Residual stresses and the rupture strength of epoxy-polymers

Abstract

These difficulties are easily overcome by the use of techniques developed in [2] to prepare specimens so that they have residual stresses of a fixed magnitude and sign. These techniques make it possible to strictly monitor the time to the beginning of longterm loading and to conduct this loading at a higher rate [4] using a method based on allowance for the scatter of experimental values of short-term strength and the static timesto-tupture obtained in tests with the same type of stress state. Presented below are results of studies of the effect of residual stresses on the rupture strength of an epoxypolymer based on the use of the method just mentioned. In accordance with the basic principles of the accelerated method, a preloaded lot of specimens with residual stresses of fixed magnitude and sign is randomly divided into two groups containing different numbers of specimens. It is assumed that if two groups of specimens belong to the same lot, then their data on short-term strength will have the same distribution law: specifically, individual values of strength ineach group, represented by variational series, will turn out to be very close. This assumption has been to a large extent validated by the results of studies of glass-plastics [5], epoxy compounds [6], and adhesive compounds based on epoxy adhesives [7, 8]. Thus, having tested the first group of specimens for short-term strength, we can reliably adopt the resulting data for the second group, destined for rupture tests. The second group of specimens is tested for long-term static strength ~at the same static load. After rupture of the last specimen in the ~econd group, the data on short-term strength (the first-group specimens) and the static times-to-rupture obtained (~econd-group specimens) is arranged in increasing order and is placed in this form in a table. We add to the table a third colu~ containing calculated values of the relative loading index