New Wheel-Based Magnetic Robot Designed for Oil Tank Inspection
A team of researchers from Qingdao University of Science & Technology has developed a new type of wall-climbing robot designed specifically for inspection and maintenance tasks on the exterior surfaces of large oil storage tanks. The robot, which combines mobility, strong magnetic adhesion, and structural efficiency, aims to address the persistent challenges of safety, load capacity, and adaptability in hazardous industrial environments. The study, led by ZOU Yujing, MA Benxiao, PANG Feng, and FAN Zhimin, was recently published in the journal China Mechanical Engineering under the title “Structural Optimum Design and Adsorption Capacity Research for Oil-Tank Wall-Climbing Robot.” The work introduces a comprehensive approach to robot design that integrates theoretical modeling, finite element analysis, and experimental validation to ensure both performance and reliability.
The demand for automated inspection systems in high-risk industrial settings—particularly in oil and gas, chemical, and nuclear facilities—has grown significantly in recent years. Manual inspection of vertical surfaces such as oil tank walls is not only time-consuming but also exposes workers to extreme dangers, including falls, exposure to toxic substances, and fire hazards. As a result, there has been a strong push toward robotic solutions that can operate autonomously on vertical or even inverted surfaces while carrying inspection tools, cameras, or repair equipment. However, existing wall-climbing robots often face limitations in terms of weight, power consumption, adaptability to surface curvature, and the balance between adhesion strength and mobility.
Most current wall-climbing robots rely on one of three primary adhesion mechanisms: vacuum suction, negative pressure systems, or magnetic attraction. While suction-based systems work well on smooth, non-porous surfaces, they require continuous power to maintain vacuum and are prone to failure on rough or contaminated surfaces. Negative pressure systems face similar limitations and often require complex sealing mechanisms. Magnetic adhesion, on the other hand, is particularly effective on ferromagnetic surfaces such as steel oil tanks, offering passive, energy-efficient attachment without the need for continuous power input. Permanent magnet systems, in particular, are favored for their reliability and durability in harsh environments.
Despite these advantages, magnetic wall-climbing robots have historically struggled with trade-offs between strong adhesion and ease of movement. Excessive magnetic force increases friction and energy consumption, making locomotion difficult, while insufficient force risks slippage or detachment. Additionally, many existing magnetic robots are either too bulky, lack sufficient load capacity, or are limited in their ability to navigate obstacles or curved surfaces. The research team from Qingdao University of Science & Technology set out to overcome these limitations by designing a compact, lightweight, yet highly capable wheel-based robot that leverages optimized permanent magnet arrays for stable adhesion.
The newly developed robot features a modular design with a central chassis, dual DC geared motors, synchronous belt drive systems, and four wheels for locomotion. At the base of the robot, four neodymium-iron-boron (NdFeB) permanent magnets are arranged in a 2×2 configuration, mounted on a soft iron yoke to enhance magnetic flux concentration and reduce leakage. The magnets, rated N42, are known for their high remanence and energy density, making them ideal for applications requiring strong magnetic fields in compact spaces. The entire magnetic assembly is separated from the tank surface by a small air gap, which is carefully calibrated to balance adhesion strength with operational clearance, particularly over weld seams and surface irregularities.
One of the key innovations in this work is the integration of static and dynamic modeling to determine the minimum required adhesion force for safe operation under various inclinations. The researchers conducted a detailed static analysis to evaluate the robot’s stability against two primary failure modes: sliding and overturning. By establishing equilibrium conditions for forces and moments acting on the robot, they derived mathematical expressions for the critical adhesion force needed to prevent slippage as the wall’s inclination angle varies. Using MATLAB for numerical simulation, they found that the maximum required adhesion occurs at an inclination of 38.7 degrees, where each magnet must generate at least 80 newtons of force—resulting in a total required adhesion of 320 newtons for the four-magnet system.
To assess the risk of overturning, the team performed a moment balance analysis around the front wheel contact point. The results showed that the highest overturning risk occurs at approximately 78.4 degrees, requiring a much lower adhesion force of about 51 newtons per magnet. This finding indicates that slippage, not overturning, is the dominant constraint in the design. Therefore, ensuring sufficient adhesion to prevent sliding inherently guarantees stability against tipping. This insight allowed the researchers to focus their optimization efforts on maximizing frictional resistance rather than structural reinforcement against rotation.
In addition to static stability, the dynamic performance of the robot was analyzed to understand its acceleration, traction, and energy requirements during upward and downward motion. The model accounts for motor torque, transmission efficiency, rolling resistance, and gravitational components along the direction of motion. The analysis revealed that the robot’s ability to climb depends on the net traction force exceeding the sum of gravitational and frictional resistances. During descent, gravity assists motion, reducing the required motor input but increasing the importance of controlled braking to prevent uncontrolled sliding. These dynamic considerations informed the selection of motor specifications and gear ratios to ensure reliable operation across different inclinations and payloads.
To optimize the magnetic adhesion system, the researchers employed finite element analysis (FEA) using Ansys Maxwell, a high-fidelity electromagnetic simulation software. The goal was to determine the optimal geometric parameters of the magnetic circuit, including the air gap between the magnets and the tank surface, the thickness of the soft iron yoke, and the spacing between individual magnets. Each of these parameters has a significant impact on magnetic flux distribution, reluctance, and ultimately, the magnitude of the attractive force.
