The bathtub curve is widely used in reliability engineering. It describes a particular form of the hazard function which comprises three parts:
The name is derived from the cross-sectional shape of a bathtub.
The bathtub curve is generated by mapping the rate of early "infant mortality" failures when first introduced, the rate of random failures with constant failure rate during its "useful life", and finally the rate of "wear out" failures as the product exceeds its design lifetime.
In less technical terms, in the early life of a product adhering to the bathtub curve, the failure rate is high but rapidly decreasing as defective products are identified and discarded, and early sources of potential failure such as handling and installation error are surmounted. In the mid-life of a product—generally, once it reaches consumers—the failure rate is low and constant. In the late life of the product, the failure rate increases, as age and wear take their toll on the product. Many consumer products strongly reflect the bathtub curve, such as computer processors.
While the bathtub curve is useful, not every product or system follows a bathtub curve hazard function, for example if units are retired or have decreased use during or before the onset of the wear-out period, they will show fewer failures per unit calendar time (not per unit use time) than the bathtub curve.
The term "Military Specification" is often used to describe systems in which the infant mortality section of the bathtub curve has been burned out or removed. This is done mainly for life critical or system critical applications as it greatly reduces the possibility of the system failing early in its life. Manufacturers will do this at some cost generally by means similar to accelerated stress testing.
In reliability engineering, the cumulative distribution function corresponding to a bathtub curve may be analysed using a Weibull chart.
Some investigations in the aerospace and other industries have discovered that most failures do not comply with the bathtub curve. It is argued that the bathtub curve is an old concept and should not be used as a stand alone guide to reliability. Most interesting in these investigations was the conclusion that wear-out issues in complex systems only count for about 4% of all failures (refer to Reliability centered maintenance (RCM); Boeing 747 - MSG2 and MSG3 investigations). According "The RCM approach" about 6 different types of failure rate curves can be distinguished. It is also remarkable that the highest contribution to failures appear to be failures that have a constant failure rate character. This mainly counts for complex systems, being highly integrated.