halt and reliability workshop · halt and reliability workshop halt and reliability workshop elite...

HALT and Reliability Workshop

HALT and Reliability Workshop

Elite Electronic Engineering Steve Laya

630-495-9770 x 119, [email protected]

Topics Covered

Reliability and Planning

Overview of Reliability Concepts‐ Distributions and life estimation

Reliability metrics: MTBF, failure rate, R (Reliability) and C (Confidence)

Accelerated Testing‐ Uses and cautions; Models for temperature and humidity

Accelerated vibration models; Other accelerated tests

Vibration Techniques‐ Electro‐dynamic, Repetitive Shock, Servo‐hydraulic

Characteristics of vibration produced, relative damage potential, recommended use.

How, When, and Why to use HALT and Accelerated Testing

How Do You Define Reliability?

“…the ability of a system or component to perform its required functions under

stated conditions for a specified period of time”

The probability of success

The capability to perform as designed

Reliability, Availability, Maintainability (RAM) , Safety, Testability

Number of failures over a period of time MTBF, MTTF, Failure Rate, Hazard Rate

Mathematical definition

Where h(t) is the hazard function or hazard rate

How Do You Evaluate Reliability?

Statistics

Probability Theory

Reliability Theory

Hazard Analysis

FEMA

FTA

Reliability Handbook Prediction

Weibull Analysis

Accelerated Life Testing

Maximum Likelihood Estimates

Markov Analysis

Physics of Failure

Design Review

Sneak Circuit Analysis

Reliability Demonstration /Growth

HALT/HASS

Which tool to use?

Which tool

to use?

Managing Reliability‐ The Core Elements of a Reliability Program.

1 Understand

Customer

Requirements

Environment

Duty Cycle

Reliability Goals

2 Feedback from Similar Components

FRACA- Test Failures, Production Failures, Field

Failures

Third Party Assessments- J.D. Powers &

Associates

Warranty Returns- Return Rates, Feedback from

Customers and Technicians

Development Testing 4 Intelligent Design

Use Design Guides

Incorporate lessons learned from previous work

Parameter Design- Choose design variable levels

to minimize effects of uncontrolled variables

Tolerance Design- Scientifically determine correct

drawing specifications

Schedule Periodic Design Reviews

Design with Information from Development

Activities

Sneak Circuit Analysis, HALT, Step-Stress to

Failure, Worst-Case Tolerance Analysis

5 Concept Validation & Design Validation

Early Design Phase Engineering Development Tests

Independent Verification Test (outside of

engineering)

Specify From List of Validated Subsystems &

Components

System Simulation

6 Manufacturing Production parts validation

Qualify production process with Cpk = 1.67

Ensure compliance with SPC program

100% sampling 1st week of production reduce

as necessary

Develop control plans for each drawing

Evaluate measurement error for in process

measurements

Qualify storage, transportation, and installation

systems

Ref: Accelerated Testing, Dodson & Schwab

3 Begin the FMEA

Update throughout the

process

7 Change Control Qualify all any changes in engineering,

production, or supply base.

Lets’ do some

testing!

Testing for Reliability

1. Customer Specified Requirements

Auto/Truck OEM, RTCA DO-160, MIL-810 (Qualitative)

2. Identify and Design-out Latent Defects

HALT and other early short duration tests (Qualitative)

3. Comparison of Products

Competitive or new products (Qualitative)

4. Estimation of Reliability Parameters (underlying distribution, point estimate, confidence

MTBF, MTTF (Quantitative)

Reliability & Confidence (Quantitative) (graphically or statistically)

5. Contractual Compliance to a Specific Metric

Reliability Demonstration (Quantitative)

Reliability Growth (Quantitative)

Estimation of Reliability Parameters

Specify the test Define Test Objective

Determine the MTTF and failure rate for…

Lab Test or Field Test Lab Test

Evaluate failure modes, failure mechanisms High impedance fault due to Electromigration

Specify Environments, DUT Configuration, Failure Criteria Temperature and Current

Assign Test Durations. Apply Acceleration Factors Arrhenius Model, Power Law Mode

Run Test and Collect Data

Record times to failure- Plot a Histogram

Fit the histogram shape to a failure distribution

Estimate distribution characteristics of interest by “parametric approach” Parametric means related to a distribution


Collect the data and plot a Histogram • Divide x-axis into intervals

• Count the number of failures occurring in each interval

• Scale the y-axis for the maximum number of counts

• Fit a curve to the plot

• Ex. Start with 100 samples on a powered elevated

temperature life test. Count remaining units at each interval.

