MPD 575 Design for Reliability

1
MPD 575Design for Reliability

• Jonathan Weaver

2
DReliability Development History

• Originally developed by MPD Cohort 3 team of
  Julie Earle, Dave Herczeg, and Jim Van Gilder in
  Fall 2002.

3
Design for Reliability
4
Why Design for Reliability?

• Reliability can make or break the long-term
  success of a product
• Too high reliability will cause the product to be
  too expensive
• Too low reliability will cause warranty and
  repair costs to be high and therefore market
  share will be lost

5
What is Reliability?

• Reliability is
• Elimination/avoidance of failure modes/mistakes
  
• The probability that a product will perform its
  intended function
• Under customer operating conditions
• For a specified life
• In a manner that meets or exceeds customer
  expectations
• A reliable product is robust and mistake-free

6
What is Probability?

• Probability is
• a measure that describes the chance or
  likelihood that an event will occur.
• The probability that event (A) occurs is
  represented by a number between 0 (zero) and 1.
• When P(A) 0, the event cannot occur.
• When P(A) 1, the event is certain to occur.
• When P(A) 0.5, the event is as likely to occur
  as it is not.

7
Reliability Failure Modes

• Two types of failure mode
• a) hard something breaks
• b) soft performance degrades
• Two root causes
• 1. lack of robustness (sensitivity to noise
  factors)
• 2. mistakes

8
What are Noise Factors?

• Noise Factors are sources of disturbing
  influences that can disrupt the ideal function,
  causing error states which lead
• to quality problems.

9
What is Population and Sample Size?

• A population is
• The entire group to be studied (e.g. all Ford
  Contours)
• A sample is
• A subset of a population selected randomly for
  analysis (e.g., every hundredth Ford Contour off
  the assembly line)

10
Common Measures of Unreliability

• Failure - of failures in a total population
• MTTF (Mean Time To Failure) - the average time of
  operation to first failure.
• MTBF (Mean Time Between Failure) - the average
  time between product failures.
• Repairs Per Thousand (R/1000)
• Bq Life Life at which q of the population will
  fail

11
Introduction to DFR

• DFR has many aliases
• Design for Durability
• Design for Robustness
• Design for Useful Life

12
When to Use DFR

• DFR should be considered throughout the PD cycle
• Early - to develop "product concepts" which are
  well suited for production (i.e., conceptual
  product design)
• Continually - to ensure that the chosen product
  concept is implemented through optimal component
  design

13
Automotive Reliability Facts

• The shortest route to higher satisfaction is not
  only through the dealership service department
• it is mainly through keeping customers out of
  the service department in the first place.
• Customers who report zero problems with their new
  cars have an owner loyalty rate of 73 percent and
  dealer loyalty of 42 percent.
• At 4 TGW, loyalty to the company drops by 1/3 to
  44, while loyalty to the dealer drops to zero.

14
Automotive Reliability Facts

• The average age of a purchased vehicle at the
  time of replacement is 5.7 years in the U.S. and
  4-5 years in Europe.
• The average lifetime of a vehicle before scrap is
  12.7 years in the U.S. and 10 years in Europe.

15
Steps in Designing for Reliability

• Develop a Reliability Plan
• Determine Which Reliability Tools are Needed
• Analyze Noise Factors
• Tests for Reliability
• Track Failures and Determine Corrective Actions

16
1. Develop a Reliability Plan

• Planning for reliability is just as important as
  planning for design and manufacturing. Why? To
  determine
• useful life of product
• what accelerated life testing to be used
• where to begin
• Reliability must be as close to perfect as
  possible for the products useful life.

17
1. Develop a Reliability Product Plan

• A Reliability Plan helps ensure that product
  reliability is optimized within the cost and
  performance constraints of a program and customer
  requirements.

18
1. Develop a Reliability Plan

• How much reliability do you need? Should you
  accelerate life testing? Where do you even begin?
  
