ARS 2005 San Diego, California, USA
Track 1, Session 7
Reliability Integration Across the Product Life Cycle Fred Schenkelberg & Mike Silverman Ops A La Carte LLC
Introduction z
“Reliability? Oh, that’s 50,000 hour MTBF!”
z
“We design for a 5 year life.”
z
“Every project passes HALT, no problem.” A successful product reliability program is the collection of goals, plans, tools, and skills focused on the cost-effective and timely creation of products that meet business objectives. Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 2
Agenda Introduction and Agenda z Overview z Reliability Goals z Reliability Plan z Reliability Execution z
FMEA and HALT Prediction and RDT/ALT
Integration z Wrap-up z
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 3
Overview Goal Plan FMEA HALT
Prediction RDT/ALT
Verification Review Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 4
Overview Goal Plan FMEA HALT
Prediction RDT/ALT
Verification Review Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 5
Overview Goal Plan FMEA HALT
Prediction RDT/ALT
Verification Review Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 6
Overview Goal Plan FMEA HALT
Prediction RDT/ALT
Verification Review Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 7
Vocabulary Reliability z Goal z Plan z FMEA z HALT z Prediction z RDT z ALT z
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 8
Reliability Goal 1. 2. 3. 4. 5. z
Function Probability Duration Environment Customer Expectation Example: Our Power Supply will provide 5V with a 90% reliability at 2 years in a home and office environment. Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 9
Reliability Goal – How?
z
Business objectives Warranty period Useful or expected functional life
z
What is it supposed to do?
z
Where and under what conditions?
z
Estimated cost of product failure
z z
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 10
Reliability Plan z
Goals, apportionment and metrics
Selected tools for discovery z Selected tools for analysis z Selected tools for verification z
Procurement activities and critical parts z Field service, maintenance, call centers z
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 11
Reliability Plan
Engineering Review Determine Reliability Targets Initial Design FMEA
Apportion Target & Determine Gaps Initial Design (Prototype #1)
Subsystem ALTs & drop testing
HALT
Competitive HALT
Competitive Teardown
Improve Design (Prototype #2)
Revise Apportion Targets and Gaps
Update FMEA
DOE (if needed)
Derating Analysis
HALT
Improve Design (Prototype #3 or Production)
HALT
Reliability Demonstration
Benchtop Lifecycle testing
Identity CTQ
Final System Reliability Analysis
Establish Control Plans
Review Results Monitor Metrics Product Launch
Mike Silverman, Ops A La Carte
Track 1
Session 7
Monitor CTQ parameters
Slide Number: 12
Reliability Execution z
Two distinct reliability schools Measurement techniques Prediction o RDT and ALT o
Improvement techniques FMEA o HALT o
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 13
Reliability Execution z
Which should we use? Both, of course... but how do we integrate them together? FMEA
Prediction
HALT
RDT/ALT
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 14
Reliability Measurement Techniques Prediction in Design Phase z RDT in Prototype Phase z
Prediction RDT/ALT Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 15
Reliability Prediction z
Reliability Prediction Definition: A method of calculating the reliability of a product or piece of a product from the bottom up by assigning a failure rate to each individual component and then summing all of the failure rates.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 16
Reliability Prediction z
Reliability Predictions are used to:
help calculate spares provide input to system-level reliability models assist in deciding which product to purchase drive design trade-off studies set achievable in-service performance standards help set test parameters for RDTs/ALTs
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 17
Reliability Prediction z
Limitations of a Prediction Predictions have inaccuracies due to: o o o o o
Data is from outdated standards Data is from different environments Data is from different applications Data is from unclean field data Technology is too new – no good models
So what can we do?