Interval

(Hour)

Remaining

Units Failures

Cumul

Failure

1 90 10 10

2 81 9 19

3 73 8 27

4 66 7 34

5 59 7 41

6 53 6 47

7 48 5 52

8 43 5 57

9 39 4 61

10 35 4 65

11 31 3 69

12 28 3 72

13 25 3 75

14 23 3 77

15 21 2 79

16 19 2 81

17 17 2 83

18 15 2 85

19 14 2 86

20 12 1 88

21 11 1 89

22 10 1 90

23 9 1 91

24 8 1 92

25 7 1 93

26 6 1 94

27 6 1 94

28 5 1 95

29 5 1 95

30 4 0 96

31 4 0 96

Estimation of Reliability Parameters Plot the Data

• Create a Relative Frequency Plot

• Relative Frequency = Class Count

Total • Evaluate the shape of the distribution

IntervalClass Count

per TotalPercentage

1 10/100 0.10

2 9/100 0.09

10 4/100 0.04

17 2/100 0.02


Probability Density Function (PDF)

• Relative likelihood for the variable to take on a given value.

• The probability density function is nonnegative everywhere,

and its integral over the entire space is equal to one

• Ex1:

• N=100

• = 0.1

• Evaluate at 10 hours

• f(10)= 0.036

• Ex2:

• N=100

• = 0.1


• f(20)= 0.0135

Probability Density Functions (PDFs) to

Cumulative Distribution Functions (CDFs)

• Sum the area beneath the PDF

• CDF provides a probability of failure relative to x-

axis (time, cycles, life)

• The compliment of the CDF is the Reliability

Function.

• Reliability (x) = 1-CDF(x)

Reliability

Expression for

Exponential

Distribution Where

= failure rate

1/ = MTTF

Evaluate the Reliability Function

• Ex1:

• N=100

• = 0.1


• R(10)= 0.36

• Ex2:

• N=100

• = 0.1


• R(20)= 0. 135

Customer Provided Reliability Metrics

• = 1/MTTF

• Failure rate = 0.1

• MTTF = 10

Examples

Reliability Point Estimates and Confidence

• Calculate Confidence Intervals for Different Distributions

• Range of values bounded above and below within

which the true value is expected to fall.

• Measures the statistical precision of the estimate

• 90% confidence interval should contain the estimate

90% of the time

• Determine the interval within which the true

parametric values lies with a given probability for a

given sample size

• Determine the sample size required to assure with a

specified probability that the true parametric value

lies within a specific interval.

Exponential: Chi-Squared distribution

Normal: t-distribution

Weibull: See referenced resources…

Reliability Point Estimates and Confidence

Calculate Confidence Intervals

Exponential requires Chi-Squared Distribution

(Normal requires t-Distribution)

t* = time at which the life test is terminated

r = number of failures accumulated at time t*

T = total test time

a = acceptable risk of error

1 - a = confidence level

T = total test time

Note: Ref: 2

MTTF= 216hrs

115hrs 459hrs

Point Estimate for MTTF with Confidence Intervals

Example: Calculate MTTF with Confidence Intervals for Fixed

Truncation Time on 100 units, C=90%

n = number of items placed on test at time t = 0



r* = preassigned number of failures



T = total test time

Fixed Truncated Test

Lower One

Sided

Confidence

Bound

Two Sided

Confidence

Bound

= 1-CL

= 1-0.9 = 0.1

/2 = 0.05

r = 96

2r+2 = 194

2T c2(0.1, 194)

2T c2(0.05, 194)

2T c2(0.95, 192)

Lower One Sided Confidence Bound

Two Sided

Confidence

Bound

MTTF = 866 =9.02 hrs Total Test Time

96 Number of Failures

1732

219.633 (7.89, )

1732

227.496

1732

160.944

(7.61, 10.76)

Interval

(Hour)

