Beruflich Dokumente
Kultur Dokumente
APPLIED
RELIABILITY
Techniques for Reliability
Analysis
with
Applied Reliability Tools (ART)
(an EXCEL Add-In)
and
JMP Software
STAT-TECH
Spring 2010
Applied Reliability Page 2
Accelerated Testing
(continued from Class 4 Notes)
Accelerated Test Example (Analysis in JMP)
Degradation Modeling
Sample Sizes for Accelerated Testing
System Models
Series System
Parallel System
Analysis of Complex Systems
Standby Redundancy
Defective Subpopulations
Graphical Analysis
Mortals and Immortals
Models
Case Study
Class Project Example
Modeling the Field Reliability
Evolution of Methods
General Reliability Model
AMD Example
Applied Reliability Page 3
System Models
Series System
Rs ( t ) R1 t R2 t ... Rn t
which we denote with the capital pi symbol for
multiplication
n
Rs t Ri t
i 1
Fs t 1 1 Fi t
n
i 1
System Models
Parallel System
n
Fs t Fi t
i 1
Rs t 1 1 Ri t
n
i 1
SOLUTION
The CDF for the two components in parallel is F2(t)
and the PDF, by differentiation, is 2F(t)f(t). The
failure rate of the system is
hs t s
f t
1 Fs t
2 F t f t
1 F 2 t
2 F t f t
1 F t 1 F t
2 F t
h t
1 F t
Class Project
System Models
A B
B
Applied Reliability Page 8
Example of Series-Parallel
System: Big Rig
G C A
H D
Trailer Cab
I E
J F B
I G E C
B A
J H F D
Class Project
Complex Systems
0.012
0.01
0.008
0.006
Single
Parallel
0.004
Standby
0.002
0
0 50 100 150 200 250 300 350 400
Applied Reliability Page 11
Reliability Experiment
Consider . . .
Defect Models
Mortals versus Immortals
Example of a Defective
Subpopulation
A Processing Problem
Spotting a Defective
Subpopulation
Graphical Analysis
Assume that a specified failure mode follows a
lognormal distribution.
Defective Subpopulations
Graphical Analysis
Plot based on total sample (mortals and immortals).
Defect Model
Mortals and Immortals
Fobs(t) = p Fm(t)
where Fm(t) is the CDF of the mortals and p is the
fraction of mortals (units with the fatal defect) in the
total sample size.
Major Computer
Manufacturer Reliability Data
Gate Oxide Fails
What Do These
Numbers Mean?
Modeling with
Defective Subpopulations
Defective Subpopulation
Models
If we dont consider mortals vs. immortals, we will
incorrectly assume that all units can fail.
Statistical Reliability
Analysis and Modeling:
A Case Study
Reliability Study
Background
Reliability Study
Design
Purpose of Study
Reliability Modeling
Modeling Procedure
Statistical Analysis Plan
Reliability Study
Bake Recoverable Failures
L i n e a r P l o t o f C u m u l a ti v e F a i l ur e s V e r su s T i m e
80%
70%
60%
Cum ula tive P e r c e nt
50%
40%
30%
20%
1 50oC 1 25oC
10%
0%
0 20 40 60 80 100 120
S tre s s Ti m e (P ow e r on H our s )
S am ple S iz e s : 1 5 0 oC = 9 0 ; 1 2 5o C = 1 7 9
Applied Reliability Page 28
Reliability Study
Bake Recoverable Failures
P r o b a b i li t y Pl o t s ( N o A d ju s tm e n t fo r M o r t a ls)
0 .5
0 1 2 3 4 5
S ta ndar d Norm a l V a ri a te : Z
- 0 .5
-1
- 1 .5
150oC 125oC
-2
- 2 .5
Ln (Tim e t o Fa il ure )
S a m pl e S iz e s : 1 5 0 o C = 9 0 ; 1 2 5 o C = 1 7 9
Applied Reliability Page 29
Reliability Study
Bake Recoverable Failures
2.5
1.5
1
S ta n d a rd N o rm a l V a ria te : Z
0.5
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
-0.5
150oC 125oC
-1
-1.5
-2
L n (T im e to F a ilu re )
NUM. NUM.
CELL ON TEST FA IL T50 LOW T50 UP SIGMA LOW SIGMA UP
THE AS SUMPTION O F QUAL SIGMAS CAN NOT BE REJECTED AT THE 95 PERCE NT LEVEL.
UNDER THIS A SSUMP TION, RESULTS LIK E O BSERVED OCCUR AB OUT 41.9 PERCE NT OF THE TIME.