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 18
Reliability Prediction z
Getting Around Prediction Limitations We use relative failure rate values rather than absolute values and feed to a FMEA. We fill in gaps of inaccuracy with RDTs and ALTs.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 19
Prediction Integrated with FMEA z
Use results of Prediction for a FMEA In a FMEA, we are looking for high risk areas. To quantify these risks, we need to assess the probability of the risk occurring. Predictions can supply this probability. Usually, a probability relative to other risks is all that is necessary, and a prediction is good for this.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 20
Prediction Integrated with RDT/ALT z
Use results of Prediction to plan an RDT/ALT Predictions will identify where a product is vulnerable to help decide which RDT/ALT to run. Predictions can give you sensitivity of components to thermal and electrical stress – key accelerants in the RDT/ALT models. Predictions can help determine stress limits of components when choosing accelerated stress levels for a test.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 21
Reliability Demonstration Test (RDT) z
RDT: Definition A sample of units are tested to validate reliability requirements. The test is usually performed at accelerated stresses to compress time. The accelerated stresses can be environmental, electrical, or mechanical.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 22
Accelerated Life Test (ALT) z
ALT: Similar to RDT but also used: to characterize dominant failure mechanisms, often due to wearout. usually at the assembly level rather than system level to target specific areas of design.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 23
RDT and ALT Parameters z
In order to set up an ALT, we must know (or derive) several different parameters:
Length of test Number of samples Goal of test Confidence desired Accuracy desired Cost Acceleration Factor o o o
Field Environment Test Environment Acceleration Factor Calculation
Slope of Weibull Distribution (Beta factor) Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 24
RDT and ALT Parameters z
Keys to a Good RDT/ALT Accelerants must be valid Acceleration factors must be measurable Failure region must be well understood o
(infant mortality, steady state, wearout)
Note that the lead-free transition will change all our old acceleration models.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 25
RDT and ALT Parameters If we don’t have a good formula for acceleration factor, we can determine through experimentation. z When wearout is a dominant failure mode, we cannot substitute units for time. z If we cannot find environmental or electrical accelerants, we can resort to duty cycle acceleration. z
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 26
Reliability Improvement Techniques FMEA in the Design Phase z HALT in the Prototype Phase z
FMEA
HALT Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 27
Failure Modes Effects Analysis (FMEA) z
FMEA is a systematic technique to analyze a system for all potential failure modes.
z
FMEA’s are used to identify highest risk items and mitigate them to reduce overall product risk.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 28
FMEA Integrated with HALT z
When planning a HALT, FMEA’s can be used for: identifying failure modes that HALT is likely to uncover. identifying failure modes that require extra planning to find. identifying non-relevant failure modes. Identifying wearout mechanisms that HALT will not be able to find. helping to decide on the number of samples. Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 29
Highly Accelerated Life Testing (HALT) z
HALT is a method of applying progressively higher levels of environmental, electrical, and mechanical stresses to a product to the point of failure in order to assess and improve design robustness and margin above its intended operation.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 30
Highly Accelerated Life Testing (HALT) z
HALT is used to:
Quickly discover design issues. Evaluate and improve design margins. Release mature product at market release. Reduce development time and cost. Evaluate cost reductions made to product.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 31
Highly Accelerated Life Testing (HALT) HALT, How It Works
Mike Silverman, Ops A La Carte
Track 1
is
Im pr ov e
ilu re
An aly s
e)
S (in tre cr s ea s s
Fa
Session 7
Slide Number: 32
Highly Accelerated Life Testing (HALT) Lower Destruct Limit
Lower Oper. Limit
Product Operational Specs
Upper Oper. Limit
Upper Destruct Limit
Destruct Margin Operating Margin
Stress
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 33
Highly Accelerated Life Testing (HALT) HALT, Why It Works Classic S-N Diagram (stress vs. number of cycles)
S2
S0= Normal Stress conditions N0= Projected Normal Life
Stress
S1 S0
N2
N1 N0 Number of Cycles
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 34
Highly Accelerated Life Testing (HALT) Limitations of HALT Classic S-N Diagram (stress vs. number of cycles) Point at which failures become non-relevant
S2
S0= Normal Stress conditions N0= Projected Normal Life
Stress
S1 S0
N2
N1 N0 Number of Cycles
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 35
Reliability Integration Between HALT and RDT/ALT Goal Plan FMEA HALT
Prediction RDT/ALT
Verification Review Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 36
Reliability Integration Between HALT and RDT/ALT z
Often times we will run a product through HALT and then run the subassemblies through ALT that were not good candidates for HALT. HALT on System
Mike Silverman, Ops A La Carte
ALT on System Fan
Track 1
Session 7
Slide Number: 37
Reliability Integration Between HALT and RDT/ALT z
And at other times, we may develop an RDT based on HALT limits, using the same accelerants but lowering the acceleration factors to measurable levels. HALT on System
Mike Silverman, Ops A La Carte
RDT on System
Track 1
Session 7
Slide Number: 38
Reliability Integration Back to Goals Goal Plan FMEA HALT
Prediction RDT/ALT
Verification Review Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 39
Reliability Integration Back to Goals z
Field Data will tell us: if our goals were accurate. If our plan was complete. If our individual tools were effective.