Remain

Units Failures

Cumul

Failure

Cumul

Time

1 90 10 10 90

2 81 9 19 171

3 73 8 27 244

4 66 7 34 310

5 59 7 41 369

6 53 6 47 422

7 48 5 52 470

8 43 5 57 513

9 39 4 61 551

10 35 4 65 586

11 31 3 69 618

12 28 3 72 646

13 25 3 75 671

14 23 3 77 694

15 21 2 79 715

16 19 2 81 733

17 17 2 83 750

18 15 2 85 765

19 14 2 86 778

20 12 1 88 791

21 11 1 89 802

22 10 1 90 811

23 9 1 91 820

24 8 1 92 828

25 7 1 93 835

26 6 1 94 842

27 6 1 94 848

28 5 1 95 853

29 5 1 95 858

30 4 0 96 862

31 4 0 96 866

2T 1732

Chi Square Distribution Table

160.944 =CHIINV(0.95,20)

DF 0.995 0.975 0.2 0.1 0.05 0.025 0.02 0.01 0.005 0.002 0.001

190 143.545 153.721 206.182 215.371 223.16 230.064 232.146 238.266 243.959 250.977 255.976

191 144.413 154.621 207.225 216.437 224.245 231.165 233.251 239.386 245.091 252.124 257.135

192 145.282 155.521 208.268 217.502 225.329 232.265 234.356 240.505 246.223 253.271 258.292

193 146.15 156.421 209.311 218.568 226.413 233.365 235.461 241.623 247.354 254.418 259.45

194 147.02 157.321 210.354 219.633 227.496 234.465 236.566 242.742 248.485 255.564 260.607

195 147.889 158.221 211.397 220.698 228.58 235.564 237.67 243.86 249.616 256.71 261.763

196 148.759 159.122 212.439 221.763 229.663 236.664 238.774 244.977 250.746 257.855 262.92

197 149.629 160.023 213.482 222.828 230.746 237.763 239.877 246.095 251.876 259.001 264.075

198 150.499 160.925 214.524 223.892 231.829 238.861 240.981 247.212 253.006 260.145 265.231

199 151.37 161.826 215.567 224.957 232.912 239.96 242.084 248.329 254.135 261.29 266.386

200 152.241 162.728 216.609 226.021 233.994 241.058 243.187 249.445 255.264 262.434 267.541

201 153.112 163.63 217.651 227.085 235.077 242.156 244.29 250.561 256.393 263.578 268.695

202 153.984 164.532 218.693 228.149 236.159 243.254 245.392 251.677 257.521 264.721 269.849

203 154.856 165.435 219.735 229.213 237.24 244.351 246.494 252.793 258.649 265.864 271.002

204 155.728 166.338 220.777 230.276 238.322 245.448 247.596 253.908 259.777 267.007 272.155

205 156.601 167.241 221.818 231.34 239.403 246.545 248.698 255.023 260.904 268.149 273.308

P

Point Estimates for Reliability at Specified Time

with Confidence Intervals

• N=100

• = 0.1


• R(10)= 0.36

Example

Fixed Truncated Test

Lower One

Sided

Confidence

Bound

Two Sided

Confidence

Bound7.61 10.758

R(10) =

R(10) = 0.268 0.395

2(866) c

2(0.95,192)

2(866) c

2(0.05,194)

2(866) 227.496

2(866) 160.994

Reliability

Expression for

Exponential

Distribution

Evaluate at 10 hours

R(10)= 0.36

2-sided 90% Confidence Intervals

Point Estimate for MTTF with Confidence Intervals

Example: Calculate MTTF with Confidence Intervals for Fixed

Number of Failures on 10 units, C=90%

n = number of items placed on test at time t = 0



r* = preassigned number of failures



T = total test time

= 1-CL

= 1-0.9 = 0.1

/2 = 0.05

r = 10

2r = 20

2T c2(0.1, 20)

2T c2(0.05, 20)

2T c2(0.95, 20)

Lower One Sided Confidence Bound

Two Sided

Confidence

Bound

MTTF = 961 =96.1 hrs Total Test Time

10 Number of Failures

Failure

Number

Operating

Time (Hrs)

1 8

2 20

3 34

4 46

5 63

6 86

7 111

8 141

9 186

10 266

Total 961

2T 1922

1922

28.412 (67.6, )

1922

31.41

1922

10.851

(61.19, 177.12)

Fixed Number of Failures

Lower One

Sided

Confidence

Interval

Two Sided

Confidence

Interval

Chi Square Distribution Table

10.85081 =CHIINV(0.95,20)

