An Empirical Equation for Tunnel Squeezing Prediction in Weak Rock Masses

Squeezing in tunnels is a major problem in tunnel engineering that involves cost and occupational safety risks as well as changes in tunnel shape. Accurate estimation of squeezing is important in the early stages of tunnel design and in terms of solving problems during construction. Previous studies on tunnel estimation evaluated tunnels under high overburden thickness. This study focused on tunnels excavated in schists with high tectonic effects, less than 300 m overburden, and weak rock masses. Within the scope of the study, an empirical expression is proposed, which provides fast and simple calculation opportunities. In the correlation, rock mass quality Q , maximum in‐situ stress ( σ max ) calculated with k value, and support stiffness ( K max ) were used as main parameters. The equation was obtained employing 342 different cases in four different tunnels constructed in schists subjected to high tectonic effects. While obtaining the empirical equation, calculations were performed with regression coefficients ranging from 0.01 to 0.99 in the “polynomial regression curvefit method,” one of the machine learning methods. As a result, it was determined that the empirical equation obtained with a regression coefficient of 0.21 predicted the tunnel squeezing condition with over 80% accuracy. In addition, it was determined that the proposed empirical equation also provides a usable approach for strain estimation. The empirical equation proposed in the study offers a new approach in predicting squeezing in tunnels where the overburden thickness and rock quality are low and tectonic effects are high, and it is recommended to expand its scope for different tunnel conditions.