Design for X Objectives of the session

advertisement
Design for X
Dr Jane Marshall
Product Excellence using 6 Sigma
Module
PEUSS 2012/2013
Design for X
Page 1
Objectives of the session
• Introduce Design for X
• Discuss Specific Design for X methods
PEUSS 2012/2013
Design for X
Page 2
1
Design for X
• Concurrent engineering is a contemporary approach to DFSS.
• DFX techniques are part of detail design
• To improve:
– life-cycle cost; quality, increased design flexibility, and increased efficiency and
productivity
• Benefits include:
– competitiveness measures, improved decision-making, and enhanced operational
efficiency.
“X” in DFX is made up of two parts: life-cycle processes x and
performance measure (ability) (Huang 1996).
• Effective approach to implement concurrent engineering.
• System design
•
PEUSS 2012/2013
Design for X
Page 3
System design
• Systems and its characteristics
– System is an interrelated set of components, with identifiable boundary, working
together for some purpose
•
A system has nine characteristics:
–
–
–
–
–
–
–
–
–
Components----------------------Subsystems
Interrelated components
A boundary
A purpose
An environment
Interfaces
Input
Output
Constraints
PEUSS 2012/2013
Design for X
Page 4
2
Interface
Environment
Components
Input
Output
Boundary
PEUSS 2012/2013
Design for X
Interrelationship
Page 5
System characteristics
• A component
• an irreducible part or aggregation of parts that make up a
system, also called a subsystem
• Interrelated components
• Dependence of one subsystem on one or more
subsystems
• Boundary
• The line that marks the inside and outside of a system and
that sets off the system form its environment
PEUSS 2012/2013
Design for X
Page 6
3
System characteristics
• Purpose
• The overall goal or function of a system
• Environment
• Everything external to a system that interacts with the
system
• Interface
• Point of contact where a system meets its environment or
where subsystems meet each other.
PEUSS 2012/2013
Design for X
Page 7
System characteristics
• Constraint
• A limit to what a system can accomplish
• Input
• Whatever a system takes from its environment in order to
fulfill its purpose
• Output
• Whatever a system returns from its environment in order
to fulfill its purpose
PEUSS 2012/2013
Design for X
Page 8
4
Logical and Physical system
description
• Logical system description
• Description of a system that focuses on the system function and
purpose without regard to how the system will physically
implemented
• Physical system description
• Description of a system that focuses on the how the system will
be materially constructed
PEUSS 2012/2013
Design for X
Page 9
Important system concepts
•Decomposition
•Modularity
•Coupling
•Cohesion
PEUSS 2012/2013
Design for X
Page 10
5
Decomposition
Is the process of breaking down a system into smaller
components in order to:
– Focus on one area at a time
– Concentrate one component pertinent to one group of
users
– Build different components at independent times
PEUSS 2012/2013
Design for X
Page 11
Modularity and Coupling
• Modularity
– Dividing a system up into chunks or modules of a relatively
uniform size. To Simplify the redesign and rebuild process
• Coupling
– The extent to which subsystems depend on each other.
– Subsystem should be independent as possible. If one subsystem
fails and other subsystem are highly dependent on it, then the other
will either fail themselves or have problems functioning
PEUSS 2012/2013
Design for X
Page 12
6
Cohesion
• A cohesion is the extent to which a subsystem
performs a single function.
PEUSS 2012/2013
Design for X
Page 13
Design for X tools
•
•
•
•
•
•
•
Design for Manufacture and Assembly;
Design for Reliability;
Design for Maintainability;
Design for Serviceability;
Design for the Environment;
Design for Life Cycle Cost;
Etc.
PEUSS 2012/2013
Design for X
Page 14
7
Design for manufacture and
assembly
• Consider during concept stage to reduce cost of
redesign later
• Involvement of manufacturing engineers and supply
chain management at concept phase
• Involve investigating waste
• Reducing number of standalone parts
• DFA time – identifies cost of assembly and efficiency
• Investigate alternatives through experimentation
• Error-proofing – poke-yoke
PEUSS 2012/2013
Design for X
Page 15
Design for Reliability
• Analyse load-strength relationship
• Design out inherent weaknesses
– Analyse potential failure mechanisms and remove or
accommodate
– Analyse potential failure modes
– past data
– suppliers history
• Plan for reliability improvement
• Plan correct testing
PEUSS 2012/2013
Design for X
Page 16
8
What is Reliability?
• Quality over a period of time
• The ability of an item to perform a ‘required
function’ under ‘stated conditions’ for a
‘stated period of time’.
