----- Original Message -----
From: Jean-Pierre <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Sunday, November 28, 1999 7:09 AM
Subject: Censured data, Optimum Accelerated Censored Life
> Hi,
>
> I'm a french student and I would like to have many
> informations.
>
> What are Accelerated tests ?
> What are censored data (Type I and II) ?
>
> Thanks for your answer.
>
> Jean-Pierre
>
.......................................................................
I thought someone would have answered this by now.
Questions relate to the field of "Reliability" in terms of the reliability
of a device or physical object with respect to a time frame under a defined
physical environment (stress). All stuff I did years and years ago.
Accelerated tests refer to a type of test in which the environment during
the reliability test is more severe then that encountered in normal
operation. Typical accelerated tests are life tests at a high temperature.
Also the device may be subjected to a more severe physical stress
environment, or exposed to a severe corrosive environment to accelerate
"failures".
Censored data refers to what happens when objects under test do not fail at
the end of the test. Type II censoring occurs when only the r smallest
observations in a random sample of n items are observed to fail during the
test period. The test plan is to terminate the test when r failures occur.
Type I censoring is when n items are put on test for a set time period of T
and at T, r items experienced failure. Type I is time censoring, and type
II is failure count censoring. (See Para 1.4.1 in J.F.Lawless, "Statistical
Models and Methods for Lifetime Data")
DAH
