You are on page 1of 555

Loughborough University

Institutional Repository

Improving the new product


development process
This item was submitted to Loughborough University's Institutional Repository
by the/an author.
Additional Information:

A Doctoral Thesis. Submitted in partial fulllment of the requirements

for the award of Doctor of Philosophy of Loughborough University.

Metadata Record:
Publisher:

https://dspace.lboro.ac.uk/2134/6941

c D.J. Stockton

Please cite the published version.

This item is held in Loughborough Universitys Institutional Repository


(https://dspace.lboro.ac.uk/) and was harvested from the British Librarys
EThOS service (http://www.ethos.bl.uk/). It is made available under the
following Creative Commons Licence conditions.

For the full text of this licence, please go to:


http://creativecommons.org/licenses/by-nc-nd/2.5/

THE NEWPRODUCT
IMPROVING
DEVELOPMENT
PROCESS
i

by

D. J. STOCKTON
(B. Sc.,

M. Sc.,

M. I. M., C. Eng. )

fulfilment
A Doctoral Thesis submitted in partial
for
the award of
the
requirements
of
Doctor of Philosophy

of Technology
of the Loughborough University
( February 1983 )

by D.J. Stockton (1983)

Acknowledgements

I would particularly

like

to thank Mr. J. Middle for the

help, advice and encouragement given throughout the period of the


research York.

In addition
Engineering staff
assistance,

my thanks are due to the Design and Production


at Herbert Morris Ltd. for their

to the Science Research Council for financial

support and to Mrs. J. Morley for the efficient


way in which this

Finally

and speedy

thesis was typed.

I would like

to thank Herbert Morris Ltd. for

allowing me to carry out the research at their


the financial

valuable

assistance

given.

company and for

11

SUMMARY

A system has been developed and is being used at H.M. -Ltd. for
the labour and overhead costs of components manufactured

estimating

by a wide variety
linear

of production

regression analysis

to develop estimating

measure the relationship

quantitatively

The system uses multiple

processes.

of a component and the factors

equations that

between the production time

that influence

this

Production

time.

times can then be converted to cost using appropriate

labour and

overhead cost rates.

The system uses design features


the estimating

only for predictor

Hence designers with little

equations.

in

variables
concept of

manufacturing methods can use the system to cost designs as they


evolve.

This feature

therefore

tool.

cost optimization

The manufacturing

times

the estimated
planning
will

and control

Software

components.
that

also

of manufacturing.
to the design

has been developed

equations

appropriate

conditions

This

generates

could

producibility

Morris

as standard
In this
features

to allow

form the basis


data for

equations

Ltd.

times for

Hence
the

way manufacturing

costs

of a product.

a computer to retrieve

and compute the production

software

estimating

at Herbert

can be used directly

linked

be directly

time data used to develop

operating

current

represents

provides designers with a powerful

times and costs


for

designers.

a larger

of

system

111

A method of allowing

designers to estimate the development times

components and assemblies has been developed.

for individual

This

enables the design process to be scheduled such that the

facility

new product development time could be minimized.

overall

element of this

scheduling method is the ability

to allocate

between components to be designed on the basis of relative


importance to. the overall

An important
resources
cost and

success of the project.

Models have been developed that allow :

(i)

The total

cost of a product to be estimated within

degree
of accuracy from a knowledge of the
reasonable
product's
(ii)

cost

(i)

(ii)

types of cost

The various
design

Finally

ten highest cost components.

solutions

design

factors

involved

to be established,

solutions

in alternative

hence enabling

minimum

to be identified.

been
has
carried
research

out into:

when developing
0

a technique

The factors

to consider

forecasting

the need for product development.

The factors

to consider when developing a prescriptive

guide frr

for

into
launched
the
that
when
ensuring
a product

market is a success.

iv

(iii)

Alternative

methods of solving the "assortment"

(i. e. determining

problem

the optimum number and specification

of

standard sizes to manufacture).


(iv)

The integration

of Value Engineering into the new product

development process.
(v)

The use of cost/design

specification

relationships

in the

generation of minimum cost designs.


The contribution

of existing

Computer-Aided Draughting and

Design systems to the new product development and original


design processes.

CONTENTS

Page

Acknowledgements

Summary

11

List of Contents

List

of Figures

xi

List

of Tables

xiv

Nomenclature included in Text

xvii

Abbreviations

xix

Introduction
Chapter

1
Product

Planning

8
f

1.0

Introduction

1.1

The New Product Development Process

9
16

1.1.1

Product Specification

19

1.1.2

Product Development

24

Criteria

25

1.2

NPDSuccess/Failure

1.3

Forecasting

for NPD

33
4 "

Chapter 2 Aspects of the Design Process

0-

58

2.0

Introduction

59

2.1

The Make or-Buy Decision

64

2.2

Value Engineering

76

vi

Page

2.3

The Assortment Problem

81

2.3.1

Solution

86

2.3.2

The Minaddition

Techniques

2.4

Design Planning and Control

2.5

Design Scheduling

2.6

Alterations

Chapter 3

91

Technique

99
109
110

to Product Designs

115

Costing Product Designs

f
3.0

Introduction

116

3.1

Total Manufacturing Costs

116

3.2

3.3

117

Costs

3.1.1

Material

3.1.2

Scrap Costs

119

3.1.3

Overhead Costs

125

3.1.4

Direct

Labour Costs

Analysis of Cost Estimating

128
at H.M. Ltd.

132

3.2.1

Function

133

3.2.2

Inputs and Outputs

136

3.2.3

HumanAgents

138

3.2.4

Sequence

140

3.2.5

Environment and Physical Catalysts

142

Work Measurement Techniques

144

3.3.1

Time Study

154

3.3.2

Pre-determined Motion Time Standards

155

3.3.2.1

MTM

3.3.3

Category Estimating

3.3.4

Synthetic

Time Standards

156
161
166

vii

Page

3.4

3.3.5

Work Measurement using Computers

167

3.3.6

Work Measurement Algorithms

172

3.3.6.1

Synthetic

173

3.3.6.2

Empirically

Derived Equations

180

3.3.6.3

Operational

Research

186

3.3.6.4

Multiple

Regression

188

of Work Measurement Method for Design


Costing at H.M. Ltd.
Selection

Chapter 4

Cost Estimating

4.0

Introduction

4.1

Classification

4.2

Techniques

System Development

198,
202
203

of Estimating

Models

204

Processes and Operations

4.1.1

Manufacturing

4.1.2

Size and Power of Machines

207

4.1.3

Manual Tasks

210

4.1.4

Component Tolerances

4.1.5

Material

and Surface

Finish

Processed

213
213

214

Data Collection
of Predictor

4.2.1

Selection

4.2.2

Data Source

217

Variables

219
226

Regression Procedure

4.3

The Multiple

4.4

Equation Validation
4.4.1

206

231

Accuracy of the Existing


H.M. Ltd.

Estimating

System at
240

4.5

Comparison with Time Study Data

240

4.6

Comparison with Existing

257

4.7

System Computerisation

Time Standards (STT's)

257

viii

Page

'

4.7.1

System Size

4.7.2

Variation

4.7.3

The Effective

4.7.4

System Expansion

of Machine Types

Chapter 5

use of Computer Time

267
269
270
272

Design Cost Forecasting

273

5.0

Introduction

274

5.1

Cost Relationships

275 -

5.2

Costing using the Pareto Distribution

285

Chapter 6

The Development of Design Scheduling Method

6.0

Introduction

6.1

Specifying

6.2

Sequencing Component Design

299
the Objective

Criterion

299
300

Component Development Times

301

between Development Tasks

307

6.2.1

Estimating

6.2.2

Interactions

6.2.3

Types of Activity

6.2.4

Accuracy

Interactions

of Estimated

308

Development Times

313

Importance of Components

317

6.3

Determining the Relative

6.4

Preventing Design Changes

Chapter 7

298

Computer Aided NPD

320
321

7.0

Introduction

322

7.1

Function Analysis

327

7.2

Data Acquisition

and Transfer

329

ix

Page

7.3

Design Creation and Improvement

331

7.4

Cost, Marketing and Manufacturing Analysis

335

7.5

Draughting

337

7.6

Integration

with Overall NPDProcess

338

Of

Chapter 8

Discussion

341

8.0

Introduction

342

8.1

Comparison of Work Measurement Techniques

351

8.2

Cost Estimating

8.3

The Make or Buy Decision

373

8.4

Cost Forecasting

376

8.5

Design Scheduling

380

8.6

CAD

385

Chapter 9

System Development

Conclusions

358

395

Chapter 10 Recommendationsfor Further Work

400

References

404

Bibliography

424

Appendix I

Development of the New EHB Range

429

Appendix II

Estimating

441

Appendix III

Design Cost Estimating

System:

Manual Method Description


Program (DESCOSTIMATE):

Operation Notes and Program Listing

453

A. 3.0

Introduction

454

A. 3.1

Running Procedure

456

A. 3.2

Program Listing

469

Page

Appendix IV

Descriptions

A. 4.0

Introduction

A. 4.1

Variable

482

of Variables

483

Description

Notes for Individual

Cost

Centres

484

A. 4.2

CommonVariables

506

A. 4.3

Variables Applicable

A. 4.4,

Illustrations

509

to Single Cost Centres

510

of Welding Positions

Appendix V

Data File for Design Cost Estimating

Appendix VI

Modifying the Regression Procedure

Program

512
521

X1

LIST OF FIGURES

Page

13

1-

Strategic

NPDDecision Flow Chart

17

Product
The
Life Cycle and Dynamic Competitive Strategy

34

Profitability

Number of Attributes

Planning Framework

36

Gap during Product Replacement


Examined per Buyer during the

Buying Decision

55

Optimum Number of Standard Hoist Sizes to Develop

97

70%, 80% and 90% Learning. Curves

Comparison of Scrap Levels Estimated from 90%, 80% and


70% Learning

124

Curves

Accuracy

Relative

122

against

Application

Time for

the MTM
158

Systems
10

Manufacturing

11

Outline

162

Time Bands

Procedure

for

Applying

the Category

Estimating

Technique

165

12

Machining Time Parameter Relationships

171

13

Method of Least Squares

190

14

Comparison of Average and Regression Lines

190

15

of Increasing the Numberof Regression Models


for a Specific Data Range

194

16

Comparison of*R with Average Proportional

Error

234

17

Comparison of F with Average Proportional

Error

234

18

Comparison of Coefficient
Proportional Error

Effect

of Variability

with Average
235

xii

Page

19

20

A Comparison of Actual and Estimated Labour Costs at


H.M. Ltd.

241

The Estimation of Negative Machining Times using


Regression Models

244

21

Frequency
ency Distribution

of 'E' Values

248

22

Frequency Distribution

of 'S'

248

23

Actual

versus

Estimated

Costs - Milling

24

Actual

versus

Estimated

Costs - Drill

Jigs

25

Actual

versus

Estimated

Costs - Plate

Gauges

26

Actual

versus

Estimated

Costs - Plug Gauges

254

27

Actual

versus

Estimated

Costs - Copy Templates

255

28

Actual

versus

Estimated

Costs - Drill

29

STT versus

Estimated

Times - Bar Lathes

258

30

STT versus

Estimated

Times - Chuck Lathes

258

31

STT versus

Estimated

Times - Bar Capstans

259

32

STT versus

Estimated

Times - Chucking Capstans

259

33

STT versus

Estimated

Times - Centre Lathes

260

34

STT versus

Estimated

Times - Horizontal

260

35

STT versus

Estimated

Times - Vertical

36

STT versus

Estimated

Times - Copy Lathes

261

37

STT versus

Estimated

Times - Radial

262

38

STT versus

Estimated

Times - Cylindrical

39

STT versus

Estimated

Times - Milling

40

STT versus

Estimated

Times - Saws

41

STT versus

Estimated

Times - Surface

42

STT versus

Estimated

Times - Gear Cutters

Values
Fixtures
(Box Type)

(Normal)

Jigs

Borers
Borers

Drills
Grinding

Machines

253
253
254

255

261

262
263
263

Grinders

264
264

Xlii

Page

43

STT versus Estimated Times - NC Chuck Lathe

265

44

STT versus Estimated Times - NC Bar Lathe

265

45

STT versus Estimated Times - Spline Cutters

266

46

STT versus Estimated Times - Shears

266

47

Descostimate Standard Input Data Form

271

48

Assembly Cost Inter-relationships

281

49

Total Costs versus Ld: D Ratio for the 3.2 Tonne EHB

284

50

The Pareto Curve of Cumulative Number of Stock Items


versus Cumulative Stock Cost

286

51

Log-Log Graph of the Pareto Distribution

290

52

Log-Log Graph of Cumulative Costs for H.M. Ltd. 's


Products

291

53

Actual versus Estimated Product Costs'(-<300)

297

54

Actual

55

Average Task Times versus


of NPD Tasks

Product

Estimated

versus

for the New EHB

Costs (>E300)

Cumulative

Cumulative

297
Number
305

56

Component NPD Network Diagram

57

Determination

of Development Times

= PARTS A and B'*,

315

58

Determination

of Development Times

- PARTS C and D

316

59

Actual versus Estimated Development Times

318

60

Pareto Curve of the 0.8 Tonne EHBComponentDevelopment


Times

318

Information

309

System

340

61

Integrated

62

Schematic

63

Load Spectrum of the New EHB Range

439

64

First

443

65

Flow Chart for

66

The Modified

Diagram of the New EHB Concept

Idler

MT Spurwheel

Descostimate

Regression

Objective

438

457
523

xiv

LIST OF TABLES

Page

Product Planning Strategies

The-Contribution
Financial

11

of Product Planning to Marketing and


Based Strategies

Product Development Matrix for H.M. Ltd.

The Design Functions Influence

22

on New Product Success/


30

in Attribute

Ranking between Product

Variability

Supplier

Attributes

Affecting

Relative

Importance

of New EHB Attributes

8.

Stages in the Design Process

Make or Buy Decision

10

Types of Batches Present

11

Manufacturing

the Purchasing

Types

Decision

65

in a Manufacturing

Organisation

73

Methods

13

Factors

79

14

Data Required

15

Optimum Product

16

Methods of Improving

17

Major

18

Factors

19

Assessment of a Work Measurement Department

the Assortment
the Minaddition

Problem

82

Technique

92

Range Data
Product

94
Profitability

105

Causes of Design Changes


Essential

70

Rope Guide Assembly

Cost Reduction

for

47
51

Criteria

Costs of Alternative
Designs (3.2 Tonne New EHB)

Affecting

43

61

12

20

Criteria

Failure

12

112

to Work Place Standardization

145

Check List

and Score Sheet

149

Operator Performance Factors

151

xv

Page

21

Accuracy and Application

22

Manufacturing Time for the New EHBRange

157

Time for the MTMSystems

160
163

Chart

23

50 Interval

24

Criteria

25

A Comparison of Estimated and Observed Machine Times

178

26

Equations for Estimating Metal Removal Times

181

27

Speeds and Feeds for Drilling

184

28

Classification

29

Manipulation

Estimating

for Establishing

176

Nd

Operations

208

of Costing Equations
Task Times for a Devlieg Spiramatic

212

. Jigmil
Factor (Machinability)

30

Material

31

Time Standard Estimation

215

Groups

for NC Flame Cutting'at


221

H. M. Ltd.
Design Factors

33

SPSS Output File

34

Examples of Altering
Estimating

224

Welding Duty Cycles

Affecting

32

.
Input

Variables

228

to Improve
243

Accuracy

245

Times

35

Minimum Operation

36

Causes of Variability

37

Component Classification

38

Cost/Hoist

Capacity

Relationships

39

Cost/Hoist

Capacity

Relationships

246

in Time Study Data


for

Determining

Assembly Times

250
276

Developed using the

0.8 Tonne and 3.2 Tonne Units

278

Costs at Various

283

40

Total

41

Cost Estimates

42

NPD Activity

using
Times

Drum Tube Lengths


Equation

47

295
302

xvi

Page

Project

Activity

Times

43

Factors Affecting

44

Objectives

of using CAD

324

45

Proportion

of Time Spent on Design Tasks

327

46

Process Cost Centres

444

47

Rules for Determining General Machining Cost Centre

445

48

Variable

49

Predictor

50

Labour and Overhead Cost Rates (f's/Hr)

51

Data File

Description

Notes for Cost Centre 202

Variable Coefficients

for Cost Centre 202

306

447
448
451

for Determining General Machining Input

Codes

513
Variable

Coefficients

52

Data File

for Predictor

53

Data File

for Labour and Overhead Cost Rates

54

Comparison of the Accuracy of Conventional and


Modified Regression Procedures

513
520

527

XV11

Nomenclature Included in Text

AI
B2.

Association

Machine Tooling Capacity

Index

C=

Cumulative Product Costs (i's)

Ca

Additional

Costs incurred

from adding one extra

Standard Size ('s)


Ce
Cm

Cp
Csv

Prime Costs ('s)

Total

Total Variable

Cost Saving obtained from adding one extra Standard


Size (`s)

Material

('s)

Costs per Unit

In-House Labour Costs per Unit ('s)

(mm)

D=

Rope Drum Tube Outside

Dm

Mean Diameter

(inches)

e=

Average Error

Ratio

E=

Average Proportional

f, g

Constants the values of which depends on the form


of the Pareto distribution

Ideal point on the ath attribute

Machining Factor

Total Machined Length (inches)

Rope Drum Tube Length (mm)

Labour Cost Rate at the"mth. Manufacturing


(f's/Hr)

Cumulative Number of Adopters in Time c

Component Complexity Factor

Nm

Number of Machines Controlled

Qb

Purchased Batch Size

Ia
Ki
Lt
Ld
Lm

Nac
Nd

Diameter

Error

by Operator

Cost Centre

xviii

Qp

Manufacturing

Recovery Period for Capital

Correlation

S
Sa

Sp
Sr2
St
Tm

Batch Size
Outlay (Years)

Coefficient
of the Error Proportions

Standard Deviation

Percentage average scrap per batch for a specific


"learning situation"

Percentage Scrap

Standard Error Ratio

Machine Setting

Production

Time (mins)
Cost Centre

Time at the mth Manufacturing


Frame Sizes

in the Interval

w=

The Number of Standard

Xi

Corresponding estimated value of the dependent


variable

A Constant equivalent to the intercept


of a straight line graph

xT, x2... xn

Coefficients

Mean Value of the yi's

yi

Actual Value of the dependent variable

The Value of the dependent variable

of the independent variables

Time at which 50% market potential


a=

Standard

on the y-axi. s

Deviation

of time

zl, z2... zn

is achieved

xix

ABBREVIATIONS

BM's

Basic Minutes

Computer-Aided Design

Critical

Compensating Rest Allowance

-"

Davy Computing Ltd.

DIAD

Distributed

DP

Dynamic Programming

Electric

Economic Order Quantity

Easily Used Computer Language for

CAD
CPM
CRT
DCL

EHB
EOQ
EUCLID

Path Method

Interactive

Automated Draughting

Hoist Block

Illustrations

and Drawings
FEM
FTF

H.M. Ltd.
LOCAM
LP
MBO
MLR

MOST
MRP
MTM
NC
NPD
OEM
PBC

Europeene de la Manutention

Federation

Floor-to-Floor

Herbert Morris Ltd.

Logan Computer Aided Manufacture

Linear

Management by Objectives

Multiple

Maynard Operation Sequence Technique

Materials

Methods Time Measurement

Numerically

New Product Development

Original

Period

Time

Programming

Linear

Regression

Requirement Planning

Controlled

Equipment Manufacturers
Batch Control

xx

PERT
PLC
PMTS
PPBS
p/t
REAP

Product Life Cycle

Pre-determined Motion Time Standards

Planning,

Prototype

Rapid Estimating

'SAPPHO

SPSS
STT

VE

and Review Technique

Performance Evaluation

Programming, Budgeting

and Planning

Scientific

Activity

Heuristic

Origins,

Predictor

Packages for

Statistical

Sales Target

Value Engineering

Time

and Systems Analysis

from Patterns with

the Social

Sciences

1
INTRODUCTION

Since the company's foundation


Herbert Morris Ltd.

in 1884 to market pull'

(H. M. Ltd. ) has become a leading UK manufacturer of

a wide range of mechanical handling equipment including


Goliath Cranes, Container-Handling

Cranes and Electric

In 1978 the company became part of Davy Corporation


international

blocks(t),

engineering and construction

In recent years the competitive

Overhead Cranes,
Hoist Blocks.

Limited the

group.

action of other European crane

manufacturers has led to many of the company's products becoming

H.M. Ltd. are still


in manufacture
competitors
enables

products

technically

acceptable they are labour intensive

and are therefore

in stock

more expensive

In addition

products.

holding

costs

than many of the leading

design

modular

and their

both manufacturers

reductions

Although the product designs of

in the market place.

uncompetitive

distributors

of competitive
to achieve

compared to the current

larger

range of the

of H. M. Ltd.

Since the products of H.M. Ltd. are financially


present recession in world markets has resulted
in sales levels

and hence profitability.

In order to recover from this


undertaking

an extensive

lines,
product

the

uncompetitive

in a considerable

decline

In consequence the company's

workforce has been reduced and product lines

ranges.

products

situation

sold off.

H.M. Ltd. are currently

programme of development of several major

which includes the wire-rope

and chain hoist

product

The need to simultaneously

develop several product lines

places

a heavy burden on the development resources of H.M. Ltd. and leaves


little

scope for further

products fail

product development should the replacement

to achieve their

expected profit

The research reported here arises


improved range of electric

decline

due to the continual

project

(New EHB) which is necessary

in sales of the existing

Ropemaster range

The EHB is also a major sub-assembly in H.M. Ltd. 's range

(Appendix I).
of Universal

from the development of an

hoists

wire-rope

objectives.

Cranes.

The constraints

surrounding the New EHBdevelopment

are:

(i)

Limited development time available

since an improved

product range needs to be launched quickly


further

serious decline

company and to restore


(ii)

in the market presence of the

for research and development

development
projects
since several
electric

are currently
(iii)

(electric

chain hoist and universal

reduced profitability

An inherent manufacturing
lack of capital
quantity

wire-rope
crane

range)

being undertaken in parallel.

Limited investment resources arising


current

(iv)

levels.

profitability

Limited resources available

hoist,

to prevent

investment

levels.
cost disadvantage since the
funds-limits

that can be manufactured.

have the advantage of cost reductions


the manufacture of larger

from the company's

volumes.

the annual

Competitors therefore
that arise through

(v)

The need to develop a product design that increases the


company's in-house manufacturing work load and hence
reduces the manufacturing

overhead burden on other

products manufactured by the company.

The present work attempts to reduce the effect

of these constraints

on the development of the New EHBrange with the further


providing

a contribution

Initially

of the NPDprocess and its

product planning,

a strategic

volume, multi-product

contribution

relating

of developing a prescriptive

projects,

element of
to the low

manufacturing environment of H.M. Ltd.

that determine the eventual success or failure


with the objective

to the strategic

planning process appropriate

Chapter 1 also reviews the literature

and future

Chapter 1 examines

Product development is an essential

planning process.

of

to the improvement of the new product

development (NPD) process in general.


the structure

objective

to the factors

of new product developments


guide for the New EHB

i. e. the product must enable the company to regain

an adequate market presence and prevent the complete erosion of its


hoist

block market by competitors.

Chapter 1 finally

reviews existing

sales forecasting

in order to develop a technique for forecasting


begin the NPDprocess with the objective
for the development process to be carried

the date

of allowing

techniques
at which to

sufficient

time

out, and hence minimize the

loss of revenue that normally accompanies product change overs.

Various aspects of the development process associated with the


have been examined in Chapter 2:

New'EHBdevelopment constraints

overall

(i)

A model has been developed and demonstrated for comparing


design solutions

the costs of alternative

and allowing

The model
the minimum cost solution to be identified.
1.0
by
those
the choice
affected
only
costs
considers
and enables the effect

of variations

in purchasing and

batch sizes and recovery period for capital

manufacturing

investment costs to be examined.


(ii)

The integration

of Value Engineering work into the NPD

improving
the
is
a
of
rate
and
method
considered
process
at which cost reduction

tasks can be carried

out is

proposed.
(iii)

Alternative

methods of solving

(i. e. determining

the "assortment"

problem

the optimum number and specification

of

standard sizes to manufacture) are examined and the


Minaddition

technique identified

as appropriate.

This

technique is then used to determine the optimum number and


the capacities
(iv)

Existing

of New EHB standard units.

methods of planning and controlling

examined in relation

to their

suitability

projects

are

for scheduling

the development process of a new product range such that


the overall

time required

to develop a product is minimized.

Chapter 2 concludes with an examination of the causes of design


changes and estimates the additional

design resources necessary to

carry them out.

Chapters 3 and 4 exmine: the existing

cost estimating

system at

H.M. Ltd. and a system is developed that can be used by designers.


major objective
estimating
therefore

The

of the present research is to develop a responsive

system that allows cost data to be generated quickly


the use of cost data in the planning and control

increasing

of the NPDprocess and the identification

and generation of minimum

cost design solutions.


IL

Whenexamined from a systems viewpoint


estimating

systems are clearly

estimating-systems

identified.

the faults

with existing

The prime causes for

being unresponsive are the work measurement techniques

used to estimate component manufacturing times since they require


personnel with detailed
methods used within
estimates.

knowledge of the production

a company in order to develop accurate cost

Hence the design function

separate estimating

must request cost data from a

or manufacture planning department.

departments often have other responsibilities,


must often queue until
the provision

Allowing

Since these

requests for cost data

service time becomes available

and so delaying

data
to the design function.
cost
of

designers to cost estimate designs would undoubtedly

improve the responsiveness of an estimating


of estimating

processes and

component manufacturing

system.

However, a method

times needs to be developed that

does not require

users to possess an in depth knowledge of manufacturing

In this

processes.

respect the technique of Multiple

has been found useful

in developing cost estimating

the causal relationship

Regression Analysis
equations by allowing

between manufacturing times and suitable

predictor

to be established.

variables

The development of a system for estimating

component manufacturing

times which does not rely on a knowledge of machine process variables


(this

type of data is not normally available, to designers)

generation of a large number of estimating

equations to be quickly

arrangement

general

drawings

for

when it

the necessity

of preparing

design

product

to prepare detailed

drawings'for

product

drawings,
Also it

concepts.

therefore,

of standard units,

development and use of cost/design

costs

since cost/design

may not be possible

the optimum number and


Chapter 5 examines the

specification

a model has_been developed


alternatively

estimated

components.

specification
that

models for estimating

allows

In

models have limitations


total

without

e. g. choosing

product costs and for generating minimum cost designs.

addition

cost

know total

determining

However there

are available.

arrangement

general

between alternative

total

designs

is an advantage-to

are instances

specification

retrieved.

system developed allows cost data to be generated

The estimating
if

Hence

models has been developed to allow

estimating

a method of classifying

the

This is because

equations.

of manufacturing processes used at H.M. Ltd.

of the wide variety

appropriate

required

product

costs

from a knowledge of a product's

to be

ten highest

The examination of existing


and PERT) indicated

estimating

planning techniques (e. g. CPM

that they were unsuitable

Chapter 6 therefore

process.

project

for scheduling the NPD

describes the development of a method for

the development time for individual

the type of task interactions

components and identifying

that exist.

The designer can use the method developed to quickly


components that would lie
hence constantly

time is available
proposed therefore

path as design proceeds and

element in a scheduling system when limited


is the effective

allocation

of this

to the overall

success of the project.

installing

a Computer Aided

Draughting system, and are planning to extend this


aid,

the opportunity

of CADsystems to a NPDproject

involved

It is

(Chapter 6) that both development time and resources

Since H.M. Ltd. are currently

important

time.

development

between components to be designed on the basis of relative

cost and importance

manufacturing

those

reschedule the NPDprocess to avoid delay.

An important

are allocated

on the critical

identify

element of the total


in the original

was taken of assessing the contribution

in. which original


design activities.

or creative

of the cost estimating

design worksformed an
The individual

design process are therefore

(Chapter 7) and the present contribution


and Design systems assessed.

into a design and

In addition

tasks
examined

of Computer-Aided Draughting
the inclusion

into a CADsystem

system developed in the present work is examined.

1
CHAPTER
PRODUCTPLANNING

9
1.0

Introduction
(2'3)

Of the established

the
methods of strategic planning
(4, S)
product planning technique
seeks to maintain an acceptable
level

profitability
and maintaining

by directing

an effective

product mix for a manufacturing

involves

The technique therefore

long term plans towards developing

selecting

long term

an appropriate

for each product type manufactured within

planning strategy

company.

an

organisation..

examining the manner in which strategic

Ansoff(3)

developed since 1950 established


strategies

that the evolution

planning had

of competitive

and corresponding managementresponses had not been one


which in conjunction

of replacement but one of proliferation,

of the future

the decreasing predictability

managementbeing confronted with strategic


complex structure

and containing

had now resulted

with
in

planning problems of

many inter-related

factors.

Product planning techniques help to reduce the complexity of


(6)
by isolating the influence of such
the strategic planning process
inter-related

(i)

factors

as:

Manufacturing
product
skills

resources,

design with

i. e. the compatibility

manufacturing

can be established.

(ii)

Changing market requirements.

(iii)

Competitive

action.

facilities

of
and labour

10

(iv)

Product profitability.

(v)

Product obsolescence.

(vi)

Product worth, i. e. the total

benefit

of a product to

For example, does the product allow total

a company.

product range to be increased or improved.


.-P

(vii)

Investment risk.

(viii)

Market environmental changes, e. g. inflation,


legislation
government

(ix)

product

type effectively
rate

for

all

Price/quality

number and variety

exist

complexity
planning
since

planning

would

the company's product


cranes,

hoists,

electric

lifting

equipment.

chain

overhead

to retrieve

planning

variables
process.

be particularly

hoists,

crane ranges,
manual lifting

within

a company, since

of inter-relationships
and hence affect
Accordingly

applicable

range is diverse,

standard

is dependent on the

strategies

the type and level

of the long term planning

custom designed

in order

products

types manufactured

of product

between strategic

strategies

the required

of one

relationships.

influence

these two factors

increases

in sales

overhead costs.

The need to adopt product

that

i. e. a decline

interactions,.

consistent

(x)

and resource scarcity.

Product

recovery

recession,

the

product

to H. M. Ltd.

i. e. includes
electric
equipment

heavy

wire-rope
and hydraulic

11

The basic types of product plans identified


in Table 1.

listed

TABLE1

PRODUCT
PLANNINGSTRATEGIES

(i)

Add new products.

(ii)

Delete obsolete products.

(iii)

Modify existing

Altering

products.

of product types within

the volume ratio

product mix could be considered an alternative


However

this

to achieve
profit

(7)
by Lawrence
are

option

such short

is usually

restricted

term objectives

the overall

product planning strategy.

to enabling

organisations

as the maximization

of annual

levels.

Alternative
orientated
as product
objectives.

strategic

planning

actions)

are considered

planning

techniques

strategies

product planning as illustrated

by Kotler(8)

in achieving

However the author

that these alternative

methods (i. e. market and financially


to be equally

an organisation's

of the work reported

can benefit
in Table 2.

effective
long term"

here considers

from an element of

12

THE CONTRIBUTIONOF PRODUCTPLANNINGTO MARKETINGAND

TABLE 2

FINANCIAL BASEDSTRATEGIES

1.
2.

Strategy

Product Planning Contribution

Find new users of the product )

Improve product specifications

Find more 'uses of the product)

to

increase the number of possible


users and/or uses.

3.

Provide more effective

Modify product to improve sales

sales

appeal, e. g. include optional

promotion

extras,

improve product design style.


4.

Provide superior
(Sales, Technical,

Improve product quality,

services
Maintenance)

and maintainability

reliability

to reduce need

for service facilities.


5.

Reduce product manufacturing costs

Provide price discounts

to offset

reduction

in product

profitability.
6.

Sell to specific

Modify product design to customer's

customers

particular

Significant
is available

literature

requirements.

on corporate and product planning procedures

by
Gilmore
being
and
published
with useful surveys

Bandenburg(9), Cunningham and Hammouda(l0) and Berry(ll).


detailed

framework for preparing

strategic

by
Gilmore
plans provided

illustrate
be
l,.
to
Figure
Bandenburg,
can
used
and
process,

i. e. the activities

Strategy"

and "Specification

constituting

The

the product planning

the "Formulation

of Competitive

be
to
Action"
underneed
Programme
of
of

taken for each product type manufactured within

a company.

13
FIGURE 1

STRATEGICPLANNINGFRAMEWORK(after Gilmore

and

Bandenburg

(9)

14

fN

a
"r
4-3

,Q
N
. "

a)
U
C

! f
r

Vf
4)
"r

"r
"r

0
4- ' -0
t0
S.. i]..
C.

V)

G)
Q1
C
-ru-

N
"r
i

co
4-3
C
rt3
G1. e
E >
'C7
0
U

r
"r

"r
H
ro
W
LL.

.4

td)

CM

N
C)

>,
rn
Sc

a)

(1)

>2

>.,

NNN
N

N
>
.
+3
"r
4-)
a)
CL
E
0
U

-U

NNNNNNNN

O
rM

Q)

.r4-3
V

4""3
C

G)
>N

.Y
(CS

++

4-)

CL

CO

"r

>

CM

rN

MMM

i
G.

Q:

dC'M

Vf

Ln

b
C
eo

O
"r

CN
+

a) >c

C
m a 0
.4Jr.
EW
U
co
Lc
'O
rn
0 ca 0

... 0,

.9

1.0

G)

R$
>
"r
-0
4-3
V
(0 V) tii 1a
C

4)

"r

i
tt)
CL
E
0
U

1/1,

a)
rn
"r+-)

a)
>

C)

C)

U
C

S.

(0
C
"r

(0
M:

4O

a)

tC

L
rn
0
50

LL.

..

00

01

C'M MMM

S_
CD)
O
S.,
a

O.
d'

r.

st

N.
r
"C:

.ti

'-I
Y

.
4J
. r.

4)

0
0.
CL
0

rn
(1)

vC

>C

cc
N

CU

+j

"r
4-3

G)

G1.4)

.a

0
"r

i
N

C
0
"r

co
0

oE

C.
.-CCo

"
+3

ouu
(L) t0
+> a.
oEE
a
-"

,a

"-

O
0
>
(O
4)
M
C
G)
4O

ooC

O.

+'
Ui

4"r

r4-"
oa
i

d-

(3
o.
-+

o
U

V1
-p

ai
3
a)
z

rrt)
"r
4..)
C
C)
OU
a
+>
C
rt3
i

t
a)
CO
"r
Lt_

N-4

>
+)
-rr"r
0
(0
L1.
ttS
G)
>U
"r
+
rt3

.'40
m i.
"r4--

NC
'C

0
Si
O.

aj

"r4-

>>
{-, 1
-rr
"r

t
fl.
tu

H
O
H
"r-

S0

c
0

a
E
0
U

.r
N
N
"r
E

+
"r

r"
. r..
.a
"r
N
to

V1
-

E
0
0
U
W

(1)
LL.

"r
LL

U0
r

W
C,'3
U.

r--

M..

t.f).

l0 .9

. C.
CO Ct1
rrrrrrrrrNN

..
NM

l, C)

l0

n.

00

011

r.

15

Figure 1 is also useful

in illustrating

important aspects of the

product planning process such as:

(i)

The need for co-operation

departments, i. e. both assessing the business

functional

of a company (activities

and manufacturing capabilities


13-19) and undertaking

(ii)

product development tasks

(activities

34-39) require

integration

and synergy.

Individual

product

the economic mission,


that

objectives

in addition

Wilson,

to considering
other

to the development

and integrated
i. e.
(12)
al.

of an organisation,

can be achieved.

of an organisation,

environment
essential

a high level of departmental

plans must be developed

the overall

such that

indicate

between

and co-ordination

et.

the total

business

characteristics

of effective

product

plans

are:

(a)

they must be future

(b)

they must be pro-active,

i. e. plans need to be

actioned now with future

benefits

(c)

orientated,

they must be flexible,

'i. e. easily

in mind, and
adapted to such

changing circumstances in the business environment


as unexpected competitive
This

need for

active

product

and flexible

to the strategic
future
effective

activities
product

action.

plans to be future
indicates

planning

is essential
plans.

the importance

process since

pro-

orientated,

of forecasting

a knowledge of

to the development

of

16

(iii)

The product development process (activities

or developing new products is

modifying
existing
.

to the basic product planning strategies

essential
listed

34-38) for

in Table 1 since even abandoning an obsolete

product normally necessitates

the development of a

replacement product.
i

(iv)

During the product development process the two basic.


elements of risk

involved

i. e. information

and performance uncertainty,

Information

in the product planning process

(13)

interact.

is associated with the uncertainty

uncertainty

surrounding the accuracy of various types of data used


(e. g.

sales

forecasts

and cost

estimates)

the prime cause of poor product


Performance

risk

the ability

of personnel

process,

plans

and is often

being developed.

due to the uncertainty

arises

involved

i. e. the relative

surrounding

in the development

success of development

is unknown and can be responsible

for

effective

activities
product

plans proving unsuccessful.

1.1

The New Product Development Process


The new product development process (NPD) has been well researched
a detailed

with Uman(14) providing

account of the stages involved,

and

flow charts being developed by Capon and Hulbert(15) and


(16).
The flow chart developed by Capon and Hulbert,
Cardozo and Ross

decision

Figure 2, is used in the present work to illustrate


between manufacturing

organisations

difficulties
the
hence
and
procedures.

involved

that exists

the variability

in the NPDprocess

in standardising

development

17

FIGURE 2

NPD DECISION FLOW CHART

's
(Ref.
1. SIZE OF POTENTIAL

It.

POTENTIAL
aouect

IURRIT
2 COMPANY KNOWLIDGI
OF THIS MARKET
3. COM/LG1INTARIT7
TTN
COMPANY STRUCTURE

WR[[T
"v14U

n
IDEA
WORTH

Im

OOOUCT.

P"
CMdTT[t

! cllv[

t0t

T117M1
IDtA?

PussumC?

11

LS
5lGM
rROOUCTIOM

sm4e
Mf

P OOVCT TIIT
$WJIA!
TO

'

TRIAL

/YTUOt

ISIST7MC
naoucTSl

Yis

1. tOFlTA4Lln
L CAPITAL
REQUIRED

non
aSCUm.!
cawrru
tW ous

oon
! OllR
r&roM tlSar
f..

Y1.
tJC, U, I?

CO..

UND

REPORTTo
eaecunre
cONrTTee

.
Rlsws.

IMITLL

tow. cu surIc"
ot. nc nrt

MAtttTIML
ttAttS
SttvlCn
at TAILED
AXIKET
IP
DATA iwtvtT

SCMMIJLJ.
ttwttrtwii

CAPITAL

O(iY120

COSTS

MlCt3YtT)

lO1IR
OIMTIO,

Gl.
IYOGIT

AMC lOOUCT
olrAUI1o,

M[t

//Oa,

cT

cu.

+mc
tGRCT,

"nunu
& wu.
D1sCUSSIi
uroaT

1. fORT.
,. c.. I".

IIOONCTIOq
T111A13 COW IKI
ruTT

Does
!I INTiva
corN rT! [
LpvB
fO1lRt

DOES I
COW AMT
SOARD
YtOr!
fOJlCTT
lO1lCT
IT

I. wawr..

u.,

: ......
. ea. en

DISCVSBD
PIODUCT

OIYIISIPICGTIOM

1W

COrwTTII

TO

WPLL
DETERMINE
CATbMS
FOR WOOL
CDMPUNT

18

The flow chart indicates

that the NPDprocess is controlled

a New Product Committee and a Product Manager.

However these formal

managementsystems are but two of several alternative


controlling

the NPDprocess.

by

methods of

Other methods are Research and

Development Departments, Commercial and Technical

Project Managers,

Commercial and Technical Line Managementand Commercial and Technical


One-ManEfforts.

The type of managementsystem appropriate

to a NPD

is according to Souder(17) dependent on:

project

(i)

The nature

(ii)

The markets

(iii)

The relative

(iv)

The capability

of the technologies

being managed.

being served.

cost

of the development

project.

of the system to handle

innovative

conditions.

At H.M. Ltd. the NPDprocess is controlled


consisting

by a Progress Committee

of membersof each of the main functional

i. e. Design, Commercial, Quality

departments involved,

Assurance, Production Engineering and

Manufacturing.

Considering the level

of technology involved

New EHBrange a multi-functional


Souder(17) prove effective.
work indicates
if

in develooing the

NPDcommittee would according to


However experience gained during the present

that the effectiveness

of using an NPDCommittee is limitel

the frequency with which committees meet and attendance at committee

is
adversely affected
meetings

by assigning

a low priority

to NPDwork.

19

is difficult

Under these conditions

it

expended with benefits

(optimality

The type and level

to optimize development resources


of the product design) obtained.

involved

of activities

in Product Formalization

developed,
being
type
dependent
the
of
product
e. g.;..
on
are

(i)ce

High sales volumes indicate

that it

is important

to

design
in terms of manufacturing
a
product
optimize
cost, ease of manufacture and range of market
specifications
(ii)

Innovative

covered.

and high technology products indicate

extensive resources may require


Formal.ization
the activities

stage.
carried

the resources allocated

a large proportion

out could involve

The need to develop a reliable

(iii)

required

to specify

basic research.

to prototype

testing

and product development.

activities.

necessary are

The individual

and develop a new product are obviously

developed
being
therefore
type,
the
and
would
product
of
on
considerably

1.1.1

of

product could influence

To launch a new product the two basic activities


product specification

to the Product

allocating

In addition

that

tasks

dependent
vary

between manufacturing organisations.

Product Specification
The tasks involved

(i)

in specifying

Finding new product ideas.

a new product include:

20

(ii)

Selecting

(iii)

Obtaining detailed

suitable

specifications

new product ideas.


design, cost and market

for the new product.

(18)
For each stage of the NPDprocess Fogg
estimated the
average number of new product ideas needed to achieve one successful

product -as:

Average Number of New Product Ideas

NPDStage
Idea

40

Technical Search

10

Technical Feasibility

Prototype Testing

Pilot

2.5

Production

Launch

Success

The high rate of project

failures

indicates

that organisations

should endeavour to maintain an adequate flow of new product


ideas in order to ensure an acceptable number of product
successes.

However the available

in an organisation

effectively

development resources withlimits

be
ideas
that
can
considered.
product
to ensure successful
adopt alternative
products,
successful

the number of new


Under these circumstances

products are developed, organisations

approaches to the problem of finding

e. g. H.M. Ltd. formerly

must

new

developing
the
of
cost
reduced

products by concentrating

on improving existing

21

products and hence reducing the technical,


marketing effort

manufacturing and
(19)
According to Cooper
by remain-

required.

ing with existing

products and markets the probability

of

marketing a successful

product is improved since the company

can acquire an in-depth

knowledge of product technology and


However the concept of market evolution

market characteristics.
8)

described by Kotler(

indicates

that since the development of

a market is accompanied by a marked increase in competitive


action the chances of launching successful-products

can be

reduced.

The obvious
two approaches

(i. e. modify

develop, new products),


Table

following.

work at H. M. Ltd.
Chain Hoist,
Universal

to adopt therefore

strategy

3 indicates
and shows that

has direct

customer requirements.
to aid the methodical

data collection

is

and

presently

of development

(New Electric

Centrelift

(Hydraulic

the

and

Crane and Piledriver)

developed.

normally the responsibility

questionnaires

status

both improved

Wire-Rope Hoist,

Obtaining a detailed

function

products

which H. M. Ltd.

the present

Crane) and new products

are being actively

this

and improve existing

a strategy

New Electric

is to balance

design specification

for a product is

marketing
department since
the
of

access to the market place and hence


(20)
Several techniques are available

collection

and surveys.

of product design data such as

However the effectiveness

of the

stage is dependent on the type of data collected

22
PRODUCT
DEVELOPMENT
MATRIXFORH.M. LTD.

TABLE3

Adapted from Berry


Increasing Technological

Effort

No Product

Product

New

Change

Improvement

Products

011

Universal
Crane
Hydraulic

+) a)
a)rn

i
O
4
W

(0

Jack

Range

Centre
Lift

Crane Range

m
. r.

C)

'Y

rO

New Electric
. r..

Chain Hoist

t0
C)
N

x V)
110W
+-J
W
0Y
W i-3 i
M 4-)
LO
C

4--+ W

Z
O N

,-m -14
s.

>b

New EHB

Hydraulic

Range

Piledriver

Hydraulic
Crane

(11

23

it,

as well as the method used to collect


that all
influence

functional

i. e. it

is essential

departments involved in the NPDprocess can

the type of data collected

information

since their

requirements are often not compatible with those of the marketing


department.
process.

Consequently problems can arise during the NPD

For example, several problems arising

during the

development of the New EHBrange were:

(i)

The maximumlength of hoist block that the market


would accept was not obtained leading to problems
drum tube and related

in minimizing
costs.

component

It was necessary to determine the drum

tube length to outside diameter ratio


overall
(ii)

design costs.

The maximumspeed of lift

acceptable to the market

defined.
not
was

This data would have allowed gear

ratios

to be minimized and hence gear box

reduction
material
(iii)

that minimized

costs.

The relationship

between height of lift

(and hence

drum
length
size) and sales volumes
and
rope
rope
was difficult

to ascertain

gathered causing uncertainty

from the type of data


during the determination

lengths
drum
tube
the
standard
optimum number of
of
to adopt for each standard of New EHBunit.

24

(iv)

The market survey was primarily

designed to obtain

data for EHBunits sold for "runway duty", i. e.


,
(21).
However the objectives of the
Class I cranes
design department required developing hoist
for "crane duty",

i. e. Class II cranes.

department
the variation
marketing
I.,

blocks

For the

between hoist

blocks for Class I and Clase II cranes is negligible,


however to the design department the variation
represents significant

design differences.

Hence

data obtained from the market survey was not always


relevant

to the design problem in-hand.

Product Development

1.1.2

Product
of design,
following

development

manufacture

is an iterative

and test-activities

sequence:

(i)

Design

(ii)

Prototype build

(iii)

Prototype test

(iv)

Jig and fixture

design

(v)

Jig and fixture

manufacture

(vi)

Pre-production

(vii)

Pre-launch market trials

manufacture

process(22)
carried

consisting

out in the

25

The iterative

nature of the NPDprocess often requires

the

or parts of the above sequence to be repeated.

entire

involved in the product development process

Since the risk

with which development tasks could be

arises from the uncertainty


performed and the uncertainty

surrounding the accuracy of data

used during the NPDprocess, the present work considers the


overall

involved

risk

(hence the effectiveness

development process) and relates

this

of components that form the product.

involved

These factors

determine

and input data

in the development process, e. g. the cost estimates

required,
parts,

to the number and type

type of activities

the number of activities,

of the product

type of design decisions,


tests

reliability

manufacture of prototype

required and the jigs

and fixtures

necessary.

since planning the NPDprocess requires

In addition

development activities

are co-ordinated

costs are minimized whilst


benefits
maximum
complexity

that

such that development

at the same time ensuring that

are obtained from the product design, the

is
development
influenced
the
process
also
of

by

the number and type of components involved.

p) 1.2

NPD Success/Failure
Identifying

ful

the major factors

and unsuccessful

developing

Criteria

prescriptive

new industrial

that differentiate

between success-

products is an initial

step to

development
for
the
process
product
new
guides

26

which might separate the "winners"

since many of the variables


the "losers"
that,

determining

additionally,

of the firm.

Cooper(23) considers

importance of these

the relative

to the effective

is essential

factors

the control

are within

from

of development

allocation

resources.

Activity
such as SAPPHO(Scientific
(24),
Predictor from Patterns with Heuristic Origins)
and that
(25)
identifying
Gerstenfeld
the
by
centred
on
problems
out
carried
Initially

investigations

facing product developers by comparing successful


the differences

new products and establishing


SAPPHO
the
project
example,

for the formulation

in which the criterion


i. e. the successful

employed a "paired

and unsuccessful

and unsuccessful

between them.

For

comparison" technique

of a pair was commercial,

product competed for the same

market.

The effectiveness
comparison
failure
and

(i)

with
that

of the New EHB project

can be-examined

main areas of difference

the five

by

between success

emerged from the SAPPHOstudy:

Successful innovators
understanding

were observed to have a much better

of user needs.

market research study(26)


the wire-rope
information

In this

carried

respect the detailed

out by H.M. Ltd. into

hoist market has provided much useful


of user needs.

Companies established

within

a market must continually

these
can
alter
since
needs
user
monitor
market environment.

with changes in

For example, at H.M. Ltd. the Electric

27

Chain Hoist range has recently

been re-designed on a

modular concept to enable distributors

to maintain an

adequate product range with low stock holding costs,


i. e. current

interest

rates on borrowed capital

are

high hence stock holding costs have assumed greater


importance.
(ii)

innovators

Successful

The level

publicity.

to employ during

of marketing

the NPD process


At H. M. Ltd.

of distribution.

to marketing

pay more attention

and advertising
is related

the majority

and

resources

to the method
of hoist

blocks are sold through a small number of major distribhence the level of marketing and publicity

utors,

required

is low.
Publicity

is of course generated by the reputation

in the

developing
the
the new product.
of
company
place
market
the case of H.M. Ltd. market reputation
(iii)

Successful innovators
and scientific
the specific
policy

area concerned.

In this

of H.M. Ltd. is to make full

the National

is high.

make more use of outside technology

advice, not necessarily

such as Leicester

In

University,

in general but in
respect the development

use of outside institutions

Loughborough University

Engineering Laboratory.

and

28

(iv)

counterparts

individual

in unsuccessful projects.
for the successful

responsible

The

to integrate

departments involved.

the various functional

of these individual

than

development

of a new product must have the authority

objectives

attempts

more senior and had greater authority

were usually
their

in the successful

individuals

The responsible

functional

The

departments

are normally at variance during product development(27)


individual

hence the responsible

must have the authority


In recognition

those objectives.

to reconcile

of these

responsibility

for NPDat H.M. Ltd. is that of the

Technical Director,

who is a memberof the main board of

facts

directors.
(v)

Successful innovators

perform their

but not necessarily

efficiently,
case of failed
determining

initiatives.

more quickly

It is in this

than in the

area that

importance of product development

the relative

would allow effective

criteria

development work more

allocation'of

development

resources.

models)

(e. g. rating

of techniques

A variety

were initially

importance

scales

used to determine

of development

criteria

and checklist

the relative

to the success or failure


test

Wind(28)

developed

a conceptual

model to screen new product

ideas for

market acceptability

of a new product.

by estimating
development

importance of
the relative
(29)
Cooper
compared the relative

subjectively
criteria.

29

importance of factors
i. e. each factor

using an arbitrary

weighted average,

was assigned a score of 1.0 if

reason and 0.5 if

a contributory

The

cause of failure.

score was then established

weighted rating

a primary

for each factor

from 101 examples of new product failures.


(23)
work Cooper
overcame the limitations

In later

previous work (i. e. basing the relative


development criteria
empirically

based weightings
of Cooper's

The application
project
that

is limited

since

that

service

estimates)

and developed

for success/failure

criteria.

work to the present

New EHB
observed

found to be significant
between industries.

as product

and competitive

price,

product

in
the market environment
changes

for

In addition

importance

the relative

criteria

success/failure
quality,

varied

importance of

the SAPPHO(24) project

factors

the individual

success or failure
(30)
Evans
observed

with

on subjective

of

of such

reliability,
prices

altered

(e. g. from energy

abundance to scarcity).

At H.M. Ltd. therefore


empirically

it

is not possible to develop

based models for predicting

success or failure

the potential

of a new product since the appropriate

data on past new product developments and corresponding


market conditions

does not exist.

30

However Cooper(23) developed a comprehensive list

from past research projects

criteria

success/failure

of

which could be used as a guide for the planning and


of the present New EHBand future

control
projects.

The criteria

listed

development

by Cooper indicated

importance of the design function

the

to the success of a

oe

new product since many of the factors

of the design department.

responsibility
particularly

This is

true at H.M. Ltd. since the design department

has the added responsibility


the NPDprocess.

of planning and controlling

(23)
Table 4 has been adapted from Cooper

to define more clearly

the influence

on new product success/failure

TABLE 4

involved are the

of the design function

criteria.

THE DESIGN FUNCTIONSINFLUENCEON NEWPRODUCT


SUCCESS/FAILURECRITERIA
Adapted from Cooper(23)

1.

Technical

(i)

and Production

Ensure compatibility
production

(ii)

of engineering

skills

and

resources with product design.

Ensure the design department carry out preliminary


technical

(iii)

Synergy and Proficiency

assessment and product development well.

Ensure that product design does not prevent


prototype

testing

being efficiently

and pilot
carried

production

out.

start-up

from

31

TABLE4 (Cont'd)

(iv)

Ensure design staff

fully

product design and manufacturing

technology,

technology used within

2.

understand product

the market and company.

Marketing Knowledge and Proficiency


(i)

Ensure that
study,

preliminary

test

provide

marketing

the design

assessment,

and market

function

which to base product

(ii)

market

with

launch
relevant

data on

understands customer's

buyer
price sensitivity,
and
wants,
needs

3.

phases

design.

Ensure design department fully

situation,

market

competitive

buyer behaviour and size of potential

market.

Newness to the Firm


(i)

Whenpotential
processes

customers, product use, production

required, distribution

method and product

technology are new to the firm the design function's


responsibility

is to ensure that product design is

factors.
these
with
compatible
(ii)

When new competitors


department
competitive

enter

the market

must have the capability


product

the design

of analysing

designs.
a

32
TABLE4 (Cont`d)
4.

Product Uniqueness and Superiority


(i)

The design department must ensure that the product


design, if necessary, has unique features

for the

to competing products in meeting

is
superior
customer,

customer's needs, allows customers to reduce costs,


carries

out unique tasks for customers and is of a

higher quality

5.

Product

design must allow

necessary,
product

in a highly

price

(i)

is important

to the buying

are presently

and Managerial

with

satisfied

The design department must ensure that

company's managerial skills


when advertising

(i)

decision

Synergy

is compatible with financial

Market

market where

products.

competitors'

Marketing

if

company to sell,

competitive

customers

and potential

7,

products.

Market Competitiveness and Customer Satisfaction


(i)

6.

than competitors

product design

resources available,

the

and aids the sales force

and marketing the product.

Dynamism
The design

department

situations

satisfactorily
frequently

must be able

introduced

change rapidly.

to handle

in which new products

to the market

and users'

are
needs

33

1.3

Forecasting for NPD


The ability

to forecast

the need for NPDprovides many advantages

including:

(i)

of R&D costs could be improved since product

The control

development is carried
to the alternative

of maintaining

planned product

strategies.

obsolescence

(ii)

out only when necessary as opposed

Adequate time can be allowed to carry out the NPDprocess


on development costs of schedule

hence reducing the effect


slippages.
(iii)

Adequate time can be allowed to ensure that the new


product design satisfies
requirements,

both market and manufacturing

i. e. in terms of product cost,

specification

and function.

Bentley
planning

(31)

(32)
have demonstrated that effective
and Wasson

strategies

can be developed using the sales or profit

life
its
during
market
of a product
forecasting
of
problem
products'

market life

for NPD involves

predicting

behaviour

Hence the

the shape of a

(PLC).
curve
cycle

Figure 3 illustrates
indicates:

as a planning tool.

the shape of a typical

product

PLC curve and

BEST COPY
AVAILABLE
Variable print quality

34

FIGURE 3

THE PRODUCTLIFE CYCLE AND DYNAMICCOMPETITIVE STRATEGY


32)

(Ref.

>
)Dt,

ioiS1

leaf Tiny product)

N
L
CL
A-Low

Learning

Products

1126,

TIME

Iesrdng
Injsalss
requirements.
loco,
and r.. nedy offering dedevelop . Idespread
facts puLkly.
and gals
vnr. oeos of benefits.
dopt. rs
vial by nrfy

ST tATEG7
DISUECTIVE

To establish a strong brand market


i. h. as quickly as
and discrlbution
possible

To

maintain

and

the marstrengttao
kilt niche achieved
through dealer and
for LLY
eooz-,

To deferd brand position against cotePilling brands and product category aBlast other potential prm. acts. through
eosstant att, taba to prod. wt Isnprrn. meet opportunles
and fresh promotional

OUTLOOK FnIt
COMPETTTIOY

is
f: ooe is 13ualy to be attracted
stages
the arty.
sprofitablit

Early

. nuance
also emulators

of numerous

aggr..

Price
tlon

her

Llmasd
hyslcel

of
sd

uodact
,

t! odular
s. Shclbn

with

models
offering

. slgss tsoth fcoossed on m1o1miaing learoing requirements.


aalgtu tx a_ and us. -englneerd
it.

t tr. o

. srL

and

snontroi

mart.
PRI21716
ONECTrVC

to east

appeal

"

tuft

. iimination

defects

rev sp(IvS

most

red.

segm

discounts

Cen

.e aliens

Oo"ect"es

1-du

JUTHIDt
POLICY

mu

ii

- iT E LLIGENCL
i OCI: 3

trans

and sampil, tf

awarenen
al Create widespread
and onto stasdieq of offering
benefits
b)Go:. trial by surly adopters
To order of slum
puytviry
Pursusi
sales
Mass e-oenewaodeuions

searo of a substa.

uN

and

A price lie. for every taste,


to. -. td to premium m. d. l.
Customary
trade discounts
Aggressive

1-1

promot

from

pricing,

f
with prices cal as fast
as coati
decline due to accumulated pro-

. eake.,
.

to substautlsl

protement

is

alert

Ine. naufid

. tt-ti..

A co.

to

improve-

opPoetsait1es

product

naus

Is the

change

tal ;. created

ab-

tea-

brand

some

_. artet

through

tithe,

t ug
py_sme.
hold cost-

ad price-penvtra!
un of ae. wartets
laraiucttoei
or major product . ha-ges
Coootase nUeotlos
of Danker products
for

iid

of deaigs

Cur-stet
line

'-tog

.1

rti. niart.

to

say r .us rot entura. ag a dl resw


prof..

tieveveemest

poxes

Re oe. vioatluo

cost cutting-

arrest

eemar

for

to possibilities

nod drops.
ping our becaaase or
decrease La Ce.

of

co-premises

.. y prod-

pricing to preserve
product
Search for Incr. category franchise.
mental pricing opport naties. tochling

pricing

private

oppe. rtunl-

Defeufto

Label

to

eowncts,

boo.

. alum.

Mal-.

re

of Pre-

level prlc Log . hl


tom, le".. disregard
ft

of an..

-!!.

es

market

slur.

end d, 1 on -prrirnee

and strengthen brad proferenre mmnng trade and final users


Stimulate general trial

Maintain consumer
franchise and
strengthen dealer
ties

S inst.
consumer and trade lot slay.
. itta atroe emphases on dealers and dinPromor. on of greater tee
trihutora.
frequetsey

Lust esauth I---Ietat- profit. blit dtstntaicw

Masa media
Personal sites
Sales pnn. oliona,

Mass media
Dealer pronotions
Pnrsoul
selllrg; n0

%Iasa md).
Dealer . nested

Cut d-.
sD wedis so
she tax-+so
ne

los

including

same-

peomolbos

Phase

We.

dealer.

out.

pr-on.

keepteg

-u.,,

a of

any k. ad

Sales promotions
Public dy

attention

to hand

. rage.

In med-l
and
M to gilortu

market

segmtntatlnn

to gnpn

snsage

of sampling

Create

Detallcd

d.

experience

Publicity

uso-

Increased attontlolt
to marke(-broadenIng and promotional
ilea

ling

e-ual de"elcying
s. s

is

went, tlght., ung up


ei line to elimnca:.
uoneceasary special
ties with little mukilt appeal

Intensive

Intensive and enensiv..


with dealer
E-lulte
or selectl. e. with
)teat high enough to keep
dlstrltsaoe
margins "rh enuuZ, to margins
Close attention to
. Lending them interested.
honey promotional
justify
rapid resupply of dlstrllaaor
stocks
u oil level,
and heavy inventories

systetss

deeis-C

neceuLry

duct

to uncover

n-

oon. Peeltioo

in design

l "his

To Ideetuy

Slmatar

stabslLad.
with fe. or w
exrases am mariet
shares taut

out the . esker

en-

approaches

Cowpetitbs
subject

of

--__bloss valuable

to so-

cry new segment and new tea. -spa


teed as fast as discovered

lntena lficatlun
(yT; O AI.
GU'DEI. UE. 5

to appeal

of variants

to quality

the wlnlnmm
of value
Ie rntng and to match
Perception of
r.! erer.

prcepcbs
the gal.
High

gaick

flaalbl.

set-

t'tlve

attention

t-rwreald

To imp_

the

rer.

to fac111tat.

design

distribution

squeeses
on the
industry.
shakl,
g

tra"

iR(. Dt. CT
7 LS; G::
OIy ECTTVC

and dlstrlbu-

and

To n: t n it. oC., t g
dry of ail passable
profit

market
miles

poshlo,.
confor

and exenad a strong


sir.,
emphasis on keeping
deader . ell supplied
but . Ito minimum
Inventory cost to hi

with . rang
and crania.,
emphasis on keet'mg dealer .. it auiplied.
but . mlaimum sen. nitory co. to him

I`tus. -. a eaten, as
they ; even, saarg, nal

Close attention

Intensifev

attertiw

product

Infornatbo

Imprvnemew

a.

potential
for
lid

to

In. pro"

product
1

tads.

to

broadening

to
em<n

marketch.,

Ccii.

fresh
to L'oasihie
Inn them..
pron..
sryf

)Moot,.

Inner-pedua
nor
. 1gns of teglnnisg

to pmsablo
Sharp

alert

for

c. "m pea lib.


decline
product

de-ally

at . roes
shu.

tslniq

the
LL

p. h. t

Prodset

Se passed

o. a

35

(i)

The rate at which a new product's

sales (or profit)

increases.
(ii)

The maximumsales (or profit)

(iii)

The period maximumsales (or profits)

level

reached.
are sustained

before they begin to decline.


(iv)

The rate at which sales (or profits)

Hence forecasting
existing

products,

decline.

the time and rate at which sales decline

and forecasting

and the maximumsales level

the rate at which sales increase

reached for new products could allow the

NPD process to begin such as to minimize the reduction


This profit

level

the sales decline

reduction

in profit

levels.

normally occurs during the period between

of an existing

product and the increase in profit

illustrated
as
product
a replacement

The PLC has been formulated


a product's

for

of

in Figure 4.

as an explicit,

verifiable

model of

sales behaviour and tested against actual data in 140

by
Polli
Cook(33).
However
products
and
of
non-durable
categories
(34)
develop
Polli
Cook
to
PLC
that
by
Doyle
and
a
and
curve
attempts
behaviour
failed
types
a
product
the
of
range
of
sales
expressed
identifiable
have
an
although most products
too irregular
tool.

describe

cycle,

such cycles are

to permit a commonPLC curve to be used as a forecasting

Presently

organisation's

life

since

the PLC concept is used to aid the formulation


short-term

objectives.

of an

For example, Fox and Rink(35,36)

how the purchasing department can control

buying policies

indicates
3
Figure
by
PLC
the
and
occupied
a
product
the
of
stage
on

based
the

36

FIGURE 4

PROFITABILITY GAP DURING PRODUCT


REPLACEMENT

4-

0
U +4
>U

------.e

I.

(Uco
4-3

.-atE
4-

s.. (1) (1)


acs..

G)
N>
r0
G) 1
>a)
Q) En

'Sf

c\

++ .
.v
y..

`c

o
iU +)
a

a)
C.
"C_
(L)
ov

-v
co
"r i

a.

C)

c Cc

"r

"r

C)

o10

C_+J C
Ec)o
tc)
o=
C)T)r

P-"

Ui
(L) =

Vf
I/f
0

i
i
i
i
I
I
I
I
iL

0
L +)
C. C)
N
L
fC3

pi
C. +3

C.

3
o"ra. Xa)
ZWZ

Profit

(E's)

Q)
aj
E
. '..
F-

37

use of the PLC curve in aiding a company to set product design and
pricing

objectives,

promotional

guidelines

and distribution

policy.

techniques such as Time Series Analysis,


(20,37)
do not have the ability
to
Moving Averages and Trend Projection
Established

forecasting

the PLC curve since these techniques merely project past sales
'Of
demand forward into the short term future in order to allow effective
forecast

production

planning and scheduling to be. carried

techniques cannot forecast

PLC "turning

these points are not related

Hence these

(i. e. points at which

points"

level
the
sales
and
maximumsales are reached

out.

begins to decline)

to past sales levels

since

but to such factors

as:

(i)

Purchasing motives.

(ii)

Product

acceptability

(iii)

Present

and future

(iv)

Strength of competitors

in the market place.


competitive

action.

in the market place.

The present New EHBdevelopment provides a typical

example of

forecasting
term
short
that
when
using
only
arise
could
the problems
The market share of the existing

techniques.
been declining
identified
during this

for approximately

as a trend,

Ropemaster EHBrange had

5 years before the decline was

i. e. the sales of the Ropemaster range increased

the
total
due
to
of
expansion
an
period

's
Ltd.
M.
H.
market share was gradually
However

market size.

being eroded (Appendix I).

38

Short term forecasting

methods gave no warning that development work

was necessary since these techniques analysed sales trends only and
not underlying

causes such as, poor product design, uncompetitive

product prices and poor marketing strategies.

By the time the need for extensive improvement of the wire-rope


'Oe
design was identified
an additional product range (Electric Chain

hoist

had also become unprofitable.

Hoist)

were available

development funds

for beginning the development of both product ranges


The Electric

simultaneously.

since the profit

priority

Insufficient

for the EHB range.

Chain Hoist (ECH) range was given


for this

potential

range was greater than

The NPDprocess for the EHB range was therefore

further.

delayed still

Diffusion

theory used to model the growth of a new product in

the market (i. e. the initial

stages of the PLC curve) suggests that

there is a time lag in the adoption of products by different


companies
(38)
Rogers
that
is
first
proposed
a
place.
product
market
a
within
by
a select
adopted

few "innovators"

adopt the product.

Thus it

non-adopters
the diffusion

Initial
deterministic
(assuming

who in turn influence

is the interaction

others to

between adopters and

that is assumed to account for the rapid growth stage in


in
seen
the
PLC
curve.
as
process

attempts

to model the diffusion

interpretation
a constant

process(39)

of the time dependent

market size)

resulting

aspects

employed a
of diffusion

in an S-shape diffusion

39

curve for the

cumulative

number of adoptions,

Nac (Equation 1).

t=C
r"

(1)

dt

Nac =

2
2a

L2a2J

exp.
'L"

t=o
oe

Where:
exp. = Exponential

Constant.

Nac = Cumulative

number of adopters in time-c.

deviation

a=

Standard

t=

Time period.

11

Time at which fifty

of time.

per cent market potential

is

achieved.

Mahajan, et. al.


potential

(40,41)

considered that assuming a constant market


since it was equivalent

for a product was unrealistic

is pre-determined
fate
"the
product
of
a
that
statement

to the

at the time of

A model was presented by Mahajan and co-workers for such

introduction".

heater,
water
machines
and
gas
washing
as,
consumer products
the
ceiling
which

on the cumulative.

in

adoption or the market potential

for a product changed with time.

Since both deterministic


variables
within

(a

u)

and dynamic diffusion

that are unknown functions

the market they specify,

diffusion

of total

models contain
marketing activity

models can only be used for

40

interpreting

data and not for predicting

historical

the initial

models cannot forecast

That is diffusion

future

performance.

stages of the PLC

curve.

Chambers, et. al.


in terms of ability
"turning

points",

and operational

(37)

compared eighteen forecasting

both long and short term events, PLC

to predict

requirements for developing the forecasting


Although qualitative

cost.

identified
Research
Market
were
and
"turning

points",

evaluations

as adequate predictors

(37)

(e. g. Regression Analysis

sales trends to their


methods are applicable

This reduces the level

on subjective

suggest that causal forecasting

techniques

and Econometric Models) overcome the need for

since these methods employ equations that relate


In addition

causes.

underlying

causal forecasting
since to be

to product planning strategies

individual
developed
for
be
models must

effective

of PLC

and hence are prone to errors.

by "experts"

evaluations

method

methods such as Delphi

these techniques rely heavily

Chambers, et. al.

subjective

techniques

of inter-relationships

product types. -

between forecasting

variables.

Hence to develop
curve

In this

sales.
since

is necessary

it

this

viewpoint,

respect

for

techniques
to determine

the criteria

the research

is an area of market
the socio-economic

forecasting

into"buyer

research

factors

that

the shape of the PLC


that

affect

behaviour"is
examines,

product
relevant

from the buyers

behind the purchase decision.

41

Current research by Newa11(42) and GrOnhaug(43) is based on the theory


that since the buyer seldom has full
full

product information

consequences in choosing a particular

decision is perceived as a risk


perceived risk

(iii)

by Newall are:

the purchasing problem:

(a)

the size of the product

(b)

the type of purchase or buying

(c)

the degree of product

essentiality,

(d)

the factors

the purchase.

provoking

expenditure,

the industrial

Those describing

task,

and

buyer:

(a)

his level

of general and specific

(b)

his level

of decision

(c)

his purchase history,

(d)

his degree of technical

Those describing

the purchase

The determinants of

problem.

buying identified

in industrial

(i)ce Those describing

(ii)

supplier

or knows the

self-confidence,

experience,
and
and professional

affiliation.

the buying environment:

(a)

the size of the buyer company,

(b)

the financial

(c)

the degree of decision centralisation,

(d)

the degree of decision

standing of the buyer company,

routinisation.

and

42

Many of the determinants of perceived risk


are beyond the control

by Newall

company, i. e. those describing

buyer and the buying environment.

the industrial

However the factors

the purchasing problem need to be considered in relation

describing
to their

of the supplier

identified

effect

on sales levels.

In this

respect the work by Lehmann

and O'Shaughnessy(44) has shown that the perceived risk


loo
of problems: likely

in adopting a particular

to arise

importance of the criteria

the relative
illustrates

affecting

product)

influences

sales levels.

Table 5

in ranking between four product types tested

the variation

by Lehmannand O'Shaughnessy, each presenting


to the potential

(i. e. the type

different

types of risk

buyer:

Type I Routine Order Products - products that are frequently


in
learning
is
there
problem
no
ordered and
products..

there is no doubt that the product will

In addition

its
out
carry

how to use such

function.

Type II Procedural Problem Products - the buyer is confident


that the product will
in training

exists

Performance

Type III
whether
ation

the product

for

which it

carry out its

function,

however a problem

personnel to use the product.

Problem Products
will

perform

there
-

satisfactorily

is being considered.

is doubt as to
in the applic-

43

TABLE5

VARIABILITY IN ATTRIBUTERANKINGBETWEEN
PRODUCT
TYPES

Attribute

Ref.

(44)

Ranking
I

Reputation

4752

Financing

Flexibility

3525

Past Experience

Technical Service

12

II

III

IV

16

16

13

13

10

137

Confidence in
Salesmen

14

15

15

16

15

17

17

17

11

11

Convenience in
Ordering
data

Reliability

43

2881

Price

Technical
Specifications

5966

Ease of Use

10

Preference of User

13

14

13

14

278

Training

Offered

16

12

11

Training

Required

17

12

14

15

10

11

12

10

Reliability

of

Delivery

1414

Maintenance

Sales Service

76

44

Type IV Political
large capital
rival

Problem Products - when products necessitate

outlays problems can arise since there are always

projects for the allocation

political

of funds.

More frequently

problems arise when the product is an input to

several departments whose requirements may not be congruent.

The variation

importance of factors

in relative

purchasing decision with perceived risk


product (Table 5) indicates

that a supplier

in adopting the
his product

must relate

and attempt to deduce from this

to the prospective application


problems the customer is likely
In this way the criteria

involved

the

affecting

the

to encounter in adopting the product.


a product's

affecting

sales levels

can be

deduced.

In terms of forecasting
Kotler(8)

(i)

identifies

the decline of a product's

sales level

two basic causes:

Changing market needs - environmental


to fundamental

changes in market requirements,

increasing

prices

for

large

importance
influenced,

oil

automobiles
of factors

is leading
.

affecting

e. g. high interest

tend to increase

to a decline

In addition

the importance

sales
rates

lead

effects-often
e. g.

in demand

the relative.
levels

can be

on borrowed capital

of product

price

45

(ii)

Competitive action - the introduction

of an improved

product onto the market by the competition


an existing

can lead to

This agrees with

products sales decline.

the behaviour of buyers observed by Cunninghamand


(45)
i. e. the comparison of product offerin. gs by
White
,
buyers

potential
e
..

industrial

buyers carry out two tasks which require

Criteria

product

carried

must be established
offerings

against

and

offerings
out for

which potential

must be evaluated.
acceptable

specification,

type of product
exactly

the collection

product data:

needed and number of products


supplier's

product

buyers is provided by Webster(46) which shows that

by potential

supplier's

(ii)

on the

account of the method of comparing alternative

analysis of alternative

(i)

influence

purchase decision.

A detailed
offerings

has a considerable

price

These include
band, delivery

to be purchased.

must be identified

Alternative

hence a search is

who may be able to provide

suppliers

needed and an examination

the

is made of what

they have to offer.