z
If not, make adjustments for next program.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 40
Wrap Up A successful product reliability program is the collection of goals, plans, tools, and skills focused on the cost-effective and timely creation of products that meet business objectives. z To minimize total Life Cycle Costs, we: z
create clear goals. choose the best tools. properly integrate the tools forwards and back.
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 41
Questions
WHAT ARE YOUR QUESTIONS?
Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 42
Presenter’s Biographical Sketch z
Mike Silverman, (408) 472-3889,
[email protected]
z
Mike is founder and managing partner at Ops A La Carte, a Professional Business Operations Company that offers a broad array of expert services in support of new product development and production initiatives. The primary set of services currently being offered are in the area of reliability. Through Ops A La Carte, Mike has had extensive experience as a consultant to hightech companies, and has consulted for over 200 companies including Cisco, Ciena, Apple, Siemens, Intuitive Surgical, Abbott Labs, and Applied Materials. He has consulted in a variety of different industries including telecommunications, networking, medical, semiconductor, semiconductor equipment, consumer electronics, and defense electronics.
z
Mike has 20 years of reliability, quality, and compliance experience, the majority in start-up companies. He is also an expert in accelerated reliability techniques, including HALT and HASS. He set up and ran an accelerated reliability test lab for 5 years, testing over 300 products for 100 companies in 40 different industries. Mike has authored and published 7 papers on reliability techniques and has presented these around the world including China, Germany, and Canada. He has also developed and currently teaches 8 courses on reliability techniques.
z
Mike has a BS degree in Electrical and Computer Engineering from the University of Colorado at Boulder, and is both a Certified Reliability Engineer and a course instructor through the American Society for Quality (ASQ), IEEE, Effective Training Associates, and Hobbs Engineering. Mike is a member of ASQ, IEEE, SME, ASME, PATCA, and IEEE Consulting Society and currently the IEEE Reliability Society Santa Clara Valley Chapter President. Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 43
Presenter’s Biographical Sketch z
Fred Schenkelberg, (408) 710-8248,
[email protected]
z
Fred Schenkelberg is a Senior Reliability Engineering Consultant at Ops A La Carte. He is currently working with clients using reliability assessments as a starting point to develop and execute detailed reliability plans and programs. Also, he exercises his reliability engineering and statistical knowledge to design and conduct accelerated life tests.
z
Fred joined HP in February 1996 in Vancouver, WA. He joined ESTC, Palo Alto, CA., in January 1998 and co-founded the HP Product Reliability Team. He was responsible for the community building, consulting and training aspects of the Product Reliability Program. He was also responsible for research and development on selected product reliability management topics.
z
Prior to joining ESTC, he worked as a design for manufacturing engineer on DeskJet printers. Before HP he worked with Raychem Corporation in various positions, including research and development of accelerated life testing of polymer based heating cables.
z
He has a Bachelors of Science in Physics from the United States Military Academy and a Masters of Science in Statistics from Stanford University. Fred is an active member of the RAMS Management Committee and currently the IEEE Reliability Society Santa Clara Valley Chapter Vice President. Mike Silverman, Ops A La Carte
Track 1
Session 7
Slide Number: 44