J U L Y 2 0 2 1 • M A T E R I A L S E V A L U A T I O N 735 Testing The development of the crawler system started with creating general concepts and initial prototyping and proceeded with bench-scale testing and engineering-scale testing of the system. This section presents the bench-scale testing used to validate the concepts and testing of the system in a tube with multiple bends and straight sections. Bench-Scale Testing Robotic crawler: To evaluate the pull force capability of the crawler, pull-force tests were conducted on the gripper and extender modules. The grippers were found to be capable of pulling approximately 84.5 N of force and the extenders were found to generate 40 N of force. These values demonstrate that the forces obtained previously in this paper are reason- able approximations of the actual pull forces. The pull-force tests were conducted using a digital weight scale attached to the ends of the modules. The value for the gripper was found by finding the maximum pull force before the gripper pads began to slip along a steel 5 cm diameter tube. The pull force for the extender was found by clamping the module to a flat surface and allowing the linear actuator to pull the scale. After the initial system was assembled, bench-scale tests were conducted to evaluate the crawler’s ability to navigate through both straight and 180° elbow sections. The elbow testing was conducted in a custom-built 5 cm diameter tube with a 7 cm bend radius. Although the system was designed to navigate through bends with a 5 cm radius of curvature, the smallest bend radius found in a transparent acrylic pipe was 7 cm. With this experimental setup, the crawler was easily able to navigate through the bend. The average speed of the crawler in straight sections was found to be 50 cm/min. The speed was slightly slower when navigating through the bends. Ultrasonic transducer module: To demonstrate the functionality of the ultrasonic transducer module, tests were performed in a clear acrylic 5 cm diameter straight tube with a wall thickness of 1.6 mm. The ultrasonic transducer gauge was calibrated to measure the thickness of PVC by adjusting the velocity of sound to 2390 m/s. Since the module was not inte- grated with the crawler for the bench-scale testing, the stabi- lization mechanism was adapted to be used at both ends of the module. Wall thickness was measured at three different locations around the inner circumference of the tube. The circumferential rotation using the spur gear set allowed the module to obtain measurements at different loca- tions along the inner wall of the tube. As shown in Figure 10, TABLE 1 Sensor specifications Sensor Measurement Range Resolution Unit Environmental • Temperature –40 to 85 ±1.0 °C • Pressure 26 to 126 ±0.02 kPa Inertial measurement • Acceleration ±2 to ±16 ±0.004 g • Angular velocity ±125 to ±2000 ±10.0 °/s Camera • Surface imaging 640 × 480 VGA pixel LiDAR • Circumferential mapping 10 to 60 ±1.0 mm Figure 10. Measurements performed on the tube’s inner surface showing the rotation of the crawler.

736 M A T E R I A L S E V A L U A T I O N • J U L Y 2 0 2 1 the measurements were consistent between 2.5 and 2.6 mm. As noted previously, the flat sensor head does not mate perfectly with the internal tube surface due its curvature. Thus, an offset must be subtracted from the measurement to obtain a more accurate reading. Instrumentation module: To evaluate the LiDAR sensor in the instrumentation module, a template ring was used to simulate a 5 cm diameter surface with a variety of irregularities. During the testing, the module is positioned at the center of the template frame and rotates to scan the surrounding irregularities on the ring. Figure 11 shows the template ring and the instrumentation module positioned at the center. Preliminary results demonstrate the potential for the detection of anomalies in tubes and pipes using a LiDAR sensor. Data from the environmental sensor is also shown and includes pressure (p), altitude (a), and temperature (t). It should be noted that the camera was not installed during this testing. Engineering-Scale Testing To evaluate the crawler’s performance in a larger-scale testbed, a mock-up similar to the superheater tubes found in fossil energy power plants was constructed. The testbed was manufactured with acrylic plastic tubes so the crawler could be observed while navigating through the straight sections and 180° elbows. As shown in Figure 12, the crawler was able to navigate through multiple straight pipe sections and bends and was limited only by the length of the tether. An additional testbed is being constructed with metal tubes and contains sharper bends, as well as 90° elbows. This setup will not allow for visualization of the crawler but will provide a more realistic testbed to evaluate its pull force capability. ME TECHNICAL PAPER w robotic inspection of small-diameter superheater pipes 270 240 210 300 330 30 60 90 120 150 180 0 Radius (mm) Figure 11. Instrumentation module testing: (a) module placed in template ring (b) template ring with anomalies (c) distance measurements obtained in template ring. (a) (b) (c) Figure 12. Crawler navigating the superheater tube mock-up with magnified images.