In the simplest terms, reliability is the assurance that a machine or a system provides its designed functions. Engineers therefore spend significant amount of time and resources during the system design, system development, and production phases of the system to ensure the system provides the desired service level.
One perfect example of system reliability is the life cycle of the popular fictional “one-hoss-shay” by Oliver Wendell Holmes, Sr. According to the satirical poem “The-Wonderful-One-Hoss-Shay”, a fictional Deacon designed this wonderful one-hoss-shay with the very best of material so that each part was as strong as every other part, and a reliability to last a hundred years without depreciation. According to the poem, the system lasted until its hundredth year. Thus reliability engineering ensures a system delivers the same services throughout its lifetime before falling with zero scrap value.
Designing a product/system may require redundancy of subsystems, introduction of newly developed components or changes in design configuration, all of which have a major impact of the system’s reliability.
A formal and more accurate definition of reliability:
Reliability is the probability that a product will operate or a service will be provided properly for a specified period of time (design life) under the design operating conditions (such as temperature, load, etc) without failure.
In other words, if T is a random variable denoting the time of failure, then the reliability function at time t can be expressed as:
R(t) = P(T>t)
No one disputes the need for engineering products to be reliable. The average consumer expects reliability in domestic products such as TV sets and automobiles. Organizations such as airlines, military and public utilities are also aware of the cost of unreliability. Consumers always expect warranty/assurance (reliability) from the manufacturer at least to last a reasonable time. However, this approach does not guarantee the reliability of the products, but that the manufacture will repair the product and the consumer will have to bear the inconveniences in the event of a product failure and outside of the warranty period, only the consumer suffers.
Why engineering products fail?
Knowing as far as is practicable the potential causes of failures is fundamental to preventing them. However, it is rarely practicable to anticipate all of the causes, so it is also necessary to take account of the uncertainty involved. The reliability engineering effort during the system design, development and the manufacture of the system should address all the anticipated and possible unanticipated causes of failure to ensure that their occurrence is prevented or minimized.
The main reasons why failures occur are:
- The system design might be inherently incapable: this means the system might be too weak, consumes too much power and so on. The list of possible reasons is endless and unique to a design. The more complex the design, the greater is the potential.
- The system may be over stressed in some way: if the stress applied exceeds the strength of the system, then failure will occur.
- Failures can be caused by variation: a system will be reliable if the applied load does not exceed the system strength. In any event of uncertainties, failure might occur.
- Failure can be caused by errors: errors such as incorrect specifications, designs or software coding, faulty assembly or inadequate maintenance can also lead to failure.
Reliability engineering in summary is more concerned with probability of failures within a time domain. This makes the difference between traditional quality control and reliability engineering. Durability as an aspect of reliability relates to the ability of an item to withstand the effects of time (operating cycles) dependent mechanism such as fatigue, wear, corrosion of products, and so on.