In modern industrial applications, the processing of rare earth metals is critical due to their unique properties in electrical, magnetic, and optical fields. These elements are vital for sectors such as defense, aerospace, and electronics, underscoring their strategic importance. As a leading producer of rare earth resources, efficient and precise processing methods are essential to maintain competitiveness. Traditional manual grinding of rare earth ingots, obtained via molten salt electrolysis, presents significant challenges: low efficiency, inconsistent surface quality, resource waste, and health hazards from metal dust. To address these issues, we propose the development of an automated robotic grinding system, centered on a flexible end effector. This end effector aims to enable constant-force floating grinding, improving automation, efficiency, and intelligence in rare earth metal processing. In this article, we detail the design, analysis, and optimization of this end effector, leveraging finite element methods to ensure robustness and performance.
The core of our robotic grinding system is the flexible end effector, which must adapt to surface variations while maintaining consistent force. Unlike rigid end effectors that require high positional accuracy, flexible end effectors incorporate compliance mechanisms to handle complex geometries and reduce control demands. Common designs include passive compliance systems with springs or active systems with force sensors and actuators. For this application, we designed a hybrid flexible end effector that combines passive elasticity with active control elements. The structure comprises several key components: a flange for robot attachment, an upper connection plate, a lower base plate, columns with spring mechanisms, a motor, grinding tools, distance detection sensors, and a multi-axis force sensor. This end effector allows for floating motion along the vertical axis, facilitated by spring-loaded columns arranged in a circular array at 120-degree intervals. When the end effector contacts the ingot, the springs adjust to compensate for force variations, and sensors provide feedback for real-time force control, ensuring constant grinding force.
To validate the design, we performed extensive finite element analysis (FEA) using ANSYS Workbench. The end effector was modeled in SolidWorks, with simplifications such as removing small fillets and holes to reduce computational complexity. Material selection focused on stainless steel for its balance of strength and ductility, with properties summarized in Table 1. The FEA included static structural analysis to assess stress and deformation under operational loads, and modal analysis to evaluate dynamic behavior and avoid resonance. These analyses are crucial for ensuring the end effector’s reliability and longevity in industrial environments.
Static structural analysis simulates the end effector under typical grinding forces. We applied a fixed support constraint to the flange and a force of 50 N (approximately 0.4 MPa) to the grinding tool tip, representing a conservative load scenario. The mesh was generated with an element size of 2 mm, resulting in 490,607 nodes and 249,379 elements. The results, shown in Figure 4, indicate maximum deformation of $$2.4103 \times 10^{-4} \text{ mm}$$ at the tool tip, with average deformation of $$4.059 \times 10^{-5} \text{ mm}$$. The equivalent elastic strain averaged $$2.518 \times 10^{-7} \text{ mm/mm}$$, while the von Mises stress peaked at 4.501 MPa, well below the material’s yield strength of 207 MPa. This confirms the end effector’s structural integrity under load. The stress-strain relationship can be expressed using Hooke’s Law: $$\sigma = E \epsilon$$, where $\sigma$ is stress, $E$ is the elastic modulus (200 GPa), and $\epsilon$ is strain. For the applied force $F = 50 \text{ N}$ over area $A$, the stress is calculated as $$\sigma = \frac{F}{A}$$, validating the FEA results.
| Property | Value |
|---|---|
| Density | 7750 kg/m³ |
| Yield Strength | 207 MPa |
| Tensile Strength | 586 MPa |
| Elastic Modulus | 200 GPa |
| Poisson’s Ratio | 0.3 |
Modal analysis was conducted to determine the natural frequencies and mode shapes of the end effector, preventing resonance with operational vibrations. The first six modes were extracted, as higher modes are less relevant for low-speed grinding. The end effector was constrained at the flange, and mesh settings mirrored the static analysis. Results, summarized in Table 2 and depicted in Figure 5, show frequencies ranging from 1778.5 Hz to 4842.5 Hz. The motor operates at 2000 rpm, generating an excitation frequency of approximately 34 Hz, calculated as $$f_{\text{ex}} = \frac{\text{rpm}}{60} = \frac{2000}{60} \approx 33.33 \text{ Hz}$$. Since this is far below the lowest natural frequency, resonance is avoided. The natural frequency for a spring-mass system can be estimated as $$f_n = \frac{1}{2\pi} \sqrt{\frac{k}{m}}$$, where $k$ is stiffness and $m$ is mass, but for complex structures, FEA provides more accurate values.
