Service Life

Abstract

The estimated service life of a joint made with structural adhesives necessarily depends on controlled laboratory or environmentally exposed trial joints for which the load applied is precisely known. In service, the applied load will in general be much lower than that applied in testing. The load used during testing may therefore be regarded as applying an accelerated test whereby the time-scale is a fraction of the service life expected or the increased load a safety factor incorporated into the estimated life. In the former case it is desirable to establish the relation between test life and service life; i.e. the degree of acceleration given by the test procedure. This is not easily done with confidence. Frequently, it is attempted by establishing the actual relation in the short-term and then assuming the relation to hold over very much longer periods of time. A few dead-load experiments, to be discussed below, have been run for up to 6 years and cyclical loading has been applied for numbers of loading cycles equivalent to those expected over the planned service life. The latter instance assumes that the performance of the joint when cyclical loading is intermittent and is interspersed with static loading applied between periods of cyclical loading, is at least as good as when the dynamic loading is continuous.