000005 0. 99 99 85
c.055 0.042250 0.231810 n.305 0.000003 0. 9999 89
[.[6[ 0.043732 ll.271137 0.310 n.000003 0. 9<; 99 92
[.[6.5 0.044519 o. 31 5095 0.315 0.000002 0. 9c;99 94
C.C7C 0.044570 o. 363291 0.320 n.000001 0. 9c; 99 95
c.C75 0.043874 0.415018 0.325 O.OO!JOOI 0 .999997
o.cac 0.042459 n. 469264 0.33'] o. oor.100 1 0. 9999 97
[.[85 0.040394 Cl. 524 775 0.335 0.000001 0. 9999 98
[.[9[ 0.037786 0.580158 0.340 0.000000 0. 9999 99
C.C95 o.034768 o. 63401 7 0.345 0.000000 0. 9999 99
o. DC 0.031491 0.685095 0.350 0.000000 0. 9999 99
c. 05 0.028100 o. 732377 0.355 0.000000 0. 99 99 99
c. IC 0.024732 0.775167 0.360 0.000000 .000000
c. 15 0.021494 0.813093 0.365 0.000000 .ornono
o. 2[ Q.018470 o. 846086 o. 370 n.oonooo .000000
c. 25 0.015711 o. 874.314 0.375 0. 000000 .OCQOOO
c. 3C C.013245 '.).898118 0.380 o.ooonoo • 0000 DO
c. 35 0.011078 fl.917943 0.385 0.000000 • 0000 on
o. 4[ o. 00920! 0.934278 0.390 0.000000 .ooonoo
c. 45 0.007596 0.947616 0.395 0.0011000 • ocoooo
c. SC 0.006237 o. 958422 0.400 0. Oor:JOOO • OCOD 00
c. 55 0.005096 0.967120 0.405 0.000000 .000000
o. 6C 0.004147 0.974082 0.410 0.0[10000 .OOOOfJO
c. 65 D.003361 0.979629 0.415 0.000000 .000000
c. 7C 0.002715 o. 984030 0.420 0.000000 .000000
c. 75 0.002186 0.987510 0.425 0. 000000 • DODO DO
0. BC c. 001 755 0.990252 0.430 0.000000 .000000
[ . 85 O.QOl404 0.992408 0.435 0.000000 1 • ocoo nn
c. 9[ c.001121 o. 994 099 0.440 0. 000000 .000000
c. 95 0.000893 o. 995421 0.445 0.000000 .000000
o.2oc i:;.nnn109 o. 996453 0.450 0.000000 .000000
C.205 0.000562 0.997257 0.455 0.000000 .ucoooo
C.21C 0.000445 J. 997883 0.460 0.000000 .000000
c.21s 0.000351 o. 998368. 0.465 0.000000 .000000
0.22[ 0.000276 0.998745 0.470 0.000000 • 0000 00
C.225 c.000211 o. 999036 0.475 0.000000 • oc:oo 00
C.230 Q.000170 CJ.999261 0.480 0. (JOQOOO • ncoo oo
[. 2 3 5 0.000133 0.999434 0.435 0. OllOOOO .000000
C.24C Q.000104 0.999567 0.490 0.000000 .000000
C.245 o. 000081 o. 999670 0.495 0.000000 • OGOO 110
c.2sc C.00006~ n.999749 0.500 o.nooono • nooo no
97

TABLE XLll

AOQ AND AFI VALUES FOR CSP-3

f = 0. 100 AND i = 40

p AOQ AFI p AOQ AFI

[.[05 0.004476 Cl. I 04859 o. 255 0. 000023 0. 9999[)9
Q.[I[ [1.008854 0.114568 0.26Q O.OO'J018 0. 9999 31
C. CI 5 IJ.013075 a.129303 0.265 0.000014 0. 9999 48
[.[2[ 0.017C82 r:J.145912 0.270 0. 0000 I I 0. 9999 6 I
O.C25 o. 020812 :J.167512 0.275 0.000000 a. 999910
Q.[3[ 0.024200 f].193322 o. 28'] 0.000006 0. 9999 78
[.(35 0.027175 0.223564 0.285 o.OOODD5 0. 99 99 8 3
[.[4[ 0.029665 o. 258383 0.290 O.DOOD04 0.999987
[.[45 o.0.316DO n. 297776 D. 295 o.DDODD3 0. 9999 91
[.[5[ [. 032 924 0.341524 0.30'] 0. DD0002 0 .999993
[.[55 0.033597 0.389146 o. 305 0. 000002 0. 9999 95
C.C6C 0.0336D7 o. 439881 0.310 O.DfJOOOI 0. 9999 96
[.[65 D. 032 974 0.492703 0.315 o.OOODDI 0. 9999 97
0.(7( [.031 752 0.546393 0.320 O.OOODDI 0 .999998
[.[75 Q.030D27 n.599638 0.325 O.OOfJOOI 0. 9999 98
C.C8C [!. 027908 0.651154 D.330 D.ODODOO D. 9999 99
C.085 D.025516 O. 69981 I 0.335 O.OOOODD 0. 9999 99
c.c 9[ o. 022 975 0.744718 0.34'] O.DOOOOO 0 .999999
[.[95 o. 020398 IJ.785280 0.345 O.OOODDD .DO ODDO
c. 0[ O.D17880 '.1.821200 0.350 0.000000 • oocooo
c. 05 0.015493 n.852448 o.355 O.DOfJOOD • DODD no
(. IC 0.013287 o. 879206 D.36D D.OODOOO .OOODOO
c. I 5 O. 0 I I 2 92 o. 901 8D6 D.365 o.oDnDDD • DODO OD
c. 2[ o.OD9520 o. 920667 n. 370 O.OOOODO .OODODO
[. 25 0.007969 0.936247 Q.375 O.ODDDOD .OOODOO
c. 3[ 0.006629 ::i. 949006 0.380 O.DDOODO • OOOD DO
[ . 35 G. OD5484 0.959377 0. 38 5 D.DDOODD .ODDO OD
c. 4[ 0.004515 (1.967753 fJ.390 D.ODODDD .oooonD
[. 45 0.003700 o. 974483 o. 39 5 D.DOODDD .ooooori
o. 5[ O.OD302D 0.979864 0.40Cl 0. DDJDDD .oooonD
c. 55 0.002457 o. 9841 51 0.405 D.DDi:JODD .OOODOD
o. 6[ C.OOI 991 !J.987553 D.410 O.ODOODD .OCODOD
[. 65 o.0016D9 o. 99[1246 0.415 D. OD ODDO .ocoonD
o. 7[ Q.DDl297 o. 992 371 D.420 n.oooDDD • 00 OD rJO
c. 75 0.001 042 !J.994D44 0.425 O.OOODDO .oooDiJo
c. 8[ Q.OOD835 i].995359 o. 430 O.ODDDOO • OOODOO
[. 85 o.DOD668 0.99~389 0.435 0. DO DODD • 0Qr10 OD
[. 9[ o. DOD53.3 0.997196 0.440 D.000000 .CJQODOO
[. 95 C.DOD424 D.997826 0.445 O.OOODDD • 000000
[ .20[ O.D00337 D. 9?8317 0.450 0. ODOOOD .DO DODD
C.2D5 0.0[10267 fl.99S699 D.455 D. nDoooD .OOODDD
C.21 [ o.OOD21 I o. 998996 0.460 D. OOIJOOD .CJOODOO
[. 21 : C.OODl66 !J. 999227 0.465 D.DDO:JDD .DOOOOO
C.22C O.OODl31 f). 999405 D. 4 70 o.Doooon .OD OD OD
[.225 D. DOOi D.3 0.999543 0.475 o.ODOODD .OCODOO
[.23[ [}. 000081 [1.999650 D.48Q 0. OD!JODD .DOOOOD
C.235 C.000063 o. 999732 0.485 D.nooDDD .OOODOO
C.240 C. CC0049 ri. 999795 0.49CJ o.OOODDO .000000
C.245 o.Dooo3a !]. 999844 0.495 o.oonDoD .000000
0.25[ O.OOD030 o. 999881 0.5DO D.oonnno .OO!:JOOO
98

