Tianjin University Team Breaks Through Millimeter-Level Deformation Control Challenge for Adjacent Tunnels Induced by Deep Excavation

At the construction site of the Binhai Station building for the newly constructed Tianjin–Weifang High-Speed Railway, a massive excavation stretching 457 meters in length and reaching a depth of 23.7 meters. Remarkably, this deep excavation is located just 8.7 meters away from the heavily trafficked Haihe Tunnel, yet the resulting influence on the tunnel has been limited to the millimeter level. Behind this engineering achievement lies the Intelligent Capsule Expansion Active Control System, developed by Professor Zheng Gang's team at Tianjin University. Acting as an "intelligent brain" for excavation safety control, the system transforms traditional passive resistance to deformation into active stress regulation, representing a significant breakthrough in geotechnical engineering through an "achieving significant effects through minimal intervention" approach.

From Passive Resistance to Active Regulation

The Tianjin Binhai region is characterized by modern marine soft soil with high compressibility, low strength, and high clay content. Conducting deep excavation in close proximity to operational infrastructure such as the Haihe Tunnel poses extreme risks. Even minor ground deformation could lead to tunnel joint misalignment, water leakage, or potential threats to traffic safety.

Conventional engineering practice typically relies on passive control strategies when strict deformation limits are required near sensitive infrastructure. These include strengthening retaining structures, staged excavation, and soil grouting reinforcement. Such methods attempt to minimize predicted deformation but cannot reverse it once it has occurred. Moreover, the inherent uncertainties of geotechnical engineering—including variations in soil conditions, material properties, and construction processes—often limit prediction accuracy and make precise deformation control particularly challenging.

Professor Zheng Gang explained that traditional methods follow a paradigm of "global control for local effects," using retaining structures to limit overall ground movement and indirectly protect adjacent facilities. In contrast, the new approach enables "local control through local intervention," directly regulating stress and deformation in targeted soil zones to achieve precise control across the entire affected area with significantly improved efficiency.

Based on this concept, the team developed the first-generation active control capsule expansion technology. The technology involves embedding specially designed capsules in the soil in advance. When excavation-induced stress relief causes disturbance deformation of underground structures, these capsules expand in a controlled manner according to preset depth, volume, and geometry, enabling targeted and precise control of deformation in the soil and structures within the target protected area. Through repeated cycles of disturbance and correction—"deformation induced, deformation corrected"—the system continuously maintains deformation of surrounding structures within allowable limits.

Building upon this foundation, the team further developed a second-generation intelligent real-time control system. By incorporating artificial intelligence, this technology combines real-time monitoring, dynamic modeling, and active control. Once the deformation of the protected structure reaches a predefined threshold, the system maintains the disturbance deformation of the underground structure in a near-stable state of "on the verge of movement but not progressing further," effectively achieving near-zero incremental deformation during subsequent excavation stages. Control accuracy reaches the sub-millimeter level (approximately 0.1 mm), meaning that virtually no additional deformation occurs, thereby minimizing disturbance to the protected sensitive infrastructure.

An "Intelligent Brain" for Construction Sites

The key innovation of this system lies not only in the reversal of mechanical principles, but also in its possession of an "Intelligent Brain," enabling directional, location-specific, and quantitative control.

High-precision sensing equipment and real-time monitoring systems are deployed on-site, continuously collecting structural and ground deformation data. These data are transmitted via high-speed wireless networks to a cloud-based intelligent analysis platform. A digital twin model then dynamically calculates the optimal expansion depth and volume for each capsule. The model is continuously updated with incoming monitoring data, allowing real-time correction of predictions and adaptive adjustment of control strategies. Finally, precise actuation commands are issued to the field control system.

This entire workflow forms a closed loop of "sensing – modeling – simulation – decision-making – control", enabling continuous feedback and intervention throughout the construction process rather than relying on pre-construction calculations alone. The result is a highly precise, exceptionally precise deformation control under highly uncertain geological conditions.

In the Tianjin–Weifang High-Speed Railway Binhai Station project, the technology was successfully validated under real-world engineering conditions. Using a cloud-based remote control platform developed and operated by Tianjin University, the intelligent control unit dynamically regulated capsule expansion on-site, achieving near-zero incremental deformation of the protected structure and ensuring the safe operation of the adjacent Haihe Tunnel. This achievement demonstrates a successful application of the integrated "industry–university–research–application" innovation model. The technology was developed by a team led by Tianjin University and successfully applied on site with the active cooperation and field support of partners including China Railway Construction Group.

Professor Zheng emphasized that the technology has broad application potential in underground engineering projects, including deep excavation and shield tunneling projects, where it can provide precise protection for adjacent tunnels, bridges, buildings, metro stations, and high-speed railway infrastructure. The technology is expected to play a significant role in ensuring the safe operation and long-term resilience of transportation infrastructure and other critical engineering facilities subjected to construction-induced disturbances.

By: School of Civil Engineering

Editor: Yu Boyang