DF 0.995 0.975 0.2 0.1 0.05 0.025 0.02 0.01 0.005 0.002 0.001

1 3.93E-05 0.000982 1.642 2.706 3.841 5.024 5.412 6.635 7.879 9.55 10.828

2 0.01 0.0506 3.219 4.605 5.991 7.378 7.824 9.21 10.597 12.429 13.816

3 0.0717 0.216 4.642 6.251 7.815 9.348 9.837 11.345 12.838 14.796 16.266

4 0.207 0.484 5.989 7.779 9.488 11.143 11.668 13.277 14.86 16.924 18.467

5 0.412 0.831 7.289 9.236 11.07 12.833 13.388 15.086 16.75 18.907 20.515

6 0.676 1.237 8.558 10.645 12.592 14.449 15.033 16.812 18.548 20.791 22.458

7 0.989 1.69 9.803 12.017 14.067 16.013 16.622 18.475 20.278 22.601 24.322

8 1.344 2.18 11.03 13.362 15.507 17.535 18.168 20.09 21.955 24.352 26.124

9 1.735 2.7 12.242 14.684 16.919 19.023 19.679 21.666 23.589 26.056 27.877

10 2.156 3.247 13.442 15.987 18.307 20.483 21.161 23.209 25.188 27.722 29.588

11 2.603 3.816 14.631 17.275 19.675 21.92 22.618 24.725 26.757 29.354 31.264

12 3.074 4.404 15.812 18.549 21.026 23.337 24.054 26.217 28.3 30.957 32.909

13 3.565 5.009 16.985 19.812 22.362 24.736 25.472 27.688 29.819 32.535 34.528

14 4.075 5.629 18.151 21.064 23.685 26.119 26.873 29.141 31.319 34.091 36.123

15 4.601 6.262 19.311 22.307 24.996 27.488 28.259 30.578 32.801 35.628 37.697

16 5.142 6.908 20.465 23.542 26.296 28.845 29.633 32 34.267 37.146 39.252

17 5.697 7.564 21.615 24.769 27.587 30.191 30.995 33.409 35.718 38.648 40.79

18 6.265 8.231 22.76 25.989 28.869 31.526 32.346 34.805 37.156 40.136 42.312

19 6.844 8.907 23.9 27.204 30.144 32.852 33.687 36.191 38.582 41.61 43.82

20 7.434 9.591 25.038 28.412 31.41 34.17 35.02 37.566 39.997 43.072 45.315

21 8.034 10.283 26.171 29.615 32.671 35.479 36.343 38.932 41.401 44.522 46.797

22 8.643 10.982 27.301 30.813 33.924 36.781 37.659 40.289 42.796 45.962 48.268

23 9.26 11.689 28.429 32.007 35.172 38.076 38.968 41.638 44.181 47.391 49.728

24 9.886 12.401 29.553 33.196 36.415 39.364 40.27 42.98 45.559 48.812 51.179

25 10.52 13.12 30.675 34.382 37.652 40.646 41.566 44.314 46.928 50.223 52.62

P

Procedure for Calculating Point Estimates

and Confidence Intervals

IEC 60505-4 Statistical Procedures for

Exponential Distribution-

Point Estimates,

Confidence Intervals,

Prediction Intervals and

Tolerance Intervals

IEC Tools For Reliability Assessment

IEC 60300-3-5 Reliability Test Conditions and

Statistical Test Principles

IEC 11453 Point Estimate and Confidence

Intervals for the Binominal

Distribution

IEC 60605-4 Point Estimate and Confidence

Intervals for the Exponential

Distribution

IEC 11453

IEC 60605-4

IEC 61649 Point Estimate and Confidence

Intervals for the Weibull Distribution

IEC 61649 IEC 61164

IEC 60605-6

Important Distributions

Exponential • Constant Failure Rate

• Mixed Failure Modes

• Most Electronics

• Mean Life R(t)= 0.368

Normal • Wear-out

• Greater than 20 samples

• Mean Life R(t)= 0.5

Weibull • Can model a variety of

different data types

• Infant mortality, constant

failure rate, or wear-out.