Buyers initially

select

possess the specifications

those product offerings

that

demanded. Those products that

cannot meet the specifications

are eliminated.

then compares the remaining product offerings

The buyer
and seeks to

determine the product that minimizes the perceived risk


involved

in the purchase decision.

Cunningham and White

(47)

46

in examining the purchase of four types of standard


machine tools sold into the UK market support this
They found that many of the supplier

observation.

thought to affect

product attributes

and

the purchase

decision did so to the extent that buyers required


an absolute level

of achievement on the part of

before offers

suppliers

could be considered for

comparison with competing products.

In order to develop a method of forecasting


therefore

(i)

it

PLC "turning

points"

is necessary to:

the relevant

Identify

criteria

affecting

product

sales levels.
importance of these factors.

(ii)

Determine the relative

(iii)

Determine the relationship

between these factors

and

sales levels.

(i)

the Relevant Criteria

Identifying
Kiser,

al.

et.

(48)

using

attributes

have imputed some importance


a definitive

developed

list

to which researchers

in the evaluation
of supplier

of suppliers,

attributes

which can be used in the form of a check list.


relative

importance

as product
all

factors

any specific

of attributes

varies

with

Since the
such factors

type and background of the purchasing


listed
buying

in Table
situation.

(Table

executive

6 could not be appropriate

to

6),

47

TABLE6 SUPPLIERATTRIBUTESAFFECTINGTHE PURCHASING


DECISION
(Ref.
Convenience-RelatedAttributes:

"'1. Deliverieswithout constant follow-up


2. Can deliverquickly
3. Answers all communications promptly
4. Advises us of potential trouble

'S. Handlesrejectionspromptly and efficiently


6. Adapts to specificneeds

'7.
8.
'9.
'10.
11.
12.

Allows credit for scrap and salvage


Offers broad product line
Accepts small order quantities
Is located in close proximity
Anticipates our requirements
Is "direct" source of supply

Economic-Financial Attributes:
01. Has competitive prices

'2. Offers volumediscounts


43. Guaranteesprice protection
'4. Has favourablefinancial position
'S. Sellsat a lower price

'6. Offers higher cash discounts


'7. Offers extended payment terms
1.
2.
'3.
'4.
'S.

Calibre- Capability Attributes:


Regularly meets quality specifications
Maintains up-to-date stock
Has good packaging. including packaging
slips
Has knowledgeable salesmen
Has high calibre management

'6. Has technicalability and knowled,^,e


'7. Has potential to expand capacity
68. Helpful in providing special handling
equipment

9. Has research and development facilities

Laagt-Depc'. ' R!! ity Atrrtuees:


'1. Reliable in quality
'2. Reliable in delivery

'3. Is fair and honestin dealings


04. Keepspromises
S. Has favourablereputation
'6. Has favourableattitude
'7. Exhibits desire for business
8. Maintains favourable fabcur management
relations
'9. Is a progressivefirm
'10. Offers well-known brands and/or
products
'11. Utilises effectiveselling methods
'12. Is a well known firm
'13. Is a large firm

48)

Inter-corporate Relations Attributes:


"1. Is a current supplier

2.
"3.
4.
"S.
"6.
7.

Sourcehas beenusedbefore
Is acceptedby our other departments
Is recommendedby our other departments
Is known to our firm
Is affiliated with our firm
Is a customer or our

Service-Related Attributes:
1. Helps in emergency situations
"2 Provides needed service information
"3. Invoices correctly -

"4. Makes salesmenavailable as needed


"S. Offers frequent delivery service
6. Helpful in overcoming our occasional
errors

'7. Maintains consignment stock at vendor


plant

'3.
9.
'10.
'1I.

Suppliesparts lists and operating manuals


Offers better warranties
Maintains technical servicein the field
Maintains repair service

12. Makes available test or demonstration


models

"13. Suppliesspecial reports

'14. Helpful in providing special handling


equipment

IS. Maintains frequent salescalls


'16. Provides information through advertising
'17. Provides information through
promotional activities

48

A more appropriate

method of identifying

is by market research since this


factors relevant

relevant

identifies

criteria

only those

From
buying situation.
(26)
the major
out by H.M. Ltd.

to a particular

the market survey carried


factors affecting

user choice of wire-rope

hoists were

found to be:
I

(a)

i. e. correct

Specification,

capacity,,

height of

and duty rating.

lift
(b)

Price.

(c)

Componentand product standardisation

through

range.
(d)

Delivery.

(e)

Life

(f)

Reliability.

(g)

Ease with which hoist

(h)

Technical and maintenance service

of product.

In general hoists
organisation
influence

are a -l ovalevel

levels required.

buying decision

and the Works Engineer usually

in determining

Because of this
important

can be serviced.

criteria

the make of hoist

has the greatest


to be purchased.

ease of maintenance and reliability


affecting

in an

the purchase decision.

are

49

It is obvious that reliability

and serviceability

be important factors

the make of product

affecting

purchased since hoists

are often located in difficult

(i. e. overhead) and when out of service

access positions
frequently

interfere

prevent or seriously

hoists

inoperative

experience of the effect

have on manufacturing processes.

with the

The author has

purchasers manufacturing operations.


practical

would

The overhead crane

furnaces
from
ingots
the
pre-heat
aluminium
unloading
to the hot rolling

at Alcan Sheet Ltd.,

mill

for several weeks prevented the

Gwent, when inoperative

cold rolling,

slitting,

annealing,

from operating,

operation

At H.M. Ltd. itself

of which could seriously

and Original

form a significant
concerned primarily

competitive

stock could be produced


much inter-

handling is dependent on overhead travelling

failure
cranes,
Distributors

embossing and packing)

since no hot rolled

to supply these processes.

variety

processes (e. g.

of the company's manufacturing

majority

Newport,

part
with

discounts

of options

affect

Equipment Manufacturers
of the hoist
quick

block

deliveries,

production.
(OEM's)

market and are


price,

and a product range with a large

(e. g. dual lifting

speeds, secondary

load brake and overload warning) and mounting positions.

50

(ii)

Determining

Relative

Importance

Several methods are available

for determining

the relative

importance of product and company attributes,


i. e. Regression
(49),
(50)
Analysis
the use of the Pareto Distribution
and
(26).
Market Research
e

determining

The regression technique involves


design attributes
multiple

of competitive

regression

between attribute

the relationship

hence
obtaining
,

sales levels
attributes

values and corresponding

the relative

is examined in detail

thesis.

of this

The use of this

since extensive data on competitive

Starkef(50)

is often difficult

model.

in Section
method is limited

of technical

back-up

to obtain.

determined the relative

design attributes

importance of

products (e. g. delivery

and level

time, product reliability


provided)

determine

in the regression

as coefficients

Regression analysis
3.3.6.4

products and using

to statistically

analysis

values for

by subjectively

importance of product
ranking each attribute

in order of importance and using the Pareto Distribution


to assess their
problems exist

relative

importance.

when using this method:

However, two basic

51

(a)

The initial

ranking of the attributes

decision and therefore


(b)

Assuming that
can be applied

prone to error.

the Pareto

(Figure

Distribution

to attribute

ranking

determined

of attributes

discrepancies

can result.

the most effective

of product

purchasing

a product

applicable

Market

attributes

since

importance

to the development

that
is normally

research

can be elicited

The relative

and the

and indicates

method of determining

importance

purchasers.

importance

by market research

use of the Pareto Distribution

the relative
the reasons for

directly

from

of attributes

of the New EHB is

in Table 7.

OF NEWEHB ATTRIBUTES
RELATIVE IMPORTANCE

TABLE 7

Market Research

Attribute

Pareto

Distribution
Ref. 50)

Reliability

0.39

0.49

Serviceability

0.24

0.35

Price

0.14

0.07

Life

0.10

0.04

Delivery

0.09

0.03

0.04

0.02

1.00

1.00

Service

50)

is not always

Table 7 compares the relative

correct.

listed

is a subjective

Levels

52

Whenusing marketing analysis

to obtain data it

is

necessary to consider the frequency with which market


studies are conducted since the relative
product and company attributes

alter

importance of

with changes in the

market environment.
between Product Attributes

(iii, ) Determining the Relationship


and Sales Levels
importance

When the relative


it

of product

is then necessary to use this


for

the customer preference

information

a particular

is known

attributes

to determine

products'sales

levels.

Models that
decision

view the choice

have
been
problem
established
making
for

buyer preference

product

choose between alternative

such as price,
Generally

process as a multi-attribute

quality

marketing

to predict

brands (i. e. a buyer has to


brands characterised

by attributes

and reliability).

studies

on multi-attribute

decision

making have concentrated on linear models of attribute


(51).
Although complex multiplicative
models
combination
(52)
have been-developed, Huber
reports that these models
have only similar
Effective
Equation
for

linear

accuracy
attitude

levels

to simple

linear

models.

in
form
the
shown
models of

(49)
2 have been developed by Beckwith and Lehmann

predicting

the relative

preferences

of individuals

for

53

various television
for predicting

programmes and Pekelman and Sen(51)

consumer preferences

for alternative

cereal brands and automobiles.

a=d
Wa (Uba -

Pfb =

(2)

a)

a=1
Where:

Pfb

= Estimated overall

Wa

Relative
=

Uba =

Perception

d=

Total

Ia

preference for product b.

importance of the ath attribute.


of product

b on the ath attribute.

number of attributes.

the
Ideal
on
ath attribute.
point
=

The research by Wildt and Bruno


attitude

(53)

models are more applicable

than consumer products.

indicates

that linear

to industrial

This is possibly

products

due to greater

care being exercised during the industrial

buying situation

than when purchasing consumer goods.


Since buyers compare alternative
the perceived
Equation
situation.

risk

inherent

in order

in the purchase

2 must be re-specified
That is

products

for

decision,

the industrial

'a would now represent

to lower

buying

the attribute

54

value for a competitive

product.

The problem begins to

becomecomplex when several alternative

products are being

considered since a value for Ia is difficult

to assess.

Pekelman and Sen(51) observed that most buyers used very


few attributes

when comparing alternative

due to the buyers poor information

This is possibly

decision.

handling

time in which to make the purchase

and limited

ability

products : (Figure 5).

Both factors

have been proposed by Gr$nhaug(43)

to explain the search behaviour of buyers, i. e. Equation 3.


of (PR, TP, AH)

Search is a function

(3)

where:
PR =

Perceived

TP =

Time pressure.

AH = Ability

performance

in information

Figure 5 therefore
multi-attribute

indicates

risk.

handling.
that the complexity

of the

decision making problem could be significantly

reduced by using only the small number of the critical


attributes

in Equation 2.

The use of multi-attribute

decision

models in predicting

buyer preference and hence product sales levels


in two ways:

is limited

55

FIGURE 5

NUMBEROF ATTRIBUTES EXAMINEDPER


BUYERDURINGTHE BUYING DECISION

of

Ref.

0.8

C)

40

0.6

C
0
4-3
So

0.4

0
S136

0.2

1234
Number of Attributes

Examined per Buyer

51

56

(i)

The method requires

extensive analysis of competitors.

According to Rothschild(54)

a disciplined

necessary to assess each major competitor

approach is
(and

interactions

between competitors)

at providing

answers to basic questions such as:

(a)

Who is the competition

which should aim

now and who will

it

be

in the future?
(b)

What are the key competitors'


objectives

(c)

and goals?

How important
competitors

strategies,

is a specific

market to the

and are they committed enough to

continue to invest?
do the competitors

(d)

What unique strengths

have?

(e)

Do they have any weaknesses that make them


vulnerable?

(f)

What changes are likely


future

(g)

in the competitors

strategies?

What are the implications


strategies

of competitors

on the market, the industry

ones own company?

and

57

(ii)

decision models do not take into

Multi-attribute
consideration

the observation

of Cunningham and

White(47) that a strong determinant of a buyer's


decision
patronage
natural

reluctance

extreme difficulty
le

is his past experience and


to change supplier.
in recapturing

Hence the

lost markets.

58

CHAPTER2

ASPECTSOF THE DESIGNPROCESS

59
2.0

Introduction
Product design can be considered essentially

process since the evaluation

design solutions

of alternative

decisions to be made. That is,

a single

as adecision-making

strategy

requires

must be chosen from

It is to be expected therefore that the


(55,56)
stages involved in the design process
represent those of
(55 57)
decision-making models,
a numberof alternatives.

of design problems presents a serious challenge

The recognition

to the designer since the rapid changes occurring


environment both create new problems and alter
of existing

(a)

in the manufacturing
importance

the relative

ones, e. g.

The need to design products on a modular basis,


i. e. the use of standard components and sub-assemblies
between products is becoming increasingly

in

important

to maintain an adequate range of

allowing distributors

product types and sizes with minimum stock costs.

(b)

Shaftel(58)

also
if

necessary

sub-assembly
standardize
all

products

there

considers

modules.

That is,

standard

it

on a few module designs


or applications

by using

against

modular

modules.

costs

is

manufacturing
to

may be worthwhile
that

will

be used in

even though a few more parts


The cost of delivering

few standard

the fixed

design

are economies of scale-in

than needed are installed.


parts

that

extra

balanced
be
modules must

of producing

more types of

60

(c)

There is a growing interest


Terotechnology(59'60)
design function

in the concept of

in which the objectives

are to minimize the total

incurred during the working life


comparedwith the current

of the

costs

of a product as

objectives

of minimizing

supplier manufacturing costs.


(d)

The design function

may be required to develop products

that can be manufactured in existing Group Technology


(61)
cells
or are applicable to modern production control
concepts such as Period Batch Control(62)

and Materials

Requirement Planning(63).
(e)

The growing use of flexible


to influence
bility

automation(64)

design criteria

since it

of use of machine tools

decision-making

solutions

effective
during

the NPD process.

insufficient
effort.
yield

to be obtained

for

However this

and hence allows greater

model (Table
each design

is normally

design resources since it would entail

Hence it
the greatest

is necessary to select
benefits

improves the flexi-

to be considered.

product design variability

The use of a formal

can be expected

8) could enable

problem that
prevented

occurs

due to

considerable

design

those problems that would

from use of a formal decision-making

process.

61

TABLE8

STAGESIN THE DESIGNPROCESS

(i)

Recognition of the design problem objectives.

(ii)

Specification

(iii)

Synthesis of alternative

(iv)

(v)
-

Evaluation

of the objective

of alternative

Deciding the solution

Ref.

(55)

constraints.

design problem solutions.


design

problem solutions.

to adopt.

The obvious method of selecting

design decisions

is to determine
(65)
design objectives.
Starkey

relative importance to the overall


(66)
identify three basic groups (Fundamental, Intermediate
and Sanders
their

in which the contribution

and Minor),

to the total

manufacturing

cost

involved in a design decision or the cost and ease of changing that


decision

is used to determine the relative

It is suggested that the relative


and Minor decisions
proportion

follow

Intermediate.

numbers of Fundamental, Intermediate

the Pareto distribution

number of decisions

of the total

importance of a problem.

are Fundamental and

Hence it would appear that these groups of decisions

could be solved using formal decision models.


or Intermediate

decision

For example, deciding


(Hydraulic

the fundamental design concept of the New EHBhoist

minor design decisions),


out.

However a Fundamental

is often the sum of numerous minor decisions.

or Electro-Mechanical)

could be carried

in that a small

involved the detailed

design (i. e.

of each concept in order that comparisons

62

During the synthesis stage of the decision-making


engineering designer is called on to use his creative

process the
ability

experience in proposing design solutions.

Ostrofsky(56)

that the higher the number of alternative

design solutions

the greater the possibility

and

suggests
proposed

of choosing the optimum solution.

During

the design of the New EHBrange it was observed that designers


difficulty

encountered little
solutions

but it was the evaluation

that restricted

(i)

in proposing alternative

design

and comparison of design solutions

the number that could be considered since:

The amount of data required

to evaluate designs

(e. g. costs, producibility,

reliability

and safety)

is large.
(ii)

Several functional

departments are often required to

hence
designs
co-ordination
evaluate

of input data is

difficult.
(iii)

The resources

available

departments

are limited,

are normally

sufficient

the various

functional

i. e. departmental

resources

within

to cope with

company data requirements


additional

The criteria
into

the following

amounts required

and cannot
during

used to evaluate alternative


categories:

the day-to-day
cope with

the

the NPD process.

designs can be grouped

63

Specifications,

c>>

i. e. design characteristics

the range of market specifications

that determine
Hence a

covered.

is defined as that characteristic

market specification

of a product required by a potential

buyer before the

product can be considered for comparison with competing


products.
i

(ii)

Attributes,
reliability,

i. e. those product characteristics


quality

and style

such as

that are used by a buyer

to compare competing products.


(iii)

Producibility,

i. e. the ease with which a product can be

manufactured and raw materials


(iv)

purchased.

Cost - of prime importance in appraising

designs since it

is used to evaluate other design criteria,


to design for,

of market specifications
quality

reliability,

ease of producibility

a major determinant

the level

of

and safety to design for and the


in terms of labour costs.

the manufacturing

addition

e. g. the range

In

cost of a product is normally

in the purchasing decision.

Several areas of H.M. Ltd. 's new product design process have been
identified

as limiting

can be economically
and/or preventing

A major factor
estimating

the number of alternative

design solutions

that

compared and/or leading to wasted design resources

an optimum product range being developed.

that causes problems is the inability

system at H.M. Ltd. to quickly

of the current

generate the large amounts of

64

cost data required during the design process.


examines in detail

the estimation

need for an estimating

In consequence Chapter 3

of product costs at H.M. Ltd.

The

system that can be used by engineering

designers is proposed and such a system is developed in Chapter 4.

Other areas of the design process called

into question are:

(i)

the "make" or "buy" decision,

(ii)

value engineering,

(iii)

the "assortment"

(iv)

planning

(v)

scheduling

(vi)

the changes that occur in product designs.

problem,

and control

of the design

of the design

process,

process,
and

These areas of the design process are now examined in greater


detail.

2.1

The Make or Buy Decision


The decision

from an outside

to manufacture a component in-house or purchase

supplier

is a frequent

problem facing the designer.

In theory each component of a product design must be assessed in this


manner.

However the decision

to make or buy is often obvious.

For example

65

either the companydoes not possess the equipment to manufacture


in-house, or the component is a standard, mass produced part,
developmentobjectives

are to design for either

or the

in-house manufacture

or purchasing.

For components in which the decision to make or buy is not


obvious the factors

to be taken into consideration

and Haas and Wotruba(68).

by Claret(67)

have been identified

These factors

are listed

in

Table 9.
MAKEOR BUYDECISIONCRITERIA

TABLE9
(i)

The need to maintain economic utilization

of capital

equipment.
(ii)

The need to absorb high overhead costs.

(iii)

Assurance of supply.

(iv)

If the component design is a trade secret.

(v)

Transportation

(vi)

Companyreputation
rather

costs and ease of transportation.


gained through manufacturing

than merely factoring

them.

(vii)

The need to diversify

(viii)

The need to smooth seasonal variations

(ix)

Lack of sufficient

(x)

The familiarity

parts

operations.

capital

in order patterns.

to finance required

investment.

involved.
technology
the
with
experience
or

66

Haas and Wotruba examined the make or buy decision


to marketing products which the potential

in relation

customer could also make,

i. e. the prospective customer is at the same time a potential

competitor.

They concluded that the decision to make or buy was dependent on the
difference

between the cost to the prospective

customer should they

manufacture the required product in-house and the prices of that


11
product from possible suppliers.

The manufacturing costs highlighted


both the direct
direct

material

indirect

and indirect
costs,

by Haas and Wotruba included


components (i. e.

costs of manufacturing

factory

labour costs and factory

direct

costs,

supplies

However Claret(67)

overhead costs).

directly
those
that
costs
only
considers
(i.
e. variable
components manufactured

related

labour costs,

to the volume of

manufacturing

should be

costs)

decision.
buy
in
the
or
make
used

is more realistic

The approach adopted by Claret

Haas and Wotruba since normally only variable


the decision
are usually

Claret

to make or buy.
incurred

considers

cannot

considered

fixed

decision.

However during

costs

and should

overheads

of the type of decision.

in certain

be used elsewhere)

resources

costs are affected'by

Fixed costs such as factory

irrespective

that

than that of

circumstances

that

direct

not be included

the development

labour

(where labour
costs

can be

in the make or buy

of a new product

range

be
to
should
produce
not
a component
the labour resources required
considered fixed

since an objective

is
to
department
design
the
of

67

minimize a component's manufacturing costs which include a labour cost


element.

be included in the make or buy

Labour costs must therefore

decision during new product development.

Levy and Sarnet(69) considered that the make or buy decision should
be based on the differential

involved

risk

of the cash flows (e. g. investment outlay


sales quantities

in the various components


for capital

and the purchase and production


However this

engendered by the alternatives.


considered if one alternative

necessitates

quantities

costs per component)

approach need only be

a higher level

investment, in which case the basic Uncertainty


and estimated costs makes it

equipment, forecast

of capital

surrounding forecast

apparent that the alternative

that requires the more expensive equipment constitutes

the greatest

risk.

Normally

component designs

is purchased

i. e. the raw material


in-house.

in-house.
designers
Quantity

total

a designer

must decide
casting

in what form to

and part

operations

moulded

to carry

approach to the make or buy decision


by using

to make such choices


(70)
to develop
theory (EOQ)

classical

a model for

out

allows

Economic Order
calculating

the

costs per unit of a component when purchased,

both.
in-house
of
or
a
combination
manufactured
control

of both make and buy,

from which a component is manufactured

and which manufacturing

The present

variable

elements

(e.
bar,
g.
or components

components)

plastic

therefore

Frequently

buy raw material

contain

the
variability
allow
also
models

The use of stock

of component cost with

into
batch
be
taken
consideration.
to
sizes
purchasing and manufacturing

68

The total

cost of a component is calculated(71)

material

using

Equation 4.

Cm =k+k.

Ic

+ Cb. Qb-'

(4)

Where:

Cm = Total material

Purchase cost per unit.

k=
Ic

costs per unit.

interest
holding
Stock
=

rate per unit per year.

Cb = Purchase cost per batch.


Qb = Purchased batch size.

The variable
by the following

Cp =C +

costs of in-house manufacturing operations


expression(71) :

Cs. Qp-1

(5)

Where:

Cp = Total variable

in-house labour costs per unit.

Cy, = Total

labour cost per unit.

Cs = Total

setting

Qp =

is given

Production

up cost per batch.

batch size.

69

Capital investment costs such as special


fixed and non-recurring

tooling

costs,

although

over the short term, are recovered over a

specific time period the length of which is normally a management


decision (at H.M. Ltd. this

period is normally three years) and is

therefore influenced by managementsattitude


the development project

and financial

Equation 6

constraints

involved

CT.2.718-r.

(6)

p=r
Ncp
p=1

Where:

Ct

investment
Tooling
cost per component.
or
=

CT

Total
tooling
=

or investment costs.

Ncp = The number of components manufactured for the


first

p years of the product life.

Recovery period

r=

for

capital

Company's discount'rate

outlay

for capital

in

on borrowed capital.

investment costs is given by

for recovering capital


(72):
Ostwald and Toole

Ct

to the risk

(Years).

investment.

70

The expression 2"718-r.

is used to take into consideration

tooling or investment costs are incurred

that

at the beginning of the payback

be discounted to give a more balanced

period and must therefore

perspective to the investment costs.

The total

component is therefore
Equations 4,5

cost involved

variable

in buying and/or making a

the sum of the individual

costs obtained from

and 6, i. e.

(7)

Cv = Cm + Cp + Ct

Where:

Cv = Total variable

According
since
for

there

to Burbidge(73)

TABLE 10

Qb and Qp are not always identical

ways in which parts

are different

convenience

component costs.

in production,

Table

are batched

together

10.

TYPES OF BATCHESPRESENTIN A MANUFACTURING


ORGANISATION

Ref.

(i)

Order Quantity

(OQ) - quantity

for production

or purchase by a manufacturing

purchase order.

(73)

of a component authorised
order or

71

TABLE10 (Contd)
Run Quantity (RQ) - quantity

of a component run-off

at

a machine or other work centre before changing to some


other component.
(iii)

by a supplier

(TQ) - quantity

delivered

at one time

to a customer, or transferred

in a factory

Transfer Quantity

as a batch between two machines for succeeding operations.


(iv)

Set-Up Quantity
necessarily

all

(SQ) - quantity

of components (not

the same) which are produced on a

machine before changing the tooling

and machine set-up.

Classical
Qb and Qp must therefore be determined separately.
(71)
be
to
theory
used
can
calculate the value of Qb
stock control
that minimizes Cm. It must be assumed that the values of Qp for each
operation

identical
to
a
component
are
manufacture
required

represents efficient

In addition
established

control

of a company's manufacturing

since this
operations.

Qp
be
Cp
for
that
then
minimizes
can
value
a

by the classical

stock control

approach, i. e. differentiating

Cp with respect to Qp,with the minimum value of Cp occurring

dC=0
d (Qp)

when:

72

Equation 7 is therefore

a useful tool for determining the costs

involved in the make or buy situation

and can be used by engineering

designers for comparing design alternatives.


design alternatives

For example, various

were proposed for the New EHBRope Guide, the

assembly which prevents adjacent rope strands from rubbing together


the rope is being wound around the drum tube and hence prevents

whilst

110,

rope wear.

IA

The alternative

designs considered were:

purchased phosphor bronze casting


existing

hoist

(currently

block range) with tooling

used on

costs included

in the component purchase cost.


II

A purchased, fully

component (machined

machined plastic

by the supplier).
III

A purchased
costs

IV

plastic

moulding

with

separate

and no subsequent machining

required.

tooling

A purchased plastic

moulding with separate tooling

costs that required

finish

machining operations

at

H.M. Ltd.
VA

purchased plastic
and finished

moulding with separate tooling

costs

machined by the supplier.

(determined
using Equation 7) of the
costs
The comparative
alternative
(the current
life

designs
in
11
for
Table
are
shown
r=3
rope guide
H.M. Ltd. payback period)

tools).
the
moulding
of

and r=

years

10 years (the nominal

73

TABLE 11

MANUFACTURING
COSTSOF ALTERNATIVE ROPEGUIDE
ASSEMBLYDESIGNS (3.2 TONNENEWEHB)

Manufacturing

Costs/Unit

('s)

EOQMethod

Design
Alternatives

10 years

r=

years

r=3

PBC System

r=3

years

10 years

r=

69.24

69.24

69.24

69.24

11

71.25

71.25

71.25

71.25

III

31.48

10.61

31.95

11.08

IV

32.31

18.33

39.55

25.51

30.74

16.70

37.80

23.76

The influence

is seen,

alternatives
expensive.
tooling

tooling
cost

life

be obtained

at a period

Burbidge(62)

approach(59)

that

a compromise

III

but recovering

approach

tends to inflate

alternative

in the PLC when competitive


rate

of design

and recovering

indicates

Of course,

However this

the growth

Hence choosing

become a marketing

of a product's

the appropriate

design

prices

sales

the
are an

in the

alternative

could

decision.

Control
Batch
Period
that
the
of
use
suggests

reduces manufacturing
organises

by using

costs

V is the least

alternative

(r = 10 years)

expensive.

over 3 years.

to increasing

place.

nominal

is the least

of a product

advantage
market

over their

III

costs

years

the Terotechnology

could

solution

i. e. when r=3

However using

costs

alternative

on the comparative

of payback period

costs since this

and plans manufacturing

production

operations

control

by controlling

(PBC)

technique
the flow

74

That is,

of stock.

the batch quantities

are determined from production

the level

sets

(on a series

actual

batch quantity

"cycle

time".

product

assigned

is dependent

is divided

to each"cycle"must

demand and is allowed

to fluctuate

into

in

orders

Hence all

'i. e. OQ = RQ = TQ = SQ (Table

termed "cycles".

equal time periods

releases

of common due dates).

selected

The year

technique(62)

control

are identical,

quantities

as opposed to the EOQmethod

of stocks.

The PBC production


product

due dates

programmes based on sales orders

received or short term sales forecasts


which controls

ordered and their

batch

10).

The

on the manufacturing

a number of approximately

The production

volume per

be in balance with

the market

with demand. The "cycle

time"

throughput
time
for
less
than
the
be
the
any
of
components'
cannot
included

in the PBC product sets and must not increase the proportion

of machine setting
required

time so that capacity

is reduced below the level

to meet demand.

Assuming a PBCsystem was operating


compare the alternative

Equation 7 could be used to

rope guide assemblies,

Qb = Qp = Ns. Te. Nw

Table 11, i. e.

($)

Where:

Ns =
TI

Nw =

Sales demand.
Longest

individual

Number of working

throughput

time (weeks).

(weeks).
in
the
year
weeks

75

For the 3.2 tonne New EHBunit Ns = 380/year, TZ =6 weeks and


Nw = 48 weeks and Qp = 47.5.

From Table 11 it

can be seen that the effect

sizes through the introduction


cost effectiveness

of lowering batch

of a PBCsystem tends to increase the


tooling

of components requiring

or investment

costs to the disadvantage of machined components. In addition


several difficulties

are encountered by the design function

when

developing products for a PBCsystem:

(i)

It is necessary to determine the "cycle time" before make


or buy decisions

and comparisons of alternative

concepts can be made. This requires


longest individual

design

a knowledge of the

component throughput time within

product.
(ii)

schedules

in order

Hence scheduling
constraints
consider

(iii)

detailed

A PBC system requires

planning

to reduce set-up

flexibility
since

arise,
the overall

is

Hence the design function


fit
to
components

times.

reduced and new design

manufacturing

to reduce manufacturing

and operation

the design

PBC systems use Group Technology

of manufacturing

function

must now

sequence of a product.

(62)

manufacturing

cells

cycle times and stock levels.


could need to consider designing

into existing

cells

in order to avoid

76

the cost of redesigning


applicable

particularly
if

a cell

within

structure

the cell

in multi-product

This is

organisations

is in existence for other products

the company's product mix.

Engineering design changes introduced to improve materials

(iv)

or manufacturing

methods could form causes of inaccuracies

in computerised inventory

systems such as PBC


(63)
Requirement Planning (MRP)_

and Materials

2.2

structure.

control

Value Engineering
During the New EHBdevelopment period a Value Engineering (VE)

Committee was formed to examine methods of reducing manufacturing costs.


In accordance with value engineering methodology(74) the members of
the
from
functional
of
each
main
chosen
were

the committee
involved

engineering,

purchasing

method of reducing

and estimating).

product

H. M. Ltd. achieved limited

(i)

development

in the new product

manufacturing

process

(i. e. design,

Although
costs

departments
production

VE is a recognised

the committee

at

success since:

VE meetings were not held regularly

therefore

the number

limited.
be
that
could
examined
was
of components

The

in
VE
order
should
meet
frequency with which a
committee
be
can
represented
product
range
examine
a
to adequately

as:

Vf

Ta.
Tp. Nw.TA

9)

77

Where:

Vf

(per
VE
Frequency
of
committee
week).
meetings
=

Ta = Average VE time per component (hours).

Ne = Number of components examined.


Tp = Product development period (years).
Nw = Number of working weeks per year.
TA = Average length of VE committee meetings (hours).

From examination of the current value e.ngineering


Ta =2 hours,. Ne = 400, Tp = 2.5 years,
TA =3 hours.

work

Nw = 46 and

Therefore from Equation 9 the frequency

with which VE committee meetings must be held is:

Vf

400

2.3 per week

=2x
46 x 2.5 x3

holding over two 3 hour VE meetings per week would

Clearly

be difficult
sufficient

since committee members could not have


time available

to
found
be
either
must

to attend.

Therefore methods

reduce the number of components

be
VE
improve
the
can
carried
work
at
which
rate
examined or
out.

Since VE is a creative

than an analytical

technique)

VE work could be carried

cost reduction

tool

(rather

improving the rate at which.

inherently
be
out would

difficult

78

whereas obviously

concentrating VEtime on high cost

components achieves the initial

option.

However to increase the rate at which VE work can be carried


(75)
out a method employed by the author during VE exercises
was found to produce advantages. The technique involved
applying an individual
components within

method to all

cost reduction

a product design.

listed
methods
reduction

Each of the cost

in Table 12 were applied to a

product design in this manner.

In this way the Value

Engineer takes advantage of a "learning


That is,

the

curve" effect.

as the number of components examined increases,

the speed with which succeeding components can be examined


and potential

cost savings identified

To reduce VE time further


this

in
conjunction
method

high cost

components,

sufficient
be ignored

during

later

stage.

is also

with

possible

the technique

i. e. the order

determined
is
are examined
time

it

also increases.

by their

is not available
the NPD process

to employ
of examining'

in which components
relative

low cost

cost.

If

components could

and value

analysed

at a

79

TABLE12

COSTREDUCTION
METHODS

(a)

Reducing the volume of material

(b)

Reducing the number of bought out componentsy

used per component.

i. e. improving overhead recovery.

(c)

Reducing-the

(d)

Converting material

(e)

Examining the possibility

(f)

Examine components for possibility

of using plastics.

(9)

Examine components for possibility

of using standard

labour cost per component.


costs to labour costs.
of using cheaper materials.

replacement parts.
(h)

Reducing the number of components required.

(i)

Reducing. the number of components required


the possibility

of components carrying

by examining

out the functions

of others.
(j)

Standardization

of chamfers, radii,

bolts,

nuts, bearings,

screws and threads.


(k)

Examining the possibility


of components.

of reducing dimensional tolerances

80

(ii)

During meetings formalised


followed.

VE procedures were not

To achieve results

programmemust be followed
(Information
stages

a carefully

basic
of seven

consisting

gathering,

planned

Speculation

phase,

Evaluation phase, Programmeplanning stage, Programme


execution phase, Recommendation stage and Implementation

stage).

Again a reduction in the number of components

examined would effectively

time for

allow additional

the formal VE procedure to be adopted.

(iii)

WhenVE is carried
is necessary
build

and test

out during the development process it

to recommend design
in order

work being necessary.

changes prior

to prevent

further

to prototype

development

However because of the inherently

VE
(compounded
by
work
can
proceed
which
at
speed
slow
the lack of regular
test

VE meetings) the prototype

took
assemblies
place before the VE
of
many
stages

designs
the
involved.
examine
could
committee
the need for design changes occurring
test

build and

stage value engineering

be completed prior
prototype

and test

preliminary

phases.

the prototype

of a component design should

to the release of the design for

manufacture.

an essential

after

To prevent

VE should therefore
step prior

be considered

to the prototype

build

81

2.3

The Assortment Problem


Whendeveloping a product to cover a range of market specifications

(e. g. size,

decisions
fundamental
to be made are,
and
power),
capacity

the number of standard sizes to manufacture and the exact specification


With respect to the New EHBdevelopment project

for each standard size.

these problems centred around deciding:

the number of basic frame capacities

(i)

standards,

and

the Safe Working Load capacities

(ii)

to develop as

of each standard size.

into an important class of problems


(76)
defined as "the assortment problem" which according to Swanson
Decisions of this

type fall

basic
two
sub-classes:
of
are composed

(i)

The "set-up

(ii)

The "job-shop

and inventory"assortment

problem.

assortment"problem.

In general the assortment problem looks at the determination


if
of what,
that

be
standard
sizes
might
manufactured such
of
system
any,

large
through
number
manufacturing
a
obtained
scale
economies of

than
offset
more
sizes
few
standard
of a
(material
is greater

and manufacturing

the lost

extra capacity

time) in supplying products whose size

than the customer requires.

With the "set-up and inventory"

have
to
is
the
a
assumed
of
manufacture
method
assortment problem

82

certain set-up cost and optimal production


it.

Therefore when different

standard the inventory,

run length associated with

size products are combined to form a

set-up and manufacturing costs must all

be

considered in deciding whether or not to combine the various sizes


During the "job-shop"

into a standard.

is assumedto be performed exactly


i
Hence there

little

is relatively

assortment problem manufacturing

to the specification
inventory

and the main consideration

costs for various methods of manufacture.

is the differing

The assortment problem involves many subjective


influence

of the solution

the optimality

to formulate

difficult
are often

problem must also reconcile


departments involved

Important

criteria

which

and hence assortment problems


The solution

objectively.

the competing objectives

to the

of the functional

in new product development.

to consider

criteria

are listed

in Table 13.

PROBLEM
FACTORSAFFECTING THE ASSORTMENT

TABLE 13

A,

of the customer.

MARKETINGFACTORS

(i)

Customer acceptance of sizes other than those historically


available.

(ii)

larger
sizes
would
Determining whether customers
accept
than their

(iii)

Competitive
price

requirements.
sizes

product

comparisons

if

'competitors.
the
than

which may prove to provide

standards

were chosen that

poor

were larger

83

TABLE13 (Contd)

(iv)

The actual demand pattern,

i. e. the preferential

sizes favoured by the market.


(v)

The range of sizes provided for the market.

(vi)

The profitability

i. e. product size versus

pattern,

profit.
lead times required.

(vii)

Manufacturing

(viii)

The need to market a full

range of product sizes,

e. g. at H.M. Ltd. product sales are strongly


dependent on customers standardizing
requirements in order to simplify
maintenance procedures.

hoist

block

stock holding and

Therefore customers tend to

buy from manufacturers who can supply their

specific

requirements.
B.

FACTORS
MANUFACTURING

>
c>

flexibility,

facilities

handle the level

design

associated

variation

of standard

(ii)

i. e. can the manufacturing

Production

Stock policy,

of product
with

and component

increasing

the number

sizes.

i. e. stock costs can be expected to rise

increases.
the
of
standard
sizes
number
manufactured
as

84

TABLE13 (Contd)

(iii)

Production control

methodology, e. g. batch scheduling,

PBC, MRP
(iv)

Machine utilisation.

(v)

Machine set-up and operation


Machine type profile,

costs.

i. e. the number and type of

automated machines.
(vii)

Type of manufacturing
volume production,
production

(viii)

C.

system, i. e. job shop, batch, high

group technology layout or automated

lines.

Cost of standardization.

DESIGN FACTORS

(i)

The development resources available

influence

the number

of standard sizes that can be designed.


1 i)

"Bought out" to "made in" component levels

in the

product design.

(iii)

The degree of standardization

(iv)

The suitability

(v)

Level of material,

between product sizes.

of the product to modularisation.


labour and overhead costs.

85

TABLE13 (Contd)

D.

CORPORATE
FACTORS
(i)

New product range development time scale.

(ii)

Corporate policy.
.le

(iii)

Long term objectives


Labour policy,

for the product range.

i. e. degree of flexibility

in labour

allocation.
(v)

Working capital
work-in-progress

constraints
levels

tend to limit

therefore

stock and

reducing the number

of standard sizes that could be manufactured.

Since many of these factors

are difficult

to quantitatively

the
around
centres
estimation
often
the assortment problem
quantifiable
criteria.

(i)

benefits

of the

in order to weight these against the subjective

The problem definition

then becomes either:

determine the minimum number of standard sizes such


that the loss from manufacturing

product sizes larger

than customer requirements is acceptable,

(ii)

or

determine the number of standard sizes such that the


disadvantages of manufacturing one extra standard size
are greater

define

than the benefits

incurred.

86

13) must be examined for its

Each of the above constraints(Table


tive

importance to the assortment problem.

scale is an important constraint

For example development time

in the assortment problem for the New

is necessary to limit

EHBrange since it

rela-

the development period. in

order that a new product range can be launched before a serious


loss in sales levels

2.3.1

Solution

of the existing

Ropemaster range.

Techniques

There are several established

techniques for solving

the

assortment problem; complete enumeration, random enumeration,


dynamic programming and the minaddition

technique.

attempts to determine the number of standard units

Complete enumeration for example involves


all-possible

the analysis

from
each, with the solution
obtained
savings
being adopted.

the number of possible

yielding

However the number of possible

increases
to
exponentially
consider
combinations

with a rise

in

sizes and the number of standard sizes.

be
's
Ltd.
H.
M.
to
used
enumeration
cannot
solve
complete

there
to
eighteeen
select
are
since
sizes
problem
assortment
(and hence 6.4 x 1015 possible
solution

of

combinations of standard sizes to determine the

the maximumbenefits

Therefore

and their

sizes that minimizes manufacturing costs.

respective

relative

Each method

would require

combinations)

an excessively

and the problem

long time to evaluate.

from

87

Randomenumeration of a limited
possible but it

number of combinations is

is difficult

to determine how close to optimal

is.

combinations based on

Selecting

any trial

solution

intuition

and experience is also liable

Several investigators
Programming (DP) algorithms

(76'77'78)

to the same uncertainty.

have used Dynamic

to solve the assortment problem,

i. e. the problem is represented as a sequential


which is solvable

in stages.

The first

(the
largest
size
one
standard

size'requiredto

problem

stage is to select
This decision

size).

normally a marketing or manufacturing


largest

decision

constraint,

is

e. g. the

cover the intended market segment or

the maximumsize that can be manufactured due to component


size limitations

on production

equipment.

The subsequent stages involve

selecting

decreasing
in
size.
of
order
standard

each individual

The underlying

is
known
"principle
the
techniqe
DP
the
as
argument of
which states that an optimal policy

optimality",
property

that whatever the initial


decision,

available)

and initial

constitute

an optimal policy

from the first

decision.

state

philosophic
of

has the

(number of sizes

the remaining decisions

must

with regard to the state resulting

88

Jackson and Zerbe

(77)

initially

solved the assortment

problem using the DP technique but according to Swanson(76) they


specified

(i)

the problem illogically

The solution

required

since:

the pre-determination

number of standard sizes,

i. e. this

of the

is a decision

variable
(ii)

and cannot therefore

The solution

determined which particular

should become standards whilst


sizes as custom designed.
of determining

objective

be pre-specified.

standards is basically

sizes

leaving the other

However, the overall


the optimum number of

to reduce the number of

sizes produced by manufacturing only a limited


number of standard sizes.

Taking these factors

decision

initial
the
necessary at
single

sizes,

feasible

item and working

At the second decision stage all

double standards were examined in the same manner.

It can be seen that extensive


required

stage to examine all

largest
beginning
the
with
standards

down to the smallest.


feasible

into account Swanson(76) found it

when considering
i. e. 5,6,7

...

amounts of computation are

even small numbers of standard

89

can be solved for simple problems using

The DP algorithm

technique developed by Swanson but for more complex

a graphical

decision

problems the sequential

process method is required.

of the problems increase the DP technique

As the complexity

becomes cumbersomeand a computer is necessary to solve

rapidly

the problem in a reasonable length of time. For a problem the


lo, H.M. Ltd. 's
approximately 11K core storage is necessary
size of
is
time
the
approximately
computing
and

A major disadvantage
problems

assortment

by Martin

highlighted

solution

and piff(78)
lift

problem

sales prices

trucks

indicated
management

The DP

such as manufacturing

manufacturing

The subjective

costs

and

truck

estimate of

that only 5 standard sizes should be


in results

The large variation

that the subjective


solution

to determine

22
that
standard fork lift
and recommended

be
manufactured.
should
sizes

produced.

non-quantifiable

to manufacture.

factors

possible,

to solve

This problem has been

when attempting

considered

range of sizes

quantities,

of including

making process.

the optimum number of fork


to this

the DP technique

is the'difficulty

in the decision

criteria

of using

20 minutes.

listed

criteria,

indicated

the effect

in Table 13, have on the

to the assortment problem.

Considering

algorithms,
can be solved

that

subjective

to the assortment

optimum solution
technique(79)

the effect

has certain
since

advantages

an assortment

in a reasonable

criteria

have on the

problem the Minaddition


over the use of DP

's
Ltd.
H.
M.
magnitude
of
problem
time
without
of
amount

the aid of

90

Hence the cost of developing computer software

a computer.

when employing DP algorithms

The Minaddition

is avoided.

technique can be used to solve the

assortment problem assuming that any required size not a


standard can be provided from the next standard size up.
This constraint

is applicable

in excess material

development and, results


being required.

The Minaddition

the particular

select

excess costs.

to the present New EHB


and labour costs

technique is then used to

set of standard sizes that minimizes

When profit

margins vary with product size

then the problem can be respecified

to minimize total

profit

loss.

Using the Minaddition


into two distinct

resolves

(i)

technique the assortment problem


sections:

Use the Minaddition


cost or profit

technique to determine the

loss from manufacturing various

numbers of standard sizes.


(ii)

Determine the extra costs incurred

from increasing

the number of standard sizes.

The optimum number of standard sizes to manufacture therefore


is that

from
the
adding one extra
savings
obtained
number where

is
considered unjustified
standard
costs and manufacturing
manufacturing

efficiency,

when compared with the extra

problems involved,

e. g. reduction

higher work-in-progress

levels.

in

91

2.3.2

Technique

The Minaddition
Wolfson

Minaddition

(79)

published a detailed

description

included
and
a flow diagram to aid computermethod

when complex problems need to be solved.

isation

of. the

The technique

New
development
determining
during
EHB
for
the
be
used
can
optiwum number and lifting

capacities

the

of standard frame sizes

to manufacture.

Initially

the data to be obtained includes the range of

frame sizes that are being considered,


"cost"
the
frame
size
and
per
The type of "cost"

per unit,
material

costs per unit,

margins are similar

per unit,

circumstances

manufacturing

labour cost per unit

During the present analysis

cost per unit.

affect

Table 14.

determined depends on the specific

costs (i. e. material

therefore

sales quantities

of each frame size,

be
profit
may
and
of each problem
direct

forecast

costs

or
the prime

and labour costs) are used since profit


throughout the product range and would not

the outcome of the Minaddition

solution.

The maximumand minimum frame sizes considered in Table 14


knowledge of the size range used within
from
determined
a
were
(26)
the total market.

The Minaddition

technique considers the range of possible

in
sections
sizes
standard
that bound each section

and temporarily

(termed "interval")

regards the sizes


as standards.

The

92

TABLE14

DATAREQUIRED
FORTHE MINADDITIONTECHNIQUE
Ref.

Sales Forecast
(Units per Year)

Prime Costs
("s)

0.16

21

241

0.20

10

252

0.25

89

266

0.32

10

284

0.40

24

306

0.50

383

333

0.63

74

367

0.80

170

449

1.00

184

466

1.25

142

533

1.60

192

627

2.00

94

734

2.50

289

868

3.20

79

1016

4.00

131

1270

5.00

74

1515

6.30

24

1886

8.00

2368

Hoist Unit
Frame Size (tonnes)

Of

(26)

93

prime costs for each interval

total

stages, i. e. the cost that results


the interval

are then calculated


when all

are provided from the larger

The prime costs in any interval

sizes.

in

other sizes in

of the two bounding

would therefore

remain

the same no matter which sizes greater or smaller than the


interval

bounding frame sizes were varied.

Costs are denoted

by:

Ce = AW (min'

max

(10)

Where:

Prime costs.

Ce,A

Sm.

lower
bounding
The
frame
size.
=

Smax = The upper bounding frame size.


w=

The number of standard frame sizes in the


interval.

The objective

of the Minaddition

technique is therefore

to minimize Ce for successive values of w.

Minimum prime costs have been determined for successive


listed
are
values of w and

(with the corresponding optimum

15.
in
Table
to
frame
manufacture)
sizes
standard

94

RANGEDATA
PRODUCT
OPTIMUM

TABLE15

v
a

.r..
.0

0
rn

4J

x
aI
a
cm
c

f0
N
+)
N

E
0

(L)
N
. r..
N
i
t0
C
du

>

N
U
v

d
CM
ct

C)
0
CD,
r--

01
-d"

co
M
N

ONN
Ol

O1
rLO

r-00
N

r0T1
--

to

q; r

C11
N

1,-
00

X
W
v

4-

U
Vf

N
O
C..)
C)
E

Cl
0
0
r

'x

ct
M

O
rO

fM

C)
C
CD

VI
C)
N

cw

9ww

O
N

C;

O
C9.0w

Lf)

LA.

LA

LI)

N"

N.

LAw
Cp

C)
E

.N-

CO
C9.

LL.

G )I
.
C
O

I
e
.,..
4-)
0

C)

06

Lj)

C.

CQ

C
w
LA
NN

C4

ww

C;
Ll)

"N.

LC)

w
Lj)

e-

Ow

Ow

r-

r-

p-.

LAw

L1')

l0

...
CV

Lj)

ww
CD

LA

r-

r-

r-

06C.

CV)

Ct

LO

..

,a
b
c
N
4O

r-

m
E

95

Since the Minaddition


costs from manufacturing
cost saving resulting

technique determines the prime

a fixed

number of standard sizes the

from adding one extra standard size can

be determined using Equation 11.

(1 1)

Cep - Cep+l

Csv =
loe

Where:

Csv

Cost
saving obtained from adding one
=
extra standard size.

Cep

Cep+1 =

Prime cost for p standard sizes.


Prime cost for p+1 standard sizes.

Having determined the cost savings it

is necessary to

in
increases
the
them
manufacturing
with
compare
are incurred

through manufacturing additional

These cost increases arise primarily

(i)

Increasing

costs that

standard sizes.

since:

the number of standard sizes within

product range effectively

lowers the average

batch
and
sizes,
purchase
manufacturing
incurring
(ii)

additional

hence

purchase and setting-up

The increased number of batches that require


processing effectively
utilization

level.

lowers the machine

costs.

96

(iii)

Increased costs are incurred

for special

(e. g. rope guide moulding tools)

tooling

and jigs

and

fixtures.
(iv)

The efficiency

of production

planning and control

systems could be lowered hence labour costs could


be increased,

i. e. extra idle time is incurred.

(v)

Work-in-progress

(vi)

The reduction

and stock levels

increase.

in labour costs and scrap levels

through the "learning

curve" effect

(see Section

is decreased.

3.1.2)

In order to determine the optimum number of standard sizes


to manufacture it

is necessary to compare the cost savings with

the added costs involved

in increasing

the number of standard

The optimum number occurs when Csv < Ca where Ca represents

sizes.

the additional

It

is an advantage to use the Minaddition

determining
estimation

costs incurred.

technique in

the optimum number of standard sizes since accurate


of Ca is not required.

Figure 6 illustrates

incurred
cost
savings
which
are
at
rapid rate
indicates
and
sizes
standard
of the additional

therefore

costs incurred

optimum number of standard sizes.


incurred
costs
extra

the

when adding extra

that only rough estimates

are needed to determine the


H.M. Ltd. estimated that the

in adding an additional

standard size lay

97

FIGURE 6

OPTIMUMNUMBEROF STANDARDHOIST SIZES


TO DEVELOP

Cost Savings

(Csv)

Ln

0
X

V1
C'e
C
"r
>

Optimum range of
Si
standard sizes to
Vli
manufacture

y. i
N
O
U

0.5

L- ----------------

Estimate of
-r
increase in costs
from_adding onp-____
extra standard
I
I

II
345678

Number of

Standard

Sizes

98

between 150,000 and 1100,000.

Hence Figure 6 indicates

that

the optimum number of standard sizes to develop would be five


number chosen would depend on a company's

The actual

or six.

That is,

long term objectives.

choosing to manufacture 6

standard sizes would aid the company in lowering material


labour costs but would require

the organisation

and

to improve or

A,

level
high
a
maintain
actual

to develop five

policy,

because the limited


constraint

H.M. Ltd. 's

efficiency.

standard sizes,

was adopted

development time available

was the overriding

on the number of standards that could be designed.

The basic
1.6 tonnes,
costs

of manufacturing

frame sizes

3.2 tonnes,
from this

arising

optimum policy

for

five

2.5 tonnes,

5.0 tonnes

of 191,000

(Table

15).

chosen by H. M. Ltd.

5.0 tonnes
policy

and 8.0 tonnes.

is equal

standard

are 0.8 tonnes,

sizes

and 8.0 tonnes)

The prime

to 238,600.
(0.5

tonnes,

gives

rise

Hence an additional

The
1.25 tonnes,

to prime costs

saving

of

238,600 - 191,000 = 47,400 is possible. by changing. to the


optimum policy.

However the range of standard sizes adopted

by H. M. Ltd. was chosen primarily

because these lifting

those required

capa-

by the market for crane

cities

are traditionally

duty.

Hence these sizes are necessary since the EHB is a major

's
Universal
Ltd.
H.
M.
in
sub-assembly
Minaddition

analysis

crane range.

has shown the additional

of adopting a non-optimum policy


judge the value of that policy.

However the

cost to H.M. Ltd.

from which managementcan

99

2.4

Design Planning and Control


Since the design of a new product range is essentially

be
design
time
the
must
allocated
process,

using subjective

i. e. the design phase is planned by apportioning

a conceptual
judgement,
design time

available

that
the
makeup the product.
sub-assemblies
major
amongst

It is

of the engineering designer to ensure that

then the responsibility

designed
is
within
sub-assembly
each
in this

design time allocated

the time allowed.

Normally the

manner is not subsequently distributed

components that form each sub-assembly.

between the individual


there is no effective

control

target
a
since
component

over the allocation

Hence

of time to design a

design time for each component will

not have

been allocated.

In addition

exists

the possibility
available

since

design

time

target

times

of engineering

for

components are not allocated

designers

between components.

mis-apportioning

Under these

conditions

the relative

design time allowed for components may not always reflect

the relative

importance of the components to the overall

design.
the product

success of

For example, during the design of the New EHB

items
bearing
housings,
such
as
minor
cost
on
time
the
expended
range,
items
to
the
the
to
overall
contribution
relative
was not proportional
the
product.
of
new
producibility
and
reliability
cost, quality,

be
to
design
allocated
resources
To enable
a project

effectively

into
that
takes
consideration
technique
planning

importance of each component within

a design.

requires
the relative

100

The conventional

are Gantt Charts, Critical

projects

development

methods of planning and controlling

Path Method (CPM), Performance

Evaluation and Review Technique (PERT), Managementby Objectives


and Planning-Programming, Budgeting and Systems Analysis

Gantt Charts, CPMand PERTnetworks

(80)

(MBO)

(PPBS).

illustrate

graphically

the chronological

order in which development tasks should be performed

and are therefore

practical

tools

for monitoring

CPMor PERTnetworks are superior

development projects.

Charts since these methods illustrate


tasks and can therefore

project

i. e. those activities

a project,
delaying

the project

the progress of

the relationships

determine the "critical

to Gantt
between
path" for

which cannot be delayed without

completion date.

PERTnetworks are useful for planning and controlling

the new

the
development
since
uncertainty
process
product
into
be
taken
consideration.
times
can
task

surrounding development
(81)
Task times are determined

12.
Equation
using

T=
act

T, + 4T2 + T3
6

Where:

Tact

Activity
=

T1

Optimistic
=

T2

time for a task.


time for a task.

Most likely

time for a task.

Pessimistic

time for a task.

(12)

101

Although the PERTtechnique was originally

designed for

controlling

development projects,

controlling

the development of the New EHBrange had several

the use of the technique for

disadvantages:

The cost of implementing PERTis high, hence the project

c>>

being planned must be sufficiently


this

large to justify

Smith and Mahler(82)

expense.

in comparing

computer CPM and PERT programs indicate


between number of activities
cost.

Using this

EHB component if
3400 events)

being planned and program

in detail

would represent

117,000 for

the development

relationship
planned

a PERT software

to increase

(ii)

To construct

of approximately
Purchasing

package.

be cost

the equal opportunity

the size

design

a PERTnetwork diagram projects

comprised of a series of identifiable

identified
be
always

tasks.

is essentially

the
tasks
resulting
process,

effective

of using

of the H. M. Ltd.

desigq
the
activity
since

of each

(i. e. approximately

a cost

PERT system would not therefore


compared with

the relationship

this

when
cash

team.

must be
However,

a creative

from the design phase cannot

in advance of design being carried

length
type
the
and
g.
of reliability
e.
out,

trials

depends on the assembly design concept adopted, hence


diagram
be
PERT
constructed
cannot
network
a

accurately.

102

(iii)

PERTde-emphasises cost (both development and manufacturing


costs) as a specific
project

factor

to be considered during a

since resources are directed

scheduled development time.

Although PERT/COST
network

diagrams are designed to take financial


Kahalas(83) indicates

consideration,
ol

that the technique

to implement since the allocation

to individual

tasks within

the projects

"one offs",

a project

into

objectives

is difficult

is,

the

at maintaining

of costs

is not easy.

That

for PERT/COST
are generally

suitable

hence the development of standard costs for


task is extremely difficult

any particular

and often not

cost effective.
(iv)

Network analysis

techniques such as PERTare concerned

with controlling

development time and the allocation

project

resources to the cost effective

development time.
the relative
overall

of each project

contribution

project

task to the

to carry out a particular

project

development time is limited

it

task.

the time required


However when

the primary aim of a planning

technique should be to allocate


tasks such that

This is necessary

resources.

Equation 12 is concerned with determining

priority

of

As such these techniques do not consider

success of the development process.

when allocating
(v)

minimization

of

development time between

is used efficiently.

103

MBO(83) is a planning technique designed to motivate management


at every level

by involving

of an organisation

process (i. e. goal setting)

and evaluating

them on their

effectiveness

in setting.

Using the

in achieving the goals which they participate


MBOphilosophy to plan and control

the development of a new product

would force managementto establish


set of development objectives

them in the planning

a consistent

and comprehensive

for the organisation

and could produce

a higher degree of commitment to the development goals.

However the MBOtechnique is normally limited

by the degree to

which the organisational

and the personal goals of individual

can be made compatible.

This is a major problem during new product

development since managers are conscious that decisions


for the life

of the product.

The tendency, therefore,

managers

are effective
with using MBO

lead
that
to departmental functions
term
focus
goals
is to
on short
concentrating

on individual

objectives

For example, the manufacturing

aims.

rather
function

to minimize labour costs rather

materials

than on organisational
could tend to choose

than total

manufacturing

costs.

for
by
USA
Department
Defence
the
developed
use
of
PPBS(83'84) was
decision-makers

to assist
projects

with varying

numerous projects
cost-benefit
benefits

levels

of costs and resulting

limited
for
competing
are

analysis

is carried

from one project

from an alternative

in choosing, planning and controlling

When

development resources a

out to compare the cost of achieving

with the cost of receiving

project.

benefits.

defence

equivalent

benefits

104

The philosophy

of the PPBStechnique is applicable

product development process since each component within


be considered a candidate for design time.

to each component based on the benefits

i. e. the relative

contribution

analysis

benefits
for
basis
comparing
common

to be established.

the outcome of alternative

can vary considerably

are often compared qualitatively

In addition

projects

the formation
analyst

can have multiple

to be gained,

development objectives.

to the overall

The need to carry out a resource-benefit

benefits

a design can

The design time must

then be allocated

projects

to the new

rather
objectives

However since
alternative

than quantitatively.
and benefits,

often
depends
basis
on the ingenuity
of a common

in using the information

requires

at hand to develop benefit

hence
of the

and cost

descriptions
comprehensive
and
outcome
meaningful
measures which provide
in
addition
and

provide a basis for rational

The basic benefit

to be gained from increasing

design resources allocated


the overall

profitability

choice amongst alternatives.

the level

of

to a component is of course an increase in


of a product design,

i. e. the total

profit

life of the product.


However overall
total
the
market
during
obtained
in
derived
be
ways, examples of which are
numerous
can
profitability
listed

in Table 16.

105

TABLE16

1.

METHODS
OF IMPROVING
PRODUCT
PROFITABILITY

Lowering product selling

price and therefore

increasing

sales levels.
2.

Designing a more attractive

3.

Improving the delivery

or desirable

product.

time of the product.

Enabling the product launch to coincide with an

-4.

increase in sales demand.


5.

Widening the range of specifications


the potential

covered to improve

market.

In examining the means by which overall

profitability

can be

improved the present work considers that the four basic activities
involved

are:

(i)

Improving product design specification.

(ii)

Improving product characteristics


reliability,

quality

and serviceability.

(iii)

Lowering the product manufacturing

(iv)

Improving the producibility


design.
product
a

such as style,

costs.

(ease of manufacture) of

106

of design resources must determine the

Hence the allocation

of resource allocation

affect

specification,
their

effect

and cost) and determine

producibility

attributes,

(i. e.

on design characteristics

product profitability.

on overall

designs manufacturing cost is the

When comparing alternative


Ile

design characteristic

that can be related

Product specifications,

with accuracy and consistency.


and producibility

are difficult

such subjective

criteria

relate

attributes
to product
depends on

of these characteristics

as the purchasing motives of potential

Hence the relative

buyers.

to accurately

since the effect

profitability

to product profitability

cost of components within

design is the most logical

criterion

a product

to use for allocating

development

and design resources.

Although it

basis of a component's relative


that nevertheless

factors

(i)

cost within

a product design, several

must be considered are:

to determine the influence

The necessity

(specifications,

design criteria

(ii)

design resources on the

to allocate

is feasible

of other

attributes

and

producibility)

on the level

The allocation

design
for
be
used
only
of resources can

improvement, i. e. it
design solution,
this

solution

of resources allocated.

is necessary to develop an initial

estimate the manufacturing

costs of

level
the
to
these
estimate
costs
and use

improvement.
design
for
to
use
of resources

107

(iii)

The necessity

to develop cost targets

order to ensure the efficient


Here. target

for designs in

use of design resources.

costs could be defined as those costs

beyond which additional

cost reductions would not

the cost of the extra design resources

justify

required to obtain them.

In practice

cost limits

practical

Brooks

(85)

margins required and quality-levels

the design proceeded.

tool a "cost trend chart"

the anticipated

An anticipated

variances

cost trend allows a greater degree

are within
actual

the pre-defined

and target

reduction
to indicate

action

indication

limits.

required.

That is,

product costs

The extent of the variance between


the type of cost

small differences

would tend

could be achieved by the engineering

indicate
large
would
variances
designer whereas

value analysis.

cost

costs) when plotted

of whether total

indicate
be
to
used
could
costs

that cost reduction

cost reduction

In addition

between actual and target

on the chart provide a visual

was

growth in product costs as

design
tasks.
in
the
of
freedom
sequencing
of
(the difference

necessary.

of costs during NPD. In order to provide

managementwith a usable cost control


that indicated

product

costing to develop a

used the concept of target

programme for the control

constructed

to assess, hence more

could be obtained using competitors

expected profit

prices,

are difficult

such cost targets

that a more radical

technique was necessary, e. g. value engineering

or

108

If negative cost variances were obtained (i. e. actual costs were


lower than target

costs) this

could indicate

the designer spends in attempting


be reduced, hence the overall

plan and control

Since existing
alternative

to improve manufacturing

costs could

design time could be lowered.

during the development of the New EHBto

It is important
effectively'

that the amount of time

project

the limited

design resources available.

planning techniques are not appropriate

developed.
be
to
needs
method

an

Such a planning method

would need to contain:

(i)

A method of allowing
data quickly.

designers to obtain component cost

This data would then be used to allocate

design time.
(ii)

A relationship

between design time and the relative

cost of components within


(iii)

A method of determining
suitable

product costs.

and controlling
In this

A
on

product costs and expected profit

A method of monitoring
against

component cost targets.

base
be
to
cost targets
would
method

competitor's
(iv)

a product.

margins.

design time

case a suitable

"cost
the
trend
be
the
of
use
charts"
would
(85)
by Brooks

method

proposed

109

2.5

Design Scheduling
Undertaking development tasks in the correct

in minimizing

development time by ensuring that tasks that form

total

path are not delayed.

the critical

However during the design of a new

path may be constantly

product the critical

changing since the design

lie
development
tasks
that
create
process may

on the critical

For example, a component may be designed requiring


period for raw materials
critical

existing
therefore

sequence is essential

or special

a lengthy delivery

hence affecting

tooling

path.

the

path, or a component may need to be long term tested


forming a new critical

automatically

path.

The scheduling of work through the design phase has important


effects

functions
departmental
the
other
on

within

the NPDprocess.

design
from
the
department in the
is
design
released
Normally
work
form of finished

product or major sub-assembly drawings which result

departmental
functions.
being
on
other
loads
placed
in high work
a Jig and Tool department

level
the
For example,

of resources within

is normally

to cope with the relatively

sufficient

low level. (as

term
tooling
NPD
of
short
requirements.
requirements)
with
compared
for
is
NPD
available
normally
requirements.
Limited spare capacity
In addition
manufacturing

tooling
term
short

to continue,

operations

NPD tooling

requirements.

high levels

of tooling

normally

overloads

tasks are needed to allow day-to-day


hence they take priority

It is therefore

over

necessary to avoid releasing

requirements during the NPDprocess since this

the tooling

department.

Hence to ensure the

development schedule-work must be sub-contracted


the
to
delay
minimum
leading

development
costs.
increased
to

110

Scheduling design work such as to minimize total


since it

time is not always the major objective


to plan work such that total
is in order to financially

is often necessary

product costs can be estimated.


justify

This

the development work with the

minimum loss in development resources.

In addition

of product costs allows control

estimation

development

the early

procedures to be established

that would allow costs to be reduced to an acceptable level.

However since development time is limited

to sequence design work could be of significant

the ability

Again existing

project

planning methods are not appropriate

method needs to be developed.

alternative

a method of allowing

need to contain
time

during the NPDproject

for

required

development

task

interactions

(in

order

hence an

Such a technique would

designers

components and assemblies

benefit.

to determine

to be developed

to identify

"critical

the total
and
path"

components).

2.6

build

Alterations

to Product

Designs

Altering

designs
after
component

their

and test

NPD
be
frequent
the
to
was
observed
process,
a
of
stages

release to the prototype

during the New EHBdevelopment process.

activity

determine
total
to
the
used
chosen at random were
that
changes

Nt

A sample of components
number of design

had occurred in this manner, i. e.

Na.Nr. Nq

(13)

111

Where:

Nt

design
incurred.
Total
of
changes
number
=

Na = Average number of design changes per component.


N-r = Total number of components within

the product range.

NA = Average number of design changes per component that


from design interactions

result

between components,

i. e. caused by Na type changes.

For the New EHBprocess Na = 4.5, Nr. = 400 and NA = 2.5


in Equation 13:

hence substituting

Nt

4.5 x 400 x 2.5

= 4500

it
designers
was established
From a sample of experienced

that

design
change
was
to
approximately
each
time
out
carry
the average
0.50 hours (includes

Therefore total

draughting

and functional

calculations).

design time needed to alter

existing

designs

15:

4500 x 0.5

2250 hours

17.6%
the
total
of
hours
2250
represents
available

for the New EHBproject.

design resources

Hence it would be an advantage

design
during
take
the
NPD
that
changes
place
of
the
number
to reduce

112

This would also reduce the effect

process.

of design changes on other

stages in the development process, e. g. prototype


fixture

build.

testing

and jig

and

From an examination of the design changes that occurred

during the development of the New EHBrange the major causes of design
changes were identified,

Table 17.

TABLE17

MAJORCAUSESOF DESIGNCHANGES

Proportion

Cause

of

Proportion

of Total
Design Resources

Changes

Cost Reductions
improvements

Functional

improvements

Attribute
Producibility

improvements

Improving specification

Table 17 indicates
carried

range

0.78

0.137

0.10

0.018

0.05

0.009

0.04

0.007

0.03

0.005

that the majority

out to reduce product manufacturing

in obtaining

of design alterations
costs arising

from
data
the present cost estimating
cost

were

from the delay


'
system. During

had
be
designs
to
to
the prototype
often
NPD
released
the
process
and test

delay
the
to
to
serious
prevent
project
stage

completion date.

been
had
designs
costed, cost reduction
Subsequently when
found to be necessary normally resulted
Table 17 also indicates
required

the proportion

design
alterations.
to carry out

build

exercises

in changes to component designs.


of the available

design resources

113

from cost reduction

Design changes arising

tasks can be reduced

in two basic ways:

(i)

Develop a more responsive cost estimating


data
quickly
cost
provide
reduction

loe

system to

in order to allow cost

exercises to be performed prior

to the

designs
design,
date
to
the
of
prototype
release
and test

build

The decision

(ii)

related

stages.
a component design should be

to alter

benefits

to the potential

that could be obtained.

Only minor savings would be obtained from reducing the


hence
these
low
savings should
components,
cost
cost of
be compared with such disadvantages of design changes as
delay in project
fixture

date,
alterations
completion

to jig

and

design and resources required to improve component

design.

Functional
revealing
functioning

design improvements arise due to prototype

design deficiencies.

Design alterations

therefore
are
of a product

ment process.
design defects

The small proportion


indicates

life
bearing
such as

a natural

testing

to improve the
part of the develop-

of changes that arise

because of

the power of mechanical engineering models;

equations, and the experience of the designer in

designs.
developing working

86'87)

114

The proportion
product specifications

of design changes arising


and attributes

that market data is collected


department.
attribute
competitive

However a certain

could be reduced by ensuring

in a suitable
proportion

format for the design


of specification

and

design improvements could be necessary to counteract


action during the NPDprocess.

Requests by the manufacturing


of designs should take place prior
prototype

from the need to alter

build

and test.

function

to improve the producibility

to the release of designs for

Hence the manufacturing function

in
design
the
designs
to
process.
early
examine
allowed

must be

115

CHAPTER
3
COSTINGPRODUCTDESIGNS

116

3.0

Introduction
Costing is essential

designs.

to the development of effective

new product

To determine the manufacturing costs of a product design it

is necessary to identify

the relevant

cost elements involved and

decide how each element is to be estimated.


Z

Equation 7 is useful to designers when comparing alternative


product designs but since this
of manufacturing

model contains only the variable

cannot indicate

it

the true cost of a specific

In
to
this
to
design
order
achieve
company.
a
product
the total

costs (which includes fixed

manufacturing

costs

objective

costs) must be

obtained.

3.1

Total Manufacturing

manufacturing costs are determined using

At H. M. Ltd. total

This model is similar

to Equation 7 but in addition

defective
for
an allowance

components manufactured and

Equation 14.
contains

elements for recovering

..

CM =

Costs

rI

both fixed and variable

---

m=n
("im.
k(1+Sk)

m=1

-1

im)

+1+

overhead costs.

Om

Ct

1+ Oa

(14)

117

Where:

= Total manufacturing

costs.

Sk- = Allowance for defective


Production

Tm

components.

time at the mth manufacturing cost centre.

Labour cost rate at the mth manufacturing

cost centre.

m=n
Tm.Lm

(See Equation 5).

Cp

m=1

Om

.= .

Manufacturing

overhead rate for the m"" operation.

Administration
=

Costs

Material

3.1.1

The material

link

the principal

involve
can

the purchasing

with

function

in up-dating

assigning

department

material

a company is initially

since

suppliers.

when product

Up-dating

costs

bf standard

costs

for

many

and purchased

However, the effort


is normally

each type of component and raw material

Lists

material

on the part

ranges contain

types of raw materials

material

they are normally

amount of effort

as in the case at H. M. Ltd.

components
involved

data used within

a considerable

of different

thousands

cost.

cost

by the purchasing

procured

costs

overhead rate for the company.

avoided

by

a standard

commonly used parts

and raw

of

118

are maintained and issued to departments requiring

materials

cost data.

material
intervals

These lists

at monthly

are re-standardised

using average percent price increases based on such

economic factors

inflation

as current

rates.

At intervals

(the length of which is determined by the cost of the raw


is up-dated using actual cost

each raw material

material)

data obtained from material

suppliers.

The use of standard cost data for costing new product


designs could represent

a possible

source of error

standard costs do not always reflect

for current

structures

order quantities.

new product could therefore


hence the discount

the true price of the

standard costs may be based on discount

In addition

materials.

since

alter

The manufacture of a

these order quantities

and

rate that was applicable.

During the development of"a new product engineering


designers
-potential.
obtaining

frequently
suppliers.
cost

However, the final

as supplier

reliability,

sourcing and vendor rating(88)

the
ultimate
choosing
when
more efficiently
the relevant

responsibility

data must belong to the purchasing

factors
such
since
multiple

cost data directly

obtain material

carried

material

from
for

department

material

availability,

need also to be considered

supplier.

This task can be

out by purchasing personnel who possess

background experience.

119

3.1.2

Scrap Costs
The basic types of scrap costs that must be considered
designs are:

when costing

(i)

Waste material

'

flash

costs such as machining swarf,

produced during forging

incurred

and waste material

during flame cutting.

Defective

component costs,

i. e. components that

are manufactured outside the accepted quality


requirements and must therefore

be scrapped.

is assumed during the present discussion

It

that

bought out components are replaced free

defective
of charge.

At H.M., Ltd. the excess cost that arises


defective
of
the
scrapping
and
increasing
the value

standard

sheet,

casting

levels

produced.

engineering

In addition

for,
other
be
used
can
then
final

this

can be taken

material

raw material

could provide essential

tasks.

cost

percentage,

type

(e. g. bar,

purposes
into

of scrap

costs should be determined from the

between the initial

this
since
shape

cost

and is based on past experience

However, waste material


difference

by a fixed

the
depends
on
raw material
of which
or tube)

loss

components is compensated for by


costs

material

from material

if

form and final

cost data for value

the waste material

or has a significant

consideration

of components.

component

produced
scrap value

when determining

the

120

Basing defective
simplifies

component costs on material

the problem since other factors

type over-

that need to be

considered are:

(i)

Component design.

(ii)

Batch size.

(iii)

Quality

(iv)

Dimensional tolerance

(v)

Operator experience.

(vi)

Complexity

(vii)

Reliability

requirements.
requirements.

of manufacturing
of production

process

used.

equipment

used.

A simple model (Equation 15) for predicting

the number of

components produced based on the manufacturing


(89):
Sims
by
is
given
size
defective

Ndef.

(15)

QP

Where:

Ndef.