The air gap, defined as the distance between the magnet surface and the steel wall, was found to be the most sensitive parameter. As the gap increases, the magnetic field strength drops rapidly due to the inverse square relationship between force and distance in magnetic circuits. Simulations showed that at a 5 mm gap, the total adhesion force reached 1,383 newtons for one magnetization configuration and 1,336 newtons for another. At 15 mm, the force dropped to around 400 newtons—still sufficient for stability but significantly reduced. Given that typical weld seams on oil tanks are 8 to 10 mm high, the team selected an operational gap of 11 mm to avoid mechanical interference while maintaining adequate adhesion.
The yoke thickness was another critical factor. The yoke serves to complete the magnetic circuit, guiding flux from the magnet into the steel wall and minimizing leakage. Too thin a yoke saturates magnetically, limiting flux transfer, while too thick a yoke adds unnecessary weight without significant performance gains. To quantify this trade-off, the researchers introduced a figure of merit called the “magnetic-to-mass ratio” (γ), defined as the adhesion force per unit mass of the yoke. Simulation results showed that γ peaks at a yoke thickness of 2 mm, after which further thickening yields diminishing returns. At 2 mm, the yoke provides sufficient cross-sectional area to carry the magnetic flux without saturation, while keeping the overall system lightweight. This optimization directly contributes to the robot’s energy efficiency and agility.
Magnet spacing was also systematically varied to explore its effect on force output. When magnets are placed too close together, their fields interfere, leading to flux crowding and suboptimal utilization of the yoke area. When spaced too far apart, the magnetic circuit becomes less efficient, and flux leakage increases. The simulations revealed that adhesion force increases with spacing up to a point—120 mm for one configuration and 90 mm for another—after which it begins to decline. The optimal spacing of 100 mm was selected as a compromise that maximizes force while fitting within the robot’s chassis dimensions. At this spacing, the magnetic flux is evenly distributed across the yoke, resulting in a broad and uniform pressure profile on the tank surface.
With the optimal parameters determined—11 mm air gap, 2 mm yoke thickness, and 100 mm magnet spacing—the researchers constructed a prototype and conducted experimental validation. A custom test rig was built to measure the actual adhesion force under controlled conditions. The setup included a rigid aluminum frame, a Q235 steel plate (30 mm thick) mounted on a height-adjustable platform, and precision load sensors to capture the vertical force exerted by the robot’s magnetic base. By slowly lowering the steel plate toward the magnets and recording the force at various distances, the team obtained empirical data for comparison with simulation results.
The experimental results showed a strong correlation with the finite element predictions in terms of trend and relative magnitude. Both curves exhibited the expected inverse relationship between adhesion force and air gap, with force decreasing rapidly as distance increased. However, the measured forces were consistently lower than the simulated values—by approximately 25% on average. The researchers attributed this discrepancy to real-world factors not fully captured in the simulation, such as minor misalignments, surface roughness, edge effects, and magnetic flux leakage into the surrounding environment. Unlike the idealized simulation, where boundary conditions assume perfect magnetic insulation, the physical setup is exposed to ambient air, which, while non-magnetic, still allows some degree of flux dispersion.
To account for this difference, the team proposed a correction factor of 0.75 to scale the simulation results for practical design purposes. This adjustment ensures that future iterations of the robot will have sufficient safety margin without over-engineering the magnetic system. The validation process not only confirmed the accuracy of the modeling approach but also highlighted the importance of empirical testing in robotic development, especially for systems relying on complex physical interactions like magnetic adhesion.
The successful integration of theoretical analysis, simulation, and experimentation underscores the robustness of the design methodology. By combining static and dynamic models with electromagnetic FEA and physical testing, the researchers were able to create a robot that is not only functional but also optimized for real-world performance. The final prototype is compact, lightweight, and capable of carrying inspection payloads while maintaining secure attachment to vertical and inclined steel surfaces.
Looking ahead, the team plans to enhance the robot’s capabilities by incorporating sensors for autonomous navigation, obstacle detection, and surface inspection. Future versions may include ultrasonic thickness gauges, cameras for visual inspection, or even robotic arms for minor repairs. The modular design allows for easy integration of additional components without compromising the core adhesion and mobility functions.
This research contributes to the growing field of industrial robotics by offering a practical, efficient, and scalable solution for infrastructure inspection. As industries continue to prioritize safety, efficiency, and automation, robots like this one will play an increasingly important role in maintaining critical assets. The work also demonstrates the value of interdisciplinary collaboration, bringing together mechanical engineering, electromagnetics, materials science, and control systems to solve complex real-world problems.
The findings have implications beyond oil tank inspection. Similar magnetic wall-climbing robots could be adapted for use in ship hull maintenance, bridge inspection, wind turbine servicing, and even space station repairs, where secure adhesion in zero-gravity or high-risk environments is essential. The optimization principles—balancing adhesion, weight, and mobility—are universally applicable to any robot designed to operate on vertical or inverted surfaces.
In conclusion, the research team from Qingdao University of Science & Technology has developed a novel wheel-based magnetic wall-climbing robot that meets the demanding requirements of industrial inspection tasks. Through rigorous modeling, simulation, and experimental validation, they have demonstrated a design that is both theoretically sound and practically viable. The robot’s optimized magnetic adhesion system ensures reliable performance across a range of inclinations and surface conditions, while its lightweight and compact form factor enhances maneuverability and energy efficiency. This work represents a significant step forward in the development of autonomous inspection systems for hazardous environments.
ZOU Yujing, MA Benxiao, PANG Feng, FAN Zhimin, Structural Optimum Design and Adsorption Capacity Research for Oil-Tank Wall-Climbing Robot, China Mechanical Engineering, DOI: 10.16731/j.cnki.1671-3133.2021.01.006