PDA

View Full Version : Reliability prediction with several different test cases


Faisal Taquie
April 18th, 2005, 02:34 PM
Hi,

I am trying to determine the reliability of a complex system. We conduct several test cases on the product and then get the data (e.g. impact testing, HALT, vibration testing,etc.). I can calculate the reliability of the product from each individual test cases.

Therefore, I have the following:

Reliability of the product from impact testing.
R(product) from HALT
R(product from vibration)...etc.

My question is that how do I combine all of these (if I can) to come up with one reliablity prediction for the product to show to the management. The management does not want to know more than 30 different reliability numbers (we literally have more than 30 different reliability test cases). The management wants to know one simple question "What is the reliability of my product after two years?". I have the answer for each test case but I do not know how to combine it together. Please help.

Pantelis
April 21st, 2005, 09:15 AM
Well…

Let me start by first saying that assuming
(a) You have the answers for each case correctly converted to use conditions.
(b) In doing so you have also normalized the time for each condition to reflect the exposure that you expect to see in the field for each condition.
(c) The failures are independent and mutually exclusive.

With those assumptions then the overall Reliability is the product of the Reliabilities for each mode, i.e. R=R1*R2….RN