Rapid masonry technology for geomechanical modelling of high arch dams

Geomechanical model tests are widely used in rock engineering design and construction, especially in water conservancy, mining, and tunneling. These tests are vital for understanding the stability of high arch dams and the reinforcement measures of dam foundations. However, traditional masonry techniques often rely on manual operation, which is time-consuming and labor-intensive and prone to inaccuracies in layout.

We proposed a rapid masonry technology that integrates rapid determination of similar material ratios, small-block pressing for model masonry, and guide-rail laser positioning. Through material experiments, we obtained 236 sets of similar material specimens. Utilizing these samples and a CPO-BP neural network, we developed a surrogate model with the rock deformation modulus, scaled according to geometric similarity, as the input and the mass ratios of four materials, namely barite powder, bentonite, water, and glue, as well as the specimen density, as output. This model enables quick calculation of the ratio of each component in the model and the density corresponding to their mechanical properties, achieving a mean absolute error of 2%. In addition, we designed compression equipment for the small blocks used in model masonry, which can simulate rock joints and fissures. The setup includes a support frame, air pressure top, block mold, and cover plate opening and closing device. Users can select molds based on their needs, allowing for automated pressing of small blocks with minimal effort, thereby enhancing compression efficiency and significantly reducing labor costs. Moreover, we developed a guide rail laser positioning device for accurate model layout. When coordinate information is input into the control box, the device uses a servo motor to drive a slider carrying a laser rangefinder to specified locations, enabling rapid and precise mapping of structural planes, valley terrain, and arch dam bodies.

Practical applications of this model demonstrate that this method significantly reduces the one-year cycle required by traditional model masonry methods to approximately four months, greatly improving the efficiency and accuracy of geomechanical model construction for arch dams. The findings provide valuable guidance for enhancing the speed and precision of geomechanical model testing while offering a reliable reference for optimizing model material usage.