Mumbai University > Electronics Engineering > Sem 8 > MEMS Technology

Marks: 10M

• There is a perception that small size can not instill a sense of reliability that generally, yet it is the small dimensions that generally increase immunity to shocks, make friction miniscale and reduce power consumption and heat dissipation.

• Definition - Reliability can be defined as the probability of failure free performance over an items useful life under specified environmental and duty cycle conditions. It is expressed as MTBE [ mean time between failures ]

• The distribution of failures over the life time of product population is critically important to MEMS reliability using these concepts distributor function can be developed and used for predictive purposes.

• A hazard rate that changes over the life time of the product starting high, reducing and increasing towards the end of the product life is termed 'Bathtube curve' The distribution function best models the early stage where log normal distribution is used to model 3rd stage. The exponential distribution is best to describe middle spam.

• The population have detective items that will fail within first few weeks to months of product life due to fabrication defects, material defects. An ideal failure behavior is to eliminate the failures due to defects in the infant mortality portion through burn in and defect reduction programme and not to operate product in wearout stage.

• The operational life is typically constant hazard rate section of the curve.

For proper lifetime distribution modeling, individual failure mechanism must be modeled independently there must be only one population. For multiple populations extract individually and analyze statistically.

a) Bath tube curve showing 3 stages over device lifetime.

b) Bath tube curve viewed as sum of the 3 failure rates.