• Planning for product reliability is just as
  important as planning for product design and
  manufacturing.
• The amount of product reliability must be in
  proportion to a product's usage and warranty
  goals. Too much reliability and the product will
  be too expensive. Too little reliability and
  warranty and repair costs will be high.
• You MUST know where your product's major points
  of failure are!

19
Some Reliability Tools

• Block Diagram
• P-Diagram
• QFD
• DFMEA PFMEA
• Design Verification Plan
• Key Life Testing
• Weibull Testing
• Reliability Demonstration Matrix

20
Reliability Block Diagram

• Three categories
• Series
• Parallel (Redundant)
• Complex (combo of the two shown below)

21
P-Diagram
Noises
Outputs
Input
J
System
Signal
IDEAL Response
(energy related)
(energy related)
error states/ failure modes
L
Control Factors
22
Quality Function Deployment
23
FMEAs

• Potential Failure Mode
• Potential Effects of Failure
• Severity
• Classification
• Potential Cause/Mechanism of Failure
• Occurrence

• Design Controls (Prevention/Detection)
• Detection
• Risk Priority Number
• Recommended Actions
• Responsibility/Target Completion Date
• Actions

24
DVPRs

• Test Specification
• Acceptance Criteria
• Test Results
• Design Level
• Quantity Required

• Quantity Tested
• Scheduled Start/ Complete
• Actual Start/ Complete
• Remarks

25
Reliability Demonstration Matrix
Robustness Assessment and Noise Factor Management
Matrix
In the development of robustness, it is
essential to provide one noise condition for each
failure mode. Don Clausing, Professor of
Engineering, MIT.
Potential Failure modes
Available Tests
Failure mode to test traceability and Noise
factor to test traceability leading to
... Reliability Robustness Demonstration
Noises 1
Noise factor management strategy
Noises 2
Noise to failure mode traceability
Noises 3
Noises 4
Noises 5
26
Reliability Demonstration Matrix
Robustness Demonstration
Battery Suspension bushing
27
2. Analyze Noise Factors

• Inner Noises
• Wear-out or fatigue
• Piece-to-piece variation
• Interfaces with neighboring subsystems
• Outer Noises
• External Operating Environment (e.g., climate,
  road conditions, etc.)
• Customer usage / duty cycle

28
2. Reduce Sensitivity to Noise Factors

• Change the design concept
• Make basic current design assumptions insensitive
  to the noises design out failure
• Parameter Design
• Beef Up Design
• Insert a compensation device
• Disguise the effect - Send the error state/noise
  where it will do less harm

29
2. Noise Factor Management

• 1. 2. 3. 4. 5.
• Change (i)Parameter (ii)Beef-up Reduce
  Comp- Disguise
• Concept Design. Design
  Noise ensate

• Piece-to-piece x x x
• Wear Out x x
  x
• Customer Use x x x
• External Environment x x x
• System Interactions x
  x x x

30
3. Test for Reliability

• How robust are the products?
• Test to Bogey assessing performance at a
  predetermined time, cycle or number of miles. It
  estimates the proportion of failures at a
  particular time. pass/fail
• Test to Failure shows when a component or system
  can no longer perform at a specified level
• Degradation Testing focuses on the key stresses
  associated with real world uses for example -
  increasing the tire load to create a tire failure
• How can you shorten the reliability test time for
  new designs?
• Key Life Test/Accelerated Test

31
3. Example Testing for Reliability

• Proportional Hazard Model to Tire Design Analysis
• Perform Root cause analysis
• Consists of laboratory tests aimed to duplicate
  field failures
• Tire geometry and physical properties are
  selected as variables that potentially affect the
  tire
• Survival data is analyzed by a proportional
  hazard model
• The adequacy is assessed by the chi-square
  goodness- of fit test and the Cox-Snell residual
  analysis
• Identify elements of a tire design that affect
  the probability of tire failure due to failure
  mode in question.

