Wireless optical displacement sensor for convergence and divergence monitoring

FIG. 1 Scenarios of convergence (Erlandsson, 2020). The automation of monitoring in general appears undeniably tempting right up to the moment at which prices are discussed. That is when suddenly manual surveys seem not so outdated anymore. Obviously, this perspective is rather short-sighted if automation is applied where its advantages can be exploited, for example: In areas where access is limited, difficult or dangerous. In projects where frequent recording rates (that is, shorter than one day) are required. Where long-term observations (that is, longer than one year) are desired. Where it takes care of repetitive procedures to release human resources for decision-making tasks. Any combination of the above. The nature of tunneling more or less represents all of the above. Monitoring work is not only repetitive, but can potentially disrupt and delay tunnel construction or operation because of constrained access conditions. Such delays can give rise to significant cost and schedule issues. A common parameter in tunnel monitoring is convergence as observation of changes in the tunnel profile perpendicular to the alignment. This can be conducted as life-cycle monitoring of ageing masonry tunnels as well as temporary observation of deformation induced by adjacent construction measures, or even in advance and construction of the tunnel itself. In all of the above cases, wireless condition monitoring (WCM) is an intelligent solution with regard to costs, ease of installation and efficiency. The optical displacement sensor (ODS) is a typical component of a modern WCM system, representing unrivalled simplicity in convergence monitoring. Convergence Tunnel deformation. Excavation of a tunnel, or close to an existing tunnel, will cause volume loss, and thus a disturbance of the surrounding medium. To re-establish the initially lost equilibrium, depending on the material, the ambient environment starts to settle with the tendency to fill the void represented by the tunnel or cavity. The design of a tunnel profile and lining type is intended to deflect external loads to maintain stability and establish a new final equilibrium with the surrounding ground. Only if this precondition is reached can operation of the tunnel be safe and provide support of the overburden, for example, in bearing urban infrastructure. Therefore, it is essential to observe tunnel deformation as a component of a project’s holistic monitoring concept. Depending on factors such as the tunnel profile, surrounding regime and overburden, the tunnel can experience various types of deformation (Fig. 1). As convergence we consider the displacement of a point toward or away from the interior: that is, the center point of the profile area. For a circular profile this would be the direction perpendicular to the outside tangent. For efficient convergence observation it is crucial to estimate the pattern of deformation in advance to distribute sensors appropriately and yield representative results. Segmental lined tunnels will deform at the joints as the segments can be considered rigid, while masonry-lined tunnels might deform homogeneously and require more thorough consideration of placement of the sensors. FIG. 2 Convergence observation utilizing total stations. Common observation techniques. The strategy of continuous observation depends on a variety of criteria, not least those already mentioned in the introduction. Generally speaking, monitoring is the observation of change, and thus the comparison of an in situ current state with an initial state some time in the past, represented by a zero or baseline measurement. In addition, the parameters observed need to be identical and comparable in both epochs, which might require some information about the environmental conditions. In the case of geometric observations like convergence, the location of the observed spot must be identical or at least be restored with sufficient precision. The latter becomes increasingly significant if the tunnel undergoes renovation, or the initial survey is carried out when only the preliminary lining is installed. The first and foremost consideration would be the observation frequency. If annual observation repetition is required and access is not an obstacle, automization is hardly economical. Manual observations can be conducted using geodetic total stations, tape extensometers or in modern days, even terrestrial laser scanning: For total station observations, the tunnel profile is equipped with reflectors, ideally prisms that are observed from the total station’s location by horizontal and vertical angle and slope distance (Fig. 2), which are converted into local coordinates at millimeter accuracy. The linear distance between these coordinates represents the observation parameter. Depending on the spacing between the profiles and the visibility along the line of sight, a number of profiles can be observed from one standpoint before moving onward. To observe convergence using a tape extensometer, eye bolts will be installed at the opposing ends of the intended chord and the extensometer temporarily hooked into them. The reading can then be taken from the micrometer. Tape extensometers are easy to use and provide high accuracy up to ±0.2 mm. The measurement will require some effort and support such as hydraulic access platforms. In any case, the measurement tape or wire has to be removed after measurement as it would impede tunnel works or hamper operation. Laser scanning covers the tunnel surface with an enormous density of measurements and can provide valuable information, particularly in the case of inhomogeneous deformation or irregular surfaces. However, identification of discrete points might be difficult and the extraordinary volume of data requires some effort in data processing and most probably, some manual intervention. All of the above methods require qualified personnel operating the survey equipment, working “through the tunnel” by observing section after section. This sequential strategy limits the observation frequency significantly and makes short observation intervals of less than one day unrealistic, or at least, highly inefficient. Automation of observation. Automation is a valid strategy to overcome the observation obstacles encountered by manual surveys. To achieve automation, the survey equipment needs to be permanently installed. Hence, tape extensometers are already disqualified as they would interfere with tunnel and construction operations. Theoretically, laser scanners could be installed, but the hardware cost would usually prohibit the permanent installation of multiple instruments. Thus, robotic