Creep tests were carried out on Inada granite under confining pressure and
on dry and wet specimens of Kamisunagawa sandstone. Distinctive points Q(A)
(the point where the axial strain rate reaches a minimum), Q(C) (the point
where the magnitude of circumferential strain rate reaches a minimum) and
R-C (the point where the magnitude of circumferential strain acceleration b
egins to increase) were defined. Dependency of the strain values at the dis
tinctive points on creep stress, confining pressure and water presence was
examined. Circumferential strain at Q(C) or P-C was not affected by creep s
tress for both rocks. Peak load points were observed in reloading stress-st
rain curves when reloading was carried out before Q(C) for Kamisunagawa san
dstone. Circumferential strain values at peak load during reloading coincid
ed with critical extensile strain, which is defined as circumferential stra
in value at peak load during constant strain rate tests and is insensitive
to confining pressure, water presence, and anisotropy, although slightly af
fected by strain rate. Load increased slightly or did not increase when rel
oading was carried out after R-C for both rocks. R-C in creep tests scatter
ed around unloading stress-strain curves from peak load during constant str
ain rate tests for Kamisunagawa sandstone. These observations suggest that
Q(C) and P-C in creep tests are closely related to the peak load points dur
ing constant strain rate tests. Circumferential strain would be used as a c
ondition-insensitive damage indicator of rock in creep tests as well as in
constant strain rate tests. (C) 1999 Elsevier Science Ltd. All rights reser
ved.