Navigation Structures: Robotic platform for the detection of corrosion and protective coating thickness above and below water

Transcription

1 9 th European Workshop on Structural Health Monitoring July 10-13, 2018, Manchester, United Kingdom Navigation Structures: Robotic platform for the detection of corrosion and protective coating thickness above and below water More info about this article: James A. Evans 1, Quincy G. Alexander 2, James R. Tallent 3, Richard D. Brown 4 and Matthew D. Smith 5 1 US Army Engineer Research and Development Center, United States, James.A.Evans@usace.army.mil 2 US Army Engineer Research and Development Center, United States, Quincy.G.Alexander@usace.army.mil 3 US Army Engineer Research and Development Center, United States, James.R.Tallent@usace.army.mil 4 US Army Engineer Research and Development Center, United States, Richard.D.Brown2@usace.army.mil 5 US Army Engineer Research and Development Center, United States, Matthew.D.Smith@usace.army.mil Abstract Inland navigation structural elements, such as lock and dam miter gates, are often made of steel and are therefore highly susceptible to corrosion, pitting and fatigue cracks after many years in service. To mitigate the corrosion issue, protective anti-corrosion epoxy coatings are applied to the steel surface before the structure goes into operation, and reapplied following scheduled inspections where coating loss is detected. The inspection process is typically a costly and potentially dangerous endeavour due to the need to inspect submerged areas of the structure where structural and environmental conditions may be at their worse. Divers are often used when dewatering is too costly or in some cases not possible. A safer, more reliable, and more cost effective method of performing the corrosion/coatings inspection process is needed. This paper outlines research currently being performed by the US Army Engineer Research and Development Center (ERDC) to address this need. The objectives of the research are to improve inspection area coverage, reduce inspection subjectivity, reduce time, reduce costs, and reduce risk to human life. To achieve these objectives, an array of sensors are being employed. Magnetic leakage flux sensor arrays and ultrasonic sensor arrays are being developed. Both technologies will consist of multiple sensor heads allowing broader tracks to be used for assessing the condition of the structure. These sensors will scan for material loss due to corrosion, and will also be able to identify cracks. In addition to detecting corrosion, the ultrasonic sensor array will measure the remaining coating thickness. Additionally, robotic platforms that can support the described sensor systems above and below the water surface are currently being developed by the ERDC. Under the direction of pre-programming, and artificial intelligence (AI), the robot will be able to climb vertical steel walls and use the on-board sensors to collect corrosion/coating thickness data. When under water (fresh water) and above water, the sensor platform will transmit the data to a central location on the surface, where the data will be assembled into a navigational site map for the structure. This will be an additional source of data to feed into the overall structural health monitoring scheme for US Army Corps of Engineers managed infrastructure. Creative Commons CC-BY-NC licence

2 1. Introduction Lock chambers, such as the chamber shown in Figure 1, are critical for navigation along the rivers, as they allow for smooth transitions in elevations. Many navigation lock and dam structural elements are being used well beyond their design life. These components cannot easily be replaced, due to budgetary constraints. Therefore, inspections are critical to understanding the condition of components so that the proper operation and maintenance decisions can be made to extend the useful life of the structures. Figure 1. Barge entering navigation chamber located along the Mississippi River Miter gates, as shown in Figure 2, are the watertight doors (or leafs) that seal off the chamber from the upper and lower reaches of the lock and allow for the control the water level in the lock chamber; thus, allowing ship traffic to safely navigate between stretches of water with significant changes in elevation. These large gates are primarily made of steel and the lower portions of the gates are always underwater during periods of operation. Inspecting miter gates below the waterline for corrosion and inspecting the conditions of protective coatings is a costly effort, since it often requires the emptying of the lock chamber, which disrupts navigation. An alternative option for inspection involves using divers to perform underwater inspections, which can be dangerous and the results are often inconsistent due to limited visibility. In many cases, the inspections are simply not being performed frequently and the condition below the waterline is left unknown. Without sufficient condition information, these valuable assets cannot be adequately managed and progressing deterioration can result in failures that severely impact navigation capability, user safety, and the surrounding economies that depend on the shipment of goods. Identifying problems early allows for maintenance prioritization and the ability to program repair funds in a timely manner. 2