• Good for limited samples

Weibull Analysis

Method for representing and interpreting data

Provides a Reliability metric directly from plot

Works well with small samples – life data

(failures) provide more information

(shape or slope), (characteristic life or scale),

(location or offset)

< 1 indicates infant mortality

= 1 indicates random failures

> 1 indicates wear out failures

Weibull Analysis

1. Acquire accurate time to failure data

2. Rank the data first failure to last

3. Plot the data on Weibull paper

4. Interpret the plot 1. Look for mixed modes

2. Measure slope to determine

3. Determine characteristic life

4. Read R(t)

Failure

Number

Operating

Time

(Hours)

Median

Rank 10

Samples

1 8 6.70

2 20 16.23

3 34 25.86

4 46 35.51

5 63 45.17

6 86 54.83

7 111 64.49

8 141 74.14

9 186 83.77

10 266 93.30

Rank Order 1 2 3 4 5 6 7 8 9 10

1 50.00 29.29 20.63 15.91 12.94 10.91 9.43 8.30 7.41 6.70

2 70.71 50.00 38.57 31.38 26.44 22.85 20.11 17.96 16.23

3 79.37 61.43 50.00 42.14 36.41 32.05 28.62 25.86

4 84.09 68.62 57.86 50.00 44.02 39.31 35.51

5 87.06 73.56 63.59 55.98 50.00 45.17

6 89.09 77.15 67.95 60.69 54.83

7 90.57 79.89 71.38 64.49

8 91.70 82.04 74.14

9 92.59 83.77

10 93.30

Sample Size

Median Rank Table

Median Rank Estimate

MR = (i-0.3) *100

(N+0.4)

i= rank order #

N=sample size

MTTF vs. MTBF

MTTF- Mean Time To Failure

Expected time to fail for a non-repairable system

Non-repairable systems can fail only once.

MTTF is equivalent to the mean of its failure time distribution.

Ex 16+12+14+6+8 = 56/5= 11.2

MTBF- Mean Time Between Failure

Expected time to fail for repairable systems

Expected time between two consecutive failures for a repairable system

MTBF= MTTF +MTTR

16hrs 12hrs 14hrs 6hrs 8hrs

Re

pai

r/R

est

ore

Re

pai

r/R

est

ore

Re

pai

r/R

est

ore

Re

pai

r/R

est

ore

Re

pai

r/R

est

ore

Re

pai

r/R

est

ore

Operational

Non-Operational

Time Between Failures

Repair Time

12hrs

16hrs

6hrs

14hrs

8hrs

StartFailure

100

20

Interval

(Hour)

Remaining

Units

(Exp)

Failures

(Exp)

Cumul

Failure

(Exp)

Cumul

Time

(Exp)

Failure

Rate

(Exp)

Rel Freq

(Exp)

1 80 20 20 80 0.20 0.20

2 64 16 36 144 0.20 0.16

3 51 13 49 195 0.20 0.13

4 41 10 59 236 0.20 0.10

5 33 8 67 269 0.20 0.08

6 26 7 74 295 0.20 0.07

7 21 5 79 316 0.20 0.05

8 17 4 83 333 0.20 0.04

9 13 3 87 346 0.20 0.03

10 11 3 89 357 0.20 0.03

11 9 2 91 366 0.20 0.02

12 7 2 93 373 0.20 0.02

13 5 1 95 378 0.20 0.01

14 4 1 96 382 0.20 0.01

15 4 1 96 386 0.20 0.01

16 3 1 97 389 0.20 0.01

17 2 1 98 391 0.20 0.01

18 2 0 98 393 0.20 0.00

19 1 0 99 394 0.20 0.00

20 1 0 99 395 0.20 0.00

21 1 0 99 396 0.20 0.00

22 1 0 99 397 0.20 0.00

23 1 0 99 398 0.20 0.00

24 0 0 100 398 0.20 0.00

25 0 0 100 398 0.20 0.00

26 0 0 100 399 0.20 0.00

27 0 0 100 399 0.20 0.00

28 0 0 100 399 0.20 0.00

29 0 0 100 399 0.20 0.00

30 0 0 100 400 0.20 0.00

31 0 0 100 400 0.20 0.00

Total 1.0

400 Lower Single-Sided Confidence Limit at 90%

100

0

90

0.1

199.80 4.0

100

227.09

235.08

167.36

Total Qty At Start of Test

% Failure Rate (Exponential)

4.8

Total Accumulated Test Time (T)

