Abstract
Relaxation tests at high temperature give direct information on the permanent tightness of bolted joints and shrinkage assemblies. They also give direct measurements of the effectiveness of strain-relief anneals. Indirectly they give, in relatively short time and small furnace space, reliable information on long-time constant-stress behavior although direct computation of either type of performance from the other is uncertain. In relaxation tests, total strain is definitely limited, whereas, creep rates based on constant-stress tests may, but do not necessarily, correspond to excessive amounts of strain, thus giving too optimistic indications of strength. Although advantageous, relaxation tests do not take the place of constant-stress tests. A simple formulation for relaxation is given which holds good for considerable periods although rather sudden changes of trend suggest the possibility of physical changes in the material which are not easily predictable in advance. Test results are given on six materials at various temperatures.

This publication has 0 references indexed in Scilit: