# Manual for Life Cost Based FMEA

Frequency values can be obtained from observing historical data of failure devices with similar designs used under similar environmental conditions. At SLAC we have 17 years worth of data on component failures in the various beamlines stored in 2 computer databases called CATER and ARTEMIS. Data from the CATER system can be used to predict failures modes and their probability for most component designs for the ILC.

What if data does not exist for a certain design in the current system? Reliability of components/system has to be modeled using different failure distributions or test data. The most common distributions used for failure analysis are Weibull distribution for fatigue related failures and exponential distribution for electronics and probabilistic modeling. A detailed description of predicting failure frequency is given in the next section.

# 4. Availability/Reliability Predictions

Mission-critical products and complex systems like particle collider require engineers to estimate the reliability of critical components and systems before they are built. In many cases, engineers rely on test data, published data, or field data to predict reliability. Reliability is defined as the probability of a device performing its purpose adequately for an intended period of time. The basic parameter to measure reliability is known as the failure rate. General expression for failure rate is expressed as the ratio of total number of failures to the total operating time which can

be expressed as follows:

# Failure Rate (λ) =

K T

(5)

where K is the number of failures and T is the total operating time.

The base model to predict the overall reliability of a single component, taking into account all operating conditions, has the generic form:

E C S R A T b p π π π π π π λ λ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ =

(6)

where λ_{p }is

the

part failure

rate, λ_{b }is the

conditions,

π_{T }

temperature

modification

modification factor,

π_{S }electrical

stress

base failure rate at some fixed operating

factor,

factor,

π π A C

application

factor,

construction

factor,

π_{R }

power

and π_{E }

environmental modification factor.

# Mean Time Between Failure (MTBF)

MTBF is the average time between maintenances. MTBF value is the ratio of total operating time to the total number of failures. It can be expressed by taking the reciprocal of the failure rate as:

FMEA MANUAL By S. Rhee and C.M. Spencer

14/26

January 2009