Qp

Number of defective
Manufacturing
=

batch

components.

batch size.

121

Sims then uses the "learning

arising

in both labour cost and scrap loss

the reduction

consideration

from the "experience

a graphical

representation

effect".

quantities,

The "learning

components manufactured) arising

skills

and experience.

The "learning
transformed

curve" is exponential

in character

and is therefore

line,,

Figure 7, if

plotted

into a straight

curve" is

of the improved performance (reduction

in the percentage of defective


from increasing

curve" concept to take into

on log-log

graph

paper.

According to Sims the reduction


the "experience

in scrap loss arising

is taken into consideration

effect"

Equation 16 to estimate the number of defective

Ndef.

Sa
=

from

by using

components.

jQp

(16)

Where:

Percentage
specific

average scrap per batch for


"learning

situation".

The- value of Sa depends on the particular


situation"
rate

'which can be defined as the relative

(i. e. -learning

"learning
decrease in the

curve performance) at which defective

manufactured.
are
components
factors
such
is dependent on

The learning

curve performance

as the component quality

and

122

FIGURE 7

70%, 80% AND 90% LEARNING CURVES


Adapted from Ref.
ae

ae
0
rn

(89)

aR
0

CC)

C
C
C

0
c

c
CD
N

C'
C

U
N

O
J

N
"r

tn

C".

0
0

C)
ko

(PS)

0
c

C)
N

(ale: )s 60l)

CD

CJ

dea: )s a6eaanya6e;

CV

u30J3d

(D
L..
.

123

dimensional tolerances

required and varies between manufacturing

For example machine shop work has a learning

operations.

curve

90% and complex assembly work one

performance of approximately
of 65% to 85%.

8,
Figure
the
comparison,
of
scrap proportions
-A

using Equation 17 for 90%, 80% and 70%

produced when calculated

curve performances indicates

learning

(Sp)

that only small differences

exist.

Ndef.
.SP =

In addition

Qp

100%

these differences

Hence a common learning


of manufacturing
to accepted

tasks

estimating

learning

curves

errors)

in the scrap proportions

that

manufacturing

curve for

involve

large

approximately

(relative

errors

evaluated.

90% and assembly work


use at H. M. Ltd.

both machining

these
two
be
the
of
average
values,
would

Manufacturing

a variety

complex machine type work usually

of approximately

operations

batch size.

be used for

incurring

approximately

in-the

increasing

without

70% a suitable

tasks)

decrease with

curve slope could

Since Sims indicates


develop

(17)

(where

and assembly
i. e. 80%.

batch sizes for the New EHBproduct average

100 components, hence from Figure 8 the errors

levels
scrap

80%
learning
by
an
using
produced

curve

124

FIGURE 8

COMPARISON
OF SCRAPLEVELS ESTIMATEDFROM
90%, 80% AND 70% LEARNINGCURVES

i.

II1
LO

-1
ct

MN

(dS) de.A:)S %

r-

I^

125

(as opposed to individual


would be +2.4% to
accuracy limits

learning

curves for specific

These error

-1.4%.

the

are within

by managementand the actual

of 10% specified

accuracy (see Section 4.4.1)

levels

processes)

of the current

estimating

system

at H.M. Ltd.

3.1. J' Overhead Costs


Overhead costs form a significant

part of the operating

expenses of a company and are recovered using either

Absorption costing

costing(90).

marginal

by H.M. Ltd. and is carried

is the method adopted

out by apportioning

costs to each component manufactured within


suitable

overhead recovery rates

(i)

A large

proportion

cost

overhead

increased
decreases.

are fixed

the volume of

Therefore

is to be recovered

maintained

margins

higher

are: -

of overhead costs

components manufactured.

place

a company using

and hence do not vary with

costs

overhead

(Omand Oa, Equation 14).

problems that exist

technique

total

new product designs using the absorption

However, when costing


costing

costing and

accounting methods, i. e. absorption

financial

of two

if

a fixed

and fixed

overhead recovery

rates

profit
must be

when the number of components manufactured


Hence if

new product

when a company's sales


overhead

recovery

rates

development

takes

volumes are low the


in operation

could

126

"overcost"

effectively

the new product design.

This

occurred during the development of the New

effect

EHB range when H.M. Ltd. increased overhead rates


by an average of 135%. However, since the new
product is intended to improve the sales demand a
product design cost could be obtained

more realistic

sales volumes to determine overhead

by using forecast

This would appear a logical

recovery rates.

since both material

method

and labour costs are also based

sales demand.

on forecast
When costing

new product designs it

only relevant

overhead expenditures

is essential

that

are absorbed by

true
to
the
to
cost
of
a
product
prevent
a product
distorted.
becoming
a company

At H. M. Ltd.
both standard

is particularly

important

custom-designed

products

are manufactured,

custom-designed

products

require

(e. g. drawing

services

which are not necessary


product
rates

ranges.
for

benefit

office,
for

Hence using

both product

the manufacturing

since

this

additional
tendering

the production

and

i. e. the
overhead
and costing)
of standard

a common set of overhead.

types would tend to inflate

cost of the standard products.

to a company in manufacturing

standard

be
then
reflected.
genuinely
not
would
products

The

127

(iii)

If

the prime manufacturing


of a product

costs)

improvements
be lowered
less

costs

are reduced through

the overhead cost

could

also

contributes

of a company.

be considered

could therefore

+ material
design

recovered

a lower cost product

since

to the overhead recovery

the design

(labour

Hence

less

desirable.

Marginal costing
product designs.

is an appropriate

Using this

method of costing new

method fixed overhead costs-are

the
from
variable
separated

elements of total

costs and the

contribution

of products to the total

determined.

However, when using the marginal costing technique

it

overhead costs is

is necessary to determine the proportion

incurred
that
are
overheads

of total

company

by the product range being examined

in order to adequately reflect

the true financial

contribution

of the product.
1 1.

annual

Since overhead

costs

manufacturing

costs,

the
by
enables
company
used

form a large

proportion

the absorption
greater

control

costing

of H. M. Ltd. 's
method presently

to be exercised

over

than
the
would
costs
marginal costing
of
overhead
the recovery
.

technique.

128

3.1.4

Direct

Labour Costs
of labour costs is dependent on the

The determination

of the time required to carry out each of

accurate estimation

the operations needed to manufacture a component. Hence estimating labour costs is synonymous with that of estimating'the
to carry out manufacturing

tasks.

Tm (Equation 14) at the mth manufacturing

The time required

time (FTF) and is

"floor-to-floor"
the
is
termed
cost centre
to complete all

the time required

When costing

cost centre.

is broken down into


that

factors

affect

whilst

Knott(92)

machinery operations

production operations

at a

the FTF time

elements, each of which have distinct


their

costs.

times to be made up of preparation


process manipulation

time

Blore(91)
tasks,

considers FTF

process tasks,

tasks and component manipulation

tasks

considers gauge and/or inspect to be an

additional

activity.

(i)

Set-up

These tasks are now considered in detail:

or Preparation

Time

The set-up time is that time required to position


the various

tools and fixtures

out the manufacturing operations


tasks as adjusting

necessary to carry
and involves

machine settings

the type of machine jaws.

such

and changing

129

that is necessary

Normally each type of operation


to set-up

a manufacturing

However Owen(93) points

time.

the process

equipment

adjustments

required

always simple
i

process

since

is given a standard
determining

out that
tools,

settings,

fixtures

and

to machine a component is not


the problems of "complete

set-up",

in which the set-up


and "adjust
already

is started

set-up",

from a stripped

in which the correct

on the production

equipment

machine

tools

and only

are

need

must be considered.

adjusting,

In addition

production

maximization

scheduling rules,

of machine utilization

such as the

also affect

set-

up times since components are sequenced to minimize


set-up tasks.
Component Manipulation

Time

Component manipulation

tasks

involve

unloading

components onto and off

equipment

and if

necessary

loaded.
when
component

loading

and

the production

re-positioning

the

The time taken to manipulate

dependent
is
on:
a component

(a)

component size,

e. g. large components often

loaded
be
to
and unloaded using mechanical
need
handling devices which result
times being dependent on their

in manipulation
availability.

130

(b)

design shape, i. e. holding fixtures


required to position

(c)

type of operation,

(iii)

components on machines,
i. e. the accuracy and hence

positioning

time with which components are

positioned

varies depending on the type of


to be performed.

operation

.41'

may be

Process Time
The process time represents the time required, for the
equipment to carry out the processing

production
tasks,

e. g. metal

The factors

pressing.

are dependent

and welding

metal

feeds,

that

forging

influence

and

forging

time

process

being carried

For

out.

time is dependent on welding

process;

availability;

welding,

on the process

welding

example,

speeds,

removal,

speed

is dependent on power

removal times are dependent on

machine power rating

and condition

tools
the
used.
and
machine
of
(iv)

Process Manipulation

Time

Process manipulation

tasks are those required

the
that
processing conditions
ensure
e. g. altering
selecting

are satisfactory,

feeds and speeds, positioning

welding conditions

and tools.

to

tools,

131

In the present research this

element contains the

time required to carry out the quality


referred

to by Knott(92)

tasks

control

as separate tasks,

e. g.

checking component dimensions and surface finishes.


checks (although often separate tasks) are

Quality

an extension of machine manipulation

tasks since

they are necessary to verify

the effectiveness

process control

In many instances

checks also form part of a manipulation

quality
task,

activities.

of

for example, when positioning

a horizontal

boring tool.

The production

time to carry out all

activities

necessary

is
the
converted to cost using an
component
to manufacture
labour cost rate (variable

appropriate

listed
Ltd.
H.
M.
are
cost rates used at

A factor

be divided

control
by the

involved

Labour

in Table 50, Appendix II.

to consider when calculating

number of operatives
operators

Lm, Equation 14).

production

in each operation.

cost is the

That is,

when

hourly
than
cost rates must
one
machine
more
number of machines controlled

*shown
in
as

Equation 18:

Llm
Nm

(1s)

132

Where:-

Lim

= Labour cost rate for the mth cost centre.


Number of machines controlled

by operator.

McGrath and Conlon(94) have examined the "interference


when an operator

effect"

two machines in order to develop

controls

schemes that ensure the operator

-incentive

to keep both pieces of equipment operating


The "interference

capacity.

through setting-up

effect"

or operating

occurs when the operator

one machine must allow the other

to increase the FTF time of a component.


controls

industry
in
common
of, increasing

effect

Situations

would tend
in which

a number of machines are becoming more

hence the "interference

effect"

will

become

importance and subsequently may need to be included

14)
for
(Equation
estimating
the
in
model
Since this

costs.

at maximumpossible

It can be seen that this

idle.
be
to
machine

one-operator

has the motivation

the "interference

type of situation

total

manufacturing.

is not commonat H.M. Ltd.

on FTF times has been ignored during

effect"

the present work.

3.2

Analysis

Estimating
Cost
of

'estimation
The
defined
system" as

processes
purpose(s)

costs may be considered a "works

of manufacturing
(95)
i. e. "a works system is the operational
by Nadler

financial
human,
the
of

combination
its

at H. M. Ltd.

input material,

desired
a
of

and physical

information

resources which

or humans to achieve the

state of a product or service".

133

fact

Recognition of this
function

(Inputs,

characteristics

is possible

it

to determine the role each system

or interactions

In addition. relationships
istics

By examining these factors

plays in achieving the overall

characteristic

can be more easily

can be highlighted

identified.

system objectives.

between system characterHence incompatibilities

and if

between

and among system features,

and system objectives,

system characteristics

seven system

Outputs, Environment, Physical Catalysts,

Sequence and HumanAgents).

individually

enables the cost estimating

a company to be described using Nadler's

within

Function,

therefore

necessary improved.

Function

3.2.1,

The function
management with

of a cost estimating

the type, quantity

and quality

plan and control

needed to'efficiently

system is to provide
of cost data:

a manufacturing

organisation.

Anthony and Lahiff(96)

use of sophisticated

techniques
estimating

therefore,
must
system
increases

a particular
normally
large

managementscience

complex operations.

to plan and control

to future

of

by managementcan be expected to increase due

cost data required


to the increasing

point out that the quantity

have the ability

in demand for cost data.

A cost
to respond

This would prove

development
during
which
product
new
advantage

requires

quantities

an estimating
data.
cost
of

system to quickly

generate

134

of cost estimating

An aspect

is that

literature

function

response from an estimating


product

development

quickly

changing

in order

during

a large

resulted

possibly

lengthy

to prevent

to

to react
price

in

in designers

sub-optimum,

design

making
decisions

to the. NPD programme.

delays

number of design

changes were incurred

the NPD process.

The accuracy
designing
when
an important
level

in new

of the New EHB difficulty

cost data often

and therefore

Consequently

lead to delays

such as competitors

During the design

subjective,

A poor

programmes or the inability

market forces

obtaining

quickly

cost data.

can provide

system could

to in published

which is the rate

of system responsiveness

at which the estimating

reductions.

not referred

of cost

an estimating
element of risk
acceptable

of risk

estimates

In large

organisation.

consideration

this

represents

system since

factor

in management decision

and hence estimate

on the volume of product

dependent

is an essential

making.

accuracy

manufactured

volume manufacturing

The

required

within

is

an

systems changes

in component design or work methods, although leading to small


in
a components manufacturing cost,
changes

could when multiplied

by the large number of components manufactured result


cost variations.

substantial
estimates

in

Hence the accuracy of cost

be
fundamental
these
circumstances
would
a
under

consideration

and could therefore

take precedence over other

cost estimating

system characteristics,

responsiveness

and estimate preparation

such as system
costs.

135

The cost
consideration

of preparing

estimates

the estimating

since

that

The amount of overhead expenditure


function

the estimating

is related

numbers of components made.

batch or job

shop manufacturing
design

are not of a standard


requirements)

the ratio
could

manufactured
estimates
represent

a significant

generate
and cost

data that

which

requires

large

high.

if

particularly

(i. e. designed

to-customers

required

products
specific

to components

Hence the number of

the manufacturing

low level

enables short

to be performed.

not sufficiently

hand, in small

On the other

system at H. M. Ltd.

the relatively

is

because of the

small

could

activities

overhead expenditure.

The estimating

activities

for

systems the estimating

of cost estimates

to control

level.

can be justified

systems,

be relatively

required

an

to the type of manufacturing

In large volume manufacturing


system.
r
cost per component would be relatively
large

is normally

function

must be held at an acceptable

that

overhead expenditure

is also an important

responsive

has been designed

of manufacturing
term planning

Consequently

to

time standards

and control

the estimating

system

to the needs of the NPD process

amounts of cost data quickly.

data
H.
M.
Ltd.
time
standards
at
cost
and
The accuracy of
depending
the
to
annual quantity
on
vary
is allowed
manufactured.
function

of components

Hence the overhead burden of the estimating

is maintained within

acceptable limits.

136

3.2.2

Inputs

and Outputs

Input data required to prepare cost estimates is provided


departments, i. e.

by several functional

(i)

Marketing provides information

and long range sales forecasts.

quantities
(ii)

on annual sales

Design provides general arrangement drawings,


product designs, parts lists

detailed

and quality

requirements.
(iii)

Purchasing
quantity

(iv)

provides
discount

Manufacturing
to estimate
machining
and jig

material/component

costs

and

data.

provides

the production

component manufacturing
feeds and speeds,

and fixture

process

data with
times,

e. g.

set-up

times,

which

costs.

Finance provides labour and overhead cost rates.

(v)

The amount and detail

of input data is dependent on the

the
estimate required.
of
accuracy
manufacturing

costs within

would require

a detailed

For example, estimating

a high volume manufacturing system


breakdown of the operations

necessary

to produce a component, whereas for less accurate cost estimates,


be
appropriate
which may
individual
detail

operations

for jobbing and small batch situations,

may be aggregated hence reducing the

data.
input
the
of

137

since the input data used to develop a cost

In addition

it

that these conditions

is essential

for a component

the manufacturing conditions

estimate reflects

are controllable.

Any

from the manufacturing conditions

permanent deviation

used
of that

to prepare a cost estimate would reduce the validity


estimate.
i

Output data must be provided in many forms to satisfy


functions.

the needs of different

uses cost data on a product and

planning and control


manufacturing

cost centre basis, whereas the production,


department use production

control
and
planning

i. e. the manufacturing

The designer
to be evaluated.
design

individual

For example, financial

times of single

requires
This

cost

or batches of. components.

data to enable design

could necessitate

features

time standards,

costing

such as machined bores,

concepts

in detail
chamfers.

and threads.

It

is essential

therefore

information
able to provide cost
detail.

That is,

that an estimating
in a variety

on a design feature,

system is

of forms and

process, component,

basis.
basis
centre
cost
or
sub-assembly, product

138

3.2.3

HumanAgents
The humanagents (estimators)

istic

system since they must be familiar

of an estimating

Since conventional

to develop cost estimates.


a detailed

methods require

is usually

familiar

are fully

estimating

breakdown of the method of

it
of
components
manufacture
,

with

methods used and types of input data required

the estimating

estimators

are an important character-

that

essential

with products,

processes and work

a company. It is normally the degree of

methods used within

background experience of estimators

that determines the speed

be
developed.
estimates
can
which
with
and accuracy

Ostwald
training

and Toole

training

have no formal

in that

the current

skills

estimators

detailed
gather

the estimating

function,

than a

need to develop

a considerable

grounding

being
the
and
methods
costed
processes
with

have the ability

manufacturing

This situation

processes of a company.

preparing
of
the
cost
enables
estimators

the majority

"more of an art

through obtaining

in the manufacturing

Familiarity

towards

attitude

that

education.

or estimating

Therefore

science"(98).

and concluded

is considered

estimating

estimating

surveyed the type of background

to estimators

given

reflects

(97)

estimates to be reduced since

to make subjective

estimates of

to
hence
time
the
required
avoid
and
costs
input data.

139

existed to the development of reliable

were found to

time allowed to prepare estimates,

lead to inconsistent

For example,

estimates.

such as high work loads and

undue pressures on estimators


insufficient

barriers

found that additional

Anthony and Lahiff(96)

and inaccurate

estimates being developed.

It is necessary to consider these factors

when costing new

data
large
the
designs
amounts
of
required
since
cost
product
during the development process often overloads the estimating
department.

Companies such as H.M. Ltd. are unable to cope with the


development
the
data
of
new
product
process
requirements
cost
"

when the estimating


relatively

low levels

day-to-day

manufacturing

required

to support

are insufficient
effective

of cost

short

experience

Therefore

estimators

estimators

are also

over

cost

term planning

activities

of estimates
of a company

estimating

methods and processes


of numerous estimating
must perform
responsible

for

other

for

function

new product

require

cost

a wide

the

engineers.

duties.

At H. M. Ltd.
short

maintaining

These tasks normally

the

to control

required

(as in the case of H. M. Ltd. ) if

operations.

estimating

to provide

The low levels

to make an individual

range of manufacturing

manufacturing

information

activities.

particularly

background

has been designed

function

term

take precedence

development.

140

3.2.4

Sequence
The number of stages involved in developing cost estimates

can influence

the responsiveness of an estimating

system.

The

time to develop a cost estimate can be considered, using


70),
to be the sum of a queueing time (the time
Queueing Theor3
a request for a cost estimate must wait before the actual
task of estimating

begins) and a service time (the time taken

to carry out the estimating


estimating

The responsiveness of an

task).

system is the sum of the queueing and service

and is therefore

times

dependent on the number of stages involved

in

developing a cost estimate.

Queueing times are dependent on:

given to requests for cost estimates.

(i)

The priority

(ii)

Humanfactors

such as motivation.
4

(iii)

The current
estimating

work load and capacity

of the

function.

Service times are dependent on:


. ,t

(i)

The cost estimating


determines the level
data.
output

method used since this


of detail

factor

of input and

141

(ii)

The accuracy of estimates required


determines the estimating

factor

since this

method used.

(iii)

The experience and ability

of the estimators.

(iv)

Capacity of the estimating

function.

)
to Anthony and Lahiff(96asedbiased cost estimates
if
developed
be
could

the person or department preparing

effected

or favourable

an adverse

this

or favourable

effect

For example,

since

there

for
EHB
New
range
the

designers

data,

estimators

of cost
cost

preparing

during

have little

system.

the design
without

incentive

of
the

in

in developing cost estimates

engineering

managementstructure.
department determine

times for components and the commerical department


data into manufacturing

therefore
formed
queues are
estimating

then

of an estimating

of a company's internal

At H.M. Ltd. the production

convert

function

data quickly.

function
be
a
can

this

the action

to make decisions

The number of stages involved

manufacturing

the estimate in
if

on the estimating

is a tendency

from the estimate

does not have an

an estimate

to reduce the responsiveness

tends

benefit

Similarly

manner.

from not preparing

resulting
adverse

the action that resulted

system.

costs.

Hence two distinct

reducing the effectiveness

of the

142

the costs of non-standard products

For estimating

H.M. Ltd. have recently

their

altered

procedure to

estimating

by
using the commercial department
responsiveness

improve its

to estimate both initial

manufacturing times and costs,

removing the production engineering

i. e.

departmental link.

Although the accuracy of estimates can be adversely


is no longer carried

(since the actual estimating

effected

most intimately

by
the personnel
out
manufacturing
to prepare

processes)
the total

allowing

connected with the


production

cost estimates

was not possible

department
load
the
on
work

the extra

engineering

could

since

not be adequately

handled.

Environment and Physical Catalysts

3.2.5

The environment (working conditions


physical

catalysts

. system are closely

(estimating

(e.
g. variety
of a company
processes,

level

over operations)
estimating

methods used) of a cost estimating

since it

related,

of a company) and

is the working conditions

of products,

of planning carried

variety

in manufacturing

out and level

of control

that determines to a large extent the appropriate

methods used.

Cost estimating

in
the manner in which the
methods vary

determined.
is
task
to
manufacturing
a
perform
time taken
are derived
differ

They

from work measurement techniques which themselves

level
the
to
according

of detail

down
for
broken
estimating
are
tasks

into which manufacturing

purposes.

143

In general the higher the level of detail


estimate manufacturing

used to

times:

(i)

The greater the accuracy of estimates.

(ii)

The greater the time required

to prepare an

estimate.

(iii)

The more familiar

an estimator

must be with the

working methods and processes in use.


(iv)

The less responsive the estimating

(v)

The higher

to prepare

the cost

system.

an estimate.

Since the method of work measurement has a significant


on the accuracy

effect

it

system,

is this

-appropriateness

and responsiveness

system characteristic

of an estimating
that

determines

system for costing product

of an estimating

designs during the new product development process.


estimating

the

influences:
method

(i)

The personnel who use the system.

(ii)

Estimate accuracy.
.

(iii)

Time to prepare the estimates.

(iv)

System operational

(v)

System development costs.

costs.

The

144

System responsiveness.
(vii)

Degree of subjective

and input data

estimating

used.
(viii)

Range of manufacturing

loe (ix)

Level of cost detail

(x)

Ease with

tasks costed.

obtained.

which computer software

can be

developed.

3.3

Work Measurement Techniques


It

is necessary

to measure the work content

a component in order

manufacturing
which

can be expected

would

be expected

to accurately

to take to carry

to develop
represent

involved

production

in
time standards

the time an operator

out manufacturing

tasks.

Time

to regulate their work performance to a


operators
allow
standards
(99)
facilitates
therefore
and
work planning,
rate
scheduling
pre-set
scheme operation.

incentive
and
to

manufacturing

regulate

determine

labour

the

Since production

time variability

time standards

help

they can be used to

cost of component manufacture.

Work measurement techniques are the methods by which standard


production
time

times can be established.

Since many factors

it
component
a
to
manufacture
taken

conditions,
subsequently

affect

the

is necessary to state the

have
time
been
developed
standards
which
and
under
to
these
conditions
adhere

the component.

during actual manufacture of

145

The development of standard production times is therefore


development
the
of standard working conditions.
with
synonymous
Defining,

standardising

and maintaining

difficult,

'particularly

in large, multi-product

can be

working conditions

organisations

since these tasks are dependent on the level

M.
Ltd.,
H.
as

such

of

exercised over the manufacturing operations.

management control
Ile

it

That is,

is necessary to optimize the level


companies with its

in multi-product
working

and production

conditions

i. e. `workplace

orientated

factors

of managementcontrol

The criteria

cost.

times are basically

that affect
of two types,

and operator performance factors

(e. g. the pace at which operators work).

Workplace Factors
in discussing

cooling(100)
maintaining
developed

TAB

18

the working conditions


indicates

in

management's responsibility

under which time standards were

the types of factors

that are important

(Table 18).

FACTORSESSENTIAL TO WORKPLACESTANDARDISATION
Ref.

at the work place on time.

(i)

Material

(ii)

Consistent

material

(iii)

Consistent

quality

(iv)

Well maintained tools

arriving

quality.
specifications.
and equipment.

(100)

146

TABLE 18 (Contd)

(v)

The avoidance of incorrect


inconsistent

(vi)

or faulty

tooling

causing

work quality.

Correct positioning

of tools,

equipment and material

at the workplace.

(vii)

Consistent

(viii)

Reduction in the variety

training

of operators.

of methods used, i. e. greater

job specialisation.
Effective

(ix)

However Cooling

and hence avoid

practice

the level

The affect

performance

organisations

to plan and control

problem situations

of managementcontrol

arising.

operations

amount of light

machine design,

conditions,

available,

layout of control

and hence production

have
in

However, in

is restrained

of work place environmental

temperature,
place
work
levels,

manufacturing

and work place problems arise as a'matter

considerations

noise

assumes that

management resources

sufficient
detail

communications.

by financial
of course.

such as,

absence of glare,

panels on operator

times are the specific

been
have
adequately examined(101).
Ergonomics and

province of

147

Factors

Performance

Operator

argues that the rate at which work is carried

Fein(102),

in a measurement sense, since in order to

quantifiable

not rigidly

out is

kind
be
there
of
any
must
a means of measuring work
measurements
make
definable

rate that is exact,

clearly

reference standards

(normal work rates)

and reproducible

and a set of

against which these measure-

be
gauged.
can
ments

Fein argues that work measurement literature

does not sufficiently

provide definitive

(standard)
of normal

all

concept
time

the relationship

that

for

work.

are adjusted

values

rating
performance
(103)
Owen
example,
measurement
that

found
work

performance

working

conditions,

between

trade

the technique
in which

rating

his own

at work against

factor

rating

of the observed work to his mental


a time standard,

by the performance

rating

as being

criticised

examined the effectiveness


in producing

realistic

the observed

factors.

However,

inaccurate.

of present
time

For
work

standards.

He

to adequately

rate

loss
of management control
of
cause
was a major

over

inability

the

requirement

are based on the

a numeric performance

In calculating

is often

techniques

standard

performance

compares the operator

of normal and establishes


indicates

which

is called

synthesis

mentally

analyst

concept

study

reference

For example,

of the analyst.

judgement

time
in
used

this

in common, i. e. all

have one characteristic

subjective

the

work rates.

used to. fulfill

Techniques

guides for the establishment

unions

of work study

leading

practioners

to the introduction

and management.

of rates

negotiated

148

Mansoor(104) found that the accuracy of time standards


was dependent on the ability

estimations

out the rating.

carrying

to develop a check list,

of the time study engineer


(105)
This dependence led Faherty et. al.

Table 19, of factors

essential

to successful

which could enable large groups of independent companies

estimating

to reduce. the variations

in time standards developed by individual

time study departments.

in estimating

The difficulty
overcome by applying
data(106,107).

study
of

Pre-determined

An extension
Time Standards

been given

a basic

determined

Time Standards

rated

rating,
also

only

work

tasks

time.

performances

applicable

techniques

statistical

have been broken down into

performance

of this

has been largely

to the analysis

of time

approach is the development

such as, MTM, in which work tasks

a set of basic

motions

However, although
data eliminates

each of which have

the application
the need for

were used in their

to large

ratings

performance

development.

volume manufacturing

of Pre-

They are

systems in which

are repetitive.

The individual

factors,

Table 20, that affect

operator performance

by
Archer
Siemens(108)
fall
and
examined
and
been
adequately
have
four

main groups.

into

149

TABLE 19

ASSESSMENT

OF A WORK MEASUREMENT
CHECK

LIST

AND

SCORE

DEPARTMENT

(105)

Ref.

SHEET

Possible
Points

A.

MANAGEMENT
X. Does the department occupy a suitable place in the company organisation
central service not tied to a particular department)?

(i. e. as a
2

Wiz. Is the status of the head of the department commensurate with his responsibilities
(Le. at least equal to that of the heads of the departments served)?
3. Are the staff deployed
Is Ile

to match the production

organisation?

there a central cell responsible for policy, research, etc?


s. Are staff interchanged from time to time?

6. Is ' there a policy manual?


7. Is there a manual of standard procedures?

24

I}

U. ` SELECTION

8. Are staff recruited from both internal and external sources?


9. Is selection by single interview, several interviews (1), board (2) or examination
(ONC, C&G,
AMIWSP,
etc. )
Io. Is a formal qualification required?

D.

$,

(3)?

TRALNING
formal
Are
course of at least 4 weeks?
a
given
recruits
untrained
II.
)
IWSP
(e.
Is
the
of
course
to
syllabus
satisfactory?
g.
standards.
and
the
conduct
12.
for
least
6
Is
trainees
period
of
at
months?
a
probationary
there
13.
2 years at least?
14. Are staff sent on continuation or refresher courses-every
from time to time as a regular practice?
held
departmental
'
Are
seminars
IS.
least once every 2 years?
departmental
Is
conference
at
there
a
16.
t
PERSONNEL
board
(t),
(2),
by
is
recommendation
or examination
seniority,
promotion
17.
for
involved?
levels
the
type
Are
of
work
reasonable
the salary
1g.
between ao;; and 5% per annum?
Is
turnover
the
staff
19.

2
3
z

(3)?

PROCEDURES
NJASUREMEN'T
WORK
before
measurement?
Are
method-studied
operations
20.
in
detail
issued?
when
standards
are
Are
specified
methods
21.
PMTS?
of
Is
made
use
maximum
22.
local synthetics, job comparisons, tables etc.?
of
Is
made
use
maximum
23.
z. }. Is the BSI rating scale used?
in
stop-watch time study?
Are
rated
all elements
25.
(z),
for
(1),
checked
consistency
sample
size
studies
adequate
time
P.
re stop-%ratch
26.
(1)?
time
recorded
against
time elapsed
data regarding
to
systematic
means
used
provide
some
Is
snore
pli
activistsa
27.
q
ether
contingencies
defined?
properly
allowances
Are
contingency
28.
for
fatigue
used
allowances?
tables
Are
guidance
29.
defined
and
standardised?
allowances
Are
personal
30.
31. Are the various allowances reviewed periodically?
issued for setting-up machines?
times
Are
standard
32.
for
the `learning curve' when long runs are expected?
made
due
Is
allowance
33"
temporary
or
used when small quantities are processed?
values
allowances
Are
run
short
34.
for non-repetitive
work?
comparative
estimating
used
Is
or
analytical
35"

3
3
2
2
z
2

2
2
2

4
2
3
4
t

3
2
I
2
I
I
2
2
I
3

Score

150

TABLE 19 (Contd)
Possible
Points

F. QUALITY

CONTROL

36. Are regular rating check sessions held, at least once every 2 months?
37. Where possible, are time studies repeated before a standard is issued:
(a) by another Time Study Man (TSM)?

5
2
s
i
I

(b) with another operator?

23. Are standards

by section (section leaders) before issue?

approved

Are
studies and/or computations scrutinised by seniors as a regular routine?
-39.
40- Do TSM

G.

Scor,

levels of operator

review

performance

in their

areas at regular intervals?

WORK LOAD
Is the ratio of TSM to operators adequate?
42. Is the number of permanent standards set per week reasonable?
(Bearing in mind cycle time, batch sizes and duration of studies. )
4t.

2}

2}

FL APPRECIATION
43. Arc appreciation courses given as a regular routine
(a) New employees
(b) Shop stewards and union representatives
(c) Supervisors
(d) Managers

SL'ALMARY
Management
Selection
Training
Personnel
Work Measurement
Quality Control
Work J; oad
Appreciation
General Impression

to:
I
I
I
I

II
7
13
7

Procedures

33
IS
5
4

(at Assessor's Discretion)

S.
I

TOTAL

zoo

TABLE 11
SSONTTORL`iG

OF A WORK

MEASL'RF-NiENT

Frequency

Items to be checked

Number of operators on PBR


Operator

Performance

Number

of permanent

DEPARTMENT

Monthly
X

(by depts.! product

centres)

standards set

Dumber of permanent standards revised


Number of Time Study Men

X
X

GraFh of results of rating check sessions


Turnover of Time Study Men
Changes

in : Management

Changes

in policy

recorded

hours

recorded

on Unmeasured

hours

Allowed

Time

Total
Total

on Measured

hours

as received

Changes in standard procedures


Total

( 6-. North

Work
Work

x
x
x

151

OPERATORPERFORMANCE
FACTORS

TABLE 20

Ref.

(i)

Job Aptitude

Rote
Factor,
-

Dexterity

Factor,

(iiY

Operator

Versatility

(iii)

Job Responsibility

(iv)

Job Conditions

(i)

Job Aptitude
(a)

(108)

Comprehension Factor,

Stamina.
Style,
-

Materials.

Repair,
-

Reject.

Equipment.
-

The Rote Factor


equivalent
"learning

described

to the Incipient

with

blueprints,

and the conditions

Learning

and Siemens is
Stage of the

in which the operator

curve"(109),

becoming acquainted
instructions,

by Archer

the set-up,

to memorize the essential

the tooling,

the workplace

of the process.
details

is

arrangements

The time required


of a job

task

depends on the experience


operator.

Carlson

and training
of the
(109)
and Rowe
considers this

influenced
is
also
stage
interruption"
is,

portion

by the "forgetting

of the learning

when a job is interrupted

That

the amount of "forgetting"

decreases as the number of units


interruption

curve.

or

occurs is increased.

completed before an

152

It

is at the final

(the Maturity

stage of the learning

Stage) where the limit

curve

to job

improvement has been reached that the remaining


factors

examined by Archer and Siemens begin to

influence

(b)

operator performance rate.

Comprehension - Howwell the job details

have been

learned.
(c)

Dexterity

The
ease with which a trainee
-

is able

to achieve and maintain the standard production


for a specified
(d)

pace

time.

Stamina - Refers to the time required to build up


the stamina necessary to perform the job at the
standard pace for the daily

working period.

Operator Versatility
Reflects

the time required

the introduction
(style)

by an operator to adjust to

of minor variations

and the effects

in job description

of changes in materials

on

machine settings.
(iii)

Job Responsibility
The time required
work operation.

to either

reject

or repair

mistakes in

153

Job Conditions
The time needed to repair

The operator

performance factors

listed

be difficult.

performance is cyclical
the initial

is low'during

in character,

that

The cyclical
use of a single

of operator

nature

time standard

for

since performance rate

rises to a peak and falls

work period,

towards mid-day, with the cycle repeating

suspect.

in Table 20 indicate

a standard work rate over the daily working period could


(110)
This is supported by Dudley
who established that

maintaining

operator

or adjust equipment.
_

off

in the afternoon work period.

performance

a specific

indicates

that

the
is

task

manufacturing

Again however, the accuracy of time standards needs to be

the
those
of
cost
preparing
standards and the consequences
with
compared

.
inaccurate
an
of using

inherent

The problems
as

estimate.

standard

maintaining

can be expected

in estimating

working

to increase

conditions
growing

with

concepts,

Enrichment("')
Job
such as

the level

of work planning carried

for

example,

that
such
their

involves

workers

and operator

over working

in developing

standard

performance,

and Job Rotationt112),

out is reduced.
tasks

such

use of manufacturing

The result
conditions
production

skills

or redesigned

and determine

is a reduction
which could
times.

since

Job Enrichment

being designed

are able to use more of their

own work pace and methods.

management control
difficulty

manufacturing

time standards,

reliable

in

increase

the

154

Within

the level

industry

manufacturing

of control

a wide variation

in

exists

exercised by managementover working conditions

performance rates. Tn consequence a number of work


,
measurement techniques have evolved with each technique being

, and operator

for specific

suitable

of: work measurement have therefore

methods

been examined in order to

system for use at H.M. Ltd.

the appropriate

-identify

Existing

manufacturing conditions.

Time Study

3.3.1

develops standards by direct

Time study(113)
work whilst

it

is being performed.

observation

The method cannot therefore

used to estimate manufacturing times and hence is unsuitable


estimating

design

considered,
technique

the pre-production

time study analysis

features

However,

be

for

design costs during new product development.

However during
NPD cycle,

of

that

the cost
since

could

manufacturing
be useful

reduce the producibility

effectiveness
time study

and therefore

Cost effectiveness

in highlighting
of a product.

of using the method needs to be


is a relatively

most suited

slow and expensive

to high volume repetitive

in low volume manufacturing

from the use of time study analysis.

In order to be cost effective:

work.

systems, as

would be dependent on the possible

in th'e case with H.M. Ltd.,


that
could result
savings

stage of the

155

'

c-Fs <N

(CI-C2)

(19)

Where:

Cry
.e

N=
Cl

Cost of time study analysis

per component.

Numberof components manufactured.


= Componentmanufacturing cost prior

to time

study.
C2

Component
manufacturing
=

cost after

time

study.

It would appear logical

therefore

to focus available

high
cost components since larger
on
resources
study

time

Cl-C2

be
possible.
could
values

3.3.2

Pre-Determined Motion Time Standards


Pre-determined Motion Time Standards work measurement

based
the
that
(PMTS)
on
concept
are
all
systems
tasks can be broken down into a limited
motions which when allocated
times to be built-up
involved.

manual work

number of basic human

basic times enable total

from a knowledge of the basic motions

To avoid the need to subjectively

rate operator

basic motion are


basic
for
times
the
each
performance
developed statistically.

job

156

Methods-Time Measurement (MTM)(113),

Work Factor(114)

and Maynard Operation Sequence Technique (MOST)(115,116)


are examples of PMTSsystems, each varying in the type of
basic motions into which work tasks are broken.

in terms of method of application

systems are similar


-application
r
the suitability

Since these
and

be
such
used to determine
one
system
can
cost,
of PMTSfor costing product designs during
MTMhas been chosen since this

the NPDprocess.

system is

used throughout the UK.

already extensively

MTM

3.3.2.1

PMTS systems in the MTM family

There are various


integrated
form
an
which
individual

systems I,

Willey(117)

indicates

III

in terms of characteristic

and accuracy

by aggregating

for

capacity

of time standards.

comparing the speed and accuracy of the

by analysing a simple work cycle,


that if

The

and V have a specific

speed of analysis,

such as

description
methods

MTMfamily

II,

to each other

relationship
features,

work measurement system.

Table 21,

the measurement system is simplified,

basic motions into larger

speed of application

groups, the

increases and the accuracy of the

9.
is
Figure
developed
reduced,
time standards

157

AND APPLICATION TIME FOR THE


ACCURACY

TABLE 21

MTM SYSTEMS

(117)

Adapted from Ref.

MTMI

MTMII

MTMIII

MTMV

12.3

16.3

% Relative Error (as


I)
MTM
with
compared
Application

Time (mins)

93

Estimated Job Time (mins)

Accordingly,

5.5
43

0.268

in deciding

15

0.283

- 0.301

the MTM system to adopt

balance
the economics of greater
must
management
ing accuracy

application

against

H. M. Ltd. 's manufacturing


production,

high level

time and cost.

conditions

(small

of custom designed

a low level of manufacturing planning),


potential

greatest

0.312

Considering

batch

products

and

MTMV offers

for cost effectiveness

lowest
the
has
application
system

estimat-

the

since this

time and therefore

cost.

The comparison of application


family,

Table 21, indicates

application

times for the MTM

the relationship

between

time and the estimated manufacturing time.

Using the relationship

for MTMV the total

time required

design
be
EHB
New
the
costs
can
product
to estimate
determined as:

158

FIGURE9

AGAINSTAPPLICATIONTIME FOR
RELATIVEACCURACY
THE MTMSYSTEMS
Adapted from Ref.

(> )

O
0i

0
Co

ta

"E
0)
E
F-

oc
LC)

"r
4J

-ra
u

.'...
CD
r
%:

C)
cY)

CD
,-

CD
N

.-

I W1W44M paaeowoo se

00

AoJa3 anPelaa

159

Test

Tap'TM

(20)

Where:

Test

Total
time
to
required
prepare
=

estimates.

MTM
V
application
=

Tap

time per hour of

estimated time.
Total
manufacturing
=

TM

time for the New

EHB range.

For MTMV, Tap = 16.03 hrs/hr.


The manufacturing

converted
labour

to labour

hours for

Hence using

hours,

Equation

Table 22.

using

H. M. Ltd. 's

20 the total
times

number of

time to estimate

the

is:

hours
6917
=

431.5

Number of man years

The total

range is 431.5 hours.

the complete

New EHB manufacturing

16.03

estimated

work measurement methods have been

synthetic

existing

costs

Total Application

Time (Hours)

No. of Working Weeks/Yr x Hrs worked/Wk


6917
46x37

4.06 man years

160

It is considered therefore
is not feasible

for costing

low volume product designs

such as the New EHB. Knott(92)


MTMV within

examining the use of

a machine shop supports this

He states that "MTMV like


family

that the use of MTMV

are more suitable

conclusion.

the other systems in the MTM


to mass production

operations

than to low volume or jobbing shop manufacturing


conditions".

TIME FORTHE NEWEHBRANGE


MANUFACTURING

TABLE 22

Labour Costs (i's)

General
ac fining

0.8
1.6

Tonne Hoist
if

11

Assembly

Fabrication

60.1

7.7

40.3

2.5

99

71.5

10.2

44.7

3.0

80.7

11.5

48.5

5.7

95.2

15.7

57.9

9.5

138.7

20.3

69.8

12.1

30.7

6.2
-

10.7

2.9

35.7

7.1

12.1

3.9

512.6

78.7

284.0

39.6

5.0
8.0
C. L. Trolley

6.3

TOTAL

(Vs/hr)
Rate
Cost
Labour
Labour Hours

Machining.

Block

3.2

3.2

Multi-

2.60
197.2

0.85
92.6

2.25
126.2

2.55
15.5

15.5
431.5
126.2
Hours
197.2
92.6
+
+
Hours
=
+
=
Total Labour

161

3.3.3

Category Estimating
Category estimating

manufacturing
as illustrated
allocated

divides

tasks into a sequence of non-overlapping


in Figure 10.

to an interval

an, individual

of

intervals,

Manufacturing jobs are then

and assigned a single

interval.
that
with
associated
estimate
.

the time range distribution

Hence it

time standard

is not necessary to

time standard-for

a manufacturing task.

A simple approach, Table 23, developed by Boehringer(118)


uses an interval
estimating

chart representing

50 equal time spaces.

time standards merely requires

an estimator

decide
judgement
to
which interval
personnel
within,

Since

to use

the job time falls

the ease with which work measurement can be carried

out makes category

estimating

suitable

for

low volume manufacturing

situations.

More sophisticated

category estimating

systems have been

analyses of job time distributions.


121)
assume that the
Secor and Kogovsek(119) and Steele(120;
developed using statistical

job
times
within
of
range

a company's manufacturing

processes

job
However
times
distributed.
examined
at
normally
are
,
(122),
turning, milling,
gear cutting,
H. M. Ltd. for welding
fitting,
and
assembly
exhibit

a log-normal

horizontal
distribution..

boring and flame cutting

162

FIGURE10

TIME BANDS
MANUFACTURING

Ref.

(g1)

tKtuV=Nl-r

U
C)

SLOT 3
()
5LL(1)

SLOT 4

-,
F--0O
7

SLOT 5
SLOT 6
17

Range of Job Times within

a Company

163

TABLE 23

50 INTERVALESTIMATINGCHART
Ref.

Std item/min
5.87-6.14

Pcs/hr
10.0

Std item/min
1.86-1.93

(118)

Pcs/hr
31.6

5.60-5.86

10.5

1.77-1.85

33.1

5.35-5.59

11.0

1.69-1.76

34.7

5.11-5.34

11.5

1.62-1.68

36.3

4.88-5.10

12.0

1.54-1.61

38.0

4.66-4.87

12.6

1.47-1.53

39.8

4.45-4.6S'

13.2

1.41-1.46

41.7

4.25-4.44

13.8

1.34-1.40

43.7

IS

14.5

1.28-1.33

45.7

3.88-4.05

15.1

1.23-1.27

47.9

3.70-3.87

15.8

1.17-1.22

50.1

3.54-3.69

16.6

1.12-1.16

52.5

3.38-3.53

17.4

1.07-1.11

55.0

3.22-3.37

18.2

1.02-1.06

57.5

3.08-3.21

19.1

60.3

2.94-3.07

20.0

2.81-2.93

20.9

2.68-2.80

21.9

2.56-2.67

22.9

2.45-2.55

24.0

2.34-2.44

25.1

2.23-2.33

26.3

2.13-2.22

27.5

2.03-2.12

28.8

1.94-2.02

30.2

974-1.01
.
931-. 973
.
889-. 930
.
849-. 888
.
811-. 848
.
774-. 810
.
739-. 773
.
706-. 738
.
674-. 705
.
644-. 673
.
615-. 643
.
587-. 614
.

i
AG
*. 064.24

63.1
66.1
69.2
72.4
75.9
79.4
83.2
87.1
91.2
95.5
100.0

164

Recent developments in category estimating


the log-normal

distribution

of job times.

Kirby(123)

the
for
category estimating
applying
a procedure
in setting

log-normal. distributions

have used this


tool

estimating

in Figure 11 requires management


in which the job time

to choose the number of categories


should be divided,

Figure 10, since the width of

the time bands determines the individual


standards

time standards for

The procedure for developing a category

system outlined

distribution

technique to

Harris and Melven(124)

technique for estimating

tasks..

repair

outlined

time standards for non-

jobbing work, Figure 11.

repetitive

have recognised

developed.

accuracy of the time

According to Fairhurst(125)

it

is

management's responsibility

to optimize the widths of the

time bands and estimating

cost since when system accuracy is

increased

the cost in terms of planning time also rises.

Although category estimating

is suitable

for developing

time standards in small batch manufacturing systems it

is generally

for costing newproduct designs because:

inappropriate

Its use requires

c>>

facturing

is suitable

knowledge of manu-

processes and methods.

need to be fully
conditions

a detailed

familiar

used within

Hence estimators

with the manufacturing

a company. Therefore it

for use by foremen and chargehands

but not by engineering

designers.

165

FORAPPLYINGTHE
OUTLINEPROCEDURE
ESTIMATINGTECHNIQUE
CATEGORY

FIGURE 11

Ref.

(123)

t
oe
Group job durations into clash which
tuocessirely increase in range by a
Common factor (e.. a 21.
Convert boundaries to togs and
t
calculate x Sic. 5.7

movie

Use Xa ten to chm* goodness of fR


to 'tog normal'

movie

Check accuracy /Avrnoc


of sample size t4

5a)

Take further

6aI

TA*

6W

Repeat 6

6c)

Check cortgatabdlity
of samples

rj

i2

`
6d)

fun. lw

<E4vTaos

Sz
S
Ef
Ef,

Add sump4esar+d

repeat 5&6
tie)

Repeat 6abcd until


sample is big enough

10

Calculate mid-values for each category


by working through probabilities at
msd-boundanes and de. iatgns.

It.

Calculate

standard

time

for

Nwage

job in each category by use of data


from 7 and 8 and relaxation allowance.
12.

Set up work standards on master heel

11.

Supervision ssumatt job eategar. es


before they are don..

14.

Obtaat sample of ategwtsed

IS.

Venfy eslomating in ample by


Comparing with work standadi

14

Calculate performance by canoarn. q


total standard fours (from ess.mated
Categories) with total clocked Wur%

17.

Prrpue

incent. ve Scheme
lxrformance.

cards

based

on

166

(ii)

The method is not designed to provide accurate cost


estimates for individual

jobs but an accurate

average time over large numbers of parts.


(iii)

Detailed manufacturing data is not used to establish


detailed

time standards therefore

cost breakdowns of

product designs could not be developed.


(iv)

A new product introduction

could significantly

job time distribution

the existing

alter

of a company.
job

Hence basing component costs on an established


time frequency distribution
inaccurate

3.3.4

Synthetic
Synthetic

cost estimates.

Time Standards
time standards

method of cost estimating


ments.

(113)

within

The method is similar

that'manufacturing

could give rise to

estimating

is an established

low volume manufacturing

environ-

to the use of PMTstandards in

tasks are broken down into basic elements

basic
times.
assigned
the
elements
and

However synthetic

time standard elements are specifically

related

products and manufacturing

and therefore

activities

only
suitable
normally
are
standards

to a company's
PMT-

unlike

for use in the company for

developed.
they
are
which

The benefit

of using synthetic

standards arises

since

individual
be
of
any
combination
standards
can
as
chosen
elements
to
that
variety
a
common
are
tasks
work

of manufacturing

processes

167

In order to develop a synthetic

and products.

system it

estimating

is necessary to identify,

measure a large number of work elements.


type and extent of work tasks included
in ensuring estimating

is essential

standards
define and

Definition

of the

in the work element

accuracy.

.ol

time standards data is presently

Synthetic
.,

designs.
to
Ltd.
H.
M.
cost
product
at
time for

estimate

development
of MTMV.

similar

to that

process

and cannot cope with

In addition

estimators

processes: and methods,


responsiveness

3.3.5

However,, the application

synthetic

Consequently
the large

time standards

is

is a slow

estimating

amounts of cost data

the NPD process.

during

required

using

being used

must'be

thus it

familiar

with

is more difficult

of H. M. Ltd. 's estimating

the manufacturing

to improve the

system.

Work Measurement using Computers


The initial

development of work measurement software

programs was directed

at using the fast

processing speed and

highlighted
the
to
advantages
achieve
computers
high accuracy of
by'Ke11er(126):

(i)

To increase the speed at which estimates can be


provided.

168

(ii)

To relieve

cost estimators

of the routine

part

of work measurement.
(iii)

To eliminate

humanerrors

that occur when large

amounts of data are handled manually.

Initially,

therefore,

software packages were developed

around conventional

work measurement systems.

Kopp(127) describes

the structure

improving time study synthesis.

For example,

of a computer system for


The computer is used to collect,

data,
in
time
the
development
study
rapid
of
and
process
store
Since the basic time study technique remains

time standards.

the
computer system examines only existing
unchanged
methods and like

manual time study synthesis

work

cannot estimate

development
for
the
new
product
process.
time standards

of time standards

Since the estimation

(e. g. PMTS) requires

techniques
work measurement
early

attempts

software

input
of
amounts

extensive
timed.

to develop

Software

was limited

be
estimated.
could

using

for

these

data to define

conventional

methods planning,

techniques

required

the work tasks

in the type of work'tasks

being

that

For example, Mason and Towne(128) developed

I
MTM
time
system
using
measurement
work
ARMAN,a computer
standards,

to measure the work content of complex electronic

tasks.
maintenance

169

The American "WOCOM"


production
includes

planning system(129l30)

both MTMI and Work Factor work measurement modules.

To reduce the amount of input data required


work, tasks the modules operate on two levels
"motion"

level

to define

complete

which are the

using input symbols that correspond to Work-

Factor or MTMI motions and the "task"

level

in which motion

and work place layouts for commonoperations

patterns

can be

from the computer memory.- The work measurement modules

retrieved

to a wide section of industry

are applicable
to different

computers including

and can be adapted

time sharing devices.

is increased,

Although the speed of estimating

through the

based
work
measurement
modules
of
computers,
on PMTSstandards
use
labour
time
standards within
estimate
cannot
batch

environments

manufacturing

is normally

description
methods
are not planned

in sufficient

since

the detail

too imprecise.
detail

job shop or small


of work
That is,

jobs

to enable MTM-I work

be
to
performed.
measurement

Steele
planning

(120)

within

recognising

low volume manufacturing

measurement software

this

be
suitable
would
system

estimate
resorts

systems, developed work

(based around MTM-2data) for measuring

and assembly work.

fitting

the imprecise nature of methods

Again however it

is unlikely

that

for use at H.M. Ltd. since to

time standards for non-repetitive


to a category estimating

technique.

work tasks Steele

(131)

170

Computer work measurement systems have gradually


since the late 1960's to become an integral
production

evolved

part of much wider

The range of manufacturing tasks

planning systems.

covered by such systems is diverse and includes manufacture of


(132)
(133)
(134,91)
tanks
storage
cardboard cartons
and
machining
,
The systems differ according to their accuracy, ability
to
lo,
optimize production plans and the level of subjective judgement
,

by the estimator

required

in order to develop time standards.

to optimize,

In terms of ability
developed

a software

planning

machining

operations

and metal

optimize

the machining

software

was limited

number of variables,
process
being

resulted

cutting

to optimizing

in a large

for

procedure

of turned

Figure

Halevi

and Stout(135)
specifying

in order

criteria

Although

components.
the "turning"

12, affecting

machines,
to

the

process

the

the optimization

amount (80K bytes)

of core storage

required.

(136)
A recent software system developed by Logan
succeeds
in reducing the level

interaction

required to

develop work plans and hence time standards.

The system,

LOCAM, consists
information
manufacturing
I

of operator

of a data base containing

all

the relevant

about a company's manufacturing facilities


logic

program which is designed to extract

and a
data

171

MACHININGTIME

FIGURE 12

PARAMETERRELATIONSHIPS

(135)

Ref.

94VE

OF

SURa6CE

PLAT

FlN1SH

POWER

DE p 1).4
CF Ni

XIOL

TYPE

CUT-17N.7.
FOPCE

E14T
i

-"

"".

.
CVrr wG
CVTTwG
FC; K--E

dr. s'E

; K--E

/\\

\NN.
OE7i+--

R<iF

OF

'CRSAOw

CW7(!

+G K'Simq

`'`, `

:/

F=ED

I
.

--

GEP'M

-------

OF

CUT

CU'

/I
IY.
Or
Cra. Cc:vG

//
CwU;

ER

T9lFR WCES
Of
LOCA'b+
Ga1Cxl, {

///

LE
OF

CUt
C7aOr.

Mn[ul\N

AE`. iU, aF,

cti'S

172

from the data base, manipulate it

and perform calculations

In essence LOCAMsimulates

with it.

the planning engineer would follow

the logic

process which

through in his mind.

LOCAM

has been designed for use by planning engineers and is therefore


unsuitable

for use by engineering

designers.

loe

Work Measurement Algorithms

3.3.6

of time standards can be performed by

The estimation
developing algorithms

that express the causal relationship

between manufacturing

tasks and the factors

that effect

Although the development of estimating

tasks.
difficult

and may require

following

benefits:

(i)

computers, their

The cost of estimating

these

models can be

use provides the

is reduced since estimates

can be developed quickly.


(ii)

The level

of subjective

judgement required

to

develop an estimate is reduced.


(iii)

The need to hold extensive manufacturing time data


files

Many factors
difficulty

involved

due to the difficulty


variables

is avoided.

affect

manufacturing

in constructing

estimating

in establishing

on the manufacturing

task times and the

times.

the effect

algorithms

arises

of individual

There are four basic

173

methods of constructing
empirically

estimating

based studies,

(i. e. synthetics,

algorithms

use of operational

techniques and the use of statistical

research

techniques)

each of

developing
is
of
models with acceptable accuracy
capable
which
levels.

Synthetic Techniques

3.3.6.1

The development of estimating


requires

models synthetically

the use of humanjudgement to alter

and the method of combining variables

until

both variables
an equation

is arrived

at that produces time standards of acceptable

accuracy.

The method therefore

on the skill

relies

judgement of the personnel constructing

A critical

in using subjective
equations.

estimating

the model.

examination of the model synthetically

derived by Mahmoud(137,138) illustrates


inherent

and

the problems

judgement to develop

Mahmouddeveloped a "universal"

21)
for
(Equation
estimating
model

the costs of turned

components.

C=
mnf

Bi Nd
.
60

Kc. Ll Dm
.
+
4.77 ,

St

BL.Qp

+k

(21)

174

Where:

Cmnf = Manufacturing cost of the component being


turned.
Bi

= Hourly labour rate

(including

Factory

Overheads).
/

Nd

= Componentcomplexity

Kc

K2.
Kmat.
K1.
=

K1

Machining

KZ

Machine type factor.

Kmat

Material

LI

length
Total
machined
=

Dm

(inches).
diameter
Mean
=

St

= Machine setting

time (mins).

B2.

= Machine tooling

capacity.

Qp

Batch
size.
=

k=

Material

Equation

21 is similar

factor.

factor.

floor-to-floor

(inches).

cost.

to the model used to estimate

H.
M.
Ltd.
costs
at
manufacturing
(Kc.
Li.
Dm
expression
Nd
+

4.77

factor.

(Equation
St

B2. Qp

time of a component.

14).

The

represents

the

175

The component complexity


by determining

factor,

Nd, is calculated

the number of discontinuities

surface and is,

according to Mahmoud,a measure of the

component design complexity.


"...
represents

on the machine

Mahmoudstates that Nd

the probable number of passes of the: tool


Therefore the more irregular

over the machined surface.


!

and interrupted
tool

the surface,

the greater

the number of

passes required to produce the component, i. e. it

is therefore

indicative

the component".

of the amount of work done on

However the method of calculating

(i. e. the types of discontinuities

Nd

considered, Table 24)

(Li.
Dm)and the type of
depending on the value of

varies

(e. g. bar, casting,

stock material

fabrication)

the

from.
is
manufactured
component

By varying the method of determining

Nd, the equation

to
designed
be
enable estimated component costs to fit
may
the observed costs rather

between component costs and their

relationship
factors.

than to express the exact causal

It is possible therefore

are not causally

related

that variables

to the manufacturing

be included in the estimating

influencing
which

costs can

model, hence poor estimating

accuracy can result.

This argument is supported by examination of other


variables

within

to be related

Equation 21.

to Nd.

That is,

For example, St is assumed


the number of surface

176

TABLE 24

CRITERIA FORESTABLISHINGNd
Ref..
Machining

1.

Number

2.

Unmachined

3.

Machined

of

set

Feature

Discontinuity
Rating

ups

1 for

parts

5.

surface
surface

4. " Threaded

Shoulders

if unmachined
before
threading

if machined
before
threading,
e. g.
turned
or drilled

before
if machined
threading,
e. g.
Drilled
and bored

s' 3

faces

or

if

1Y6-

if

>

1"
16

.
6.

7.

Chamfers

_
Radii

if

if>

if
-

(138)

0
1

1"

1"
00, ts

.x

1
8.

9.

Drilled

holes

Bored holes

10.

Reamed holes

11.

Recesses

1
-if

bored

- if

bored

- previously
- previously

if

if>

<

1"
1"
4

wide

2
and reamed

cored
drilled

or cast
or

bored

3
1
1

1
3

each

177

features

is related

to the number of tools

Although this

those features.

instances when a single

tool

design
features,
of
number
operation,

could apply there are


is often used to machine a

particularly

during the turning

hence St could be independent of Nd.

According
multiplying

used to form

to Mahmoud(138) St must be corrected

by Nd

by

Again however the methods of calculating

B2
the actual

value of Nd vary:
Bs

(i)

When Nd > Bz then Nd


B2
determine

the setting

Using this

=i

time correction

does not take into consideration

rule

to

factor

cases in which

single tools machine numerous surface features


and therefore

Nd must be less than one to


B2.

reduce the setting


(ii)

time to the correct

value.

WhenNd.< B2-then St is reduced by a factor


Nd

This rule does not consider components

Bz
that require

a different

tool

for each surface

feature.

The variable
tolerances
tightened

required.

Ki is dependent on the dimensional


If dimensional tolerances

Ki increases to reflect

the increase

are

178

in machining time experienced due to the need for


depths of cut,

smaller

increased number of machining

passes and use of slower machining speeds.


the use of a single
tolerances

to account for dimensional

variable

required is unsuitable

since KI for components

with various dimensional tolerances,


Of

is machined to different

feature

is difficult

causally

is clearly

times

machining
actual

in which each design

dimension tolerances,

to determine.

The problem involved


not truly

However

related
illustrated

times estimated

observed times.

with using variables

that

are

to component manufacturing
by Table 25 which compares
using

Equation

Poor estimate

21 and the

accuracy

is

indicated.

TABLE 25

A COMPARISON
OF ESTIMATEDAND OBSERVED
MACHINETIMES

Estimated Times
Equation

Observed Times
(H. M. Ltd. Time Study Data)

%
Error

297.2

290.7

+ 2.2

35.4

39.0

9.2

55.0

81.7

32.7
-

69.0

48.0

+ 43.8

45.7

53.8

15.1
-

261.9

61.4

+326.5

52.1

52.6

0.9

179

TABLE25'(Contd)
Estimated Times
(Equation

,,

Observed Times
(H. M. Ltd. Time Study Data)

%
Errors

115.8

78.0

+ 48.5

15.0

66.2

77.3
-

147.0

182.9

19.6
-

35.2

31.1

+ 13.2

25.7

33.0

22.1
-

9.5

18.0

47.2
-

102.8

90.0

+ 14.2

22.3

37.5

40.5
-

65.1

34.3

+ 89.8

14.8

8.0

+ 85.0

1.2

1.5,

20.0
-

2.5

12.0

79.2
-

4.7

14.0

66.4
-

2.3

4.2

45.2
-

19.7

13.5

+ 45.9

Estimating

Accuracy:

E=-O.

C8

S=0.55

Equation 21 could not be used to estimate New EHB


manufacturing

times since the variability

is
developed
not within
estimates
limits

of

,S=.

in the accuracy of

H.M.. Ltd. 's current accuracy

'O.314 (See Section 4.4.1).

180

3.3.6.2

Empirically

Derived Equations

derived estimating

Empirically

equations of the form

shown in Table 26 are developed under "laboratory"


That is,

conditions.

the variables

times are individually


influence
exact
r

manufacturing

therefore

controlled

of each variable

affecting

allowing

on the total

the

manufacturing

times to be determined.

The majority

derived equations estimate

of empirically

(e.
times
g. metal removal times for turning,
machine process
boring,

and drilling)

milling

since these operations

to a wide range of industries

applicable

and the factors

machine process times can be easily

that influence

are

varied

to
each other.
respect
with

However empirical
exact

conditions

for

although
to carry

out turning

inapplicable

conditions

since

cutting

cutting

is,

under the

to be maintained.

For example,

the time

use with

according

NC machine tools

speed facility

between the spindle

its

operations,

(NC) machine tools

Controlled

only

which they were derived.

22 is used to estimate

Equation

constant

models are valid

Numerically

to Moore(139)

are provided

enabling

required

with

optimum cutting

Hence a linear

relationship

speed component (used to calculate

speed) and Dl(Equation

23) does not apply.

181

The cost effectiveness


for estimating

of using empirical

is questionable

models

since the variables

used

to determine machine process times in Equations 22 to 26


are themselves dependent on a wide range of factors.

EQUATIONSFOR ESTIMATING METAL

TABLE 26

REMOVALTIMES

Turning (Manual Operation)

Ref.

(140)

(22)

tm = n. Dt. Lz. Vc-1. Fl-1

Where:

tm = Metal removal time (mins).


DI

Lz=

Diameter

of work surface

(mm).

Length of surface to be machined (mm).

Vc = Cutting

speed (m/min).

F1; = Feed rate (mm/rev).


Turning (Numerically

Controlled

Machines) Ref.

(139)

Lm = -ff.(D1-dt). (F1. Vc)-1. (D1. dt). (D1+d1)-1

Where:

di

The
turned component diameter (mm).
=

(23)

182

TABLE26 (Contd)

Ref.

Drilling

(141)

tm = L3. VS-1. Fj -1

(24)

Where:
Ls

hole
(m).
Length
of
=

VS = Spindle speed (rpm).


(Reciprocating

Face Milling
tm

(142)

Cutting Operations Ref.

-1 (V1 + V2-1 )
F2Lz.
63.
=
.

(25)

Where:
B3 = Width of surface to be machined (mm).
Fi

vi
V2

(mm/min).
Feed
Stroke
per
=
=

Average forward

Average return
(Roughing)

Gear Cutting

tm

=nt.

cutting
cutting

Ref.

M=

speed (m/min),

(143)

Mod.(Fw + 20). 600-1. F3. F. Fv. Fu. Fm

Where:
RE

speed (m/min) .

Number of teeth.

Module (mm).

(26)

183

(mm).

Fw = Face width
F3 = Tooth

umber coefficient.

F = Helix angle coefficient.


Fa = Coefficient

of the material.

of the strength

Fu = Machinability

for influence

Fm = Coefficient

12 illustrates

Figure
influencing

material

ships-between

increase

reduction

parameters

Since the relation-

machinability.

shown in_Fiqure

12 are

empirically

the estimator

tables

such as Table

to the use of reference


consequent

of machine size.

the process

the variables

to derive

difficult

coefficient.

in estimating

in estimating

must resort
27, with

speed and therefore

costs.

The accuracy of empirical

estimating

dependent on the degree of control

models is

exercised over machining

specifications
and quality.
(145)
found, during an examination
Bhattacharyya and Coates

conditions,

material

of machining conditions

used within

industry,

variation

in both the machining characteristics

materials

(both between and within

of

batches) and machining

(feed
used
rates and cutting
parameters
addition

a wide

speeds).

Bhattacharyya and Coates established

In

that in

124

SPEEDSAND FEEDS FOR DRILLING

TABLE 27

Ref.

OPERATIONS

cn)
O
`

[o

cJ

.p
O

"

rl

SI

I ( o
- -, --

In

Cl

CIC to
g
.o

,nt
=,.

O1

o
fi l

in

L'

I 1,
,
s

"0

.0

LIL

c_J

ln
.aT

:
\

Iv

\
n

--

,.

-5

;,

``
00 i
n

':

pi

IA\;

o
.s
G Ir

.o

(
o

o
of

.o

a;

sl

of

^'

-D

1-;

gl

GC

a!
O

,
o
.

! n^
,",
,Z

O
i

^?

gl

o!

r-

oI
iO

c~

cx ( V?

'
tc
r u; l

ol

o I
ip
-0
_

1 !o
N

`a

IO

Q !

cc ,

l! 1

OI

'
co'

a
+

a;,,

...
0

"r

,o

'

r"4'
'01

,
of 1

0:
Oi

rl

-$;

1
i0
-1)
rJ

a,
0'

!1"

;
"D

o
fr

,. r f

'

e:
C1:

,;

Ll ro
c
ti :
c

Vl ;

In

aI
O

of
GI

P-.

N'

;
.
ai ^- : i 00

ro o
''
N!
;,.

,
.
,
', '

4!

.0
L`1

C I

J:

V1

OI

" I
O

2i

-i

-l I

`D
lI

0!

:(

r)
O(

"

1 oi

_'

oj

Ci

:
(

4:

8i

O
Q

?i

Or

o}

nt

lrl

:Oi

0 1I

,
01
$;

-+ :0N
o f60
O

S,.;

t
!

40
I

M
L

- ----0

oc
o
o

ln) 1OI
O

o o n o a
i

o.

.
1
46

O;

04 "';
.

0N
OO

-.

si
Q

"

Ji

_'

r? t0!

!1 .
-4

Q;

1n;I s

a0

o!
c:
CI

N
^

V'

.a

ao

,, '0

? `'

,,,

04

o,
oi

Ca

to I

DI

O)
O

OI

W,

caI
N
O

ap

--

n1
0
p
r1 i

1O
I r:

O!

p1

01

o!

`'iC

Oc
o
9 (

tn

o)

10

C-4

o(

O' lI

cc

.0

l {
o

(An
cr1

O
'I
:.:.

ov

- od

; o

oi

In

tn

I
r-

ln

;'

_I

of

OI

1`)

a
f

"J
tn

to

LSv

I I'

'H

'

'

(144)

fi

185

60% of the samples studied the selection


parameters was controlled
indicate
empirical

of machining

by the operator.

that in the majority

These results

of companies the use of

modus to estimate manufacturing times could

be suspect since insufficient


the machining conditions

control

used on the shop floor

specification

is exercised over

However the increasing

of raw materials.

improve the

use of NCmachine tools would effectively


models for estimating

value of empirical

(i)

and the

since:

Machine process parameters can be closely


controlled.

(ii)

Equations

for

estimating

of machine processes
and tool

In addition
optimizing

manipulation

specific

the manual content

(i. e. loading,
tasks)

are not required.

equations can be developed for

machining operations,

e. g.

of the turning

(i)

Cost optimization

(ii)

Optimizing machining conditions


(147)
demand

(iii)

Selection

unloading

operation(146).
with sales

of the optimum machine tools

(148).

186

Operational

3.3.6.3

Research

The operational

research techniques that are suitable

for developing estimating

(i)

Simulation. - Manual simulation


the synthetic

,or

equations are:

is idhntical

to

development of estimating

equations in that both variables


of combining variables

and method

are altered

until

an

equation is obtained that produces accurate


time standard estimates.
The use of a computer to carry out the simulation
improves the accuracy of the

technique greatly

models developed since a larger


variables,

variable

combinations and actual

time standard observations


the estimating
(ii)

number of

(used to construct

model) can be considered.

Queueing Theory - Queueing Theory has been used


(122)
to examine the waiting times
by the author
of welders

using

overhead

heavy fabrications.
the indirect
welding

costs

cranes to manipulate

The data obtained


of the handling

to be quantified.

allowed

content

of

187

(iii)

Linear Programming - Observations of


manufacturing tasks can be formulated

into

a series of simultaneous equations (constraints)


which when solved using linear
yield

programming (LP)

between manufacturing

the relationship

and manufacturing times.

variables
11,

Ballard and Hichkad(149) and Oliver

(150)

formulated LP models of the form shown in


Equation 27:
=n
Aq. Xqj

Bi

E2

.-

= Yi

(27)

Where:
Aq

Estimated

Frequency count of the qth work


element in the jth

Eij; Ezj

to. perform

of the qth work element.

one unit

Xqj

time coefficient

operation.

= Error elements which represent the


amount of time per unit of the jth
operation
the total

requiring

an adjustment in

time for that operation,

i. e. these elements are the times


either

unaccounted for or in excess

and result

from the chosen set. of

work elements.
Y

= Total standard time per unit of the


nth operation.

188

To solve Aq > 0, Xqi > 0, Ei


J>0,

Ezi > 0, Yi >0

The objective

of the LP program involved the

minimization

of the sum of the error elements


is then a set of time

E1j and E2j.

The result

coefficients,

Aq's, for each work task.

The benefits

of using LP to construct

estimating

equations are:
(i)

data can be introduced

Managementcontrol

into the model in the form of extra constraints


that influence
(ii)

Negative

Multiple

3.3.6.4

regression

extension of the "line


simple linear
establish

of leas

regression,

the relationship

dependent
variable.
a
and

Y=

Aq's are avoided.

is essentially

an

squares method", i. e.
technique used to

between an independent variable


Multiple

on a dependent variable

of the following

(151)

a statistical

however, determines the effect


variable

for

coefficients

Regression

linear

Multiple

the time coefficients.

linear

regression

(MLR),

independent
than
one
of more
using a first

order equation

type as a model, i. e.:

X0 + x1Zl

+ X2Z2 +

... .

XnZn

(28)

189

Where:
Y=

The value of the dependent variable.

X0'=A

constant equivalent
on the y-axis

to the intercept
line

of a straight

graph.
or

Xi, X2... Xn = Coefficients


variables

of the independent

Zl, Z2... Zn respectively


the individual

and represent

effect

that each independent variable

has

on the dependent variable.

When used to develop


technique

regression

minimizes

the dependent variable,

Figure

13.

and actual

values

Estimates

of

can

be derived from the equation of the "least

squares" line through actual

observations.

technique guarantees that no other line


actual observations

The regression

through the

of the dependent variable

have a smaller sum of squares as illustrated


It therefore

yields

the coefficients
to predict

the

equations,

the sum of the squares'of

between estimated

the differences

therefore

estimating

an unbiased and efficient

of the independent variables

the value of the dependent variable,

the values of the independent variables.

would
in Figure 14.
estimate of
with which
given

190

FIGURE13

METHOD
OF LEASTSQUARES
Ref.

(114)

Dependent
variable
berm

Constant

Vzrahla

m
1

si -minimum
i-1

FIGURE14

--m

OF AVERAGE
AND REGRESSION
COMPARISON
LINES
Ref.
14

(114)

RANGEOF DATA

12

e
1$

10

'=
1 ..,;

ix

2I
0

.-

CaE'"i'
?:,
' _.

:
i

4y

:]E!
I-c. Y. S x Ma

11

191

Although a linear
without

model is produced (with or

a constant term Xo) it

can be used to predict

values for a dependent variable


the exact functional
variable

when it

either

expresses

between a dependent

relationship

or when it

and independent variables

is an

acceptably accurate approximation of a more complex


Ile

e. g. a non-linear

relationship,

Using the multiple

linear

relationship.

regression technique

be
from
data
to
used
an existing
enables
it

system (while

operates) to develop the regression model.

