The evolving complexity of components and technologies makes ensuring reliability of electronics designs increasingly difficult and drives the need for Design for Reliability (DfR).
Along with bringing better performing, more reliable products to market faster, DfR allows defects to be caught early in the design stage, minimizing cost and disruptive impact compared to those same defects being found later in the process.
There are a number of DfR testing tools that assist reliability analysis and prediction, each with their own strengths. However, most of the tools are too broadly focused to fully address the unique challenges of the electronics industry, and the process you choose could limit or skew results and impact costs. Knowing the benefits and drawbacks of DfR qualitative and quantitative testing will help you make an informed choice.
Generally speaking, qualitative tests expose small product samplings to a single severe level of stress, to multiple stresses or to a time-varying stress like temperature variance. If the samples survive, the product passes the test. If not, the design is re-evaluated and improved until satisfactory testing results are achieved.
While highly effective in revealing probable failure modes and increasing reliability, an improperly designed qualitative test could result in product failures due to factors they would never encounter in real life under normal conditions. For this reason, qualitative tests are best used to provide input and feedback during test design instead of application.
Two popular qualitative tests are Highly Accelerated Life Test (HALT) and Failure Modes and Effects Analysis (FMEA).
Highly Accelerated Life Test (HALT)
HALT is an excellent, low-cost tool for assessing electronic product robustness – not reliability. HALT is a test that is not considered DfR as failure analysis is too far down the path to be considered part of the design process.
- Complex environments with varying temperatures and non-predictive vibration loads from repetitive shock prevents direct extrapolation of results to life data
- Not everything can be modeled or predicted (screw loosening, for example)
- There are no HALT failure modes that cannot be addressed through strong DfR activities
- Best in Class organizations will predict HALT failure modes before performing HALT testing
Failure Modes and Effects Analysis (FMEA)
FMEA is a step-by-step approach for identifying all possible failures in design, manufacturing or assembly processes, or those of a product or service. FMEA data translates to the ability to prioritize failures based on the seriousness of their consequences, occurrence frequency and ease in detection.
While comprehensive, FMEA is also limited:
- Failure modes are identified but not corrected, necessitating outside action
- Failure mode identification, testing and documentation is time-consuming and can increase project scope without improving results
- Testing effectiveness is weakened if not implemented early in the design phase
Quantitative Accelerated Life Testing (QALT) supplies reliability information based upon usage rate acceleration and overstress acceleration tests that quantify the life characteristics of a product. QALT also often incorporates HALT data.
Not surprisingly, accelerating usage and stress quickly provides probability of failure information and insights on mean life under normal use conditions, projected returns and warranty costs, risk assessments and design comparisons. While highly accurate, QALT tests require analysis and interpretation by an expert after testing which can leave some margin for error or skewed results.
Automated Design Analysis
Unlike HALT and FMEA “test-in reliability,” DfR is a process that ensures product or system reliability prior to prototyping. As such, modeling and simulation software that provides an independent design assessment without the need for a physical prototype is a good fit.
DfR Solutions’ Sherlock Automated Design Analysis™ Software is a unique tool that closely aligns with DfR goals. Using Physics of Failure (PoF), Sherlock allows designers and reliability engineers to:
- Determine the expected life of the design
- Characterize PCB behaviors before product prototype testing
- Annotate potential problems within the bill of materials to prioritize focus areas
- Collaboratively develop design recommendations to reduce uncertainty and risk
Automated design analysis results and design reviews are a powerful combination in determining how much testing is needed with the added benefit of culling out unnecessary assessments.
Sherlock results deliver a clear and comprehensive estimate of useful product life under the expected operating conditions. This level of reliability provides manufacturers with the ability to better estimate warranty costs, and ultimately provide lower lifecycle cost to end users.
For more information on DfR processes and tools, download the slide presentation from our Design for Reliability webinar. Just click the button below!