Life Analysis Using Limited Field Data

[John Shelton]

I am trying to determine if I have a sufficient data source with which to conduct reliability analysis from field data using the weibull plot. The problem I have is that the data that is collected is not tracked by serial # or any other "smart #". The mfg date is more often than not omitted. The only data that I have usually is purchase date, product, and failure mode. They're also are sometimes multiple failure modes. I also do not know if this is the 1st repair or the 3rd. I really could use some help here on what I should do.

[RS Support]

John,

Do you have date failed?

[JimV]

If there are relatively few failures (vs. number of units sold), then 1st vs. 2nd vs. 3rd failure on same part won't much impact the analysis.

The bigger issue is whether or not you can incorporate all the UNfailed parts. Those are suspensions that will GREATLY impact your Weibull fit. A common error is to not include those data points.

[Andrew Rowland]

I have a similar situation. System was originally placed in service in late 70's, but computerized maintenance records not available until June 1991. Also, the records are not typically detailed enough to bin by failure mode. Essentially, the statistic I have available is "time of repair request".

I have considered cumulative time to repair request with t0 = June 1, 1991. Using truncated portions of example data from MIL-HDBK-189, I find this approach unacceptable. I also find that goodness of fit tests of my actual data typically fail with this approach.

The other approach I have considered is time between repair requests, since these times would be independent of actual system age. This would, then, allow me to determine the recurrence rate. I could use this recurrence rate in subsequent modeling with the understanding that times must be accurately specified.

Is there an optimal methodology for handling truncated system repair data such as mine?

Thanks.

[Shaique Ali]

I have some warranty data for the last 1, 2, 3, 4 years for a automotive component. I have access to all of my failures points distibution from the warranty data in terms of MIS and Mileage. Unfortunately I have no access to track the non-failure points. Or in other words that I dont have access to track my suspendid distribution in terms of MIS and Mileage. I know the volumes that I produced over the above defined period of time. Please advise I could come up a very good estimate of Reliability