The rapid development of the economy and society has led to a consistent improvement in living standards and a correspondingly growing demand for dairy products. Traditional, semi-mechanized milking operations on farms are increasingly unable to meet daily production requirements. Consequently, the demand for fully automated solutions is rising year by year. The introduction of more intelligent and efficient milking robots has become imperative. These systems accelerate the process of farm智能化 (smart transformation), and the number of milking robots deployed is a significant indicator of a farm’s level of technological advancement.
Contemporary milking robots primarily utilize a clamping mechanism to grasp teat cups for attachment. The standard workflow involves a cow entering the milking stall, where a laser positioning system identifies its presence. The robot then controls a robotic arm to initiate the milking sequence. The arm moves to a predefined position, where a gripper (a component of the end effector) picks up a teat cup from a rack. The laser system provides real-time positioning of the cow’s udder, guiding the robot to maneuver the held cup beneath a specific teat for attachment. A significant limitation of current systems is their use of a single-gripper end effector. This design necessitates repeating the entire pick-and-place cycle four times per cow—once for each teat—which severely limits operational efficiency. Prominent models, such as those from certain American manufacturers, often employ industrial 6-degree-of-freedom arms, prized for their fast response and high precision, yet they still suffer from this sequential workflow bottleneck.
To address this core inefficiency, I have designed a novel end effector based on inventive problem-solving principles. This new design features dual gripping capability and integrated fine-angle adjustment mechanisms. The primary goal was to enhance the cup attachment success rate and overall speed. The end effector is responsible for retrieving teat cups from the rack and performing the attachment maneuver on cows in the stall. For successful attachment, the teat cup must be positioned directly below and coaxial with the cow’s teat. The designed end effector incorporates two perpendicularly arranged steering servos (yaw and pitch axes) that allow for micro-adjustments of the grippers’ orientation. Furthermore, a lead screw mechanism at the base enables precise adjustment of the distance between the two grippers. This combination allows the system to adapt to varying teat distances and udder conformations across different cows. A schematic representation of the mechanical design is provided below.
The operational concept relies on continuous real-time positioning data from the laser system feeding into the robot’s controller. As the robot arm moves, the two servos and the base motor on the end effector are actuated dynamically. This closed-loop control ensures the two held teat cups remain aligned with their target teats throughout the approach trajectory, significantly improving the first-attempt attachment rate. As the most critical component directly interacting with the cow and responsible for the core task, the reliability of the end effector is paramount. Any failure leads to immediate operational stoppage, resulting in economic losses for the farm. Therefore, a thorough structural reliability analysis is not just beneficial but necessary.
Structural Design and Material Selection for the Dual-Gripper End Effector
The mechanical design prioritizes stiffness, lightweight construction, and corrosion resistance. The core functional components and their selected materials are detailed below. Key considerations included the need to support the weight of the teat cups and associated milk tubes, and to withstand a small upward force applied during the final “lift” phase of attachment to ensure a proper vacuum seal.
| Component | Material | Yield Strength $\sigma_y$ (MPa) | Tensile Strength $\sigma_u$ (MPa) | Elastic Modulus $E$ (GPa) | Poisson’s Ratio $\nu$ | Density $\rho$ (kg/m³) |
|---|---|---|---|---|---|---|
| Flexible Gripper Jaw | AISI 304 (0Cr18Ni9) | ≥ 205 | ≥ 520 | 193 | 0.29 | 7930 |
| Gripper Body & Servo Frame | 6061 Aluminum Alloy | ≥ 110 | ≥ 205 | 68.9 | 0.33 | 2750 |
| Servo Housing & Connectors | S136 / Structural Steel | ≥ 250 | ≥ 460 | 200 | 0.30 | 7850 |
| Lead Screw & Base Plate | Structural Steel | 250 | 460 | 200 | 0.30 | 7850 |
The choice of AISI 304 stainless steel for the gripper jaws ensures durability and resistance to the humid, wash-down environments of a dairy farm. Aluminum alloy 6061 is used extensively for non-critical structural parts to minimize the overall mass of the end effector, thereby reducing the inertia load on the robotic arm. Structural steel is reserved for high-stress connection points and the servo housings. The interface bracket connecting the two perpendicular servo units is identified as a potential high-stress area due to complex loading from combined moments and is a focal point for analysis.
