In the first part of my new blog series I want to focus on one of the key elements when talking about reliability: prediction.
Reliability prediction for electronic components or systems is a very important part of the development or design phase and provides help to the design team to select suitable components. It has on the one hand a major impact on the robustness of the developed product, but also on maintenance and service costs. It is used to determine failure rates of electrical or electronic components, depending on their application and environmental conditions.
The results flow into the failure analysis methods like FMEDA, FMECA, or FTA, but also into the maintainability and availability analysis (spare parts).
The bandwidth of failure rates
Failure rates can be represented in different ways. For reliability prediction analysis the usual units FIT, MTBF and MTTF are used.
The Failure in time (FIT) rate of a component or system is the number of failures in one billion (10^9) hours of operation, that can be expected.
Mean time between failures (MTBF) defines the expected operational time between two consecutive failures.
Mean time to failure (MTTF) represents the average lifespan of a component and defines the exception of the time to failure.
It’s important to highlight, that all these values are statistically collected and evaluated based on probability calculations. There is no guarantee that components will fail or not fail exactly after calculated time or certain number of failures will occur after certain time.
Handbooks and standards
For evaluating the failure rates, different standards/handbooks are available on the market. The most used are:
- IEC TR 62380:2004 (Reliability data handbook, withdrawn bus still used in practice)
- SN 29500 (Failure rates of components, Siemens standard)
- FIDES Guide 2004 (Reliability Methodology for Electronic Systems)
- MIL-HDBK-217 (Reliability Prediction of Electronic Equipment)
- 217 Plus (Reliability prediction methodology, should replace MIL-HDBK-217)
- IEC 61709:2017 (Reference conditions for failure rates and stress models for conversion, has replaced IEC TR 62380:2004)
All handbooks mentioned above can be applied to different industries. You will learn more about them in my upcoming blog posts.
Finally, it is important that the results of the prediction calculation will be compared with the results of the stress tests (qualification & validation) to figure out the systematic failures.
In the next part of the blog (How reliable are you? Part 2: All roads lead to Rome) we will discuss different failures and calculation, as well as different standards and handbooks for reliability.
By Dijaz Maric, Quality Management & Reliability Engineering Consultant