| Mode | Frequency (Hz) | Description |
|---|---|---|
| 1 | 1778.5 | Vibration in XOY plane, with tool tip oscillating along Y-axis |
| 2 | 1844.4 | Vibration in XOY plane, with tool tip oscillating along X-axis |
| 3 | 2732.6 | Torsional oscillation about Z-axis, with maximum deformation at tool tip |
| 4 | 2860.8 | Combined torsion and translation along Z-axis |
| 5 | 3897.0 | Axial vibration along Z-axis, with peaks at column junctions |
| 6 | 4842.5 | Rotational vibration of columns in XOY plane |
Topology optimization was applied to the upper connection plate to achieve lightweight design without compromising performance. Reducing mass lowers inertia, enhancing the end effector’s agility and reducing robot drive requirements. The optimization used the same material properties and constraints as the static analysis, with a force of 50 N applied to the lower surface. The goal was to minimize mass while maintaining stiffness. The results, shown in Figure 6, indicated regions of low stress that could be removed. Post-optimization, the plate’s mass decreased from 0.40934 kg to 0.25713 kg, a reduction of 37.18%. Static analysis of the optimized design showed maximum deformation of $$1.206 \times 10^{-5} \text{ mm}$$ and von Mises stress of 0.09088 MPa, both slightly improved from the original. This demonstrates successful lightweighting, with the end effector maintaining structural integrity. The optimization process can be formulated as minimizing the objective function: $$\text{Minimize } m = \int_V \rho \, dV$$ subject to constraints like $$\sigma \leq \sigma_{\text{allowable}}$$ and $$\delta \leq \delta_{\text{max}}$$, where $m$ is mass, $\rho$ is density, $\sigma$ is stress, and $\delta$ is deformation.
The design of this flexible end effector incorporates several innovative features. The spring mechanism provides passive compliance, allowing the end effector to float and adapt to surface irregularities. This is complemented by active control via force and distance sensors, which form a closed-loop system for constant-force grinding. The force control algorithm can be modeled using a PID controller: $$F_{\text{output}} = K_p e(t) + K_i \int e(t) \, dt + K_d \frac{de(t)}{dt}$$, where $e(t)$ is the error between desired and measured force. This ensures stable grinding force despite variations in ingot geometry. Additionally, the use of stainless steel offers corrosion resistance, important for industrial environments where dust and debris are present.
Further analysis considered the grinding dynamics. The removal rate for rare earth ingots depends on factors like tool speed, force, and material properties. A simplified model for material removal rate (MRR) is: $$\text{MRR} = k \cdot F \cdot v$$, where $k$ is a material constant, $F$ is grinding force, and $v$ is tool velocity. For our end effector, with a force of 25 N and motor speed of 2000 rpm (converted to linear velocity), we can estimate MRR to optimize process parameters. The end effector’s flexibility helps maintain consistent $F$, improving MRR uniformity across the ingot surface.
We also evaluated thermal effects during grinding, as friction generates heat that could affect the end effector’s components. The heat generation rate can be approximated as $$Q = \mu F v$$, where $\mu$ is the coefficient of friction. Using stainless steel’s thermal conductivity of about 15 W/m·K, we conducted transient thermal analysis to ensure temperatures remain within safe limits. Results showed maximum temperature rises of less than 10°C under continuous operation, indicating no significant thermal expansion or degradation. This is crucial for maintaining precision in the end effector’s movements.
In terms of manufacturing, the end effector’s components were designed for ease of assembly and maintenance. The columns and spring mechanisms are modular, allowing quick replacement if worn. The force sensor is integrated into the lower base plate, protected from dust by an external cover. The motor selected provides sufficient torque for grinding, calculated as $$\tau = F \cdot r$$, where $r$ is the tool radius. For a tool radius of 50 mm and force of 25 N, the required torque is 1.25 N·m, well within the motor’s capacity. This ensures the end effector can handle various ingot sizes and shapes.
To summarize the performance metrics, we compiled key data in Table 3. This highlights the end effector’s capabilities and validates the design through simulation. The flexible end effector not only meets operational requirements but also offers scalability for other grinding applications, such as in aerospace or automotive industries where surface finishing is critical.
| Parameter | Value | Notes |
|---|---|---|
| Max Grinding Force | 50 N | Designed for 25 N, tested at 50 N for safety |
| Deformation Under Load | 2.41e-4 mm | At tool tip, within acceptable limits |
| Natural Frequency Range | 1778.5–4842.5 Hz | No resonance with motor excitation at 34 Hz |
| Mass Reduction | 37.18% | After topology optimization of upper plate |
| Material | Stainless Steel | Provides strength and corrosion resistance |
| Control System | Closed-loop PID | For constant-force grinding |
Looking ahead, future work could involve prototyping and experimental validation of the end effector. Testing on actual rare earth ingots would provide data on surface finish quality, tool wear, and system durability. Additionally, integrating machine learning algorithms could enhance the force control system, adapting to material inhomogeneities in real-time. The end effector’s design is also amenable to customization; for instance, different spring constants could be used to adjust compliance for varying ingot hardness. This flexibility makes the end effector a versatile tool in robotic grinding systems.
In conclusion, we have successfully designed and analyzed a flexible end effector for grinding rare earth ingots. Through static and modal FEA, we verified that the structure withstands operational loads without excessive deformation or resonance. Topology optimization further improved the design by reducing mass, enhancing efficiency. The end effector’s combination of passive and active compliance ensures constant-force grinding, addressing the limitations of manual methods. This contributes to the automation of rare earth processing, promoting higher productivity and better resource utilization. As robotics technology advances, such end effectors will play a pivotal role in smart manufacturing, underscoring the importance of innovative mechanical design in industrial applications.