TABLE XLill

AOQ AND AFI VALUES FOR CSP-~

f = 0. 150 AND i = 40

p AOQ AF! p AOQ AFl

r:.cos 0.004216 0.156864 0.255 n. DODO I 5 0. 99 99 4 3

C.CIC 0.008295 0.170472 D.260 0. 0000 I I 0. 9999 57
C. CI 5 [:.012158 0.189476 0.265 D.000009 0.999967
o.c2c 0.015132 0.213424 !J.270 0.000001 0. 9999 75
C.025 0.01891!5 0.21!2184 0.275 n .UOLJ005 0. 999981
0.[3[ 0.021729 0.275689 0.280 0. 000004 0.999986
[.035 0.0240!7 0.313804 0.285 0.000003 o. 9999 89
0.[4[ 0.025751 o. 356230 0.290 0. 000002 0 .999992
C.C4'5 0.026890 0.402445 0.295 o .oom102 o. 9999 94
c.csc 0.027416 D.451680 D.300 0.000001 0.999996
[.[55 o. 027339 0.502930 o. 305 0.000001 0. 9999 97
0.[6[ 0.026699 o. 555020 0.310 0. DODOO I D.999998
C.065 0.025565 0.606693 0.315 0.000001 0.999998
[:.[7C 0.024029 0.656724 0.320 0.000000 0. 9999 99
[. C75 Q.022197 0.704033 0.325 0.000000 0. 9999 99
0.[8[ 0.020119 !]. 74 7767 o. 330 0.000000 0.999999
[.[8,'; 0.018075 0.787349 0.335 0.000000 0. 999999
C.C9C 0.015977 0.822483 0.340 0.000000 .000000
[.[95 0.013953 0.853126 0.345 0.000000 .ocoooo
o. DC 0.012056 0.879438 0.350 0. 000000 • 0000 00
c. fJ 5 0.010319 0.901726 0.355 0. 000000 .000000
o. IC 0.008758 o. 920383 0.360 0. 1100000 .000000
[. I 5 C.007378 o. 935841 o. 365 0.000000 • 0000 LlO
o. 2C 0.006176 o. 9405.37 D.370 0.000000 .000000
[. 25 0.005139 o. 958889 0.375 O.OOIJODD • DODO DO
c. .3[ 0.004254 o. 967275 0.380 0.000000 .000000
[. 35 0.003506 o. 9no32 o. 385 0.000000 • 0000 OD
rJ. 4[ 0.002877 o. 979451 0.390 o. oonooo .000000
c. 45 0.002352 0.983780 0.395 0. 000000 • 000000
c. SC 0.001916 0.987227 D.400 O.ODOOOO .000000
[. 55 Q.001556 0.989962 0.405 0.000000 .000000
o. 6[ 0.001260 0.992127 0.410 0. 000000 .000000
c. 65 O.fJOI017 0.993836 0.415 0.000000 .000000
o. 7[ 0.000819 o. 9951 83 0.420 0. 000000 .000000
[. 75 Q.000658 0.996242 0.425 0.000000 .000000
o. 8[ 0.0[]0527 o. 997073 0.430 0.000000 • 0000 00
D. 85 0.000421 0.997724 0.435 0.000000 • ODDO 00
c. 9[ 0.000336 0.998233 0.440 0. 000000 .ocoooo
[. 95 0.000261 0.998630 0.445 0. 000000 I .000000
0.20[ 0.000212 o. Q98940 0.450 0.000000 1.000000
C.205 0 •. 000168 0.999180 0.455 0.000000 1.000000
c.21c [.000133 D.999368 0.460 0. [100000 I .000000
[.215 Q.000105 0.999513 0.465 0.000000 I .000000
0.22[ 0.000082 0.999625 o. 470 0.000000 I • 0000 00
C.225 0.000065 0.999712 0.475 0.000000 I .000000
C.23C o. 000051 n. 999779 0.1!80 0.000000 1.000000
C.235 c.000040 o. 999831 0.485 0.000000 I .000000
0 .240 0.000031 o. 999871 0.490 0.000000 I .000000
[.245 0.000024 o. 999902 0.495 0. 000000 ! .000000
c.2sc 0.000019 o. 999925 0.500 0. 000000 1.000000
99