3 Figure 2. Example Miter Gates. (a) Upstream view of miter gates in dewatered chamber and (b) downstream view of miter gates in operational chamber 2. Miter Gate Operational Conditions Miter gates are divided into three (3) zones defined by the lower and upper pool elevations, namely: the atmospheric zone, the splash zone and the submerged zone, as shown in Figure 3. The atmospheric zone is where the gate is only exposed to air, the splash zone is where the gate is exposed to both water and air as the water in the chamber is raised and lowered, and the submerged zone is where the gate remains constantly underwater. Studies indicate that material loss due to corrosion is normally greatest in the splash zone, but the submerged zone cannot be ignored, since structural stresses on the gate are greatest here. 3. Inspection System Design Figure 3. Miter Gate Inspection Zones A safer, more reliable, and more cost effective method of performing the corrosion/coatings inspection process is needed. In response to this need, the US Army Engineer Research and Development Center (ERDC) began an effort to develop an 3

4 automated inspection system. The inspection system design consists of three primary components; namely, the corrosion inspection sensor, the protective coating thickness measurement sensor, and the Multi-Sensor Inspection Platform (MSIP) Corrosion Inspection Sensor The ability to quantify the amount of corrosion that exists on these partially submerged steel miter gates in the splash and submerged zones will help with maintenance planning, and help extend the life of the structures Operational Criteria The following list of key operational criteria was developed for the inspection sensor packages. Sensor Design Objectives: Measure without making contact with steel surface Measure while moving over steel surface Measure both steel surface and internal steel structure Measure over relatively wide area (not single point) Minimum Sampling Rate (10 Hz) Relatively low cost sensor array Relatively low power Robust Operable in both air and water Operable on dirty steel surface (organics) Based on a review of existing non-destructive test (NDT) methods for corrosion assessment, magnetic flux leakage technology (MFL) was selected for development of the corrosion sensor. This technology satisfied operational criteria outlined above Magnetic Flux Leakage Sensor Development MFL techniques use permanent magnets to temporarily magnetize a steel test substrate. The basic concept is shown in Figure 4. The magnetic flux flowing through the steel substrate is uniform when there are no flaws in a metal substrate. If surface or internal flaws are present, such as pitting, corrosion or other forms of damage, the magnetic flux lines are distorted. This distortion or leakage can be measured by a magnetic field sensor located adjacent to the test point. MFL methods can be used to detect both nearand far-surface defects in ferrous materials. Due to the nature of the proposed method, obstructions caused by non-ferrous materials do not affect the measurements. MFL is currently recognized as an industry standard for metallic pipe wall assessment. Many commercial products are available that employ this technology in dry environments. 4

5 Figure 4. MFL Theory of Operation The skin plate of typical miter gates used in navigational structures is approximately 12.7 mm thick. To assure that MFL penetrates this material thickness, the Finite Element Method Magnetics (FEMM) software package was used to simulate the interactions between the magnetic fields and the steel. FEMM solves two-dimensional models using Maxwell s equations. A thick arched shape was selected for the magnetic flux bridge. The arched shape of the flux bridge keeps magnetic fields from crossing between the flux bridge and the test surface. This increases the magnetic flux density inside the test plate under the sensors and reduces the pull force from the device. A thick material was selected to provide more cross sectional area for the magnetic flux to pass through, giving an easier path for the flux to travel. Flatter vs taller magnets were found to produce the most increase in magnetic flux density in the steel while minimizing the pull force from the device. The large flux bridge did however increase the size and weight requirements of the overall platform, as compared to other options. Numerous commercially available magnetic flux sensors were evaluated for performance, stability, and range of sensitivity. The selected sensors were further analysed using a 6.35 mm thick plate with machined defects, as shown in Figure 6. This test plate was scanned using an X-Y table to maximize stability as the sensor is moved across the steel surface. Three-dimensional sensor data was collected and analyzed for each of the magnetic flux detector types tested, and a sample output from the test series is shown in Figure 5. After extensive laboratory test scan data collection and analyses, the Melexis MLX90393 Triaxis magnetometer was selected as the sensor of choice for determining plate corrosion. This sensor supports selectable resolutions, wide magnetic operating range, and expandability due to Inter-Integrated Circuit (I2C) or Serial Peripheral Interface (SPI) communication protocol. An additional test series was performed using the selected sensor on a plate sample with various types of welds. The Melexis MLX90393 Triaxis magnetometer successfully detected complete penetration welds, connecting two plates, as well as spot welds on the back side of the plate. A sample of these test results is provided in Figure 6. 5

