(A) Reducing confidence levels

(B) Reducing variation

(C) Increasing sample size

(D) Increasing test time

Answer

The key word is most, thus we should look for the most effective, most useful, etc. method. There may be more than one correct answer so use your judgment to determine the answer that provides the best or most improvement to product reliability.

A, C, and D all involve prediction or testing. Confidence levels deal with the estimate of reliability which is unknown and doesn’t change the reliability. Increasing sample size permits a better estimate of the population, yet doesn’t change the actual product reliability. And, increasing test time may increase the testing results ability to highlight either the failure modes or time to failure, yet doesn’t change the product reliability directly.

B, reducing variability, creates a product that is more robust or further from specification limits. It is the only answer that changes the product not our estimate of its reliability. Products that are weaker or closer to specifications are more likely to fail. Reducing variability creates more products that are within specification or well away from limits, thus able to survive the expected stresses longer.

B is the correct answer.

Reference: Grant, Eugene L., and Richard S. Leavenworth, Statistical Quality Control, 7th ed., New York: McGraw-Hill Publishing Co., 1996, pp. 686-687. ISBN 0071142487

2.  Which of the following methods can be used to quantitatively identify items that are risk-critical?

(A) Design review

(B) Fault-tree analysis

(C) Concurrent engineering

(D) Human reliability analysis

Answer

I’m focused on finding a method that is quantitative (numbers) and associated with risk or probability.

D Human reliability analysis looks at one of the aspects of product reliability and does not consider hardware, software or system reliability. While the analysis can provide quantitative results is does not consider the overall product.

A is a review process and may create a list of important elements to consider or address int eh design, It is generally not quantitative. C is a method of design practice and not associated with risk assessment.

B FTA is a reliability modeling method and quantitatively permits one to determine the highest risks to product reliability.

B is the correct answer.

Reference: Gryna, Frank M., Richard C.H. Chua, and Joseph A. DeFeo, Juran’s Quality Planning and Analysis for Enterprise Quality, 5th ed., New York: McGraw-Hill, 2007, p. 338

3.  At the design stage of a product, the first safety focus should be on which of the following?

(A) Regulatory requirements

(B) Production tools

(C) Controls and equipment

(D) End-user applications

Answer

‘Design’, ‘first’, and ‘safety’ are all key terms. ‘First’ indication that more than one answer may be correct and one is held as primary.

A regulatory requirement should never dominate the design as they often provide only a baseline criteria. Some regulatory requirements are safety related and have to be met and the design stage is appropriate, yet it probably isn’t the right ‘first’ answer.

B production tools and C controls and equipment refer to variation control during the production stage. Some consideration to these items should occur during design and may not be the ‘first’ concern.

D End-user applications is correct. Safety is about the end user and the safe use of the product under design. The customer’s safety should be held as a primary safety consideration in all stages including the design stage.

D is the correct answer.

Reference: Ireson, W. Grant, Clyde F. Coombs, Jr. and Richard Y. Moss, eds., Handbook of Reliability Engineering and Management, 2nd ed., New York: McGraw-Hill Professional, 1996, p. 9.3. ISBN 0070127506

4.   Which of the following statements is true about response surface methods?

(A) They can eliminate day-to-day variations in a manufacturing environment.

(B) They are more efficient than two-level factorial design techniques.

(C) They do not require technical considerations to implement the method.

(D) They determine how an output is affected by a set of variables over a specified region.

Answer

‘Response surface’ is a type of design of experiment method.

A is incorrect as the variation in any factory is going to occur. One of the aims of DOE is to make the product or process less responsive to this expected variation (robust).

B is incorrect as response surface experiments are generally less efficient needing more levels (at least three to map a plane for example) to map the plane when compared to a two-level design.

C is incorrect as all DOEs require technical considerations to implement. Understanding the factors likely to change the output, how to construct and measure the items, and making design or process changes all take considerable technical knowledge.

D provides a basic definition of a response surface DOE.

D is the correct answer.

Reference: Breyfogle, Forrest W. III, Statistical Methods for Testing, Development, and Manufacturing, New York: Wiley, 1992, p. 250. ISBN 0471540358

5.   Which of the following assumptions is true about degradation?

(A) It is inherent and irreversible.

(B) It is inherent and reversible.

(C) It is unacceptable and irreversible.

(D) It is unacceptable and reversible.

Answer

I like this kind of answers. Really in only two questions is degradation inherent or unacceptable. Then is it irreversible or reversible?

Degradation is the deterioration of performance. It could be light output slowing fading in intensity, or the wear of a brake pad reducing braking force. Degradation is generally unacceptable at some point, yet may also have a region of performance that is acceptable and still undergoing degradation. Inherent means it is part of the product or material or process. It is what happens during normal use. For this first comparison, I would select inherent as the better answer.

The second question is about irreversible or reversible? Reversible means the product or process ‘gets better’ sometimes. By definition, degradation is irreversible which permits us to model and predict the time to failure (transition from acceptable to unacceptable performance.

A is the correct answer. BoK III. B. 2.

Reference: Nelson, Wayne, Accelerated Testing, Statistical Models, Test Plans and Data Analyses, New York: John Wiley and Sons, Inc., 1990, p. 523. ISBN 0471522775