TABLE XLIV

AOQ AND AFI VALUES FOR CSP-3

f = 0. 200 AND i = 40

p AOQ AF! p AOQ AFl

r:.co5 0.003957 o. 208591 D.255 o.nooD10 0. 999960

c.c1c O.fJ07745 '1. 225486 0.26'J o.oonoo8 0. 99 99 69
C. [I 5 Q.011268 IJ.248783 0.265 0.000006 D. 9999 77
c .c;~r: 0.014447 0.277660 o. 270 0. 00000 5 D. 9999 82
(.[25 c.Dl72D9 D.311646 0.275 o. oonoo1i D. 9999 87
[.(3( O.Ql9490 r:J.350318 0.280 O.OOOD03 D. 99 99 90
C.C35 0.021240 n.393151 0.285 o.oD0002 0. 99 99 92
C.C4C 0.022423 fl.439434 D.290 0.000002 0. 99 99 94
C.C45 o. 02.3028 0.488257 0.295 o.ODODDI D. 9999 96
r.c5c 0.023074 0.538529 0.3DD O.DDfJDOI 0.999997
c.cs: o. 022602 !J. 589047 0.305 0.000001 0. 9999 98
(.(6( 0.021684 n.638598 0.310 0.000001 0 .999998
C.C65 o. 020406 0.686055 0.315 D.OOOOOD 0. 9999 99
[.C7( o.rJ18867 0.730476 0.320 0. 000000 0. 9999 99
C.C75 0.017163 0.771162 0.325 n. ooooDo n. 9999 99
[.(8( 0.015385 0.807686 0.330 0.000000 0. 99 99 99
C.C85 r:.013610 n.839879 o •.335 o.oooooD • IJOOO OD
Q.[ 9( 0.011 899 o. 86 7791 D.340 D.ODnDoo • OC 0!1 DO
C.C95 0.010294 o.891643 D.345 O.OOOODO .000000
o. 0( 0.008823 ri. 91 I 769 D.350 0. ODO DOD .QCOOOD
c. oc C. OC75DI f].929566 0.355 o. oonooo .000000
c. I ( 0.006330 o. 942452 0.360 o.OOODDD • OCOD 00
c. 15 0.0053D8 o. 953840 0.365 O.OOODDO • DODO DO
c. 2( 0.004426 n.963115 0.370 D.000000 • OCOO OD
c. 25 C.003672 0.970625 0.375 D.OOOODO .ODODOD
[. 3[ 0.003032 D.976675 0.380 0. ODO DOD • OCOD OD
c. 35 o. 0024911 o. 981 528 0.385 o.ODOOOO • 00001]0
c. 4( 0.0['2043 o. 9854117 0.390 o.OOODOO .OCODDO
(. 45 C.001668 0.988496 0.395 D.OOOODD .000000
(l. SC 0.001358 0.990950 0.400 D. OOODDD • !JOOO DO
c. 55 0. 001 I 02 o. 992893 0.4DS D. ODO DOD .000000
c. 6( o. 000891 I]. 994 4 30 0.410 O.ODOODO .000000
C.165 C.OC0719 o. 995641 0.415 o.OOODOD .DOOOOO
C. I 7C o.D00579 o. 996595 0.420 o. oDnoDo • 0000 00
c. I 7 5 (1.000465 0.997344 0.425 D.OOODOD .nooooo
C. I 8C C.000372 0.997932 0.43D D. nnrmDo .OCODOD
[.185 o. OC0297 o. 998392 0.435 o.OOODDO • 0000 no
[.19( O.OC0237 0.998752 0.440 o. oormoD .OCODDO
C.195 O.fJGOl89 o. 999032 0.445 0.000000 • OOODOO
c.2oc 0.000150 o. 999251 0.450 D.oonoDo .OOODOO
[.205 r:.000119 O. 99942 I 0.455 0. ODODOO .nDoono
[.21( O.OGOC94 D. 999554 'J.460 D. 000000 • OCOCl 00
C. 2 I 5 C.OG0074 '.1. 999656 0.465 o.nDonoo • OOODOD
c.22c 0.[]00058 o. 999735 o. 4 70 o. rmoDDD .occooo
r. .22 5 C. OC0046 o. 999797 0.475 o.OOODOO .000000
0 .2 3( 0.000036 0.999844 0.48D 0. OOODOD I .OQQOOO
C.235 C.OOD028 o. 999881 'J.485 D.nDOODO I .DODD OD
C.24[ e.000022 0.999909 0.490 D.oonooo I .OCODOO
0.245 C.000017 o. 99993D 0.495 o.ODOODD 1.000000
c .2 5[ O.OODOl3 0.9999117 o.sno l].000000 I.• OCOODO
100

TABLE XLV

AOQ AND AFI VALUES FOR CSP-3

f = 0. 250 AN D i = 40

p "-00 AF i p AOQ A FI

C.C05 0.003700 o. 260041 D.255 o.noooo8 0. 999970

O.CI( o. 007204 0.279631 0.260 0.000006 0. 999977
C. CI 5 c.010405 o. 306 309 0.265 0.000005 0 .999983
Q.[2[ 0.013223 0.3~8851 0.270 0. 000004 0. 9999 87
[.(25 0.015589 0.376424 D.275 I]. 00000 3 0. 9999 90
C.C3C 0.017452 0.418251 0.280 0. 000002 0. 99 99 9 3
C.035 o. 01 8779 0.463464 0.295 0.000002 0. 9999 94
C.C4C 0.019550 0.511054 0.290 0.000001 0. 9999 96
C.C45 0.019805 o. 559887 0.295 0.000001 0. 99 99 97
C.G5C 0.019562 0.608760 D.300 0. DODOO I D.999999
c.C55 C.01899.3 o. 651i494 0.305 0. DODOO I 0. 9999 98
Q.[6[ 0.011010 0.702026 0.310 0.000000 0. 9999 99
C.C65 0.016608 0.744487 0.315 0.000000 0. 9999 99
C.C7C o. 0151 72 0.793252 0.320 0.000000 0.999999
[. [: 75 C:.013653 0.817957 0.325 0. 000000 0.999999
C.[8[ 0.012122 o. 848490 0.330 0.000000 .000000
e.ca5 0.010633 []. 874901 0.335 0.000000 • 000000
[.[9[ 0.009229 o.897454 0.340 o.oonooo • OGDODO
C.C95 c.001935 0.911l470 0.345 0. 000000 .000000
[. DC 0.006767 o. 932 334 0.350 0.000000 .000000
c. 05 0.005728 0.945450 0.355 0.000000 • 000000
c. IC 0.004817 o. 956209 o.36o o. rmoooo • occooo
c. 15 c. 004028 o. 964976 0.365 0.000000 • 0000 DD
c. 2[ 0.003351 0.972079 . o. 370 0.000000 .000000
r:. 25 0.002774 n. 977806 0.375 0. 000000 .000000
c. 3C 0.002288 o. 982404 0.380 0. 000000 .000000
c:. 35 Q.001879 o. 996 092 0.395 0. 000000 .000000
c. 4[ n.001538 0.999015 o. 390 0.000000 • 0000 DO
c. 45 0.001255 D.991347 o. 395 o.nooooo .000000
c. SC 0.001020 0.993197 0.400 0.000000 • DO OD OD
c. 55 C. OCD828 o.994661 0.405 0.000000 • 0000 no
[. 6[ 0.000669 n.995816 0.410 0.000000 .000000
o. 6': 0.000540 0.996727 D.415 0.000000 .000000
c. 7[ 0.000435 0.997444 o.420 0.000000 .0(0000
c. 75 c:.000349 o. 999007 0.425 n.000000 • 0000 DO
c. 8( 0.000279 0.998449 0.430 0.000000 .ornooo
c. 85 0.000223 n.998794 D.435 D.DOOUOD .000000
c. 9( 0.000178 0.999064 0.440 0.000000 • orno rm
c. 95 C.HOl42 o. 999274 0.445 o.ooaooo .000000
c.2oc o.oco112 0.999439 0.450 o.oonoon .000000
[ .205 C.OOOC89 n.999566 D.455 0.000000 • 0000 no
c.21c 0.000010 0.999665 0.460 0.000000 .ocoooo
c. 21 5 [. 000055 o. 999742 0.465 0.000000 .000000
C.22C 0.000044 0.999802 o. 470 0.000000 .(100000
r:. 22: C.000034 o. 999849 0.475 0.000000 .cooooo
C.23( 0.000027 o. 999883 0.490 0. 000000 • ocoooo
0.23~ c:.cooo21 f].999911 0.485 0.000000 • ocoo 00
C:.24[ lJ.OG0016 0.999932 0.490 0.000000 • !)00000
C.245 0.000013 IJ. 999948 0.495 o.oonooo .000000
c.2sc 0.000010 o. 999960 o.500 0.000000 • OCOD OD
r