This therefore

data
from
designed
to
the
generate
need
experiments
removes
in which independent variables

have to be controlled

to determine the influence

individually

of each on the

Whendeveloping regression models

dependent variable;
for estimating

production

direct
from
obtained

time standards,

observation

if
estimates
or previous

data can be

accurate.

of the regression

depends on the assumption

of "cause and effect",

the independent

chosen actually

variable

distinguish
cannot
occurring

variables

being examined.
between a natural

by chance.

tasks

of manufacturing

sufficiently

However the effectiveness

dependent

-'

technique
i. e. that

influence

the

The regression

model

relationship

or one

192

Crocker

(152)

observed, "the ease with which noncan be found is surpassed only by

relationships

existent

the ease of missing important


essential

is,

It

ones".

to choose independent variables

therefore,

with care.

The use of personnel familiar

with the work being studied

can aid in the identification

of independent variables.

MLR in Cost Estimating


The regression
in the field
20 years,

technique has been known to be applicable

of production

time standards estimation

although the applications

been used, though diverse,

for which the method has

have been limited

in highlighting

critical

the development of estimating

Canning

(153)

factors

is

that influence

models.

used the regression

models for estimating

in number.

literature

Nevertheless an examination of available


useful

for over

technique to establish

the man-hour requirements for a jobbing

transformer
assembling
shop

cores.

Regression models that

degree
of accuracy were developed for
acceptable
gave an
individual

sections of the job shop and for each product


The work undertaken by Canning

made within

a section.

illustrates

how the accuracy of estimating

improved.

By establishing

models can be

models for each section of

the job shop and for each product type made within
section

both the range of variablity

of assembly times

193

and the number of possible variables


times were reduced.
variability
error

assembly

affecting

Reducing the assembly time


lowers the maximumpossible

effectively,

obtained from the estimating

model and therefore

improves estimate accuracy.


i

In addition

if

the actual

dependent and independent

variables

models may be used to

of estimating

Canning effectively
that influenced

between

relationship

is non-linear,

a number

improve estimate

in Figure 15.

illustrated
as
accuracy,

factors

causal

reduced the number of variables

assembly times by removing the effects

such as:

(i)

Variation

in production

methods between

stations.
(ii)

Variation

in production methods between

products.
(iii)

Variation

in productivity

levels

between

in productivity

levels

between

stations.
(iv)

Variation
products.

of

194

FIGURE15

EFFECTOF INCREASING
OF REGRESSION
THE NUMBER
MODELS
FORA SPECIFIC DATARANGE
Ref.

(154)

...
C)
-0
t0
S..
tC

C)
. O,

C'

C)!
C)

4O
C)

10.

Independent Variable Values, (x, x2, x3, n)


A

Bi , B2, B3

data
for
the regression model
complete
range,
max. error = e2
the regression
max. error

models for sections


= el

<
el

e2

of the data range,

195

(155)
Cochran
states the limit

to reducing data

in this manner is the point where the errors

variability

obtained from using the average value of the dependent


are acceptable and the use of a regression model

variable

improve this

would not specifically


i

In practical

accuracy.

terms, however, the number of models developed would be a


managementdecision

since the greater the number of models

developed the higher the application

cost.

Hence application

cost would need to be optimized against estimate-accuracy


in determining

the number of models-to develop.

Salem(156) determined the indirect


required to load pallets
variables;

predictor

labour time

with packed material

using the

number of cases packed per pallet,

number of orders per pallet

load, weight of packed material

load and the volume of cases per pallet

per pallet

load.

It was found that the number of cases packed and number of


orders per pallet
whilst

other-

the model.
residual

load were the best measures of job time,

predictor

variables

The criterion

standard deviation

model should be minimized.,


existed

did little

for choosing predictors


resulting

was that

from the regression

A high inter-correlation

between the two remaining variables

of cases and number of orders per pallet


bought into question the need for
the model.

to enhance

(i. e. 'number

load) which

both variables

Therefore Salem used Student t-tests

determine which of the two variables

within
to

should be dropped

196

These tests

from the equation.

indicated

of the number of cases packed per pallet

coefficient

had the best statistical

and in addition

precision

gave

of packing times per pallet.

an adequate estimation

Doney and Gelb(157) used the multiple

,or

that the

linear

regression technique to estimate the cycle times of


turned and bored components on an automatic lathe.
Componentdesign features
on cycle

effect

finished

times were stock

factor,

machinability
turned

selected as having the greatest

outer

overall

improved the regression

length

diameter,

Several

bore diameter.

outer

diameter,

of the component,

hub outer

variables

material

diameter

when transformed

and
however

model:

Variable

1=

Stock outer diameter xn


Material Factor

Variable

2=

Stock OD - Finished
2x Material

Outer diameter
Factor

(29)

(30)

Although the model developed gave low percentage


estimating

errors

(+2.3% to -2.6%) the variability

times used to construct

in cycle

the model was narrow (1.2 minutes

to 2.4 minutes).

Regression equations can only be assumed

to be applicable

the data range from which they were

constructed.

within

Hence-the regression model developed by

197

Doney and Gelb would be limited


It is essential

therefore

to a narrow data range.

when developing regression models

to provide actual data points to limits


all

work likely

that would include

to be encountered.

An important advantage of using MLRto develop


r

estimating

models is the wide range of manufacturing

tasks for which the technique is applicable.

For

MLRhas been. used to develop models for estimating

example,
time

standards for cost control and scheduling of Flexowriters


(158),
in a sales order office
setting incentive payments
for

vehicle

developing

deliverers

tooling
estimating
costs
(162)
gross time standards
and for estimating

assembly and fitting

cranes

(159),

times

for

heavy,

(160,161)

custom-designed

(154)
.
(136)
Logan
recognised the cost-effectiveness

regression

based estimating

systems.

of using

Regression analysis

was used to reduce thousands of individual

time standards

to simple formulae hence improving the estimating


ation time, removing the high level of skill

applic-

and background'

experience required to develop time standards and therefore


easing the problem of training

Software for carrying

new estimators.

out the MLR solution

technique

now forms part of several work measurement computer systems,


e. g. WOCOM(129),LOCAM(136)and REAP(91) where the

198

technique's

ability

to estimating

3.4

Selection

new product

(i)

in jobbing shop and low volume

systems.

of Work Measurement Method for Design Costing at H.M. Ltd.

When selecting
developing

time standards

models ensures that these systems are

commercially viable
production

to reduce individual

work measurement system for

the most appropriate

an estimating
development

system capable
process,

during

the

methods can be ignored:

several

since this

Time Study

designs

of costing

time

method cannot predict

standards.
(ii)

PMTS

since

these time standards

shown to be suitable
manufacturing
appropriate
facturing

(iii)

Comparative

Estimating

have been

in high volume

systems and are not


to the small-batch

manu-

of H. M. Ltd.

environment

since this

method cannot estimate

accurately

the individual

manufacturing

costs of components.
Synthetic

(v)

since these are difficult

Algorithms

prone to high inaccuracies.

Operational

normally operational

to develop and

research techniques

Research

are used for optimization

Algorithms

their

problems, hence

develop
to
estimating
use

equations

199

(v)

Contd

...

i. e. the computing

could prove costly,

cost to determine optimization

solutions

is in excess of that required to find the


between dependent and

relationships

independent variables.

systems used in low volume manufacturing

Estimating
normally
task

times)

process

Therefore,

times).

of using

suitability

standards

be
cost
not
would

involves

work

the

models or synthetic

of the present

and cannot handle efficiently

of the New EHB development

time

Synthetic

algorithms.

to increase

machine

a suitable

determining

regression

manual

estimating

models are the basis

estimating

effective

the present level,

design

derived

and empirically

requirements

costing

estimating

(for

algorithms

multiple

system at H. M. Ltd.

estimating

(for

the problem of selecting

costing

either

empirical
and

standards

derived

and empirically

for
method
measurement

time

time standards

employ both synthetic

systems

project.

estimating

the
Since it

resources

over

these work measurement methods must also be

discounted.

When examined with respect to the estimating

system requirements of

the new product design process, the development of equations using


regression

(i)

analysis

i. e.:

appears possible,

Accuracy
regression

level
The
-

of accuracy

using

models can be improved by reducing

variability,

i. e. increasing

developed.

Accuracy

since

obtained

the number of equations

therefore

the number of equations

against system operating

data

cost.

is a management decision
developed

must be optimized

200

The present work develops estimating


the accuracy limits

(ii)

system (Section 4.4.1).

Setting-up

Cost-

time

The cost

can be expected

of developing

and carry

out a costly

cost

regression

to be low since

study data can be used, i. e. it

to design

(iii)

of H.M. Ltd. 's existing

estimating

equations

equations within

existing

is unnecessary

time study programme.

Operating Costs - The cost of using such a system would


be low since time standards retrieval
data files

from extensive

would be unnecessary enabling estimates to

be prepared quickly.
(iv)

Responsiveness

The
responsiveness
-

be dependent on the personnel


The present

estimating.

who carry

reducing

out the

work establishes

by
design
use
of
system capable
the effect

of the system would

engineers,

of queueing time

an estimating
therefore

on system

responsiveness.

(v)

Type and Detail

of Cost Data - Previous literature

diversity
the
shown

has

of manufacturing tasks which have

been estimated using regression models.


can be expected that regression

Therefore it

equations could be

developed for the wide range of processes used at


H. M. Ltd. and at various

levels

of cost detail.

201

(vi)

Subjective

Judgement - Since design engineers are to use

the estimating

system, the level of subjective

degree
of familiarity
and consequently

judgement

with manufacturing

methods and processes required to develop time standards,


must be minimized.
types of predictor

This factor
variables

quantitatively

evaluating

is dependent on the

chosen and methods of

these variables.

202

CHAPTER
4

COST ESTIMATING SYSTEMDEVELOPMENT

203

4.0

Introduction
The development of a cost estimating

analysis

to develop estimating

equations

(i)

equations necessitates

tasks to which each model relates.

manufacturing
to identify

relevant
jare

defining

for which specific

Factors to consider are:

valid.

is essential

that

equations

equations can be accessed


can be located

both the wide range of manufacturing

processes

at H. M. Ltd.

estimating

levels

the

This allows estimators

equations and the conditions

The ease with which estimating


It

and the need to achieve

S=0.314,

a large

developing

levels

the current

within

(i. e. E=0.096,

each design

see Section

estimating

of the small

retrieval

operating

number of equations

used to code or classify

(ii)

system that uses. regression

The number of predictor

out

carried

accuracy

at the company

4.4.1)
(>200).

requires
The system
the

models must allow

number of equations

to be carried

since

quickly

required

to cost

out quickly.

variables

within

each estimating

equation

This number should be minimized where possible to reduce the


effect

of interactions

to develop estimating

between variables

models of greater accuracy.

reducing the number of variables


effectively
total

and therefore

per estimating

Hence it

However,

equation

increases the number of equations within

system.

help

the

is necessary to balance estimating

accuracy with system size.

204

(iii)

The type of factors

chosen as predictor

Since the estimating


engineering

system is being developed for use by

designers it

is necessary to minimize the effect


of those factors

on manufacturing cost estimation


data is inaccessible

variables

Such factors

to designers.

for which
would

(e. g. machining feeds,

include machine process variables

speeds and depths of cut) since the generation of this


type of data lies
training
(iv)

The level

outside the normal experience and


designers.

of engineering
of. detail

by the estimating

of cost data that

can be provided

system

The system needs to estimate the cost of machining individual


in order to allow designers to carry out

design features
detailed

4.1

cost comparisons of alternative

Classification
Initially

feasibility

of Estimating Models

an examination was carried

of using existing

out to determine the

component coding systems to classify

Coding systems. such as VUOS0(163), Opitz(164)

cost centres.
Brisch(165)

designs.

enable design features

could possibly

and

to be coded and therefore


equations to be quickly

enable estimating

referenced.

However existing
the rapid

retrieval

components with

systems are primarily


of drawings

respect

concerned with

either

groups of similar
(166,167).
requirements

or selecting

to production

been
have
not
they
specifically
such

designed

for

As

cost estimating

205

Since GombinskV(165) points out that the use of

purposes.

systems of classification

for purposes other than those

intended is not always possible


existing

it

systems would be suitable

appears unlikely

that

for costing purposes.

This assumption is confirmed by work carried


(1
and Blore91).
and Schilperoort68)

out by Remmerswaal

Remmerswaaland Schilperoort
systems in terms of their
information

evaluated component coding


to provide the type of

ability

required to enable various departments to function.

They found that although classification


information

systems provided

for planning and scheduling,

the manufacturing

time standards used were assumed to be known. Blore(91)


coding systems during the

abandoned the use of existing

development of the REAPsystem (Rapid Estimating

During the present work the classification


models required the exclusion

of variables

and Planning).

of estimating

considered inaccessible

to designers since this would then-overcome the need to request


data from other functional

departments and therefore

response of the estimating

system.

to be that which is involved

Accessible data is considered

in the design decision making

process, e. g. component dimensions, design features,


materials

improve the

and the manufacturing

operations

required.

tolerances,
Inaccessible

data is associated with the manufacturing input to the estimating


task,

e. g. machining speeds, feeds and manual manipulation

tasks.

206-

Examination of relevant

machining literature

indicates

is caused by:

that process variability

(i)

(169)

The process being carried

out, e. g. machining,

welding and flame cutting.


(ii)

The operation

being performed, e. g. turning,

boring

and milling.
(iii)

The size and power of the machine used.

(iv)

The manual work content involved.

(v)

The tolerances

(vi)

The material

4.1.1

and surface finish

required.

being processed.

Manufacturing

Processes and Operations

The manufacturing
have been identified

processes carried

and classified

out at H.M. Ltd.

for ease of access into

four main groups:

(i)

Pre-machining - processes used to prepare material


for subsequent machining operations.
are

sawing, shearing,

flame cutting

Examples
and

fabrication.
(ii)

General Machining - those processes, such as


turning,

boring,

drilling

that can be carried

and thread cutting,

out on a wide range of

machine types, e. g. capstan lathes,

centre

207

lathes,

bar lathes,

numerically

horizontal

borers,

lathes and vertical

controlled

borers.
(iii)

Special

Machining

carried

out on specially

tools.

Examples are gear cutting,


drilling,

radial
grinding
(iv)

those
processes
-

Post Machining

are

designed machine

spline

and surface

that

cutting,

milling,
cylindrical

grinding.

- processes

such as painting

and assembly.

Estimating
carried
for

equations

out within

a process

which estimating

Table 28.

were developed

This table

also

were developed

includes

cost items for which estimating

4.1.2

each operation

The process

group.

equations

for

other

and operations

are listed

in

manufacturing

equations were developed.

Size and Power of Machines

At H.M. Ltd. "General Machining" and gear generating


operations
capacity

represent approximately
and are carried

The variation

out on a range of machine sizes.

of power input of machines results

range of process variables

In order to effectively
process variables
divide

70% of the machine shop

in a broad

being used.

remove the influence

on manufacturing

costs it

of machine

is necessary to

the range of general machining and gear cutting

208

TABLE28

CLASSIFICATIONOF COSTINGEQUATIONS

PRE-MACHINING

SPECIAL MACHINING

Flame Cut

Gear Cut

Saw

Mill

Shear

Cylindrical

Weld

Surface Grind

Grind

Spline
GENERALMACHINING

Radial Drill

Turn
POSTMACHINING
Face
Drill

Assembly

Bore
OTHERCOSTS

Ream
Chamfer

Milling

Fixtures

Radius
Drill

Jigs (Normal)

Drill

Jigs (Box Type)

Groove
Circlip

Groove

Copy Templates
Thread
Plate Gauges
Tap
Plug Gauges
Rope Groove
Recess
MATERIALCOSTS
MANUALTASKS

Rigid

Steel

Joists

Rigid

Steel

Girders

Load/Unload Machine
EN8 Round Bar

Machine Manipulation
EN8 Plate

Machine Set-up
Hollow

Bored Bar

209

machines into groups.

a group does not have a significant

within

variability

on manufacturing

effect

In this way the machine process

times.

Machines that performed general machining operations


were therefore

grouped (according to machine size and power

by machine type,

availability)

i. e.

Bar Lathes
Bar Capstans
Chuck Lathes
Chuck Capstans
Centre Lathes
Borers

Horizontal
Vertical

Borers

Copy Lathes

NC Bar Lathes
NC Chuck Lathes

Within each group of machines there is insufficient


variation

in machine size and power to significantly

the accuracy of cost estimates.

These factors

influence

can therefore,

be ignored.

Although

gear cutting

and type an examination


equipment

equipment varied

revealed

that

by both size

the particular

dependent
is
is
on which a component
machined

on the number and module of gear teeth.

210

Estimating equations could therefore


for specific

be developed

size ranges of spur and pinion gears

(Appendices IV and V) enabling the variability


process factors

to be effectively

to
classify
not necessary

in machine
It was thus

removed.

the gear cutting

equipment used

at H.M. Ltd.
r

In addition

the material

type from which the gear or

found
be
to
was
related
was
manufactured
pinion
module so that a Material

Factor equation

to the

(see Section 4.1.5)

was not required.

4.1.3

Manual Tasks
The model developed by Mahmoud(138) does not estimate

separately

the manual content involved

in machine processing.

The usefulness of this model is therefore

limited.

For

example, the manual work content removed by process


be
determined.
cannot
automation

The present work therefore

was designed to enable manual task times to be estimated


independently of machine process times.

Halevi
independently

and Stout('35)

determined

of machine processing

of machine process
approach unsuitable
dependent on factors

time.
since
other

manual task times

time but as a proportion

The present
manual tasks

author
times

than the machine

considers
could
process

this

be
time.

211

Tool manipulation

times for example, are often dependent

on the distance moved between machining operations

rather

than on the time spent machining.

manual tasks a detailed

Whenclassifying

such as illustrated

of each operation,

considered inappropriate

in Table 29, was

since a knowledge of methods

be
then
required for their
would
planning
adopted by Blore(91)
their

main objectives

manipulation

breakdown

The method

use.

of grouping manual tasks according to


(i. e. machine set-up tasks, machine

tasks and machine loading and unloading tasks)

has been used in the present work since sufficient


design
for
be
costing
provided
could

Although Knott(92)
further
a
as
of components
this
work
present
integral
level

purposes.

regards"inter-process

measurement"

manual task objective,

for the

form of measurement is regarded as an

part of machine manipulation` tasks.

of quality

detail

control

checks required

The low

during the

H.
I.
I.
Ltd.
at
components
also allowed these
of
machining
be
included
in
to
the
tasks
machine manipulation
measurement
tasks without

loss of estimating

The manual content involved

accuracy.

in gear cutting

less
5%
than
the
total
of
represented
was therefore

operation

operations
time and

included in the machine process task time

since a significant

loss of accuracy would not be incurred.

212

TASK TIMES FORA DEVLIEG


MANIPULATION

TABLE 29

SPIRAMATICJIGI4IL
Ref.
List
S.

of

B. Ma.

Elements

Load and Unload


Load/Unload
Component to Table
".

50 lb to
10 cwt

Up to
50 lb

35
.

Machine Manirulation
Fit/Remove Tool. to Flash Change
Fit/Remove Tool to Quick Change*
to Cut
Position/Retract
Index. Table
Change Feed or Speed
Position
to Axis (Tape or Manual).
Apply Tapping Grease
Deburr by Hand

Clear

Spindle
Workplace

Move In
of

Cut.

Cuttings

Out of Bore
Blow Cuttings
Set Depth of Cut'
Set 'X' Axis to Zero
Set 'Y' Axis to Zero
Gau
Micrometer

Internal
External
-

Rule
Depth Vernier
Caliper Vernier
Plug Gauge
Internal
Caliper
External
Screw Gauge - Plug
Air Gauge - Plug

B. Ms.

Basic Minutes

Over
10 cwt

2.43

(No
Fixture
True)
Table
Position
or
on
(True
Table
by Clock).
Position
on
(True
by Scribe Line).
Table
Position
on
Gland or Clamp (1'Nut or. Screw)
Tighten/Release

Retract

(144-)

...

40
.
2.33
1.62
58
.

25
.
20
.
28
.
13
.
16
.

52
.
14
.

"13

59
.
26
.

14
.
29
.
9.49

3.49

32
.

31
.

20
.

45
.

41
.
30
.
32
.

35
.

46
.

213

However, the number of estimating

equations within

system would be reduced hence lowering operating


and increasing

the
costs

the speed at which estimates could be

prepared.

ComponentTolerances and Surface Finish

4.1.4

equations were developed for

Separate estimating
those machining operations

with dimensional

<+ 0. lmm and those operations

tolerances

of >+ 0.1mm.

with tolerances

This took account of the variation

in machining speeds,

feeds and depths of cut caused by differences


tolerance

a surface finish

average and dimensional tolerances

not influence

the variability

since these tolerances

factors

operation.

grinding

In this

+0.20mmwith tolerances
to be left

of <2. Oumcenter

of <+ 0.05mm did

in General Machining process


were achieved by a subsequent
case a grinding

on the allowance of s+ 0. lmm had

Processed

Obviously machining times are influenced


with which a material
material

allowance of

on the work during the General Machining operations.

Material

4.1.5

in dimensional

requirements.

Components requiring
line

of

is termed its

can be machined.
"machinability".

can be used to quantitatively


machinability.

by the ease

This property

of a

Various criteria

measure a materials

(170,144)

Of the commoncriteria

the metal removal rate per horsepower is

machinability

for use during the present work.

suitable
indicates

the relative

materials

and can therefore

This criterion

metal removal rates for different


be related

to machining times

The relative

and hence manufacturing costs.


'Of

used to define

rates of

metal removal are determined under constant power


These conditions

conditions.

exist

during the present

work due to the method'of developing estimating


for specific

machine sizes.

Using metal

removal rate

this

to their

relative

materials

machinability

is not influenced

classification

during the present work.


of a materials

data obtained

were classified
(Table 30).

Since

by the specific

of machine shops it can be used

manufacturing conditions

effect

per horsepower

Group(144)

from the P-E Consulting


according

equations

To take into consideration

machinability

the

on machine process

times, the machining times estimated using the regression


factored
be
must
equations
equations listed

4.2

Factor

in Appendices IV and V.

Data Collection
The collection

initially

using the Material

required

and a suitable

of data for developing the estimating


predictor

(independent) variables

data source identified.

equations

to be chosen

215

MATERIALFACTOR(MACHINABILITY)GROUPS

TABLE 30

Ref.

(144)

Material
Factor Group

MF1

EN9

EN18

EN36C

EN43B

EN24T

EN15

EN19

EN36T

EN43C

EN5B

EN16

EN19C

EN39A

EN45

EN16J

EN23

EN39B

EN56A

EN16S

EN24

EN42

EN56B

EN16T

EN36A

EN43

EN58

EN16U

EN36B

EN43A

EN58A

High carbon, Alloy and Stainless

MF2

EN6

EN8A

EN14A

EN20B

EN6A

EN8B

EN16B

EN22

EN7

EN8D

EN16H

EN33

EN8

ENGE

EN19A

EN34

SAE 8620H

Medium Carbon Steels

MF3

Cast Iron
Cast Steel

MF4

Steels

ENIA

EN3A

EN32A

EN2

EN3B

EN32B

E7112A

EN3C

EN352

EN2C

EN8t4

EN355

EN2D

EN32

216

TABLE 30 (Cont'd)

Material
Factor Group

MF4
Cont'd

Free Cutting Low Carbon Steels


loe

Low Carbon Steels


Nodular Iron

MF5

PB1

LMG

LM9

PB2

LM4M

LM9M

PB3

LM4WP

Phosphor Bronze,
Gun Metal,
Free Cutting

Leaded Bronze

Hard Brass,

Soft

Brass

Aluminium Alloy Castings


Plastic

Materials

Brass

217

4.2.1

Selection

of Predictor

Selecting
the regression

predictor

Variables

variables

is a critical

procedure since the variables

part of

chosen

(Appendix IV) are assumed to have an actual causal effect


on the-dependent

Although this

variable.

often be proven by empirical


laboratory

evidence obtained from

studies of manufacturing

Equations 22 to 26), many situations


is not possible.
manipulation

assumption can

processes (e. g.
exist

Examples are, welding,

tasks and tooling

costs.

where this

machine
Hence care must

be taken when choosing independent variables.


the present work predictor
reference to personnel,

variables

During

were identified

with

such as foremen and production

engineers, who were familiar

with the manufacturing tasks

being examined.

Estimating

accuracy and system complexity were

improved as follows:

(i)

Various types and combinations of predictor


variables

were examined to determine the most

accurate estimating
variable.

model for each dependent

For example, it was found that during

the development of estimating


boring operations

equations for

that the use of length of bore

and diameter of bore as separate predictor


variables

gave more accurate results

the product of these two dimensions.

than using

218

(ii)

When estimating
levels

models gave unacceptable error

both the independent variables

and the

data source used to develop the models were


re-examined for suitability
dependent variable.

in estimating

the

For example, the use of

H.M. Ltd. 's standard welding time data was


.-Ir

found unsuitable

and other more reliable

(see
had
be
data
found
Section
to
of
sources
4.2.2).
(iii)

When small differences


levels

existed

between alternative

in accuracy

equations,

consider-

ation was given to the standardization


predictor

variables

of

between equations and

hence the simplification

For

of the system.

example, a number of estimating


developed for estimating

equations were

the times to turn


i. e.

components on a range of machine types,


bar lathes,

bar capstans, chuck lathes,

chuck

capstans, NC lathes and centre lathes.


small proportion

For a

of machine types using the

dimensions
"turned length"
the
of
product
"depth of cut" as a predictor
the greatest

estimating

"turned length"

accuracy.

gave

However,

and "depth of cut" were used

as separate predictor
not seriously

variable

and

variables

impair estimating

allowed the predictor

variables

times to be standardized

since this

did

accuracy but
for turning

between all

machine types.

219

4.2.2

Data Source
A library

of measured time study data was. held by the


department.

engineering

production
by carrying

This data was collected

out observations

on operators with rating


(113).
This
80
the
0/100
on
scale
of
rating
performances
0/100
is
than
the
the
average
slower
performance
on
pace
scale,

i. e. the average rate at which a large group of

qualified

and motivated workers could be expected to

achieve.

However it

operators

can be expected to work under -the manufacturing

conditions

at H.M. Ltd.
result

manufacturing)
become familiar
jobs.

represents

the standard rate at which

These conditions

in operators

with the production

Hence overall

(low volume

frequently

having to

requirements of new

work performance is lowered.

In order to represent current

operating

conditions,

data to be suitable

for the present


(113)
(CRT)
allowance
compensating
rest
study, a
was added
and hence allow this

manual task times.

This allowance represents the

non-working time required

to take account of such factors

to all

as

operator

fatigue

and personal needs.

The standard

CRTallowance used at H.M. Ltd. was 17.5%.

Manufacturing
been collected

tasks for which time study data had not

by the production

dealt
with in the following
were

engineering
manner:

department

220

(i)

Sawing, Spline Cutting and Shearing Operations


The standard production
shop floor

for control

times issued to the


purposes at H.M. Ltd. are

termed Sales Target Times (STT's) and have been


estimated using synthetic

work measurement

techniques.

These times were used to develop

estimating

equations for sawing, spline cutting

of

and shearing operations

since:

(a)

control

Lack of effective
recording

by shop floor

of job time
personnel resulted'

in recorded job times having little


comparison with the actual times taken
to perform tasks.
(b)

The STT values used were initially


examined by production

engineering

personnel and considered appropriate.


(ii)

Numerically

Controlled

An existing

procedure employed by the estimating

Profile

Flame Cutting

department (Table 31) for developing flame cutting


times was adapted (Appendix IV) for use within
estimating

system.

of estimating
and still

The adaption involved no loss

accuracy over the existing

allowed the individual

in flame cutting

tie

method

tasks involved

to be costed separately.

221

TABLE 31

TIME STANDARD
ESTIMATIONFOR
NC FLAMECUTTINGAT H.M. LTD.

OPDE2.
No.

TA?E No
Nd.

..

SETuP.

A*

No

TAPES
aF

No

CP NOZZLE SET?tNGS
.

X-

00
S.

shtis =

3.00 sds =

_x

PoEtL

C.

C ST= -1

FZ.
LOAD ATE

S.OOsts -

L'uNtv-.
ALJ
-:
D

C
TutCKNESS
mm]
FLATS

No OF PIZEHEATS

=E

&U

UN1L0i1D COWONENTS

&tt?NtNG

VALUES.

Tl-iK.
ZNtM

LMn-t/nuN

C55

ID
12
15

560

x*0.10
CV,
x )

510
405

SM VALUE

SET UP
A+B

2.080

3.855

305

3"B55

32

255

4. co2b

35

255-

4.626

50

2o-3

5.72

75.

i50

7.7t0

SM's =

VJofZK VALUES.

1 850
.

2.313
2 "Eeo

VS=

2.
Y. oo srns=

305
. 305

2-2.
25

x CDI__,

CLCIM.OF PP.
CMl
rFtLE1

I9

x Is.Go

VALUE ISSUED =

RZOFILE

3. $S5

C+E+F+G=
VAL.'UE

Standard Minutes

tSSuED

222

is used to prepare edges for

When flame cutting

welding then the time estimated

from
the
shape
component
of

flame cutting

the
doubled
to
be
obtain
must

stock material
total

(iii)

for initial

time required.

Welding
The method of estimating

labour time standards

(STT's) for welding tasks at HJI. Ltd. involved


the total

calculating

joint

rate factor.

and applying a single


influence

of factors

welding position,

length to be welded
Since the

such as, type of joint,

electrode

wire diameter,

thickness was not

handling time and plate


taken into consideration,

the use of STT times

for developing regression

equations was considered

That is,

inappropriate.
inconsistent

STT times would be

with the actual work content


Examination of welding data confirmed

involved.
that similar

fabrications

had widely varying

STT's issued (up to 200% difference).


The approach adopted by the present author
initially

involved developing models to determine


between the "basic arc time
(171)
issued by the Welding Institute

the relationship
standards"

and weld design factors


Notes, Appendix IV).

(see Variable Description


Since the "basic arc time"

223

for a weld is that amount of time the arc is


burning,

actually

it

is necessary to add on a

time allowance for such tasks performed by the


welder as weld joint
the fabrication

preparation,

to gain access to the weld joint

and adjustment of protective


O

manipulating

In

clothing.

since the standards issued by the

addition

are "basic times" they must


(113)
be converted to "Standard Times"
by adding

Welding Institute

appropriate

allowances.

relaxation

The'basic arc time" is related

to the total

time required to weld a fabrication

as shown in

Equation 31(172,173)
Tarc
Or =

(31)
Twe1d

Where:
OF

The operating

Tare

Basic arc time

factor
(mins).

Total fabrication

Tweld =

or duty cycle.

time (mins).

From Equation 31 it*can

be seen that the

cycle is the proportion

of the welders time in

which he is actually
metal.

welding or depositing

Duty cycles are influenced

conditions

duty

weld

by the working

in the welding shop, the welding

224

involved
type
the
and the
of
work
used,
process
help given to the welder in the form of unskilled
labour.

of these factors

influence

The actual

by
be
time
duty
the
measured
only
can
cycle
on
A previous study(122) of

and motion studies.

the welding of standard fabrications


/

at H.M. Ltd.

that the average duty cycle was

indicated

0.35 for a combination of 80%

approximately

Gas Metal Arc welding and 20%Manual Metal Arc.


However, an average duty cycle for a range of
is unsuitable

fabrications

for the present. work

since duty cycles can be expected to vary


between fabrication
Duty cycles
design

TABLE 32

(i)

must thus be related

features.

influence

types and welding processes.

The principal

duty cycles

to specific
features

are listed

that

in Table 32.

DESIGN FACTORSAFFECTING WELDINGDUTY CYCLES

The number of planes that welding is carried

out on.

This factor

determines the number of times the

fabrication

must be re-positioned

to enable the

joint.
the
to
to
weld
access
gain
welder
(ii)

The number of individual

weld lengths.

determines the number of weld start


ancillary

operations.

This factor

locations

and

225

TABLE32 (Cont'd)
(iii)

Location of welds and their

(iv)

Shape, size and weight factors

ease of access.
that influence

the

of the

time for manual or mechanised manipulation


fabrication.

The primary process used to weld standard


at H.M. Ltd. is semi-automatic

fabrications

Gas Metal Arc welding and is therefore

the only

welding process considered in the present work.


The range of the duty cycle for the Gas Metal
Arc welding process is considered to be 0.25 to
0.60(172).

Although Myers(173) considers that

for small amounts of welding the duty cycle could


approach 1.0, a figure

of 0.6 is considered during

the present work to be a practical


since the maximumrelaxation

upper limit

allowance

that

35%. To this figure must


_
be added allowances for assembly and tack welding
is.
be
could
added

and equipment handling.

of the fabrication
Although

actual

accurately
listed

time studies

estimate

are required

the influence

in Table 32 on the welding

subjective

estimates

are considered
sufficiently

during
accurate

(using

for

of the factors
duty cycle

experienced

the present

to

personnel)

work to be

estimating

purposes.

226

(iv)

Chamfers, Circlip
Little

Grooves, Radii and Tapping

was observed in the machining


55)
times for these features.
Cochran('
points
variability

out that in such circumstances,


accuracy levels

can be obtained by using the

average of the operation

adequate

times.. Henceregression

equations were not developed for these operations.

4.3

The Multiple
Ezekiel

carrying

Regression Procedure

and Fox

(174)

and Hawkins

out the regression

procedure

models)

using a desk top calculator.

rapidly

lose their

variables

within

provide

methods of

(i. e. developing

estimating

However, manual techniques

as the number of independent

effectiveness
a regression

(175)

model increases.

In addition

when large numbers of estimating

models need to be developed

the use of a desk top calculator

becomes both tedious and time

consuming.

However since multiple


statistical

linear

regression

is a standard

procedure used in a wide range of business and

research applications
for most micro-,

mini-

computer software packages are available


and main frame computers.

The regression

independent
in
the
variables
vary
of
packages
number
software
that can be dealt with and the type and range of supporting
statistical

data produced.

227

The regression

package at the Loughborough University

Computer Centre is a module of the Statistical


Package for the
(176)
Social Sciences (SPSS)
and has been used during the present
work since the package:

(i)

can deal with 50 independent variables

and hence

the maximumnumber of independent variables


expected to be required

(i. e. 20)

to develop estimating

equations,
(ii)

provides

a wide range of supporting

data

statistical

(see Table 33).

(iii)

combines standard
regression

procedures

to be exercised
variables

(iv)

and stepwise

has the facility

model,

to test

inter-correlation

independent variables

work since this

and

independent variables

for

from the

This overcomes the need to test


for inter-relationships

to running the regression

The stepwise regression

control

of independent

and exclude variables


basis.

linear

which enable considerable

over the inclusion

in the final

model-on this

multiple

prior

procedure.

procedure is suited to the present

procedure allows those independent variables

to

be chosen which provide the best prediction

model possible with

the fewest independent variables.

the method recursively

constructs

a prediction

That is,

equation one independent variable

at a time.

BEST COPY
AVAILABLE
Variable print quality

228

SPSSOUTPUTFILE

TABLE 33

oI

M
n

10,

Q
W

.
I

III

.
.s

LL

0-

..

C
WNN
a
<C

M1

p
M1
^

CNN
IA

J
O
H

"

Cl
0
C
0
c

W
f. i
2

Vf
W

F
K
f
6

6
f

"
C

zr
<

M
M

C
C
C
C
C

J
f

z
c

C
K

vNC
Wp
aC
<MQ

(.

J
6
f

IC

a
t

.r
I

V.

_r

W
C
T
Y.
W

~
`C.
A

I-

..

c
z

rcc

n, r

d
J
h
C

..

x-

L6

I'
s

<J
C6

F6G-

C
YX

<
CS

GC
Z4
t\

>

V.
V
V

<C'
1V
r

<ss

*C

W
L

I-

W
Z

11

"W
"

Nr.

CCNGry
N"

"A

.'_
Z.

I.

n
":s

.LAN

uC

"z

Yi

P
J

_
`3

rC
pCNpGA

rr
rcrJ

4CJ
CC
<hN1
N
CS4P

MNr
rJ

V.

J"A

VI
2

<

Y
Y

HW

Nh
vY

fe
JY

r"
J

r"

1V

s.
V^3

CZ
2N

2N

Sf

Ir
1L

.. _;

3LW>
<f
MNG

CSC
""2<

<

I,
"I
II
$1

Is

-r
1 h

-r

J
r

z>
I

...

.u<

rR<Y.

z
I-

..

ls

7=

t
t

I
9

u.
C-

GC

fn
%

f6f2

is

CI

2V

^j

U.

W2

-Iz
WH"G2

"f2

C_

I..
_-.

;1

""<F
""QY.

iV W
LN

U.

"".

"V

iI

"f

Kr
f

OC

""'JY

"ra

n.

.C

NC
22
CW

.r"
"C

~
y7.

S
t

""C

QWK
p=<
Y<6

AC

.r

4C

Hv

1r'"0.

.'c
2
Ca

CCh;

"n

"

Gr

1G

W
t

r.

aC
,`

U.

P-

II
II

W-. arn

tr
>WG
JT
<vV.

W
G"

`I

>

'~
C

<<
2IS

II
I

rw
CANCA
aC"G

T!

:.

,
I

229

TABLE 33 (Cont'd)

I
Wr

JM
IC

a
"
"

-0
N

0
N
IC
If

>
0

."

am
tv

O[

"
J
N

.Cp
om

M
M

"
"

aC A

'O

4i
V'
2OO
w

CC

NO
C y
"
CA
OO

cc
1

a
W4nM

Y/
LL

<d
4
OAAt
Yi

<A
eC Y+
"..

f0

C9

<

.^.

ACR
KAC
CC{

Q.

C
I.:

Wt

't
m
4O

cc

a
J

G
2

z
C

'
C

J
vr

{Y
I-'
z
t
aa

Cr
CdG

. a0 C

z
v

C
7

v.

CC

ce
:.

zsw
'"

r-n

on

cc

r'

W\f
2S

U.

r
WN
z"W
W

-rr<
c
<nfa
0

-c
2W
C

zrz
G
LL

"

.c.
LL

""

Cc
tJ

.L
"J

H.
6
{:
Ix
O2tt
rf""

""

I<W
-a"G
OW

Y.
J

aC

.
N2<<

0--

z2

.a
62

C
LL =

LG
Ot .
< LL GOZi

"+ WN
236GCC

WK""

Nc""

a0

CD
O

O2
HV"""
G
<

LL

""W

CL
ZxX

2ZZ

"""2
1V

N.

n
i
r.
N

230

TABLE 33 (Cont'd)

II

.
.
N
1:

r
W

t
G
O
O

--------------------------

.;

.
.

CC

z
c

r
.

V1

CO

tL'
a
C

G
Y.
N

WWY.

Y.

JCCPd.

CL

z
i

pMM

r
J

NrANC
y
It

IV
Y.
C K";
PNN
It
NCOdwW.
NNM*
aC 0C

N"^
J
aC MM
dNOPN"P
aC CNCCAd
ACfbCNMA0ddCC-A
r' hNdaC
NrOMMNN"O
'
""""""_____"""""
IlttttItttIIIIItIIVZF.

Cr
MMdC

:;

1L W
zM0MACCdCCW.
aC d
C. NNdrX
NdGP
O. OP
1' -ACM
e M
ACCAO.
AA
N^
AOlC
OM
MM
"dN

NNNdM.
4'
ANMNdCN.
N.
p NP
"A
CC r"
dzNJAO
aCr0C.
f MC.
dddCIVKACCC
M"NA
"

""

"

apN

.O

E(
AGAAM6
.

cc AO;
NM
MNd
7C N1A

p
""-""MH

.
W

W
H
t

~
Y
V

""""

MNC
NK
p
NNM7A
t
ti
K<"'
K
CrrP.
r.
CV
C
C.
C C O
AKOCJC
M
CCV
YN
JAV
rCs
pJ
C

NX
JH
AL
A^

"

e"

orrr

"

: r.

NN
W4J
C N.
O
A
V` hJApJKMJ
A
CL
r.

r.

JA

O Jp
M N

OA

Ahr.

fV
JO
AM

JJM

aa

r.

Ar!

OK
Arti.

P
G

Yi

ff

r.

+rar.

r.

-z

U.

r
h

H
N
W

rr

=W
Ca

..
46

a.:

..

O[

"}

WK

s<

CGCCCCGCCCCGGCGCCCCC_GGCCCCCCCCCC
CGCCCCCCGCCGGGCCCCGGCGCCr'CCCCCCC
G
JOCG
GrCCCOOCOOGCCC070rGGCCOOCCCC
C0
/1C
t0
CGCGCCC
000CCCCCCCCCC
>
.
e=

"= C r,
V` C.:
h
. _
N
JN.
f
MNA
aCh'"'h

. 1. t
7"

.-.

".....

-...

".

Nr
C

.r.

Y V.

CG
C""'

GGC
Jf<SC.

} Vhrrhf

0[C

RrJ

M`

CCC
Jt

M r.

VI
{

CC
h

.-

2
r=

""

"a
v

:NMJ

V O f.

eC P

C.

NKJ

NG

I eC 0O

r` hMJ

N'G

4' PGn.

C
C.

r s.,
< "
"i
r`
a-

z
C
1L:
J"""

C.

WN"

.. r

.
V".

I-

Ok

WrY.
C'Z Si
`uG

M
a

Ot

v.
u.
Y

a
W"

..

v
uC

Z
O

0tW

-=

yG""
_=

t:

C1

FW
t
a

oVe"f

t:

Y.

LL

" "

cs

tL
JMC
O NNNKO
J9pPJ
JN
CK
GIMC
NK
MNO

A 1. ' CCNvGNNJ
V' riKhA0C
0v
JeCC
0 0
C
CaCCP:
CA
ANC.
r"K
."
. "Ar4C
C
i`
C. M'.:
JJ
N
C CA
JM'r'.
"KJ
.

<
W=G
Z
=C
r..

SH
yt

'
C

>

WM
2

H
N

MN

0.

x^
C
z

F
C
r

CCCCCCCC

7=

231

The first

step is to choose the single

the best predictor.


to the regression

The second independent variable

then proceeds in this


by-step until

recursive

no other variable

The program
step-

make a significant

to the prediction
is selected given

package allows the data for each'machine

to be held in separate sub-files.

can then be carried

are in

in the equation.

The SPSSregression

The regression

out on individual

procedure

or combinations of these

Hence equations can be developed for single or ranges

sub-files.

of machine types enabling the total

4.4

will

At each step the optimum variable

the other variables

within

variable.

fashion adding variables

(normally a user decision)

equation.

tool

with the first

the desired number of independent variables

the equation or until


contribution

to be added

equation is that which provides the best

in conjunction

prediction

which is

variable

number of estimating

models

a system to be minimized.

Equation Validation
Since the multiple

linear

used when the exact relationship


variables
estimating

is unknown, it

regression

between dependent and independent

is essential

that the accuracy of the

models is confirmed to be within

of H.M. Ltd.

technique is to be

(E = 0.096 and S=0.314,

the current

Section 4.4.1).

limits
To indicate

formula accuracy the most commonmethod adopted(174) is the


Correlation

Coefficient

(Equation

32):

232

i=N
iY;

R=1

II

N.0'y. 0"x

- Y)(X;

(32)

- X)

i=1

Where:
R=

Correlation

Coefficient.

deviation
Standard
=

of actual values.

(Tx

deviation
Standard
=

of estimated values.

yi

dependent
Actual
the
value
of
variable.
=

xi

= Corresponding estimated value of the


dependent variable.

y, x

Mean
values of y and x respectively.
=
Number of data points.

N=

Coefficient

The Correlation

the proportion

indicating
by
accuracy

for.
been
has
that
accounted
y
to
1
the
unity
closer
and
and
of the model.

is said to measure. fornula

The value of R lies

variation

the greater the estimating

measure the difference

(termed
"residuals")
values
estimated
a model's estimating

accuracy.

since little

it

accuracy

between actual and


cannot truly

measure

The present work supports this

correlation

is observed (Figure 16)

between the value of R and the Average Proportional

in

between 0

However, Cochran(155) points out that since R

does not directly

observation

of the original

Error

(Equation 36) which is a measure of formula accuracy based

233

Hence the use of R to measure

on the size of residuals.

formula accuracy could be misleading


high residuals

to unity

be obtained and the model

could still

would have poor predictive

since when R is close

powers.

Other commonly used methods of measuring the estimating


lo,

of regression

accuracy

of Variability

Coefficient
with

the size

dependent

out that

simply

data has a narrow range.


1-e could

fails

that

be an unreliable

17 and 18.

to consider
foundation

and

for

therefore,

Coefficient.

the estimates

the use of the quantity


it

and is,

on the use of the Correlation

variable

correlation

in Figures

accuracy

be
to
accurate,
appear
could

formula

and

33) proposed by Cochran makes direct

Equation

However, Cochran points

since

also possess little

to measure formula

use of residuals

the F-ratio

1-e (where e is the average error ratio

using

an improvement

is,

as illustrated

of residuals

The quantity
is calculated

models that
(176),

(174)

obtained

from a

because the actual


In these circumstances

be a poor measure of accuracy

a narrow base of experience


projecting

future

results.

However, the present author considers that the use of an


is
data
a pre-requisite
range
adequate
building

hence removing this

objection

xi)
Ny

to regression model
to the use of 1-e.

(33)

could

234

OF R WITH AVERAGE
ERROR
PROPORTIONAL
FIGURE16 COMPARISON

1.0
01x

0.8

cx

x..

iX

0.6

xx

XXx

-P
a
'v

0.4

xXX

. r44-

14 0.2

xX

C-)

-0.5

-1.0

+0.5

Average Proportional

Error

+1.0
(E)

AVERAGE
PROPORTIONAL
OF
F
WITH
ERROR
COMPARISON
17
FIGURE

70

xX

60
. -.
v
0
.

50

(Ki

40

30

t0
. r.
i
(0

x
it

)c

IL

yl

XYX

19

20

xx
x

10

it

-1.0

-0.5

Average Proportional

+1.0

+0.5
Error

(E)

235

FIGURE 18

OF COEFFICIENT OF VARIABILITY WITH


COMPARISON
AVERAGEPROPORTIONAL
ERROR

41

"r
r
"r
-0

70
60

X
X

"r
L
tC

50
4O
d-)
aJ
"r
U
.r
44O

x
X

ix

xx

40

xx
I

30

20

Xx
xx

10

1t111111

-1.0

-0.5

Average Proportional

+0.5
Error (E)

+1.0

236

To take into consideration


an estimating

construct

model, Cochran suggested using the

Index (AI, Equation 34) which again increases in

Association
direct

the range of data used to

to the degree to which the formula accounts

proportion

in the cost data.

for variability

AI

(34)

=1-T

TY

Where:
e2

=1

iY;

- xi)

21

zI
Qy

=N2

Q'e2

Standard Deviation

of the Residuals.

Standard Deviation

of the Actual Values.

0y

However, although
in actual
in a factor
estimating

the Association

data to be considered,
(y) other
values

by doing

than the difference

influencing

AI is not, therefore,

Index enables

the calculated

so it

the variability

effectively

between actual
accuracy

results
and

of a model.

a true measure of formula accuracy.

In this

is
1-e
disadvantage
the
of
a
use
since this measure of
respect
formula accuracy also includes the term y in its

determination.

237

An index of formula accuracy, the Standard Error Ratio


(Sr2, Equation 35), is provided by Ezekiel and Fox(174).
index is suitable
it
since
models

Sr2

for measuring the estimating


uses only the size of residuals

=N,

I(Yi

This

accuracy of costing
in its

determination.

- xi)`-

(35)
----

Sr2 is a measurement of the range of estimating


therefore
equation
and
an
using
obtained

indicates

errors

the closeness

estimate may approach the unknown actual

with which any specific


value.

The present author considers that an indication


estimating

overall

Proportional

accuracy can be obtained using the Average

Error (Equation 36).

takes into consideration


large,

of the

The use of this

equation

circumstances where residuals,

be
however
a relatively
only
may

small proportion

although
of

corresponding y values.

(36)

E=1
N

yi

Where:
E=

Average Proportional

Error

238

The Average Proportional


average error as a proportion

Error is a measurement of the


of actual values and is appropriate

for product costing equations since the accuracy of cost estimates


is normally determined as a proportion

of actual costs.

be used to indicate

For the present work the value E will


e

the accuracy
this

of estimates

method of computing

of the regression
using

regression

error

proportions

despite

the apparent

estimating

accuracy

procedure.
analysis

Minimizing

between

conflict

and the objectives

the values

(yi

does not necessarily-result

- xi)

in the

being minimized.
Yi

However the regression


accuracy

estimating

of high cost

to form the majority

(Section

5.2)

products

manufactured

with

procedure

the ability

improves the

effectively

items which have been shown


of the total

at H. M. Ltd.

This

costs of the
together

advantage

to improve the estimating

accuracy

of equations

data
low
the
the
of
end
range using minimum operations
at
(see Section
suitable

4.5)

make the normal regression

method of developing

costing

technique

times

the most

models at H. M. Ltd.

It is possible to modify the regression procedure to


%1)
`
into
to
take
values
consideration
and
yi
xi
minimize
-

y; /
the relative

numbers of components manufactured.

(Appendix
VI)
be
would
more applicable
procedure

This modified
to high volume

239

systems in which the large batch quantities

manufacturing

involved could magnify small costing errors

into large cost

variances.

To indicate

in individual

the variability

hence the effective

accuracy of regression models, the present

Ile
author proposes the use of the quantity
represents

the Standard Deviation

hence indicates

(Pi - E)2I

S (Equation 37).

of the proportional
in individual

the variability

s=

errors and

errors and

errors.

i
(37)

Where:

S=
Pi

Standard Deviation
Proportional
=

of the error
yY

error

x )

proportions.

for the ith

observation.

E and S values can be used as descriptive


accuracy, location

and variation

frequency distribution

irrespective

of the labour times.

to determine confidence intervals


that the random variation

since this

measures of
of the type of

They cannot be used


requires

an assumption

of labour times is normally distributed.

240

4.4.1

Accuracy

of the Existing

Estimating

System

at H. M. Ltd.

Using the quantities

system at H.M. Ltd. was determined.

estimating

(Figure 19) indicated

results

S=0.314

in the individual

estimating

errors

since

(i. e. 31.40).

Comparison with Time Study Data

4.5

The accuracy
was tested
set

the +10% accuracy limits

However, there is a high

that managementspecify.
variability

The

that the Average Proportional

This is within

Error is 0.096.

E and S the accuracy of the

against

time study

at H. M. Ltd. 's current

data with

accuracy

(Appendix

the acceptance

limits

V)

criteria

of E=0.096

and

Time study data used to develop the equations was

S=0.314.

to
test
employed
not subsequently
In this

model developed

of each estimating

way a more valid

indication

the accuracy of these equations.


of estimating

accuracy would

be obtained.

Those equations that failed

to meet the acceptance criteria

imoroving
for
possible
methods
of
estimating
were examined
The causes of poor estimating

(i)

Incorrect
Frequently

accuracy were found to be:

combination
either

predictor

variables

variables

and/or

accuracy.

of predictor

changing

the combination

used and/or
removing

variables.
of

adding new

variables

improved

the

241

FIGURE 19

A COMPARISON
OF ACTUALAND ESTIMATEDLABOURCOSTSAT H. M. LTD.

E=0.096
S=0.314
le

3.0

N,

d-
O

2.0
O
L]
.
tC
J
r

t0

1.0

1.0

2.0
Estimated Labour Costs (VS)

3.0

242

estimating
typical
(ii)

accuracy of models.

Examples of

out are shown in Table 34.

changes carried

The use of negative coefficients

the

within

A disadvantage of using the

estimating

models.

regression

procedure is the possibility

of

developing equations which contain independent


Hence the

with negative coefficients.

variables

of negative manufacturing times is

estimation
possible.

If the negative sign is contrary


causal effect

of the independent variables

the
that
variables
suggests
model without

had to be re-run without

variables

variables

The regression
variables

procedure

producing negative

included.
equations exclusion

of the predictor

with negative coefficients

equations that failed


criteria

accuracy.

with many equations developed

during the present work.

For certain

Thelwell(151)

can be excluded from the

serious loss of estimating

This proved correct

coefficients

to the suggested

to meet the stated acceptance

for accuracy levels.

of these

Inclusion

however improved the overall

equations but resulted

produced

accuracy of

in negative operation

being estimated at low values of predictor


(see Figure 20).

times

variables

In order to overcome this difficulty

243
TABLE 34

EXAMPLESOF ALTERING INPUT VARIABLES TO IMPROVE


ESTIMATING-ACCURACY.

Ec
"r
FN
W
C'S
Z
S

"r
_C
C)

co
Q

TJ

(A
CO

Q)

CU
C71
tC
i
Q) (l)

C
O
"
+)
rt3

O>

O
+)
O"C7

4- Wi
a)
O. C S- E
+J (3) S-

eo
m

i
CU
+ .C
U
CU +)
E "

Cl O
Q1 O 4-

s..
X O.
C C
a
i-) "r4f"r
g
aJ i-) oa
(n
CE
cZ

>

sai
(1) c.1 .nO
+J r- E "r
= 4-3

o "r
'O
C:: UC

co

NO
C1 O
Oi
c3: CD

-C.7

C3O
to

G)
Cl)
>
0 .C>
4-) O

v-v
-a C
ro

E aO
CU Ni
Cr cz c7

SCU

.N

-C0

.r.
4-3

4J
CL
a
t+

OO

OU
O
F--

'o

c
.
.n
E

G)
>i
O

G)

}
--

G)

., _
-0
I

G)
CL
C1.

N
C)
4-)
O

+
-

r*l
N

C)
01

r=
1

N
Z
W
M:
r
Zd
QZ

w
-+
f-

F-

.-

C7

tn

0W

--

WO
dO
Ow
CY`i

r--4
2Q

F-

XU
QO
H

O
"r.

0.

.C
i
CL

a
0

. -.

C)

t
41
0
0
---

C)
I-

a.

C)
.0

:N~
C.7
v

Z
l0
N

.,...

.:
1--1
ci
CV

ZZO

t
C)
C)
N
H
C)
.0
t0
"r..

I--,

f0

>
O
rr
FP-4
co
M:
O
U
J
-
1r-+

+
z
+++
zzz
o00
V)

V)

ZZZ
WWW

i
1--n

C.7
1--4

00

XXi

c
"

U
iO

Rs

ZG

(I)
a
O

5a)
.a
E

..4Jc

O.
(1)
G

If.
C.
(1)
a.
0

(1)
>

r-"r-

l0

OaO
Nu00NU

M:

C)
>

O
O
C7

au
rn
(c
S-

0
o
S..

a)
>

cc

O1

r
T7
O

to
O
C7

a
m
a4-.
0
a
.0

LL
4-)
"r

a
V
to
LL.

v
>
i

co

a)
4-)
a)
E

a)
-0
., _
N
4J

U
. '..
,a
C)
i
0
FO
O
"r"

0.
. r...
i
U
N
a

O
i
O
LL

rd'

4O

. -

NUQ.

=w 0-

4O

-1

000
>G
ddd=
Z

w F--

v)
G)
>
0
O
S.

t4-3

C
e0
-m

td
GJ
C7

i
+j
0

244

FIGURE20

THE ESTIMATIONOF NEGATIVEMACHININGTIMES


USINGREGRESSION
MODELS

.le

4
Regression

Line
oe

(n
C
. r.

"r
IC,
C
"r
C
I

UI

+ve
0

-ve

ii

.-I

Negative Machining Times

Independent Variable Values (xl, x2, x3


xn)
...

245

the dependent variables


minimum operation

times shown in Table 35 below

which estimated operation


Since the allocated

concerned were allocated

times could not fall.

times are based on an

examination of actual minimum operation times


this

procedure effectively

improves the accuracy

oe

of estimates from equations with negative


coefficients.

TABLE 35

TIMES
MINIMUMOPERATION

Cost Centre

Equation
4.7.1)

(see Section

Number

Minimum
Operation

Times

(rains)
201

203

1,2,17,18

204

1,2,3,4

214

216

219

216

202

1,2,3,4

204

18

209

0.5

1.0

2.0
210

17

236

1,2

214

the

246

TABLE35 (Cont'd)
Cost Centre

Equation Number

Minimum

(see Section 4.7.1)


202

Operation Times
(mi ns)

9,13,18

)
3.0

219

1)

210

1-54.0

206

5)
)

209

5.0

18

The time study data used to construct


for
accuracy.
checked
was
also
equations
in this data source were identified

the estimating
The causes of variability

and are listed

in Table 36.

TABLE 36

CAUSESOF VARIABILITY IN TIME. STUDY DATA

1.

in machine process data (e. g. machining speeds

2.

Variation

identical
between
feeds)
and

machining operations

from the absence of detailed

methods planning and control.

Variation

in the dimensions and quality

e. g. the dimensional tolerances

arising

of materials

of cast materials

and the

presence of casting defects.


3.

The rating

of the work performance of operators is a

subjective

decision

and would therefore

between time study engineers.

used,

tend to vary

247

Models were not developed for estimating


since preliminary

attempts indicated

to H.M. Ltd.

(Figures 21 and 22) for the E

The frequency distributions

equations developed indicate

and S values of the estimating

of the equations is within

the accuracy of the majority


accuracy levels

current

of H.M. Ltd.

levels
had
accuracy
equations
limits

component or material

This type of data is normally

manufactured by suppliers.
not available

costs

that a major predictor

was the volume of a particular

variable

material

the

A number of estimating

that lay outside the current

of H.M. Ltd.

for

These were equations

estimating:

(i)

assembly task times,

(ii)

tooling

(iii)

machine manipulation
vertical

costs,

times for horizontal

load/unload

(v)

boring times (tols.

(vi)

turning
vertical

(vii)

turning

and

borers,

(iv)

vertical

times for horizontal

! O.lmm) for chuck lathes,

borers and horizontal


times (tols.

and vertical

borers.

borers,

>0. lmm) for chuck capstans,

borers and copy lathes,


times (tols.

borers,

and

<0.1mm) for centre lathes

that

248
DISTRIBUTION OF 'E' VALUES
FREQUENCY

FIGURE 21

100

N
4)
N

O
i
d)

50:

0.25

0.20 0.15 0.10 0.05 0

0.05 0.10 0.15 0.20 0.25

DISTRIBUTION OF 'S'
FREQUENCY

FIGURE 22

(E)

Error

Average Proportional

VALUES

100

N
IN
C..)
4O

50

CU
.aE

//
//

/
/
/I

-I- I

0.05 0.10 0.15 0.20 0.25 0.30 0.35


Standard

Deviation

of Proportional

Errors

249

(i)

Assembly Task Times


Since time studies of assembly tasks had-not been
out at H.M., Ltd. data was not available

carried

develop
to
estimating
which

models.

repetitive

The use of MTM


to highly

time standards are more applicable


'Oe

from

work tasks which are normally not found

in-low volume manufacturing

systems such as H.M. Ltd.

An attempt was made to use STT times for various


sub-assemblies and determine the relationship

between

STT times and the types of components that formed


the sub-assembly

(Table

37).

(e.
Equation
g.
models
produced
of which clearly

procedure

38) the coefficients

did not reflect

of the predictor

effect

The regression

the actual

variables.

causal

Hence the

be
included
in the estimating
could
not
equations
system.
At

4.48a1 + 93.87a2 + 48.01a3 - 101.09a4


44.86a6
130.73a7
80.50a5
71.89a8
+
+
+ 76.61a9 - 38.75a10 - 9.49a11 - 514.66

Where:
At
al, a2, a3... all

(mins).
Total
time
assembly
=
types
The
number
component
of
=
(Table 37) within

the assembly.

(38)

250

COMPONENT
CLASSIFICATION FOR DETERMINING

TABLE 37

ASSEMBLYTIMES

COMPONENT
DESCRIPTION

VARIABLE
NAME

EQUATION
38
CODE

Disc shaped with outside diameter <150mm

a4

DISCB

Disc shaped with outside diameter $,150mm

a10

SHAFTA

Shafts with length <250mm

a6

SHAFTB

Shafts with length p250mm

a5

BRINGS

Bearings, keys

a2

SCREWS

Screws, plugs,

CCLIPS

Circlips,

dowel pins,

retaining

rings,

DISCA

PLATE

Plate sections

SMALLFAB

Castings,

nipples,

nuts,
dust

springs,

bolts

caps,

washers

SEAL

Castings,

a1
a9

Pressings and Fabrications

dimension
maximum
of <150mm
with a
LARGEFAB

all

a8

Pressings and Fabrications

with a maximumdimension of >150mm

a7

Seals, gaskets

a3

251

Equation 38 illustrates
analysis

regression
That is,

a disadvantage of using

to determine causal relationships.

as the number of predictor

to construct

used

equation increases it

an estimating

appears that the possibility


technique finding

variables

of the regression

non-existent

also

relationships

increases.
The technique

models was that

regression
author
for

developing

adopted for

by the

used previously

models for

when establishing

assembly time

estimating

the. assembly of heavy custom designed

The technique

involved

using

the assembly of existing


H. M. Ltd.

product

to choose predictor

assembly times

for

with

ranges within

variables
levels

various

cranes(154).

familiar

personnel

times

and estimate

and combinations

of variables.

From this

data, estimating

models were developed that


between assembly time and

expressed the relationship

the dimensions of various component types, such as


disc shaped components, pins,
plate and fabrications.

castings,

shafts,

In addition

numerous

(Appendix
IV) were allocated
components
times shown in Appendix V.

the standard

252

The accuracy of the estimates developed by


these equations could not be measured against current
in

STT assembly times because of inconsistency


the STT times.

is particularly

when comparing STT times for assembling

noticeable
similar

This inconsistency

In these cases maximumdifferences

products.

of 200% have been observed.


(ii)

Tooling Costs
Attempts to develop equations for estimating
tooling

of various special

as illustrated

successful
Whilst

examining

suggested

the data used. to construct

tooling

special

The primary difficulty


manufacture of jigs,

in sub-contractors
increases,

are difficult

fixtures

numerous sub-contractors

firms.

since the design and


and special

levels,

tooling

of variations

overhead rates,

labour rates and material

to isolate.

that

costs was inappropriate.

arises

profit

the

models for

Hence the influence

sub-contracted.

factors

were highlighted

the use of regression

estimating

price

items were only partially


in Figures 23 to 28.

models factors

estimating

the costs

In addition

costs

since

could be employed these

would also vary between sub-contracting

is

253

800

t/f

700

W
4-3
N
O
U

600

. 500

400

400

500
Estimated

600

700

800

Cost (i's)

700

Vf
W
....
+1
N

0
U

600

600
Estimated Cost ('s)

700

254

100

80
. --4

(N
W

"

t/1
0

60

U
Q

- 40

20

20

40
Estimated

60

80

100

Cost (S's)

14

12

Vf

10

W
v
4-)
N
0
U

2468

10
Estimated Cost (I's)

12

14

255

107
W

4-,
1/f
O
U
r

i-,
V

150
.
Estimated

200
Cost

250

('s)

700

N1
W
v

600

4-3
N
O
U

4-)
V
C

500

400

300

300

400
500
Estimated

600
Ws)

700

256 '

Therefore it was decided to remove the jig


fixture

costing

and

equations from the estimating

However in the event that H.M. Ltd.

system.

decide to manufacture jigs


the procedure for calculating
been retained

in-house

and fixtures

costs has

tooling

in the estimating

software program

(Appendix III).
of estimating

At the present moment the lack


equations in the data file

a default

initiate

message informing

that equations for estimating

will,

the user

tooling

costs are

not yet available.


(iii)-(vii)

Manual Task Times and Machine Process Times

The equations

for

estimating

machine manipulation

times for

horizontal

and vertical

times for

horizontal

borers;

n. lmm) for
horizontal

chuck lathes,
borers;

borers;

boring

times

for copy lathes and vertical

(tols.

(tols.
and

<O. lmm)

borers represent a

(0.3)
of the total
percentage
small
equations developed.

times
borers

vertical

turning

load/unload

number of

The causes of inconsistency

between estimates and time study data could not be


found.

It is therefore

arises

from the factors

assumed that this


listed

inconsistency

in Table 36.

257

4.6

Comparison with existing


Although

the estimating

equations

are within

determine

whether

times

this

objective,

estimates

been compared with


time for

machining
tools

it

for

complex designs
This will

operation.

using

STT's since

to

which require
ensure that

In order-to

additive.

the regression
these times

a component processed

and are therefore

is necessary

of the

can be used to build-up

are not wholly

errors

estimating

limits

the equations

more, than one manufacturing

of the majority

accuracies

the stated

manufacturing

accurate

Time Standards (STT's)

equations

represent

on individual

the sum of the individual

achieve
have

the total
machine

operation

times.

The accuracy with which the regression


is
illustrated
STT's
estimate
ideally

4.7

all

points should lie

equations can

by Figures 29 to 46 in which
on the 45 lines.

System Computerisation
Although the system can be used manually as described in

Appendix II,

software has been developed('77)

to carry out the tedious calculations


computational

errors.

to allow a computer

involved and hence reduce

Operation notes for the computer program

in
III
Appendices
are contained

to V:

258

20
I-N
"r

Ftcn

15

a)
E
"r
H

10
Q"
i..
cc
F-N
G)
N

10

Estimated

15
Time

20

(Mins)

50

N
c

"r

..

40

.
IFN

Ei
E

30

a
(ii
I-

20

m
N

10

10

20

30

40

50

259

5
N

FN
N
E
"r",
F-I

3
Of

+-)
N
cn
LI

NI
4)

r0
tJ')

234

Estimated

5
Time (Mins)

25

...
N

20

F-

d
E

15

F4-3
N
0)
L
to
H

10

N
C1

10
15
20
Estimated Time (Mins)

25

260

200
tN
C
.

160
F-FN
E

120,

"r

4-3

F--

80

4)
r

NI

40

40

80
Estimated

120
Time

160

200

(Minn)

100
Ni
CI
.r

80

f-N

60

"r

F
4-J

(I)
S..
t0
f"
V1
G)
r

40

t0
V)

20

20

40

60

80

100

261

50
N
C
. r.

FI-N
v

40

a
E

"r
h-

41
ai

30

rn
cC

Vf
G)
r

20

10

10

20
Estimated

.t/1-

30
Time

40

50

(Mins)

25

... i

H
H
N

20

C,
E
"r
Fi-
G)
C,
S-.
r
Fvf
C)
r

r6
Cl)

15

10

10
Estimated

15

20

25

262

25
N
. r.
Z:

IH
N

20

Q)
E
."

"r

I-

15,

4-,
a
rn
S.
-" U,
N
r
f',

10

tN

10
Estimated

15

20

25

Time (Mins)

25
0N
"r

FFcN
()
E
. rF-

20

15

4-3
G)
O)
t.
r0
FN
G)
r

L 10

IO
N

10

15

Estimated Time (Minn)

20

25

263

Vf
C
"r

15

,I.--,
FN

.. -I

C)
E
H

ti

+J

rn

10
e

t/)

10
Estimated

FIGURE40

15

Time (Mins)

STT VERSUSESTIMATED
TIMES

COSTCENTRE236
SAWS
.tN-
C
"-.
i
..

10

1
II-c/)

"r.
I-
4-2
N
L

1-`

cn
a)
r

b
/)i

5
Estimated

10
Time (Mins)

264

30

"r

M:

. -,
. -

r! 2
C)

20

H
4J
C)

rn
SN

R3

10

20

10
Estimated

30

Time (Mips)

200

"r

160

120

%80

40

40

80
Estimated

120
Time (Mips)

160

200

265

25
t/1
C
. r.

0F
N
v

20

a
E

If-

4-3
a

15

,or

Sto
FN
. G)

r-
tN

10

15

10

Estimated

20

25

Time (Mins)

STT VERSUSESTIMATEDTIMES-

FIGURE 44

COST CENTRE258

25

NC BAR LATHE (BATCHMATIC-) xx

N
C
"r

20

f-

N
dJ
E
.,.
H

15

4J
d)
RV

Xx

al
(If
C)

10

V)

Y.

Xx

I1

10
15
20
Estimated Time (Mins)

25

266

FIGURE45

STT VERSUSESTIMATED
TIMES

COSTCENTRE270

SPLINE CUTTERS

t/1,

C'
I

'ix

-.
F-N
...
W
E
"r
F-

xx

4-)
N

xx

Q
L

N
r

a)

Estimated Time (Mins)

5
N

c
.v
. -,
II-

I)

a)
P

"r

4-3

(1)
r. P
S(0
F-CA
a)
rrd
N

234
Estimated Time (Mins)

267

(i)

Appendix III

contains the operation

the program and a full

notes for using

program listing.

Included is

of the input codes for the manufacturing

a list

processes that can be estimated using the system


(Flow Chart Code 3).
(ii)

IV contains

equation

codes and the order

variables

(iii)

definitions

Appendix

should

be entered

of predictor
in which values

into

data required

for

the computer program.

Appendix V is the program data file


relevant

variables,

and contains the

to determine the machine type

when general machining processes are being costed


(Table 51), a list
ing coefficients

of the predictor

variable

for each operation

estimat-

that can be

(Table
52) and the labour cost rates,
costed
manufacturing

overhead rates and administration

overhead rates for each cost centre

(Table 53).

System Size

4.7.1

The system consists


models.

of a large number of estimating

A method of numerical coding was required to allow

the appropriate

equations to be quickly

initially
was
equation
machine tool

retrieved.

Each

coded according to the applicable

group (using H.P. Ltd. 's managementaccounting

cost centre classification)

and secondly allocated

an

268

operation reference number in order to allow each individual


equation to be located in the data file,

e. g.:

AA

L
i

Equation number for a rough turning

operation

Cost centre coding for bar lathes

Specific

equations can therefore

the cost centre classification

be isolated

and equation number.

Exceptions are cases in which different


have identical

using

cost centre codings.

types
machine

At H.M. Ltd. these

are:

Cost Centre Coding

Cylindrical

and Surface Grinding

216

NC Bar Lathe (Batchmatic)


258

NC Chuck Lathe (Churchill)

It has been necessary therefore

to add an extra

the
to
input
to
form
cost
centre
code
codes (see
numerical
Flow Chart Code 3, Appendix III)

for the computer program.

The use of H.M. Ltd. 's coding system enables the


be
to
simplified
software

for the conversion of production

times to costs since both labour and overhead cost rates

269

vary with machine type.


is
not required
number
costing

4.7.2

Hence an additional
to identify

classification

each equation for

purposes.

Variation

of Machine Types

The range of machine groups capable of carrying


general machining operations
to develop decision

It was necessary

machine group, i. e. normally

engineers choose machines for components.

This type of data is therefore

inaccessible

to designers.

Examination of the component limitations


factors
the
that
main
revealed

affecting

of machine tools

the choice of

machine type are:

(a)

rules to enable engineering designers

to choose the appropriate


production

is large.

out

Whether cylindrical

or prismatic

operations

are required.
(b)

Whether a boring operation

(c)

The component diameter.

(d)

The component length.

(e)

The annual quantity

is required.

of components.

270

In the present study prismatic


"those
operations
as,

only with horizontal

out about the axis of

not carried

borers (input

is not necessary to include this

the decision

are defined

of the component" and are associated at H.M. Ltd.

rotation

it

operations

rules for determining

Hence

code 0,209).

group of machines in
the relevant

input

code.

The length,

diameter and annual quantity

have been used to develop decision


Appendix V).

rules-(see

of components
Table 51,

Two groups of rules are necessary since the

length
is
dependent
type
per
machine
maximumcomponent
boring
operation
a
not
or
on whether

The annual quantity


the cost

effectiveness

is to be performed.

of components machined determines

of using

numerically

controlled

tools.
machine

4.7.3

use of Computer Time

The Effective

To obtain effective

use of computer time it was

to
enable estimators
necessary
for the estimating
estimators

system.

This involved

to determine the relevant

Table 47, Appendix III


variables

to collect

and to identify

using the appropriate

in
Appendix
IV.
contained

the input data


allowing

input code using


the independent

Variable

Description

Notes

A standard form (Figure 47)

271

FIGURE47

DESCOSTIMATE
STANDARDINPUT
DATAFORM

PAGE:
MANUFACTURING TINE/COST

ESTIMATIONS

bF

DATE:

PRCDUCT:
D.: tING
No.

F:/C

CENT.'2L

OPERATION

A
CODE

Il?/C

F. T. _

TIME

/
TIi: 7--,
COST

272

has been designed in order to aid the collection

and

input of data.

Although the decision


appropriate

rules for determining

the

General Machining input code have

been included in the system software (for use when


is
limited)
time
not
computer
Description

Variable

the estimating
Design system.

the inclusion

of the

Notes could only be justified

software formed part of a Computer Aided


Cost estimating

could then take place

operator intervention.

without

Estimators with a detailed


techniques

(e. g. production

knowledge of manufacturing

engineers) would know the

machine code for estimating

relevant

determination
for
code
machine
module

equations,

the

can therefore

the
hence
system application
reducing
passed

4.7.4

if

be by-

time.

System Expansion
Since in the course of time new manufacturing

are likely
be easily

to be introduced
expanded.

In addition

of two extra machine groups.

equations to be included.

since the flexibility

enables estimating
developed.

space has been left

each machine group contains vacant data lines

to enable additional
beneficial

into H.M. Ltd. the system must

Hence data file

introduction
the
for
vacant

processes

This is

of the regression

equations for specific

operations

procedure
to be

For example, the machining of grooves for rope

drum tubes (Equation Code 20609).

