Warranty Cost

[amnon-g]

Hello,
I would like to verify my assumptions as follows:

Basic assumptions - In a year there are 8760 hours
Unit MTTF (non repairable) = 8760 hours
The failures distributes exponentially
100 units were sold on 1st Jan.

I would like to calculate the expected number of failures till 31st Dec the same year in two cases:
Case - 1: A failed unit is not replaced
I expect 63 (?) units to fail.
Case - 2: A failed unit is replaced
I expect 100 (?) units to fail.
Is it right :confused:

Thanks,
Amnon Ganot

[Pantelis]

Yes. By the mean you expect 63% to fail. When replacing them the population gets larger because the replaced ones can also fail.

[amnon-g]

Many Thanks,

Amnon