101

TABLE XLVI

AOQ AND AFI VALUES FOR CSP-3

p AOQ AF! p AOQ A FI

p AOQ AF! p AOQ AFI

p AOQ AF I p AOQ A FI

C.GIC 0.008226 o. 177 354 0.260 0.000000 .000000
C. CI 5 o. 009541 o.363943 0.265 D.DDOODO • OCOOOD
c.c2r: G. 007776 O. !>I I 216 D. 270 O.DOODDO .OOOOOD
c.c25 O.OQ4640 0.814417 0.275 O.DOOOOO .ocooao
C.C3C 0.002245 O. 92 5 I 63 D.280 O.DDDDDD .DO ODDO
C.C35 o. OOD971 a.972268 D.285 D.DDDDOD .DCOOOD
[.[4[ o.OD0396 0.99DIOD 0.290 O.DDODDD • OQOO 00
C.[45 D.000156 o. 996527 0.295 O.DOOODD • DO OD no
c.csc (1.000060 0.998793 0.3DD D.DDOODD .DO ODDO
C.055 o. 000023 0.999584 0.305 D.OODODD .OOODOO
C.C6C 0.000009 0.999857 0.310 0. DO DODO .000000
[.[65 o.OODC03 o. 999951 0.315 O.ODOODO .DO ODDO
c. [7[ o. 000001 D.999983 0.320 O.DDODDO ,, .DO ODDO
[.[75 0.000000 0.999994 0.325 0. DDODDD .OCODOD
C.CBC Q.OODODD 0.999998 D.330 D.OODOOD • DODD OD
c.C85 Q.OODOOO D.999999 D.335 D.DQOODO • DOOO DO
[.[9[ 0.000000 • ODODDD D.340 O.OOODOD • OQODOO
C.C95 0.000000 • DODDDO 0.345 D.DDODDO .DO ODDO
C.IOC c.000000 • OOODOD 0.3SD D.ODDDDD .DOODOO
C.ID5 0.000000 • DDDOOD 0.355 O.DDOOOD .000000
C. I IC 0.000000 • DDOODD 0.360 O.DDODDD • 0000 OD
C. I I 5 0.000000 • ODO DOD 0.365 0. OODDDD .000000
C.12C c.000000 • ODOODO o. 370 0. OD ODDO .OOODOD
[. 25 0.000000 • ODDODD D.375 D.000000 .DOOOOD
c. 3[ c.000000 • 000000 0.380 0.000000 .000000
c. 3!: 0.000000 • ODODOD o.385 O.DOOODD .DOOODO
c. 4[ 0.000000 • 000000 0.390 o.OOODDD • 0000 00
o. 45 0.000000 I • 000000 o. 395 D. 000000 .000000
[. SC 0.000000 I. DOD ODD D.400 O.OOODDD • DO OD OD
o. 55 0.000000 I. OODODO 0.405 0. OOODDO .ocoooo
c. 6[ c.000000 I • 000000 0.410 0.000000 • DODD 00
[. 6!: o. r:rnoooo • 000000 0.415 D.OOOODD .000000
('. 7[ c.ooocoo • ODO DOD 0.420 O.OOOODD .000000
c. 75 0.000000 • 000000 0.425 O.ODDODD .OOOOOD
c. BC c.000000 • 000000 D.430 o.ODOODD • 0000 OD
c. 85 0.000000 • DDOODO 0.435 0.000000 .ocoooo
[. 9[ 0.000000 • 000000 0.440 O.DOOODD .000000
c. 95 [•. 000000 • 000000 0.445 D.000000 .000000
c.2oc C.QOOOOO • 000000 0.450 0.000000 • ODDO DD
0.205 0.000000 • 000000 0.455 O.OOOODD I .000000
c.21c 0.000000 • 000000 0.460 0.000000 1.000000
C. 2 I 5 c._000000 • 000000 0.465 0.000000 I .000000
C.22C 0.000000 • 000000 0.470 O.OOODDO l.OOODOD
0.225 0.000000 • ODD DOD 0.475 0.000000 I .000000
0.230 (l.000000 • DDODOO D.480 O.OOOODD I .000000
C.235 0.000000 • 000000 0.485 O.OOODOD I .OCOOOO
C.24C c.000000 • 000000 0.490 0.000000 I .OCOODD
[.245 0.000000 • DDDDDD 0.495 O.OOODOD 1.000DDO
C.2SC 0.000000 • DD DODO 0.500 0.000000 I .000000
.~·