273

CHAPTER5

DESIGN COST FORECASTING

274'
5.0

Introduction
based estimating

The regression
to quickly

system developed allows designers

estimate product costs if

general arrangement drawings exist

from which to obtain the values of predictor


having to carry out more detailed

This avoids,

variables.

design work which is necessary for

costing purposes when using conventional

costing techniques.

The

based
estimating system permits economic consideration of
regression
number
designs
of
alternative
and therefore has considerable
a greater
advantages over conventional

The ability

still

the likelihood

by reducing

accomplished

designs.

concept

and costing

costs

the necessity

without

to undertake

total
be
to
enable
technique would

(i)

this

objective

limited

be economically
to cost

required
task

using
of designs

of designs
estimation

of total

high levels

of

of a cost forecasting

The benefits

to be gained

are:

design resources could be allocated

more

efficiently,
(ii)

in the

product costs to be determined as

in the design process as possible.

in achieving

or best

only

the costing

to be, "the

The objective

stage would improve

could

Cost forecasting

considered

design and cost estimating".

early

This

techniques

task.

is
therefore
work

product

needs.

that

alternatives

the ideal

the amount of resources

forecasting
become
a
would
present

of emerging with

Rather than being an estimating

drawings

established

at the concept

considered

defined
to
meet
system

possible

techniques.

to add to the number of design

could be economically
further

costing

the level

of design resources employed on cost

ineffective

design concepts could be reduced,

27 5

(iii)

specification

for the optimum standard sizes could

be identified

early in the NPDprocess, and

data
on competitors
cost

(iv)

products for sales

techniques could be more easily

forecasting
provided.

5.1

Cost Relationships
An obvious

method of cost

forecasting

when designing

a new

is
the
of
cost/design
use
specification
relationships.
range
product
(178)
that
has
relationships
shown
can exist between
Pedder-Smith
design

such as product

characteristics

reference

to the New EHB development,

developed

from selected

hoist

sizes

causal relationships

capacity.

capacities

using

and costs.
cost

Hence, with

relationships

the costs

of remaining

can be predicted.

Since regression

relationships

hoist

size

analysis

is an effective

method of determining

the technique has been used to quantify

the

between manufacturing costs of components and hoist


Examples of costing equations are shown in Table 38,

i. e. only the coefficients


listed.

The predictor

capacity

(Tonnes)".

of the costing model (Equation 28) are

variable

in all

cases is "basic hoist

276

TABLE38

COST/HOISTCAPACITYRELATIONSHIPS

Component

Type of
Cost

Equation 28

Estimating

Coefficients
Xo

Accuracy

ES

('s)
i

Brake

Material

9.3

29.1

-0.002

0.077

Gearbox

Material

13.2

58.6

0.003

0.030

Gearbox Assembly

Labour

20.6

5.6

0.002

0.056

Hoist Block

Material

204.3

278.0

-0.008

0.042

Material

207.9

115.3

0.007

0.040

Material

56.2

36.4

0.020

0.090

0.021

0.097

0.000

0.023

Labour
Dual WoundMotor
Squirrel

Cage Motor

Rope Band

Material

Moulding Tool for


Rope Guide

Tooling

0.451

5393

0.234

504.6

It can be seen that the accuracies of the cost models listed


Table 38 are well within
S=. 0.314.

the present accuracy limits

of E=0.096,

However these cost models have been developed using the

total
the
of
costs
cost estimating

product range.

required

hoist
of
minimum number

In order to reduce the amount of

cost models must be constructed


sizes.

hoist
the
number
of
minimum
model

In order to construct
sizes is two.

It

develop
to
two
hoist
models
cost
using
only
'necessary
determine the overall

accuracy of these models.

using the

a linear

cost

is therefore
sizes and

in

277

Whenusing regression analysis


necessary to employ the full
encountered.

hoist capacities

range of cost data expected to be

using the smallest

(0.8 tonne) and largest

(8.0 tonne) that form the product range.

However this

is not always possible

since the order in which

units are designed must often consider marketing factors.


the schedule for the New EHBrange required the initial
1.6
3.2
tonne
tonne
0.8
tonne,
the
and
units-since
of
total
the
92.0%
of
represented
profit.

In addition

forecast

Developing

costs

outside

since the cost models may not represent


data range.

hoist

This is illustrated

development
these units

sales and hence the largest

to be obtained.

cost models from a limited

to
predict
use
subsequent

For example,

the development and launch of a limited

product range would allow early profits

levels

Therefore,

general arrangement drawings prepared must be for these

the initial

potential

is

Hence in terms of the New EHBdevelopment cost estimating

equations must be constructed

units.

to develop cost models it

this

data range and their


range is not recommended

the causal effect

in Table 39 which lists

of the total
the accuracy

developed
the
0.8
tonne and 3.2 tonne
using
models
of cost
units.

278

COST/HOISTCAPACITYRELATIONSHIPS
DEVELOPED
USING
THE 0.8 TONNEAND3.2 TONNEUNITS

TABLE39

Equation 28

Type of

Component

Cost

Estimating

Coefficients
x0

AccuraSy
xi

Es

(i's)

Brake Assembly

Material

13.7

28.8

Gearbox Assembly

Material

18.0

Labour

is

Dual Wound Motor


Cage Motor

54.0

-0.036

0.038

22.3

4.3

-0.056

0.060

Labour

269.2

245.1

-0.024

0.040

Material

196.5

114.8

-0.018

0.036

Material

33.4

51.5

0.121

0.123

-0.014

0.272

-0.001

0.022

Material

Rope Band
Moulding

0.077

Material

Hoist Block

Squirrel

0.049

0.333

Tool for

Tooling

Rope Band

5400

Comparing Tables 38 and 39 indicates


is
reduced.
models
still

0.333

within

500

that the accuracy of the cost

Although the accuracy levels

shown in Table 39 are

accuracy levels

the present estimating

of H.M. Ltd. this

Hence
be
the
is
assumed.
use
of
cost
always
models
restricted
not
could
in
do
which
costing
development
programmes
requirements
not conflict
to
with

of the marketing function.

those

Cost models are also useful


designs.

in the cost optimization

For example, the wire-rope

the wire-rope

during operation

of product

drum tube is used to wind and store

of the hoist

block.

The outside diameter

279

drum
is
tube
the
machine grooved to prevent adjacent
steel
of
mild
and hence causing premature rope wear.

together,
rubbing
strands
rope

The main dimensions to be determined during the design process

are:
loe

(i)

length

-(ii)
and

(t)
thickness
wall
(Ld)

(D)
diameter
outside

(iii)

These will

in
detail.
be
considered
now

Wall Thickness (t)


The wall thickness
since it

required

should be sufficient

hoist
the
maximum
under

load.

is related

to the hoist capacity

to prevent the drum tube buckling


The required wall thickness

be determined from the appropriate

can

crane design standards(179).

Since the present example looks at the design of the 3.2 tonne
drum tube the wall thickness
it

can be considered constant.

Hence

is not a design variable.

Length and Outside Diameter


Since the drum tube stores the wire-rope
operation
rope.

it must be capable of containing

The relationship

during hoist

the full

length of

between length (ld) and diameter (D)

for a hollow tube is given by Equation 39.

280

H. N. 1.125G

Ld =

(39)

D
ir.

Where:
(Metres).

H=

Height of Lift

N=

Number of Rope Falls.

1.125G = Rope Groove Pitch (where G is the rope.


diameter in millimetres).
Drum tube length allowance to allow the

locate
to
the rope on the
guide
rope
outer rope grooves of the tube,

be obvious

It will
the tube length

required

However, as D increases
also

increases.

dimensions

(a)

that

as the tube diameter

to hold unit
the cost

length

per unit

as Figure

In addition,

influence

(mm's).

the costs

of other

increases,

of rope decreases.

length

of drum tube

48 indicates,
components viz:

The length of the drum tube influences


of the drive shaft,

tiebars,

drum tube

the length

rope anchorage and

support tubes.
(b)

The diameter of the drum tube influences

the

diameters of the rope guide assembly and drum


tube end plates,

and the gearbox reduction

required to provide the specified

hoisting

ratio
speed.

28.1

COSTINTERRELATIONSHIPS
FORTHE
ASSEMBLY

FIGURE48

NEWEHB

a)
v'

a:

ai
Y
(10
i

X
O
t:
Q)
CJ

.n
FE
L
0

co
aV
n
0

CL
0
w

U
0
r--

4-3

V-

E
0
4-)
4-)

"r

0
co

>

a)

rn
c
rts
If.
+1 (1)
O
a)
C
CL U
= (0
OC
OrM: d
Q: Q

N
S.
C)
. fI-

Motor
Brake

Gearbox
*

*
*

Drum Tube

Rope
Rope Guide
Bottom, Block

Drive Shaft

Tiebars

Support Tubes
Mounting
Plate

*-

Cost Relationships

282

Since both D and Ld influence

the total

components the problem of cost optimization


designer is to determine the Ld: D ratio

cost of these
faced by the

that minimizes total

Costs.

Using regression

analy
analysis

the relevant

cost relationships

le
determined:
were

Components

Cost

Cost

Identification

.
Drive

Shaft

+ Tiebars

Relationship

Support Tubes + Rope

Ct, = 56. OLd

(40)

Anchorage

Ct2

Drum Tube End Plates


Rope Guide Assembly +

Ct2 = 46.0 + 143.94D (41)

Gearbox Casing

Ct3

Drum Tube

Ct3 = 6.17 + 0.00233

(TT(D2_(D_2t)2)Ld)(42)
4

The total

(Table
40)
were determined over a range
costs

lengths.
tube
drum
of

283

DRUMTUBELENGTHS
TOTALCOSTSAT VARIOUS.

Table 40

Drum Tube
Diameter (D)

Drum Tube
Length (Ld

Total
Component Costs ('s)

Costs

(mm)

Ct1
.

Ct2

Ct3

Ct1 + Ct2 + Ct3

0.458

39.20

111.82

42.41

193.43

0.423

42.00

106.89

41.92

190.81

0.80

0.393

44.80

102.57

41.49

188.86

0.85

0.367

47.60

98.83

41.11

187.54

0.90

0.343

50.40

95.37

40.64

186.41

0.95

0.324

53.20

92.64

40.44

186.28

1.00

0.306

56.00

90.05

40.14

186.19

1.05

0.290

58.80

87.75

39.87

186.42

1.10

0.275

61.60

85.58

39.54

186.72

(M)
0.70
0.75

oo,

Using the total


(Figure
cost curve
that

Ld: D ratio

(i)

in Table 40 the minimum'

49) has been constructed.

"minimum" costs are incurred

(i. e. 2.6 - 3.3).


actual

costs listed

The curve indicates

over a range of Ld: D ratios

Factors to consider when deciding the


to adopt. are:

The minimum drum tube to rope diameter ratio


design
the
hoist
under
relevant
permissible
(180)
Using these standards the
standards.
drum
tube diameter is 320mm.
minimum

284

FIGURE49 TOTALCOSTSVERSUSLd: D RATIO FORTHE-3.2 TONNEEHB

a
4-)
Q)

(0
r

~OM

M
M

.-O

E
:3
E

-0
+
J

"r

"

"
'O

cn

C)

LO

O"
O

"

M
tu

JCi
CD

I^

er
"O

ko
N

LU,
LLC)

IIIII1III

0
rn

LO
Q1

F-N
E
.
ipW
s-I

n
-P
Vf

oI

U
C.)

Ln

Co
r

.r.
4-)
ro

vo
J

cm

285

The maximumlength of hoist

(ii)

block acceptable in the

(an increase in the length of the hoist

market

block reduces the working area of the hoist).


Since the market survey(26) did not determine the
maximumhoist

length acceptable in the market it

is

necessary to choose a drum tube length at the low


end of the minimum cost area.
As the drum tube diameter is reduced the lifting

(iii)

speed for constant drum tube revolutions


decreases.

per minute

Thus to maintain a constant lifting

speed the gearbox reduction

ratio

In order to lower the reduction

must be lowered.
ratio

the number of

teeth and outside diameter of the spur sears must be


reduced.

However reducing the outside diameter of the

spur gears forces the gear centre lines


together.

Therefore,

the limit

closer

to the minimum

drum tube diameter which is feasible

occurs at the

point where the gear centres cannot be moved any


nearer to each other.

5.2

Costing using the Pareto Distribution


If the cumulative

its
that

cumulative

percentage of an element is plotted

percentage effect

the resulting

curve will

distribution,
Pareto
The

then the Pareto distribution

be of the form illustrated

although applicable

against
predicts

in Figure 50.

to many areas of industry,

286

FIGURE50

THE PARETO
CURVEOF CUMULATIVE
NUMBER
OF STOCKITEMSVERSUS
CUMULATIVESTOCKCOST

,or

80

..
v

60

G)

G)
>
.r

fl7
r

40

c.)

20

11i

-1

20

40

Cumulative

60
Item Count (%)

80

287

used to design selective

is principally

only the minority


(181
of stock costs

stock control

systems which

of stocked parts that account for the majority

control

The equation for the curve shown in Figure 50 is of the general


type:
i

Cstk

9
f'Nstk

where for stock control

Cstk
Nstk
f, g

(43)

purposes:

cumulative

value of stock costs.

cumulative.

number of stock items.

constants

the values

of which depend on the

form of the curve.

In terms of costing products Equation 43 finds


the
electronic
-in

and aerospace industries

where the evolution

design of a product to a stable configuration


manufacturing

of the

takes place during the

In these circumstances Equation 43 also


of learning(182)

represents

the effect

to predict

the reduction

therefore
and can

in manufacturing

be used

costs that arises due to

,
an increase

in the number of products manufactured.

cumulative

is
be
log-linear
to
assumed
curve
cost

unit

stage.

a wide application

determined
be
Hill(182)
to
according
can
cost

Whenthe
the resulting
using Equation 44.

288

(Nu
C=f
hh

1+9

1+g
Nu
h-1 J

(44)

Where:
= The cost per unit

Ch

The cumulative

Nuh

for the hth unit.

units
number of

= h.

Nuh-1 = h-1 units.

Examination of the component costs of a wide range of products


that
followed
Ltd.
H.
M.
component
revealed
costs
manufactured at
Pareto distribution.

typical

It could be expected therefore

determine
the
total
to
be
43
used
Equation
could
if

the coefficient

and Slope respectively

C-intercept

that

costs of a product

f and indice g could be determined.

log-linear
be
to
is
assumed
equation

Since the

f and g would represent the


of the log-log

straight

line

curve

45.
Equation
of

(45)

Log10C = Loglof + gLogloNu

Where:
C

(S's)
Cumulative
costs
product
=
.

Nu

Cumulative
of
components
number
=
.

Therefore

using only a small proportion

both
C-intercept
the
product
a
components of
straight

line

curve may be determined.

of the high cost


and the slope of the

Hence if

the total

number

289

a product are known the total

of components within

cost of a product

may then be estimated.

The obvious advantage of this

method is the reduction

in the total

to be accurately

product costing

number of components that need

costed.

However this

method when used to cost various products manufactured

at H.M. Ltd. resulted

in large inaccuracies

being obtained.

That is,

be
to
found
three
to
times
than
up
were
greater
actual
costs
estimated
product costs.

A model of the type expressed by Equation 43 was found by Swan


to be useful in solving

and Worrall(183)

problems.

organisational

to
found
account
were

for

be improved.

sales

could

using

Equation

various

industrial

For example, a small number of shoe models


the majority
The "over

43 can be explained

that the slope of the straight

of annual

costing"
since

line

sales

effect

indicated
's
Ltd.
M.
H.
products
of

Determining the coefficients

Swan and Worrell

greater
seriously

when

indicate

curve obtained from Equation 44

a similar

effect

Examination

(Figure 52).

f and g using various percentages

(10-50%) of Nu did not significantly

than fifty

hence forecasting

experienced

(Figure
Nu-axis
the
51).
distance
along
decreases at a

developed.
of the models

and

improve the estimating

In addition

developing cost models using

percent of a product's

detailed
the
of
amount
reduce

accuracy

components would not


cost estimating

required.

290

51
-FIGURE

LOG-LOGGRAPHOF THE PARETODISTRIBUTION

Ref.

(183)

"- C-axi s
%rber

?Wotiows

at

CTatirr

(i. e.,

""

t"l,

: "1s2.3"1":

s3. etc. ).

soo
I,

I
I.

ISO

dc

too

dNu

/;, i
;J
ed

J6

10
t_34S6

7$

9 10

IS

"

:o4;

rwy

"Organizationproblems. Croup priorities.

rri.,

rity

Nu-axis

291

FIGURE 52

LOG-LOG GRAPHOF CUMULATIVECOSTS


FOR H. M. LTD'S PRODUCTS

100

50
N

4--l

rn

4-J

U
a)

-.

10

't
.

Q)
>
"r
4--3

r6

t.
- ---r ....

1
5

Cumulative

A-1.6

Tonne New EHB

B-3.2

Tonne New EHB

C-0.5

Tonne Electric

10
Percentage

Chain Hoist

50
Components

100

292

Values for coefficients


estimating

f and g that produce an acceptable

accuracy are difficult

dc (Figure 5)
Uffu

to determine since the point

at which the slope of the straight

be
predicted.
cannot
changes
that the relative

However, it

values of f,

g and Nu affect

a relationship

the position

levels

of point

and values of f, g

In order to test this

may exist.

determine the relationship

hence
quantitatively
and
assumption
involved,

curve

is reasonable to assume

dc Visual examination of previous error


.
BNu
and Nu indicated

line

actions were taken:

the following

H.M. Ltd. products the values of f and g

From existing

c>>

10
determined
highest
the
using
were
cost components.

(ii)

Using the regression


between total
and f,

product

logarithmic
regression

the relationship

(the

cost

(the
Nu
and
predictor
g

To ensure that

(iii)

technique

the variables

dependent
variables)

variable)
was examined.

are linearly

form of each variable

related

was entered

the

into

the

procedure.

Various combinations of independent variables

were tested

in order to determine the model with the highest


estimating

accuracy.

Log10 C=

Log10 f+g

Log10 C=

Log10 f+g+

Log10 C=

Log10 f+

The combinations tested were:

Log10Nu
Log10Nu

...............

Log10g + Log10Nu

(46)

293

Using the method of determining


previously

(S and E) an expression of the form shown in Equation 46


the highest correlation

was found to yield

accurate for product costing

had
errors greater
estimated
costs

following
the
causes
possible

since 401oof the product

than +15%. When the products


accuracies were examined for

producing unacceptable estimating

(i)

between dependent and

However the model developed was not considered

independent variables.
sufficiently

equation accuracy developed

reasons were found:

The highest cost items used to determine the regression


were found to be inappropriate

coefficients

since the

those
of sub-assemblies and not of
were
used
costs
individual

components.

Hence cost items need to be of

a commonform.
(ii)

Whendetermining
circlips,
split

Nu items such as washers, screws,

dowel pins,

locknuts,

included.
been
had
pins

precision

nuts and

However, such items

are normally considered consumables and are therefore


included

in H.M. Ltd. 's indirect

overhead costs.

the inclusion

of such items when calculating

to inaccurate

estimates being obtained particularly

products that require

Hence

Nu can lead

large numbers of such items.

in

294

(iii)

f represents

The coefficient

expensive component within


regression

the cost of the most


Although the

a product.

procedure estimates the value of f it was

found that using the actual value of the highest cost


component improved estimating
Upon correcting

accuracy.

the input data the costing accuracy of

the subsequent regression model (Equation 47) was


improved.
f0.9948 101.209g Nu0.162

C=0.5916

(47)

Comparisons of actual and estimated costs using


Equation 47 are shown in Table 41 and Figures 53 and
54.
Table 41 shows that:
65.2% of estimated product costs are within
91.3%
8.7%

11

11

11

11

11

11

If

11

11

11

Although the small Proportion


levels

it

+10%of actual
+15% "

+15%

11

11

to

of products with high error

It is therefore

concluded that although

Equation 47 has improved the estimating


precisely

since

it

were re-examined no obvious reasons could be found

for these levels.

overall

is

values

reflect

the point of deflection

accuracy is however within


E= -0.015 and S=0.096.

accuracy it

cannot

of the slope.

H.M. Ltd. 's current

The
limits,

295

COST ESTIMATESUSING EQUATION47

TABLE 41

Nu

fa

Actual
Cost

0/

Estimated
Cost

Difference

57

27.9

0.441

99.1

106.6

+ 7.6

174

64.6

0.529

354.7

375.8

+ 5.9

58

7.6

0.682

60.1

57.4

7.7

7.7

4.5
0.0

0.471

522.8

529.7

+ 1.3

25
175

,1 .50.594
107.1

29

1.8

0.639

10.2

10.5

+ 2.9

178

198.1

0.453

907.3

931.1

33

1.9

0.642

11.5

11.9

+ 2.6
+ 3.5

40

20.0

0.519

80.2

89.7

+11.8

67

152.5

0.814

1637.2

1659.6

+ 1.4

37

80.4

0.798

732.3

769.1

+ 5.1

27

24.3

0.810

226.4

229.1

+ 1.2

19

36.5

0.445

117.4

117.5

+ 0.1

19

48.7

0.408

142.7

141.3

1.0
-

25

7.4

0.247

16.4

14.5

202

63.0

0.397

279.0

280.5

-11.6
+ 0.5

40

74.0

0.653

434.6

478.0

+10.0

26

46.5

0.440

169.6

154.9

29

4.6

0.509

19.3

19.3

8.7
0.0

79

36.5

0.835

434.6

122

240.10

0.522

12

23.2

127

436.5

+ 0.4

1381.0

1288.2

0.369

57.9

56.2

6.7
2.9
-

339.2

0.580

2220.4

2138.0

+ 3.7

12

29.0

0.360

69.8

69.2

0.9
-

291

78.7

0.521

648.8

486.4

45

5.6

0.535

29.7

26.9

-25.0
9.4
-

60

8.4

0.666

70.8

60.8

-14.1

81

62.0

0.311

207.8

173.8

31

1.0

0.732

7.7

8.1

-16.4
+ 5.2

34

15.4

0.223

30.0

29.5

1.7
-

63

9.2

0.67,1

80.0

68.4

-14.5

296

TABLE41 (Cont'd)

Nu
-'

Actual
Cost

fg
-

Estimated
Cost

%
Difference
+10.7
+ 2.5

39

140.6

0.664

840.8

931.1

73

0.526

16.2

16.6

60

3.3
"9.5

'0.727

95.0

81.3

20

2.9

0.676

15.7

18.2

-14.0
+15.9

29

0.7

0.858

6.8

7.4

+ 8.8

22

1.1

0.671

5.9

6.6

70

0.6

0.799

6.8

6.4

+11.9
5.9
-

77

9.7

0.719

106.9

84.7

91

15.5

0.578

83.8

93.3

566

45.8

0.546

390.8

338.8

63

11.2

0.785

138.7

113.8

31

1.0

0.617

6.9

6.0

202

69.0

0.383

244.9

274.2

70

27.5

0.590

166.4

164.1

52

67.4

0.509

315.6

305.6

-20.8
+11.3
-13.3
-18.0
-13.0
+12.0
1.4
3.2
-

297
FIGURE-53

ESTIMATED
ACTUALVERSUS
PRODUCT
COSTS <300)

250

i/f

200

W
v
(A
4-)
t/f

0
V

150

4-)
U
0
C3-

100

r
t
7
4-,
U
Q

50

50

FIGURE 54

100
150
200
Estimated Product Costs (i's)

250

ACTUALVERSUSESTIMATEDCOSTS (>E300)

200
N

N
d-3
Vf
0
U
4--3
U

150

-0
O
i
CS.

+I
U

100

50

500

1000

1500

Estimated Product Costs (Its)

2000

298

CHAPTER6

OF A DESIGN SCHEDULINGMETHOD
THE DEVELOPMENT

299
6.0

Introduction
The scheduling of components through the design process is similar

to the sequencing of operations within

a machine shop where the

approach to developing effective


(184)
described by King
centres upon:
theoretical

(i)

Specifying

an objective

scheduling rules

criterion

by which alternative

schedules can be judged.


Developing a method of ranking components in the order

-(ii)

with which their


(iii)

design should be completed.

Developing

a method of determining

importance

of each component to the overall

of a product

thus enabling

to be effectively

design

success
resources

allocated.

scheduling the design process requires

In addition,
determining

available

the relative

the
probability
reducing
or

affect
could
these
changes
since

a method of

of design changes occurring

any design schedules developed.

is analogous to the problem of re-working

defective

This

components during

manufacturing.

6.1

Specifying

the Objective

The conflicting

Criterion

aims of the various

departmental functions

involved

development
process can make the task of specifying
in the new product
an objective
financial

criterion

difficult.

For example, a major concern of the

department is the minimization

the manufacturing
time of a product.

function

of product cost whereas that of

could be the minimization

of the manufacturing

300

However, the time available


by marketing policy.

was limited
therefore

the minimization

The choice
importance

of this

of development
e
(23
11

in the market place


on its

profitability

cost effectiveness

The objective

Sequencing

criterion

chosen was

of development time.

obective

criterion

reflects

time to the ultimate

the growing

success of a new product

However the success of a product also depends

and-design optimization.
of any reduction

on design optimization

to the effect

6.2

for development of the New EHB range

Therefore the total

in development time must be related


and profitability.

Component Design

The sequence with which components are designed is important to


the minimization

of development time and forms the basis of conventional

project

PERT
CPM(80'83).
techniques
as
and
such
planning

correct

design
requires
of
component
sequencing

time required

a knowledge of the total

to develop each component, i. e. the sum of the individual


development
the
stages, and the interof
at each

times required
relationships

Accordingly

that exist

between various development tasks.

involved in the new


large
components
normally
of
are
numbers
Since
it
is
to remove the dependence of
necessary
development
process
product
development
data
from
time
to
the
component
request
need
designers on
functional
other

departments.

301

6.2.1

Estimating

Component Development Times

Since the tasks required to develop each component within


a product are a direct

result

of the design process it

is

reasonable to assume that the development time per task can be


related

design factors.

to specific

In order to develop a method

component development times it

for estimating

the design criteria

identify
to
necessary
development task.

Hence ascertaining

is therefore

associated with each

the development time per

task would allow designers to directly

estimate component

development times.

From discussions

relevant

times
average
and
maximum
minimum,
the various
is possible

for

design

involved

determine
hence
and

The tables
all
it

(Table

designer

the appropriate
the total

Using Table 42 it

development

development

time

tasks

the interactions

required

per component.

to ensure that
Initially

however

between activities

in order to determine those tasks that can be carried


simultaneously

for

to examine a component

development tasks are considered.

is necessary to identify

the

42) were estimated

can be used in the form of a checklist

relevant

of the

Chain Hoist),

in the NPD process.

an engineering

and identify

departments

and examination

(i. e. Electric

and past NPD projects

tasks

functional

Department),

the Purchasing

(particularly
present

with

and that would thus reduce the total

out
development

time of a component.

In this

"critical
would allow

path" components to be identified.

way a knowledge of such interactions

302

NPDACTIVITY TIMES

TABLE42

rN

CO

LC)

00

NO
rd

tD

to

rN

rr

LC)
r

O
LD

Lt)

re)

c+') O

(Z
LC)

O
00

LC)

F_ W.
vz.
4 H

Lt1
n

.zr

CO

LC)

tC)

rd

Z
r
Z

CD

L1

Mr

ON

rr

NO

Lo Ln
r

11
U

U
O
.}
Vf

N
4-3
(A
43
t

a
+

CL

EN
a) Y

O
CL

"-

CCCC
L
"r
Q)
t
"r
3

XrX
a)

r-tio vv

-0

"r
t
4-)

S..
Q)
a-

a)

a)
>
. r.
4J
. r.
4-3
(L)
Q
E
O
V

.
tu
.v

N
N
r0
O

O)
N
4)

tu
+1
de
-v
c
"r
N
d)
-0

rn
.
i
-0
(L)
4-3
O
O
L

"r

S.

"r
cr
u
Q

cr

u
Q

+)

0
-p

C)
"r-

a)
"

C-

-C.
C_

r0

.GU

d
(0
+)

a)
NtNtC
a-1
r'
fO
C
"r
L
"r
4)
.C
4-3
r+36
3

"
N

C
fC

a)

C)
N

e-+)
CCC
G!
L

a)
Y

G)

a)
.

S.
C)

r-mmm

C]

r-"
fCT
"r
Si
(U
4J

4J
C
()

>1
r-+)

4-J

NM

sC)
"r
r
CL
C-

'fl

N
r"

EL

sC)
"r
rCC_

+1 0
Kn -v

R)

a)
tN

a)
L.

(A
C)

4)

C
(0

0
O.

L)
0

vvvv

a)

CN
"r
L
O
"3

0
0

t0
"r
C

4)
C)

O
C

-. CU . -. . -.
.e6

"
"p

+-

Uf
Y

4--)

C)
SSCCC
t)
NN

U0U

N
C
0
N
"r
Sco
a"
E
0
V

C1
vf

VN
o
a)

a
CL
10
EE

h
N
--

YN
VIC.
s=
O
OC
+
NCNN

L
4-)
.

'Q

r0
vvvv

N
,G

0
C

O
M

7
VCD.

303

TABLE42 (Cont'd)

X
Q

O
d'

LP)
r

U)

U)
r

LC)

LA

LO

N
U)

M}-4
rr

f- LLJ
U=
cc P-4
F"

00

r
r-

U)

LO

Y
U
0
4-3
t/N

-v

V1

>

Nrr

NO
Y
UN
0O
O
N

.0

LL
NU
"

"r

aa
s.
aa
UNN
tu

-0

Z7

c
0

co
r
N

(A

C)

Co

v)
C)

Lin
N

c.
O
i
C1

lp
MM
N

01
04

01

er
O
NN

ww
a1
Crr
ONN
d

. -.

vv

ww

o
l. i
Vf

a
LM

ww
C)

NN

., _

c
b
a
.
L

rd

3
rt'

aJ

a)

Vww

4-3
V1

"r

N
OO
NN

aJ

0
U

a,
CD
CNN

"r

a)
F..

r
R:
m

Up

01

ww

t0

-0
v

Q.

N
V

rc

N
G

304

Maximumand minimum times for each development task have


been included in Table 42 to enable designers to gain. a realistic
outline

of the limits

of the possible

development time required.

Although Table 42 provides a quick estimation

of development task

times the actual development times should be obtained when


for
potential
searching

material

suppliers.

If the cumulative percentage activity


tasks is plotted

against number of tasks a Pareto distribution

(Figure 55) is obtained indicating


Stock Control policies)

the possibility

(as with

of ranking components in order of

length of development time required.


identify

time of the various

This procedure would

those components that could form the "critical

isolate
to
be
it
possible
Hence
would

the critical

path".

components

be
times
development
must
estimated.
for which accurate

305

FIGURE55

AVERAGE
TASKTIMES VERSUS
CUMULATIVE
NUMBER
OF NPDTASKS
CUMULATIVE
0

Lr)
M

LO
NN

Y
t/1
iJ
G
d
Z

4O
i.
a)
.0

LC)
a)

. r..

O
r-

LO

CD
Co

C)
t.0

C)
cr

CD
N

(%) sawl Jsel anTpLnwno

306 -

In addition

to the development tasks listed

there are other NPDactivities


difficult
are
which
factors

that affect

in Table 42

(Table 43), the task times of

to obtain from past NPDprojects


these tasks vary considerably

since the

between

projects.

TABLE43

.le

AFFECTINGPROJECT
ACTIVITY TIMES
FACTORS

Influencing

Activity
1.

Prototype

(a) parts marshalling


(b) assembly

Factors

Number of standard sizes,


complexity of product design
and type of products

2.

Pre-production

(a) parts marshalling


(b) manufacture
(c) assembly

3.

Launch (a) parts marshalling


(b) assembly

(c) stock build


4.

Launch quantities required,


sales forecasts, product type,
design complexity
Sales forecasts,
design

product

type,

complexity

up

Test area organisation

Type of product, type of tests


carried out, length of testing
period

5.

Plant layout/rearrangement

Level of changes required,


of plant layout adopted

Removal of obsolete stock

Present stock levels,

type

rate of

sales
7.

Design and Manufacture of


Production/Assembly aids

Sales quantity required, cost


feasibility
of manufacturing aids

307

6.2.2

between Development Tasks

Interactions

The two types of interactions


scheduling(55)

(i)

involved

in project

are:

Precedence Interactions

Task
A

Ii

Task
B

F--*

Task I-a
C

Here task B is preceded by task A (which must be


and leads to task C (that must be

carried

out first)

carried

out next).

are therefore

scheduling objectives

to determine the order in which tasks

should be carried
required

The project

out and the total

is the sum of the individual

development time
task times

involved.
(ii)

Simultaneous Interactions

Task
A

Task
B

Task
c

-01

Here tasks A, B and C can be carried


(i. e. in parallel)

and it

out simultaneously

is the task that requires

the

longest time to perform that is used to determine the


total

development time for a component.

308

It was possible to classify

the tasks involved in the

New EHBdevelopment process using the two types of activity.


interactions

and hence construct

a network diagram (Figure 56)

for the development of individual

components.

Task interactions

6.2.3.
in
Section
described
are
'

An element not shown in Figure 56 is the need for an

information

feedback system from each of the NPDtasks to the

design stage since the design function


corrective

6.2.3

action is required.

Interactions

Types of Activity

(1)

should be involved if

Design Tasks of individual


be carried

sub-assemblies can normally


(except when combining

out simultaneously

sub-assemblies)
available.

if

In general however lack of sufficient

resources requires

the design of sub-assemblies to

In addition

be sequential.
component within
interaction

design resources are

sufficient

the design of each

a sub-assembly is a precedence

since this

is the basis of creative

design.
(ii)

Procurement Tasks can be carried


involved
time
the
since

out simultaneously

in placing

is minimal when compared with

procurement orders

the delivery

period of

the supplies ordered.

That is, procurement task times

are basically

periods"

simultaneously

"waiting

since they require

development resources.

which can occur


small amounts of

309

NPD NETWORK
DIAGRAM
COMPONENT

FIGURE 56

4"
a
v

"r
4-J

.r
4J
C.I
fC3

CCC
O00
41

(L)
CL
>,
}
41

OOO

N
--

-0

O.
"
iLi

"r-

""-

"OII

0
i

td U
UUUi
rN t0

Cl.

rn
c
."V1

t6

": -) i

"r'
CGN

F-

Otz
Z"r"

a
S.4)

u`
-1b.
(0
4.,

rn
.
N
ei
--'

(A
N
F"

rti

mO
N

Q.

U). RS
-0E
+1

(41 4-)
t3l
td
+i
r-

NN

CL
nG

ZbZU

310

(iii)

Checks (Incoming Supplies)

Quality

since the time required

as simultaneous activities
to test materials,

can be regarded

is small

components and tools

when compared with the overall

development period.

Componentsthat determine the critical


this

stage are those that either

path through
long term

require

tests or must be tested outside the

reliability
company.
(iv)

Prototype

to precedence

hoist

period

of the length

since

had to be carried
and bottom block

(v)

the prototype

EHB
had
be
New
developed
the
to
of
assembly

at the beginning

short

must be

of the prototype

For example,

required.

are subject

Sub-assemblies

constraints.

sequenced in order
test

of Sub-Assemblies

Manufacture

long term reliability

trials

out whereas the suspension


(hook)

term function

sub-assemblies

required

a prototype

only

testing.

Prototype Manufacture of Components within


Sub-Assembly.

frame

The total

time required

component consists

each

to manufacture

of a "waiting

period"

(the time a component must wait before a manufacturing


operation

(the time to

begins) and a "service-time"

carry out an operation).


in excess of "operation

Since "waiting
times",

periods"

i. e. measured in days

as compared with hours for "operation

times" the

latter

total

can be ignored when estimating

manufacturing

times.

are

311

Manufacturing activities

between components can

(even though more than one

occur simultaneously

the use of the same equipment,

component may require

since in these situations

all

"development components"

are normally batched for manufacturing

purposes)

unless components need to be processed together


lo,

(e. g. precision

machining and welding fabrications)


are imposed.

when precedence constraints


The "waiting

for each operation

periods"

on an individual

out

component must however be added

interactions
since precedence
incurs
"waiting
a
component
manufacturing operation
waiting

carried

i. e. the

exist,

between each

period"

required.

Hence total

time is the sum of the individual

"waiting

periods".
(vi

Post Manufacturing Quality


each operation

required to manufacture a component

and are therefore

subject to precedence constraints.

This is particularly
quality

Checks can occur after

important

to consider when
However between

checks are sub-contracted.

components, quality

checks can be regarded as

for
similar
simultaneous
of incoming supplies,

reasons to the testing

i. e. testing

when compared with overall


when reliability

trials

time is minimal

development time except

are required.

312

(vii)

Testing

Prototype
since

determines

it

are designed

are designed

of the prototype

assemblies

not directly

trials

reliability

The order

task

in which

is dependent on the
Those sub-

period.

involved

can therefore

design
the
end of

(viii)

test

precedence

in which sub-assemblies

the order

and manufactured.

sub-assemblies
length

is an important

in long term
be placed

at the

schedule.

The actual prototype

testing

be
carried
of course

out simultaneously.

of sub-assemblies can

Jig and Fixture

Design are precedence tasks when

limited

departmental resources are available.

tooling

The actual manufacture of jig

and fixture

identical
to
subject
are

task interactions

manufacture of prototype

components.

Whenjig

and fixture

are sub-contracted

components
as the

design and manufacturing


then they could be carried

tasks
out

simultaneously.
(ix)

Procurement of Pre-production
similar

to the procuring

are therefore

Materials

of prototype

subject to identical

are tasks
materials

and

task interactions.

313

(x)

Pre-production

Manufacturing

tasks are similar


That is,

tasks.

prototype manufacturing

to

the tasks

are normally simultaneous between components but


operations

carried

out on individual

components are

subject to precedence constraints.


Product Launch is a major precedence task since it

(xi)

could determine the order in which the standard sizes


within

a product range are developed.

the New EHBproject

is desirable

to develop a

product range (e. g. 0.8,1.6

limited
units)

it

In the case of

and 3.2 tonne


development

in order to reduce the total

before
that
elapses
revenue can be obtained
period
from the new products.

Accuracy of Estimated Development Times

6.2.4

It

is important to recognise that Figure 56 represents the

that
task
diagram
must
undergo
component
a
each
of
network
involved

in the development process.

design
will
a product

require

Not all

components within

every development activity.

For

hence
bearings
components
purchased
and
are
require
example,
manufacturing

and jig

and fixture

building

no

tasks.

Since the order in which product sizes and sub-assemblies


by major precedence constraints
determined
is
developed
are
(e. g. marketing constraints
period)

and length of prototype

testing

the designer need only consider the order in which

314

components within

sub-assemblies are designed.

In addition

the order in which these components are designed is determined


by the amount of development time required to prepare components
for prototype

Hence the use of Figure 56 in conjunction

assembly.

with development activity

development time of each component within

estimate the total

Examples of development time estimation

design.
product
illustrated

times (Table 42) allows designers to

are

in Figures 57 and 58.

Upon completion of the prototype testing


drawings are the tools

stage engineering

that allow development tasks between

be
to
carried
components
individual

out simultaneously

(although for

components each development task must be carried

out in the correct

order) except when components need to be

together.
manufactured

The validity

development times (using

of estimating

Table 42 and Figure 56) was tested using the actual development
times for the 0.8 tonne and 3.2 tonne New EHBcomponents.
results

indicates
59
in
Figure
shown

The

that close agreement is

development
between
times.
and
actual
estimated
obtained
However a proportion
exist

of the points indicate

that differences

between estimated and actual development times.

large variances arise primarily

since in actual practice

been
have
that
carried
could
tasks
i. e. development activities
total
minimize

These

out simultaneously

were not,

were not scheduled such as to

development time.

many

315

FIGURE5.7

DETERMINATION
OF DEVELOPMENT
TIMES

- PARTSA AND B

PART B

PART A

DESIGN

GEAR
OUTPUT

BEARINGHOUSING

i.aterial
Procure B2a

Procure
Metric B3c
Hobber 30

__

Procure
Metric B3c
Cutter 30

4r

Examine

1I.

Dla 1

Procure
Material

4r--

.+
nle

JTurn,

Turn
,,.

Faro
MV\.

1Cut Gear
Clb 15
Teeth 217

9A

217) C1b15

Examine Dlb
f
..A
1
1/

Examine D2 5

ri..

Bore 2fl4 Ala iu

Cut
Splines

Face
Bore(202) Cla 10

Bla
0.5

Task and
Task Code

t
I
t
L_

Task
Tine

Surface C2 10
Harden

Examine D2 5

Total = 76 days

PROTOTYPE
ASSEMBLY
AND
TEST

Total = 56.0 days

316

FIGURE58

OF DEVELOPMENT
TIMES -- PARTSC AND
DETERMINATION
D

PART p

PART

-C',
GEARCASE

Examine

Total

DESIGN

ENDPLATES

Dla

days
122
=

PROTOTYPE
ASSEMBLY
AND

TEST

Total = 47.5 days

317

The cumulative distribution

of development times

estimated for each standard unit were found to form a typical


(Figure 60) indicating

Pareto distribution
level
the
reducing

of managementcontrol

by
"management
exception"
a
directing

the possibility

of

resources by operating

This would involve

policy.

managementresources to the control

of those

development
longest
times.
the
with
components

6.3

Determining

the Relative

Importance

According to Ostrofsky(56)
development of alternative
these solutions
this

of Components

engineering

design solutions

design requires

and the comparison of

in order to choose the most appropriate.

However,

to adopt in projects

design methodology is difficult

with limited

is
there
frequently
design
since
resources
development and
sufficient

design time to develop an initial

solution.

design methodology.

alternative
to
an
adopt
necessary
involve:
would
such a methodology

(i)

developing the initial

(ii)

determining

only

Hence to

design resources are used effectively

ensure that available

the

it

is

The stages in

design solution.

the benefits

to be gained by allocating

improving
design,
to
design
the
resources
extra
(iii)

determining
benefits

and

(iv)

the relative

to the overall

use of this

relative

design resources.

importance of these extra


success of the product,

importance to allocate

available

318
FIGURE59

TIMES
DEVELOPMENT
VERSUSESTIMATED
ACTUAL.

100
N

80
E
"r
4-3
cl
UJ

60

O
N

40
4-)
U
Q

20

20

40
Estimated

FIGURE60

ae

60

80

100

Development Time (Days)

DEVELOPME.
1T TIMES
CURVEOF THE 0.8 TONNEEHBCOMPONENT
PARETO

80

a)
E
t+J
c
a)
E

60

0
a)
a)
a)
>
"r
r0
r

40

20

80
40
60
20
Cumulative Number of Components ()

319

Determining the relative


necessity requires

importance of design solutions

a commonbasis of comparison with which to relate

designs that could perform quite different


planning technique

the PPBSproject

functions.

reduce manufacturing

market specifications).

cost,

However, manufacturing

to quickly

assess

To allocate
cumulative

system previously

manufacturing

the available

and widen

cost is the most


since the regression

developed allows designers

costs.

design time the Pareto curve of

product costs versus number of components is useful since

the cumulative

cost axis can be converted directly

The amount of design time to allocate

time.

(e. g. improve

improve quality

convenient for comparison purposes particularly


based cost estimating

Examination of

(Section 2.5) has shown that the

design process can produce many types of benefit


producibility,

by

direct
be
a
as
thus
obtained

to available

design

to each component can

measure of the relative

cost of a

component.

The Pareto distribution


allocate

could only indicate

to each component since factors

considered.

However it

the design time to

other than cost need to be

is an advantage to use such a chart since the

the
then
additional
weigh
designer must
design resources)
(in
terms
of
cost

benefits

required

to be gained with the

to achieve them.

320

6.4

Design Changes

Preventing

Design changes have been estimated

amount of design resources (2250 hours).

considerable
proportion

(78%) of design changes resulted

design cost effectiveness


ability

cost estimating

occur to be reduced.
be ascertained
functions.
benefit

it

That is,

the cost effectiveness


to their

estimates can be quickly

of designs can

release to other departmental

based estimating

system could be of

prepared and help to

on the design department to prematurely

design drawings in order that prototype

In addition

from the need to improve

would enable the number of design changes that

Hence the regression

reduce the-pressure

Since a high

is reasonable to assume that an improved

and improved prior

since cost

the possibility

for
releasing
not
reason
prime
releasing

(Section 2.6) to consume a

release

manufacture can begin.

of having to change designs is a


design work in a smooth flow.

design work in a smooth flow of "stabilized"

However,

component designs

do
in
high
that
development
workloads
not
arise
subsequent
ensure
would
jic
as
areas such

and fixture

manufacture.

The estimating

be
design
to
established
costs
developed allows
stabilized)

earlier

in the design process.

system

(and therefore

Hence the ability

designs
to

for
flow
in
subsequent development activities
designs
smooth
a
release
improved.
be
would

321

CHAPTER7

COMPUTER
AIDED NPD

322

7.0

Introduction

The growing use of computers to aid various manufacturing,


(185,186)
will in the future undoubtedly
planning and control tasks
include the NPDprocess. The immediate challenge according to
(187)
is to devise systems and procedures that provide the
Hancock
links

to integrated

manufacturing

systems since progress with sub-

systems has reached the stage where the majority

of the tasks involved

in manufacturing

industry

can be automated.

With respect to the development process it


that has received most attention

is the design phase

with the development of a wide variety

(CAD)
Draughting
Design
and
systems.
of Computer-Aided
been developed by both academic institutions

Systems have

and commercial enterprises

drawing
the
have
a
of
office,
needs
analysed
who

studied the manner in

CAD
(e.
design
Distributed
produced
packages
and
evolves
g.
a
which
(DIAD)(188),
Draughting
Applicon(189), Polysurf(190,191)
Automated
Interactive
(193)
)Comdraw
to
the
IV(192) and Ferranti
meet
majority of the require,
ments possible

in keeping with a reasonably priced total

H. M. Ltd. intend to install


although
with

an Autotrol

AD/380 system(194) which

not the lowest cost CADsystem available

the Ferranti

system.

(i. e. 150,000 compared

system cost of llO, 000), has the advantage of being

Ltd.
Computing
Davy
by
marketed

(DCL) who in commonwith H.M. Ltd. are

Companies.
The
Group
Davy
of
the
primary
advantage
of purchasing
of
part
the Autotrol

system is the ability

for training

purposes at a relatively

to use DCL's own CADsystem, EUCLID(195)


low cost.

H.M. Ltd. use the

Univac
DCL's
mainframe computer through an existing
on
EUCLID system

323

computer link.

Since the actual

plotting

of drawings is carried

have
H.
Ltd.
M.
DCL,
only to purchase and install
at
equipment such as

The Autotrol

an interactive

out

low cost peripheral

graphics unit.

AD/380 system however runs on a dedicated mini-

is
the
by
the user company.
of
package
purchased
part
which
computer
The system supports peripheral

storage/reader
libraries

and since it

and teletypes

digitisers

unit

equipment such as workstations,

facilities,

plotters,

has magnetic tape and disc

the system can compile in-house

of data.

The drawing language. of the Autotrol

system is similar

"every
day"
EUCLID
words.
and
uses
the
system
EUCLID can be accomplished quickly.

Features that make the Autotrol

Hence the changeover from

In addition

be
transferred
EUCLID
can
the
system
using

to that of

data files

to the Autotrol

system preferable

generated
system.

to the EUCLID

system are:

(i)

The ability

to zoom into or out from a particular

area

drawing.
a
of
(ii)

The ability

to draw in refresh,

commandis invoked a particular


around the screen and positioned

i. e. if

the refresh

feature

can be moved

as required.

324

(iii)

The ability

to invoke a series of commandsby the touch

of a single

key.

position

The AD380 workstation

has a 240

keyboard which can be overlayed by a menu card

and then touching one of the 240 keys invokes the set of
commandswhich have been assigned to that
there is no limit

key.

to the number of different

240
the
be
up,
position
set
can
which
considered as multi-layer

rather

Since

menu cards

keyboard can be
layer.

than single

Although the advantages obtained through using CADsystems have


level
the
documented(196,197)
been well
dependent on the specific
considers
(Table

it

implementation

TABLE 44

1.

A better
reduction

before

idea of the objectives

embarking

on the selection

OF USINGCAD
OBJECTIVES

Ref.

design.
improved
through
product
in design errors

a computer.
representation

(198)

CADmay result

in a

and in improved performance at lower

optimization

of the design parameters.

be
design
cannot
easily
which
problems
Solving

circuits.

and

of a CAD system.

better
through
a
costs

2.

achieved are
(198)
Hence Laxon

design application.

have
to
a clear
necessary

44) to be achieved

of benefits

Examples include finite-element

tackled without
analysis,

computer

integrated
design
and
computerised
of
surfaces
of

325

TABLE44 (Cont'd)

3.

Reduced design costs through a reduction

in manhours. The

replacement of engineers and draughtsmen by a computer system


can lead to substantial

savings in direct

increase in time as salaries

will

rise

These savings

costs.

and computers become

cheaper.
4.

design capacity.

A greater
full

productive

potential

Many companies cannot reach their


for a product because of a bottleneck
is impossible

If it

in the design office.

CAD
draughtsmen,
may offer
and
5.

Reduced design
stage

time.

to recruit

designers

a solution.

The use of CAD at the conceptual

design

lead
time
to
the
needed
get a new product
reduce
can

the market.
response

For detailed

to customer's

design

and faster

queries

Determining the benefits

CAD can result

onto

in a faster

delivery.

to be achieved is not straight

forward

during
NPD
the
CAD
an
aid
the
as
system
process,
a
since
using
when
of resources expended on the individual

relative

proportion

involved

in the design process varies

This is influenced

(i)

by such factors

The innovativenessof
original

considerably

activities

between NPDprojects.

as:

the product design.

the product design the greater

design
time
on
expended
creation
of

The more
the proportion

activities.

326

(ii)

The type and number of functions

the product carries

out.

Products such as automobiles and heavy cranes require


large amounts of stress,
(iii)

The complexity

of the design.

the amount of detailed

Hence for a particular

This factor

could increase

draughting work required.

NPDproject

it

in the design process.

is necessary to determine

The approach adopted during the New

EHBdevelopment process for achieving this


Activity
the
of
activities

and performance analysis.

amounts of resources expended on the various activities

the relative
involved

function

objective

involved the use

Sampling work measurement technique(113).

required

to prepare various assembly designs from their

the
to
preparation
conception
intervals
at random

The

of production

drawings were examined

in order to determine the proportion

of time

(Table 45) a designer spends on each type of task.

When examining the advantages of CADsystems to the development


45
Table
the
EHB
New
allows
relative
range,
of the
CADfacility

to be recognised.

importance of each

327

TABLE45

PROPORTION
OF TIME SPENTON DESIGNTASKS

Proportion
Function Analysis

0.46
and Transfer

Data Acquisition
e

of Design Time

0.23

Design Creation and Improvement

0.12

Cost, Marketing and Manufacturing Analysis

0.08

Draughting

0.06

Integrating

7.1

with overall

NPDprocess

0.05

Function Analysis
In their

basic form CADsystems are merely automated draughting

hence
do
have
the
data
and
systems
not
storage
capability
and
performing

the complex mathematical calculations

functional

and performance analysis

and can be included

modules are available

to allow advanced analytical

Unfortunately
engineering

for

However software

in the total

CADpackage

design to be performed.

the development of software in the mechanical

area has tended to be piecemeal in nature.

diversity
the
from
general
analytically

of designs.

required

of

predicting

of the topic
the overall

and the difficulty

This results
in

performance of a mechanical

the
its
in
terms
performance
of
of
system

individual

components.

In certain fields such as finite


element analysis(199) and
(200) "off-the-shelf"
However,
bearing design
software is available.
during the development of a new product there is often a need to
develop "customised"

software for specific

analytical

For example, the design of trim cages for controlling

of flow within gas valves


for four wheeled vehicles

(201)

requirements.
the velocity

and predicting stability

criteria

in the Railway Industry(202)

With respect to the New EHBdevelopment software would need to


be designed for:

rope twist

during lowering of hoist

(i)

Modelling

(ii)

Estimating

(iii)

Determining motor and brake sizes.

(iv)

Modelling

(v)

Determining rope diameters required.

drum tube thicknesses.

hoist deflection

under load.

However since such software would be useful


therefore
development
and
product
its

load.

infrequently

only during new


it

is doubtful

that

the
be
compared
with
when
present
manual
economic
would
use

methods used.

329

7.2

Data Acquisition

and Transfer

Since a major function


facilities

of CADsystems is to provide data storage


and transfer

the acquisition

be
to
be
greatly
expected
can
extensive data files

of data during the NPDprocess

improved, i. e. CADsystems allow

to be maintained and information

retrieved

rapidly.

-A wide variety

of data can be stored including:

product data,

(i)

existing

(ii)

competitive

(iii)

parts lists,

(iv)

material

(v)

details
component

product data,

specifications,
(e. g. material,

dimensions and

tolerances),

and

(vi)

manufacturing

(vii)

cost data,

(viii).

data,

component, assembly and product drawings.

The ability

to store and quickly

development of an effective
be
electronically
can
be filed.

This facility

retrieve

product design.

Since engineering

stored a complete definition


is of benefit

data aids in the


drawings

of a component can

since the capturing

by the

330

computer of the definition


instant

facilities

data transfer

for associated manufacturing

For example, the Autotrol

business operations.
package has the ability
used within

of a component at the design stage provides

a particular

to create a full
assembly.

The need to write

of the components

This information

when placed on

to the main frame computer


and product structure

out a parts list-remote

files.

from the drawing board and/or

into punchcards is therefore

transpose the parts

lists

Hence the interface

between the design function

established)

AD380 basic software

parts list

disc
be
transferred
tape
can
or
magnetic
Oe
handling such processes as stock control

and

unnecessary.

(where information

is

and the data processing department (where information

is processed) could be improved.

The amount of data stored on an engineering


significantly

increased producing benefits

Using manual methods, everything


occurs on a single
deal of information

written

drawing can be

to the engineering department.


or displayed

layer namely the sheet of paper.


which could be usefully

on a drawing
Hence a great

stored on the drawing has,

be
to
lack
stored using other means.
space,
of
because of

However the Autotrol


storing
utilised

layers of information.

different

by the designer,

designers.
Tool
and
engineers

AD380 system produces a total


layers

layer 5 could be made available

Layer 6 could then be allocated

to store data on production

information
of

Hence if

could then be compiled.

methods.

drawing by
1 to 4 were
to the Jig

to the production

A centralized

library

331

data it

To transfer

would be necessary to provide access to the

the
for
various
system
-CAD
that appropriate

functional

departments

involved

and to ensure

systems allow design changes to be made by

security

design personnel only.

functions

The manufacturing
it

individual
for
*take place
Presently

or a welder and a machinist,

information.
structured

However using the layering

technique,

drawing number system, 'it

is possible

drawings
at no extra
machining
and

also benefits

can be recalled

7,3

design.

of creativity

hence the
only
and an
to

to the designer.

machine tools

within

from CAD systems in that component geometry


of NC tapes.

and Improvement

is
design
essentially
Product
level

effort

controlled

for use in the preparation

Design Creation

to

and machining drawings or separate

The growing use of numerically


H. M. Ltd.

to

casting

drawings attempt to convey information

fabrication
produce separate
casting

the operations

or welder must search drawings and identify

machine operator

appropriate

from a CADsystem since

types of processes such as, welding,

both a founder and a machinist

relevant

benefit

to produce drawings reflecting

be
possible
could

and machining.

will

involved

process although the

is dependent on the innovativeness

The amount of creative

did
EHB
not represent
New
the

a creative

of the

design involved

in the development of

a major proportion

of design time since

332

the level

of product innovativeness

design is similar
a deliberate

was limited,

to that of several

policy

competitors

i. e. the New EHB


This was

products.

to enable development time to be reduced.

Design improvement also takes place primarily


synthesis although various optimization

by creative

techniques such as Linear

Programming(55) and Dynamic Programming(78) can also be used.


Indeed

it

difficulty

is through the use of optimization


in transferring

the creative

techniques

that the

element of design from a

be
illustrated.
task
to
task
can
computer
a
manual

'For example,

to the "assortment problem" (Section 2.3) indicated

the solution

that there are many subjective

involved

factors

into
be
taken
consideration
cannot
which

in the design process

when using optimization

techniques.

human
Masters(203)
to
creativity
According
design
in
projects
be
necessary
always

and judgement will

where:

(i)

the design work is not a mathematically

(ii)

there are many variables

and influencing

exact process,
factors

involved,
is difficult

to define when a design is "acceptable",

(iii)

it

(iv)

the number of decisions


intuition

to make are immense hence human

or judgement is necessary to solve a problem,

333

(v)

where the work being performed cannot logically

be

defined,
(vi)

where the problem in hand cannot be specified

to a

computer,
input
data
for problem specification
the
of
volume
when

(vii)

and

is large,

i. e. input specification

may be as great. a

task as producing the design manually.

Since the large proportion

of original

it
the
elements
above
one or more of
to contribute

little

Once a design solution

design problems contain

to the initial

can be seen that CADsystems have


creation

of design solutions.

is generated however computers are ideal tools

for design optimization providing


(204)
logically
be
can
and
correct

the optimization

procedure is

defined.

to CADsystems being of limited use to the creative


(205)
"the
that
Cooley
suggests
crude introduction
design process,
of
In addition

into
computers

the design activity

design
quality".
the
of
introducing
of
the

into
rate
to alter

in a deterioration

This argument is based on the known effects

high-capital
manufacturing

equipment with an increasing


environment.

the working conditions

high-capital

...

may result

of their

hours
24
per day.
equipment

design
the
impose
to
staff
on
seek
could
(205)) the following
conditions:
Bristol

Essentially

obsolescence
employers seek

workers in order to exploit


In consequence companies
(as occurred at Rolls Royce,

334

(i)

the acceptance of shift


capital

and

work in order to exploit

high

equipment,

(ii)

the acceptance of work-measurement techniques,

(iii)

the division

of work into basic elements and the setting

of times for these elements, such times to be compared


lle

with actual performance.

Overall

there would be a tendency for designers to become paced

by the machine and since CADsystems can produce quantitative

data at

designer
be
the
the
could
enormous.
on
high
stress
rate
a

Clearly

the conditions

environment
(206)
could
designer
a
a working

described

by Cooley(205)

in which factors
be generated.

level

that

would not provide

improve the creativity

These factors

are:

(i)

A designer's

(ii)

Adoption of a positive

(iii)

Adopting an open mind.

(iv)

Adopting an attitude

(v)

A designer having the courage of his own convictions.

of self

confidence.

attitude.

of constructive

discontent.

of

335

The use of sophisticated


detailed

functional

CADsystems capable of carrying

of designs could be expected to reduce

analysis

the design resources required to generate a design solution.


result

two strategies

resources required

are possible.

of the introduction

process the latter

strategy

can be reduced or the number


Given

would appear difficult

to achieve.

Cost, Marketing and Manufacturing Analysis

essentially

a data collection

this
of
the actual examination
a small proportion

of the total

development time available

analysis

of design work is

process with the relevant

by
the
appropriate
at
present
generated

departmental

data by the design function


design time.

restricts

In addition

the effectiveness

own cost,

Hence

represents
the limited

the number of design alternatives


marketing

that needs to be undertaken.

is the rate at which data can be generated that influences

Since it

to manipulate

analysis

data being

function.

hence
the
be
reducing
amount of cost,
generated
can

and manufacturing

their

the amount of design

of CADsystems on the creative

Cost, marketing and manufacturing

that

As a

generated could be increased.

the adverse effect

7.4

Either

during a NPDproject

design solutions

of alternative

out

of the design process, the ability

of a computer system

large amounts of data could allow designers to generate


marketing and manufacturing

delay to the design process.

data in order to avoid

336

system developed helps achieve this

The cost estimating

designers
with a powerful tool
providing

for generating

cost data for

designs as they evolve and hence enabling cost effective

their

to be more readily

involved

of manufacturing

data is difficult'since

the

The solution to the problem


(208)
is the establishment
and Coates, et. al.

are complex and varied.

Colding(207)
to
according
of manufacturing

designs

achieved.

The generation
factors

aim by

data bases or manuals.

The effectiveness

of data

bases would increase with the growing use of NC equipment since they
allow optimum manufacturing

"

The estimating

production

to be selected.

developed
by the author could form the basis
system
data base.

of a manufacturing
difficult

conditions

operations

be
removed.
avoid could

For instance,

the costing

that the manufacturing

the use of a particular

manufacturing

to be displayed.
times required
the display

wish to

department before

the product design.

In addition
appropriate

function

Format statements could then be inserted

the manufacturing
to
designer
consult
the
asking
stabilizing

equations for

costing

equation could cause

requirements and cost reduction

suggestions

For instance the use of an equation to estimate the


to bore a hole with tolerances

of such statements as:

of <0. lmm could initiate

337

(i)

Avoid blind

(ii)

Can these tolerances

The generation

holes.
be relaxed?

of manufacturing

data bases and their

a long term solution

designers is possibly

use by

to the problem of analysing

designs in'terms

of producibility.

More immediate solutions

the construction

of lists

"producibility

and the use of a "data file

designer(209)

layer"

to carry out the multi-disciplined

the ability

for a Value

analysis

of designs

Draughting

CADsystems are mechanical draughting


they relieve
creating
hatching

and creating

friendly

interactive

and accurately

manufacturing

symbols.

scaling,

dimensions, cross

and arrowheads, calculating

In addition

the user

graphics system enables drawings to be quickly

generated.

draughting
task
of
the
time.

systems in the sense that

the draughtsman of such tasks as dimensioning,

lead lines

effect

for the

design improvement effort.

during a co-ordinated

overall

tips"

The advantage of using a VE committee is

Engineering committee.

7.5

providing

would be

However during the design of the New EHB

represented a small proportion

of a designer's

Hence CADsystems could be expected to have a minor

in terms of reducing required

design time.

338

CADsystems are suitable

for designing small numbers of complex,

custom-designed products (e. g. heavy cranes) where design is


accomplished by using various arrangements of standard components
and assemblies.

In these circumstances the ability

to hold large amounts of component drawings in file


retrieve
addition

and quickly

them enables design time to be significantly


reduced.
a
storing cost and manufacturing data will allow other
For example, materials

departments to function.

In companies that

manufacture

during
CAD
system
the use of a
lead to the generation
if

may occur

files

both special

the design

can be purchased

created

of designs

for
for

and standard

of a standard

of ineffective

initially

development
in
the
used

In

schedules prepared.

and tenders and manufacturing

could

of CADsystems

product
say jobbing

higher

products

product

designs.
quantities

range
This
were

volume component

manufacture.

7.6

Integration

with Overall

The responsibility
that

of

NPDProcess

for co-ordinating

the designer and a multi-disciplined

NPD tasks at H.M. Ltd. is


NPDcommittee.

The

the
allows
effective
NPD
technical
system
control
of
management
present
but
level
be
the
to
achieved
of planning carried
development process
be
development
time
to
minimized.
does
enable
not
out

339

A CADsystem could improve this


effective

by allowing

situation

communication between departments.

more

Also the use of a project

design
the
technique
such
as
planning
scheduling method developed in
the present work would allow overall

The future

direction

development time to be minimized.

of CAD/Computer-Aided Manufacture integration

Knox(209),
be
directed
to
should, according
information
cohesive
Knox suggests that
corresponding
requirements

bills

system" for combining a series

of business tools.

for a company's products and the

a forecast
of material

should be used to create part

(Figure 61).

A Pareto cost analysis


fication

at developing a "coherent,

and segregation

both by a family

classi-

system and manufacturing polycodes should then be used to

determine both the type of manufacturing


tools
machine

to perform the work.

required

and the correct

340

FIGURE 61

INTEGRATED-INFORMATION
SYSTEM

Ref.

PRODUCT

BtU.

(209)

OF MATERIAL

341

CHAPTER8

DISCUSSION

342

8.0

Introduction
In recent years the increased action of competitors

development of manufacturing

technology have increased the rate at which

companies must undertake product development.

larger

increasingly

proportions

of the technical,

Therefore,
company.
of
a
resources
management
implementation
to actively

process and demands


financial

and

the managementand

development
of product
must change to reflect

the need

minimize development costs and improve the planning and

of the NPD process.

control

,,

The new product development

becoming a "continuous"

is gradually
process
-(NPD)

and the rapid

An important

projects

effect

of the increase in the rate at which NPD

is
the
be
need to consider a greater number of
undertaken
must
product design concepts.

alternative

has_shown that

the major obstacle

designs is not the synthesis


with which alternative
limiting

factor

generate quickly
required

The work carried

out at H.M. Ltd.

to the economical comparison of concept

of such designs but the generation of data

concepts can be compared.

is the inability
and economically

of conventional

In particular
estimating

a major

systems to

the large amounts of cost data

during the NPDprocess.

Two approaches are available


of alternative

for increasing

the economic number

design concepts that can be costed during NPD. Both

have been considered by the author and they are:


approaches

343

(i)

cost data, and

reduce the cost of generating

reduce the amount of cost data required to compare


designs.

In order to reduce the cost of generating cost data the existing


Ltd.
M.
H.
was examined using Nadler's(95)
at
estimating, -system

The examination revealed that the work measurement

system descriptors.

technique used to estimate manufacturing


the overall

seven

and estimating

effectiveness

times of components controlled


accuracy of an estimating

system.

the work measurement technique employed determines

In particular,

the type of personnel who could estimate manufacturing times.

Conventional

the
type
H.
M.
Ltd.
techniques
such
as
used
at
work measurement

(i. e.

synthetic

extensive
with
personnel

forms only

cost

for

duties.

The large

data.

detail

NPD and the extensive


costs

manufacturing

of an estimators

for

requests

during

required

Cost estimating

using

conventional

the NPD process

Hence the responsiveness

system is dependent on the priority

of an estimating
department's

part

require

background knowledge of the methods and processes

components.

used to manufacture
normally

derived algorithms)

time standards and empirically

given

to the design
data

amounts of cost

required

techniques

to estimate

exacerbates

this

situation.

It
system
design

be
greatly
can
evolves.

examined
adapted

how the responsiveness

has been explained

existing
for

measurement

improved when cost

In order

to achieve

this

of an estimating

data can be generated


objective

identify
to
work measurement systems

designers.
by
use
techniques

the present
those that

The examination

of existing

more fully

in Section

is discussed

as the

work
8.1.

work has
can be

344

system has been developed that reduces the time and

An estimating
cost of generating
develop costing

cost data.

The system has used regression

analysis

the manufacturing

equations capable of estimating

to

times

and hence labour and overhead costs of components for a wide variety
The system which can be used by designers is

processes.

of production

in Section 8.2.

discussed more fully

costs quickly

manufacturing

High cost items can therefore


exercises

and economically as the design evolves.


be quickly

In addition

initiated.

It is capable of estimating

identified

the estimating

types of cost data to be generated,

and cost reduction


system enables many

i. e. labour costs. of products,

assemblies and components, process and operation

costs,

the manual

for
manufacturing processes and costs arising
content
work
and machine
in individual

cost

Hence the cost data requirements

centres.

such as management accounting,


engineering

production

budgeting,

tendering,

control,

of H. M. Ltd. 's current

estimating

system

data
designs
component
cost
were often released for
generate

to quickly

had
before
been
they
costed.
manufacture

prototype

functions

can be satisfied.

and marketing

Because of the inability

production

for

This procedure was

development
delay
the
to
The
to
process.
prevent
ability
of
necessary
based
the
estimating system to cost designs
regression
designers using
therefore,
can,
as they evolve
design

that
changes
for

accounted
EHB range.
changes

that

It

arise

be expected to reduce the number of

in this

manner.

These design

13.7% of the resources

available

be argued that

reducing

cannot

take place during

amount of resources

required

for

designing

the New

the number of design

NPD would necessarily


to design

alterations

reduce the total

the New EHB range.

This

is

345

because the resources required to achieve acceptable manufacturing


for component designs cannot be accurately
dependent on such subjective
of the engineering

factors

scheduling

determined since they are

as the experience and ability


out the design work.

designers carrying

of design changes occurring

However the possibility

for designs are prevented from being draughted.


be effectively

fixture

frequently
departments
-loads on such

occur

the

can effect

of development tasks adversely since production

design cannot therefore

costs

drawings

Tasks such as jig

and

scheduled, and high work


due to the inability

of the

design' department to release component designs in a smooth flow of


finished

drawings.

Hence the reduction

would
take place as a result
that

in the level

of design changes

of using the estimating

be
to
by
the
significantly
can
expected
author
developed
scheduling

of development tasks with an overall

reduction

system
improve the
in lead time

to manufcture.
During the concept and detail

design phase of the NPDprocess cost

importance
to
that
not
only
ensure
manufacturing
is
paramount
of
data
costs are within
of decision

(i)

tools

pre-defined

but also to allow the economic use

such as:

buy
analysis,
or
make
value engineering,

(iii)

targets

determining
respective

and

the optimum number of standard units

and their

sizes to manufacture in a product range.

346

At H.M. Ltd. the extensive use of such design decision


at the present moment due to the inability

limited

system to quickly

estimating

The implementation

of the estimating

improve this

situation.

Design

to enable cost

designs to be developed.

effective

In particular,

increased use can be made of the model (Equation 7)

for
by
the
author
allowing
proposed

designers to carry out "make or buy"

The model which enables both the purchase and in-house

analyses.

manufacturing

costs of designs to be quickly

is discussed more fully

Value engineering

is by its

nature

a relatively

slow process.

numbers of components need to be examined during

large

process,

if

To avoid

design
of
wastage

a complex product

design changes arising


build

to the prototype

such as a crane hoist

resources

it

from value engineering

the NPD

is being developed.

is necessary

to enable

to be implemented prior

phase of the NPDprocess.

by
developed
the
that
author,
as
such
system
this

estimated by designers

in Section 8.3.

Normally

achieving

of the existing

system developed by the author

can then be used more extensively

tools

is

and economically generate cost data.

can be expected to significantly


decision

tools

A responsive estimating

is therefore

essential

in

objective.

Value engineering must be integrated


to enable cost effective

into the overall

designs to be developed.

of the VE function.

effective

control

(Equation

9) can be expected to assist

NPDprocess

This aim requires

The model developed by the author


in this

area by allowing

calculation

347

of the number of components that can be value analysed taking into account
the time and resources available

for the NPDprocess.

VE could then be selected on the basis, of their


importance to the overall
responsive estimating

Candidates for

relative

cost or

success of the product design.

system is essential

Again a

to the selection

of

candidates for value engineering.


/

During the present work it was observed that the rate at which
reduction
cost

exercises

could be accomplished was increased by taking

"experience
the
of
advantage
in Table 12 was selected
total

effect".

Each cost reduction method listed

in turn and then each component within

product design analysed with respect to a specific

the'experience

method.

the
Hence

gained through examining a previous component enables

be
to
more rapidly
components
subsequent

The alternative

analysed.

approach to the problem of increasing

the economic

be
that
during
design
NPD
the
can
considered
concepts
of
process
number
involves

data
the
of
amount
cost
necessary to compare product
reducing

designs.

The research carried

out by the author in this

for
forecasting
models
of
the development
designs.

(i)

costs of product

The basic methods examined for developing costing models were:

the determination
specifications
of electric

(ii)

the total

area involved

of relationships

such as hoist

between cost and design

capacity and the horse power

motors, and

the use of a small number of the high cost components of


design.
a product

348

The models forecast

the total

discussed in Section 8.4.


is not required,

costs of a product design are

Using these models detailed

hence the time and resources necessary for costing

be
to
be
significantly
expected
can
purposes
factor
is
major
a
since cost
knowledge of competitive

effecting

reduced.

This is of benefit

been shown to be suitable


market life

product costs through the ability

"turning

to quickly

techniques to be

to a company since causal techniques have


for forecasting

(PLC) of a product.

cycle

In addition,

price and sales demand, a

forecasting
these
would
enable
causal
costs
estimate
Used.

cost estimating

the decline

The ability

stage of the

to forecast

the

of the PLC of a product enables the optimum time,

points"

to be determined for developing a replacement or improved design


fora

demand
is
declining.
sales
product whose

The model developed using a modified Pareto distribution


the manufacturing
if

cost of competitive

product designs that would result

Ltd.
H.
M.
at
manufactured
was
the product

in manufacturing
products

of competitors

the NPDprocess.

competitors
The market price

into
the decision
not
enter
would
products
This is beneficial

product

development work.

product

does not always reflect

design.

'For example, marketing policy

be reduced artifically
market.

Hence only differences

's
H.
Ltd.
between
M.
and their
costs

initiate
to
be
used
would

estimates

to carry out

since the price of a

the manufacturing

costs of the product

may require

in order to capture a larger

product prices
share of the

to

349

The inability-of

current

forecasting

of the decline

predict the start

stage of the market life

period of time.

be
required.
may
resources

over the NPD process.


during
examined
development

of a new product

techniques

planning

Existing

the present

project

techniques

planning

have been

work and found to be inappropriate


The disadvantages

range.

to the

of such project

as PERT, MBO, PPBS and CPM are:

High cost of implementation.

(ii)

Inability

to integrate

the important factors

i.
e. optimizing
process,
The need to predict

of the NPD

product design.

the times required

development tasks prior

to carry out

to the design phase.

