Quote:
Originally Posted by tharbert
Chris: "With used components it is not (?) difficult to make a decision with regard to the amount of fatigue damage that has been accumulated unless the part has been used in very controlled conditions and monitored in terms of the stresses that have been generated in service. Many Aircraft components are monitored in this manner but the costs of producing the base line data to make this approach work is very high."
|
Well that was a
'dumb' typo on my part - Thank you for pointing out I have now edited this mistake.
The problem with metal fatigue is that up to the point at which the crack physically initiates there is no real way to test the part to understand how much damage has accumulated.
The rate at which cracks propogate once they have initiated is very important and the conservative decision is always to assume that crack initiation is effectively the failure point.
With some components it is possible to develop a test programme to decide when failure will occur and simply replace the part at some safe proportion of a components life.
Ti rods used to be systematically replaced after around 50 racing hours and many military aircraft parts used to be treated in a similar manner.
This is because for many components the life between crack initiation and failure is very short and the risk of failure is too great to tolerate.
If parts can support reasonably long cracks it is possible to use an inspection based technique as long as the crack length does to reach a critical length between inspection periods. In this way if parts were crack detected and found to be defect free they could be re-used.
I would generally agree that 911 rods are conservatively designed and under most circumstances will never suffer from fatigue failures.
The decision is a simple risk/reward analysis.
If I were building a 2.0 litre engine to run at 8500rpm I would almost certainly shot peen the rods.
The cost is small compared to the damage that could occur if a rod failed.
If I rebuilt a standard 911T engine I wouldn't generally bother.