147

TABLE XCll

AOQ AND AFI VALUES FOR CSP-3

f = 0. 100 AND i = 200

p AOQ AFI p AOQ AFI

c.co: o. !:104208 o. 158408 0.255 0.000000 I • 0000 OD
c.c1c C.GC6872 'J.312778 0.26J 0.000000 I .000000
C .CI 5 0.006794 o. 54 7091 Q.265 0. 000000 '.ocoooo
c.c2c o. 004631 I]. 768461 n. 211 0.000000 I. 000000
[.[25 o.OG2436 0.902576 0.275 0.000000 ' • 0000 nn
C.C3C C.CCll07 o. 963[]97 0.280 0.000000 I • 00 DO OfJ
C.C35 o. 000467 0.986669 0.285 0.000000 I .000000
C.C4C O.COOl89 J.9952Fl6 0.290 0.000000 I .OCQOOO
C.C45 0.000074 0.998352 0.295 0.000000 '.000000
C'. C SC c.000029 0.999429 0.300 0.000000 '.000000
[.[55 0.000011 0.999803 0.305 0.000000 1.ocoooo
C.C6C c. 000004 o. 999932 0.310 0. 000000 1.ocoooo
C.C65 G. OG0002 o. 999977 0.315 0. 000000 I .000000
[.[7[ c. 000001 o. 999992 0.320 0.000000 1.oocooo
[.[75 0.000000 0.999997 0.325 0.000000 I .000000
c.csc 0.000000 0.999999 o. 3.30 0.000000 I • OCOO 00
C.[85 (. ocoooo ' • 000000 0.335 0.000000 .000000
C.C9C c.000000 I. 000000 0.34J 0. 000000 .000000
[.[95 0.000000 I • 000000 0.345 0.000000 • 0000 DD
c. oc O.DGDOOO I. 000000 D.350 0. 000000 .000000
c. 05 (. ococ:oo ! • OOOOOD J.355 0. 000000 .ODOOOO
[ . IC C.OGOOOD ' • 000000 0.36:::! 0.000000 • 0(!0000
c. I 5 o.ocoooo ' • OOOOOD 0.365 0.000000 .ocoooo
c. 2[ 0.000000 I • 000000 0. 37[) 0.000000 .ocoooo
c. 25 o. ooooco I • 000000 0.375 O.ODOOOO • ooco 00
c. 3[ o. 000000 I. 000000 0.380 0. 000000 .ocoooo
c. .35 O.OODDOO I • 000000 o. 385 0.000000 .000000
[. 4[ C:.OQOOOO !.000000 :J.390 D.000000 • ocoo no
c. 45 C. OCOQOO I. OOOlJOD 0.395 0.000000 .000000
[. SC c.ccoooo I. 000000 0.400 0.000000 • 0000 no
c. 55 0.000000 I. 000000 o. 405 n. 000000 • 0000 00
o. 6C 0.000000 I • 000000 0.410 0.000000 • DC OD 00
c. 65 c.ocoooo ' • 000000 0.415 0. 000000 .000000
c. 7[ c. OOOODO I • 000000 0.420 O.ODOOOO .000000
C.175 O.OOOODO ' • 000000 D.425 0.000000 .OCOODO
C.18C 0.000000 I. 000000 0.430 0.000000 .ocoooo
"! C.185 0.000000 ! • 000000 0.435 o.onnooo .DOOOOO
C.19C c.000000 I. 000000 0.44i1 0. 000000 • 0000 00
[.195 0.000000 '. 000000 0.445 0.000000 .ocoooo
r. .20[ O.OODOOO I • OOOOOD 0.450 0.000000 .000000
(.205 c.ocoooo '. 000000 0.455 0. 000000 • 0000 00
C.21 ( 0.000000 I • OOOOOD 0.460 D.000000 .000000
0.215 o. ocoooo I. OOOOOD 0.465 0.000000 .000000
C.22( O.OGOOOD ' • 000000 0.470 0. 000000 .000000
c.22: c.000000 I • 000000 o.475 0.000000 .OOOOOD
(.23[ o.cooooo I • 000000 o. 48'.l 0.000000 .ODOOOO
C.235 o.ccocoo ' • OOOOOD 0.485 0.000000 .000000
0.24C O.OOOOOD I • 000000 D.490 0.000000 .000000
(.24: O.OOOOOD I • 000000 0.495 0.000000 • 0000 00
c .2 sr: o.ocoooo I. 000000 o.soo 0.000000 .DOOOOO
148

TABLE XClll

AOQ AND AFI VALUES FOR CSP-~

f = 0. 150 AND i = 200

p AOQ AF! p AOQ AFI

o.co5 0.003849 0.230144 0.255 0.000000 .000000

O.CIC o.0058D4 0.419570 0.260 O.OOODDO .000000
[.[15 0.005140 0.657359 0.265 D.DOODDO .DODO DO
[. [2[ o.c:o3t89 o. 840542 o. 270 D. oonono .DODO DO
c.c2: 0.001 591 n. 936 363 0.275 O.ODODOD .OCDODO
[.[3[ D.000707 o. 97!>1143 0.200 O.ODDDDD .DODD DD
[.035 0.000295 0.991565 D.285 D. OD DODO .OOOOOD
[.[4[ C.000119 !J. 997027 0.290 O.DDOOOO • DODO OD
[.[45 Q.OG0047 0.998962 0.295 D.DOOODO • 0000 00
[.[5[ O.OOOOl8 'J. 9996110 0.3DCJ 0.000000 .000000
[.[55 c. 000007 0.999876 0.305 D. OD ODDO • OCOD [10
[.[6[ c. 000003 0.999957 0.310 0. OD ODDO .000000
[.[65 c.0000D1 0.999985 0.315 D.OOODDO • 0000 00
[ . [7( 0.000000 0.999995 D.320 D.000000 .ocoooo
[.[75 c.000000 o. 999998 0.325 o.oODOOO .000000
[.[8[ C'.COODOD 0.999999 0.330 O.ODDDOO • ODOO DD
[.[85 0.000000 I • 000000 0.335 O.OOOODO .000000
C.C9C 0.000000 I. ODD DOD 0.340 0.000000 .ocoooo
[.[95 c. 000000 I • OODOOD 0.345 0.000000 .000000
[. DC Q.QGODOO ' • OD ODDO 0.350 0. 000000 .000000
c. 05 c.000000 I. 000000 0.355 O.OOOODD .OCODDO
o. IC o.ocoooo I • OOOODO 0.368 O.ODOODO .ocoooo
(. I: O.GOOOOO I. OODODO 0.365 D.ODDDOD .DOODOO
c. 2( C.OODOOO I • DOOODD o. 370 O.ODDDDO .000000
o. 25 o.oooooc ! • DODDOD D.375 O.ODOOOD .000000
[. 3[ 0.000000 '. 000000 0.38J 0.000000 .000000
c. 35 Q.000000 I • QD0000 i].385 D.000000 .DO OD OD
c. 4[ Q.OODOOD I • ODO ODD 0.390 0. DDDDDD .000000
[. 45 c.ooooco I • 000000 o. 39 5 0.000000 • OCOO OD
[. SC 0.000000 ' • OODOOO 0.400 0.000000 .OGOOOD
(. 55 c.oooDOD I • D00000 0.405 O.ODOOOO .DO DODD
c. 6( c.000000 I • OOODDD D.410 D. DDOOOO • 0000 OD
o. 65 0.000000 '. ODDOOO D.415 O.ODOOOD .OOOOOD
[. 7[ D.COOOOO ! • OOODDD !J.1120 D.ODOOOD .000000
c. 75 O.QOOCOD ! • ODOOOD 0.425 D. ODDO DO .000000
c. 8( O.OODOOO I. OOOODD 0.430 D.ODDDDD .DO OD OD
•' (.105 c.000000 ' • ODODOO 0.435 O.OODOOO .000000
C.19( o.OOODOO ' • ODO DOD 0.44J D.ODODDD .ocoooo
C. I 9: O.OCDOOO l • ODD ODO 0.1145 0. 000000 .OOOODD
[ .20( c.c:ooooo I • ODDOOO 0.115D 0.000000 .OOODOO
0.205 c.000000 I • ODO ODO 0.455 O.DDODOO .DQOOOO
0.21 [ 0.000000 I. 000000 0.460 0.000000 • OC:OD DO
(. 21 5 Q.OQOOOO I • DOOOOO 0.465 D. OD ODDO .DO ODDO
c.22c c.000000 I • ODD ODD 0.470 0. OOODDD • DODO DO
c.225 0.000000 I. DOODOD 0.475 O.OOODOO • OCOO OD
[. 2 3( o. 1100000 I. ODDOOO 0.480 D.DDDOOO .ocoooo
[.235 0.000000 ' • DODOOO 0.485 O.OODDOD .OOODDO
C.240 C.OQOOOO I • OD ODDO D.490 O.DDODOO .OOOODD
c.245 Q.['00000 I • OODOOO 0.495 0.000000 .OOOOOD
( .25( 0.000000 I • OODC100 o.soo o.DDODDO .000000
149

