Abstract: Rock quality designation (RQD) index is an important parameter for evaluating the quality of fractured rock mass and this index has been widely used in rock engineering. It is difficult to express the actual rock mass of the whole space region by RQD obtained from geological exploration drilling in specific location because of structural and random distribution of RQD in engineering rock mass. In order to make up the deficiency, the spatial interpolation method in statistics provides powerful means. By considering the anisotropy of RQD spatial distribution and the difficulty to describe the structure of RQD distribution in engineering coordinate system, a standard coordinate system could be established according to the space rotation matrix and zoom ratio matrix after getting the principal Hessian direction (pHd) of geological exploration drilling RQD in engineering coordinate system. Then Kriging interpolation method is introduced. Variation function is used to describe RQD structural changes in standard coordinate system, a mathematical model which is conforming to spatial variation is established and the interpolation calculation is made. At last, the prediction of RQD distribution in engineering coordinate system will be acquired by inverse transforming standard coordinate system. Using R programming language, the method is applied to engineering. The results show that this method can effectively predict the distribution of RQD in engineering rock mass, and make up the limitation of local drilling.

Key words: RQD, principal Hessian matrix, Kriging interpolation, variation function, R programming language