𝔖 Bobbio Scriptorium
✦   LIBER   ✦

Analysis of creep failure times of cured-in-place pipe rehabilitation liners

✍ Scribed by R Nassar; M Yousef


Publisher
Elsevier Science
Year
2002
Tongue
English
Weight
168 KB
Volume
17
Category
Article
ISSN
0886-7798

No coin nor oath required. For personal study only.

✦ Synopsis


In experimental studies involving accelerated time until failure, each specimen is usually tested until failure or a given time t whichever occurs first. In analyzing such data, the conventional method used is least squares regression. The recommendation Ž . Ž . ASTM D2992 is to include non-failed observations censored data as failures in establishing the regression line. Such an approach could lead to a serious error in prediction. To remedy this situation, we recommend instead a maximum likelihood approach in which non-failed observations are incorporated as censored and not as failed observations in the analysis. We compare the two methods of analyses using data on accelerated time until buckling of cured-in-place plastic liners.