To allow a greater degree of planning and control


during

development

must be exercised

and control

(i)

(iii)

occurs to achieve an

However to minimize the required design

degree of planning

resources a'greater

If this

the NPDperiod additional

design
within
product
optimum

cycle of

in which new products must be

lead
to
a situation
can
a product
developed in a limited

techniques to accurately

to take place

by
developed
the
been
has
author to allow designers
NPDa method

to estimate

individual
times
development
of
the

components.

Hence the

design
the
through
the
process
with
objective
components
of
sequencing
In
development time required can be accomplished.
total
of minimizing
be
basis
the
allocated
on
can
of the
development resources
addition
to
the
importance
the
components
of
overall
success
of
or
cost
relative
through
design
to
sequence
components
and
The
allocate
ability
design.
limited
enables
development resources
effectively.

development resources to be used

350

to schedule design allows use of cost targets

'_Theability

to the reduction

a greater contribution
ment of a cost effective

design.

to make

of design time'and the develop-

For example, if

before
the allocated
for
a
component
reached

a cost target

was

resources had been expended

then the unused resources could be employed more effectively


If cost estimates for a new component design fail

elsewhere.

to meet pre-defined

then the size of the cost variance could be used to determine

targets

,
the appropriate
would indicate

to be taken.

action

High cost variances for example

that a value engineering

exercise was necessary whilst

the recovery of small cost variances would be the responsibility


The scheduling method developed by the author is

the designer.
,

of

discussed more fully

in Section 8.5.

In order to develop an optimum product design the tasks performed


by-the various

functional

integrated.

be properly

departments involved in the NPDprocess must


That is, all

sequenced with compatible objectives.

correctly

the integration

ment for

of the activities

barriers
the
of
is the removal
departmental

functions.

out must be

A fundamental require-

of the departmental functions

(interfaces)

An alternative

carried

to information

solution

flow between

which the cost

developed
by
the author achieves
systems
scheduling
and

estimating

information
for
the
need
is to remove
integration

interfaces.

The total

accomplished

using a CADsystem.

system,

activities

the opportunity

to flow across departmental

of the NPDprocess could be


Since H.M. Ltd. are installing

was taken of examining the contribution

a CAD
CAD

A
NPD
valuable part of this contribution
is
the
process.
the
to
makes
in
integration
the
CAD
the
departmental functions
aids
of
to
which
extent
in
this
is
discussed
The
out
area
work
carried
NPD.
more fully
during
in Section

8.6.

351

Comparison of Work Measurement Techniques

8.1

work measurement methods are concerned with the development

Existing

tasks with which to control

for
times
production
of standard
operations.

times are affected

Since production

the
is
estimation
performed,
which work

manufacturing

by the conditions

of standard manufacturing

under
times

development
the
and maintenance of standard working
must also consider
11
conditions.

It

is the level

of precision

standard working conditions

in the development and maintenance of

that is the limiting

be
that
can
obtained.
accuracy

of estimating

resources required

to the degree

factor

The high level of management

for precise planning and control

of a working environ-

in
low
volume,
multi-product
is
uneconomic
normally
ment
Ltd.
M.
H.
such as

Hence variability

out specific

listed

in the times required to carry

tasks.

The variation
out within

carried

in the working conditions

in variability

20
18
result
and
Tables
in

organisations

in the degree of planning of manufacturing methods


industry

has led to the development of a variety

of

These vary in the degree of methods planning

systems.
measurement
work
time standards.
to
estimate
required

to
estimate
used
the
Of
methods
synthesis
cannot
products

ignored
be
can
predict
during

during

manufacturing
the design

time standards,

the present

work since

time study
this

hence
is unsuitable
times and
phase.

technique
for

costing

352

PMTS systems such as MTMare accurate


detailed

but require

standards,

design of the working


slow and costly,

and of little

For example,

H. M. Ltd.

of PMTS systems,

family

including

methods planning,

These systems are therefore

area.

practical

use for

using MTMV, which


the time required

level

of resource

expenditure

MTM standards

unsuitable.

measurement systems

to estimate

(120,129).

the manufacturing

is clearly

function

work

the application

time

methods planning

Hence these systems would be unsuitable

required.

This

of computerised

However, although

at

of the MTM

4 man years.

estimating

form the basis

also

relatively

is the fastest

detailed
through
using
a
computer
reduced

is greatly
is still

on the cost

the detailed

cost estimating

be
EHB
New
the
would
approximately
range
of

costs

time

methods of estimating

for use by

designers.

Estimating

using synthetic

Ltd.
M.
H.
used at
is therefore

similar

This technique uses pre-specified


in application

for
synthetic
tasks
which
the work
related

specifically
a company.

time standards is the method currently

to the use of PMTSsystems.

synthetic

However

standard times are measured are

to the manufacturing

In addition

time standards and

activities

carried

out within

time standards are normally larger

PMT-standards.
The
level
than
basic
motions
of
work
of
aggregates
to estimate manufacturing times is
development
required
methods
lowered.
time
the
application
estimating
and
reduced
therefore
be
cost
estimates
can
prepared
using
which
at
the
rate
However,
synthetic

time standards is not sufficient

cost
data requirements.
NPD

In addition

to cope with the current


their

use requires

extensive

353

knowledge of manufacturing
in the production

to identify

the relevant

work tasks involved

of a component. This method cannot thus be used

by designers and the responsiveness of the estimating

directly

given by the estimating

is dependent on the priority

function

system
to

data.
for
cost
requests

1 -Category estimating,

although a work measurement technique

to low volume manufacturing

appropriate

systems, cannot be used to

designs
is
the
labour
the
of
product
since
costs
method
estimate
designed to provide acceptable estimating

statistically

The time standards for individual

components.
large
of
number
a
therefore
could
components
to determine the appropriate
It

accuracy over

be poor.

This method requires

estimators

time band (Figure 10) for a component.

is reasonable to assume that narrowing the time bands would improve

individual
of
the accuracy
allocated

the specific

However Fairhurst(125)

manufacturing

with the manufacturing

therefore
cannot
and
company

A suitable
manufacturing
allow

to each time band.

time and reduces system responsiveness.

Again category estimating


familiar

time allotted

components are

points out that narrowing the time band increases

application

the estimating

estimates since individual

requires

personnel (foreman, chargehands)

methods and processes used within

be used by designers.

method of estimating

time standards for a low volume

is
the use of work measurement algorithms.
system

developed
be
to
quickly
time standards

the causal relationship

between manufacturing

and accurately

These

by expressing

task times and the factors

354

In addition,

those times.

that affect

work measurement algorithms

developing
level
both
the
the
of
cost
estimates
and
could reduce
subjective

judgement involved,

data files

of time standards.

of

and overcome the need to hold extensive

To achieve these benefits

the variables

used within

algorithms

must be easily

to designers.

accessible

be used by designers inaccessible

That is,

predictor

if the system is to

variables

such as machining

feeds, speeds and depths of cut must be avoided.

Four basic methods can be used to develop work measurement


algorithms:

Synthetic

c>>

Development - Humanjudgement is used to select

both the predictor

variables

and the method of combining

variables.

The obvious disadvantage with this method is the

possibility

of developing an algorithm

in its

development rather

actual

causal effect

predictor

variables.

to fit

than a model that expresses the

between manufacturing

by the model (Equation 21)

developed by Mahmoud(138) for estimating

the actual

times and

This disadvantage of the synthetic

development technique is illustrated

components.

the data used

the costs of turning

The model employs a discontinuity

factor

(Nd)

value of which varies with component volume and

the form of the raw material


fabrication).
and

used (e. g. casting,

Nd is intended to represent

solid

bar

the probable

355

number of passes of the tool over. the machined surface and


therefore

the design complexity.

that design complexity

However it

is not always related

volume or the form of the raw material


is not causally

related

is obvious
to component
Hence Nd

used.

to manufacturing time and a variety

Nd are required to allow estimated


of methods of calculating
'Oeturning times to fit the
observed data. Therefore poor
estimating
(ii)

Empirically
conditions
individually

accuracy can result

as illustrated

in Table 25.

Derived Equations - are developed under laboratory


hence the predictor
controlled

to be assessed.

variables

the effect

allowing

Empirical

examined can be
of each variable

techniques are normally used to

develop machining time algorithms

such as Equations 22 to 26,

which could provide rapid estimating

facilities.

Usually

however the many parameters (Figure 12) used to estimate


be
times
cannot
causally
machining
to
construct
necessary
27.
in
Table
shown

related,

hence it

is

standard parameter tables of the form

The use of such tables reduces the speed

be
can
prepared and prevents designers
estimates
with which
since familiarization

using the algorithms


is
needed.
methods

In addition

empirically

the
conditions
under
only
are useful
constructed.

Hence the variability

with machining
derived models

from which they were


in working conditions

low
lead
volume
manufacturing
systems
with
could
associated
to variances

between the estimated and actual manufacturing

times for components.

356

(iii)

Research Techniques - are capable of developing

Operational

work measurement algorithms:


(a)

Simulation. - This method is basically

a synthetic

development procedure although the use of a computer


allows the examination of a large number of potential
However since models are synthetically

models.

developed it

is possible

ing predictor

variables

to obtain a model containthat are not causally

to the independent variable


(b)

related

being estimated.

Queueing Theory -This method has been used to estimate


the waiting

time involved

in fabricating

crane sub-

is
interference
useful
when
and
assemblies(122)
occurs between manufacturing operations.
(c)

Linear

Programming - This

linear

models that

develops
method

it

important

regression

of equations

(iv)

cost

is an expensive
during

similar

equations

in terms of computing
tool

specific

optimize

linear

by multiple

method can be used to develop

because LP is an optimization
This factor

work since a large

is
number

needed to be developed.

method of determining

between dependent and predictor

The method employs partial


the coefficients

However,

analysis.

Regression Analysis - is a statistical


relationships

to those developed

alternative.

the present

The

objectives.

differentiation

of predictor

variables

variables.

to determine
(Equation 28)

357

that minimizes the sum of the squares of the differences


between estimated and observed values of the dependent
The technique therefore

variable.

constructs

between a dependent

that expresses the relationship

Hence

variables.

and any number of predictor

variable

a model

time standards can be

individual
large
of
number
a

into simple formulae.

transformed

it

is possible to use data

Using regression

analysis

from an existing

system without the need to isolate

effects

of each predictor

variable

the

involved under laboratory

conditions.
"

Although

investigators

by chance,

predictor

variables

requires

naturally

manufacturing

cannot distinguish

between variables
(158,159,162,160)

effect

causal

an actual

if

technique

the regression

and one that

have shown that

the use of personnel familiar

tasks being examined) effective


In addition

They are however essential

experienced

in maintaining

should manufacturing

several

based estimating

models:

models.

a model's

conditions

alter.

An examination of previous research(153,156,157)


factors

with the

estimating

personnel are not required when using estimating

important

occurs

are chosen with care (i. e. which

developed.
be
can
models

applicability

between

highlighted

to consider when developing regression

358

Although the estimating

(a)

accuracy of models can be

improved by reducing the data variability


in Figure 15, additional

illustrated

necessary to enable the total


covered.

range, as

models are

data range to be

Hence the estimating

complex and less responsive.

system becomes more


Accuracy must therefore

be balanced with estimating


(b)

It

regression

to an estimating

equations are applicable

data range used in their

predictor

variables

other predictor

variables

System Development

The development

of a cost

to identify
equations
estimating

the

only within

development.

that are inter-correlated

Cost Estimating

centres)

model since

The complexity of equations can be reduced by removing

(c)

models requires

time and cost.

is necessary to define the manufacturing tasks and

data range appropriate

8.2-'

application

the definition

estimating

the relevant
are valid.
equations

in a model.

system that

of the manufacturing

to which the estimating

models apply

equations

In addition,

within
used
are
equations
numbers of

since

a classification

this

(termed cost
allows

estimators

under which these


system must allow

particularly

a system.

employs regression

tasks

and the conditions

to be accessed rapidly

with

when large

359

The number of equations required for a system is dependent on


the range of manufacturing

tasks that need to be costed and must often

be increased in order to improve the estimating

accuracy of individual

decreasing
the
factors
i.
number
of
e.
equations,

considered variables

within an equation
requires additional

(whilst

improving estimating

models to be developed.

the present work is to balance estimating

accuracy) 'naturally

The objective

during

accuracy with system size

costs.

and application

The development of an estimating


types of predictor

variables

system must also consider the

used and ensure they are accessible

to

In this respect
designers, hence maximizing system responsiveness.
.
factors such as machining speeds and feeds, depths of cut and number
from
the
be
estimating
excluded
must
of cuts

In addition,

equations.

a system must enable cost data to be estimated on a

that allows designers to carry out detailed

level

design solutions,

alternative

design features

comparisons of

hence cost data is required for individual

such as splines

An examination of existing

and gear teeth.

component coding and classification

indicated
Opitz(164)
Brisch(165)
VUOSO(163),
these
and
as
such:
systems
the
to
they
designed
present
work
since
are
applicable
systems were not
(165)
is
design
It
Gombinski
by
than
costing.
pointed
out
other
for purposes
that

their

supports

be
for
may
purposes
not
costing
possible.
adaption
this

conclusion

since he abandoned the use of existing

the development of a computerised estimating


during
systems

.
system.

Blore(91)
coding

and planning

360

to the designer is associated with the manufacturing

Data inaccessible

task and variability

input to the estimating

of this

type of data was

found to be caused by:

The process being carried

(i)

out, e. g. machining, welding

and assembly.
(ii)

The operation

being performed, e. g. turning,

boring and

drilling.
(iii)

The size and power of the machine used.


The manual work content

(v)

The tolerances

(vi)

The material

involved.

and surface

finish

being processed.

The removal of manufacturing variables


of processes and operations

identification

initially

required the

(Table 28) carried

of developing estimating

the
objective
Ltd.
with
M.
H.
process/operation

required.

out at

models for each

combination.

Since the variability

of machine process criteria

arises due to

it
the
develop
to
used
machines
was
of
necessary
power
the size and
estimating

basic
type
for
each
of machine used at H.M. Ltd.
equations

General
of
In the case

Machining

operations
Hence it

types.
machine
of
range
wide
a
on
a method

appropriate

(Table

47, Appendix

II)

of enabling

for
components.
machine

a component could be processed


was also necessary
designers

to select

to develop
the

361

Allowing the manual work content involved in machining operations


to be calculated

separately

from the machine content could both improve

estimating accuracy and allow the cost estimating


in areas other than design costing.

system to be useful

For example, cost comparisons

could be made between manual and automated machining systems since


the manual work content saved by automating a. process could be
e

specifically

determined.

The method of estimating


(135)
involved estimating
Stout
time.
the
process
machine
of
since

unsuitable

manual task times as a direct

a tool

time,

proportion

However this method was considered

manual task times are dependent on factors

than the machine process


distance
the
on

manual task times adopted by Halevi and

e. g. tool

manipulation

is moved between machining

other

times depend

operations.

involved
the
in estimating manual
To
of
number
equations
reduce
,:
the
in
overcome
addition,
times
need for detailed methods
and
task
planning

being required,

to Blore(91),

individual

i. e. load/unload

manual tasks were grouped according

machine, machine manipulation

and machine

During the present work an additional manual task group


(92)
by
inspection)
Knott
is considered an
recognised
(gauging and/or

tasks.
set-UP

integral
often

part of machine manipulation

difficult

manipulating
location

to separate.
a tool

tasks since the two types are

For example during horizontal

to the required position

are often carried

out simultaneously.

boring,

and measuring tool

362

Although the type of dimensional tolerances


affects

machining operation

depths
of cut,
and
speeds
Henceestimating

required from a

machine process variables


this variation

such as feeds,

is normally a step function.

equations had only to be developed for two tolerance

bands, i. e. <+0.01mmand !O.Olmm). For dimensional tolerances=<0.05mm


an additional

operation was necessary for which a grinding

grinding

allowance (+0.1) had to be left

times are influenced

Since machining
materials

it
be
machined
can

factor
define

or measure machinability

rate
effect
it

available
classified
obtained
equations

(Table

The collection
involved

(i)

used to

the index of "metal


This

measure.

is constant.

30) according

removal

index, measures the

on machining

time since

Materials

were therefore

machinability
using data
(14 4)
Factor
Group
and Material

to their

from the P-E Consulting


included

a "machinability"

at which metal can be removed when the

rate

horsepower
machine

which

of the common criteria

machinability

relative

the
relative
measures

to include

identified

as a suitable

of a material's

by the ease with

was necessary

An examination

in the system.

per horsepower"

on the component.

in the estimating

system.

develop
to
from
data
estimating
which
of

equations

two basic tasks:

Choosing the predictor


had to be primarily
to use the system.

variables,

design criteria

i. e. these by necessity
to allow designers

363

(ii)

of the predictor

Collection
variable

and corresponding dependent

data from a suitable

With respect to choosing predictor


bear in mind that the regression

source.

variables

it was important

to

procedure could not distinguish

between variables and a


between an actual causal relationship
Of
by
Hence
the
to
that
chance.
occurs
need
choose
relationship
predictor
literature

with care required

variables

the examination of relevant

and the use of personnel familiar

with the manufacturing

tasks under examination.

various types and combinations of predictor

In addition

to
in
obtain
effective
order
examined
were
dependent variable.

When small differences

levels
accuracy

of alternative

standardization

of predictor

simplify

the overall

estimating

estimating

models for each

(E-<0.05) existed in the

equations consideration
variables

variables

was given to the

between equations in order to

system.

(MTM
V standards,
data
sources
Several

synthetic

time standards

time
programme)
were
examined
study
and
considered
planned
a
and
library
the
of
use
an
with
existing
extensive
compared
when
unsuitable
M.
Ltd.
With
H.
held
to
data
respect
sawing,
at
spline
time
study
of
data
time
study
was not available.
operations,
shearing
cutting'and
that
the
it
considered
standard
was
production
operations
these
For
times

floor
to
the
issued
shop
(STT's)

source,
relatively

particularly
low.

could represent a suitable

data

since the times associated with these tasks are

364

An existing
times' for steel

procedure (Table 31) for developing flame cutting

Since no loss of estimating

accuracy over the existing

the equations did not require

incurred,

loss of estimating

plate was adapted without

such as chamfers, circlip

grooves and radii


The arithmetic

exhibited

adequate accuracy

for

developing

data issued

by the Welding

personnel

(Equation

factor"

arc time"

estimating
"basic

used since

to total

the approach
times

welding

arc times"
and (ii)

Institute(171)9

the affects

to estimate

for

to estimate

equations

to convert

was therefore

tasks were inaccurate

welding

equations

(i)developing

"basic

a machine type.

within

could be obtained(155).

levels

Since STT values


for
adopted

variability

mean of each operation

method was

validation.

The times taken to machine features


little

accuracy.

using

developing

time using

welding

of variables

involved
standard
equations

experienced

on the welding

"operating

31).

Since a large number of estimating

equations needed to be developed

the
to
Hence
actual
regression
perform
analysis.
used
was
a computer
therefore
avoided
and
possible
was
errors
eliminated.
effort
tedious
much

The regression

package(176) available

at the Loughborough University

during
it
the
present
work
since
Centre
used
was
Computer
supporting
of
range
wide

statistical

data (Table 33) with which to

of the estimating

determine

the quality

addition,

the regression

provided a

equations developed.

package could recursively

construct

In
prediction

365

equations using stepwise regression


maximumestimating

procedures.

accuracy to be obtained using the minimum number

variables.

of predictor

The package also has the facility


for inter-correlation
from the final

This enabled the

to test predictor

and when necessary remove correlated

The data handling facilities

is
part)
procedure
regression

of the SPSSsystem (of which the

enabled data to be held in sub-files

be
developed
individual
for
that
could
equations
such

in a single

This facility
regression

enabled estimating

run for individual

equation
an
when
reduced
accuracy

for

equations to be developed

the system could be

a group of machines gave acceptable

results.

Since the regression

technique has been used when the exact

relationship

between dependent and predictor

is essential

that the accuracy of the estimating

within

or combinations

or combinations of machine

'Hence the number of equations within

sizes.

variables

model.

estimating

of sub-files.

variables

the current

accuracy limits

required

The commonplacemethods (R, R2, F-test,

.
Variability)
of
inappropriate

of indicating

is unknown it

variables

models proved to be

by H.M. Ltd.

t-test

and the Coefficient

formula accuracy have been shown to be

to the measurement of a model's estimating

direct
a
not
they
are
since

measure of the difference

accuracy,

between actual

366

(i.
e. residuals).
values
and estimated

The present work supports this

18
16
to
Figures
show that R,
opinion since
of Variability

are unrelated

directly
determined
is
which

F-value and the Coefficient

to Average Proportional
from residual

Error

values.

1-e where e is

Cochran(155) proposed the use of the quantity


determined using Equation 33 but states a limitation
of dependent variable

data is involved.

when narrow ranges

In these circumstances the


accuracy since it

1-e is a poor measure of estimating

quantity

(Equation 36)

fails

to consider that a narrow base of experience could be an unreliable


future

foundation for projecting


when'using regression

results.

However a basic limitation

models is the fact that models are only applicable

derived.
from
they
data
were
which
range
over the

Hence Cochran's

to the use of 1-e is not valid.

limitation

The present objection


an equations estimating

to the use of l-e and a further

accuracy proposed by Cochran the Association

inclusion
is
the
34)
Index (Equation
in the indices

measure of

"completely"
being
not

y
term
the
since this
of
directly

related

results

to residual

values.

index of the estimating accuracy of models provided


(174) is
Standard
Error
Ratio
(Equation
the
35).
It
Fox
and

A suitable
by Ezekiel
requires
indicates

the
size of residuals
only

for its

determination

and therefore

the closeness with which estimated values may be expected

to approximate

to the true but unknown actual values.

individual
the
of
accuracy
the
of
range
of
a measure

It is therefore
estimates.

367

A measure of the estimating


the proportional

errors yi-xi

accuracy of a model needs to consider

since the accuracy of cost estimates

yi
is presently measured in this manner.
an improved measure of the estimating

By using the index E (Equation 36)


accuracy of models is obtained

since this value is a measure of the average proportional


arises through using an estimating

Unfortunately

equation.

the conventional

regression

the sum of the squares of the residuals


the value of E, as illustrated
effective

estimating

conventional

procedure of minimizing

does not necessarily

minimize

in Table. 54, so that the most

equation is not always developed.

regression

that

error

procedure is suitable

However the

for use in the present

work since:

(i)
.

The low volume manufacturing environment at H.M. Ltd.


prevents proportional

errors

being magnified into large

cost variances.
/

Analysis

of product manufacturing

distribution
Pareto
a

exists

costs indicated

(See Section 5.2).

that
A small

high
cost components accounted for the majority
of
number
of a product's

total

manufacturing costs.

use of the conventional

regression

because the technique effectively


high
of
cost components.
accuracy

In this

way the

procedure is suitable
improves the estimate

368

For high volume manufacturing systems in which the reduction


average proportional

error

estimating

Appendix VI also illustrates

errors.

manufacturing
/

individual
of
range
errors

the average error

proportion

equation and does not therefore

errors

that could arise.

arising

from

indicate

the

The range of individual

that could arise from an equation is indicated

Deviation
Standard
the
is
which

of proportional

The values of E and S (and R2, F-value,


descriptive
are normally

errors

Student t-test,

measures of a model's estimating

times examined are normally distributed.

manufacturing

form a negative exponential

times usually

intervals
of confidence

calculation

a simple statistical

distribution

not practicable

cannot therefore,

1-e and AI)

accuracy and
when the

However labour
(122,123).

The

be achieved using

package such as SPSS. They could however be obtained

a
sophisticated
more
using
error

distribution

by S

(Equation 37).

can only be used to provide confidence limits(213)

therefore

routine

how the relative

accuracy.

the use of an estimating

skew

regression

volumes of components can be allowed to influence

The value E indicates

estimating

the conventional

(Appendix VI) to minimize proportional

procedure can be modified


estimating

is important,

in

package such as GLIM(214) and specifying

such as Log Normal.

since the proposed estimating

However this

approach is

method is intended for

industry.
in
use

E
the
and S were obtained using new data (i. e. time
of
values
Since
the
to
it
is acceptable
estimating
construct
model)
used
data
not
study
by
judge
to
the
which
as
norms
values
estimating accuracy
these
to use
of equations.

369

Examination of the equations that failed

(i. e. the acceptance levels were set at H.M. Ltd. 's

'acceptance criteria
estimating

current

accuracy limits

of E=0.096

that the prime causes of poor estimating

(i)

control

were:

Variation

in operator experience.

Variation

in the quality

and dimensional accuracy of the

are characteristic

systems since economics require

although

have
to
poor estimating
appear
equations

the
affects
minimize
would

within
for

of predictor

In the latter

variable
outside

variables.

below which estimated operation

errors

therefore

of negative coefficients

To overcome this

problem

to the manufacturing
times could not fall.

tasks in
Hence

low
data
the
the
end
of
at
range of a dependent
accuracy

improved
be
whilst
could
this

type

in negative labour costs being generated

times (Table 35) were allocated

the estimating

on the

accuracy (i. e. incorrect

and estimating

case the inclusion

low values of predictor

question

they

and the use of negative predictor

variables

were corrected

an equation resulted

operation

accuracy levels

in working conditions

of variations

causes of poor estimating

coefficients)

reduced.

Hence

of task times.

variability

variable

of management

to be exercised over working conditions.

planning

and combination

of low volume manu-

a low level

and control

Additional

from lack of

used.

However these conditions

inherent

indicated

over work methods used.

raw materials

facturing

and S=0.314)

in machine process data arising

Variation

(iii)

to meet the accuracy

area.

not influencing

the accuracy levels

370

The frequency distributions

for the E and S values of the equations

developed (Figures 21 and 22) indicate

labour costs of a component, as illustrated

to estimate the total

from each equation were "additive"

create large estimating

therefore

errors.

H.M. Ltd. components and these total


Sales Target Times (STT's).

lie

The comparisons (Figures 29 to 46) indicated

on or near the 450 lines which indicate

direct
a
manner

i. e. 11.

the probability
relationships

exists

to compensate for this.

between design criteria

the use of a relatively


The resultant

of the regression

then
In

and the manu-

working conditions.

assembly

large number of predictor

model (Equation 38) indicated


procedure finding

increase
increasing
to
with
appears

variables.

Hence it was beneficial

into-groups

the
number of variables
reduced
which

tasks.

accuracy.

time standards and the

develop
to
equations for estimating
attempts

involved
times
task
variables,

link

can be altered

time standards used to control

Initial

perfect

consider* these standards over estimated,

function

the equation coefficients

facturing

times for existing

times compared with the corresponding

If equations are used to develop manufacturing

this

and would

accuracy could be achieved since a large proportion

that good estimating

manufacturing

in

Hence the estimating

determine
the
to
used
manufacturing
were
equations

of the points

required

was necessary to determine whether the individual

it

errors

estimating

of the equations

However since more than one equation is usually

are acceptable.

Appendix II,

that the majority

to classify

that

non-existent
number of predictor

manufacturing
influencing

operations
manufacturing

371

in estimating

A reduction

time of approximately

's
Ltd.
M.
H.
current
compared with
regression

based estimating

estimating

methods, by using the

The response of the estimating

system.

be
to
be
much greater
expected
system can
time is involved

(i. e. cost estimating

since little

le
own cost

their

the estimation

data.

the lack

In addition

"queuing"

for the design fynction

departments)
loads
the
of
other
work
on

reliant

70% was observed

when designers generate

of a subjective

of design costs reduces the risk

is not

in

element

of "user bias" entering

into the process.

Although the estimating system can be used manually (Appendix II)


(177)
to allow a computer to carry out the
developed
been
has
software
improves
This
the
involved.
rate at which estimates
tedious calculations
the'occurrence
of
computational
developed
reduces
and
be
can

errors.

large
is
of
a
number of equations,
the
composed
system
Since
method of retrieving

appropriate

machine

classification

converted
cost rates

at H.M. Ltd.

for the particular

number which,
each equation

centre

type since this

to costs by application

Equations within

grinding

equations from the data file

was

The obvious method was to group equations according to the

required.

centre

the appropriate

corresponded with the existing


Manufacturing

of appropriate

times were easily


overhead and labour

cost centres.

a cost centre were then allocated

in conjunction

cost

an identification

with the cost centre classification,

to be individually

retrieved.

With cylindrical

allows
and surface

(i. e. cost centre 216) and the NC bar and chuck lathes
identical
had
258) processes

cost centre classifications.

(i. e. cost

372

were included in the system operating

Hence instructions

to enable users to retrieve

(Appendix III)

procedure

the appropriate

group of

equations.

Since the computer program had to be interactive


data needs to be input)

variable
carried

i
those
by
the
and
users
out

was necessary

Hence it

predictor

In this

to allow

users to identify

that

respect

sharing"

the relevant

the appropriate

input

could be obtained from a component's engineering drawing prior


program.

the'computer

avoids delay arising

variables

predictor

Using Appendix IV, therefore,


whilst

An advantage of using design criteria

departments within

To fully
transfer

dimensional
procedure

variable

of predictor

estimating

to identify

for estimating

manufacturing

available

to all

In this

system would require

the direct

data from component drawings to the cost

respect the ability

in CADsystems to retrieve

data from stored component drawings could enable the estimating


to be partially

of the estimating

predictor

these equations.

computerised.

system would require

the
relevant
to enable
organize

to using

an organisation.

computerise the estimating

module.

data

using the program.

of cost data being freely

is
the
possibility
costs

it

the computer

on a "time

were obtained

in order

involved

variables

by the computer.

department

was necessary

was necessary to determine the tasks

use of the computer since

to make effective

of the design

facilities
basis.

it

(i. e: predictor

estimating

variable

To complete the computerisation

additional

software developing

equations to be identified

data in the correct

and

form and order for

373

8.3

The Make or Buy Decision


The decision

outside supplier

to manufacture "in-house" or to purchase-from an


(the make or buy decision)

development of a product since all

is fundamental to the

components need to be examined in

this manner.

A model (Equation 7) has been developed during the present work


that allows designers to cost designs and therefore
of alternative

design solutions.

involved
buy
elements
make and

to compare the costs

Since the equation includes both the


in a component's manufacture (i. e.

include
both
designs
often
an
element
of
make and buy),
component
involving

practical

decisions

facturing

can be undertaken.

included

since tool

a significant

the economical amount of "in-house"


In addition,

a tooling

(which
include capital
costs

influence

manu-

cost element is

equipment costs) have

on the economics of alternative

design

solutions.

The model has been adapted from the classical


determining
for
used
manufacturing

the variable

of alternative

batch
sizes
economic

costs involved

batch sizes.

stock control

models

in stock holding and

The affect

(under various manufacturing

of adopting
situations

such

or,
occur
when
simultaneous
costs
manufacture
and
shortage
when
as,
the
design
economics
of
alternative
is
solutions
applicable)on
usage
be
In
this
taken
determining
respect,
care
must
when
be
examined.
can
since purchase and manufacturing
batch
sizes
economic
differ.

batch sizes can

374

The use of Equation


costs highlighted

7 to compare the New EHB rope guide assembly

the influence

design

on the economics of alternative

expected varying

rope guide designs.

Table 11) but tooling

of the tools

for H.M. Ltd. this

As can be

the relative

The choice of which

department therefore

the cheapest

is chosen (Alternative

period is 3 years,

III,

then the initial

cost

the cheapest design solution

(Alternative
term
chosen
were
the short
need to decide if

V, Table 11).

the higher costs are acceptable

is
sales
growth
required)
rapid
the market and

will

If an increase in price is not advisable


profit
to
reduced
be
a
accept
would
In this

investment costs can be more easily

In addition

to

be reduced.

an alternative

solution

margin during the investment cost

manner. production

the purchasing policy

over

The marketing

(at
the
is
introduced
a
when
period
product
sales
product

recovery period.

costs

costs recovered over the company's normal payback

higher
be
if
than
could
of a product

capital

11).

rope guide assemblies as examples, if

design over the nominal life

or if

(Table

solutions

tooling

to adopt could then become a marketing decision.

Using the alternative

period,

for

the tool cost recovery period influences

economics of alternative
design solution

of the payback period

technology requiring
introduced

high

into a company.

of a company towards production

introduction
its
long
until
period
maximum
a
utilization
with
equipment
benefit from this type of investment policy.
also
could
was achieved
include
industrial
would
robots where a previous
This type of equipment
(122) indicated
that
a considerable amount of redesign
the
by
author
study
before the maximumbenefits could
fabrications
was
necessary
of existing
from
arc
welding
continuous
robot.
a
be achieved

375

Equation 7 can also be used to determine the effect

of alternative

systems on the economics of design solutions.


(62)
Period
Control
Batch
basis
is
the
the
of
systems
example,

For

production control

determination

of a manufacturing "cycle"

be a'function

of both the time required to manufacture the component

with the longest production

the analysis

time and the forecast

sales demand. The

batch sizes (which are equal under PBC systems)

purchase and manufacturing

from these two factors

be
calculated
can

time for a product which can

of the alternative

and as illustrated

by

rope guide assemblies can be reduced

Reducing batch sizes increases the labour and purchasing costs

in size.

tooling

per component whilst

and investment costs per component remain

i. e. these cost items are not dependent on batch sizes but

constant,

The outcome of this

change in costs

on annual quantities

manufactured.

therefore,
be,
would

an improvement in the chances of adopting design

that require

solutions

tooling

and investment costs.

design
a
product
of
could not take place until
Accurate costing
manufacturing

be
designs
could
not
costed until
product

therefore,
design

and manufacturing
to identify

estimators

time.

throughput
arrangement
required

batch
known.
Under
sizes
are
a PBCsystem,
purchasing
and

planning had taken place to enable cost

the component with the longest individual

This would present difficulties

drawings.

to identify

detail

sufficient

In addition

in costing

the sequence of detail

the component with the longest

the
be
with
overall
compatible
not
may
time
put

general

design

individual

through-

development plan.

376

-Normally
order quantity

when costing

not considered.
these quantities

to determine

design

minimum costs

solutions

7, however,

total
could

the effects

of varying

on manufacturing
designers

requires

manufacturing

costs.

be compared at their

which occur at the optimum values

and manufacturing

8.4.

batch sizes

and manufacturing
Using Equation

alternative

designs,

product

costs

are

to use

Hence
respective

of order

quantity

batch size.

Cost Forecasting
The regression

work-allows

based estimating

system developed in the present

designers to estimate component labour costs from general

drawings.
arrangement

However during the concept design stage of the

NPDprocess (particularly

if

design concepts are

several alternative

being compared) the need to develop general arrangement drawings in


determine
manufacturing
to
order
determine product manufacturing

costs is undesirable.
costs without

The ability

the need for such design

drawings would:

(, )

enable unattractive
early

design alternatives

to be abandoned

in the design process and therefore

allow design

be
to
used effectively,
resources
(ii)

increase the number of design alternatives

that could be

analysed,
(iii)

design
decisions
fundamental
allow

(e. g. the "assortment"

be
taken, and
to
problem)

-(iv)

allow a greater
to be costed.

to

number of competitive

product designs

377

An obvious method of cost "forecasting"


identify
is
to
product range

the cost/design

specification

hence knowing the design specification

that exist,

cost could be found.

the appropriate

(e. g. capacity,

in determining

has been used to construct

the technique

relationships
size)

In the present work since regression

has been shown to be effective

analysis

when developing a new

cost/design

causal relationships
specification

models.

To ensure model estimating


to
be
construct
used
costs must
the smallest

means that
costed.

depends on the order

during

to enable early

in preference

launch of a limited

models.

in a product

units

in which product

range of product

specification

to develop

the New EHB project

were developed

units

and largest

sizes

effective

costing

costing

the 0.8 tonne,

This

range must be

are developed.

for

however are not always suitable

schedules
example,

cost/design

way the ability

In this

the total

accuracy

models
Development
For

purposes.

1.6 tonne and 3.2 tonne

to the 5.0 tonne and 8.0 tonne units

product range for which the greatest

forecast.
was
market share

Although during the present work cost relationships


in the optimization
disadvantage

(e.
designs
g. drum tube assembly) their
of

total
to
estimate
used
when

detailed
to
carry out
need

product design costs is the

This need limits

hence
the
be
estimating
and
costed
can

respect

to the minimization

of individual

initially
designer
to
for
a
advantage
an
relationships

basic

design and costing of those product sizes

the
develop
models.
to
cost
used
that

have been useful

the number of products

accuracy of models.

With

product design costs it

is

identify

the existence

between the various assemblies of a product.

of cost

Representing

378

these

in a table of the form shown in

cost relationships
a useful

provides

design aid.

Figure 48

For example when the drum tube assembly

designed,
being
Figure
48
EHB
New
provided a quick reference
the
was
of
to the assembly cost relationships
order

to enable total

that needed to be determined in

product manufacturing

costs to be minimized.

As might be expected examination of a wide range of products


indicated
Ltd.
H.
H.
at

manufactured

that when cumulative

costs were

distribution
Pareto
the
number
of
components
a
cumulative
compared with
is

Although the form of the curves varied between products it

formed.

could

be expected that for each product the curve is log-linear.


both cumulative cost and number of components to their

Converting

form (Equation 45) would produce a straight

logarithmic
If

line

curve.

(which
highest
the
cost
components
are
easily
of
number
a small
since they form the basis of the conceptual design of a

identifiable
product)

the whole straight

to
then
extrapolate
used
are

the constants

f and g (Equation 43) required

line

to determine total

curve,
product

that
found
small
be
only
a
number
of
components
would
so
could
costs
The
detail.
in
costing
of
products
also
requires
a
be
to
costed
need
total
the
knowledge of

number of components within

this
which
with
the accuracy
the relative
rapidly

contribution

increasing
with

figure

a product although

is determined need not be high since

to the cumulative

product costs decreases

number of components.

form
in
Equation
43
the
shown
of
were used
However, when models
these
's
models
grossly
Ltd.
products
M.
H.
to cost
actual

product costs.

The over costing

effect

over estimated the

can be explained with

379

out by Swan and Worrall(183).

reference to the work carried

by these researchers indicated

curves established

that the slope of the

Nu-axis.
distance
the
decreases
along
a
at
curve
by
constructing
confirmed

log-log

Log-log

This explanation

was

cost curves (Figure 52) for several

's
Ltd.
M.
H.
products.
of
0

Increasing

the number of components used to determine the values

total
50%
the
(up
to
of
f
and g
of

accuracy of cost models to an acceptable

did not improve the estimating

i. e. to the accuracy of H.M. Ltd. 's existing

level,

greater

Using proportions
be an effective

method of improving estimating

required

estimating

From visual

system.

to determine total

observation

accuracy since this

reduce the amount of detailed

cost

product costs.

of the relative

values of the variables

to
it
43
reasonable
assume that they were related
was
Equation
form

the
changes.
slope
which
at
the
point
to
test
to
used
their

estimating

than 50%was not considered by the author to

significantly
not
would
procedure

that

number'of components was examined)

this

logarithmic

theory since conversion of the variable


form ensured that they were, if

Although significantly
initial
the
developed
accurate

(i)

estimates.

Hence regression

analysis

was

values to

at all, linearly

related.

improving the accuracy of the estimates

models developed did not provide sufficiently


Poor estimating

accuracy was found to be caused by:

Not using a commoncomponent structure


both component and assembly costs.

level,

i. e. using

380

(ii)

Using items classed as consumables (e. g. washers) in the


determination

of the total

number of components within

a product.
(iii)

Not using the actual value of f to determine product costs.

the input data the new model developed, Equation 47,

Upon,correcting

accuracy, i. e. E= -0.015 and

was found to give an acceptable estimating


S=0.096.

8.5

Design Scheduling
When the time available

to develop a new product is restricted

in
to
tasks
development
order
minimize
overall
of
scheduling
time is important.

In addition

the efficient

allocation

the

development

of design

i.
be
be
to
to
e.
resources
considered,
need
allocated
needs
resources
importance
tasks
to
the
the
of
to
overall success
relative
in relation
design.
of a product

Existing

network analysis

be
to
unsuitable
found

planning techniques have been

project

for controlling

the development of a new product

range since:

(i)

They are expensive to implement (PERT implementation


for
17,000
is
a project
cost

of similar

size to that

development)
inferior
EHB
New
the
and
are
an
of
for the expenditure

of development funds when compared

with the equal cost opportunity


staff.

choice

of increasing

the design

381

(ii)

To construct

a network diagram requires

identifiable

tasks.

a series of

Due to the creative

nature of the

design
process the remaining tasks required
new product
to develop a component cannot be accurately
until

le
In addition

the design process is complete.

project

existing

task whereas when limited

Since

should
design
major

such that

constraints

scheduled

(i)

through

allocation

are initially

is here that

That is,

begin.
stage

on effective

it

(PERT, CPM) are

a planning technique

of the available

created

resources.

at the design
of development

the scheduling

overall

the design

development

the sequence with


process

is minimized.

time

stage
work

be
to
sequenced through
need
components

influencing

The order

techniques

design time is available

component designs

NPD
the
process
of

planning

the time required to perform a development

determining
concerned with

should concentrate

identified

the
The

which components are

are:

in which products are developed which can

be dependent on marketing or financial

considerations.

For example, H.M. Ltd. need to market a "mini-product


be
to
to
obtained quickly.
revenue
enable
range"
(ii)

The length of the prototype


considerably

testing

period which is often

longer than the sum of the remaining project

task times particularly


needs to be undertaken.

if

long term reliability

testing

382

Although these constraints

determined the order in which sub-

it
be
developed
to
are
assemblies
design of each component within

is necessary to sequence the detail


In this'respect

each sub-assembly.

scheduling components through the design phase is considered in the


present research analogous to the scheduling of operations within
batch manufacturing

system.

For the design process the main steps are:

(; )

Specifying

the objective

schedules

can be judged.
departmental

the various
product

development

of a single

increasing

(29)

a new product
(ii)

Sequencing

the interactions

it

project

difficult,

time during

importance

that

exist

project

success of

the obvious choice.


In this

the time required


tasks

for

respect
to carry

it
out

each component and

between development

involved
of
components
are
number

tasks.

in the New EHB

is essential that designers are able themselves to estimate

functional

present

the importance

the present

delay
in
the
to
times
task
avoid
acquisition
development
other

in the new

to the ultimate

makes this

development

the individual

aims of

could make the specification

criterion

to determine

the conflicting
involved

the design of components.

is necessary

Since a*large

functions

development

of minimizing
and its

Although

process

objective

by which alternative

criterion

departments.

research related

design criteria

In order to achieve this

development activities

and developed a check list

of data from
situation

and activity

of activity

the

times to

times (Table 42).

333

Using the two types of task interactions


development (precedence and simultaneous)
to develop individual

required

that arise during product

the interactions

between tasks

components were identified

This table used in conjunction

(Section 6.2.3).

with the NPDnetwork diagram (Figure 56)

for a component allows a designer to determine the interactions


tasks and hence calculate
component

within

between

development time for each

the overall

a sub-assembly.

of the method established'for

The validity

estimating

component

development times when tested against actual components from the New EHB
(Figure 59) indicated

project

that close agreement existed between

development
times.
estimated
and

actual

The cumulative

distribution

of development times estimated for

found
form
typical
to
were
a
unit
each standard

Pareto curve (Figure 60)

indicating

the possibility

of reducing the level

of managementcontrol

resources

by concentrating

on the small proportion

of components with

long development times.


"length
the
in order of
updating
isolate
time.

this

list

In addition,

maintaining

a list

of development time required"

of components

and constantly

(as design proceeds) enables designers to quickly

those components that could affect


The construction

of such a list

the overall

development

could remove the need to develop

diagram.
update
a network
and

When limited

the conventional

design

design solutions

cannot

development time is limited.

Under

design resources are available

creating
of
methodology
be
undertaken,
always

and comparing alternative


particularly

if

384

these circumstances an appropriate


to create an initial

solution

improvement of this

solution.

on the relative

design resources for the

and allocate
Allocation

of resources could be based

to be gained from the overall

contribution

the product design.

design methodology to adopt would be

If the initial

cannot be improved

solution

by the designer then it must be the value engineering

sufficiently

to create alternative

teams responsibility

design solutions.

importance of alternative

Pleasuring the relative

PPBSproject

planning techniques illustrates

design

use of additional
the NPD process

resources

additional

although

since

is an increase

by which this

of routes

the overall

is
easily
always
not
route
of each
The effect

to be gained

The effect

there

are a variety

on profitability

could therefore

to a

be used as a

Design
therefore
would
resources
basis
measurement.
of
common
allocated
respect

the relative

cost of a component or assembly.

the potential

for cost reduction.

for
the
allocation
is also convenient
regression

be

In this

cost of components would-be used to allocate

resources on the reasonable assumption that the greater


cost the greater

from

quantified.

This factor

on the relative

to

applies

of manufacturing cost changes can be related

profitability.

product's

benefit

also

in profitability

can. be achieved.

in

obtained from the

This difficulty

resources.

with introducing

the problems involved

benefits
basis
the
of
measuring
a common

finding

design solutions

difficulties
basis
The
of
measurement.
a common

requires

success of

based estimating

the initial

Manufacturing

cost

of design resources since the

system developed in the present work allows

385

designers to generate cost data rapidly

and therefore

enables design

as the design process proceeds.

resources to be allocated

In addition

as new designs are created cost comparisons can be made with existing
of design resources carried

components and re-allocation

out when

necessary.

oof
To allocate

resources

therefore

of a Pareto distribution

the construction

a simple method would involve

curve for component costs and

cumulative cost axis to cumulative design improvement time,

the

convert

design

design time to be obtained as a direct

hence allowing

measure of

component costs.
Since cost is but one of several important product characteristics
the allocation

of design improvement time based on relative


the need to improve other characteristics

be weighed against
increase

8.6

the overall

which

worth of a product.

CAD
Through the installation

have the use of a sophisticated


designing
parts,

cost needs to

custom-specified

and extension

Engineering

of the Autotrol

system(194) H.M. Ltd. will

Computer-Aided-Draughting

system for

products assembled from existing

standard

into the Computer-Aided Planning and Computer-Aided

(e.
NC
part programming and stock control).
g.
areas

The initial

introduction

of DCL's EUCLIDsystem(195) can be

low cost through a time sharing facility


at
accomplished
hardware
digital
as
expensive
such
a
to
purchase
the need

which overcomes
plotter.

Hence

386

both the experience gained and the initial


using this

system will

preparation

of data files

enable the payback period for the more expensive

system to be reduced.

Autotrol

Since Autotrol
the system in its

is essentially

an automated draughting

basic form has numerous limitations

system,

when used to

NPD
the
design
the
process.
of
phase
aid

In order to assess the

of these limitations

an examination of the

significance

relative

tasks involved

design of the New EHBwas undertaken.

in the original

This examination identified

design tasks and determined the

devoted
designer
to the accomplishment of each task
time
a
of

proportion

during the preparation


the results

of a product design.

It must be recognised that

(Table
45)
the
particularly
examination
of

the proportion

task
to
the
design
to
devoted
New
the
EHB
applies
each
only
time
of
of
between
types.
vary
product
would
and
product
design
require
could
product

For example, an innovative

a higher proportion

of creative

design

design
New
the
the
EHB
of
with
range since
compared
when

effort

based
is
those
design
's
on
of competitive
Ltd.
M.
H.

In determining

the applicability

of the Autotrol

EHB
New
the
the
contribution
development
of
the
involved

products.

system in aiding

of CADto each activity

in the design process has been examined.

387

(i)

Function Analysis
basic form CAD systems are merely automated draughting

In their

therefore
data
systems
and
storage
and

require additional

software modules in order to carry out the functional


analysis

designs.
of product

types
within
product
of
nature
is possible

Because of the diverse


the mechanical engineering

that a wide range of functional

industry

it

analysis

software packages for such applications

design and stress analysis


design of wire-rope

hoist

would be applicable
blocks,

as gearbox
to the

and would be available

"off-the-shelf".
In those cases where "off-the-shelf"

software is not avail-

developing
depends
the
software
on the
of
economics
able
of the analysis

complexity
its

Since new product ranges are developed at H.M. Ltd.

use.

intervals

at infrequent
analysis

required and the frequency of

it

is unlikely

that functional

software would be an economic alternative

involve
that
methods
manual
present
calculators

and scaled models.

to the

using desk top

However use of software

that could be applied to heavy, custom-designed cranes


(which represent

the major portion

of the design work carried

be
in
)
terms
Ltd.
H.
M.
economic
more easily
could
out at
justified.

It would be of benefit,

the functional

analysis

therefore,

to examine

requirements for the complete range

in
H.
M.
Ltd.
identify
to
at
order
manufactured
of products
be
that
economically
would
packages
software

viable.

388

(ii)

and Transfer

Data Acquisition

of CADsystems to hold large quantities

The ability

design data (including

of

a complete component definition)


of data by the design

and transfer

enables the acquisition


department to be greatly

improved.

The advantages of a CAD

include:
would
system

(a)

An increase in the amount of data stored on an


drawing.

engineering

data transfer

(b)

Instant

(c)

The possibility

to produce drawings

The ability
for

of preparing NC part programmes

from the CADsystem.

directly
(d)

facilities.

individual

manufacturing

Since the design process relies

on the acquisition

CAD systems can be expected to be useful

large quantities
by the various
to limit

of cost

of data it
functional

the present

data using

of data

in the NPDprocess,

in the development

However although CADsystems can store

the usefulness

For example,

information

operations.

departments involved

from other functional

designs.
optimum
of

hold

that

is the generation

departments

that

of this

could

be expected

of CAD systems to the design


research

has shown that

H. M. Ltd. 's present

estimating

data

process.

the generation
system is

389

slow and cannot provide the amounts of cost data required


by the design process.

Hence the development of the

system in the present work could be considered

estimating

to the economics of CADsystems when used to aid

helpful

the NPDdesign process, in that the time delay involved


in data acquisition
e Linking

eliminate

interaction

was not required would then completely

the time delay between data acquisition

Design Creation

have an adverse effect

could

arise

work.

of shift

CAD systems could


of design

solutions

in the

These changes,

work and measured work

from the need of a company to quickly

the high investment

cost of purchasing

The present author considers


original

that

would take place

under which designers

such as the introduction

recover

(205)

on the creation

because of the changes that


conditions

and use.

and Improvement

There are reasons to suppose

rates,

reduced.

system to a CADsystem such that

the estimating

operator

(iii)

could be significantly

it doubtful

CAD systems.

that effective

design would take place at a CADterminal


is an unpredictable

since

the creation

of designs

incompatible

with the types of tasks such as, draughting

normally carried
it

out at visual

display

units.

process and

In addition

has been shown that a relaxed atmosphere aids the creation

of design solutions
possible

which is an environment normally not

at man/machine interfaces.

390

Conventional techniques for creating


are also incompatible

solutions

with the use of CAD.

for example, requires

Brainstorming(206)

as many design solutions

the creation
to-their

as possible prior
it

To carry out the technique effectively


quickly

design

alternative

of
analysis.

is necessary to

sketch design ideas then concentrate on generating

'Of
design
i.
flow
ideas.
to
ideas,
e.
create
a
of
more

Hence

the need to record using a keyboard(which requires

translat-

ing design ideas to the input language of a CADsystem)


would constantly

interrupt

Although having no direct

flow.

input to the creation

of design

the use of CADcould be expected to increase the

solutions
proportion
tasks.

the creative

of time a designer allocates

That is,

to the creative

the increase in speed with which designs

design
be
could
more
allow
solutions
analysed
could

to be

considered.
The use of CAD systems"to
by the subjective

nature

actually
of design

create

In addition

specify

decisions
of

to be made can be sufficiently

design solutions
time it

large

the number
that

could be generated manually in the same

takes to develop software to specify

to the computer.

is limited

problems and the ability

to logically

requirements.

designs

the problem

391

(iv)

Cost, Marketing and Manufacturing Analysis


At the present moment the major difficulty

in carrying

out the cost, marketing and manufacturing analysis


product design is the need to obtain relevant
the other departments in the NPDprocess.
involved

are usually

of a

data from

The departments

not designed to generate the large

amounts of data required

during product development.

This is a major reason for the development of the


based cost estimating

regression

the "day-to-day"

tasks,

In addition

system.

such as production

scheduling

out by these departments, often

and purchasing carried

over requests for data by the design

have priority
department.

of CAD systems to this

The contribution
process

when systems are developed

arise

will

designers

to quickly

facturing

data.

estimating

area of the design

generate

cost,

The development

system is a significant

marketing

estimating

with

the CAD software

extracting

and manu-

of the regression
contribution.

system to the CAD system will

this

to allow

based
Linking

be possible

data from the stored

drawing of a component and transferring

data to the estimating

module.
The eventual generation

of manufacturing

data by the designer

lists
the
construction
of
of producibility
around
centres
(207,208)
(209)
that are developed for
and data bases
tips

use by designers.

An alternative

would be the allocation

392

layer to a VE team, who could then analyse

of a data file

the manufacturing elements of a design and propose changes


be implemented to the original

that could quickly


(v)

design.

Draughting
Since CADsystems can include mechanised draughting
f

be expected

they could
design

to have a significant

However, the time a designer

process.

ideas to paper is not a significant

committing

design

of the total

impact on the draughting


only

activities,

(763 hours)
60110

is
CAD
approximately
of

spends
proportion
CAD

in terms of their

solely

since

of the total

Hence the maximum saving

expended.

on the

For the New EHB project

activity.

not be justified

systems could

represented

effect

systems

possible

these tasks
design
through

time
use

8,000.

of using CADto reduce the time required

The feasibility

to

is
drawings
dependent
design
the
on
engineering
prepare
of the product since this

complexity
detailed

functions.

design customer specified

This makes the use of CADto


heavy cranes economically feasible.

with the overall

Integration

of design

The integration
primarily

the amount of

design work necessary to prepare drawings for the

manufacturing

(vi)

affects

NPDProcess

with

the remaining

a problem of data generation

departments.

The contribution

limited
be
would

if

departments

NPD tasks

and transfer

of CAD systems in this


are unable

to quickly

is

between
area

393

generate cost, manufacturing and marketing data with which


to progress the development of a new product.
contribution

to the integration

of the overall

An effective
NPDprocess

if CADsystems were developed

could be made, therefore,

to aid the generation of product design data.

The transfer

of data between departments can already be effectively


I-eaccomplished using CADsystems.
In addition

to aiding the generation of NPDdata it

important
considered

by the present author,

is also

that CADsystems

must be expanded to enable:


(a)

A coherent,
developed
tools

information

cohesive
that

allows

Programming,

(e. g. Linear

Make or Buy models)


this

information

a series

optimization

(b)

this

of design

optimization

Dynamic Programming,

to be combined.

The basis

system would be the analysis

common data requirements


to transfer

system to be

and the development

data to specific

of
of the

of software

decision-making

and

tools.

The planning and control

(including

the allocation

of

design resources) of the design process using an


appropriate

planning technique.

project

such a system would be the necessity


to generate their

The basis of

to allow designers

own planning and control

data.

Hence

the importance of the scheduling technique developed


in the present work.

394

(c)

A system that ensures successful

product designs will

be developed.

is likely

At H.M. Ltd. this

to be

achieved by an expansion of the existing Quality


(215)
Assurance scheme
since product quality is but
one important design characteristic
The draft

of a product.

guide to design managementprocedures for

the assurance of quality issued by the British Standards


(216)
lists with respect to quality assurance
Institute
the important factors
department.

These procedures could be expanded to

include other essential


a product.

to be considered by the design

design characteristics

of

395

CHAPTER9

CONCLUSIONS

396

1.

system has been developed and is currently

An estimating

used at H.M. Ltd. that enables designers to quickly


designs as they are created,
designer.

using data readily

The system therefore

which are within

the current

cost product
to the

available

allows estimates

accuracy levels

being

(the accuracy of

of H.M. Ltd. ) to be

design
during
the
process, hence cost data can be used
generated
to greater

for:

effect

(i)

developing minimum cost designs,

(ii)

design
the
of
number
changes that occur after
reducing
the prototype

build phase has begun and so reducing the

design resources needed to develop a product,


(iii)

allocating

limited

design resources based on the relative

cost of designs,
(iv)

allowing

the designer to maximize product worth, and

(v)

allowing

the greater use of design optimization

techniques.

Methods have been proposed for modifying the conventional

regression

in
to:
order
procedure

(i)

Minimize the sum of the squares of "residual"


measured as proportions

This modification

values.
calculating

corresponding

"observed"

(which enables the method of

the accuracy of estimates and the objectives

of the regression
suited

of their

values when

procedure to be made compatible)

to high volume manufacturing

H.
M.
batch
Ltd.
of
environment
small

is more

systems than to the

397

Enable the volumes of individual

(ii)

the accuracy of their


is again suitable
it
where
magnified

components to influence
This procedure

cost estimates.

for high volume manufacturing systems

can help avoid errors

in cost estimates being

into large cost variances due to the volume of

components involved.
i

2.

A model, which allows the total


from the product's
by modifying

ten highest cost components, has been developed

the equation for the Pareto distribution.

3.

the need to prepare general arrangement drawings.

A check list

of development task times has been constructed

which the time required

to develop individual

The check list

estimated.

lie
that
would
components
to construct

with the data

allows designers to identify

on the critical

path without

be constantly

costs of alternative
classical

manufacturing

has. been constructed

using

EOQtheory.

to
to
be
apply
a variety
modified
The model can
situations,

Planning

updated as design proceeds.

A model which allows designers to compare the variable


design solutions

those

the need

the whole task network for the NPDproject.

therefore,
can,
schedules

from

components can be

when used in conjunction

interactions
task
on
provided

4.

Product

costed at the concept design stage

designs can be accurately


without

cost of a product to be estimated

of manufacturing

e. g. simultaneous purchasing and usage, when shortage

398

Period
Batch
Control
the
using
and
costs apply
control

technique.

facturing

In addition

that minimizes total

(PBC) production

the amount of in-house manumanufacturing costs can be

determined for a product design.

5.

A method of improving the rate at which the manufacturing costs


design can be reduced has been proposed.

of a'product

technique takes advantage of the "learning


requiring
all

6.

cost reduction

components within

The

curve" effect

methods to be applied

by

individually

to

a product design.

The tasks required to allow H.M. Ltd. to develop a method of


forecasting

the need for new product development have been described.

The development of such a forecasting

technique at H.M. Ltd. is

by
lack
data
the
the
moment
of
present
on past
at
prevented
development projects.

7.

The cost effectiveness

of using CADsystems to aid new product

development at H.M. Ltd. has been found to be dependent on:

(i)

the amount of draughting

(ii)

the availability
analysis

(iii)

involved,

of software for performing functional

of designs, and

the innovativeness

of the new product design.

399

In addition

it

adversely effect

is considered that the use of a CADsystem could


the creative

design process but would benefit

the NPDprocess by enabling effective


the various

departments involved.

of integration,

data transfer

between

This would improve the level

planning and co-ordination

achieved between

hence
improve
tasks
development
the
and
chances of
product
marketing a successful

product.

400

10
CHAPTER

FORFURTHER
WORK
RECOMMENDATIONS

401

work should include:

In the short term further

1.

Extension of the range of processes and operations


area work is presently

In this

costed.

Polytechnic

Leicester

that can be

being carried

to develop regression

out at

based models for

the cost of cast components.

estimating
i

2.

Aggregation of individual

to estimate the manufacturing

required

estimating

as sales quotes and tendering.

data base for

necessary

to identify

attribute

(reliability,

carried

out,

quality

costing

during

operation

in use, tolerances

type of material

the estimating

the design

the design

as type of design

be
to
considered.
needs

This data must


feature,

required

process
of
and

The use of a particular

from
the estimating
equation
a list

data that

type and size

finishes

and surface

being processed.

display
to
be
used
then

and product

process.

being performed,

system into

Research is

process.

and serviceability)

to such factors

then be related

for

the type of manufacturing

be
to
considered
needs

equipment

developed

of the software

a comprehensive

being

Hence

system could be developed for use by sales personnel

for such functions

Expansion

costs of designs.

to estimate costs would be reduced and an

the time required

3.

The objectives

equations.

area would be to reduce the number of design variables

in this

"

estimating

system developed could

of relevant

In particular

this

design data that


concept could be

402

designers
to
with relevant
provide
used
Additionally
containing

manufacturing data.

coding of design features would enable data files


software for the analysis

of functional

and solution

design problems to be referenced.

work would include:

In the medium term further


le
1.

The development of a comprehensive cost and design optimization


design
the
for
process.
system
for
the
available
software

This would involve writing

cost optimization,

resource allocation

the
development
techniques
design
and
of a common
scheduling
and
data base.

would be to allow designers to

The objectives

data into the system and call-up

input
relevant
and
prepare
appropriate
scheduling

optimization

The cost forecasting

tools.

and design

techniques developed in the present work would

in
included
be
such
a
system
whilst
undoubtedly

the estimating

data and transfer


both
developed
cost
generate
could
system
this

2.

data to a suitable

data file.

link
to
development
software
of
The
developed to a CADsystem.
CAD system to retrieve
engineering
estimating
presently

The objective

relevant

drawing and transfer


system.

the cost estimating

the time required

would be to allow the

design variable

data from an

this data directly

Hence removing the interactive

the
to
estimating
use
required

system

to the
element

system and reducing

to develop cost estimates.

403

In the long term further


of all

functions

to remove all

involved

current

work would need to consider the integration

in the NPDprocess.

barriers

would be

to the development and transference

data between departmental functions.

area by allowing

to be stored and quickly

retrieved

large

by all

departments.

In addition, a scheme would need to be developed for

co-ordinating

the activities

objective

of

A computer-aided engineering

in
important
this
role
an
play
system would
information
of
amounts

The objectives

of departmental functions

with the

design
is
the
that
success
of
a
product
new
of ensuring

considered at all

stages during the NPDprocess.

be
of
existing
extension
an
undoubtedly

Quality

Such a scheme would


Assurance methodology.

404

REFERENCES

1.

WAINWRIGHT,D.
Morris

2.

of Herbert
1974.

Benham Ltd.,

Hutchinson

1st Ed.,

Ltd.,

The Story

Cranes and Craftsmen:

W.H. and LOGAN,J. P. Strategy, Policy and Management,


NEWMAN,
7th Ed., South-Western Publishing Co., 1976.
Ile

3.

ANSOFF, H. I.
J. of General

4.

P. M.
GUTMANN,

"The Changing Shape of the Stategic Problem",


Summer 1979, pp. 42-58.
Management, 4,4,

"Strategies

for Growth",

IBM Product Strategy

Manual,
Seminar
University
Management
and
Centre, Oct. 16-17,1967, pp. 10-11.

of Bradford Management

5,

BROWNE,W.G. and KEMP, P. S. "A Three Stage Product Review Process",


Marketing Management, 5,1976,
Industrial
pp. 333-342.

6.

for
CRAWFORD,C. M. "Strategies
Dec. 1972, pp. 49-58.
Horizons,

7.

LAWRENCE,R. J.

"Planning

New Product

the Product

Development",

Strategy",

Business

Marketing,

Dec. 1970, pp. 34-37.

8.

KOTLER, P.
4th Ed.,

Marketing

Management:

Analysis,

Planning

and Control,

1980.

Prentice-Hall,

9.

R.
"Anatomy
G.
BANDENBURG,
F.
F.
and'
of Corporate Planning",
GILMORE,
Seminar,
Management
University of Bradford
Strategy
and
Product
IBM
16-17,1967,
15-25.
Oct.
Centre,
pp.
Management

10.

A.
A.
M.
"Product Strategy for
HAMMOUDA,
T.
M.
and
CUNNINGHAM,
J.
Management
Studies,
1969,
The
May
of
Goods",
pp.
Industrial
223-242.

11.

BERRY, D.
of Adversity",

"Strategic

and Growth in the Face


- Stability
Marketing Management 5,1976,
pp. 209-211.

Planning

Industrial

405

12.

WILSON, I. H. et.

15.

Planning

"P_owto Cope with Uncertainty",


1978, pp. 67-70.

BELL, P.
April

14.

"Strategic

for

Marketing",

Dec. 1978, pp. 65-73.

Horizons,

Business

13.

al.

ManagementToday,

OMAN,D. B.

"New Product Programs"; 1st Ed., American Management

Association
f

Inc.,

1969.

"Decision Systems Analysis


CAPON, N. and HULBERT, J.
Marketing Management, 4,1975,
Industrial
-Marketing",

in Industrial
pp. 143-160.

16.

"New Product Decisions by Marketing


CARDOZO,R. N. and ROSS, I.
Jan.
1972,
10-16.
J.
Marketing,
pp.
of
Executives",

17.

SOUDER, W.E.

"
18.

"Effectiveness

Marketing

Industrial

"New Business

FOGG, C. D.

Industrial

Process",

of Product Development Methods",


Management, 7,1978,
pp. 299-307.
Planning:

Marketing

The Internal

Management, 5,1976,

19.

Products FaiZ",
COOPER,R.G. "Why New Industrial
315-326.
4,1975,
Management,
pp.
Marketing

20.

RAWNSLEY,A.
J. Wiley

21.

B. S. I.

"Manual of Industrial
1978.
Sons,
and

Electric

22.

23

Ed.,

pp. 179-195.

Industrial

Research",

Cranes for

Cenral

Standards,

and Creating SuccessfuZ Engineering


Van Nostrand Reinhold Company, 1980.

LOVE, S. F.
1st

Overhead Travelling

Marketing

Workshops and Warehouses, British

Factories,

Development

COOPER, R. G.
Failure",
and

Planning

1st Ed.,

Use in
BS466,1960.

Designs,

"The Di.:.ens^.,ons of Industrial


New Product Success
43, Summer 1979, pp. 93-103.
J. of Marketing,

406

24.

25.

ROTHWELL,R. et.
Research Policy,

"SAPPHOUpdated - Project
3,1974,
pp. 258-291.

GERSTENFELD,A.

"A Study of Successful

al.

SAPPHOPhase II",

Projects,

Unsuccessful

in Process in West Germany", IEEE


and Projects
Transactions
on Engineering Management, Aug. 1976, pp. 116-123.

Projects,

26.

ANON,, "EZectric
for

27.

28.

Wire Rope Hoist


H. M. Ltd.

Range",
EHB
New
a
"Designing

H.
W.
-REA,
Engineering

for
Oct.

Designer,

"A New Procedure

WIND, Y.

Market Survey and Recorrnnendations


Internal
Report, Aug. 1977.

the Engineering

Industries"
1969, pp. 16-18.

for

Concept Evaluation".

The

J. of Marketing,

1973, pp. 2-11.

37, Oct.

29.

COOPER,R. G. "Introducing Successful New Industrial


Products",
European Journal of Marketing, 10,1976, pp. 298-329.

30.

EVANS, R. H.

"Assessing

Industrial

Product",

Introduction

Marketing

Factors

for

a New Industrial
Management, 7,1978,
pp. 128-132.

31.

BENTLEY,T. "ldeasuring Product Profitability';


July/Aug. 1979, pp. 17-18.

32.

WASSON,C. R.
Industrial

ManagementAccounting,

"The Importance

Marketer",

of the Product Life Cycle to the


Industrial
Marketing Management, 5,1979,

pp. 299-308.

33.

POLLI, R. and COOK,V. "Vatidity of the Product Life


The J. of Business, Oct. 1969, pp. 385-400.

34.

DOYLE, P.

of the Product Life


Summer 1978, pp. 1-6.

"The Realities

Review of Marketing,

Cycle",

CycZe",

Quarterly

407

"The Product

RINK, D. R.

35.

Strategy",

Cycle in Formulating

Life

Marketing

Industrial

Management, 5,1976,

Purchasing
pp. 231-242.

36.

RoZe Across the Product


FOX, H. W. and RINK D. R. "Purchasing's
Marketing Management, 7,1978,
Life CycZe", Industrial
pp. 186-192.

37.

"Row to Choose the Right Forecasting


J. C. et. al.
CHAMBERS,
Technique", Harvard Business Review, July/August 1971, pp. 45-74.

38.

ROGERS,E. M. "New Product Adoption and Diffusion",


Research, 2, March 1976, pp. 290-301.

39.

Diffusion,
Adoption
LANCASTER, G.A. and WHITE, M. 'Industrial
J.
Marketing,
European
10,5,1976,
of
Communication",
and

J. of Consumer

pp. 280-298.
40-.

MAHAJAN, V. et.
Market

41.

Potential",

al.

"A New Product

Growth Model with Dynxnic


Long Range Planning, 12,1979,
pp. 51-58.

MAHAJAN, V. and MULLER, E.


Growth Models in Marketing",

"Innovation

Diffusion

J. of Marketing,

and New Product


43, Fall 1979, pp.

55-68.

42.

NEWALL,J.
11,1977,

43.

44.

45.

"Industrial

Buyer Behaviour",

European J. of Marketing,

pp. 166-211.

in
Organisational
"Search
Behaviour
Buying",
K.
GRONHAUG,
4,1975,
15-23.
Management,
Marketing
pp.
Industrial

J.
O'SHAUGHNESSY
"Difference
in Attribute
R.
D.
and
LEHMANN,
J.
Industrial
Products",
for
Different
of Marketing,
Importance
38, April 1974, pp. 36-42.
J.
G.
WHITE,
"The Behaviour of Industrial
T.
M.
and
CUNNINGHAM,
for
Suppliers
The
Search
Machine
J.
in
Their
Tools';
of
Buyers
Studies,
Management
of

May 1974, pp. 115-128.

408

46.

WEBSTER,F. E.

"Modeling
2,3,

Research,

Marketing

the Industrial

Buying Process",

J. of

Nov. 1965, pp. 370-376.

47.

M.T. and WHITE, J. G. "The Determinants of Choice


CUNNINGHAM,
J.
European
Supplier",
of Marketing, 7,3,1973,
of
pp. 189-202.

48.

"Clues to the Design of a Marketing


8,2,1974,
European J. of Marketing,
pp. 168-179.

49.

BECKWITH, N. E. and LEHMANN,D. R.

KISER, G. E. et al.

Weights

in Multiple

Attribute

"The Importance

of Differential

Models of Consumer Attitude",

J. of Marketing Research, 10, May 1973, pp. 141-5.


50.

STARKEY, C. V.

"Profit

by Design",

the Engineering

Designer,

June 1969, pp. 12-17.

5T. -

PEKELMAN, D, and SEN, S. K.

"Mathematical
Weights",

of Attribute
the Determination
1974, pp. 1217-1229.
8, April

52.

HUBER, G. P.
and Field-Like
pp. 1393-1402.

53.

"Multi Attribute
Studies",

Utility

Programming Models for


Management Science,

Models:

Management Science,

WILDT, A. R. and BRUNO,A. V.

A Review of Field
20,10,
June 1974,

"The Prediction

Linear
Attitude
Equipment
using
"Capital
Research, 11, May 1974, pp. 203-5.

of Preference for
Models", J. of Marketing

54. -

The Missing Link in


ROTHSCHILD,W.E. "Competitor Analysis:
Quarterly,
Autumn
1979,
McKinsey
The
pp. 42-53.
strategy",

STARR, M.K.

56.

OSTROFSKY,B.

Product Design and Decision Theory, 1st Ed.,


1963.
Prentice-Hall,

ist

Ed.,

Design,

Prentice-Hall,

Planning
1977.

20,

and Development Methodology,

409

57.

April
58.

"A Decision Making

ARCHER, E. R.

Engineering

1975, pp. 27-29.

SHAFTEL, T. L.
Arizona

Industrial

Model",

"How Modular

June-July

21,6-7

Review,

Design Reduces Production

Costs",

1972, pp. 1-5.

59.

"Life

CycZe Costing: A Review of the Technique",


ManagementAccounting, Oct. 1976, pp. 343-347.

60.

OF INDUSTRY, "Terotechnology:
An Introduction
DEPARTMENT
HMSO,
1975.
Resources",
Physical
Management of

61.

TEDFORD, J. D. and FERRIS, T. J.

HARVEY,G.

"The Application

to the

of Group

at the Production Planning Phase of a New Product",


Engineer, Nov. 1976, pp. 643-646.
The Production

Technology

62.

BURBIDGE, J. L.
Heinemann,

The Introduction

of Group Technology,

1975.

63.

DEVARAJAN,D.A. "Care and Feeding of MRP", Industrial


Aug. 1978, pp. 22-27.

64.

BELIANIN, P. N.

65.

1st Ed.,

The Industrial

Robot,

STARKEY, C. V.

"Profit

The Engineering

Robots in Production

"Industrial
Sept.

Automation",

1978, pp. 141-143.

By Design:

Designer,

Engineering

Part

II

Design Decisions"..
May 1969, pp. 12-16.

66..

"The Long Range Planning of Research", British


Range
Planning
Long
for
Profits
Management,
and
of
Institute
Planning,
1968.
Development
July
Market
Product
Growth,

67.

CLARET, J.

SANDERS,G. S.

Accountants

More than a Question


"To Make or Buy - It's
1978, pp. 20-21.
Weekly, 28th April,

of Economy",

410

68.

T. R. "Marketing Strategy in a Make or


HAAS, R. W. and WOTRUBA,
Marketing Management, 5,1976,
Industrial
Buy Situation",
pp. 65-76.

69.

LEVY, H. and SARNET,M.


General Management, 4,1,

70. "

TAHA, H. A.
Collier

71.

