None of us would like to experience failures in non- reliable products. Engineers argue for different reasons for these failures, however, O’Connor and Kleyner concluded that failures are always due to people, whether they are designers, suppliers, assemblers, users or maintainers. As a result, reliability is a managerial task where management is responsible for putting the right resources in place to prevent the failures (O’Connor & Kleyner 2012, p.17).
Reliability is defined as:
“The probability that an item will perform a required function without failure under stated conditions for a stated period of time” (O’Connor & Kleyner 2012, p.1).
In order to generate a comprehensive reliability plan, the concept of Product Life Cycle can be used as shown below. In this way, each phase will need a different set of reliability tools.
Before applying this diagram on a selected case study, it is important to address the limitations and weaknesses for some of the reliability tools as the following:
- RBD (Reliability Block Diagram): this type of diagram is difficult to be constructed for complex systems with different interactions and sub systems. Additionally, different RBDs should be used when the failure is expected due to events that are not related only to the components themselves, e.g. interface problems (O’Connor & Kleyner 2012, pp. 144,153).
- Load- strength relationship analysis: this analysis may not be accurate because it depends on different assumptions. For example, it might be difficult to consider the variability in loads and strength values for certain electronic circuits (O’Connor & Kleyner 2012, pp. 121,131).
- FMECA (Failure Modes’ Effects and Criticality Analysis): this analysis takes a long time to identify all possible failures, difficult to trace the effect of low-failures correctly in complex systems and even when it is done using specialized software. Additionally, it can be inappropriate method for some applications such as digital electronic systems (O’Connor & Kleyner 2012, p.187).
- FTA (Fault Tree Analysis): different FTA charts have to be constructed for different top-events. In some cases, software is recommended to create an effective and economic analysis (O’Connor & Kleyner 2012, p.157).
- HALT (Highly Accelerated Life Testing): reliability/ durability values cannot be measured during HALT. Additionally, it does not tell how a design can be more robust and hence can be used only if failure is caused by some- known reason. Finally, this kind of testing should be conducted in a very special environment (O’Connor & Kleyner 2012, p.320).
- SPC (Statistical Process Control) Chart: it should be applied to a stable process only and the type of chart should be carefully selected and interpreted (O’Connor & Kleyner 2012, p.389).
- ESS (Environment Stress Screening): it can be very expensive (O’Connor & Kleyner 2012, p.402).
This post provides a suggested reliability plan for any product. The proposed plan is divided into three phases:
- Firstly, tasks that should be done to manage the reliability plan proactively;
- Secondly, reliability tasks from product design until product use as shown above; and
- Finally, tasks to monitor and control the reliability effort.
The application of the above phases, alongside with the breakdown of the reliability tasks over the product life cycle for a suggested case study, coffee maker, is available here.
O’Connor, P. and Kleyner, A. (2012). Practical reliability engineering. 5th ed. West Sussex: John Wiley & Sons, Ltd.
WMG: QRM Course Note, section 14 (b), pp. 16. Available at the WWW at http: //moodle.warwick.ac.uk/course/view.php?id=15228. [Accessed in 1st December 2015].