Reliability Assessment Using Physics-of-Failure Principles, Modeling and Simulation
One of the most important aspects of Reliability and Maintainability is understanding how things fail. Engineers can only make systems reliable if they understand what makes them unreliable.
Physics of failure is an engineering-based approach to reliability that begins with an understanding of materials, processes, physical interactions, degradation and failure mechanisms, as well as identifying failure models. The approach uses modeling and simulation to qualify a design and manufacturing process, with the ultimate intent of eliminating failures early in the design process by addressing the root cause. The physics-of-failure analysis proactively incorporates reliability into the design process by establishing a scientific basis for evaluating new materials, structures and technologies.
While physics of failure is practiced throughout the reliability industry, it is especially important at NASA throughout the design and build process and new technology development, where hardware is subjected to a complex set of stress interactions on complex designs and systems.
“In simple terms, reliability is the ability of a product to perform the function for which it was intended for a specified period of time (or cycles) for a given set of life cycle conditions,” according to Bhanu Sood, a Commodity Risk Assessment Engineer in the Quality and Reliability Division of the Safety and Mission Assurance Directorate at Goddard Space Flight Center. “When we say something has stopped being reliable, it has stopped performing the function for which it was intended. It has reached the end of its life. That could be in 10 years, 20 seconds or a fraction of a second, depending on the use scenario.”
The physics-of-failure approach involves understanding the technical engineering perspective of when, why and how things fail. It builds in reactions between the components of a system and includes design architecture trade-offs, design analysis validation, operational environment validation, integration of human factor concerns and technical evaluations of related program risks.
Tools of the Trade
Using the system- or subsystem-level electronic assembly information such as a drawings, layout, bills of materials and materials of construction, a physics-of failure practitioner can create a virtual model to simulate different interaction of the materials and the effects of stress conditions on the assembly.
For example, a physics-of-failure practitioner can conduct a preliminary analysis to evaluate how a specific design performs under thermal load without having to physically build a test sample and perform a test. This virtual assessment highlights the weakest points in the design, thereby informing the design and engineering team about the weak links. In many instances, design decisions are made based on the weak links in the interpretive design, which can help in making the hardware more robust in its intended end use condition.
Similarly, at NASA, Sood and his team are working on what’s called a virtual reliability assessment. For items such as electronic circuit board assemblies, the team inputs physical attributes, such as material properties, circuit layouts, bill-of-materials and solder materials, into a physics-of-failure-based reliability assessment tool. The team then uses a physics-of-failure-based analysis tool to simulate various stress conditions, such as temperature or vibration, on the assembly. The tool implements math-based physical models and allows for the creation of a computer model of the design. The off-the-shelf software tool provides an estimate of whether a part/system can meet defined life cycle requirements based on its materials, geometry and operating characteristics. Examining the physics of failure on a single board means examining just one part of the overall mission, and there are more boards and other components that make up a system. Therefore, physics of failure also involves virtually creating models of an entire system or mission in an approach called Model-Based Mission Assurance.
“We have the ability to go into the software and pull up specific slices of data that pertain to a system so we can see how subsystems interact,” Sood said. “This is critical because it gives the team an ability to simulate the effects of a single component in an assembly or the system of the entire mission to see where the components interact and where they could fail.”
Benefits of Virtual Physics of Failure
Virtual physics-of-failure assessments provide cost savings during the iterative design process because engineers can beat the learning curve by reducing the design time and cost associated with building physical models for design verification and tests. This is especially helpful for programs and projects with quick turnarounds, such as CubeSat missions.
“If you have a program where you can take 10 years to design and build a mission, maybe you can take your time designing, building and testing the physical models,” Sood said. “But if you need to be up and running in a year and a half, you don’t have the luxury of iteratively building and testing those models.”
Virtual physics-of-failure modeling also allows engineers to determine if new technological node can be added to an existing system. Using the software, they have the ability to plug in a component and run diagnostics to determine how the insertion of the new node may affect reliability.
“We run a readiness check with our virtual tools to see if we are prepared to adopt the new technology,” Sood said. “This lets us build better missions and conduct better science.”
For more information about this topic, please contact Bhanu Sood. OSMA will be issuing a report soon and it will be made available in this article.