WILD, R.
1st Ed.,

Operations

MacMillan,

"The Make or Buy Decision",


Autumn 1976, pp. 46-50.

Research:

An Introduction,

2nd Ed.,

1976.

of Production and operations


Rinehart and Winston Ltd., 1980.

Essentials
Holt,

J. of

Management,

72.

OSTWALD,P. F. and TOOLE, P. J. "How Do You Handle Tooling Costs",


Engineering, Nov. 1975, pp. 19-23.
Industrial

73.

BURBIDGE, J. L. "The Pseudo Economic Batch Quantity",


1981,
Jan/Feb.
3-8.
Control,
Production
pp.
of

74.

The Organised
CRUM, L. W. Value Engineering:
1st Ed., Longman Group Ltd., 1971.

75.

"New EHB Range: Value Engineering,


STOCKTON,D. J.
Reports, June 1980, pp. 1-23.
H. M. Ltd. Internal

76.

SWANSON,E.

"Determination

of Standard

77.

"Determination

be Manufactured Using Dynamic Progrru77fling",


401-406.
1968,
Aug.
19,8,
pp.
Engineering,

78.

Search for

Parts

Value,

I-III",

Sizes to be Manufactured

An
Extension",
Programming
Dynamic
Using
Sept. 1970, pp. 191-202.
R.
D.
ZERBE,
M.
B.
and
JACKSON,

Institute

AIIE Transactions,

of Standard Sizes
The J. of Industrial

C.
I.
"Case
PIFF,
Study:
C.
Determining
J.
M.
and
MARTIN,
be
Goods
Capital
to
Manufactured",
Sizes
of
Standard
of
443-456.
9,4,1971,
Research,
pp..
Production
J. of

to

the Range
International

411

79.

80.

M. "Selecting the Best Lengths to Stock",


WOLFSON..
-L,
Research, 13,1965, pp. 570-585.
REUTER, V. G.

"Using Graphic Management Tools",

Management, 30, April

81.

Operations

J. of Systems

1979, pp. 6-17.

Modern Production Management, 5th Ed., John Wiley


466-467.
1977,
Sons,
pp.
and
BUFFA, E. S.
/1

82.

SMITH, L. A. and MAHLER, P.

"Comparing CommerciaZZy Available


Engineering,
April
1978,
CPM/PERTComputer Programs", Industrial
.
pp. 37-39.

83.

KAHALAS, H.
Implementation,

"A Look at Major Planning


Strengths

Methods:

and Limitations",

Development,

Long Range Planning,

pp. 84-90.

11,4,1978,

84.

CLELAND,D. I. and KING, W.R. Systems Analysis


1968.
Ed.,
Ist
McGraw-Hill,
Management,

85.

BROOKS, D. T.

"Designing

for

and Project

ProducibiZity:

Design Influenced
Automotive
Society
Engineers,
Progrcun",
Cost
of
Production
and Space Engineering and Manufacturing
National Aeronautic
Sept. 28-30,1971,
SAE Report
Meeting, Los Angeles, Calif.,
No. 710747, pp. 1-6.

86.

JOHNSON, R. C.
John Wiley

Optimum Design of Mechanical


1961.
Inc.,
Sons,
and

Elements,

1st Ed.,

87.

SIEGEL, M.J. et. al. Mechanical Design of Machines, 4th Ed.,


1965.
Company,
Textbook
International

88.

HERZOG,R. E. "Engineering and Purchasing:


1973,
46-50.
Aug.
23rd
Design,
pp.
Machine

Partners in Design",

412

89.

SIMS, E. R.
Society

90.

"Precision

of Automotive

SIZER, J.

An Insight

Costing

Operations",
of Manufacturing
Engineers,
Report No. 680575, pp. 1-16.
into

Management Accounting,

1st Ed.,

Penquin Books, 1977.

91.

BLORE, D.
Production

"Taking a New Look at Work Measurement", The


Engineer, May 1981, pp. 34-36.

92.

"MTM-VApplication",
Measurement, 4,3, pp. 22-27.
.

93.

OWEN,E. J.

KNOTT, K.

"Estimating

The J. of Methods-Time

in Light Engineering",

Work Study,

Oct. 1978, pp. 30-35.


94.

MCGRATH,M. and CONLON,J.


Interference

AZtouaances",

"Use Probability
Industrial

to Estimate

Engineering,

July

Machine
1970,

35-36.
pp.
95.

NADLER, G.
Richard

96.

Work Systems Design:

D. Irwin,

Inc.,

The Ideals

Concept,

Rev. Ed.,

1970.

J.
M.
LAHIFF,
F.
T.
ANTHONY,
and
AchiZZes' Heel of Decision
Aug. 1976, pp. 25-29.

"Estimate

Making",

Acquisition:

Managerial

The

Planning,

July/

97.

OSTWALD,P. F. and TOOLE,P. J. "IE's and Cost Estimating",


1978,
Feb.
40-43.
Engineering,
pp.
Industrial

98.

VERNON,I. R. Realistic Cost Estimating for Manufacturing,


Engineers,
Manufacturing
Michigan, 1968.
Society
Ed.,
of
ist

99.

of Production Norms in Manual


HITCHINGS, G.G. "Reliability
Dec.
1978,
Work
Study,
30-37.
Tasks",
pp.
Repetitive

100.

COOLING,W.C. "Understanding Production


22-25.
1978,
Dec.
pp.

Standards",

Work Study,

413

101.

SHACKEL, B.

and Technology
102.
.

103.

Handbook, 1st Ed.,

Ergonomics

Applied

Press Ltd.,

FEIN, M.

"A Rational

The J. of

Industrial

IPC Science

1974.

Normal in Work Measurement",

Basis for
Engineering,

18,6,1967,

pp. 341-346.

OWEN,E. J. "Realism in Work Measurement"3 Work Study, April

1981,

pp. 25-27.
104.

MANSOOR,E. M.

The J. of

Practice",

105.

FAHERTY, B. T. et.

al.

Industrial

Certain

Engineering,

Aspects

of Rating
Feb. 1967, pp. 184-190.

"The Quality

Study and Management Services,


106.

into

"An Investigation

SMITH, S.

"Better

Analysis",

The J. of Industrial

Work

of Work Measurement",
June 1970, pp. 467-470.

Work Standards

Through Statistical
Engineering,

July/Aug.

Time
1959,

pp. 261-265.

107.

in Work Measurement",
KLEIN, M. "Double ScunpZeEstimation
1959, pp. 193-196.
May/June
Engineering,
Industrial
of

108.

Evaluation
for
ARCHER, E. R. and SIEMENS, N. "Quantitative
Industrial
Engineering,
Jobs",
1971,
March
Production
Repetitive

The J.

10-15.
pp.

109.

A.
J.
"How
ROWE,
Much Does Forgetting
J.
G.
and
CARLSON,
1976,
40-47.
Sept.
Engineering,
pp.
Industrial

Cost',

110.

Work Measurement: SomeResearch Studies,


1968.
Ltd.,
Co.
MacMillan and

Ist

111.

G.
G.
HITCHINGS,

DUDLEY,N.A.

"Job Enrich'n nt",

Ed.,

Work Study, Aug. 1978, pp. 28-35.

414

112.

CARNALL,C. A. and BUTTERISS,M. "Job Design in the Process


Chartered Mechanical Engineer, Sept. 1977, pp. 57-59.
Industries",

113.

CURRIE, R.M.

114.

J.
HEWARD,

115.

THRELFALL,J.

"Making the Most of Work Study", Management


April 1979, pp. 19.

MAYNARD,R. B.
Study,

"Maynard Operation

Sequence Technique",

WILLEY, D. E.

The J. of Methods-Time

"MTM--V: A User's Report",


Measurement, 5,2, pp. 19-31.

118.

J. R.
BOEHRINGER,

1.1

Incentive

119.

E.
P. "PDS Work Measurement",
KOGOVSEK,
W.
H.
and
SECOR,
April
1967,
Engineering,
254-258.
Industrial
pp.
of

120.

STEELE, P. M.

Rates",

Work Study,

123.

Industrial

Chart Helps in Calculating


Engineering,

Fitting

June 1978, pp. 31.


The J.

and Assembly Estimating",

Feb. 1981, pp. 10-16.

"Case Study of Computerised Fitting


Work Study, Feb. 1981, pp. 34-39.

STOCKTON,D. J.
MSc. Thesis,
(Unpublished)

"50-Interval

"Computerised

STEELE, P. M.
Estimating",

122.

Work

Nov. 1979, pp. 10-14.

117.

121.

1978.

"Computers and Work Study in the USA"., Work Study


and ManagementServices, Feb. 1970, pp. 102-114.

Accounting,
116.

Work Study, 4th Ed., Pitman Publishing,

and Assembly

Robots and Work Handling for Welding,


Loughborough University
of Technology, 1979.
Industrial

KIRBY, P. W. "Category Estimating".


Services, April 1970, pp. 302-306.

Work Study and Management

415

124.

HARRIS, D. G. and MELVEN, D. G.


Easy Way", The Production

125.

FAIRHURST, J. H.
International

"Tootroom Work Measurement the


Engineer, Jan. 1974, pp. 26-30.

"The Accuracy

of Comparative Estimating",
Research, 12,3,1974,
pp. 399-407.

J. of Production

126.

KELLER, L. W. "The Use of a Computer in Cost Estimating",


ASTMETechnical Paper No. MM67-80, pp. 1-13.

127.

KOPP, K. K.
The J.

"A Computer Program for


Engineering,

of Industrial

Time Study Analysis",


18,2,1967,

pp. 147-152.

128.

MASON,A. K. and TOWNE,D.M. "Toward Synthetic


Engineering, 18,1,1967,
The J. of Industrial

129.

"Developing Standards by Computer",


WEAVER,R.F. et. al.
Engineering, Jan. 1978, pp. 26-31.
Industrial

130.

"Techniques for Measuring Work Yrainned in Computer


HOLLINGUM, J.
236,12th
April,
1973,
Engineer,
The
pp. 54-57.
System",

131.

O'NEAL, M. H. and MOORE,C.


Standards
Industrial

Provides

Tool for

Engineering,

"Mutti-Plant
Overall

Computer System for

Manufacturing

Control",

Aug. 1980, pp. 54-62.

Computer Management, Nov. 1976,

132.

ANON. "A Packaging Package",


27-36.
pp.

133.

Task just
HILDEBRANDT,H. and WOOD,H. "Mamvmoth
1977,
July,
7th
Europe,
pp. 14.
Computing

134.

ANON. "Estimating
10,2,1978,

Methods Analysis",
pp. 52-56.

System for

pp. 149-150.

Vickers",

a Mini Job",

Computer-Aided Design,

416

135.

"A Computerised Planning


HALEVI, G. and STOUT, K. J.
Engineer, April
Machined Components", The Production

Procedure

for

1977, pp.

37-42.

136.

LOGAN, F. "Let LOCAMTake the Paperwork out of Planning",


Engineer, June 1981, pp. 15-18.
The Production

137.

M.A. M. and PUGH,S. "Turned Component Costs",


MAHMMUD,
Engineering, Jan. 1981, pp. 35-37.

138.

'MAHMOUD,M.A. M.
Stage,

139.

The Costing

Ph. D. Thesis,
l(Unpublished)

MOORE,S. J.

of Turned Components at the Design


Loughborough University
1979.
of Technology,

"Calculation

of Machining Times", Work Study,

Oct. 1978, pp. 12-20.


140.

CHISHOLM,A. J. et. al. The Action of Cutting


Machinery Publishing Co. Ltd., 1959, pp. 57.

141.

to Workshop Technology,
KEMPSTER,M. H. A. Introduction
Press Ltd., 1972, pp. 372.
The English Universities

142.

M. Manufacturing Technology, 1st Ed., The English


HASLEHURST,
Press Ltd., 1969, pp. 32.
Universities

143.

ANON. MAAGGear Book,


1963, pp. 215-216.

144.

ANON. Machine Shop Synthetics,

145.

COATES,
J.
S.
K.
B.
"The
Assessment of Manufacturing
and
BHATTACHARYYA,
J.
Production
International
Research, 12,2,1974,
of
Technology",
159-177.
pp.

Ist

Tools,

Ist

3rd Ed.,

Ed., MAAGGearWheel Co. Ltd.,

PE Consulting

Ed.,

Group Ltd.,

Zurich,

1979.

417

K. V. and LAMBERT, B. K.
RAMASWAMY,

146.

Pass Turning

Single

International

Costs",

Operation

"Cost Optimization

Including

Inventory

J. of Production

Research,

for

a
and Penalty
12,3,1974,

pp. 331-344.
RAVIGNANI, G. L.

147.

Machining

for

International

Optimization",

a Fixed Demand - An Approach to


J. of Production Research, 12,3,1974,

pp. 361-376.
BARTALUCCI, B.

148.

"On the Optimum Selection

Operations",

Turning

of Machine Tools in
Annals of the CIRP, 17,1969,
pp. 445-452.

149.

BALLARD,C. and HICHKAD,R. "An Equation to Estimate Assembly


Engineering, Aug. 1969, pp. 20-22.
Time", Industrial

150.

OLIVER, R.

151.

THELWELL, R. R.

'An ORApproach to Work Center Standards",


Engineering, Dec. 1969, pp. 21-24.

Regression
-Industrial

152.

"On Evaluation

Industrial

of Linear

Programming and Multiple


The J. of
for Estimating Manpower Requirements",
18,3,1967,
Engineering,
pp. 227-236.

CROCKER,D. C. "IntercorreZation
and the Utility
of Multiple
The
J.
in
Engineering",
Industrial
of Industrial
Regression
18,1,1967,
pp. 79-85.
Engineering,
4

153.

CANNING, F. L.
The J.

154.

of Industrial

J.
D.
STOCKTON,
International

155.

"Estimating

in a Job Shop",
10,, 6,1959,
pp. 447-449.

Load Requirements

Engineering,

"An Approach to Improving Cost Estimating",


J. of Production Research, 20,6,1982,
pp. 741-751.

"Using
Techniques
in
Regression
E.
B.
Cost Analysis:
COCHRAN,
J.
Research,
Production
14,4,1976,
International
of
I",
part
"Using Regression Techniques in Cost Analysis:
465-468.
pp.
J.
Production
Research,
14,4,1976,
International
of
II",
Part
489-511.
pp.

418

156.

157.

SALEM, M. D.

Linear

Regression

Analysis

for

Work

The J. of Industrial
pp. 314-319.

Measurement of Indirect
18,5,1967,
Engineering,

Labour",

DONEY, L. and GELB, T.

"Regression

Industrial

Estimating",

158.

"Multiple

Engineering,

Short-Cuts

CycZe-Time

Feb. 1971, pp. 22-23.

RICHARDSON,W.J.
Regression",

"Measurement of Indirect
Work using Multiple
J. of Production Research, 9,4,
International

1971, pp. 481-486.

Work Study and Management

159.

MELLOR,R. 'Mlultiple Regression';


Services, July 1970,, pp. 572-574.

160.

"Faster Better Tooling


KLEIN, R. S. and TAIT, H. J.
Engineering,
Dec. 1971, pp. 12-17.
Industrial

161.

COATES, J. B.

"Tool Costs and Tool Estimating",


April 1976, pp. 201-208.

Engineer,

162.

MARTIN, D. D.

"Instant

Time Standards",

Industrial

Estimates",

The Production

Engineering,

Nov. 1971, pp. 24-27.


163.

KOLOC, J.

"The Use of Workpiece Statistics


to Develop Automatic
International
J. of Machine
Progrcvrnning for N. C. Machine Tools",
Tool Design and Research, 9,1969,
pp. 65-80.

164.

OPITZ, H. et.

al.

Application",

International

9,1969,

165.

"Workpiece

and its Industrial


J. of Machine Tool Design and Research,
Classification

pp. 39-50.

GOMBINSKI,J.

"Fundamental Aspects of ComponentClassification",


Annals of the CIRP, 17,1969, pp. 367-375.

419

166.

FOGG, B. et.

al.

of Press Tools",

167.

168.
,

as an Aid to the Design


Annals of the CIRP, 19,1971,
pp. 141-151.

A. H. "Classification
for Automatic
SWIFT, K. G. and REDFORD,
Assembly of SmaZZProducts", Annals of the CIRP, 27,1978,
pp. 435-440.
J. L. and SCHILPEROORT,B. A. "An Evaluation
REMMERSWAAL,
of
Systems in Production Engineering",
Classification
Existing
Annals

169.

"Component Classification

of the CIRP, 20,1,1971,

CRAWFORD,S.
Stoughton,

Basic Engineering

170.

171.

ANON. Standard Data for


1978.
Publication,

172.

ANON.

173.

Welding

2nd Ed.,

Hodder and

2nd Ed., John Wiley and Sons

Arc welding,

"Welding Economics:

The Welding

Some Basic Practical

and Metal Fabrication,

MYERS, L. W.

Processes,

1972.

BOSTON,O.W. Metal Processing,


Inc., 1958.

Welding

pp. 91-92.

"How to Calculate

Design and Fabrication,

Institute

Aspects",

March 1977, pp. 85-91.


and Control Welding Costs,
Nov. 1978, pp. 78-86.

174.

A.
FOX,
K.
M.
Methods of Correlation and Regression
and
EZEKIEL,
John
Wiley
Ed.,
3rd
and Sons, Inc., 1959.
Analysis,

175.

HAWKINS,J.
Accounting,

176.

"Applying Matrices
April

to Regression",

Management

1979, pp. 37.

Statistical
SPSS:
Package for
H.
N.
al.
et.
NIE,
Hill
McGraw
Ed.,
Book
Co.,
1975.
2nd
Sciences,

the Social

420

177.

MISTRY, B. and STOCKTON,D. J.


1981.

Report,

Internal

"DESCOSTIMATE", H. M. Ltd.,

178.

D.W. "Price, Cost and Product Size",


PEDDER-SMITH,
Accounting, April 1976, pp. 136-137.

179.,

AUSTRALIANSTANDARDS.Drwns and Sheaves, AS1418 Part 1,1977.

180.

FEM.

Federation
181.

HILL,

the

of

Stock Control

Press Ltd.,
L. S.

Design

"Towards a Co-ordinated

Industries,

1st Ed.,

Approach to the Structuring

A Case Example",

Operations

"A New Quantitative

Research,

Technique

for

Problems
and Organisational
of Industrial
J. of Production Research, 12,2,1974,
pp. 313-330.

Identification
International

184.

Equipment,

Section 9.

in Manufacturing

SWAN, A. W. and WORRALL,B. M.

KING, J. R.

Lifting

1968.

Models:
Design
Cost
and
of
13,1965,
pp. 619-631.

183.

Serial

Europeenne de la Manutention,

THOMAS, A. B.
Gower

182.

for

Rules

Management

Gap in Job-Shop Scheduling",


March 1976, pp. 137-143.

"The Theory-Practice

The Production

Engineer,

185.

"Computers - The Effective Way to Manage


NUTTMAN,L. et. al.
Engineer,
July/Aug.
1979,
Production
The
pp. 21-26.
Manufacturing",

186.

is a job for
"Keeping Tabs on Shop Floor Activity
17th
Engineer,
Jan.
1980,
The
43,45.
pp.
microprocessor",
the

187.

HANCOCK,C. "Integrated Manufacturing Systems -a


1979,
Dec.
Engineer,
pp. 49-50.
Production
The

188.

MORTIMER,J. "Computer Draughting a TooZ - Not a Takeover",


47-48.
1981,
Feb.
pp.
Eureka,

HOLLINGUM,J.

ChaZZenge",

421

189.

ANON. "CAD to be used in the Development of Nuclear


Today, 3rd August, 1981, pp. 8.
Engineering

190.

SIMKINS, P. "The Computer as a Production


TooZ';
Mechanical Engineer, Jan. 1980, pp. 53-55.

191.

NUSSEY, I. D.

Engineer,

PROLAMAT'76",
and Computers:
April
1977, pp. 51-52.

ANON.

"Easy Automation

Today,

22nd June, 1981, pp. 11.

for

the Drawing Office",

193.

WALKER,B. Designing by Computer", Engineering,


1104-1108.
pp.

194.

HEWITT, E. C.

195.

Oct. 1980,

Design and Draughting at


Managing Computer Aided Design, I. Mech.
1980, pp. 9-20.

Publications,

BOND, A.

Engineering

"Computer Assisted

Davy-Loewy Limited",
E. Conf.

Chartered

"Manufacturing

The Production

192.

Reactors",

"Starting

CAD at a Lower Entry

Fee",

Computer Systems,

June 1981, pp. 78-81.

196.

HOLLINGUM, J.
The Engineer

197.

"Computer Graphics
24th July,

Survey,

Playing

a More Vital
1980, pp. 28-29.

RoZe"3

WATTS,G.A. and VEENMAN,P. "Computer-aided Design in the NonComputer-Aided


Design, 8,1,1976,
Industry",
Metals
Ferrous
pp. 13-16.

198.

Design,

199. -

"Selecting

LAXON, W.R.
9,4,

Oct.

SHOPPEE, G. J. V.
development
Guildford,

and Evaluating
1977, pp. 233-237.

"Current

tends",

Structural

CAD Systems",

Computer-Aided

analysis
systems and
CAD 76, IPC Science and Technology Press,

1976, pp. 206-212.

422

200.

HEATH, H. H.

"MELBAProgram Solves Bearing

Engineering,

May 1977, pp. 53-55.

Design

Problems",

201.

"Design of Low-Noise Gas Valves",


F. J. et. al.
NEWTON,
Aided Design, 9,2, April 1977, pp. 135-140.

202.

COATES,P. J.
1,1,

203.

Computer

"Design:

Ends and Means", Computer-Aided Design,


Autumn 1968, pp. 27-32.

MASTERS, E. H.

"The Engineering

Effectively
most
can
.

Aid"

Design Functions

Computer-Aided

Design,

which the Computer


Autumn 1968,

pp. 33-38.
204.

PITTS, G.
Design,

205.

"Computer Aided

Design:

Use and Misuse",

Computer-Aided

Summer 1970, pp. 41-43.

COOLEY, M. J. E.
Function",

"Impact

of CAD on the Designer and the Design


Computer-Aided Design, 9,4,
Oct. 1977, pp. 238-242.

206.

ALGER, J. R. M. and HAYS, C. V. Creative


1964.
Ist Ed., Prentice-Hall,

207.

COLDING, B.

Economics and Industrial


Annals of the CIRP, 17,1969,
pp. 279-288.

208.

COATES, J. B. et.
Technological

Synthesis

"Machining

al.

"Improve

Data Base",

Productivity

in Design,

Data Manuals",

by Establishing

Management Accounting,

Dec. 1976,

pp. 437-440.
209.

"CAD/CAMand Group Technology:


The Answer to Systems
Industrial
Engineering,
Nov. 1980, pp. 66-73.
Integration",

210.

Report:
ANON. 'International
Inc., 31st Jan., 1980.

KNOX, C.

Sverre Munck A/S';

Dun and Bradstreet,

423

211.

BRITISH STANDARDSINSTITUTION.
Stresses

in Cranes and Design Rules,

BS2573, Part

212.

"How to Make Better Decisions",


Jan. 1976, pp. 21-22.

213.

A.
ROMANO,

SIMPSON,L.

Allyn
214.

Applied Statistics
1977.
Inc.,
Bacon
and

BAKER, R. J. and NELDER, J. A.


The GLT.4 System,

Modelling:

for

Specification

Permissible
2,1980.

ManagementAccounting,

1st Ed.,

for Science and Industry,

Generalised

Linear

Interactive

1st Ed., Royal Statistical

Soceity,

1978.

215.

BSI.

"Guide to the Operation

Systems",

21.6.

BSI.

BS5179, Part

"British

Standard

and Evaluation
3, December 1974.

Assurance

Guide to Design Management Procedures

the Assurance of Quality",


79/62358 DC, May 1979.

for

of Quality

Draft

for

Public

Comment,

424

BIBLIOGRAPHY

LILIEN, G.L. and LITTLE, J. D.C. "The ADVISORProject:


A Study of
Marketing Budgets", Sloan ManagementReview, 17, Spring 1976,
Industrial
pp. 17-31.
SAUNDERS,A. and SMITH, N. J.

"Hydraulic

A Product
EZswick Works Z846-2936:
May 1970, pp. 55-70.
IMRA Journal,
WILLSMER, R.

"Long-range

Management Decision,

Corporate

Life

Cranes from Armstrongs


Cycle in Capital

Planning

Goods",

and the Marketing

Men",

Autumn 1967, pp. 12-15.

REISMAN,A. "Forecasting
May 1976, pp. 38-45.

Short-Term Demand", Industrial

Engineering,

PARKER, G. G. C. and SEGURA, E. L. "How to get. a better forecast",


4arch/April
1971,
Review,
Business
pp. 99-109.
Harvard
'
J.
W.
COAKER,
The Importance of Price as a Choice
J.
P.
and
KELLY,
Industrial
Decisions",
Purchasing
Marketing
Industrial
for
Criterion
Management, 5,1976,

pp. 281-293.

D.
A.
ROBERTS,
"The
T.
M.
RoZe of Customer Service in
and
CUNNINGHAM,
J.
Marketing,
European
8,1,1974,
of
Marketing",
pp. 15-41.
Industrial
"Customer Service in Industrial
European J. of Marketing, 10,3,1976,

BANTING, P. M.

Marketing:

Study",

pp. 136-145.

A Comparative

in
Preliminary Design", Society of
Making
"Decision
C.
GOTTWALD,
730889,1973,
No.
1-8.
Report
Engineers,
pp.
Automotive
'A Mechanised Assembly System:
TEWARI,
N.
K.
W.
B.
and
HEGINBOTHAM,
International
Tolerances",
J.
Design
Component
and
of
on
Its Influence
16,1,1978,
63-75.
Research,
pp.
Production

425

DOGRAMACI,A.
Inventory

"Design of CommonComponents Considering Implications


AIIE Transactions,
Costs and Forecasting",
June 1979,

of

pp. 129-135.
BURBIDGE, J.

"What is

Wrong with Materials


Requirement
Engineer, Oct. 1980, pp. 57-59.

The Production

"Trends in Production
PENDLEBURY A. J.
Inventory
and
i
Jan. 1979, pp. 18-22.
Management Accounting,
SATO, N. et.

al.

27,1978,

Management",

"Group Technology
for

Methodology

An Integrated

Planning",

Requirements Planning:
and Materials
Production
Control, ", Annals of the CIRP,

pp. 471-473.

SERPA, J. R.

"Manufacturing

and Design Engineering:

IEEE Transactions

Cost Reduction",
pp. 13-15.

on Manufacturing

BIRCHALL, J. "Marketing and Design:


Sept/Oct. 1975, pp. 13-16.
for

"Designing

HERZOG, R. E.

Foreign

Partnership

in

Technology, Dec. 1972,

Part One", Engineering Designer,

Markets",

Machine Design,

Aug. 22,

1974, pp. 73-77.


MILITZER,
Design",

"Needed:

R. W.

Manufacturing

Manufacturing

Engineering,

Engineering

82,2nd

Feb.,

M.G. "On VaZue Added by Design",


HODSKINSON,
15-16.
1978,
pp.
Nov/Dec.
G.
BOOTHROYD,
28,1,1979,

to Product

1979, pp. 2.

Engineering Designer,

Economic Manufacture",

Annals of the CIRP,

pp. 345-350.

and Its Use to Managers",


Feb. 1980, pp. 144-158.

GORB, P.
Arts,

"Design for

Input

'Design

J. of the Royal Society

of

426

ARCHER, B.
Industrial

"Design as an Element in a New Approach to Britain's


Quarterly
Review of Marketing,
Future",
Winter 1976,

pp. 1-7.
RAO, V. R. and SOUTAR, G. N.
Decision

Decisions",

OXFORD,T. R.

"Subjective

Sciences,

6, Jan.

Evaluations

for

Product

Design

1975, pp. 120-134.

"Designing for Production"3

Engineering Designer, Feb.

1976, pp. 1'1-14.


L. and RENEGAR,B. "Federal Express
Labour Sector",
Measurement Index for Indirect
PHILBIN,

Corp. Develops Performance


Industrial
Engineering,

Aug. 1980, pp. 48-53.


"Application

KANON, D.
Engineering",
ALLAN, W.G.

to Industrial
of Nonparametric Statistics
Engineering,
The J. of Industrial
18,4,1967,
pp. 272-276.

"Decision

Systems and Work Study",


Nov. 1970, pp. 882-889.

Work Study and Management

Work Measurement",
"MTM-3 and Integrated
Oct. 1970, pp. 833.
Services,

Work Study and Management

Services,
ANON.

D.
WECHSLER,
B.
"Estimating Electronics
R.
W.
and
LASHER,
1972,
17-21.
Sept.
Engineering,
pp.
Industrial

Assembly Costs",

"A Simplified Approach for Estimating


BLANCHFIELD,D. E. et. al.
Society
Level",
Part
the
Cost
of Automotive Engineers,
at
Manufacturing
1-9.
730908,
No.
pp.
Report
MURPHY,R. "A Computerised Standard Data System", Industrial
Oct. 1970, pp. 10-17.

Engineering,

R.
DOSSETT,
J.
"Computer
L.
C.
Use and Non-Direct Labour
and
BRISLEY,
in
Profession
transform
the new decade", Industrial
will
Measurement
Engineering,

Aug. 1980, pp. 34-43.

427

PARK, W.R.

"How to Make Quick Manufacturing


Jan.

Engineer,

KIDDNEY, R. P.

1972, pp. 44,46

Sept.

Consulting

and 48.

"A SmaZZ Company's Guide to Computerised

Engineering,

Industrial

Cost Estimates",

Shop Standards",

1978, pp. 54-55.

LIPTON, P. R. and DUFFY,D.J. "An Application of Statistical


Control to
The
J.
Industrial
Problem",
Measurement
Engineering, 18,9,
Work
of
a
01
1967, pp. 526-529.
"Time Study on TriaZ",
June 1970, pp. 462-465.
DAVIES, E. C.

ALLEMANG, R. D.

"New Technique

Engineering,

Industrial

KINACK, R. J.

"Activity

Evaluation

Developing Time standards";


1967, pp. 13-16.

Industrial
BECK, C.

CouZd RepZace Learning

Technique

The J. of

"Comparing Productive
Engineering,

Curves",

Aug. 1977, pp. 22-25.

for

ALLEN, D. S.

Work Study and Management Services,

-A Quick and Easy Procedure


Industrial
Engineering,
18,5,

Activities

at Different

Times"

March 1978, pp. 40-43.

"Work SannpZing Sets Direct

Labour Standards",

Industrial

Feb. 1969, pp. 22-25.

Engineering,

LIPTON, P. R. "An Application of Factorial


J.
Industrial
The
of
Process",
Measurement

Experimentation
Engineering,

to the Work
18,8,1967,

449-455.
pp.
Measures for Predetermined Time
"Developing Variability
Engineering,
J.
Industrial
18,1,1967,
The
of
pp. 120-122.
Systems",
NANDA, R.

CHRIST, C.
Industrial

"MRT: An Analytical
Engineering,

Way to Measure Indirect


June 1969, pp. 36-38.

Work",

428

"Activity

NADKARNI,R. P.

Sequence Analysis",

Work Study, June 1978,

pp. 35-38.
"Standards for Engineers",
Oct. 1975, pp. 1217.
SMERILSON,H. H.

MODER, J. J.

Engineering,

"Activity

Sampling with Applications


to Time Standard
Engineering,
The J. of Industrial
18,1,1967,
pp. 24-29.

Estimation",
ANON.

Industrial

"Employing

Synthetics

in the Workshop",

Work Study,

24, Feb. 1975,

pp. 32-36.
Synthetic

Values for

ANON.

"Obtaining

Study,

24, Nov. 1975, pp. 32-42.

EILON, S.

"Job Shop Scheduling";

use in Grinding

Work",

Work

The Production

Engineer, Jan/Feb. 1977,

Cost Behaviour",

Management Accounting,

47-48.
pp.
FREEMAN, J. M.

"Factors

Affecting

226-227.
1976,
pp.
June
HARLOW, I. P.

"Cost Models HeZp to Answer the Questions"3


1976, pp: 266-268.

Accounting,

July/Aug.

GUTMAN, N.

"How to Keep Product

22nd Jan.,
RIST, L.
American

Management

Costs in Line",

Machine Design,

1981, pp. 50-53.


"From CAD to CAMvia Expanded Factory
March 1980, pp. 134-137.
Machinist,

Applications

of Computers",

429

APPENDIXI

OF THE NEWEHB RANGE


DEVELOPMENT

430

an electric
the hoist

Ropemaster hoist concept was designed as

the existing

Initially

hoist
unit

block for runway duty.

During this

period the UK market was supplied from a large number

holding
Ltd.
H.
M.
the
predominant market
with
manufacturers

as suppliers

originally
entering

type
the

1960's European competitors

with units

enter the market seriously

their

strategy.

In the late 1950's/early

share.

1950's

then in production was adapted for use on cranes in an

endeavour to match competitors

of local

However in the early

the crane field

penetration
crane.

began to

designed to FEMstandards(180)

blocks to end users but gradually

of hoist

through minor manufacturers and increasing

because of the competitiveness

The Ropemaster during

this

period

of the hoist block


was modified

to increase

5.5
to
10
tonnes
tonnes
from
a
maximum
of
of
maximum
a
range

lifting

capacity.

's
Universal
Ltd.
M.
H.
The success of
this
of
the acceptance
increase

type of crane within

in crane capacities

Crane range launched in 1962,


the market and the gradual

and spans resulted

in further

extensions

in
1969,
4-Fall
10-tonne
that
a
standard
unit
so
Ropemaster
range
the
of
20
to
tonnes
stretched
capacity.
subsequently
was
which
introduced
was

increase
in
the
1960-1974
from
an
market
showed
The period
requirement

for electric

hoists

so that whilst

the market share of

imports,
high
foreign
by
volumes
being
were
eroded
Ltd.
was
H. M.
1300
levels
in
1969,1200
in
1970,
of
order
being sustained with

431

900 in 1971,900
due. to production
customers turning

in 1972 and 1250 in 1973.


constraints

during this

to other suppliers

Extended delivery
period resulted

to meet their

Since 1974 Ropemaster sales have declined


sold for runway applications)

of "hoist

700 per annum. The ratio


"blocks

with current

constraints

requirements.

(particularly

sales levels

for units

being approximately

crane" is approximately

created market acceptability

their
the
due
to
of
concept
which

in regular

blocks sold for runway duty" to

Universal
of
a
as
part
sold

The delivery

times

1.5: 1.

for competitors

design H.M. Ltd. have subsequently

been unable to overcome.

The delivery

constraints

arose due to:

methods being used.

(i)

Complex production

(ii)

Poor component supply.

(iii)

Lack of priority

given to the manufacture of the

Ropemaster.

Thus, whilst

in the early 1970's 40% of the machine shop capacity

Ropemaster
production,
with
occupied
was
22%.

this

is currently

running at

432

Traditional

home orientated,

markets, predominently

have shrunk

and due to the Ropemasters design philosophy H.M. Ltd. have been unable
by export sales.

to replace this

The design is labour intensive

coupled with the low volume is uncompetitive


hoist.

a straight

identified

but, whilst

still

acceptable in the home market the price content of the

technically
within

the crane necessitates

selling

led to a change in the hoist

concept within

is now an essential

Ropemaster unit

EHB product

and not a secondary


cranes

through

extend

the capacity

at a lower profit

of the Ropemaster as a straight

The uncompetitiveness

hoist

has

The

crane philosophy

The development

is again emphasising

of the current

margin.

the organisation.

of the total

part

range.

market acceptance

be
stretched
cannot
Ropemaster

It

in the market place as

Whenused on a crane, the cost of the hoist unit

to the customer is not separately

hoist

and

of hoist

block

the need to
However, the

range of cranes.

to meet these objectives.

is not possible to increase hoist

sales volumes appreciably

design
is
longer
Ropemaster
the
no
to
acceptable
present
other
since
distributors
crane

and assemblers who form the major portion

and export markets.


by
clients
commitment
unit

must effectively

The design philosophy means either


or alternatively
be built

long delivery

to contract.

a high stock

times since each

Current design parameters

do not allow a ready change of suspension method, lifting


extras such as speed variations.

re-reeving

or any optional

be readily

by
accommodated competitors

obsolete

products.

capacity

by

These can.

It is therefore

design, expensive to produce and inadaptable

demands.
market

of home

an

to meet current

433

Since the hoist

block is an integral

part of the Universal

crane

that H.M. Ltd. continue to market such a product.

it was important

This would also enable the company to retain


decades)
as materials
several
up over

Four options existed,

its market image (built

handling experts.

each of which could allow H.M. Ltd. to

hoist
block.
to
a
market
continue

Option 1
To factor

complete units

from another manufacturer.

have been achieved by buying low cost hoist units


Kone (UK) Ltd.
current
price

selling
structure

prices but with reduced profit

margins.

their

However, the

of Kone (UK) Ltd. represented a "dumping exercise"

up of their

This eventually

own manufacturing

facilities

's
Ltd.
in
H.
M.
decrease
crane sales.
to a gradual
limited

from the distributor

This would have allowed H.M. Ltd. to maintain

the
to
market.
them
penetrate
enable
the setting

This could

could result

to
in

which could lead


This option also

H.M. Ltd. 's export potential.

The decision

not to manufacture a hoist

following
the
produce

(a)

problems:

Since the machine tool

investment

1200,000 over the last

6 years

at improving
majority

block range would also

programme involving

has been primarily

Ropemaster manufacturing

of this

capital

expenditure

productivity

aimed
the

would be wasted.

434

(b)

Approximately 37% of the machine shop operatives


need to be made redundant.
the flexibility

would

The redundancy would lower

of the manufacturing

facilities

across

the company's whole product range and prevent a quick


response to unusual market demands.
(c) ' Redundancy on such a scale could have a repercussive
effect

on the workforce in other areas with a consequent


in performance generally.

fall-off
(d)

The other product groups would need to absorb the


remaining manufacturing expenses and would become less
profitable

(e)

as a result.

There are many components commonto the Ropemaster and


other product ranges.
materials

(f)

Inevitably

therefore

the cost of

would increase due to the reduction

The profitability

of the spares division

in volumes.

would be

lowered.
(g)

Increased stock holding costs would be incurred.

(h)

Market reaction
place continuing
product.
foreign
direct.

could be unfavourable

since many customers

business simply because it

These customers may be reluctant

is an H.M. Ltd.
to purchase a

block through H.M. Ltd. and may prefer


This attitude

to purchase

could also extend into the Universal

Crane business with consequent reduction

in order input.

435

Option 2
To obtain a licensing

This is not feasible

manufacturer.

since all

foreign

current

manufacturers

due to a down turn in the world market and are therefore

are suffering

in-house manufacturing

seeking to increase their

at the present time in this

not interested

levels

and hence are

type of arrangement.

Option

Collaborative

with

whom to collaborate

revealed

to hoist

respect

Consequently

An examination

manufacturer.

another

with

of a composite

manufacture

with

place

agreement from a suitable

that

of possible

manufacturers

many were strong

block manufacture

any deals would give

range in conjunction

and design

them minimum advantages

in the market
expertise.
and a strong

hand in any negotiations.

However one company Swerre Munck A/S (Munck) of Norway showed


greater

potential

a difficult
financial

in this

position
difficulties

area, since they themselves appeared to be in

in terms of a rapidly reduced turnover and recent


(210).
Numerous discussions were held with

to
establish
Munck management
existed.

the extent to which mutual advantages

However, although there were many advantages to be gained

(i. e. reduction

in the new product range introduction

time due to

developed
having
a new product range, reduction
part
Munck already
the
extension
of
a
possible
product range
and
development
cost
in
through joint

8.0
tonne unit)
an
of
manufacturing

division
the
be
of
reached
not
amount of royalty
work carried

agreement could

of the in-house work content and the

paid by H.M. Ltd. for the previous development

out by Munck.

436

Option 4
The fourth

the only feasible

choice to

i. e. to develop and manufacture a new range of wire rope

follow,

This option would then give maximumutilization

hoists.
plant,

option was therefore

maximumoverhead recovery,

conform to the company's image as

in-house manufacturers of the total

range of crane products,

in export markets and future

interests
is retained

of existing

protect

growth and ensure that control

of own manufacturing costs and selling

prices.

However the disadvantages are long development time, current


low sales volumes of company's products negating capital
investment of the technical

limited
a
on
compromise

design

and a possible

range.

From the detailed


basic

resource at the expense of other areas,

design
block
hoist
in
philosophy
experience

limited

investment,

market survey

specifications

for

carried

out in 1977(26)

the

the New EHB range were considered

to

be:

(i)

The hoist

block should be cylindrical

standard flange mounted squirrel


driving
reduction

through a central

drive

in shape with a

cage motor at one end,


shaft

gearbox at the other end.

to a triple
The gearbox is a

separate module making assembly and replacement easy.


(ii)

A standard failsafe

brake is required,

back-end of the gearbox and therefore


by removal of a weatherproof

cover.

mounted at the
easily

accessible

437

(iii)

The construction

of the frame must be open for ease of

assembly of the rope and rope guide and to lend itself

to

varying drum lengths with the minimum of components.

The

design of theframe must allow units


to either

to be easily

normal or low headroom monorail units,

versions or fixed suspension units,

converted
crab

as illustrated

in

Figure 62.

-(iv)

The unit must be designed to be easily


single

(v)

to two or four rope fall

convertable

from

versions.

The units must conform to FEM Ib(180) and BS 2573 Part 2


M3(211) standards.

(vi)

The range of capacities,

heights of lift

and lifting

follows:
are
as
required
speeds

Safe Working Load (SWL)


(tonnes) 1 Fall of Rope

Height of Lift
(Metres)

Speed of Lift

0.8

10

20

1.6

11

20

3.2

12

20

5.0

12

20

8.0

12

20

(Metres/min)

This is to allow the range to be extended as illustrated


Figure 63 to other FEMgroups at lower capacities.

in

438

FIGURE 62

SCHEMATICDIAGRAMOF THE NEWEHB CONCEPT


Ref.

1,/'

(26)

h:

1,

439

FIGURE 63

OF THE NEWEHB RANGE


LOAD SPECTRUM
(26)

Ref.

\\`11\. \

operating
in
dai
wcrkiny

Light
Medium

<1

1I1

411

i
>16

1-2 ; 2-4

4-8

8-16 f >1t
1
4-B
>9

160

320

640

200

40 0j__

800

250

500

Hill
Ivv
II
a

I6

640

1000
-

---

1250

800

00

IT
__

-j i

2000

1000

2000

L500
3200
4000

v
^\

O;

6400

5000
6400

8000

W
i

8000

4000

"5000
!nj'.

20000

6300

12500

8000

16000

32000

10000

20000

4-'

25000

j 'J

Ji
/

K,
itc

n.

xc
IN

'
--- ---- -
..

----___----1'
;:
I

v
:,

Y.
iti

I ---

2500

10000

ist

10000

5000

..

1600

.:

50 vO

3200

i:

%941

Z;

4000

2 500

1250

*
-L.
71c

3 2OO

1 2000

2500

1250

1000

-4

kg

k9

320

8-16

pa it y

Hoisf
kg

F E.M
roup

p? r

I4-8
2-4

0.5
0-5-1
<

Heavy

211

time
hours

Average

Lcad
Spectrum

440

(vii)

For ease of service


design feature)

(which is intended to be a particular

standard proprietary

motors and brakes

must be used as well as metric bearings,

oilseals

and

threaded components to meet export requirements.


From H.M. Ltd. 's marketing department a forecast

(viii)
e

future

sales of the hoist

of the

are:
Sales Forecast

(ix)

SWL

Market

1st

2nd

3rd

(tonnes)

Demand

Year

Year

Year

0.8

37 %

259

444

740

1.6

34 %

238

408

680

3.2

19 %

133

228

380

5.0

8%

56

96

160

8.0

2%

14

24

40

The unit must be designed for economic manufacture and


designing
by
on a modular principle.
assembly

The

be
to
motor, gearbox, wire-rope
are
modules

drum tube,

frame, suspension unit,


block (hook arrangement).

brake, rope guide and bottom

I
441

APPENDIXII

ESTIMATING SYSTEM: MANUALMETHODDESCRIPTION

442

The sequence of tasks

in estimating

labour

using the example of the 30 tonne Goliath

illustrated
spur gear

involved

(Drawing Number 4/14162/2)

costs

is

Crane idler

which is machined from an

EN24T forging:

()

(Figure

From the component drawing


machining

operations

required

64) identify

to manufacture

the
the part.

These are:

(2)

Turn outside

(b)

Face both flange

(c)

Face both boss faces

faces

Bore centre

(e)

Grind centre

(f)

Grind outside

(g)

Cut gear teeth

(Tols. )O.

lmm)

(Tols. > O. lmm)


(Tols. )O. lmm)

hole

(d)

Determine

(To1s.,: ; O. lmm)

diameter

(a)

hole
diameter

the appropriate

cost

centres

using

Table 46.

443

FIRST IDLER MT SPURWHEEL

FIGURE 64

A
A

il
i'
1

a
i,
i
;
i,
i;

i
1

139.70

O
Ql

O
co

co
CO

O
O

CC)

L(7

00

Q1

aM
co

i
1

nfl

I
I

I
I
I

92.10
Depth

Tooth
Module
All

14.29mm

Number of Teeth

= 45

6.35

Machining

= +0.40mm

dimensions

Material

EN24T

in mm's

(unless
Annual

Limits
otherwise

Quantity

stated)
= 50

444

PROCESSCOST CENTRES

TABLE 46

Pre-, Special and Post Machining

General Machining

Cost

Processes

Machine Type

Cost

Centre

Centre

Flame Cut

248

Bar Lathes

201

Saw

236

Bar Capstans

202

Shear

238

Chuck Lathes

203

Weld

239

Chuck Capstans

204

Gear Gutting

217

Centre Lathes

206

Milling

214

Vertical

210

216

Horizontal

Borers

(Prismatic

Operations)

(Cylindrical)

Grinding

216

(Surface)

Grinding

Borers

209

Spline

Cutting

270

NC Bar Lathe

258

Radial

Drilling

211

NC Chuck Lathe

258

227

Copy Lathes

219

Assembly

(3)

If general machining operations


appropriate

are carried

out and the

machine on which the component is processed is

determine
the
known,
relevant
not
is
extracted
which

cost centre using Table 47,

from Appendix V and shown here for

convenience.
The user initially

chooses the correct

depending on whether
It

is then necessary

centre

or not a boring

section
operation

of the table
is required.

to move down the columns until

is found where all

conditions

a cost

(i. e. component diameter,

445

RULES FOR DETERMININGGENERALMACHINING

TABLE 47

COSTCENTRE

ComponentVariables

Diameter (mm)
e
..

<

<

Length (mm)

Input Code

Boring Operation Required


51

152

70

0,203

92

305

70

0,201

Cost
Centre
es

76

152

-e1000

NC Bar
1,258Lathe

92

305

ie1000

0,201

394

178

Z 70

0,204

686

305

70

0,202 *

610

229

z1000

K
Chuc
NC
0,258
Lathe

686

305

,ei 000

0,202

<1778

610

191000

0,210

< 914

<5639

191000

0,206

No Boring

Annual Quantity

Operation

Required

51

178

70

0,203

92

610

70

0,201
NC Bar
258Lathe

76

e 254

21000

92

610

21000

0,201

394

178

70

0,204

686

305

70

0,202
NC Chuc1c
0,258 Lathe ,

610

279

21000

686

305

1000

0,202

21829

2 762

1000

0,210

22134

t5639

21000

0,206

General machining

cost centre

for

the spur gear

446
length and annual quantity)

reference to the spur gear the first

cost centre that can

machine a component 396.0mm in diameter,


and with an annual quantity

Hence with

are satisfied.

92.1cm in length

of 50 is 202.

Cost centres for machining the spur gear are:


Cost Centre

Operations
le

(4)

a-d

202

e-f

216

217

For operations
the relevant
abstract

carried
predictor

from this

out on machine group 202 identify


variables

appendix

for

using Appendix
cost

centre

IV.

An

202 is shown

in Table 48.

(5)

From the component drawing determine the values of predictor


variables:
(i)

Turn (Tols.

!o. lmm)Trurned lengths


Depths of cut

= 92.1mm
= 5. omm

No. Turned Diameters =1

(ii)-

(iii)

FTool

Face (Tols. sO.lmm)

Travel Lengths

No. of Facing Ops.


(iv)

Bore (Tols.

Flange

Flange

Boss

Side 1

Side 2

Side 1 Side 2 Total

28.6

28.6

23.6

Boss

28.6

11114

EBored
lmm).
Lengths
!O.
= 92.1 nn
57Bored Diameters = 88.9 mm
.
9o. of Bored Holes =1

114.4

I
C. o

c-,

C.C,

c:

(:

C.

N
L1

V9
L.

.._

C)

CJ

'

'L

i)

W
f-

t--)

1-

OW
UU

I
.

i
! -

"
S

Ell

O
=

L-

_
Z-
W

O
O

LLJ

[t

W
C)

W
r? "
C: )

=
O
-

N
W
J
C

,
W

Vf

C7

O
C

i.

W
}
O

W
C)

N
O
N

N
Q
O

p
W

'

f_

.a(

IW

N_

IN
O
U
m:
O
LL..
t/)
W
IO

.1

F-Z

co

9-+

.::

Ci
Z

C;

.
""

r_-

C )
.

Y
U

W
If - -

(. )

co

(1

O
C)

: r
..

C)

!.i1

N
N

rU

F-p

W
F"

-J

lV
LA

..

C)

C7
W

V)

N
..

..

Ln
W
C

W
}

4
W
J
co
Q

.
t.. i
-'

0-4 I

J
J

Ca
M

1--

\
Q

S
-

ME
CJ

== / v //Y

CC
W

G
U

W
C9

N
N
LJ
<)
C7
R

`)

u
f
v --

LL.
U

'-

. 1.

I.

J
C7

LA-

4J

C-

laJ

I--

Vj

_
G

C_)

ta.

.
v

PC
--

W
f:
'

J
-J
CC

'7

J
J
--

tf\

- VI
C
v

W
-

W W W

W
J
m

W
}

CO
qe"

F--

(il

_C
C
U

O
r-

""

A
.7- L

W
(_
Z-

T
-!
O

FK

_1

a
O

v
J
"
_
!v

. 'J
-

L..)

%D

>
O
O
CC

LT

()

Cr

O
"-

. "!

Lr

:V

h'1

U1

%D

-Z
O

r"

-"
" 's

(t:

F,
W

448

(6)

Insert

the coefficients

of the independent variables

(Table 49, Appendix V) in the relevant


extract

from this

equations.

An

appendix for cost centre 202 is shown

here for convenience.


Equation Code
i

20201 (0.113 x 92.1)

+ (0.041 x 5.0) -(2.24

x 1)

20203 (0.058 x 114.4) - (0.18 x 4)


20206 (0.049 x 92.1)

+ (0.037 x 88.9)-(2.09

x 1)
Total

TABLE49

8.37

5.92

5.71
20.00 mins.

VARIABLECOEFFICIENTSFORCOSTCENTRE202
PREDICTOR

111:
COSTCENTRE202 - CHUCKING
COMBINATIONS

112:
113:
114: 20201

0.113

0.041

-2.24

0.00

0.00

0.0

115: 20202

0.147

0.041

-5.77

0.013

0.0

0.0

116: 20203

0.058

-0.18

0.00

0.00

0.00

0.0

117: 20204

0.081

-0.68

0.00

0.00

0.00

0.0

118: 20205

0.032

0.065

-0.92

0.0

0.0

0.0

119: 20206

0.049

0.037

-2.09

0.0

0.0

0.0

120: 20207

0.095

0.032

-1.81

0.0

0.0

0.0

121: 20208

0.50

0.00

0.00

0.0

0.0

0.0

122: 20209

0.147

0.041

0.013

0.0

0.0

0.0

123: 20210

1.00

0.00

0.00

0.0

0.0

0.0

449

TABLE49 (Cont'd)

124: 20211

0.00

0.0

0.0

0.0

0.0

0.0

125: 20212

0.50

0.00

0.00

0.00

0.00

0.00

126: 20213

0.095

0.037

-2.09

0.0

0.0

0.0

127: 20214

0.00

0.00

0.00

0.0

0.0

0.00

20215
128:

0.0

0.0

0.0

0.0

0.0

0.0

129: 20216

1.10

1.00

1.00

1.00

0.36

0.0

130: 20217

0.003

0.016

-0.10

0.0

0.0

1.2

131: 20218

0.002

-0.009

0.95

0.14

0.0

-0.88

0.00

0.00

0.0

0.0

0.49

132: 20219

(7)

14.08

Multiply
(Equation

appropriate

The total

machining

1.1
20.00
x
=

time for

consideration

group.

the spur gear

22.00 mins.

To estimate the "floor

to floor"

for machine set-up,

and machine manipulation


the relevant

Factor

Material

into

Factor

Use Table 30 to

of the material.

determine

required

Material

Code 20216, Table 49) to take

the machinability

(8)

time by the relevant

machining

time (FTF) time, the time

loading/unloading

tasks must be calculated

equations and predictor

from Tables 48 and 49.

components

variable

using

coefficients

450

(i)

Load/Unload Machine
Equation Code 20217

Max. A Dimension

396.0mm
=

Max. C Dimension

92.1mm

Total

Ops

=6

Load/Unload time. =
(0.003 x 396.0)+(0.016

x 92.1)-(0. l x 6)

+ 1.2
(ii)

3.26 mins

Machine Manipulation
Equation Code 20218

Machine Manipulation

396.0mm
=

Max. C Dimension

Total Ops.

=6

Machine Time

x 92.1)+(0.95

+ (0.14 'x 22.0) - 0.88

22.00 mins

x 6)

7.86 mins

Machine Set-up

Equation Code 20219

Set-up time

No. of main ops.

=3

Batch Size

= 20

(14.08 x 3) + 0.49
20

Total Manual Task Time =


(9)

92.1mm

Time =

(0.002 x 396.0)-(0.009

(iii)

Max. A Dimension

The total

2.14 mins
13.26 mins

FTF time for the general machining of the spur

is:
gear
Machining Time

22.00
=

Manual Task Time

= 13.26

Total Machining Time = 35.26 mins

451

To estimate the labour and overhead costs the total

(10)

time must be multiplied

by the appropriate

overhead rates for cost centre 202.

FTF

labour and

Labour and overhead


in Table 50,

cost rates for all

cost centres are listed

which is extracted

from Appendix V and shown here for

convenience.
35.26 x 2.60
60

(i)

Labour Cost

(ii)

Manufacturing Overheads = 1.53 x 8.25 =

(iii)

Administrative

=1.53

Overheads =

(1.54 + 12.61) x 0.2

=12.83

Total Labour + Overhead Costs

TABLE 50

12.61

16.97

LABOURAND OVERHEADCOST RATES (f's/Hr)

Cost

Labour

Manufacturing

Administration

Centre

Rate

Overhead Rate

Overhead Rate

248

2.55

6.70

0.20

236

1.30

8.25

0.20

238

2.60

7.00

0.20

239

2.60

8.25

0.20

201

2.60

8.25

0.20

202

2.60

8.25

0.20

203

2.60

8.25

0.20

204

2.60

8.25

0.20

452

TABLE 50 (Cont'd)
Cost

Labour

Manufacturing

Administration

Centre

Rate

Overhead Rate

Overhead Rate

206

2.60

8.25

0.20

209

2.60

8.25

0.20

210`

2.60

8.25

0.20

219

2.60

8.25

0.20

258

2.60

8.25

0.20

214

2.60

8.25

0.20

216

2.60

8.25

0.20

215

2.60

8.25

0.20

270

2.60

8.25

0.20

211

2.60

8.25

0.20

217

0.85

7.00

0.20

227

2.25

4.90

0.20

(11)

Stages 1-

10 are then repeated

and overhead costs


operations.

for

to determine

the grinding

the labour

and gear cutting

453

APPENDIX III

(DESCOSTIMATE):
DESIGN COST ESTIMATING PROGRAM
LISTING
OPERATIONNOTESAND PROGRAM

454

Introduction

A. 3.0

of the program is to allow both the time. and cost of

The function
manufacturing

a component to be estimated.

the time and cost for each individual

by calculating

building
and
a
component
on
performed
times and costs.
appropriate

out this

function

machining operation

these into total

floor-to-floor

Machining costs are then determined using the

labour and overhead cost rates.

to be carried

The manufacturing operations


be analysed in the following

(i)

It carries

Initially

order:

the total

machining

Machining

be calculated.

out on a component must

time per cost

centre

must

can be analysed

operations

in

any order.

(ii)

The total

machining time calculated

must then be input as

a variable

into the "Material

correction

to be made for the type of material

Factor"

equation to allow a
being

machined.
(iii)

If Load/Unload and Machine Manipulation

times/costs

these must be carried

the total

required

out after

are

have
been
times
adjusted using the Material
machining
Factor equations.

To enable the estimating


filed
are
type,

under their

equations to be quickly

appropriate

i. e. Pre-Machining,

accessed they

works cost centre number and process

General Machining, Special Machining and

455

The cost centre that the component is to be machined

Post Machining.

be
determine
known
input
to
the
code that allows access to
must
under
the relevant

equations.

estimating

However, General Machining operations

can be carried

out on a range of cost centres.

fore available

to select

the correct

A sub-program is there-

input code if

Operation times and costs can be calculated

not already

known.

for each operation

for
the
the
total
to
component
and
manufacture
used

number of operations

used.

The information

(i)

required to run the program is:

The cost centre or machine type if


identification
relevant

(ii)

known.

The cost centre

forms part of the input code for retrieving

estimating

equations from the data file.

If the cost centre is unknown, i. e. for General Machining


only,

the component diameter,

length and number of components

machined per year.


(iii)

The operations

(iv)

The design variables

and costs required

to be estimated.

for the relevant

estimating

equations

in
the
which values for these variables
order
and
into
the program
entered

Many design variables


However to simplify
is available

variable

must be

(See Appendix IV).

are commonto a range of estimating


identification

a variable

in Appendix IV for each cost centre.

equations.

description

To gain the full

sheet

456

(and
the
reduce computer costs)
system
of

benefit

a standard form has

been prepared (Figure 47) to allow the values for the relevant
variables

to using the program.

to be recorded prior

Unless indicated

to the contrary

all

component dimensions are

in
mm's.
measured
r

The equations for each machining cost centre can be used to calculate
total

floor-to-floor

times and costs and machine set-up times and costs.

Manipulation
Machine
Load/Unload
the
or
Where
small proportion

floor-to-floor

of the total

included,
been
have
not
separate equations
for
calculating
the
equation
of

A. 3.1

times represent only a


time of the component,

i. e. these times form part

machining times.

Running Procedure
The system flow chart

'form the operating

(Figure 65) and the following

instructions

"Descostimate".
'computer program

for the estimating

procedure

of costs using the

The example used to illustrate

illustrate
to
the
is
that
used
component
of
procedure
II.
Appendix
in
estimating method

the manual

the

457
FLOWCHARTFOR DESCOSTIMATE

FIGURE 65

START

2
i

61

"I

15

21

1
11
T

12
22

23

16
13
17

END
18

19

20

14

ZO

1
I

BEST COPY
AVAILABLE
Variable print quality

458

I_`.:
(r

F! F...
Ea
F-_,
L`:
t
I.L.
1.
.
..

1i1_"

r
0

.:,
", ,
H

>,
r-C-

_
,.
:"H .-a
i-"

,"r

C)

....

0
a

'-."'

O
U

G.

J.
;^`..

I r,..
_4

c
0
. I4-)
CL
0
>
N
U
., _
3

a)
H

O
v

v1

Fi

an
(L)
L]
.
E
rO
i

O
i.
G.

cu
0

F-- F"

.0
N

La
f%

,a
CJ
. r..
E
b
x
C)
N
C
O
41
S..
C)
C1
0

L
O
F41
Q.
r0-4

459

i
U

N
. '..

c
O
4N
a)
-0
O
U

0
U

-y

. t-

"y
I11

C)

tij I-

H-"r
(..

-A
1!
f 't

N1

C)
C)
r

t0

to
-0
C
co

P
G)
-P
"r
N
O
U

C)

f0
S..
Cn
pU
SL7.4-C)
s
F-

WI
O

1=

/"{

(i

C.

`j"-

LLI

1_t

_1

13

WI-1=
=
1=

1_1 1=1

i2_
H

1-. 1= 1=

i=i

7'

xi i tD
c '-f C"J ,-" i

('"-

'? " .l}


+",q . --, ---r

k,L'i im
(""

'-

-'
-+

i' l t:'"i i: '"1i; '*i i'+! i: ""1i: '1 ': '"1 ;

t^i
i..,} ID 171 i"yi Fri 0

'"i

iri

IM U-5 i: J
rI
LL

'=1

J
W

a.
cn
c
"r
L

rt

c
3
0
c
Y

ch

..

J.
i

_
<L
4

I..
t: iC.
-t..
L:

.._ L.

{yj0
rF--
J

HA
c4
^_
=t
tz
U, otHir:
Jy
'-4
W
_I
JIi
ij: ?W
J F-i F-/
1 -a'-''='
....
Jt==LL'L;
L:
=
1:
ULL CO ti--:. CO CtoJ
;
1"1
I IX 1--IL

N
V1
C)

C
(o
E

F-

LL-4! 1-j L'eF-Jitj

i_1

I--

i'{

O
O

w
LLj P=W

f11

CL

bO
Q
N
"r
U

CC

11 ,r tl *'1r{

Ckf 4G
1_' E': 11
L L: :L1.
L. G
1'J W Cl '=' 1_t IT 1J 1-1Z 12 ==
+1i wl 1_J 1..J {_1 _: _
0')

v1
"r
r

-.j 1:'j t:"j 1t


1'j 1,

11)

I=C;

07,

=1!
J=IJjLck
I- L
1-- L

Q-1

"r

iT

I{[

L_

{T"{ IXI cc t 17"1


_:uj Ij 1 IJ"1 . -1

>_+1j 1:'j 1:'j 1: j r'j 1:'j


11
1_11: 111y11; 111; {

1:'11 17-11_r { Sj 1;;


lyt 1=t 1-t i=l --r

=
-

460

a
O
u

0
4-)
E
rts
S-

+-)

C)
a.a
N
C)
N

(A

'L7
C)

r
i-3
N
C)
Q'
C)
L
O
d-)
O

4J

i
C
0
N
"r

U
C)

C
"r
C
"r
to
C
O
U

-0

C
0

.
C)

0
a
0
4-)
C)
v
0
CJ

CL

4-3

a)
SCL

1
a,
C
O
"r
4-)
U
C)
N

()
C
"r
4--)
0
S..
.0

. a
I -.
r.
.
I-1.:

L
C)
N
C
Q

4:r

f0
O
4-)

L.
L"
S

r-

b
L

rn
0
i
a

C.
E
"r)

to

C.

L:

L.

}-"

LY:
l. 1

C)
4-A
(n
O
U
C)
E

"r7
C
()
t
4--)
rC

Oi
i-)

()
i

Cl

0
i
O.
C)

Lt

.C
+)

C
rt)

0o
C)

V1
G)
>1
v
>(4 .
1-

Lin

(A
C
0
",..
4. )
t0
C)
d

N
. r.

a
a

r
c
a

()
V)

L:

N
C

"r4-A

ON
4-

"r
t
V
A

.
0

a
a

"N

4-)

.
c
c
"r

tu
L
a
0
a,
co
a

4.3
C
OC
CO
O
a
EU
O
tJ

to

S-

V1
"e-

0
Eo
0
ca
4

a
s.

c
a

rl%

461

_}!
F-I
IIi

E
I

0
4-)

E
=3
. In

0
4-)
Es

-)
J

C)
V

C)

=t

OC)

.
C

II

. r.
L
U

u
p.

cm

S(L)

O.

C7

C)

s-

a. i

4-

n
C)
v,

C)
'fl
o

rt
U

1L1J
-j

E
L
O
44-)
V
C)
L
i
O
t
C)

I
rt:
W

i-

L'i

.,...

a,".,D
II

.0
o
U

G
41

4J

LJ

"LJ

t1,

cu
O

(1)
tl
41

C.

RS
i-

LLJ

OC',
"r
Vi

G)

"r"I
'r.

4-3

E
b

CI 1

C
CC
"r"
L
a)

.0

li..

"r

CL

A
C
t

fb

}-O
S..
C_

LL

r
L
d

4-)
(L)

F--

F-

0
So
4-

E
ro
L

.o U
U

C1

co

1l
N

CD
r-

462

}W

LL

vn

0
L
a

u
-

ftif

I-

4-1
C3.
"r

1=4

Vf
"r

>,.

CD,
r-
4

4J
M
C
(U
d-3
C
(U
C
O

O
U

<LD
w

r
N
01

Li!

V1
.
C)
4

C)
E

4-)
c
(L)

c
0
G
E
0
U

LL
t=.

4-J

C.

a)
V1
V1

`
i
.

a)
S""

Ls

N
.

a)

t
t

O
r`ti

ui
t

cC
tJ

a-
C: t4-1
E
1.
Q1
O
i

a
1.11 F-f
rr

"..

a
a

CL

"r

i
C1.
0
i
C1
C1.
to

a)

a)
3

4J`
G
O

1=4
L::
-r_":
w
-j
Lu
ij;

a)

CL.

-.

--

0
"
4-)
V1

L:S

r.

rr-

N
r-

N
fo
rCL

. '..

M
--

463

e
eo
OW
s..
CL

CL

++
A-3

'-

!W
i --: F-

C]..

.0
LO
cc
o

"r+'
U
N

LL

4-

1=1

`"'

o
4-

"

a..
U

N
O
U

J=
1:4 1-1

t0
N

tJ1
CL

LU i-

CWJ

CL
t. J

"n
O

OC

ci
s-

C
J

LL..

c.

O
.
4J
b

We.
1-r

tn

C,

h- W
_tQ
t-+

"p

t-. i t!
t,[

O
i
O.

F-

O
Q
C
. '..
4J
t0
r-V
r-t0

WW

F-

O
V-

2J L:

_
v

E
. r.

iJ

Lo

...

LLi

tz

E
(0
Str)
O
L
d

464

_Iw
H
Ll. l

w
0

a)
[]
. E.

.J_..
LL-

I'd
lal
CA

a)
s
+'

W
IW
II-I., III
Ii
i_I '. '
CI
H
f=
(l

yW

Cl CL

Lt. W
Wi.

LL W
Cl W
t:t: W
W=
wiN

Li

ID

r
N
"r
E

i
c7)
O
L
a

4-+

W sii
=W
f- A

(o
1.

OO
C)
CO

'.

N
N
C)

L1

4-3
N

i"C
0
4-

C)
V

4-3
C)
C)
i

01

O
()

00
r

COH
0O
"r
+
OCNS
Q'
W

C)
""

.C
4J

+J
OrOOU
CL
"r

4(A
+J
.r
r0

Oi

-t()
4-)

4-)
Nd

10..

NN

E
i
O0
4-

C
"r

"

C
r0

C
OL

C)

I1i

rp

t0

wON

OE
CJ
C)

t0

OEO
wLN
Cy)

'a
C)

t
+j

.0L.
+
d..

C)
a0
0
u-

C)
C.

Cd
O

4-

C)
.0

"r

C)

0
"r
+1
C

C)

rO

{. t

r
O

cr

C)

4-

. Li
O
.

OC

NO
-r
.
Ft

-r-

--

I---

Lc)
r

(a
bwL

S-

Cl.
O

F--

t0

r-

465

cu

c
a

a)
N

Q)

ON
C

.aC
TJ

Ll;

E0

(1)
i
S-

l t'_
L;
HON3
-,

t: tf

{:i{
_1

F-

_C

:V
iJ
i-

f-r

C, LL
l"'d

_ 0I

?W
1-+ 1.ID

!-

):L

(i ) Cj )-ti
Lt_
W
...
LL 'J
--I
N

f
r
N
. '..
"a

W
I=

CJ III
Ld

rOU

a)

"

C)

.C
4-

"v
i
a)
O

(1)
-0

Qj
L.
a)

a)
(A

4-)
rC

a)

a)

a)

S_
O
t1..,

+
V)

"

a)

Vi

"r

dw

W
CL WW
if)
LL
-

rr

1: J '"1

iQ

>C

4O
N

O
"r4J
V

a
N
"e

aJ
rro
>

3
O
.N

C)
I-

,--

t0

Co
r-

L.
d
lt')

rt7

O
iN
a)

tA

CN
"r
C
"r
rt3

C
"r
fC
4-1

5C

C)
CL

o
i3
"r"

L. V

a
E
(ts
S
aI
0
SCL

c>

;
s.

(1)

4-J

""
+j

Cc 0
__- ""-_

J3

-r

. r.

i::

(0

UNa"

RD

WHC

a)

a)

o
,i,
LL

"r"

466

c
b

,;c,
. -

c
0
.1"

ti
t-

,
i
RQ

Iil

...:,

ct

M-+

Vf
T
r0

t
t

N
.Q

7-Li
CL
17J

"r

a)

(L)
S..J

_I
Cl
IW4
J

0
u

II
Ihrr}
r=r
J
J
CI
r
F-

0
..
+.)

a
i
d
dw

ov

4-

"

"

N
4-J
N
ovx
aj

om
"*
rC
cr
W

C1

r--

Oi
C

iP-4

y-.
.,

. -,

W
C4

LL!
-
J

CL
"
tt

"r

cn
ro
N

Vf
(Li

e
L.
a

LL
Lti
W

H
i=
N
F-

cu
J3
r-

IJ

F-

>,
,-

o
444

t92-1

J
LL

"a
(1)
4-1
N
0
c.

a
C)
---

4-3
N

"r

E
r0
i

O
i

c.

C)

W
C
O
V
w

{-=i

.0

=t
f--r
!--

a)

a
V

Cn

I-

F-

W
;=,

r"_

O
i
O_

m
S41

4={

: G:

W
-j
6-4
LL

rq
c

EE
u

_}

+_
+11

t! i

CL
N
. r..
'L7

1N

1M

al
IN

Lt!
i:: l
IJ
t

467
rw

r- 1
e d

C
..4-1
OO

S- a 'C7 Ol
OCC
4" tC "rC)

N
4J
(A

eC)
+"
b
,a

" r- "r
At
C)
"r
"` U #O
C)
vl
(A C) Q
C). V) Nr
10
A
CA

'fl
U
C=
V1 cp 'C
C)
al
S. C)
'0
O :3
-0

C)
:.:

C)
Ua
C rt3 >
-+ ,. J 0

wS.
i-) (1)

UONC
C S- O
"-y
ddS?

C)

i
r.. ..,
.:

4-)

tic
i17
N
C)
E
"r
)N
C
O
"r
4-)
f0
L
C)
d
0

(t3
r--

r"

V)
-
. II

t4

a)

U
C)
N
0
N

i.

G1

LL.

ILL,

a1

I lI

4-J
C)
S_

2
fJ
It

t0
r
O

C)
4-J
V
a1
V1

S.
O
S_
C1
i-
N

'

i..

C)

1-:

4-)
an

fo

fit

E--

,_.
it:
w
7*
w
Cl
w
2
t-

41
0

C{-

I %1 0-)
W
'Tf

r_n ..
-+

0
u
C

C)
-0

"r

-0
c

eb
c

O
. r.
i-3
V

Cj

T
-
'_l

F-

lJ

--

1=L

! d-.

'r'

F=.
+f

!_-

-+

F--

.. L-

J
CL.

J
..
W

._W
'J

W-+

C
0

E
0
c

0
41

C1
b
4-

LL

CL
N

i_r) W

. r.

s.
Cf

0
.bi

.0
C)
S..

Q.
N

0CJ

F-

ii

1_

L++

LU

ta

F-

, -

F-

LL

;:;

F-

O
L

LL.

a.
aj
4-J

I.j

W
F-

O
.r..
U

w
1-4
t
=

'
r

.L.
_
,"'

CV

-
w

i::

.ir

c
N

LW
2
F-

. 7.
"

''t

Cl
S-

W
7
W
J-D

1.il

0w

.
.

"'
`
1-4

"r

'-'

"rC)

W
2

va1
+-A

FW

V
L

Li

. r.

O
. r..

LL

a)

u.
l
i
I
ii

i
-
ic..

F-

I
I.

C)

H
w

C)

to

1 '!

O
i
Q.

I. i. l

N
C)
N
N
C.1

r.
,
,
H
r-

rN

J
_0"
:."1

468

G)
f
tn
N
G)
So
SSCL)
rrt

O
"r
4-)
U
N
O
4-3
V1

s
i
Cn
O
SG
Q)

4-)

-0

N
. r.

.N-"
GJ
T
v

i'
4-

N
"r

E
co
50
i
L]..

U
3
O
J
Li.

N
N

Li
i....

EW
LLi

SCl.

CL

C)

ti-J
4-)
0'r'

t/1

4-)

CL
.r..
N
"r
r
O
C
v
Z
.4--

LL! )
O
O
C.

p'f
O

C.
"r

t_t
tZ

.a

S_
4-3
Q)
i

N
b

M
N

.
a
CL
a
4J

rts

4-

--.

"

w
"

"_

469

A. 3.2

Program Listing

1T,
IT'1
1.

e
..

H
cr

r_l
rJ
r".