total stations (RTS) remain a widely used means of deformation observation. Here, by far the biggest share of hardware investment lies in the total station itself. To minimize these costs, multiple convergence sections would be observed from one location. This seems efficient, but it has drawbacks. The quality of observation will decrease with increasing distance as dust, fuel emissions and other sources of haze cause the line of sight to be blurred and physical obstructions such as traffic or machinery block it entirely. Measurement at distance is further complicated by increasing distance, making targets appear closer to each and therefore harder to discriminate (Fig. 2). To make things worse, the RTS itself will most likely be located in an unfavorable position on a bracket mounted to one side of the tunnel. Hence, alternatives might be required. In recent years, miniaturization has led to the development of devices referred to as “nodes,” that combine sensor, data transmission and power supply in a single compact housing. The key term here is WCM, and their ease of use makes them attractive alternatives in many of the scenarios described. Wireless observation constellation. WCM systems essentially consist of three components: (1) the sensor node, (2) the gateway and (3) the data-management platform. Compact design and low power consumption of the node allow easy installation of the field equipment even by nonexperts with only minimal maintenance required for up to 15 years depending on the sensor. The nodes establish communication between an implemented sensor, such as a MEMS tilt sensor and/or a laser-based ODS, or a connected external sensor (such as a potentiometric crack meter, vibrating wire strain gauge or PT100 temperature sensor). The data is transmitted either via 868/915-MHz LoRaWAN-based signal or meshed 2.4-GHz Wi-Fi frequency to a gateway. It is temporarily stored at the gateway and transferred either via cellular network or ethernet landline to a remote central server. Wireless communication minimizes installation effort and reduces the risk of accidental damage or vandalism causing loss of data that cabled systems can experience. The long battery life reduces the need for maintenance to almost zero and allows for redeployment on multiple short-term projects if not required for long-term lifecycle monitoring. Intelligent systems enable remote system modification according to the stage of works, related expectations of the development of changes, or requirements of contingency measures. Last but not least, WCM allows easy modification and expansion of the system by gradually adding more and even different sensors, all communicating within the same platform without the need for additional cables. It is relatively simple to relocate or replace elements of the system; for example, to focus resources close to the tunneling face as construction progresses. Even the subsequent extension by additional sensor types is reduced to the mere hardware installation. The data-management platform provides access for multiple users with adequately defined user roles. The data are stored and further processed if required. In the event of predefined threshold values being breached, automated alarms can be sent to approved stakeholders. FIG. 3 A section-diagram screenshot from the data-management platform. Case studies Costa Blanca — Martorell Tunnels — Spain. As part of the Mediterranean Corridor project, the Spanish rail infrastructure operator ADIF needed to upgrade three old tunnels in the Catalonia region for use by bigger, faster trains. Located on the line between the towns of Martorell and Castelbisbal, each tunnel was several hundred meters in length. Main contractor Dragados was tasked with strengthening the linings, lowering the track and installing overhead electrification. By lowering the invert, the overall tunnel geometry was altered and there was a need to closely observe whether the redistribution of loads would jeopardize structural integrity. Continuous monitoring was instructed to compare predicted and actual movement. This was viewed as essential to protect the workforce and provide timely warning of any significant deformation that would require a response from the construction team. The requirement for near real-time monitoring throughout the works made automated monitoring the obvious, and probably only, choice. Conditions for longdistance optical observations were expected to be highly dusty, with sight lines blocked by construction machinery and materials. The use of total stations was therefore rejected. Dragados and its assigned monitoring consultant, INSTOP, opted for a Senceive WCM system using laser distance sensors (ODS) in a triangular arrangement with a node at each spring line and one at the crown (Fig. 3). Although also relying on optical observations, the distances observed in this configuration are much shorter and therefore more reliable than in any applicable total station arrangement. The ODS nodes include a three-axes MEMS tilt sensor enabling observation of rotational movement at each location regardless of the node orientation, as well as measuring the distance to a reflective target. Arrays of ODS nodes were installed at 25-m intervals such that the Costa Blanca tunnel featured 30 ODS profiles, Martorell tunnel 27 profiles, and Castelbisbal tunnel 35 profiles. Achieving the nominal submillimeter accuracy not only depends on the environmental conditions — for example, the medium the laser beam has to travel through — but also on the quality and orientation of the target. Aiming at rough and/or oblique surfaces decreases accuracy as the incidence point might shift inadequately. Therefore, specific targets were installed to provide a reflective, smooth surface at 90° incidence angle. In addition to the distance measurements, the ODS’s three-axes MEMS tilt sensor allows the observation of rotational movements without the node being restricted to horizontal oriented installation. Besides the challenging environment, one more reason for choosing WCM was the tight schedule, which provided a very short window for installation. Because the system is configured before installation, and because of its modular characteristics, a wireless system is easy to install, even by nonspecialists. At four sections per hour, an entire tunnel installation was completed within a single working day. At the Costa Blanca and Martorell tunnels, the system included instruments to automatically observe longitudinal settlement. The system consisted of fixedlength beams daisy-chained in cardanic wall mounts with wireless tilt nodes fixed to the beams. Via trigonometric functions, any longitudinal tilt could be converted into vertical