TABLE XCIV

AOQ AND AFI VALUES FOR CSP-3

f = 0. 200 AND = 200.

c.co.: 0.002096 0.580891 0.255 o.nooooo I .000000

C.CIC C.002291: 0.770191 0.26Q 0.000000 I .OCOOOO
C.C I 5 0.001516 o. 898938 D.265 0.000000 1.000000
c.c2c 0.000786 D. 960691 o. 270 0.000000 I .000000
. C.C25
C.C3C
0.000361
[.000155
J. 905553
!]. 994 82 3
0.275
0.280
0.000000
0.000000
1.ocoooo
.DGOOOO
( C.C35 o. 000064 0.998169 D.285 0.000000 • 000000
C.C4C C.000026 0.999357 0.290 0.000000 .000000
·' C.C45 c.000010 0.999776 0.295 0.000000 .ocoooo
C.C5C C.OOOC04 o. 999922 0.300 0.000000 .oocooo
C.C55 o. 000001 o. 999973 o. 305 0.000000 • ooco 00
[.[6[ c. 000001 o. 999991 0.310 0.000000 • 0000 00
[.(65 o.c:ooooo 0.999997 0.315 0.000000 .ocoo 00
C.C7C 0.000000 0.999999 0.320 0.000000 .ocoooo
c.C75 o. ocoooo I • 000000 0.325 0.000000 • 000000
c.csc c. COD ODO ' • 000000 D.330 0.000000 .000000
c.ca: 0.000000 I • 000000 o.335 0.000000 .000000
C.C9C c.000000 ' • 000000 o. 34'..l 0.000000 • OCOO OD
c.C95 O. GOD COO ' • 000000 0.345 0. 000000 • OGOOOO
c. DC c.000000 I • 000000 D.350 0.000000 .000000
[. 05 c.000000 ' • 000000 0.355 0.000000 .000000
c. IC c.000000 I. 000000 0.360 0.000000 • DODD 00
c. I 5 o. ococoo I • 000000 0.365 0.000000 .000000
o. 2C c.ocoooo I • 000000 o. 370 0.000000 • DODO OD
c. 25 c.000000 I. 000000 0. 375 O.OOOOOCl .ocoooo
c. 3( c.ocoooo I • 000000 0.380 0. 000000 • OQOO DO
c. 3: o. ocoooo I . 1)00000 0.385 0. 000000 .oocooo
c. 4[ c.ocoooo I • fJOOOOD D.390 0. 000000 • 0000 OD
c. 45 0.000000 I • 000000 0.395 0.000000 .ocoooo
c. SC o.ocoooo I • 000000 0.400 0.000000 .000000
c. 55 O. OOOCDO ' • 000000 0.405 0.000000 '.ocoooo
c. 6( C.GCOOOO I • 000000 D.410 0. 000000 .000000

,..,
.. c • 65
c. 7(
c. 75
c. 8(
Q.QOOOOO
c.000000
o. 0[0[100
o. ocoooo
' • 000000
l • 000000
t • 'JOO ODD
' • 000000
0.415
0.420
0.425
0.430
0.000000
0.000000
0.000000
0.000000
• ocoo 00
.ocoooo
.OOQOOO
.000000
ii c. 85 [1.000000 l • 000000 0.435 0.000000 • DCOO DO
c. 9[ c.000000 ' • ooonoo o. ~4f'l 0.000000 .000000
c. 95 o. ocoooo ' • 000000 0.445 O.OOr:JOOO • 0000 00
( .2 DC c.ocoooo I • 000000 0.450 0.000000 .000000
C.205 O.IJOOOOO I • IJOOOOD n.455 n.nnnnnn .nnmnn
C.21[ o.ocoooo I • 000000 J.460 0.000000 • ocoo 00
0.215 O.OC:OOOO ! • 000000 o.~65 0.000000 • 0000 OD
C.22C c.000000 I • 000000 D.470 O.OOrJOOO .000000
c .22 5 c.000000 I • 000000 0.475 0.000000 .ocoooo
(.23( c.000000 I. 000000 0.480 0.000000 .000000
0.235 O. QOOOOO I • 000000 0.485 0.000000 .ocoooo
C.24[ c.000000 I • 000000 D.491J 0.000000 .OOOOQO
c. 24 ': C.QQOOOO l • 000000 o.495 0.000000 .ocoooo
c .2 5[ c.000000 I. 000000 o.soo 0.000000 .000000