Number of Failures

Number of Suspensions

Confidence Limit

Alpha

2r

Lower 2-Sided Confidence Limit at

90 %

Upper 2-Sided Confidence Limit at

90%MTTF

r

Lower Single-Sided

Lower 2-Sided

Upper 2-Sided

3.5

3.4

0

10

20

30

40

50

60

70

80

90

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Co

un

ts

Hours

Exponential

y = 0.25e-0.223x

0.00

0.05

0.10

0.15

0.20

0.25

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Re

lati

ve F

req

ue

ncy

Hours

Exponential (Relative Frequency)

y = 0.2e-0.2x

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0 10 20 30

Pro

bab

ility

De

nsi

ty

Hours

Exponential PDF

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20 25 30

Cu

mu

lati

ve D

en

sity

Fu

nct

ion

Re

liab

ility

Fu

nct

ion

Exponential CDF & Reliability Function

Reliability Bathtub Curve

Exponential Distribution

= failure rate = constant

Weibull Distribution

< 1 indicates infant mortality

= 1 indicates random failures

> 1 indicates wear out failures

System Reliability

Rsystem = Rsubsystem1 *Rsubsystem2 *Rsubsystem3… *Ri

Block Diagrams

Series

Parallel-Series

.97

.93 .98 .95 R System = 0.84

.97 .93

.93

.98 .95 R System = 0.90

Success Run Test to Establish R & C

Success Run Test, Test to a Bogey

Based on a Binomial Distribution

Test Results are Either Success or Failure

Prove a Target Reliability with an assigned Confidence Level

Don’t care about continuous measurement or calculating a parametric value, ie

MTTF or failure rate

Define the test conditions to represent 1 or more lives

Operate without failure for a specified time

Reliability, Confidence, and Sample Size related by Success Run Formula N= Sample Size, R= Reliability, C= Confidence Level


Calculate required number of samples based on R and C

Example

R= 97%, C= 50%

Example

R= 97.7%, C= 90%

Binomial Distribution

Nomograph for

R/C/N

HALT/HASS and Accelerated Testing


Define “One Life”

R = 97.7%, C = 90%, N=100

8 hours/day x 365 days/year x 5 years = 14600hrs

Test 24 hours/day 14600/24 = 608 hrs on test

Apply Arrhenius model for

Temperature Acceleration TAF= 11.5

Time on Test = 53 hours

Ea= 0.8eV = 1.28 x 10-19 J

k = 1.38 x 10-23J/K

Tmax = +50C

Ttest= +85C

Time Dependent Failure Mechanisms

Loss of signal Silicon Diffusion Temperature

Power Failure Dielectric Breakdown Electric Field

Loss of signal Electromigration Temperature & Power Cycling

Intermittent Output Corrosion & Oxidation of Fractures Humidity, Voltage, Temperature

Loss of signal Dendrite Growth Humidity, Temperature

Water Intrusion Seal Leaks Pressure

Cracked Solder Joint Fatigue Thermal cycling & vibration

Failure Mechanism Accelerating Factors Failure Mode

Overstress- ESD, Mechanical Shock, Thermal Breakdown

Time Dependent- Fatigue, Wear, Corrosion

Accelerated Stress Testing

Acceleration Factors

o Temperature

o Arrhenius Model

o Humidity

o Lawson Model

o Coffin-Manson

o Vibration

o Power Law and Miner

Criteria m= S-N slope

o Voltage

o Inverse Power Law

o Product Life Cycling o CALT Testing

o Test to Failure & Apply

Weibull Analysis

MIL-HDBK-338 Table 8-7.1

Product Life Cycling

Calibrated Accelerated Life Testing (CALT)

Suggest primary fatigue mechanism

Simulate loads at three stress levels

90% of foolish load (first test)

80% of first test load

Third stress level

Depends on first two and ultimate life

Test all units to failure

Plot S-N curve, Determine AF’s

Generate Weibull Plot

Product Life Cycling

Accelerated Life Testing

•“Accelerated Testing: Statistical

Models, Test Plans, and Data Analysis”

•By Wayne Nelson

•CALT GMW 8758

•Example

Automatic Lubricating System

CALT Test Example

•Simulate loads

at three stress

levels

•Monitor test

counting cycles

to failure

CALT Test Example Stress Cycles To Failure

36 3121

36 1075

36 629

36 9452

31 11386

31 1104

31 6624

31 1577

25 11044

25 15405

25 19257

25 28723

Pump S-N Curve

y = 3050953219559.39x-5.93

100

1000

10000

100000

10 100Applied Stress (PSI)C

ycle

s t

o F

ailu

re

•Collect Failure Data

•Plot and determine Inverse

Power Relationship

•AF = (Saccel/Snormal)b

Determine AF's

Condition

High Stress

Mid Stress

Confirm Stress

Normal Stress

Accel Factor

180

74

21

Stress Value (PSI)

