Abstract:
One of the remaining tackles in rock mass or soil grouting is the postulation of appropriate grouting parameters for an economical execution of works, considering the limitations of fracturing the ground. From an economic perspective, a minimised usage of material to complete within time is desired. From a geomechanical perspective, minimised usage of material to penetrate the ground effectively at an appropriate pressure is desired. Following this simplified approach, ensures completion in a minimum of time and economical usage of material. The lack of a comprehensible determination of appropriate parameters is still the driving factor for either too conservative execution or opposite false outcomes with fracturing or cracking the ground. Based on reverse engineering, grouting processes were analysed and grouped within a machine learning to identify processes which lead to undesired and standard execution of works. Atypical grouting is typically related to fracturing of the ground, which can be predicted, based on the algorithm and counter steered in real-time with the new system. The ability to identify the process allows now for stoppage upon early detection during execution. The overall process of steering the operation is based on the initial zero flow of the grout. The appropriate steering of the process is being ensured by focusing only on the two factual parameters, grouting pressure and flow rate, rather than focusing on sophisticated geological input parameter detection. The system processes the grouting data in real-time and adjusts the grouting parameters accordingly to ensure an optimal grouting operation. Therefore, it is capable of detecting the maximum feasible pressure about the saturation, while grouting. The appropriate penetration is found the zero flow criteria, which can be controlled by a relation of maximum tolerable pressure and the flow rate. The device can be attached to any grouting unit. It allows for appropriate steering of all grouting operations. In combination with a modern data management system, the device represents a preconditioned self-learning system. The operator selects the appropriate grouting position and starts the grouting process. Controlling these two parameters allows to tackle all operations within project limits and to ensure a proper grouting process. The device furthermore determines the optimal end of the operation and stops the grouting pumps and therefore allows for autonomous grouting processes.