"
<}~~\

w.
155

TABLE C

AOQ AND AFI VALUES FOR CSP-3

f = 0.500 AND i = 200

p AOQ AFI p AOQ AFI

[.(05 D.DOl856 D. 628807 0.255 D.000000 •ocoo 00

O.CIC 0.001 962 o. 803 776 0.260 0.000000 • OODOOO
C.015 0.001264 0.915765 0.265 0.000000 .000000
[.[2[ 0.000648 0.967606 D.270 0.000000 .ODOOOD
~ [ .02 5 0.000296 0.988149 D.275 0.000000 .000000
,.. [.[3[ 0.000127 0.995761 0.280 0.000000 • ocoo:-rn
( L > C.035 c.000052 o. 998501 D. 285 D.DDOOOO • 0000 OD
[.[4[ 0.000021 0.999474 D.29Cl 0.000000 .ocoooo
...~ [.[45 D.ooooD8 0.999817 D.295 O.ODOODD • 000000
C.050 C.DOD003 0.999936 !J.300 0.000000 .000000
C.055 o. 000001 o. 999978 o. 305 0.000000 .ODOOOO
C.C6C O.OOODOO o. 999992 0.310 0.000000 .ocoooo
C.C65 0.000000 IJ. 999997 0.315 0.000000 • 000000
[.[7( O.COOOOD 0.999999 0.320 0.000000 • OOODOD
(.[75 O.DOOOOD I • OOOODD 0.325 D.ODOOOO .000000
c.cac O.OOODDD I. 000000 0.33CJ 0.000000 .000000
[.(85 O.OODOOO I• D00000 0.335 O.ODOOOO • 000000
0.[9( c.ooooDo 1 • 000000 0.340 0.000000 .000000
[.[95 0.000000 ! • 000000 0.345 0.000000 .000000
c. DC c.DDoooo 1. oDonoo 0.350 0. ODODOO .000000
[. 0.5 O.OODODO I. 000000 0.355 0.000000 .oocooo
( . IC c.ooooDo I. 000000 0.36D 0.000000 .000000
[. I 5 C'.OOOODO I • 000000 0.365 0.000000 .000000
c. 2[ [.(00000 I. 000000 o. 370 0.000000 .ocoooo
[. 25 0.000000 I. 000000 0.375 0.000000 .000000
c. 3( c.000000 • 000000 0.380 0. 000000 • oocooo
o. 35 o.DOOOOO • 000000 0. 38 5 0.000000 .DO CO OD
c. 4[ o.ooooco • 000000 0.398 0.000000 • ocoo 00
c. 4.5 O.OODDOO • 000000 0.395 0.000000 .nonnnn
[. 5( C. CGODOO • 000000 D.400 0.000000 I • OCOOOO
c. 5.5 c.000000 • 000000 o. 405 0.000000 .000000
Q. 6C c.oooooD • 000000 0.410 0.0000.00 .OGOOOO
0.000000 • 000000
... '. c. 6.5
[. 7(
O.OOODOO
c. ocoooo
• 000000
• 000000
0.415
0.420 0.000000 .000000
--~cF
c. 75 c.ooooDo • 000000 o.425 0.000000 .000000
.. c. 8(
C.13.5
c.000000
D. OCOODO
• 000000
• 000000
0.430
0.435
0.000000
0.000000
• 000000
• 000000
0.19( G.000000 • 000000 0.440 0.000000 • ocDono
C.195 c.oDoooo • 000000 0.445 0.000000 • 000000
( .2 oc Q.OOOODO • 000000 0.450 0. 000000 .000000
C.205 0.000000 • 000000 0.455 0. 000000 .000000
[.21( O.GODOOD • rmoooo 0.460 0. 000000 .000000
Q. 21 5 O.OOOODO • 000000 0.465 0.000000 .ODOOOO
c.22c c.DooDoo 1. 000000 0.47CI 0.000000 .000000
0. 22 5 O.DOOODO ! • 000000 0.475 0.000000 .OOODOO
C.23C c.000000 I • OOODOO 0.480 0. 000000 • Dooono
0.235 0.000000 I • 000000 o. 485 0. 000000 • oocono
C.240 0.000000 I • CJOOOOO 0.490 0.000000 .000000
C.24!: D.OODOOO I • 000000 0.495 0. 000000 • ocoooo
C.25[ C.OOOODO 1 • 000000 0.500 0.000000 • ocoo 0[1