displacement by the known length of the beam. The accumulation of the incremental vertical displacements along the chain yielded longitudinal settlement or upheaval. As the components of the system were located relatively far apart, the long-range LoRaWAN system was the chosen communication platform. Each tunnel was equipped with its own gateway serving a system with the furthest node 1.2 km (0.8 mile) away. Gateways also received signals from the nodes located in the other tunnels — at distances of up to 8 km (5 miles). However, to maintain redundancy, the multigateway constellation was retained. Specific diligence was required on the Martorell tunnel as the city of Martorell and its high-rise buildings were located directly above. A number of rod extensometers were therefore installed to observe vertical ground movement at three levels (3 m, 6 m and 9 m), each communicating its readings to the solarpowered gateway via multichannel vibrating wire nodes. All of the systems were delivered preconfigured and with all sensors functioning. It was therefore possible to check system performance within minutes of installation, avoiding the need for potentially disruptive repeat visits to the site. The renovation of the three tunnels, including their intermediate galleries, was completed within three months with a monitoring system of more than 400 nodes in place for about four months. After demobilization, the components were redeployed on other INSTOP projects. Chipping Sodbury Tunnel — UK. A comparable situation to the previous case was encountered in the Chipping Sodbury Tunnel near Bristol, UK. The structure, which was completed in 1902 and thus is more than 100 years old, is 4,000 m (13,100 ft) in length and features a brick-lined arch profile 8.4 m (28 ft) wide and 6.4 m (21 ft) high. To comply with modern requirements there was a need to lower the track by 150 mm in places to achieve the required clearances for a new alignment design including electrification. Again, a tight schedule and challenging site conditions, plus the need to keep trains running, necessitated a solution using automated monitoring methods. The monitoring consultant, AECOM, had gained experience in similar renovation projects and could therefore use the proven technology of WCM. Tunnel geometry, obstructions by traffic, and construction activity as well as dusty, humid atmospheric conditions prohibited total station observations. Unlike in the Spanish tunnels, it was possible to access cellular signal reception throughout the entire length of Chipping Sodbury Tunnel. Therefore, Senceive’s proprietary communication platform FlatMesh could be employed. Due to the greater bandwidth of the 2.4 GHz signal, the platform provided optimized synchronization and short message pulses for high-frequency observations. The wireless mesh platform provided significant redundancy and therefore a high level of system reliability. Hence, establishing communication proved to be relatively straightforward by simply installing gateways at regular intervals to gather the data from adjacent nodes. As stated, three-axes tilt sensors can be installed in any orientation without the necessity of prior leveling. Therefore, the initial approach was to install daisychained tilt beams within the tunnel section representing deformation arrays. AECOM chose miniaturized Senceive NanoPlus MEMS tilt nodes for compactness and confidence in the IP68 performance under the expected wet and dusty ambient conditions. The use of shape arrays had been considered but rejected because each chain would have needed its own gateway, whereas data from numerous tilt sensors could be gathered at a single gateway, saving time and money. As the work progressed it became impractical to install beam chains at some locations where tunnel curvature meant they would have projected into the tunnel and caused an obstruction. Hence, individual tilt beams were installed only at crucial locations, still employing more than 200 nodes throughout the tunnel. Where convergence observation was required, tilt nodes were substituted by ODS nodes positioned to measure the distance to the opposite wall. Permanent online access to the data-management platform allowed remote temporary deactivation of the inherent alarming during blockage of the line of sight by construction works to avoid false alarms. The tracks were lowered one after the other; the behavior of the remaining track was observed using rail-specific vibration resistant tilt node. A total of 843 of these high-G tilt nodes were installed at 3-m (9-ft) centers within and adjacent to the immediate area of impact. In addition, it was considered crucial to ensure proper operation of the tunnel drainage, realized by a subsurface culvert located in the center of the tunnel. It was expected that any significant impact on this structure would result in measurable deflection of the invert and thus the track, so additional tilt nodes were installed on sleepers to observe movement related to the invert. Conclusion The monitoring programs described here found that neither of the presented projects encountered significant deformation. However, because the works were conducted under the umbrella of risk mitigation by meticulous observation of structural changes and deformation, they could be progressed with confidence and efficiency. The use of the optical displacement sensor in particular enabled analysis of convergence in situations that would not have been possible with other methods. The use of automated wireless systems in general demonstrated that their capability encompasses the entire workflow from data acquisition via transfer and processing toward multiuser data visualization and thus provision of the observed parameters to all involved disciplines without delay. Looking forward, the obtained information can be fed into the “observational method,” the feedback loop suggested by Peck (1969), confirming or refuting the design assumption in a “learn-as-you-go” procedure. This allows timely modification of processes to optimize construction progress and reduce the hazard risks. An additional advantage of the presented solutions is that the initial installation (“ab initio” system) can be expanded to multisensor, multipurpose systems by successively adding sensors in response to changing conditions and requirements (a “best-way-out” method). The ease of use of the system promotes the common and efficient utilization of automated methods in physical and commercial risk mitigation, even at early stages, by making “deformation monitoring an integral part of decision making in the design-construction-supervision maintenance system” (Kavvadas, 2005). (References available from the author.)

