Your critical component failed unexpectedly. How will you diagnose and resolve the issue?

If you're about to perform any assessment of a component failure, the absolute minimum information you should be requesting will be the circumstances of failure, how long the component had been in service versus it's expected lifetime, the material composition and a drawing of the part. In my experience, not often, but on occasion, the person sending the part for investigation can claim that something has or hasn't happened to the part in relation to the circumstances, but on later assessment the opposite of their claim proves to be the case. The people sending the failed part don't always have the full picture of the circumstances themselves, or they may have a motive to not disclose, or worse still, to hide the full picture.

Observation is key point. Consider you are investigating a failure, go to the area ask everyone involve, take pictures and compare with reports. Be aware and screen facts from opinions. I have a colleague who say "go to the site and watch with your feet" meaning that when someone is in charge of the investigation is important to be present and recover evidence directly.

First, I gathered all relevant information about the component, including its operating conditions, service history, material specifications, and any prior incidents. With this data, I conducted a thorough visual inspection to identify any obvious signs of failure, such as cracks, deformation, discoloration, or corrosion. Next, I performed a fractographic examination using optical and SEM to closely examine the fracture surfaces. This allowed me to determine the mode of failure, whether it was brittle fracture, ductile fracture, fatigue, or creep. Additionally, material characterization tests to ascertain the material properties. This involved hardness testing, tensile testing, and chemical composition analysis using XRF and EDS.

Keep track of the component's specifications and run the routine testing methodology to determine where the failure has occurred. Other information that can be helpful are the operational conditions (temperature, humidity, nearby environment), the operational hours, days and years of the component and the defect history of the part/component at the factory, if any (this includes defect rates over time and specific manufacturing issues that may have been short-lived in nature). Many parts can fail due to being operated out of specification; so for example if a component is not tested or specified to work under water or dust ingress conditions then this should also be checked thoroughly.

Selecting the right material analysis technique is crucial for accurate diagnostics. For fatigue failures, SEM-based fractographic analysis can reveal crack initiation and propagation, while X-ray diffraction (XRD) or FTIR might be better suited for identifying chemical degradation. Tailoring the analysis method to the specific failure mode ensures a more effective diagnosis, leading to more reliable solutions.

Using advanced technologies is great to prove hypothesis. Reality is that's no available for all. In developing countries sometimes is hard to find an updated laboratory which can provide services and when found it's common it will be expensive. Be aware of it. Sometimes Fractography analysis can be executed using optical microscopes, I have done a couple of analysis using an USB microscope. I found it very useful and comfortable to use. That kind of microscope usually has low prices.

Enhance stress testing by integrating real-time data acquisition and monitoring, such as using sensors and digital twins, to observe component behavior under controlled conditions. Accelerated life testing (ALT) can also be employed to predict long-term performance, ensuring the durability of the implemented solution. These approaches help validate the component’s resilience before full-scale deployment.

As a comprehensive approach, RCA is not only the combination of all testing results, but also requires the utilization of the engineering principles to reach a reasonable discussion conclusion, similar like a academic research paper.

Industry has advanced, processes have been continuously improved and technology have been developing so fast because knowledge have been created over important quality tools: let's mention PDCA and it's variants and RCA. Textbook example is post war growth of Japan.