36

31

25

15 N/A

CALT Test Example

Stress Level Test Stress Accel Factor Rank

High (IG) 3121 180 9

High (IG) 1075 180 4

High (IG) 629 180 2

High (PP) 9452 180 12

Medium (PP) 11386 74 11

Medium (IG) 1104 74 1

Medium (PP) 6624 74 8

Medium (IG) 1577 74 3

Confirm (PP) 11044 21 5

Confirm (PP) 15405 21 6

Confirm (PP) 19257 21 7

Confirm (PP) 28723 21 10

Cycles at Normal

Stress

560979

193224

113059

1698933

398246

594009

843189

81757

490540

116785

228397

318585

Median Rank

5.61

13.60

21.67

29.76

37.85

45.95

54.05

62.12

70.24

78.33

86.40

94.39

81757

228397

594009

843189

1698933

Sorted Least to Most (Resort these numbers for each

change to spreadsheet)

318585

398246

490540

560979

113059

116785

193224

Sort and apply median ranks

Generate Weibull Plot


CALT Test Example

Weibull Plot

•Obtain distribution parameters

•Reliability metrics

•B1, B10

•Reliability vs life

Vibration Testing Techniques

Servo-Hydraulic Electro-Dynamic Repetitive Shock

• Frequency Range 3Hz-2,500Hz

• Programmable vibration

characteristics; Sine, Random, Sine-on-

Random, Random-on-Random, Field

Data Replay, Mechanical Shock

• Displacement generally limited to 2-3” p-p

• Single axis motion

• Frequency Range 0.5Hz-300Hz

• Programmable vibration

characteristics; Sine, Random, Sine-

on-Random, Random-on-Random, Field

Data Replay, Mechanical Shock

• Displacement generally up to 12” p-p

• Multi-axis motion from multiple cylinders

• Frequency Range 20Hz-10,00Hz

• Vibration output quasi-random with

limited PSD shaping

• Six-axis simultaneous vibration

• High G peak levels • Displacement generally limited to 0.5”

Electro-Dynamic Vibration Machine

Armature Field coil Field

Current

Thrust

Armature

Center Pole Base

Body

Thermotron armature and

cut-away illustration here???

Vibration Time & Frequency Domain

G2/Hz

0.1

Hz

G pk

2

Hz50

1

2

Hz100

1G pk

Power

Spectral

Density

(PSD)

Random Vibration

Probability Density Function

o 2 3 -2 - -3 -4 -5 4 5

1 accelerations occur 68% of the time



>3 accelerations occur less than 1% of the time

= Grms

Random Vibration

Power Spectral Density Plots

Which is the more severe test?

G2/HzG2/Hz

G2/Hz

0.1 0.1 0.1

0.2 0.2 0.2 0.2 0.20.2 0.2 0.2 0.2

Hz Hz Hz

Power Spectral Density Plots

Vibration Testing

Vibration fatigue failures are caused by stress reversals

Vibration at resonance amplifies damage

High accelerations generate proportional Displacements,

Velocities, and Forces, and damage

A higher concentration of High G peak accelerations has the

potential for greater damage

Most ED vibration testing limits peak accelerations to 3-sigma

RS vibration generates a greater proportion of High G peak

accelerations


Which Tests To Run

Input from all departments

Determine failure modes (FMEA)

Consider complete life cycle of product

Suggest stresses that will precipitate failures Maximum Stress vs Time Dependent

Develop test plan

Execute test

Failure of Electronic Equipment

20 year U.S. Air Force Study

55% of failures due to high temperature and thermal

cycling

20% of failures due to vibration and shock

20% due to humidity

New Product Development Testing Screens

Qualitative Testing

Qualification

Testing Qual Retest

Quantitative Testing Manufacturing

Screen

New

Product

HALT, HAST, ESD,

Power Cycle, EMI RTCA DO-160

MIL-810,

SAE J1455

Temp, Vibration, Shock,

Waterproofness,

Altitude, Humidity

HASS Analysis

Phase

Development

Phase

Essential Reliability Reference Documents

Vibration Analysis For Electronic Equipment, David S. Steinberg, Third Edition

MIL-HDBK-338B Oct 1998 Military Handbook- Electronic Reliability Design Handbook

The New Weibull Handbook, R.B. Abernathy

Practical Reliability Engineering, Patrick O'Connor

IEC 60300-3-5 Reliability Test Conditions and Statistical Test Principles

GMW 3172:2010


HALT and Relaibility Workshop

Any Questions?

Thank You!