Success through failure with accelerated stress testing
Breaking a product earlier in the development process with accelerated stress testing can yield many benefits.
When dealing with new materials, environments or requirements, how does someone get valid information? Testing the way people traditionally have can give the tester questionable information. Using a properly designed accelerated stress test to assess failure points can help identify non-intuitive failures that would normally require months or years in the field to identify. By doing this earlier in the development process, people can save time and potentially avoid significant issues.
Having the right information to make good decisions can ultimately keep a product from failing validation testing and can provide more useful information throughout the design and production process. The key is highly focused information that doesn’t come from conventional, statistics-based tests but through accelerated stress testing (AST) that provide specific failure mode information that can drive design change, improve warranty periods and extend life spans.
Each step of the development process requires decisions to be made: Can the design concept be brought to market successfully? Does the design meet engineering, customer and industry objectives? Does the production process produce the design? Testing is all about providing information to make decisions. The right test not only produces the key information but produces the information at the right time.
Conventional tests determine the fail/pass point for a group of samples when subjected to a single stress point and need to be repeated for each different state of stress. They give statistical data and are necessary to demonstrate compliance with customer requirements and industry regulations. However, they don’t necessarily identify weaknesses in materials, designs, construction or end-products. Accelerated stress testing determines failure modes for a small group of samples when subjected to different levels of multiple, simultaneously applied stress sources. AST can be a valuable tool in product development and exceeds statistical verification for requirements and regulations.
There are several methods for running AST on products:
- Highly accelerated life test (HALT): HALT is sometimes used as a generic name for accelerated testing to failure. It is structured to provide context to failure modes relative to operational and destruct limits of a product. During the HALT process, a product is subject to increasing levels of stress specifically related to the operation of a product. Structured to provide context to the failure modes relative to the operational and destruct limits of a product, HALT looks not only at what will fail and what cases the failure, but also how the failure ranks against specific specifications.
- A failure mode verification test (FMVT) reveals inherent design weaknesses by determining multiple failure modes and root causes. It measures both the “time to” and “number of” failures to measure opportunities for improvement or maturity relative to the technological limit. FMVT is intended to provide context to the failure modes relative to the maturity of the design or in comparison to similar designs. It examines what will fail, what will cause the failure and how the failure ranks against other failures.
- Key life test (KLT)/full system life test (FSLT): KLT/FLST determines a product’s reliability for a given combination of stresses on a full system. These tests evaluate product changes and ensure a product’s reliability. By putting objects through simulated real-life conditions, these tests determine if a product can meet a determined minimum life.
- Accelerated reliability test/accelerated life testing: Several sets of parts are tested at stress levels much higher than expected until enough data is collected to fit a math model to the stress, time to failure and probability of failure space for a given failure mode. These tests ultimately determine the expected minimum life of a product.
AST provides a list of observed failure modes for a given product, and this information can be used in redesign efforts or for general product information and expectations. To determine which test is best suited for someone’s needs, it is important to consider: information goals, key data needed to move a project forward and the information that would allow validations of assumptions that decisions are made on, both for business and engineering. Knowing this will determine the best test and next steps.