Abstract: The non-uniformity and discrete nature of rock mass are among the most significant characteristics of rock mass. Jointed rock masses are complex block structure systems composed of different scales of rock masses. Different indexes and standards for the division of rock mass structure and integrity in various industries, along with qualitative descriptions, hinder the generalization of obtained data and conclusions. Using unweighted undirected graphs, the study achieved the rapid and batch construction of random joint network models. Employing algorithms such as connectivity domain detection and breadth-first search, the constructed simulation models were analyzed for block degree and equivalent sieving. With the original closed block area, the independent effective block, and the reconstruction of the effective block amplification area as a parameter source of the block degree deconstruction, a quantitative index for the degree of blockiness of jointed rock masses was defined (the rock block index, RBI_2D). Diverging from traditional rock mass quality assessment bases on structural planes, this study proposes a normalized treatment and quantitative evaluation method for the deconstruction of jointed rock mass blockiness. Simulations were conducted on jointed rock masses with varying degrees of integrity based on relevant industry standards and actual engineering site conditions, establishing a correlation between rock structure types and RBI_2D. The integrated development of a normalized processing evaluation system for jointed rock mass blockiness has been implemented and applied to the rock mass quality evaluation and surrounding rock classification of the Gandershan Tunnel on the Baotou-Yinchuan Railway, showing good consistency with onsite measurements. The research provides an alternative effective approach for obtaining integrity indices in surrounding rock grade evaluations.

Key words: jointed rock mass, blockiness, deconstructive analysis, surrounding rock classification, engineering application