Influence of stress concentration at rock sampling on the in-situ rock stress value estimated by tangent modulus method, which is one of the oriented core methods for in-situ rock stress measurement, was experimentally investigated. Cylindrical specimens of Kimachi sandstone were preloaded at 30% UCS for 24 hour simulating in-situ stress and then a triangular shape-loading up to 40% UCS simulating stress concentration which took approx. 1 min. were carried out. Loading up to 50% UCS was applied to the specimen twice after certain delay time. Bending points were observed at 40% UCS for 0 to 1 hour delay time, at 30% UCS and 40% UCS for the 3 hour delay time, and at 30% UCS for 1 and 3 day delay time. The bending point was not observed for 1 week delay time. It is considered that the memory of concentrated stress acted at rock sampling for a short time can be lost and in-situ rock stress which acted for geological long-term can be accurately estimated if the test is carried out after an appropriate delay time.
Tangent modulus method (TMM) is one of the oriented core methods for in-situ rock stress measurement. The following procedure will be used to determine in-situ rock stress.
Oriented rock cores are sampled from the site.
Cylindrical rock specimens are made.
The specimens are uniaxially or triaxially compressed twice to a certain stress level.
The stress value of the bending point in the stress-tangent modulus curve in the first loading cycle or the point where the first and the second stress-tangent modulus curves begin to separate (this point is also called bending point later for convenience) is regarded as the normal component of the in-situ rock stress in the direction of the specimen.
In order to check the validity of the tangent modulus method, rock specimens were compressed to a certain stress level and the stress was kept for certain time (preloading). The specimens were cyclically compressed to a higher stress level twice after certain delay times. Comparison between the stress value at the bending point and the preloading stress value was made. It was confirmed that bending point appeared at the preloading stress level for dry Shirahama sandstone (Fig. 1), Shikotsu weleded tuff, Inada granite and Kimachi sandstone. Bending points became vague with delay time. However, bending points were observed at the preloading stress level even the delay time was 6 weeks for 17 hour preloading (Fig. 2). The mechanism of tangent modulus method can be explained by nearly irrecoverable closures of such voids as microcracks and pores in rock. Let's assume that A in Fig. 3a is the in-situ condition. Some voids which are tabular enough and large enough are partly closed at the in-situ stress level. A few voids would slightly open with stress relief by rock sampling (B). Rock is stiff during the first cyclic loading up to in-situ stress level (B to C) because nearly no more void closure.
