Reliability predictions estimate the failure rate of components, sub-systems and systems during their infancy, useful life and wear out periods and may be expressed in the number of failures in a given set period of time. Some of the most common methods used to predict an item’s reliability is its Mean Time Between Failure commonly known as MTBF and the use of Weibull Analysis.
The main methods available for determining the reliability of a product are:
There are two standards widely used by the reliability engineering world to predict product reliability. These are Mil-HDBK-217F and Telcordia.
In the case of MIL-HDBK-217F there are two methods, a parts count and a full stress:
- Field Failure Data
This method is used to determine the reliability of a product based on its historical operational performance in the field, collecting accurate data is of paramount importance for this to be a success.
This method is used to determine what the reliability of a new product will be based on “known” reliability information from a similar product with similar technology and complexity.
Weibull Analysis can be used as a method of determining where a population of modules are on the bathtub curve. The Weibull distribution is a THREE parameter distribution. The three parameters that make up the distribution are β, η, and time. The β parameter is the slope and signifies the rate of failure and η the Characteristic Life.
Social Media Links DE
send a request aaaaa
Case Study – Emerson Network Power
How Anecto developed a number of innovative solutions for the leading multinational Emerson Network Power.