(THE S MA LLER THIS PE RCENT, THE LESS LIKEL Y THE ASSUMPTION.)
MAXIMUM LIKELIHOOD ESTIMATES
NUM. NUM.
CELL ON TEST FA IL T50 LOW T50 UP SIGMA LOW SIGMA UP
Reliability Study
Bake Recoverable Failures
M od el Fit to Ac tua l
80%
70%
60%
C um um ative Pe r ce nt Fa ilur e s
50%
40%
30%
1 50oC
1 25oC
20%
M LE F it: 1 50oC
M LE F it: 1 25oC
10%
0%
0 20 40 60 80 1 00 1 20 1 40
Tim e (P o w er o n Ho ur s)
Applied Reliability Page 32
P r o je ct e d F ie ld F a llo u t w it h V a rio u s M o rt a l
P e r c en t a g es
2 0%
1 8%
5%
1 6%
1 0%
1 4% 2 0%
C u m u lat iv e P e r ce n t
3 0%
1 2%
4 0%
1 0% 5 0%
6 6%
8%
6%
4%
2%
0%
0 2 4 6 8 10
T im e i n F ie ld ( K H o u r s )
Applied Reliability Page 34
A Note of Caution
Analysis When Mortals Are Present
A Side Benefit
Screening a Wearout Mechanism
Class Project
Defect Models
50 components are put on stress. Readouts are at
10, 25, 50, 100, 200, 500, and 1,000 hours. The
failure counts at the respective readouts are 2, 2,
4, 5, 4, 3, and 0.
Note: Percent Failure scale on Weibull Probability paper is faint. Values are 99.9, 98.0, 90.0,
70.0, 50.0, 30.0, 20.0, 10.0, 5.0, 2.0, 1.0, 0.5, 0.2, 0.1, etc.
Applied Reliability Page 39
Class Project
Defect Model Estimates
t / cm
F (t ) 1 e
How could we confirm that the Weibull model for
the mortal population fits the data? We estimate
the CDF at three times and compare to
observations.
Mortal
CDF Model Empirical
(Weibull Mortal CDF for CDF All
Time Model) Fraction All Units Units
25 0.221 0.4
100 0.632 0.4
1000 1.000 0.4
Applied Reliability Page 40
Defective Subpopulations in
ART
Enter failure information (readout times, cumulative
failures) into columns. Under ART, select Defective
Subpopulations Enter required information. Click OK.
Applied Reliability Page 41
System Models
Assumptions
Constant failure rate
Single overall activation
energy
Ambient temperatures
No separation of failure modes
Applied Reliability Page 43
Primitive Method
Problems with Calculations
Example
Primitive Method
Comparative Calculation
Later Method
Problems
An Alternative Model
Dead
Defective
Deficient
Applied Reliability Page 48
Quality issue
Defective Subpopulations
Competing Risks
FT Fe Fd 1 FN
where
FN = 1 - R1R2. . . RN
Appendix
Applied Reliability Page 55
Class Project
System Models
Class Project
Complex Systems
C
G H
Class Project
Defect Models
1. Estimate the proportion defective p and the
number of mortals in the sample. Fill in the mortal
CDF column in the table below.
Cum # CDF Est All CDF Est
Time Fails Units (%) Mortals (%)
10 2 2/50 = 4%
25 4 4/50 = 8%
50 8 8/50 = 16%
100 13 13/50 = 26%
200 17 17/50 = 34%
500 20 20/50 = 40%
1000 20 20/50 = 40%
Class Project
Defect Model Example
n = 50
Cum # CDF Est All CDF Est
Time Fails Units (%) Mortals (%)
10 2 2/50 = 4% 2/20 = 10%
25 4 4/50 = 8% 4/20 = 20%
50 8 8/50 = 16% 8/20 = 40%
100 13 13/50 = 26% 13/20 = 65%
200 17 17/50 = 34% 17/20 = 85%
500 20 20/50 = 40% 20/20 = 100%
1000 20 20/50 = 40% 20/20 = 100%
Class Project
Defect Model Example
Model Check
t / cm
F (t ) 1 e
Mortal
CDF Model Empirical
(Weibull Mortal CDF for CDF All
Time Model) Fraction All Units Units
25 0.221 0.4 0.088 0.08
100 0.632 0.4 0.253 0.26
1000 1.000 0.4 0.400 0.40
Applied Reliability Page 61
Class Project
Defect Model
p x Weibull CDF Plot
0.9
0.8
0.7
0.6
CDF
0.5
0.4
0.3
0.2
0.1
0
0 100 200 300 400 500 600 700 800 900 1000
Times (Hrs)