X hits on this document





17 / 26

Manual for Life Cost Based FMEA





us how many components were working. This information is used to analyze the type of failures in chronological order. The start time of the failure is required to determine how long it took to detect the failure, how long it took to fix, and to see if there were any delays. Beam Time Loss This information tells us how long the accelerator was shutdown due to this failure. This value should be equal to the sum of detection time, fixing time and delay time. We are interested in failures that actually forced the accelerator to shutdown. Failure component This information tells us which component has failed. You may have to work out the type of component to put it in the correct category. Location of failure This information tells us where the failed component resides along the beam lines. It indicates if the failed component was in the LINAC, North Damping ring, South Damping ring, HER, LER, and etc. Observation The technician who gets to the scene of the failure reports what is observed


when they arrive. control operator. Action taken

It is entered into the CATER/ARTEMIS by the main

This information describes what the technicians have done to fix the problem. There are no categories in the trouble report systems for entering the root


cause of failure. Thus, it is the FMEA team’s responsibility to interpret root cause of failure from Action taken and Observation descriptions. Date and time of finished repair This information helps us interpret how long the repair took and how long beam line was shut down for.



Once these pieces of information are gathered from the CATER or ARTEMIS system and analyzed a few other pieces of information need to be gathered to carry out the reliability analysis as follows:

  • 1.

    Determine the time period to investigate in the CATER system From a statistical point of view, 30 data points are said to be the minimum number to do any statistical analysis. Obviously we are able to make better predictions if there are more data points. However, analyzing hundreds of data entries may take a long time. We recommend observing a period that contains more than 50 data entries (failures) and several different types of beamlines. A 5 year period is recommended for the initial investigation.

  • 2.

    Beamline running schedule

The beam line configuration changes overtime and the number of components change with respect to the configuration. Thus, we need to know which beamlines were running during the investigation period and how many hours it was suppose to run (Runtime).

FMEA MANUAL By S. Rhee and C.M. Spencer


January 2009

Document info
Document views113
Page views114
Page last viewedFri Jan 20 08:13:18 UTC 2017