"'·
•
~

The Desk Reference of Statistical Quality Methods PDF
100% (1)
The Desk Reference of Statistical Quality Methods PDF
560 pages
American Board of Industrial Hygiene Code of Ethics
No ratings yet
American Board of Industrial Hygiene Code of Ethics
2 pages
Chapter 24
No ratings yet
Chapter 24
8 pages
OSHA Construction Operations
No ratings yet
OSHA Construction Operations
49 pages
Waste Class Code Selection Flow
No ratings yet
Waste Class Code Selection Flow
6 pages
Clean Harbors EPA
No ratings yet
Clean Harbors EPA
17 pages
Beyond Texas City
No ratings yet
Beyond Texas City
108 pages
Industrial Hygiene 1
No ratings yet
Industrial Hygiene 1
28 pages
CSP Complete Guide
No ratings yet
CSP Complete Guide
36 pages
Introduction To Industrial Hygiene
No ratings yet
Introduction To Industrial Hygiene
34 pages
Occupational Hygiene Framework
No ratings yet
Occupational Hygiene Framework
21 pages
American Industrial Hygiene Association
No ratings yet
American Industrial Hygiene Association
18 pages
Hygiene
No ratings yet
Hygiene
62 pages
Full Download The Design, Implementation, and Audit of Occupational Health and Safety Management Systems (Workplace Safety, Risk Management, and Industrial Hygiene) 1st Edition Ron C. Mckinnon PDF
100% (3)
Full Download The Design, Implementation, and Audit of Occupational Health and Safety Management Systems (Workplace Safety, Risk Management, and Industrial Hygiene) 1st Edition Ron C. Mckinnon PDF
62 pages
Principles, Steps For Supervision and Methods of Supervision
100% (1)
Principles, Steps For Supervision and Methods of Supervision
2 pages
CH4101 - 5 Industrial Hygiene
No ratings yet
CH4101 - 5 Industrial Hygiene
57 pages
Principles of Leadership
No ratings yet
Principles of Leadership
283 pages
Prudent Practices For Handling Hazardous Chemicals in Laboratories
No ratings yet
Prudent Practices For Handling Hazardous Chemicals in Laboratories
300 pages
Lecture 2 Occupational Hygiene Principles
No ratings yet
Lecture 2 Occupational Hygiene Principles
42 pages
17 - Guide To Asbestos For Industry
No ratings yet
17 - Guide To Asbestos For Industry
73 pages
Flashcards 25
No ratings yet
Flashcards 25
6 pages
Flashcards 44
No ratings yet
Flashcards 44
6 pages
3.12 Hearing Conservation Program: HSE Manual
No ratings yet
3.12 Hearing Conservation Program: HSE Manual
2 pages
5.2HCS-Safety Data Sheet (SDS)
100% (1)
5.2HCS-Safety Data Sheet (SDS)
35 pages
CSP 3
No ratings yet
CSP 3
24 pages
Guide - Industrial Hygiene Program Management
No ratings yet
Guide - Industrial Hygiene Program Management
3 pages
CandidateHandbook ABIH
No ratings yet
CandidateHandbook ABIH
21 pages
Dept of Labor: Occupational Safety and Health Administration
No ratings yet
Dept of Labor: Occupational Safety and Health Administration
18 pages
IDSE Unit 1 Sample Exam
100% (3)
IDSE Unit 1 Sample Exam
2 pages
Health and Safety Training Matrix
No ratings yet
Health and Safety Training Matrix
1 page
Get (Ebook PDF) Fundamentals of Occupational Safety and Health 7th Edition Free All Chapters
100% (4)
Get (Ebook PDF) Fundamentals of Occupational Safety and Health 7th Edition Free All Chapters
42 pages
Flat Roof
No ratings yet
Flat Roof
68 pages
UIL Safety Training
No ratings yet
UIL Safety Training
46 pages
03 Industrial Hygiene
No ratings yet
03 Industrial Hygiene
50 pages
Industrial Hygiene Baseline Hazard Assessments: MSC-PRO-17916
100% (1)
Industrial Hygiene Baseline Hazard Assessments: MSC-PRO-17916
13 pages
M.B.A. Degree Examination - 2012: 240. Industrial Hygine and Occupational Health
No ratings yet
M.B.A. Degree Examination - 2012: 240. Industrial Hygine and Occupational Health
2 pages
UNIT 7 de Recruitment or Separation
No ratings yet
UNIT 7 de Recruitment or Separation
48 pages
Director Environmental Health Safety EHS in Philadelphia PA Resume Akhil Buch
No ratings yet
Director Environmental Health Safety EHS in Philadelphia PA Resume Akhil Buch
3 pages
Introduction To Surveillance
No ratings yet
Introduction To Surveillance
58 pages
(Topics in Physical Chemistry 1) Prof. Dr. Klaus Christmann (Auth.) - Introduction To Surface Physical Chemistry-Steinkopff-Verlag Heidelberg (1991)
No ratings yet
(Topics in Physical Chemistry 1) Prof. Dr. Klaus Christmann (Auth.) - Introduction To Surface Physical Chemistry-Steinkopff-Verlag Heidelberg (1991)
283 pages
1 120i PDF
No ratings yet
1 120i PDF
60 pages
Asbestos Control - Surveys, Removal, and Management (PDFDrive)
No ratings yet
Asbestos Control - Surveys, Removal, and Management (PDFDrive)
108 pages
American Conference of Governmental Industrial Hygienists (Acgih) Threshold Limit Values (TLVS) For Lifting
No ratings yet
American Conference of Governmental Industrial Hygienists (Acgih) Threshold Limit Values (TLVS) For Lifting
6 pages
JSA National Safety Council
No ratings yet
JSA National Safety Council
2 pages
[FREE PDF sample] Welding Health and Safety A Field Guide for OEHS Professionals 2nd Edition Michael K. Harris E Michael R. Phibbs ebooks
100% (2)
[FREE PDF sample] Welding Health and Safety A Field Guide for OEHS Professionals 2nd Edition Michael K. Harris E Michael R. Phibbs ebooks
40 pages
Safety Engineering Guidelines. Pneumatic and Electric Solutions
No ratings yet
Safety Engineering Guidelines. Pneumatic and Electric Solutions
80 pages
Safeguarding of Machinery and Plant PDF
No ratings yet
Safeguarding of Machinery and Plant PDF
66 pages
Industrial Hygiene (IH) - Concepts and Principles
No ratings yet
Industrial Hygiene (IH) - Concepts and Principles
81 pages
Job Hazard Analysis Template .Docx Version 1
100% (1)
Job Hazard Analysis Template .Docx Version 1
2 pages
Blueprint References: Title & Auxiliary Information
No ratings yet
Blueprint References: Title & Auxiliary Information
2 pages
Ahmed Fouad CV
No ratings yet
Ahmed Fouad CV
7 pages
3 UEq Book Excerpt PDF
No ratings yet
3 UEq Book Excerpt PDF
38 pages
Industrial Hygiene & Safety Word Puzzle
No ratings yet
Industrial Hygiene & Safety Word Puzzle
2 pages
Inlet Line Replacement Procedure
No ratings yet
Inlet Line Replacement Procedure
6 pages
OHST References
No ratings yet
OHST References
1 page
OSHA
No ratings yet
OSHA
2 pages
The Desk Reference of Statistical Quality Methods PDF
100% (1)
The Desk Reference of Statistical Quality Methods PDF
560 pages
Minimum Average Fraction Inspected For CSP-M Plan: Chung-Ho Chen and Min-Tsai Lai
No ratings yet
Minimum Average Fraction Inspected For CSP-M Plan: Chung-Ho Chen and Min-Tsai Lai
4 pages
Continuous Sampling Plan
No ratings yet
Continuous Sampling Plan
11 pages
Zero Defect Sampling
No ratings yet
Zero Defect Sampling
7 pages
04 - Engine Mechanical
100% (1)
04 - Engine Mechanical
48 pages
Marketing Strategies Used by Kotak Mahindra Bank
No ratings yet
Marketing Strategies Used by Kotak Mahindra Bank
3 pages
Unikl Miat Pamplet 2009
No ratings yet
Unikl Miat Pamplet 2009
4 pages
(Ebook) Transportation Experience: Policy, Planning, and Deployment by William L. Garrison, David M. Levinson ISBN 9780195172508, 0195172507 - The complete ebook set is ready for download today
100% (1)
(Ebook) Transportation Experience: Policy, Planning, and Deployment by William L. Garrison, David M. Levinson ISBN 9780195172508, 0195172507 - The complete ebook set is ready for download today
58 pages
Latest Ballast Water Management Implementation Dates
No ratings yet
Latest Ballast Water Management Implementation Dates
2 pages
1 s2.0 S0016510715030370 Main
No ratings yet
1 s2.0 S0016510715030370 Main
10 pages
Parameter List EPA Commander SK (English)
No ratings yet
Parameter List EPA Commander SK (English)
2 pages
Narrative (85) Aa
No ratings yet
Narrative (85) Aa
20 pages
Ilide - Info Soal Procedure Text PR
No ratings yet
Ilide - Info Soal Procedure Text PR
8 pages
Type of Monolithic Refractories PDF
No ratings yet
Type of Monolithic Refractories PDF
4 pages
Chapter 1 Book Questions With Answers
No ratings yet
Chapter 1 Book Questions With Answers
5 pages
Underwater Welding Seminar Report
100% (1)
Underwater Welding Seminar Report
19 pages
JVC GV Chassis Av30w475 TV D PDF
No ratings yet
JVC GV Chassis Av30w475 TV D PDF
24 pages
Astm B3-95
No ratings yet
Astm B3-95
5 pages
Assessment of Gadgets Addiction and Its Impact On PDF
No ratings yet
Assessment of Gadgets Addiction and Its Impact On PDF
5 pages
Jósmenta
No ratings yet
Jósmenta
3 pages
RB014 - Launching A2-P2 Rev 1
No ratings yet
RB014 - Launching A2-P2 Rev 1
15 pages
Visa Platform
No ratings yet
Visa Platform
3 pages
SailPoint InstallationSteps 6.4version
No ratings yet
SailPoint InstallationSteps 6.4version
2 pages
Abhishek Kaushik, 202010, FBD, 3 Yrs, B.Tech Mech
No ratings yet
Abhishek Kaushik, 202010, FBD, 3 Yrs, B.Tech Mech
2 pages
WEEK1 Q1 Module ActSheet w2
No ratings yet
WEEK1 Q1 Module ActSheet w2
4 pages
Hull OFOD11 e Solutions CH 13
No ratings yet
Hull OFOD11 e Solutions CH 13
13 pages
Classical Theory of Management
No ratings yet
Classical Theory of Management
8 pages
Garvey Et Al-2006-Research in Nursing & Health PDF
No ratings yet
Garvey Et Al-2006-Research in Nursing & Health PDF
11 pages
Safety Net Fixing at CUT OUT JSA HSE Professionals
No ratings yet
Safety Net Fixing at CUT OUT JSA HSE Professionals
1 page
Expoling OEEin Ind MFG Plant
No ratings yet
Expoling OEEin Ind MFG Plant
12 pages
CINEO C4040 Installation Manual
No ratings yet
CINEO C4040 Installation Manual
98 pages
Artificial Intelligence For Business - Question Answer Booklet
No ratings yet
Artificial Intelligence For Business - Question Answer Booklet
17 pages
Porsche Cayenne 2003 - 2010 Fuse Box Diagram
100% (2)
Porsche Cayenne 2003 - 2010 Fuse Box Diagram
14 pages
Microwave Installation and Commisioning
No ratings yet
Microwave Installation and Commisioning
54 pages