DfR Solutions Reliability Designed and Delivered

The Best Method to Calculate Risk of Failure and System-level Effects

Posted by DfR Solutions on Oct 25, 2018 10:52:00 AM

One of the key problems in today’s electronics industry is the constant changes in needs and deliverables. Today’s electronic devices are smaller and faster and are constantly exposed to changing environmental conditions. With more people putting electronics closer to a human body in the form of wearables such as iPhones, Fitbits, or heart monitors, electronics designers and manufacturers need to ensure the safety and reliability of these devices to avoid costly mistakes.

Here, at DfR Solutions we work with hundreds of electronics manufacturers across industries and have noticed an increasing number of companies reporting early life failures in the field or unexpected failures in tests due to solder fatigue. They're noticing that the classic solder fatigue calculation models do not seem to capture all the possible risks of failure.

What are the most common models to calculate risk of failure and what are they missing?

The Engelmaier model uses a strain range to calculate your cycles to failure under thermal and power stresses. Since the cycles to failure often depend on components’ materials, you need a lot of experimental data to compliment this approach. Inevitably, it becomes a challenge to do so. For example, assumptions for mechanical properties at lower temperatures may not work well at higher temperatures.

Darveaux’s model is the most common. It combines strain energy calculations and finite element analysis (FEA) and enables the quantification of crack propagation through a solder joint. Using this approach requires extensive, package-dependent testing since your crack growth correlation constant is going to change as your package and solder properties change. While this approach is very popular and involves FEA, it is time-intensive and requires an in-depth understanding of different geometries and material properties.

Blattau’s model is the most recent concept and involves computing the strain energy when calculating the cycles to failure. Deciphering the strain energy allows you to understand the relationship between your materials, and how the temperature dependency of material properties will affect your final cycles to failure for your components.

Which of these models are best suited to evaluate system-level effects?

In other words: How do you handle housing interactions with your board under extreme environmental conditions?

The Engelmaier model allows you to understand local interaction between a component solder joint, or components themselves, with the board, but can’t calculate solder fatigue of the board when there is complicated housing. To do that, you would use  FEA-based Darveaux method. 

You would perform a finite element analysis to deal with housing using any FEA tool, such as  Abaqus, ANSYS, Nastran, or SolidWorks. These tools allow you to build very complicated housing geometries that you can attach to your board, and then determine the strains and stresses that are impacting the board and the components attached to them. Once the strains and stresses are identified,  Darveaux’s method to could be used to determine cycles to failure.

 Finite element analysis  can capture modality and influence of system-level effects outside of the package/PCB. While it is time-consuming, it does produce results.

Blattau model, which is based on the closed-form equation, captures the impact of environmental conditions faster and easier, but is situationally unaware of the system-level effects.

So, which approach should we do?

The best approach is to combine the power of FEA with the speed and accuracy of Reliability Physics Analysis using any FEA engine and the Blattau model. This will enable users to learn if and when the board will fail or if it will pass test requirements.

New Sherlock 6.0 for SOLIDWORKS® 3D CAD integration allows users to bring in very complicated housing geometries from SOLIDWORKS or from any other FEA tool like Abaqus, Ansys, or NX Nastran into Sherlock to run a Reliability Physics Analyses on the board and the housing.

In Sherlock,  you can not only analyze the interaction of the printed circuit board and parts, but also the interaction of the board and the housing under various thermal conditions before the product is built. You can accomplish all this in on one tool and without being an FEA expert. Sherlock analysis provides insights into how mechanical structures, such as enclosures, batteries, thermal solutions, chassis, displays, and stiffeners influence and affect the robustness of electronics exposed to thermal and mechanical loads. This information is vital for simulating circuit card assemblies as close to reality as possible and improving overall product performance.

 

Topics: Reliability Physics, system level effects