• Reliability is an aspect of engineering
uncertainty and therefore measured using
probability
PEUSS 2012/2013
Design for X
Page 17
Why is reliability important?
•
•
•
•
Costs of unreliability or unavailability
Safety
Competitiveness
Examples:
– Airline costs
– Automotive reputation
– Military equipment availability
PEUSS 2012/2013
Design for X
Page 18
9
Why do products fail?
•
•
•
•
•
•
•
Design inherently incapable
Item overstressed
Variation
Wear-out
Specification, design or coding errors
Noise
Interaction
PEUSS 2012/2013
Design for X
Page 19
How do Products fail?
• Discuss how products can fail?
• Choose different product types and give ideas
for how they may fail?
PEUSS 2012/2013
Design for X
Page 20
10
Wear-out and Over-stress
OVER-STRESS
WEAR-OUT
Fracture
Fatigue
Yield
Creep
Thermal
Erosion
Electrostatic
discharge
Corrosion
PEUSS 2012/2013
Design for X
Page 21
Electronic component failure examples
Mechanical
failure
• Chip capacitor overstress
• PCB PTH barrel crack
Solder
joint
fatigue
• Chip resistor
PEUSS 2012/2013
Design for X
• BGA active device
Page 22
11
Corrosion
Fretting Corrosion
Galvanic corrosion
PEUSS 2012/2013
Design for X
Erosion Corrosion
Page 23
Load and strength
• Load - Key Characteristic of interest in
environment
–Temperature / Pressure
– Vibration / Humidity
• Strength - Key characteristic of interest in
manufactured product
–Tensile Strength
–Current Rating
–Propagation Delay
PEUSS 2012/2013
Design for X
Page 24
12
Requirements & Environment - Engine Controls
FADEC
• Typical Civil Engine environment
• Temperature extremes
• operating
-55 to 100
• non-operating
-65 to 125
Reliability
Requirements
• Vibration extremes
• low frequency
<5g
• general
20g
• blade passing
50g
• Typical environment
• much more benign
• Service requirements
• Life 25 yrs 100,000 hrs operating
• Reliability 30,000 hours MTBF
failures/million hrs) Design for X
In a Safety Critical Harsh
Environment Application
PEUSS 2012/2013(33
Page 25
Load and strength
Generally failures occur if
Load exceeds Strength
PEUSS 2012/2013
Design for X
Page 26
13
Probability
Load and strength
Load
Strength
No overlap of distributed values ….. No failures
PEUSS 2012/2013
Design for X
Page 27
Load and strength
Probability
Failure
Strength
Load
PEUSS 2012/2013
Design for X
Page 28
14
Over-stress
Load
Strength
Stress
PEUSS 2012/2013
Design for X
Page 29
Wear-out
Load
Strength
Stress
PEUSS 2012/2013
Design for X
Page 30
15
Hazard function
Bathtub curve
Useful Life
Time
Infant
Mortality
PEUSS 2012/2013
Wear Out
Design for X
Page 31
Design for Reliability
Aim to maximise reliability during service life by:
– Measurement & control of manufacturing quality /
screening
– Optimized design & build process to improve
intrinsic reliability
– Assure no systematic faults present in product
– Provide sufficient margin to meet life requirements
PEUSS 2012/2013
Design for X
Page 32
16
Hazard function
Early - life
Useful Life
Time
Infant
Mortality/early life
PEUSS 2012/2013
Wear Out
Design for X
Page 33
Early Life Failures –
Environmental Stress Screening
• Sometimes called burn-in – electronic
components
• Stresses that cause defective production items
which pass other tests to fail
• Normally 100% test but can be done for batches
• HASS – high combined stresses
– Test time shorter than ESS
– Cheaper
PEUSS 2012/2013
Design for X
Page 34
17
Early-life failures
Quality in
Manufacture
Life Margin
Service Life
Hazard function
e.g. 20 Years
Screening
Effectiveness
Reliability
Life Failures
Time
Perceived
Quality
PEUSS 2012/2013
Design for X
Page 35
Early-life failures – Too little
product screening
Service Life
Life Margin
e.g. 20 Years
Screening
Effectiveness
Degraded Reliability
Increased Low Hour Failures
Improved 1st Time Pass
Yield
PEUSS 2012/2013
Design for X
Page 36
18
Early-life failures – Too much
product screening
Service Life
Life Margin
e.g. 20 Years
Screening
Effectiveness
Higher Service Reliability
Reduced Low Hour Failures
Reduced 1st Time Pass Yield
Reduced Life Margin
PEUSS 2012/2013
Design for X
Page 37
Screening effectiveness
• At what point is Manufacturing Quality measured?
