CRE Live Preparation Course

S13 Block Diagrams, PoF, Predictions, Design Evals & Stress-Strength

Embedded recording of the session

Resources & Links

Contact me anytime with questions or concerns.

Fred Schenkelberg
fms@accendoreliability.com
(408) 710-8248 or via Skype (I’m fmsrel or you can find me by my name, I think)
**Please arrange a time for a call or Skype to make sure I’m available and expect your call.

Additional Reading/Listening

Reliability Modeling using Monte Carlo

Design for Assembly

8 Factors of Design for Maintainability

Descriptive Models of the Design Process

Understanding the Design Process

Reliability Modeling Options

Reliability Block Diagrams Overview and Value

Series System

Parallel Systems

k out of n

Standby Redundancy, Equal Failure Rates, Imperfect Switching

Standby Redundancy with Equal Failure Rates and Perfect Switching

Stress Strength Normal Assumption

The Stress – Strength Concept in Practice

CRE Live Preparation Course

S14 Design of Experiments

Embedded recording of the session

Resources & Links

Contact me anytime with questions or concerns.

Fred Schenkelberg
fms@accendoreliability.com
(408) 710-8248 or via Skype (I’m fmsrel or you can find me by my name, I think)
**Please arrange a time for a call or Skype to make sure I’m available and expect your call.

Additional Reading/Listening

Reliability Improvement with Design of Experiment, Second Edition, (Quality and Reliability)

Interpreting Design of Experiments

Taguchi Design of Experiments Approach

Design of Experiments and Reliability

Design of Experiments

CRE Live Preparation Course

S15 Design of Experiments & Design Considerations & lost connection

About an hour into this session my mouse locked up and froze me out of the session – restarted twice and still got disconnected. So, decided to spend some time troubleshooting rather than continuing.

We’re scheduling a makeup session on Tuesday, March 10th from 3 pm till we finish. If this will not work for you, please let me know and we’ll sort out a time to meet one-to-one, if necessary.

Embedded recording of the session

Resources & Links

Contact me anytime with questions or concerns.

Fred Schenkelberg
fms@accendoreliability.com
(408) 710-8248 or via Skype (I’m fmsrel or you can find me by my name, I think)
**Please arrange a time for a call or Skype to make sure I’m available and expect your call.

Additional Reading/Listening

Reliability Improvement with Design of Experiment, Second Edition, (Quality and Reliability)

Interpreting Design of Experiments

Taguchi Design of Experiments Approach

Design of Experiments and Reliability

Design of Experiments

CRE Live Preparation Course

Session 9 – Intervals & Data Management

Embedded recording of the session

(still some audio drift, yet hopefully not too bad)

Resources & Links

Contact me anytime with questions or concerns.

Fred Schenkelberg
fms@accendoreliability.com
(408) 710-8248 or via Skype (I’m fmsrel or you can find me by my name, I think)
**Please arrange a time for a call or Skype to make sure I’m available and expect your call.

Additional Reading/Listening

Confidence Interval Interpretations and Misunderstandings

Tolerance Intervals for Normal Distribution Based Set of Data

Confidence Intervals for MTBF

CRE Live Preparation Course

Session 8 – SPC, Capability, and Intervals

Embedded recording of the session (coming soon)

(still some audio drift, yet hopefully not too bad)

Resources & Links

Contact me anytime with questions or concerns.

Fred Schenkelberg
fms@accendoreliability.com
(408) 710-8248 or via Skype (I’m fmsrel or you can find me by my name, I think)
**Please arrange a time for a call or Skype to make sure I’m available and expect your call.

Additional Reading/Listening

A slew of articles or tutorials on SPC and Process Capability

Confidence Interval Interpretations and Misunderstandings

Tolerance Intervals for Normal Distribution Based Set of Data

Confidence Interval for a Proportion – Normal Approximation

Confidence Intervals for MTBF

Confidence Interval for Variance

Point and Interval Estimates

The Many Ways We Use Variance

CRE Live Preparation Course

Session 7 – Sample Size and SPC Concepts

Embedded recording of the session

(still some audio drift, yet hopefully not too bad)

Resources & Links

Contact me anytime with questions or concerns.

Fred Schenkelberg
fms@accendoreliability.com
(408) 710-8248 or via Skype (I’m fmsrel or you can find me by my name, I think)
**Please arrange a time for a call or Skype to make sure I’m available and expect your call.

Additional Reading/Listening

SOR 473 Reliability and Sample Size (podcast)

Fundamentals of Sample Size Determination (webinar)

Determine Success Testing Sample Size

Three Elements of Sample Size Concerning Decision Making

SPC Sample Size

Three Considerations for Sample Size

Sample Size Success Testing

A slew of articles or tutorials on SPC and Process Capability

Module 4 Estimation of Reliability Metrics

Lesson M04-10

Text: Section

Duration: 12 minutes

Other Methods for Reliability Prediction

We close out this module with a short introduction of standards-based approaches for reliability prediction.  We also briefly discuss a more theoretical approach called Physics of Failure.