Comprehensive Static Structural Analysis Using Finite Element Method
Static analysis is an indispensable step in the design and validation process of any mechanical component. I employed the Finite Element Method (FEM), a powerful numerical technique for approximating solutions to boundary value problems. The fundamental principle of FEM is to subdivide a complex continuum into a finite number of smaller, simpler discrete elements (mesh). The behavior of the overall structure is then approximated by assembling the behaviors of these individual elements. The governing equations for linear static analysis are derived from the principle of virtual work and can be summarized as solving the global system equation:
$$ [K]\{u\} = \{F\} $$
where $[K]$ is the global stiffness matrix, $\{u\}$ is the vector of nodal displacements, and $\{F\}$ is the vector of applied nodal forces. The strain $\{\epsilon\}$ and stress $\{\sigma\}$ are subsequently calculated from the displacements using the strain-displacement matrix $[B]$ and the material constitutive matrix $[D]$:
$$ \{\epsilon\} = [B]\{u\} $$
$$ \{\sigma\} = [D]\{\epsilon\} = [D][B]\{u\} $$
The commercial software ANSYS Workbench was utilized for this analysis. The 3D CAD model of the end effector was imported, and material properties from Table 1 were assigned. A high-quality tetrahedral mesh was generated with a global element size of 2 mm and a high relevance setting (90), resulting in 280,789 elements and 173,979 nodes, ensuring solution accuracy.
Boundary Conditions and Load Application: A fixed support constraint was applied to the bottom face of the lowest servo mounting bracket, simulating its rigid connection to the robotic arm’s wrist. The critical load case was defined based on operational requirements. Each flexible gripper jaw must hold a teat cup and withstand an upward force during the seating action. A conservative极限载荷 (limit load) of 30 N was applied perpendicularly to the inner gripping surface of each jaw, representing a worst-case scenario. This simulates the combined weight and dynamic seating force.
| Condition Type | Location | Value / Type | Justification |
|---|---|---|---|
| Fixed Support | Base of lower servo mount | 0 displacement (all DOF) | Connection to robot arm wrist |
| Force Load | Inner face of each gripper jaw (2 locations) | 30 N, normal to surface | Simulates cup weight + seating force |
Results and Discussion of Static Analysis: The solver calculated the deformation, equivalent (von-Mises) stress, and elastic strain. The von-Mises stress $\sigma_{v}$ is used to predict yielding in ductile materials and is calculated as:
$$ \sigma_{v} = \sqrt{\frac{(\sigma_{11}-\sigma_{22})^2 + (\sigma_{22}-\sigma_{33})^2 + (\sigma_{33}-\sigma_{11})^2 + 6(\sigma_{12}^2+\sigma_{23}^2+\sigma_{31}^2)}{2}} $$
The results are summarized below:
| Parameter | Maximum Value | Location of Maximum | Allowable Value (Material) | Safety Factor |
|---|---|---|---|---|
| Total Deformation | 0.304 mm | Tip of flexible gripper jaws | N/A (Functional limit) | N/A |
| Equivalent Elastic Strain | 6.9e-4 mm/mm | Tip of flexible gripper jaws | N/A | N/A |
| Equivalent (von-Mises) Stress | 45.52 MPa | Connection bracket between the two servos | 110 MPa (6061 Al yield) | 2.42 |
The analysis confirms that the maximum deformation is negligible (0.3 mm) and occurs at the gripper jaw tips, which does not affect functional accuracy. Crucially, the maximum stress of 45.52 MPa is well below the yield strength of the weakest structural material in that region (6061 Aluminum, $\sigma_y$ = 110 MPa). The minimum safety factor is approximately 2.4, indicating a robust design with significant margin for unexpected overloads. This validates the static structural integrity of the dual-gripper end effector for its intended duty cycle.