32
3. Example - Testing for Reliability Contd

• Type of failure mode analyzed tread and belt
  separation

33
3. Example - Testing for Reliability Contd

• Tread and belt separation can be considered a
  sequence of two events
• Failure crack initiation in the wedge area
• Crack propagation between the belts
• Design characteristics that could be variables
• Tire age
• Wedge gauge
• Interbelt gauge
• End of belt 2 to buttress
• Peel force
• Percent of carbon black (chemical in rubber)

34
3. Example - Testing for Reliability Contd

• Testing procedure
• Dyno testing
• Warm up over 2 hours at 50 mph
• Cool down over 2 hour at full stop
• At 1300 lbs of load speed steps starting at 75
  mph and increasing by 5 mph every half hour till
  90 mph and then every hour till failure
• At 1500 lbs of load all the above speed steps
  are half-hour duration

35
3. Example - Testing for Reliability Contd

• Test speed profile

36
3. Example - Testing for Reliability Contd

• Vibration and sound pattern of tire before tread
  and belt separation failure

37
3. Example - Testing for Reliability Contd

• Test data set used in proportional hazard analysis

38
3. Example - Testing for Reliability Contd

• Estimates of proportional hazard model with
  covariates identified

39
3. Example - Testing for Reliability Contd

• Estimates of Proportional Hazard Model with
  statistically significant covariates

40
3. Example - Testing for Reliability Contd

• Exponential probability plot of Cox-Snell
  Residuals

41
3. Example - Testing for Reliability Contd

• Cumulative Hazard function predicted from the
  estimated model based on some typical values of
  covariates for poor and good tires

42
3. Example - Testing for Reliability Contd

• Conclusion
• Wedge and interbelt gauges as well as the peel
  force are significant factors affecting hazard
  rate of tire and belt separation failures in an
  inversely proportional way
• Agree with hypothesis

43
3. Test for Reliability

• Component design and manufacturing technologies
  are becoming increasingly complex.
• As geometries shrink and development cycles
  shorten, opportunities for defects increase.
• Testing for Reliability is becoming increasingly
  important.

44
4. Track Failures and Determine Corrective
Actions

• This process involves
• Data collection and selection
• Set up databases for tracking failures
• Warranty, Early Warranty, Things Gone Wrong
• Analyzing trends
• Performing closed loop analysis/corrective action
• Calculating observed reliability parameters
• Assessing reliability growth.

45
4. Track Failures and Determine Corrective
Actions

• Brake warranty is on track with targets and
  achieves more than 60 warranty CPU reduction
  since 1994
• Brake health charts were instituted in 1995 to
  monitor key performance index and drive design
  competency
• Supplier business unit reviews (BURS) quarterly
  to address key quality and manufacturing issues

46
4. Track Failures and Determine Corrective
Actions

• Time-to-Failure Curve

47
Challenges in DFR

• Many CAE models have limited capability to
  represent real-world noise therefore, surrogate
  noise based on engineering knowledge is required.
• Precise reliability estimates require precise
  knowledge of statistical distributions of noise
  factors. 
• As a contrast, comparative reliability
  assessments and robust design require only
  approximate knowledge of statistical
  distributions.

48
Challenges in DFR

• Many CAE models are computationally expensive
• preparation time to set up the model
• computing time
• Many CAE models focus on error states (e.g.,
  fatigue, vibration, noise) therefore, a
  multi-objective optimization is often needed.
• In early product development, when the impact of
  robust design can be greatest, design objectives
  and constraints are still imprecise.

49
References

• Reliability - Ford Design Institute
• Ford Reliability Class T. P. Davis, V.
  Krivtsov/ VKRIVTSO
• http//www.reliabilityanalysislab.com/ReliabilityS
  ervices.asp
• U.S. R.L. Polk Vehicles in Operation Report
  June, 1997 Europe New Car Buyer StudyEuropean
  Buyer - Big Five Survey 1995
• Ford's Strategy in Reliability (Prof. Tim Davis)
• http//pms401.pd9.ford.com8080/arr/concept.htm