What Is Failure Analysis?

Failure analysis is a systematic engineering discipline concerned with determining why a component, system, or process failed to perform its intended function. The goal is to identify the root cause of the failure and to develop corrective actions that prevent recurrence. Failure analysis draws on materials science, mechanical engineering, electrical engineering, statistics, and chemistry, with the specific methods employed depending on the type of system under investigation and the nature of the observed fault.

The discipline gained formal structure through military and aerospace programs in the mid-twentieth century, where the cost and safety consequences of fielded system failures demanded rigorous post-failure investigation processes. Those practices have since been standardized and adopted in semiconductor manufacturing, power systems, automotive engineering, and consumer electronics.

Analytical Methods and Tools

Failure analysis proceeds through a sequence of investigative steps: symptom characterization, non-destructive examination, destructive examination, and root cause determination. Non-destructive examination includes visual inspection, X-ray imaging, acoustic microscopy, and electrical characterization, all of which preserve the physical evidence for subsequent analysis. Destructive techniques such as cross-sectioning, scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDS), and focused ion beam (FIB) milling expose the failure site at the microscale.

Physics of failure modeling, developed at NASA and widely adopted in defense electronics, uses knowledge of the degradation mechanisms governing a given technology to interpret physical evidence. Common degradation mechanisms include electromigration in metal interconnects, time-dependent dielectric breakdown in gate oxides, corrosion of bond pads and lead finishes, and thermomechanical fatigue in solder joints. Identifying the specific mechanism narrows the investigation to the process, design, or use-condition factors that accelerated the failure.

Structured Analysis Frameworks

Several formal frameworks guide failure investigations. Root cause analysis (RCA) is a broad methodology that uses techniques such as the five-why interrogation, fault tree analysis, and cause-and-effect diagrams to trace the observable failure back through contributing factors to the underlying system cause. Six Sigma DMAIC (Define, Measure, Analyze, Improve, Control) provides a data-driven complement to RCA by quantifying process variation and identifying which variables most strongly predict failure.

Failure Mode and Effect Analysis (FMEA) and its criticality extension FMECA are often used upstream, during design, to anticipate likely failure modes before they occur in the field. Field failure data then feeds back into these analyses, updating failure probability estimates and revealing failure modes that design-stage analysis missed. IEC 60812, the international standard governing FMEA and FMECA procedure, provides the framework for documenting failure modes, their effects, and the severity rankings that guide prioritization.

Environmental and Stress Screening

Failure analysis often reveals that environmental exposures, such as temperature cycling, humidity, vibration, and electrical overstress, accelerated a latent defect into an observable failure. Environmental stress screening (ESS) applies controlled stresses during manufacturing or acceptance testing to precipitate and eliminate weak units before they reach the field. When a field failure occurs, comparison of the failure site to the known degradation signatures of specific stress types helps analysts determine whether the failure was caused by design margin deficiency, process variation, or an extraordinary use condition.

Weibull analysis and other parametric life data methods allow analysts to characterize the statistical distribution of time-to-failure across a population of units, estimating the fraction of units that will fail before a given mission time and identifying whether the failure pattern corresponds to early-life defects, random in-service faults, or wear-out. NIST's handbook on assessing product reliability provides reference methods for applying these statistical tools to component and system failure data.

Applications

Failure analysis has applications in a wide range of disciplines, including:

• Semiconductor device manufacturing defect screening and process control
• Aerospace and defense component qualification and mishap investigation
• Automotive powertrain, electronics, and structural component evaluation
• Power grid equipment post-failure forensics
• Consumer electronics reliability improvement and warranty cost reduction