6 Figure 5. Sensor Response Figure 6. 3-Dimensional Image of Magnetic Flux Data 3.2. Protective Coating Thickness Measurement Sensor Understanding the thickness of protective coating that remains on submerged steel structures, such as the miter gates, will help with budget development for future maintenance. As such, a sensor system is also being designed to measure the thickness of anti-corrosion coating on a metal substrate Operational Criteria The operational criteria outlined in Section also applies to the anti-corrosion thickness assessment sensor. Of the existing technologies for anti-corrosion thickness assessment of non-metallic coatings on metal substrates investigated, only two (Ultrasonic Pulse-Echo sensor, and Low-Coherence Interferometry sensor) do not 6

7 require direct contact with the substrate. Ultimately, the Ultrasonic Pulse-Echo sensor was considered to be most appropriate for the above- and below-water conditions, and the even harsher splash zone where material loss due to corrosion is known to be greatest Ultrasonic Pulse-Echo Sensor Development The proposed system includes an array of piezo elements for emitting an ultrasonic signal into the coating, and an array of piezo elements capable of receiving the returned signal, as shown in Figure 7. For ultrasonic coating thickness measurement, a high-frequency ultrasonic pulse is generated by the pulse echo system. A portion of the signal reflects off the coating, and a fraction penetrates the coating and is reflected off the coating/steelstructure interface. The thickness measurement is then based upon the time difference between the return waveform, seen as the first echo in Figure 8. Because the coating is very thin, a very short pulse is required to avoid blanking-out the initial pulse. To assure the separation of the front and back of the coating, a 10 MHz target frequency was chosen for the initial design of the ultrasonic scanning system. The proposed frequency is still being evaluated, as it may require a higher frequency (shorter pulse) due to variations in the coating thicknesses. Figure 7. Overview of the Coating Thickness Measurement System. 7

8 Figure 8. Received voltage from piezo-film versus time shows echo response Multi-Sensor Inspection Platform To properly deploy the corrosion inspection sensor and the protective coating measurement sensor, a platform must be developed that is capable of traversing the vertical surface of the steel structure below and above water, while maintaining a set clearance from the structure Operational Criteria The following list of key operational criteria was developed for the MSIP MSIP Design Objectives: Operate above and below water surface (2 bar) Climb steel walls (in and out of water) on dirty surface Make lateral moves or small radius turns (skid-steer like) Maneuver over small objects (12.7 mm) Move at a constant speeds (non-stop) Semi-autonomous operation MSIP Development The current mobile platform prototype design, shown in Figure 9, was designed around the requirements of the MFL sensor package. The payload of this platform is approximately 23 kg and the linear inspection width of the MFL sensor package is approximately 15 cm. This mobile platform was designed to operate primarily on inclined and flat surfaces, such as the case of the upstream-side of the miter gate. The primary obstacle on the upstream-side of the miter gate is the rivet; therefore, the floor 8

9 clearance of this design is set at 1.25 cm. The floor clearance of the mobile platform can be increased while maintaining near 100% substrate saturation for steel plate thickness of 2 cm. or less. Figure 9. Multi-Sensor Inspection Platform Prototype with MFL Sensor Magnetic wheels were designed to allow for the platform to move vertically along the steel surface. The current magnetic wheel design (Figure 10) is a modified version of a commercially available off-the-shelf product. The commercially available components were modified by inserting magnets and reinforcing material inside the wheel. With this modification, sufficient robustness and magnetism was attained. No conventional machining is required for the wheel to be fitted with magnets and all parts snap into place with the force of the magnets and without the use of adhesives or fasteners. The structural wheel inserts are 3-D printed using an ABS material. Furthermore, the inserts do not require a tight tolerance nor do they require large quantities of material to print. The wheel design is low cost, easy to assemble, and available in various input shaft sizes and durometers. Figure 10. Magnetic Wheel Design 9

10 4. Conclusion The US Army Engineer Research and Development Center began development of a more cost-effective and safer means to inspect hydraulic steel structures, such as miter gates, for corrosion. The research indicated that magnetic flux leakage technology and ultrasonic pulse-echo technology were preferred for development of the corrosion and coating thickness sensors, respectively. These systems were tested and proved to be be a potential solution below and above the water line. A multi-sensor inspection platform was developed to deploy the sensors on the vertical steel components using magnetic wheels, while maintaining a consistent clearance. This project has shown the feasibility of the system, and further development will be required. 10