– Following screening ? …. following x hours in service?
– Processes optimised to just pass ESS?
– Quality engineers aware of early life (service) failure?
• How well is E.S.S. Effectiveness measured ?
– Failures being detected during ESS?
• What type of failure, at what point and in which test.. ?
• Optimised cycles, levels,etc ?
• Measures of performance ?
PEUSS 2012/2013
Design for X
Page 38
19
Product screening – the answer?
Service Life
Improved Build Quality
Improved Quality
in Manufacture
More Stringent Quality Control
Less Reliance on Screening
Higher Service Reliability
Reduced Low Hour Failures
Improved 1st Time Pass Yield
Increased Life Margin
PEUSS 2012/2013
Design for X
Page 39
Hazard function
Useful- life
Useful Life
Time
Infant
Mortality
PEUSS 2012/2013
Wear Out
Design for X
Page 40
20
Useful- life
• Characterised by failures occurring randomly
over time
• Due to non-systematic, minor anomalous
process variations, for example:
– Marginal design / tolerance build-up
– Random periods of reduced manufacturing process
control
– Component part batch problems
– Maintenance induced failure
PEUSS 2012/2013
Design for X
Page 41
Minimising failures during useful life
• Learning (and using) lessons from current
products
– Determine those factors that have a significant effect
on product reliability by analysing in-service with
manufacturing data.
– Root Cause analysis of service issues enables
generic process improvements in design and
manufacture
PEUSS 2012/2013
Design for X
Page 42
21
Hazard function
Wear-out
Useful Life
Time
Infant
Mortality
PEUSS 2012/2013
Wear Out
Design for X
Page 43
Wear-out
• Characterised by failure mechanisms broadly repeated
across a product type, often following a similar period
of use.
• Due to systematic weak links in design / manufacture.
– Processes that consistently stress product during
Manufacture
– Design / Selection of inappropriate parts or function
– Incompatibility between design and manufacturing capability
• Not necessarily an ‘end of life’ issue
– Could occur at any point during product life
PEUSS 2012/2013
Design for X
Page 44
22
Wear-out - solutions
• How can systematic problems be detected prior
to use ?
– Physics of Failure (PoF) analysis,
– Finite Element Analysis,
– Design Of Experiments,
– Reliability Enhancement Testing (RET)
PEUSS 2012/2013
Design for X
Page 45
Design for Reliability
• Design for useful life
• Design out inherent weaknesses using tools such as:
–
–
–
–
–
–
–
FMECA
Fault tree analysis (FTA)
Physics of Failure
ESS
Finite element analysis
Development Testing
Data analysis
PEUSS 2012/2013
Design for X
Page 46
23
Product life cycle
Design
FMECA, FTA, PoF,RBD
FE,accelerated life test
Development
Development Test
Use
Field data analysis
FRACAS
Test
Manufacture
ESS, Burn-in
PEUSS 2012/2013
SPC
Design for X
Page 47
Design for maintainability
•
To minimise:
–
–
–
–
–
•
The downtime for maintenance,
user and technician maintenance time,
personnel injury resulting from maintenance tasks,
cost resulting from maintainability features, and
logistics requirements for replacement parts,
backup units, and personnel
Maintenance actions can be preventive,
corrective, or recycle and overhaul
PEUSS 2012/2013
Design for X
Page 48
24
Design for serviceability
• Ability to diagnose, remove, replace, replenish,
or repair any component or subassembly to
original specifications with relative ease
• Poor serviceability produces warranty costs,
customer dissatisfaction, and lost sales and
market share due to loss loyalty
• Have serviceability personnel involved in the
early stages, as they are considered a customer
segment.
PEUSS 2012/2013
Design for X
Page 49
Design for the environment
• Addresses environmental concerns as well as postproduction transport, consumption, maintenance, and
repair
• The aim is to minimize environmental impact, including
strategic level of policy decision-making and design
development
• DFE usually comes with added initial cost, causing an
increment of total life cost
• Economic evaluation is required both for maximum
economic benefit and to estimate what the expected
financial savings (or losses) will be
PEUSS 2012/2013
Design for X
Page 50
25
Design for life-cycle cost
• Real cost of the design
• Includes the associated costs of defects, litigations,
buybacks, distributions support, warranty, and the
implementation cost of all employed DFX methods
• Activity-based cost (ABC)* is a method for estimating
life-cycle design cost
• ABC assumes that the design, whether a product, a
service, or a process, consumes activities.
• ABC objective is to identify activities in the design life,
and then assign reliable cost drivers and consumption
intensities to the activities
PEUSS 2012/2013
Design for X
Page 51
26
Download