Accelerated testing typically targets - life estimation and defect or design weakness identification. The intent is to obtain more information from a given test.
Challenges in Accelerated Life Testing Pradeep Lall, Associate Professor Auburn University Department of Mechanical Engineering and Center for Advanced Vehicle Electronics (CAVE) Auburn, AL 36849 Tele: 334-844-3424 E-mail: lall@,ena.aubum.edu Acceleration factors correlating the accelerated test conditions to field conditions are often derived based on the Arrhenius relationship. The component itself is assigned an activation energy, which may be a weighted average of the failure mechanisms. Convenience of implementation is a prime advantage of this approach. However, the computed acceleration factors implicitly assume a dominant dependence on steady state temperatcm for acceleration of the failure mechanisms. A more complete discussion of the failure mechanisms and their temperature dependencies are available in Lall, et. al [1997]. In general, failure mechanisms may be accelerated by a various forms of temperature stresses including temperature cycle, time dependent temperature change and spatial gradient of temperature or non-temperature stresses including humidity, shock and vibration. Variability of field use conditions often make the task of developing correlationsmore difficult.
SUMMARY Miniaturization has stimulated the emergence of new packaging technologies often accompanied by new materials. Further environmental regulations have accelerated the use of lead-free solders and bromine free laminates. There is need for methodologies for development of accelerated tests to target new failure mechanisms and modes. Accelerated testing typically targets - life estimation and defect or design weakness identification. The intent is to obtain more information from a given test time than would normally be possible, Time compression is typically achieved by acceleration of single or combination of stresses - temperature, vibration, humidity, etc. Short product development schedule, time-to-market requirements and the development timelie often dictate a tixed time under test. In a test-to-ked life approach, given number of the product are tested to a tixed life, with zero failures allowed. The test demonstrates the product's minimum reliability, while allowing for a fured time to conduct testing. In absence of life-test information, the challenge is to decipher if the product may be over-designed or marginally designed.
REFERENCES Lall, P., Pecht, M., Hakii, E., Influence of Temperature on Microelectronic and System Reliability, CRC Press, Boca Raton, Florida, 1997.
Optimized stress magnitudes for new technologies may be hard to derive. A relative goodness of design approach is often used, where newer technologies are subjected to accelerated test conditions used for the previous generation of products. High value of stresses will precipitate overstress failures instead of wear-out failures impairing any attempt for life correlation or derivation of acceleration factor. Failure mechanisms and modes observed in accelerated test may not be observed in field life. Accelerated tests such as thermal shock from -55 to 125°C may stress the material beyond the equipment operating range, such as above glass transition temperature, where material properties may be significantlydifferent.
727
2004 Inter Socleiy Conference on Thermal Phenomena