17-1
J

-",
li1

J
J

1tTl

IT, fH

t_J 111
. '0
F-- --.

rl ief
I_t1J

Cl

-1

i_i
I
t t j0

C-:
CID

cl GJ P--i-f tJ.
f_:
r-

i::

t
CID
I-A

. -!

1J
! il

174
L'J W
I- Fy f=
1=4

W
t-

r-a

ll-

I=
a

'"

1-j-, I
W
iCOi CC' W L't
M^JJ J LL
tt; L) "-+

LLJ

i'-'

LL

v) jC,

J1= -j
I= L.t lJ
WW

W
ti

I-i
(-.

WW
;Z4 _

_l

i-J

-1
":

r'i

.I,..

'""

1r_1

LL W. =' i_} i } 0 i0. L'1 it: W CL c, c.


J 4=1 CL i. J I.i"I li..

I--

i-i

'=i

o_I

IiI!

Ci

(0

"W

. Lt CL

. "J1,

LLJ: "

It 0

.0

. yltil

4%j

N Cl CACAr--j
.. ...
0..

ti
--

r".
:, -. W Wcj
c;
Li-.1 u L1J cJ 1.3 WW
i c,
"".
""
^" i= LL' lt: lL

""w

. J 17,
4"
FF"FFC".
1,

- fl::

"'-'

I=

".,:, 0.0

".

Il I^
-'
(-ii

I-

F- I"- I-

W
n

F= =a

F=
rIII i I-=
r
_I
"-r JiiHJ

i'1 s" it %- "_. +-. "v ". rf..


0 Lt:, ! 1+ Cl I:r: y 11
W LL LL LL LL LL r^ r
i_J I:J
iZ
1: 4_._L'L'Lr r-+ I--. H n"' r-"

, --I
fri
i

i_t '" '"


-" " '- iy
%. IIII11iL`IIWUW(A

sln_.
_,

Iy
LI

-..

1_1I_

t-!

_. _.., _.. , - ra . -+

".
_i

t-".
^.
-6
ii't JW.
iyi I-i,
C. 0 '-- Ltl i'"_ "".
. T".
CL 5 F- .! t.; 1 '" L31 i"t
IJ "y 1=4 J 'T

U.

LL

t1J
a

ti"

_:.

111

! jl

"

H
I-t"'

"I

LL

i_i i_ j
r=tl

"

L;t: "LJ::: : rJ'1IJ 1


F3
..
r_r AC4F-F-wtiF-

Ui
44

4.4t

1 IJ

`='

. -,
44

0i

IyI
Iyt

ti

: ",
; it

Lt1 .L
CL """
-- W
I.._ :
C-1 F-.

F- -'

., i

F:

"%

JJ

-. L'L LL'

F-

D-- i_I

I_t

c;,
.
.
WW

F.1

;.
'-

f_i

-="

FW

I_

. .., .,
_

fi

_(

ii iW0_

r",J

-: - 'A (
_(
.. ,

ij (". +.

i1

: -if

{F- "" ii.


r_, !y F- :'d JJJ
iij
i. Ll.W Ll.
CC,

f:.

ID

f' 1
Ls.!""1
h+!
I

1--

't :J

11 C.
", IM
( .
l, " i "3" L_j"",
_L
1 i"
I.-. 1 1'".1 I'" ( C"i

-- ".i

LO

(")
71

i_-,
-+

t=l

1=1

T, ^i .rf
0i ',
i, 1i

.. r
11

. ,',!

'

IJ .lJ '
:.
. ':. 1 T' i -"

'"5 I' 1I "'I '1 i' i YI1Y

: 1'

470

'

_#-

"
CID

LL
"
a,

,
"

Lt1

111 ^
F-i 1_f
_

"
i I1
i

I-- ls

F-

yW

IJ

, "'

I--

LL
11
IJ

111
L

::

--i

i
--j
'
-'

+.
+

-r

CY: 1.1)
l'L"' 'J
t= t_t

F-

--

4-Ld

tu
'`

14
'"
-` Ld
".
LLJLt:

-"J'

1-, "

. IrI
t-.'-" 1=1

_"
l. L

# #

J:K

Iyl

ly

t!
_r

lu

#t

'# r1
"".: "".

L! '..

#_# t i.
LL '-"

IT"#
1'T"

1)

1_
.

-.

11 #

tI

.T

"': #

^#

""#

*7"

_i"

-r

'.

'!

F-_
; "' ~I
_. !
_.

rJ"J

+..

_..

F-.

r t'LLJ
l1J

i. i

Ir'I
IT1

'Z

+I

F--

FM

I" i

_}

iJ

F-"

-+1,1
Cu

!--4

in

_-q

ID

1y',1 ",. e

't
I, ",

L C;

.L

#-.. I

...

1
'"i.

iI

i'I

1 I [-

CL C,

F-r

,.

"J

'_' HC;
.

"
LLJ "1
t+

IZ

#"

''"., ht
"
I, J i. '!

.J

-5

'" # "!
.,

1,!::,

17D

I ".

#,:

1,.! .+.:. I
#" .t-,
.r

1". J

17:

. #

. +,

1'I

ID

#;

# 's

L:

' !

:II

*-"

""!

1 -1

17i

1=:

#!

11;:

'"

#7, # --a '. { #`- 'T .Jt.,


it
tl'1 i"
h#:!
-t t:
,

"

1r.

1 7_

r_"!

It

'#"

..,

1 t't
I. E. F1I

I'1

!
_1
LL

'.

^
+'

"

: ._.

f...

"

-+

I`,

."".

'
.

{.
'.

F-

"

L:

11
11
ii

1# I
''
i

_I

1
,
'y

.
T

1^,

"'1

'ii
! i,J
I; _I

F-

...

=F..

"
IT,

1 1-+
1 -" ."

f,

.!.

_
iJ

+wx".

L.

jy
LL '- t% 11.
i

,-

".

1, -.

.-"

"wr.,

I:Li

'..

='i

_.
r.

-"
r", J

Il !
"-".
_.

`'

-r- -

W"J
LLJ
171
1,1
F- F-- 1-"" i. 1"1 I- F- F_I #--"
F-- F-

1-'
-1

-
..
f_. # '

it.

CL

: Il

1--,

#11,
Li, J

'"""

i.

I
#_.

Va -s*.
.

lL'
F'

1".1.# 1... 1

1j

I
1.
'7'
_1: i '1
1_
#-#
i
ri
-. j1
FF'.

I L{ .-.,..
....
-7

#=I

'-'

LJ #-1
F- i! !

_-

F,..
'"

r4.t*t

=4
=

iJ

: "

' "
`-' _
1 #_J -" i

"J'_#

+_+

11

--#

++-

_:

"i

=-r

Lu 1' LL
'_
r

CD

"-"

i1I

" yl

+-+

., "

0..

Cl

.
F- +

''
Q) t.0 ll. l
t_i 1_:
J =

}--

F- +.
,y1=

: -".
Cl

F-

P-- #:

I-C 1y+
i-I

I. 'r1+.

#+

r"

tl

""e-#-.

.r1-#.
I
1y
LLI _.

LL

'J +-I
=0

tt: ,
"
1:.. -

#L

_J

ii

T-0

IT-

.
r-

t' +
.,, ".
LL: +

U.,

L1:

LtJ

I. il

'!
1 I.. #

^+'....

J L,

"t

, !

II

"+'.

17J

-4

'-

I.. I
r1'

I.. 1
.

"
_" I _I
",

lyl

1)

1"",

tj
"t+.!
1
1#1 iTy 1""",
'. I, I 1"". ("". T"
1 ! 11 .: ', '. 1'1':

1.. I

'.

"
"

-.

471

L_
_.

+
+
;#,
+

F-

CL

#
#

I
1-

+J
*
;,P,
-0-

1J
0

f.

l"

'-

F- .A.

0-1,

jiy

1"

't

rsj
,

-,

'..
T4

LL
t-

[J t1

CI

I=

{A

iIJ

-S

oi_i

"

CL

AH
f=tiif'"-W
Cr- 0 i"'.j i.t_

.. wC,
r, = w

, - iJ '""_
1=1 F- C
, jii",

=F-W1J_J-J t
. {A
I'D im t". j
F-- ' "..
F.
.4-.

1 - _
J C-.
1=1 I: -,i LL
Cl

', j-,.

t."t t=

n-"}

C,

Lt

Li1
J=
W F--

LSi
k1;

-*

W
A
0

i--+i-f-#
'tt

WW

= 4"
'_ '""WW=i.
#
F..tiJ5J0

"f

u
f-..
.

CL CL

I.J

+{'J

#rJ....

=i.

".

io

ID =F-- -I.
H

'"

CAj""_ Ci

:Z....
I.

F-1

-4#

L+IJIJt
= LL LL +
0
IJ

21=Lt1F-+
z
`

rJl 41--+ V+
174 1-- IJ +=

C.L'

ti
1
:
ti1
,
JWt="rJ
W
to
ff
JH
-"
fJ
= rr,
0.
-. C'J
F;
..
;
..
r:
f!
-.
W

fJJ CL
Z 0
T

=1
W

rrJ
CIO
rr
".
J W w

.
"
I..L'.
1=

F-'"Ia
= rr
1=i

"

I
11

1:1.

W
=-1-

=1:- I: i

11

LL

LX, +

La!
L2: 1.L
L11

F- W
`". 1=

F- i-

JJ

CL . -t1_
LU LJ s=
lk: - LL -+ Cl ==
#-

Li 1.1 =sra?1
111

.i

F[t:
F-

'+

4-4
(M

Z:
W

W
= lrJ

Co '_

Im _

J r"" =
{"J J _J U. CE: Cl
.Ji
=r o-"".'_
1""M
If_f-tt= t:"1
VLLW
.,
. -.
?7'F"-+ LL
".

- r f!

,1 t
LL

=I ii}CO
lL

-+'"". _ii

F-f

"
-ry

*-

_ {-- +. "
W

t
N. r.".1 _

rrs

._+0

i_l i.-r) CA
t:t:
.:
,
of co r,
0 U-LL

1-..

tt:

J;; .J
LO
. -'..CJ

LD
"-
,.,

Crt
O %, r.
:L,
CU C+J
+
r",
iJ.- iji.
rJJ
[t: l:.
s+'.
""
ChT-+J
t=
t= ri f f"' W=1
lJ! tL f- '"
J
-"+
J t:'J
i+J 0
ctWW
--'lnn
=-"fjJi'J-li
CL irWZrr0
:-W
:
+
CJJ
('i22+
""
F".
.
4
r.
4
, tr,
W
F-_
L"'rS'
WW
=
W0
i=+
H
1=1rI 1=4
F=H
i!
,
+2
Q.
+
`.
F- r_r +.
.
11
ti1.
W
ti
Or

1
.
H
.
C.'
.1
i_rj r_r
-"r
ff2ff#
F- 3 .= CZ '- -..
_,
N
_.
f
tr7ck:
v",
IzWWF-1E--ifiCrJ
-"v-tv+I
tr/
'`"". F-F-F--F-2i-I-+=J
WvF-lr1
11
=ti.:
IZ"i
F-Ar,
ct-+ 1- W^...
r= r= r= t_t r= r= +O
fJ
C+J
Ui
J
WW
1
1,171
LLJ
w
T2r'r=
CL
CL
WNC
=f_vvv
J
N

1-4
tt:
l"i
L2.
00
:
t:
rn
-+
-LLLLLL
r_r2LLl.
lt',
i_i=
-"
Cl rJ +J
LJ t_r
= tt: :L
1J.. Jf, J4-a.
-tJ
+Hb-4
r~
CA
74
1:
;n ti

tij"

":;.o

ti

+. 1":

"JJ

".!

F-
Wr_rr=r=.
".
"=
-e
__tJ".
:
S#
tC
V

rrJ
r=
..:
-2
F-1--"
W F- }- + {_ . IJ O_ $U
10 Q
f_I
FI-re JrJZ2-+i-I-+r,
w
-6
;i,,
F- UO
I: tz iCLW'_
W '_
+:
tUUi_1
2ii`%.
WW
(-f
F-}- -'-

'"

"

{S

"%

: %;

"1

{rJ

"11

LL LL LL '"
iif
.}; 1 ": %:

r W

rr

rrA

--r

U'i r.lr f'-

LL LL#"U1
=ic

': fj

+
r' j rr
r

in
-i t1
.r

rrr i +
r^r 8

ci
7 '7
"S'

Irl

II

Im

in

in

I-_I

11 II

Irl

lyl

tJ1

c
,
.

i
CI

'al
1..

G 11')
i ,ij"
. "-.. -

I:: 1 Irl

j
i1
1.
(""
l,
IT',
j
It
y
I.
tJ
'
I-'.
+-/
1, I(iJ
("` 111 I"1 lyl . -
Cr,"IT,
ccc.
i
co
1?
j
1
1
I:
iI1.1
.-1I1
IJ
Ij
l". t"". f""- I I"I
_

Cl

trj

^. I Im
1..

T. IT ITr ITl lit

Cl

IM
l.r/
-. 4 1"iJ 1"'1
0

l=-

4=1 Ia

IiI

1? 1 In t 1 II
IT. 1 . + ... ..

I,rJ
-..

1--I as
(JI l"L

IM
(`

IM
tn
i-. l lil

fJ

l=

1y1

..

Iy1 11 111 1-1 Ir1 1;. 1 . -. .


. ... . . . . ,., /
.

472

1:11
i: '.l

iJ
F-

IJ
i_}

,
-: -. S.

1-1

LLi
_

. }.

W
7-

y. ...

'

*
t

a
"-,
"..
H

ci:
J
LL

tt

iit
t; ",1

"
tit
ts. t
"0

t
+

f
W

-+J
t

""1

F. (I'
r_,

r.

t=titjt=t

--,

:.

F-,
U_r:
tiLL
--. r
-"1=1 JLY.
+.
r10 = --#. ILi
LL. %"" il _=' " 1I= -'WiJ
Jiri
--ti 0=-J
._I
'. / J.
J

" ,

jW
-.

. -4
h-; I'".

f_!

: 1-

N0Cl
yy
i_i
I= tT,. r.
J

..

rr

ljj
rti

""

IFi"
;

,., 1r

80
Cl f0
. r. . -% WI--

U'i

CO j
. - . -r am am -4O
17,1 CC' P- 17'1-u
(-i)
U-)
-4
Cl.. CL
"--1 i

w t--;; i "J i'"1 1'".


j Oj iJ 4J

W
."".
"
. IJIiIiIJ
i ". i:
's
}'.

WWU.

l".

\-..

++
W F-

i W
.v

""

F-1 ~-

W15a

F- f_t

. -s

~'

r"
W
=J!
" -+J

,
"-,
V
'J

11 -- _l CL 1- 11 0 `- _r
11 ol
il
LLl
U
W
O1
:
i_t'1S
_t11
'.
11
t-. il , t=
H W tCL LL_
Ni=;

=
W'
s
ss
..,
WF-C'lF-JJF-=
*-. 4 IL J
J r_Ct:
: W'=
`` J t=+ C J

I"L
U-11
r.D
C.

tY ' W - W
. 1-- r..

s_c s=i s_s Cl


'-.
.
-". 11
_. _
ty, 0 r-_: LL LL LL L.
r+ H r r. i 1.-. r-+ -+

J .. W
L
lX.

,=

ifssJ

CL.C=o 1J
11 11
W 1-

.v

W
-_
r '' I. L! LL r.
F-I'}
11

r_I

-W
WH

(,L

W
CL LL L--,
1J H 1.J

+1

T--+,
_",
"-r
r"rr:
r"1.; -rUi.
--. 1"'--r
r.,.
r. r. . _. -"...

'- '

O,

r"
?.
"J ', "-t",.
-4-

--"

, ; )

C7f 0
'.''l .+1
in

a;

. -.

C1

I I"": ,r.,.
. -d

w. +_.

In-

_- T
.

I{'}I,
_I

-Q

-4

.jl'_if
Ir/

1M

17'1

U-.. ty,. f".

': I II '
_1

1-- "
.`
F-- t
:=

{-1 Y
%J ,1
r{'

"7:

-.

L LL
"%

1:

:! "!
--

utua

I;, I'D t: J
# 6r/ rr t1 i+t

-r

_.. ...

h_J
J
r
=SJ

ti i VLL
V!

U
:J

I:i1 J
.^ J

FLL

.:
Lj.
I
4i 2
'_i h- - 1=

"5i

"1
{'1

tz
W u
1-.+

11

h
\"
i1

H+

++ t

,, . , f..
r"__.
Oo
.J r_t J U. I.1, W
LL. LL. 171
r_j I.; J rJ I:. j - r_) -+ - L.t
r_s
b- r-.

4717,
9"1

1.1)

...
.:.

IZJ

1- r i-- M- r,_

1""t
CA

LL

r.y

iJ +
lL
+ F-

-.
lG

"W
W "

+
IL

I- U_'ltl
U) wT
0 CL
rJ

w+-a
i-- +w w
i= + I_ H
F- -.: r
tJZa. [ r,J
+ f_
F- + :; L
+W

Ui--

FCL

F-}-1)
O.
-0--e
-Jlu; dw rn
i
-- W

UJ
CL

t7

CZ

Wi-j

CO

11
J
_.

i.e
0

W
WL
iai
LLJ -,

"ww
"
-j
c..1
--'

A
0

ow
-+ l.: r1
F- F
.
= Wr

LL

t_t
t_t
iy

""

0W
F-. =

LL
"-+

S. /1

..

1=0 F- L
CL JiW
IJLL
J J U

-4

.,

LK

-,

C)
P-1
F-

Ll. ) 4-i

..-

!'J

-4

(j)
W

r_t ia
CL CA
t=t
ytJ
" F"4:-4

C4

-\

Cri

U-

r-

't

--

tJ

4r
.(t

": ' '.


t. Lt f'

"1 -T
-" T -t
-_.
...
...

473

Fr)
Co
e
w

u
t. L..
LLt
1
...,

1J

F-

CL

-j
v
Li
1-

J
=.J

i
F-

.0

L=i

(J
-..
~J

li

ttil

T1

LLI i_ U
i_ I=

F-

,=-v
d-

r _}
.., ^

-. i'1

"t
a"

. LL, r-1-4
CL
Z
!y ft..

" W
11 J
1

..

t-" Ii
1
Ffjj!
'ti
CL
1_ L, i11yt
Ci in = I'"J
_ ",1 W

i_ W CL
-- H Ci
Fi

i--

AL

W L.t.
c:r ij
Fl_ c'
4=1 FLiJ LL
-+ 'F- F-iCL.
ir: CL
+WCc

-:

'_
tJ tC
F-- - it:
i_ W '=
-,-.
ti.
-!1? -"W

LL
:

-"
Im

t.3

1.71)

-o,

ti

CA
*t

.._.
-"

2J

. .
-

1' -

r.

---

v -. j
i--' .
e= A

ff)

1=)

r.

-4-

-"

1_j.

-, r,
=y

/1

o-y
."

ce ..

I: D

CA tr J",
p7 -"" _r p

vT".

-r ffj

:.
. . "
-. n.
W"
O
fJ 1=0
iy = U1 '
W=
'.LH
=)
rty
r"-- .`
.
.
==. -=W
-.
I1 H '-'l .
In
J
"^
U-) 1:3
fJ = U7 r". U"! w
ryl
-q J-Jr
-0
C+l
CO)W -NCo
F-- -. *
:"i J 11 J"
--*
-- U7
'L" :=
-6 ii:
11
al
'"ri
i*. l
-.
Ct',
% .4 It 11 =.
""
'-'
1= V O'. JD C.l
LL
L
+'"J:
I.
J
CA
C 1'71 CA
WJ'^-eU
-"' fJl --4 "'"'_
17'1""" L.2: .0=.%-*
.=.
11 v 11 -+ 'J 11 '1 11 F-" ,t- as '.. e
W " W
(D
i Q.
2 '-+ W
==N
WJ1-E=1F-JWt=tWJF-W1-1WJUiA
EE rii
= 12
I.- I_1-. I=-, JiL'=
": '. = F-.:.
0W
LL t a CL` O
LL` W !:L 2V
-y --' '-0 O 'J -. - (j'1
=W
J... 1JAi.
iY- -2' r) wLL J_i1'wLL
C1-+Q
: f.'j
-+iC=
_
--

lJ)

. -"`
:L U')
: l7 i
=' l.
1--;pU7
C. .
11 1"0
CC v
i- LU

FW

Ui o:
: A:.?.

":

-' r"'
1^, 1

1rJ

"C'

07-1

11 Iri
IM
..
1It

- ru

r+

U+N

11
t. '"!

lyl

Irl

t'-rl
{tJ

trj

1:'"1

1.''J

lyl

IrJ 1r1

1 I, lj

U-.

1-11 'r

r..

1J
'L'
UI

1ri

_..

uI o; U-IU- o-I
r
.
-,

Ir

{''1

1r1
1T'1 1=1 "'t
U-1 1.
1.

trt

IrJ

1m

0,

i'"J

lrl

k11
-T U1
1i1

1.

ti1

L1 t1 L i,
--". -.

Ir1

IrI
1i1

at
U -j

Irf

C''1

U)

IJ 1

IrJ

i'1

$'I)
I.

i1

Irl
'-To
1

1M
Ir<

.. +

1
1
i
1'",1 I: I'7'
.

11

Ui

1^J

U r1

Irt
If1IJt;

trl

IrJ

t ; tt r"_r r ryr". r". r". r. r..




.

474

*
*

*
*
*

*
-

-.

*11:

t-

".

W
I
t-

IJ
t_I

LtJ

W
=S
F-

'd! '

_I-

.c_
1=1
z7.

"
"

"
"
"
=+ '_

-.

-- Z
"?I:' F4-

-d!- u_

W
i-

's I_'
L)

J)

lJ

--

IZ
LL1 I--1
_1 T
CL CL
i
i.

",

#"
; "

-. #
'..CI -9k"
r-i

ti :H

I-- I-'yr"Cc
: I'LI
34 Lt:
"'p" hI ' ";

ilk.

#:

.r '

"#. la,l
'. fl
.."
HjF" 1, '1
I)
.. r.
.1
fj
!JjI:,

iri

1111

1~I
. ,

...

ii

ii~i

F-

Ll.. "

+". Ltl_i
F-t t-."i.. -"i
'-5CL1-

Q
i
tJ'

-. i= +=
W
_S
FW
+=
-_-.

QW
*-+
-. F=; '"1
+J F-

I I"1
... f

1'

1- _ _- W

1"I
I

ww
iH^'_tffjrJA

-+ '- f^f
F- }- f L: '-"

LI

IJ

HI

1-1

i-

c", J . -4

a, 1=i
. 1J

_1
t=i

C, cl
i ;, i_t

i"M.11*-+

E--

.-". .-.
'J *-+

aa
WW

CIi L:J

-, - p..
-.. .`- co
-W W-?,
T

cl T
i_l

Oil:

. r".
LL W r"-

f--
lL f-f J

I::.

lt:.
i= W

LL 03

ry

:_ii
- _,
h'-'-

+ W
-f C. ff ; '
fif t=
-_ _ i_f f=. f =t 0W

tl.,

.I

17"f

-I "".: I
il
,
_I
IyJ
1--1 a
1='

.=Z:

,_I

I1. LL

LI r-+ F"-. LL

-., li

. ^I ITy I=I

V',

1.._ 11 CT, I"".i

F-

t.! F..

I.j; l
r.
"-" 1't ... 0

i i!

LL Hi".
i_!

W I.- IT",

I. J

-I-

.' r". 1
IT"I
C1 Cl

yi

Iit

Lj

IM

+-fi iJ~_ r"


"{ i

,r

.:

tJ_II.

I_ f l
,:

.II

iyI

'

"ii

+rl

I^I

1I

+1+
i, j ! '".
'

iJ t- r-" Cr- LL '"". *-+ }--'_

al
1.. 1'"'

iJ

J'"",

f=f1J1-r_t

LL. U. f1:
: -

i1
IT,

r. "_
i :l'.
1J~1

I'.
1 : _!

F--

LL. '"

ll.

FIT'

rT L
r= t'
:_". 'f._=i"
ICFI.
h
L.:: 1_! L
1.= I. _ t211 It 1=1 !. L CzI 1_I
I-'_'

t)

LO U C). '"
iJ i} i
JW',
L"G ..-,,
r"
7-- CY: LL) L0
_'

1:'

'"rL-WW'.

IPS

WJ
.
--''_+
JJ'=
W i-e r". +_++=

I-Cl "!}; "


.
#.
.
1_'4

_.

i7

f=i

'-.CI

L4

""-

-. : b:"

4"=4
WCl

"

IW-4.

i--

fi+CL

F- }-a
I_t "`
!!! : -: i: "

iis_T_
:+F-"-"
'" "i:
" I--= F- F- Ct F`' =d"
LLJ
b
L'' =:
L1
FF: .,
".
"
.
.
W

.,
.,
:.
""J
="'
'. "
.

tt

F- A
+_+
F- +_i

t--1

I--

F--

=I
L.L. F-

'-'
CL
1D rl
Fj-1 '-'

i_

WW
wZg.
F- LL'

=i
cI

r_"

F-

- I'y!

iJ
W!

6.

W
1-#
I`I
lt
Ii : '".

.:a H i_i t

LWQ:

it.

~J

."

"ti F-'
CI

cr.: it
W0I.

II

L: J

--4

Hrl
W--

1=

I_I
i-

fl:

Ch: F-" "


" 1JJ J --

C!
uf

Z+W
W

F-'r
. -_
fl.
1_J

t-
-:.

1.0 All. H
_- +Mr
, r: FI
WAr

F-*=}1
4

ti

LtJ
w

iW
Fw
; ;r=
#iL
,-"

Fr=A.

Ct
(L

W }= cc .
'"1
-H1,

i11
LL

I_I

t_t

Lt:
E-+ LL
Lti

*
W
:.
:. _, I

L.

F-

IJ
CL
LI

1.

iJ

1,1 1 III?
-.. 1 I

1yi

,yI

T'

, "".
1

1M, 1_I
.rY1.1

IhI
1

.. 1

-j

,t.

-l
I

i 1 -"
IT.,

...

-1

-4

+_i
,yi

..,

...

, yi
7/.

1 -. i

11
j1

C z; 17J t =l
f'"-

i I

IS,

171

r"1

1 "1

..

On. ; 7:. I-, j 17,,i 1=1


:1
1

r
-1
.

1_
I1
11i1

i'1
:;;

IM

I'D
1 I.

rr ' J: Pt

1 la

1 1

1i

I.

LiJ

I ". i I'". i

ll
I"1II

'"'

i, y1

i:: 1

.. d
i

Ii1
fI

fy, i
i"

i"f

1"1 i

475

'it
i
It
H
F-i

C:LtJ
L
J
I.,

lti

1;ri

+-1

I. JI.

,_l
rr
CI
F-

"+1

'DC)
F- !11
1V

...
s1 WW
i
L
i

11
J

1
J

11
f_1

1{

"'".
11
1ji
1IJ

.^.
11
I. i)
1.,L1

"1-

! 7'
1,'

Irl
' y.

,r.
I+j'1
11
1}i
1;"j
t. "

..
1J W

U'1

'"-

Hli
i_t '
IT',

t t

1'11

co
Ut f,'! C-t
! '- ;ft: 1 '"

LL
-+-+ LL J
; -,

,Ti;
Lt

p
J
jl
. -

: -".
b-1
T.,
LO
.L C.)
0
'"t
'.. U"a 11 (' 1
Cr., '.Y"

ir'"

')
17.
Ci''1

!) t-i +_t
NL'J_lA

ti)
.
LJ
+
-.

i iJ _

"

tT"t

t_t i't ""f t_t

11

;-" ;_. (. H

+_; LL

1-: 1

U-)
1=1 1=1 1=1 r=: 1=1
LO I". r'"1_r ri,
r.. l --+ ---r r"I
. -,

''II""Ji.

(I-, (i"

1i

_;

+-" LL t_;

LL J

f^i

I=I 1=1 IM Iy1 Iy1


-. j '''+ "1"
r--!
f+: ''J I,
}"- ;I
"f r'1(

ii,;:

-fi.

i=l
fJ'I
j
1;4

s..,

. r:

. /'.
11

U'j 0 Lt:
'T I+I i- )

Lt:

ro-i

H
i_I "_ I1I, .j i_ .o
tj ij F"' i-J ir"il
4 WJAWJI-.., %. H
-..
1
+:.

CI

F--

V..

F1H1I

fJ
i"'1
-r

WW

e.

;r

1-

i_II_1
Ii

'.t-j

LU iJ

-i JW

, -,

IT

--

11

LL

"

ti

1.J
Irl

I1

i:I-t:

"'J
I.

--

j13

C.I

'+ I i!
17) '"
-+ =1 iJ '_II
-+
AWJ
-L=1
''-"'.

+.
M-A =

i_, -+ Lt_ JW

I^I

LL

Ir

i_I I-. ;.1. J'slj

!j
h

ly, '1 1"', 11 I


I 1;
Irj
I1
11
Ijl
I
1,1 1
`I
y1 Iyi
I"JJ '".. 1I"1 ITy {^I
U: I
't'
, --r I'". J I''!
': j I"^.j L' :I 1"'r 1:' "1 1 'I 1'1 I''I i; ! I. ''. f i: ''i
1;'" (1jI;

1i''11''II'JI`''ii''fI'''II''"Ji''(I'ii.

" . +

U"o
l

,;.:
1r1
I. i.

+,'. I r
-r

fi". I1''. (i". ji". f'': (I"': fI'eIII'"I';

"

II
. .

476

',

_..

'..

r'"-

'jj

171

'J

F-

r,
E-,_,
-

I_,
'_,

Cl
,f,

cl
,;,

wwwwwww
iyt
- {- { "_

i i i"1

--yAW_!
.. i..

..
'- L
1-(it

'Ty

. -,

.. -1

i--

I--

0
l'",
ti

IM,
a

"+-".
ID

; -.

i.: '
,.i,
trJ

'=,
'f"'
': j

.....

^""

"-"

''

"

{", 1
j""_
Co
(r
1r)
IT,
't
. kLi
. '1"
, -4
" i=j
f'- ii CtJ
li 1 I` "1!
i: +; ii ll= :'"
js i"; J ti+
lk: ii
:'"1 '-
F- .,j "11i
i" . --" i^i Fr . - IM .4;:
CI 1- +. + -T, *-+ C i
'71
I--J
1.
r
1'",
,
-"e
_
,J

i ('J 1-)'% I"L C,1 (J


I.',J i_i , -"4 111.":j i_i".
i, Jit".
fri II :,J
.tIIt1j!
1- IJy
I"- 17"1*"^
t--" ".ry , -+
F- CT.,-"
I- 17"1'.",1
F- 17'1.
--:
i}
j'"..,
io-a
11
11
IIiiiij11
ij.,
if
ii1.0',
+-4 i=t
1i
+-+ i_t o) "'" 11 F."'t
Hci
F-+
r"
h"-_:: LU_.. 11-AWJI-GttlJ1-:::
'NWJI-:
4
-.

:J '-+ I-, JW

'"1 +... H'


=' .. +.. `
i_i LL -" !_i LL

i_, -

Ui
,...
IM IM
I1 'L1
-f'

,y,
i'J
,: j
':. J

,yz

ci

'''rj

'": J
1,

lrj

';
17-I lyi
In
U-1 I. LI r-

lyl
ii1

LU '_J -+ LL JW
-1

L.

i_I -+ LL J

Iji
kiJ

, Ii
I=I
Irl
In
I'y 11 .. r I: J I: 'J

."+.

r"_
I=I
'r"

1=1 i'm
U. I --.3

I=1
('".

-i
i-

1_"aWJF-`".
1=4 WJ1^"":' +.. i.
'"_' r:. i--
''
..
"+_ .T
LL
LL r -+ i=i LL + "-
#_I
,,
_
L1.

LL
WI-J
L111::I 1""-+
"J

i' ' Ij} IJ :I Ur' U-1 U""1 U-1 Lil U 1J"'I 6-. 1 !
'j" 'r
-t
'j- .t
"jj
"11!
'"Ii.
Ii",
ji""j1'".
j!
Ii,
j1`.
li".
I1,
Ji",
I{",
Ii,
'"")1'j
I1"'":
i, Ji""II.

LU CiH

LL JW

'Ti

i:

: -j !,'Z ':. -i I.jl


I:: t i=i
!m
' / ... r I. ",J 1''1
I i; l IT"1 I,..

F-+

!rJ

I=
.I

I: j
(".
-

Iwt

I=,

I1

IT',

11

i I I, "I I
Il
i". J i'. I ?U:!
+-+

IM

i"I ti) 1.1--l1"LI '" LI '.il


fl !. G LLj ("... j_ ('"- ('". j ". I'".
yI'"111', Ii. If. 1i", I! '",
I1'. II`"ji.
{
Ii, )1'"")! ')i.

! J
I"LI

!:. I
; ""

('

(.

477

y
.

I '1

. -,

-1

l; i

, -+

ii
t:'-J
Ci
F-

if

G-1

1_il
ciJ

'*

I!

cl
r--

i--

F-

Ci
F-

i_[

r
rt

1_t
i_t

+=1
t, l

t_t
+_t

Iyl
Cl

11

It
Cl

r"_
-i

,y,

+-i

r.

J
_'

C'd

!'"1
..
WW

''d

-. 1.11
.
Lt i: ': t

,-.

Cr.'

lil

Li)
l'

o!
,
WW

WW
:

r.

....
1,' "1
U-1

U't
. r.,

1:1.

Ci

F--

*.j-

p._
.

. ? ",
al
111
1-, lyi
IT.
*:
1
1
1
ck
Lt
t:: N
, j)
C9 1.: 1
1 1D H
1A
U-1
LI
(',
Cl
1-.
1
i',
1D
i;
FIi'".
I
1;
)
'",
I. ij
F-- .'"4 IT"1 i; ". 1^I
CA
I
I
i'(
r"t1
IIt
iiII
i"1
i't
"".
U+.
t
!i1:
1,
1_i
i_i "". t:'"1
,1
tI11
,ii
'"1
"
,
f- Ir"1 (._t .:
=
1- 1?'+ 1;,I1IT-, 1:",1 r-F- Cr ;!
-.
a11
CQ1..
ht
il
ist"(I

+1ii!
-
ii;
1"-4
tt
if t-+ 11
1
t"
C') ""_ i1 t--t 1_t
W_1
FJ! - t= t11 J
"-"
Lt1
z
{i-j
i-1
1=4
-
-4
iT
...:
--.
.
LL_I
_"
.+
LL I-r
!
C'
L.
t-+
.

t(
LL
1
_{
1
_''.:.
1'-H
L.
'
t--1
i
1-"+
r"t
F-f
^(

.i::
1 LL",
U.
J
F-1
_1
_L -.
!! _ _! '. in
Ir,

I-"
-4

-i
r4

ryl
4: 0

r'"_

, -f-1

iy:
IT,

"

_
,;
1 ; i"ri

t`.
Ir

im

Ir

hI

rI

1, 'I 1' 1

iiiiiiit:
! I. {i
:i

C'
"

i ., i

ri

=, 1M Iyl
U-1 LJ_. r"

1: 7! i
i

I,rl

'! iIt

Crl

f^y1 r; v
1D

171 1,71 lt
t
--I

1_i"1lji ". ;i.!


ti. 1i, ":

1
V.

i;
I. LI

lyi
r".

hi
C!
'='
1M
!, 1; a11. "" *""g
_

_'I Ti: 1T"! , T";


1_I1i;
.
j
':: i"1'.
ii,!!

r""_

v
..

. t'".J

1=1

t+tF-- t: ! t:' J t;i iry


i""- t.! tTt t; j t"t

i_t '" 't II: .i i5.


,Lt II, J _;i
1-- CT*1t'+
t't ". " 11 0-4 1-'1'. ""1 fI -, i-a ,!
W
TWJF-;
74
A.
LL -+ =
lL F-. -+ -= t=, U.
J Li ii-, r-.
F+Lt- .JWi:;
t
..
1:i'1
''j
,

-4

1i
rt

_
}--

t--

Ch

y. In
,
i",
j i. '

1". i I
i':
!

i,'"i
1i
t

r'"'1 I=1 1^I


jJ i I. LI r"_

_ I.;.
r!. 1 ha-i";
}.; i'ii"I
1t:11,

I=:
I 11

1=1
IT,

"

478

T
11

ly
t
"!
i

1-""
l i, l

iJ
I.
-:

1-i
H
iy

T
V

ti

'1""
!

I; 1
-..

e-a

--

f---

tL

r r,

,
..

ii-:;

Tt

1`1

I
F-

T. 1

t.
.,

ki:,

I-L)

CI
I-

cl

C,

I_J

fJ

. ry

11

I1'

"i

:-

!11

.1

Cl"CA
I. A

_'
Ct 1:
J L-1
__:

W
P4

P-}

Cu-Wo
:

L'!
UJ 1'
,
"
It -
WJHW

::. {+} i+f h- "r -+ "i rJ Fi_j ". "i l i" j i_


Hij
F-" iT} -
C'i=
.j "_+ _ iI }t _ 'iJ
!:.f
AW
Jh:
:Z=
". :=
-'A"..
.
o
LL

i_i
- Hi N--iL _JU
Ww

t:

r-+ J

" .: " =

`t

'T
_' l'J "L7LI J
.r. LY:
j F- "i'
't
-'=1_ - iIf
-- ?'i '
iii "_ 1 -r
4-E

WWTi

-..
.`i

--4

f-ia

=t ".:

'
c
0

"'1

-'
In'.

^4

1 J
..

I"1

i'

tt

-T
.

Ii

1'"1
J'} 1"CI (-

-i ('"1

: _". : =.

U7 0
-+
'"r"
'y.

4' 1=1

i""-

F- I-t
-i

" :
w
in

cl

""

'"t

.
1-) Wt''J

i: 1 WW

T
".
"=.,
_
LL H
1-1 =

--J

1--1LL: F"-+01

-3t,

I=i

r-e

1 j-4

1=1

a-.

r.-

I-,j

UJ

iJ

-i

1=-c-

11
D

tl

11

t i

"_ . 11 -

4 , ra -4
..
'r
1'"!
'" Ii
;

Z: C, -+

CL
Wu
W
l-Wto
= ,yF-.
: 11-1A1-1!
U'T-F"-h1--f"-'.
ijF-""+hi-=
_
J!
"..
'-".:
F..

HH
Ir-'-'.
V
"0W
lti` IX 1-1 G i_i I.D [:. LL L1: ! _, LL LlLL

It

1I

1_1
(r,

=1

1=1

L;i

W
=u

"La
to
(".J :'J V. i ' il

Ci
W

_
I-1-

J"W-.
, , -, ., W11 -,
CL
-.
=
,
_1
_1
f! (i , = T k.0 k.u - ti J =t il _I. yt p
-.
ITl
F",yH
=W
:A 7="
IFF%.
...- : . -1.
,J

LJ i*3 -+ LL
-W

! T'1

""-r '
I-

+= W

CA
CA

.=+ 'i

-. t "_
l IM

-"
_j
'"

'"

ii",
im
i-I, --.

IJ ""ti
fJ b..
. _,. p..+

iy

wu
- -..
"

1-)

-. Hr-.

SS

+?y m I:k: 1=,1

Z=
f
LL- ;

in

Li
LC'

1=

-4

i, " i

lyl

ly, i

}'j y
I, IJ

Iml

T
W

ijl

I^y1

II

1-1

L)
i--

_
Cl
t. J

"

"

trJ

IM

-r .
.j . 1 f-

.-{.
11"I I''

I'D

t : -" I i

j
t' i

1,1

Ij'e

0:

1 I wl

Lj; t (".

Isl

"

i t"I

}::

I I7j

lyl

_1
""
1 1; -1 '. 3 I "! ( '"J 1,
I' itr:
I+ !!

17D

.I

_". + 1 rl'

1''11:

' : 1 1 I 1 1"'t i '? I'

I,

1-i1

1=1

'T

il

IM

I IJJ

I y1

1=1

1=1

('"-

1i

rTy

tJ
I; J

1=
.r"!

i rI ;.: t" I' 1?i 1,i1, ri I""I 1""'1 -r ,11'I


i
I'"1
t'

f'i

t" i 1"''1 1""t

i"' o r""

J'1
1.' 1Y

1 "" r"

-7" .

"'

1" 1{ "I

:
I-! "I 1 t. 1 (

t_t.
1:

I'"1

...
1'"1

479

-;
tH.
(!-j in W

Vs C-s it- I-!

in

1-

t_1
(LI

.
1,J
V'1
.,.. 1

11

11

.r

..

-"

Cl

WLi1
"
1_,1 .J t=S

111

*-,
-: :
- -
F-; ".
-.

-'

''"'

.,_

:
;=
1_i
-W
t_,

LL
0

t_J } -; ""

..
i-1 i
T

.
i=ii

Li
i

? f'

t.Lt

1yt

1j1

11

''J 1:V 1'"Jt_;


tU
"""
t="i'S t; . t_t

I
-j
1-

ti!

iii

ff tl.

i
.,

LLJ

ti

tUJ
"

t-,
i"'l

I.I
*"
'"'
Iyl

cl

C,

"W "I:

"

"Lt=
II

V-I

~,

" CI
{'"j "
." II I. y
C.
i, L! C.
C.
O1 C-i 1r1 f'. j
I Cl
r1
.
1 'i yj
""jjJ

'. -r

"0

-i
"1
y"
".
I-

II

14

W""

"
L

i_I {11 '. L3 . . IZI

W
-It

.J

I
11 II

ttJ

Iy1

-". I=I tyj I--:.


:,

I ;"..
I'i

1:: 1 Il
II
it
1:
1
1
"F-"
.
.
l7I
0: ' Lt;
1:J
U-1 " U71
W CL I:X: u-,, p"- "
i... "
""J _-"I F, _:
1'"! iil
{1
1:1
FI-''r ++
I"`I
.
1.
1.
11
11
II
il
lr 1::.1
Lt1 1'"J W --" LU "-
17', it --+ LJ
F-+
WW
1=1 F-'"_+
1F-- I=t "-^ ttj
F-"i
11
FF1=
i
FH
t::
i
-1
CI
CI
Q) I7-I I1 :-.. 0
i.
:
WI -'
t, '"j
1=1
1.1 II
fl
Cl
111-1
""j
LL.
en
t
t
.
-^
:t"
C) z
r-+ I'.,

lL

i4

.:
LL; LL _ +LL. iJL tL
I_,I t--.

! '

If'
C)

yti
.,

..

"

I.J.. Ll. IJ.! +-L C, LL IU LL ILL I_' 1'L 1=1 il


-4

uu

F. +
"-i

::
tit

I-a

LU IT', Cl L1J
4=M r+ i-t ::
tl 1;'. 1 1L 1=1

"

11
""
(. J +yi 1r1

Lti Ltj I
.'
"'-

.
Il

l: W
" 1..
1=1 . "

" iyt
.
t --. J

1:

-. I,_I -, "--
-.
-..
-.
-".
~y
"ti: I"-"

Lt.l
"--r I.:
i "t ri l''! 1'i+ +.''J 1--1
C: 'I

--+ *i'

t:'! -1- i "j

1:'
11 i'y:

.+
tit

1yt t1"t
--r

1=

+..

"1`'i

y.

.Lf

tZtrtsiit
..
tit
it tit
ID
s-1 ft

_ "":.

HH
1=1
1...

i_

,r. "
11171
U -j 1-: 1 -4
1".j "r.. IT,
! "r" IJI lyl

"

CL CL 1_1 47"1

r,'4 '

'(_

!-_j1=;
t.U Cl
Ia.J

-i 1- i '": t+a
1_1 W '"f
I_i 11 H , -!

al

:
t
Fii

1 "
" I'.J
" 1. t-L.t
-1

'1

W""

"

*'

W! t1 "0
_' !t
_1 _J Ct!

"'

. tom

rl

1 '1 :y'.

""1,

i_t t_,t i-j !:=4


LL. Li.. CL i=l
tt

-r

i'"1

""

_1

iw l

e"1

. -r

-1
I-Vj
Ij!

Li1

"
it

!
i. t:
--,
!

IT,

1 '1

H
...:

IJs :JH

U-)

s_s t_s ._f


C, 4- t-- ILL,

h,

:-;
,
""
Lt.l

'
tt

ii
_-,

1:
s',!

IJ. t(I}DaIJ.

C:

t , '! t'. ", }: t t', "s t', "t t-tt"11

1--+ I-"
J

C-1 1: 1 'J
_,

4: " st
siiiit

s""'_,,
t's
s ',!

1UJ[ts.

"t t ^: ,11i

"t t. ,": i:

l. L.
L. r

U) U'1
i-ii=1
"r

rf,

"i t,.

1 71 -"-

-I

ss t; ys ts :T;,: : ',
''': ' #' Is 1 `"I . ("

t11}1!

" [_I
i

1=1 CL 'Z
1: I i--. =

ts : y,
,,,...

t:
I! ',

tyt
'. [:

s! i, L. "
, t..
, '. i'

sr"i
f"..

t.Lt.

:. " 1'" ; t' i" f

-r

-*

sI ts
'i itf
.

:
'a

s,ti

.", i""'i

'

1...

. .tr!:,

s s"t
_j_ , j..

f_i

f".

6",

1:,:

ts

sys

'

.
"'. ; ,. t :

480

.4
rt

47't

7'

:
t

.
.:

>

-:

:
:

-71:
+
:

rt.

111 "'+":
f

L_ ',

-d.

CL
'r

-q".

}-:

'"

t
. ..
F--i
IL

ly

F-

'tb'
": '

"

w
fl

1--i

iir"
!- -

!;
-.

t-

t L' "-

:.::
-.rJ.

IT

t-
=--i
r.

1=
4i

Li_

"
4:.

"
" 1

i1;-'t.. T=
-+ j- Cl 1D.'"
-l .. :r t
1_I
17,
t'; t
-+i:.!
LL
I.il 1=t !.L

LL L

+
+

:.

:t

ti_
_

't

"

M _ i
. -t"ii

f^

(j

!.
I-r,

1 I.

LJ
:
L"#
L

i_c 1i i_7 !_} i


Lt. Lj- 4_

111
.

ii

_".
1:

Cl

I_i

l_

=i:

L*
--.'6"
L:

r.. :"I

r,

i_.'t

_ 17-

4T
-LIT:.

:
,
'-

4" ,,
-.ip-J. _-_4
:r t_.t t_:1

i--:
w LL
C,.

LL

_L. L'

L2 LL

rr

"--.
W4
: " Lt:
1_I
l_
-" ..

1--- 1_ 1e
12:
LL
LJ 1_,
.

;__

i:;

_
__.

-"

yr"

".: .. ": "._. "..


F-- F- i-- !- f-- =
ly t,
}y=

-ic- 01 CA

!! __

_.
"_.

w!

-"

,`

! Z: T
r"' : -". 1;_t
= !i!
"

ja i --+ W
; "j-F... .
-'
-+'.
4+s :::
t,

IT' "_

1--: I i. L.
1.'77 ly

Li:.

t -L L::
1.511.
ti[

ll_
F-i

'"' 'f=' 2
-,'
-lyi

I.
tIt

Let

'_'
i-L!

IZ FLi_ t.t =`
+-. L. ..-

Fw
E-

Ct::

!y

-U
*
:,,:

ft

Ij:
:

1'

hh
,<- .r

,_, 0
LL
.

+
*

:..

.4
-fi

.4

11 ^

*
-a---r -+ - -r.
--

f::,
Li i

i.

= t-

. .{..

1=

-C-

-
*

Q-

ii

= y...

4-

i: '

F- .:4:
t-

'r

.t.
!a..

--
, s; .-

'"

ii1

1. ".

"if':

1I 7 ii

, ._-'.

tv
( "_ t_.
, ii f'f

481

i
+
++

++
+

+
="-

.r.

.9-

CA -+
.+
tfl

LL

". ;-++;
U, i

ICI

CL

. -, tl -+

+_s

("-v,
Lt-J+

+_1

.
1-N

ii'1

tti+
!-

lt:
tt

11

It

II

[i

Li.i
F-1=

t_i
: t.
'11
r

"
,r. "
iyi

-"

ii

Wrii
i

.. (..

Ft! iii11I=
u LL
! _!
LuF-
5-si: S
"ri
Lr

f-1

ti

.-., . -
J jl cl (_I
r ;i II F-

ii. Ci
IZPiJW

t-

y-=

. `...

j. 1(

-i'U'II.
.t

rt.

tf...

-+ LJ i~. 't=t

I,wl

1::,

r+.

IT,

r" r. r .
'}
r!

t"y'r r'

:J:

--t(.
::.

+"

. ...

-G...

"-.

-4.;.
1, -T"M-

1-+

-. I_1_1 +

-e- X. !- b_. +

-" N
P'! "
0

r'

off r.

"r

ZO;0-4 It
+
.. _1 _J -"

.+- " U'i

-M
f
: .. ",, r "v. `-'

= u0_91...
-... FCl- = Lt Lt1
li t;=t _7 t.
t:

+L
+

r. i

4:1
rl+
. -.

ir

-i-

lsj

-.
-r"

. 1,.

II

G, +I,
-JF--r
+ LL
lri

w"
rCt

C.'

"'!

1'-_i

." i ". " -+ r'd


jF--t
" rI I-. _t . -.
"-i".
'f' Cl -t JII.;
- I-- W -. Of CL CL .
11 -- =I '"fj -- I=. 11 %A)
i r- u n.=t
Pi
"
tiLWr..
F- IT,
131-....
-f=
IT1
i
{'

Lij
tJ:

-I;. f

%=
"+.
l rt Lr -+-

a':.. f':
Lt: +iI1

WW"',.

". r
!i
+
+
if, if- !_ r- L' { !
F- W
AWJ=....
LL
1:4- = -' F- h- FwiijF= -+
LL1
r-4
1_
-Cl: 0
U
`, c-5 o-=
LL 0-4 -r
--t ,_l L' J
.T
=. 1J
_
-+
1
-+
i; i i _ i_t ... _, L.LJWJ..
iic:
'1,1

IT 11y.

CC) W

t= lL -

r-r
"
tilt V lt:
'"".,

-. 04 iL 8
w --:- E- r-;
. r.. lIJ

--

i4
F- ++

"W
^

LsJ

I^1
1?

k go {_"2

" '. ' -90.11 -. -4r.. f ". j

0 D1:
W
LL
L W

,. +

"'h

-4-lb.f.

`-- v)
u ZZ.
f--C N
C+_
L
=J J

Ci
_
L

J--.:

=Z
l. U
F-

L:.1

". _{

..

[i
1-

1oi
111'+

..

t
W#

W
"_i

#
i
ef
t
*
i. :.

rr

r. a: r:"r

r. r": .. r V"r L} Iirr


. . - +.,..
. r.
'? .t
.t
.r!
.t

.
"f

tL ! -a
&'i
LL LL Lt. -4- l.
,

-W

,-,
-:.
,',
-- ..
r

r"

,=-, 1-1,
+1 .
...4
..
. "- :
r

:"

rT " rr "- {. r, , .r
... t "t 1 .; "". t r" J!
+? +t t
.r

tr
.t

482

APPENDIXIV

DESCRIPTIONSOF VARIABLES

483

A. 4.0

Introduction
purposes design variables

For description

have been placed into

four main sections:

A. 4.1

Variable Description

Notes for Individual

I'll This section contains for all

list

Cost Centres

cost centres: '

(i)

an individual

(ii)

the order in which values for predictor

of predictor

variables,
variables

should be entered into the computer program and,


the equation codes for each estimating

(iii)

for specific

In addition
variables

A. 4.2

Those variable
A. 4.4

CommonVariables

alphabetical

costs in more

used to estimate manufacturing


In this

than one cost centre.

section variables

are listed

order.

Applicable

Variables
Variables

cost

section.

of

not present can be found in Sections A. 4.2 -

Variables

A. 4.3

cost centres descriptions

can also be found in this

descriptions

equation.

centres.

that

are specifically

Variables

centre

number.

A. 4.4

Illustrations

to Single Cost Centres

are listed

related
in their

of Welding Positions

to individual
particular

cost

in

484

lf

A. 4.1

Variable

Description

Cost Centres

Notes for

Individual

485

C>

C)

C.:

C)

C)

L.

C)

J
O
S

-2

la.

cn

C
W
_

C]

lL
U

CL
-D
t7
cc
O

li..

r
,
CV

Q
1-

lt

ft:

C.

M
ly
:[

Cl)

N
l. 1

t. .

rJ
n

CJ

1-

4J

oW
UU

..

r_S

J
O

tat

Vf

L7

G
r"

tj

lJ

~L-

C.

--

CL

C..i

G4

RA
W

U
4

W
J

b7

1-

F-r'
G
W
J

i. w
J

1=
I.-

-yW

tJ
J

. -.

"

^
.
t .r

N
S
l

^I
II

'

!1
J

L)

1-

LL.
J

p
O

II

V)
S

A)

f
'

CLr

V1
W
J

r"1
1-t;

:*
(A

l..
W

C.)

1,

'
/ `I

--

L
lV

/
Y
J

/n V /n V /
-

LI:

'

%Aj

C)

; lc
C
.

L-

>

W
L

N
W
f. J
O
crC{..

C.)

v^

r-

c\

".

tn

C)%

--

-c

[p

tV

t.1

tV

N\

!T

tt1

...

4b-E

c'

C)

C', u

c:

c:

c:

L, f

C
W

O
N
C
N

C)

_
N

W
K

F- f-N
O t+. F
UU

LI

.i

t't
._

tn
laJ
N

C)
Z

C
w

-o

O
LLJ
J

<

lu

(L'

r.

. "C

C)

O
G

". )

o
O

y"
C7

L.
M

la

...

C)

1. J

uf

".

cn
W
<(
C(
-.

.1
O

t")

cn

F-

co

....

tT

1"6..

l !

'Li

t"

t.!

r"
t7

AI

I `I

I,

N
L.
. ;

W
Ga

'w'

41

r-)

17.1

O
O

C3

-^1

"

r7

r
"-

^
41
J

"=

>

.'

t
f

LJ

F-

7
N

U
4

C)

N
Li

u
''"

IJiJ
?
:

(.
"
lJ
J

C'f
W

_
J
4

W
_

J
J

.M
O

0.
yJ

W
:: -

ua.

/ 'f

o_ .
/AV /A V

.z
J
"S
er_
W

C)

CJ

LL.

L.

La.

)
46.

:3

2 '

F-

"]

-.

a
t-

1-Y

t:.,

..

It
=I

o
",

"7

'2

:J

rS

tr

C.1

l.

'
_

ln

t tJ

:o :,

--

C:

..

".

tz

lyJ

;F-

N
L. l
C)
CJ
CT.
O-

vi
n
/A W W
J

u
1"

C-

W
c.7

Y'

.L

<.
"J

!)

O
=

S3

1=3
-

;;

..

Lf

Cy

.. "

`-

"

l
t

in

PT

Ul

^J

N1

V1

%O

p"

ct
.

487

C)

(=;

C3

.::

l..

C.

C.

la
W

C)

=
K1
G
N

F--

.J
V.

W
- cn w
C)
UU

4,
O

U.

Ell

[3..

C)

N
...

-S

LAJ

?
.

0.

W
r, 3

O
Z

4-

p
=

--

7
W

is :

fC
O

U.

;,,.
-

^.
'1

c
cm
rJ

LLJ
J

rn

W
Q

W
J

fi J

..,

1--

W
J

C7

N
W

W
".

C. )

K
T

O
C.1

U
Co

LL-

:3

j.;

. j
-.

th-

>
W

t`

%.)
"y

C=

,,,-

O
-

I\

I\

tJ

Li

rn

V
O
F-

F-

in

CD

OC

F-

2
.

F-

C7

/
tD
Q

-r

L
U

2
U

W
Z

0
O

co

fn

sQ

C7

""

"

Ca

v
a

in

V in

s
CJ

I..i

c^

L--)
F-

--

/n

%S

- - -

;
C

N
C)
"'

``'
__.
..

CJ
CS

CJ

"C
I.,

c-

_C

rl

.r
-.
2

4/i

=)

l.

t;

f:

"C

r/1
w
c:.

-C
Jr

W
C)

1,^1

u)

C,

-.

tV

co

tr

488

C.

Cr

CD

CD

v
c
N

LAJ
J

'''
D

Li.
O

f-N
pw
UC.

U.

C.

L,

Z
)

Hv
C
~

CN

F-

CL
C

c3

<L

N
Q

CO

LAJ

F-

W
O

J
f=.`

Li.
=

: -J

O
._

.. t

C:

O
.i

"2

U
---
17C-)

W
Z

F-

V)
J

L
U
C

UO
G
J

C7

-1

D
.

W
co

F-

C3

CJ

Zi
'2

N
W

p
4
C

ts.

ac

l7

:s
:;i

lJ

L-i

4N

p
C,

G.

c3

laJ

O..

.J

w
"
tL

rJC
J

L"i

"

%Q

r"

:4

[
:J

ea

10

14

Q
O

.d

I-

C1.

--

=
i- N
N
W
fJ

W
L

C1:
w..i

U.

r.

.n

'-

LV

)n

U
...

t?

4by

C)

C;

Cs

Cl

w
"C
C

%M
c
N

C)

C:

t )1

w
c=

-N I-G !
UU

N
W

JO

L)

La.
C_
O

lu
C

CL

ca

_
N
C

LLJ
N
N
'"j
W

L`

'-

o
^

rn

t,

:.

C:
S
>.

C7
"t

1-

L.

C9

CO
N

li.

_
`:

"r
n

0
!o

6'

w
1-'

.r

tr1

W
W

c '
-

ry
w

W
-. j

'

t. 3

G".

C:

.t

I
N

w
z_
s
v
4

L
N

^
Ci

LJ

C7
_WJ

L..

(>

J
J

\
1:. 1

O
JJ

=3

U.

U
L.
0

O
G.
0

-c

.a

P-

S4

t.. l
J

>

K-

G.
`

LJ
N

W
W

L.
0

W
.

1-.

[^

yr,

C6.

J
Q

C.

O
0

/ri A

J
4

/n
/jj\
/,
fn

s
W

J
J

1-

C7

W
C

W
L)
=
LV

, L.

C;

f"-

l+-

W
Cl-

w
"

L
1

v
W
I)

1_
1
`C

t.

W
t j

Cf.

:. s

r)

..

CL

O
._

IN

M 1V

J
.

tY

V
!

,.

ls:
al

.
.

'

.S

t. )

t7

r..
L.

tn

o
%,

t-

C17

111

%p

490

C:

co

CG

C:

Lt

C.

..

L/)
L. i
J

O
LA.;
La.

C,

C3

C]

Ca

L:
W

C)
[il

LL

r_
N
Q
UU

C
Z

N
W

M
La.

C/I
C
Q "

CL

LL

O
Z

cc

12

L--

co
C/)

Cl

LL
C

G..

ta.

LL

r')

S
W

J
J
C

K_
CD
C'

N
_
-LD

C.^.

_
v

1-W

.
41
J

O
LaJ
J

"

-'

C"U
C)

(Y

WC

`C_'
L.J

f, r^

N
LJ

:. 7
G

J
.:
O

Li

+C

CL

C-)

:. )

H
v

'

LL

P,

l/f
=

I-

L
U

l/1

Ci

J
7

"

-K

11

::f

"

V
_[

"

:n

inv inv in W
N

"

l/1

LC')

O
-

W
J

C)

L.J

cc
1-

L. J

C:

W
J

CD

zC

1-

W
\

L+J

)
C
LL.

'r
O
CO

r
L.J
tt:

V7

: 1.

K
v

tj

W
.1
W

r
W

.
:2

_
W

"

."

">

--

N
n

W
CA-

p
rI

lJ

LNJ

Lt1

O
-

1V

M
"

of

ui

491

t. i

C:

C.`i

Co

C:

C7

C:i

C3

C:

C,`

i,

C.

N
: -t
J

W
J

O
t

C
C-)

cv

Cl

r_-

W
J
J

tJ

M-

%0

t.

t- N

O
Z

O
UU

-14

f1C
V

=
%.. r

W
O
Z

tA

cc

Ci

i`
a

tL
0

LLC.)

O
Q

iv

. J

4-4

v
:i

9
_.

_i

N_

Vf

W
J

0
4

''

1- -

..

41
J

I--

r-

.. +

J
f-)

<

L. J

La

-C

t..

..

. -.

tA
W
-j
0

i.

.i

tA-

O
%q

ttff

*i

:L

W W W
-

:C

O
i

cr
92

(A

1-

J
Z

Lj

p".

CC

vs

eG

e.

"c

"

tJ
r"

"
"

..

J1

.`

W
V

"t

4-;

/AV /AV
N

J
J

W
4i

',

/A

"

"

L
4

J
v

=
l

!I
W+

=L

()

i.
.

r+1

%r

t':

t/1

-j

tJ

tL ,

. a.

LAJ

c`.

tJ

lJ
-

ti1

.1

1%

!,

tV

Mt

"1

UN

ttY

CO

Cj%

492

C/5

N
W
N
OW
UU

r_
1_

C
...
N

Q
O

G
N

"

Cn

J
C

W
J
U'

L,

_
C"r
C

ft,
.[

J
KL
H.

F--

N
Q

ln
W

Z
,

CD

o
W

J"

N
J

C
C

tn

C]

W
lil

x
U

<L

V7

V)

`
w
J

C
t,,,

J
J

CL

"d

/-

t--"

r
W
"=
Q

WW
ca

G
_

tL
C:

t. i

cc
0

"

t=

W
v

"S

41

'l

41

e.",

'L

-t

1-

. rg

ttN

.D

r-

tJ
-

CL

L..;

f. 1
(il

W
Cn

4.
0

493

v
N

NZ
taJ
O
(, U

FZ
O

N
CL
O
O

1QC
J

C)

C-0
J
J

"
1.

O
Z
QC

"

1
W

c7

ZZ-"
=
UmOL

U
O<<
z

4.

ae

"
7

J
W

z>

N_

CL.
L'i

<

U
<
2

!-C

0
zzCz

co

ca
a
OztO
J
I

Z,

d'

L.
CaJ

c_

{.7

U.

cC0

co

Z
W

F
0
N

Fl:
O

/-=
s,j

r`.

t)

z<

OR

XO
d.

._

1-_

IL
O

F"
_2G

GOJ

NN11
2
u
rx

zu

W
Z

11v

u
j

Q`

tC

77
C-

O
Z

S
FC7

CO

O
JL

XX

1-

rU
2z

"

z0

C7

H
O
J

f=

i..

4
CJ

Z
O

N
2
Fr^
Z

=
4J

W
J

J
J

..

4. I

I.c

0
N

Z
O

CD

J
J

W
ClV

r'

J
J
(N
tN
W
U
O

G.

fl.
=
1
FW
N

r0

O
O
Z
4

Li
tO

s*
W

-q

ll1

v7

sT

l! 1

.....

-NM

F-

<<1
=sU

; RX

494

W
F-
N

ow
UU

Cl

N
C
C/)

t,",
..

O
Zn
Z

M
G
C7
C

J
C

W
Z

crm
C.
^
-

Sc-,
6

x
i-

J
4

C)

r^

_ - ....

a
c

O
J

V7

Cr
W
Cl.

<

w
-C

J
-t

co

C%

(/7

tN
l. J
U
U
C:
G.

tO

lLl

"O

r-

495

.O
N
W

r.:
--

UU

z
0

0
N
N
X
Q

N
W
J
U

G
1-

Z
C

:Z
N
W

S;

1"

W
CL

"

.. t

t)

C7
Cl

C7

4J

W
U
4'
U.
(i

<

Cn

-Z

v`

7 .

'<

III
2

C`

n'
LA
'Z

N
7_
W

/
4:

Z:

"

'J

`:

.rr

III

/n V

`
C'7.
tu

Z.:

-_

I
C

. i

67
(/)
tJ
!)

C-)
(::
C...

r!..

r"-
41

i
I-

496

C.

C'
G-t,

F-

47

-,4 -"
_m
F-

x
LCi

w
c
F---

u"1.

N<
O l. J
C U

t+.

-
- -

c
W`^
.3.?
Cv

C]

.r

w
. .J

7-1

r=

c.?

t3 W

UV
u

--

6.
FCC
Ku
..,

_
1
W

L.

C"
=

NC
^

QW
"

i1
C.
NJ
=

1W

C]

C~

=i

G'

<-

a
W

<

.: ...

b-

Z.

CV
H

`
W

L'

p.

r2C
S
4

M' C

sG
W

1Y

h"

_;:

41
W

.im

17

<r

G
W

4.

...

r
v

` /
r

=
F-

l! 1

U1

: _
'
Sf

-'

=_

V1

"J

V- - n - -

(! 1
1, -i

y
V
W

n cc

inAr

I.

-m-

, "
o
-'.
cr

W\O
f"

\
M

'cr

_ +
rL

taJ
: '

C'
`V

li.
C

Vf

C
F"

r"
W

f. Z

>-

V7
W

-
n

.:.

C_

!
M

It

-a

4
IX. I..

c.
=
-

14

497

UN o
n

C,
C"i
w
CC

w
U

M
.

N
W

Ct

s3

N
r

a
J

ri
e
a
.,

>CC-,

U
S

c7
4

N
W

L}
S

?
Q

--

c^

,-

o
o
z

W
J

W
. -.

"`

s
..

.
J

v1

cn

...

:z

..

... f.

;_

17
V)
W

LA.

Av

...

0
W

f-

4.

In

Cf

, D
---r

.;

-n

1-0

r-

iz

CN

C
-

^J
.

rn

'

...

..,_
.r

v1
w
r
,;

498

TI

K,

al

c"I
N

4J

a:
r- i-oW
v c.

,.

Vf
W

O
_.
C
C
"

..

Q
t_

>W

F
4
Y
<!)

Y
'=

C7

N
W

J
co

. rJ

CL

c7
N

'

L-

--

N
-J

'

-J

--

V)

.
N

F-

Z
O

cc

7zi
Q

Cl

Q
_

N_
_

Cl,
H

l..
"^
V)

-C

.:

-C
Y"

::

VI
H

Q
t
c.

rn

Li

r
.J
W

U

".

y
N
N
Z

C/7
V7
W

pr: ..

4.
S
_

C^

1-t

'

F...

";r
C.

"C

f..
"f
J

rn

J
J

W
H

J
'7

-t
1,,.

-s

CC

C7

V1

'J

St

C7

499

.O
K1
N
J
La.

cn z

oW
c

-)c-)

O
N

z
a
0
x

"
W

2
0

Li
a

N
z
u

w`

z
O_

N
N
O
a
U

N
O
U

co

cr.
V.
1.O

?C
S

J
J
Q"
O
O_
N
cr.
W
W
W
C:

N
'1.
W
M'
C3

Sc
x

00

W
F
c: j

F"
C.-.
W

[
'v
d

U
t
Cn

HW

cN

N
(/1
lJ
lJ
4.

N
J

r1

lC%

19

1-

co

C%

bUU

Co
rn

lJ

.
N C) W
UU

ZZ

Ljj

C3

n
N

aJ

i
.;

:..

. -.
W

F""

-N

N
f/1

cn
W

v
-

L. J

_.

rn

ul

-o

rn

Li

Co
1
FN

L..
C
F'-

OW
UU

N
LW
OO
O

F+
'O 0
cii L
r U
..F+
c7
F+ W

u
t,o p,
CZ
...

x
v

ub

x
v

N
"r1

n7

C!
l+

ea u
ca
... ...
U

C]

Y
3

F-

Vf

ti
...
(n
bf

CL
1-W

C=
0

Q
J

X
C>

C)
U
Fu

I-

W
W

u
1<
J

J
C..?
6

J
N

4.
0

'

-tI',

L^

w
a

tn
cn
W
r
U
G.

O
M

L'1
I

N
'M

C
r.

-C
J

:s

r_ )

"

_"

"

p..
"1
J

1
.

C.
n

'

"

,,)

w
Z

1-.
C]

I+)

LJ

. tO
tJ

==

-0

f-

(: )

d
CO"
"C N
41

i
rn

....

4.
t.7

lt,.

CJ

.r
w r+
co
r

F-

W
(1)

v
o

ri d
w
. N

C.7

r;

U,

I
O

Co
LC1
N

L:

w
N
OW
UU

C--

LLJ
4

t+..

_
W

t)

C
"

"'

C
M

W
O

0-

O
=

C.

.
.

O
"

C
t`
.

W
O

F-

v
N

W
L

-j

O
I-

O
I-

C?
v

Cl)
LJ
J
G
KL

=L
IC
U

[C

N
1.
CD

LL
Cl

v'

N
_
H
R

<
C)
W

\
+...

VI
'

"

C:J

t7

~9
'..
'L

>

.J W
?
<

.
'7
7

N
T
1

W
J

F--

L
C

O
Z

W
J

N
S
F-

C7
W
J

]..
.
J

C
t,. J
Ii

-i

CJ
^'
C7

d
w

.J

N
Cl.
G

_
v

`'
ti

_. +

W
(T

"

i"

UU

'
..

'

:.

TA

W
L

41
N

^
W

I-

C'7
-.
J

T
v

N
a

='J

4J

IAJ

N
_,
,

cC

C-)

=
*

L.

<t

CT3

cW
O

'

Y
=

II
I

_
S

"

`iI

.: c

lA
J

` '
`y

A
\\
//

"

IA J Z:

V
"

y_

W
c.7

2E .

K , o

:T-

O..

A.
W

G
(N
C/)
lJ
C, )

ta.

:
-
4"-

<
:

`r 4
-

W
-C

4i
:

-L
W
_w
I-

.''
=

a
O

.J

1r

nti
U
J

c
\

CL.
CL

Ci

-W

_:I c.

:V

v
-

"

wh

503

C3

c.

U-1 a

U
N

i -.
"
O

C)

C)

C:

L:i

Li

U
i

Co
LN
O
W

ClV) z
o
UU

W
Q

oe
W

J
O

G`

Ow
CU

C-C
1-

C)
0
Z

c7
z
H

C)

''_]'
L'i
J
J

Li

La

o
o
Z

C)

rr
O
co

r_
C)
:1

C)

N
j
W

7.

F""
o

M
U.
.

`L
L+"

N
W

Cr

xU

U
F-

U
C
N
7

cV

..

<L
G

M-

u
co

1.
C

., J
w

C)
C]

W
N
'

C.7

2
J
C

:C
J
C
LJ

O
F-

"
W
J

.JW
-

C3
W
J

J
G
!-

Vf
Li

_
c

1 `/

G7

CO

C
U

L7

1"-

r
L!i

<

-"

1--

N
}

1--

_1

U
Q

C)

y
o
G
CJ,

J
_
cc
C]

O
=

N
'=
O

'O

C')
<

G
LL.

C.J

d
O

,,

J
J

vN
a.

C/)
CL'

<i"

c:

U
CC

O
C

4.
C)

lwi.
_

LJ

W
=

-.
-

Q
C:

CL.

1--"
i ,"
'.
W

J
fm

U.
O

L.
CD

L.J
J

LAJ

W
1)

W
C"

v
W
_

yz

F...

!-

CC

ri

LWa.
LaJ

o
'2

. -.
i-+

"

`--^
Ci.

ui
c.7

71
-

nv nv
J

J
J
K

'

rL

nl

;
W
@

F-

Li.
"
"".

'

U
cn

Cr

O
:)
7

Vf
N
ki

L
lb

y
L

Q:
cc

F-

R:I.

"I

-: c
J
=

--<

`
iJ

Lr.

2
`V)..

L-

L:)
1-

'. J

"t )
c.

hl

tl

aJ

U-i

%0

(X)

Lh

504
E

a
S

r+"

,a

-t
%0

0
a

C'1

00

I-

o O
N

yp
N

0
--

a a a a
co -Ir

zr co u,

tn

-4

CV

a
a

r,

(31,
Ln in

.40

41
U
O
S.50
U

ro
s
+

C
. r.

N
N
CJ

rUl
C

"r

vp

.7

r-1

Oi
U

r-4

d
4-)
to

"

94

a a

a a

4j

aj

aj

a
=

.G

C:

a
1

co0
OU

+J

PQ

u
Q

H
x

0
00

11
A

41

JJ

ai

a a wl
v
.c

.c

H
G

+L-

OOr.u
'

C)
(L) NN

tL

04)
L
()
41 Vf
U

a
+

a.

a a
.

.
.C

.
H
.

C)-~+
S-CD

H
.

C,)

~
E+

>

a
H

t
H

cD
zH

a
0
P4
H
9

>

A
6

C.
C) C

WH

-+

...

4O4--0

vi
tn
DG
O

L
()

#.A
". -

"c

tn.

i=^Q1S..

`"

". - .b.+ Nf (1)


Cr-"
Y
41
(1)
3WLn

rt7 -.
GNIb
41
Q1 LA Q1 C)
()
C)

C)
r-

11 L

r-

C)
C

Y "C
Ur-

i""
.CLC C)
++
to NE
471
d- r-

L () O. ]L
C) EO

C) iJ
rC) 4- Ln
3C".
VL
-0
W"C
...
4J
C) i
Cl WU
.-O
Ei-.
br-S

C) "rb4.
J
"r"
. r" C) O rO -0 11 C. VII

II

O
GLL. J+1

O=
CCm

11

Ce
.

"