Wireless optical displacement sensor for convergence and divergence monitoring

To view the entire article you must be a member of UCA.

- Coming Soon -

Tunnel Demand Forecast – June 2023

New UCA chair sees a bright future for underground construction industry

New UCA Chair looks ahead to an exciting two-year term

Environmental Product Declarations —What they mean for the tunneling industry

Gas extraction and in situ oxidation for TBM tunneling of the Purple Line Extension, Section 1, Los Angeles

41 workers trapped for 17 days are rescued in India

Breakthrough at DC Water’s Northeast Boundary Tunnel

Atlas Copco and ABB collaborate on mobile integration system for underground mines

United States to host World Tunnelling Congress in 2016

Anglo American’s Grosvenor project launches its tunnel boring machine

Environmental Product Declarations —What they mean for the tunneling industry

Hybrid tunnel boring machine takes on Australian decline tunnel

To view the entire article you must be a member of UCA.

- Coming Soon -

Caterpillar announces R2900 XE Diesel Electric Underground Loader

Amtrak applies for $7.3 billion in federal grants for infrastructure upgrades

Related Articles

Environmental Product Declarations —What they mean for the tunneling industry

Gas extraction and in situ oxidation for TBM tunneling of the Purple Line Extension, Section 1, Los Angeles

Rescue efforts slowed by difficult ground conditions at collapse in India

Liebherr R 930 Tunnel crawler excavator launched on the world market

Tunnel Demand Forecast – June 2023

New UCA chair sees a bright future for underground construction industry

New UCA Chair looks ahead to an exciting two-year term

Environmental Product Declarations —What they mean for the tunneling industry

Gas extraction and in situ oxidation for TBM tunneling of the Purple Line Extension, Section 1, Los Angeles

41 workers trapped for 17 days are rescued in India

Breakthrough at DC Water’s Northeast Boundary Tunnel

Atlas Copco and ABB collaborate on mobile integration system for underground mines

United States to host World Tunnelling Congress in 2016

Anglo American’s Grosvenor project launches its tunnel boring machine

Environmental Product Declarations —What they mean for the tunneling industry

Hybrid tunnel boring machine takes on Australian decline tunnel