SAE J3168_202112

This recommended practice has been developed for use in any EEE system used in the AADHP industries. RPA is especially important to AADHP systems, which are often safety critical applications that must operate for long times in rugged environments. These EEE systems often use EEE components that were originally designed and produced for more benign consumer applications. Although the focus of this recommended practice is on AADHP applications, the process described herein is not limited to AADHP and may be used for EEE systems and components in any industry. This recommended practice has been developed jointly by the SAE International Automotive Electronic Systems Reliability Standards Committee (AESRS) and the SAE Avionics Process Management Committee (APMC) to describe a standard process for reliability physics analysis (RPA) of electrical, electronic, and electromechanical (EEE) systems, equipment, sub-assemblies, modules, and components used in the aerospace, automotive, defense, and other high-performance (AADHP) industries. RPA is a science based, engineering discipline that augments classical reliability prediction methods that are based on part counting tabulation of averaged, historic failure rates of generic component categories. RPA combines failure mechanism models developed by physics-of-failure (PoF) research with lifetime load profiles to calculate durability life and likelihood of failure over time of specific designs incorporating various EEE parts, modules, or assemblies operating in specifics applications. RPA leverages knowledge of material properties and failure mechanisms to identify failure risks, predict reliability-durability, and improve design robustness through durability simulations and models executed in a computer-aided engineering (CAE) environment. RPA interacts with and augments traditional performance related EEE simulation methods, such as electrical, electromagnetic, thermal, and mechanical analysis. RPA uses the outputs of performance CAE tools to capture stresses due to usage conditions and environmental loads (e.g., shock, vibration, temperature, electric field, etc.) and uses this information to predict time to failure based on the material science principle of stress driven damage accumulation of materials for determining the rate and degree of damage accumulation over time or usage cycles. This document standardizes and defines RPA best practices and provides a clear process for exchanging RPA results up and down the supply chain. RPA results may be used for: 1 Assessment of new component technology and packaging: Unlike traditional failure rate tabulations using handbooks of historic generic component failure rates, RPA can evaluate the potential reliability performance of new component technologies (i.e., 45 nm SOI, 22 nm FinFET, 10 nm planar, etc.) and packaging (i.e., BGA, QFN, CSP, SiP, 3D-IC, etc.) with minimal physical testing or field failure data. 2 Design verification/validation of EEE modules: Traditional qualification methods rely on physical tests that are expensive, time-consuming, and too late in the process to effectively identify all potential reliability issues/failure risks. RPA durability simulations allows for a virtual validation process that provides the Tier 1, OEM, and Airframe Integrator (for avionics) insight into reliability performance and failure risk identification so that the risks can be eliminated or mitigated early in the design process. 3 Inputs to certification analyses of EEE equipment and systems (e.g., ISO 26262, Functional Safety or FAA Airworthiness Requirements). 4 Change evaluation and acceptance: Modifications to fielded products can be stymied due to cumbersome and ill-defined processes for reviewing and approving change requests from suppliers, and the potential for expensive physical retesting. RPA allow for a standardize evaluation process.