If your company is still making electronic reliability predictions using handbook methods, you could be losing a competitive edge. These methods, which make predictions based on extrapolation of very specific test conditions, can lead to over or underpredicting the life of components due to differences between the test’s conditions and your product’s operating environment.
Taking advantage of technology advancements, the industry is moving toward Reliability Physics Analysis, or RPA, to predict electronic reliability more accurately; instead of trying to extract a performance metric from a previous application, RPA looks at how a specific circuit board will operate in its specific environment.
Challenges with Handbook Methods
The Handbook Methods are problematic for several reasons:
- Because they are based on information gathered in similar operating conditions, some problem areas associated with exact operating conditions can be missed. This can lead to under or over-prediction of failures, which can cast doubt on the reliability of your product.
- A prediction of failure does not necessarily tie to a specific failure mode and therefore can lead to imprecise solutions being implemented.
- New and emerging technologies do not necessarily have handbook models to predict failure, which can lead to a hesitancy to adopt new components
Industry Specifications:
- SEA J3168 – SAE has outlined the methodology types it is looking for when subcontractors and manufacturers are trying to understand the reliability of their assembles. They are looking for reliability physics analysis of electrical, electronic, and electromechanical equipment, modules, and components. This document sets the common standards for RPA predictions across the industry
- GMW3172 – GM is moving toward complete virtual validation of their equipment before any kind of physical testing. This standard calls out the types of analyses to be run on a lot of GM subcontractor-type components as well as components included in their cars. It includes general specifications for electrical/electronic components, environmental and durability
- ISO 26262 and VW (Volkswagen) 8000 – Both of these standards are an attempt to try to understand electronics reliability from an RPA perspective
The Solution – Building the Board in a Virtual Environment and use RPA
Building and testing boards in a virtual environment saves companies time and money by allowing them to easily test the electrical reliability of the component, identify where problems exist, test different solutions, and finalize a design before building the physical board using proven RPA methods. These methods can identify the weakest components and the specific failure mode that is most likely to occur. These predictions inform design changes well before the first boards are spun.
Based on years of working with companies in all different industries, we recommend Ansys Sherlock simulation software to eliminate physical prototyping and testing for electronic hardware. Unlike other tools on the market, Sherlock uses files created by the design team to build 3D models of electronic assemblies for trace modeling, post-processing, and reliability predictions. It comes with a library with more than 900,000 common, off-the-shelf components that can be used to easily create PCBA models that incorporate more details possible than with other methods. It also offers tools for defeaturing, so users can easily omit components not needed for the testing.
Want to learn more and see some demonstrations? View the recording of this 60-minute webinar, An Introduction to Ansys Sherlock for Electronics Reliability. Contact us to start a discussion around your specific situation.
About the Author
Follow on Linkedin More Content by Adam LeMay