Module 4 Estimation of Reliability Metrics

Lesson M04-09

Text: Section 7

Duration: 16 minutes

In lesson 9, we introduce an estimation approach that does not require the use of a specific probability distribution.  It is a more general approach which may be useful when adequate models are difficult to fit.  There are some shortcomings of this approach.  For example, a large amount of data is needed and extrapolation is not possible.

Module 4 Estimation of Reliability Metrics

Lesson M04-08

Text: Section 6 pages 26 – 32

Duration: 35 minutes

Handling Multiple Failure Modes & Exercise

In this lesson, we show a demonstration of modeling multiple failure modes independently and then combining the results to produce overall system level estimates (and graphical output) that account for all failure modes.

Exercise

Here are the exercise files mentioned:

Dataset_5.MTW

Module 4 Estimation of Reliability Metrics

Lesson M04-07

Text: Section

Participant Guide: Module 4 pages 43 & 44

Duration: 24 minutes

Exercises

Here is a chance to practice reliability estimation with censored data.  You also are asked to explore how making an additional assumption (regarding the shape parameter) affects the precision of the estimates.  A second exercise asks you to compare two groups by analyzing them together.

Exercises

1. 20 switches were put on a rig test. 15 units failed and 5 were right censored (Data is found in Dataset_2.MTW)

a) Using the best fitting (Weibull) distribution:

i. Generate and interpret a plot of the failure (hazard) rate

ii. Estimate the time at which the reliability drops to 0.95

iii. Provide the 95% lower confidence bound for the above estimate

b) If we assume that b is known (at its estimated value), how does the lower 95% confidence bound on the above estimate change?

2. The diaphragm material was changed to try and improve the Pump Reliability. Compare the reliability of the two materials (Data is found in Dataset_4.MTW)

Specifically:

a) Use Distribution Fitting by Design to find a reasonable distribution to model each set of data

b) Generate a Reliability Curve by Design

c) Does there appear to be statistically significant difference in the reliability performance

Here are the exercise files mentioned:

Dataset_2.MTW

Dataset_4.MTW

Module 4 Estimation of Reliability Metrics

Lesson M04-06

Text: Section

Duration: 11 minutes

Comparing Multiple Groups

In this lesson, we illustrate how to perform an analysis on multiple groups of data for easy comparisons with regard to reliability performance.  The groups are independent data sets that could be from different designs, plants, suppliers, etc.  We also discuss how to assess whether the groups are statistically similar or not.

Module 4 Estimation of Reliability Metrics

Lesson M04-05

Text: Section 6 pages 12 – 25

Duration: 14 minutes

Reliability Estimation with Censored Data

Here, we provide another demonstration in Minitab which focuses on handling censored data.  In addition to the use of a censoring column for right-censored date, we also provide an alternate more general format which can accommodate right, interval, or left censoring, if present.

Module 4 Estimation of Reliability Metrics

Lesson M04-04

Text: Section 6

Duration: 12 minutes

Exercise

This exercise in this lesson gives you an opportunity to analyze a dataset and estimate some specified reliability statistics.  You also are asked to perform a sensitivity analysis be using an alternate model and compare the results.

Excercise

8 units were tested until failure and the failure times

(in cycles) are: 5323, 2133, 16709, 7746, 8782, 14033, 2406, 2699 (Data is also found in Dataset_1.MTW)

1. Using a Weibull distribution:

a) Estimate the B10 and B0.5 life

b) Estimate R(3000)

c) Provide the 95% lower confidence bounds for the above estimates

2. Repeat the above estimates using a Loglogistic distribution

Here is the exercise file mentioned:

Dataset_1.MTW

Module 4 Estimation of Reliability Metrics

Lesson M04-03

Text: Section 6 pages 7 – 11

Duration: 26 minutes

Using Minitab to Estimate Reliability Metrics (Demo)

In this lesson, we walk through a complete example (in Minitab) of estimating reliability metrics using models found via probability plotting.  We learn how to specify the model, how to obtain specific estimates for reliability statistics (including confidence intervals/bounds), and how to obtain the desired graphical output.

Example dataset

Module 4 Estimation of Reliability Metrics

Lesson M04-02

Text: Section 6 pages 5 & 6

Duration: 18 minutes

Confidence Intervals & Bounds

Most reliability targets specify a confidence level that must be achieved in order to meet the target.  Here, we review what confidence intervals/bounds are and how they are used to when estimating reliability metrics.  We also discuss factors (like sample size) that affect the precision of the estimates.