Reliability of a System - Estimation of R(t) having constant failure rate "Lambda"

[Fredo405]

​

Dear professionals,

Different sources provide different equations for estimation of R(t) having constant failure rate "Lambda"

which of the following is right one?

1) R(t)= 1-(l/t) or;
2) R(t) = exp(-l/t)

In fact the results are ALMOST the same with the consideration of minor differential and i understand this is because of coefficient of "exp"

Example:

Having l(sys) = 500 (failures per million hours)
following the Equation 1 above the R(sys) = 0.9995 however if i use Equation 2, result will appear like 0.999500125.

Which of the Equations above is the right one to convert estimated "Lambda" to Reliability.

Thanks.

 

[Steve Prevette]

Re: Reliability of a System - Estimation of R(t) having constant failure rate "Lambda

For "small" values of lambda relative to t, the two equations will give close to the same value. The "correct" equation for the case of constant failure rate (no burn in, no burn out) is the exponential equation, number 2. Equation 1 was useful in the days of slide rules and calculations by hand. Now with the ready availability of computers, it is less necessary, but it always is useful to be able to do a quick calculation in your head and verify the fancy computer solution.

There is a good Wikipedia explanation of the exponential function at http://en.wikipedia.org/wiki/Exponential_distribution

 

[kenvinlee89]

Re: Reliability of a System - Estimation of R(t) having constant failure rate "Lambda

Great post! It's very nicẹ Thank you so much for your post