Modal Analysis for Dynamic Performance Evaluation
While static analysis ensures strength, modal analysis is essential to understand the dynamic characteristics of the end effector. It determines the natural frequencies and corresponding mode shapes—the inherent ways in which the structure will vibrate if excited. Operating near a resonant frequency can lead to excessive vibrations, causing control instability, premature fatigue failure, or failure of the attachment procedure. The governing equation for undamped free vibration is an eigenvalue problem:
$$ ([K] – \omega_i^2 [M]) \{\phi_i\} = 0 $$
where $[M]$ is the global mass matrix, $\omega_i$ is the $i$-th natural angular frequency (rad/s), and $\{\phi_i\}$ is the corresponding mode shape vector. The natural frequency $f_i$ in Hertz is related by $f_i = \omega_i / (2\pi)$.
I performed a modal analysis in ANSYS Workbench using the same material definitions and mesh as the static study. The fixed support constraint at the base was maintained. The solver was set to extract the first four modal frequencies and shapes, as lower-order modes are typically most relevant for external excitations.
Results and Discussion of Modal Analysis: The first four natural frequencies and a description of their mode shapes are presented below. The primary sources of vibration excitation in the operational environment are expected to be low-frequency disturbances from the servo motors and the base lead screw drive, typically below 100 Hz.
| Mode Number | Natural Frequency $f_n$ (Hz) | Description of Mode Shape | Primary Location of Maximum Deformation |
|---|---|---|---|
| 1 | 115.11 | Lateral (side-to-side) swaying of the entire upper assembly. | Gripper jaws and top of gripper body. |
| 2 | 121.86 | Fore-aft rocking motion about the inter-servo connection. | Gripper jaws. |
| 3 | 317.83 | Lateral bending with a twisting component, focused on the gripper arms. | Tips of the flexible gripper jaws. |
| 4 | 343.21 | Complex torsional and pinching motion of the two gripper jaws. | Tips and sides of the flexible gripper jaws. |
The most critical finding is that the fundamental (first) natural frequency of the end effector is 115.11 Hz. This is significantly higher than the anticipated major excitation frequencies from the drive systems (<100 Hz). Therefore, during normal milking operations, the end effector is unlikely to be excited into resonance. This separation between operating frequency and resonant frequency ensures stable and predictable dynamic performance. The mode shapes themselves reveal that the most flexible parts are, as expected, the slender gripper jaws, but their deformation in these high-frequency modes does not pose a concern given the frequency margin.
Synthesis and Conclusions on End Effector Reliability
Based on the integrated static and modal analysis performed using ANSYS, I can conclusively evaluate the reliability and performance of the newly designed dual-gripper, adjustable end effector for milking robots.
From a static strength perspective, the end effector demonstrates a high degree of safety. Under a conservatively estimated极限载荷 of 30 N per gripper jaw, the maximum induced von-Mises stress is 45.52 MPa. This value resides in a connection bracket made of 6061 Aluminum, which has a yield strength of 110 MPa. This results in a safety factor of 2.42, which is more than adequate for agricultural robotic applications where load variations can occur. The maximum deformation of 0.3 mm is functionally insignificant for the套杯 (cupping) operation.
From a dynamic performance perspective, the modal analysis confirms favorable characteristics. The first natural frequency of 115.11 Hz provides a substantial margin above expected operational vibration sources. This frequency separation is critical to prevent resonant amplification of vibrations, which could lead to control errors, increased wear, or failed attachment attempts. The end effector is therefore dynamically stiff enough for precise, high-speed positioning by the robotic arm.
The design successfully addresses the key limitation of single-gripper systems by enabling the potential for simultaneous or rapid sequential attachment of two teat cups. The integrated micro-adjustment capability via servos and lead screw ensures the design can adapt to biological variation among cows, directly targeting the improvement of the套杯率 (cup attachment success rate). The identified high-stress area, while safe, could be further optimized in future iterations with topological optimization or slight geometric reinforcement to potentially reduce mass or increase the safety factor further.
In conclusion, the finite element analysis validates that this novel end effector design is structurally reliable and dynamically suitable for the demanding environment of automated milking. It meets all required strength and stiffness criteria while incorporating the necessary flexibility for adaptive positioning. This work provides a solid theoretical foundation and a proven model for the subsequent physical prototyping, testing, and eventual deployment of this enhanced end effector system, contributing to the development of more efficient and intelligent milking robotics.