In modern manufacturing, gear hobbing stands as a pivotal process for the production of high-precision gears. The shift towards high-speed dry gear hobbing, which eliminates cutting fluid and significantly increases cutting rates, has introduced new challenges. These challenges primarily revolve around managing thermal deformation errors, extending tool life, and ensuring superior gear quality, all of which are critically dependent on the selection of optimal machining parameters. Traditional optimization methods often focus on single objectives or use less efficient algorithms, leading to suboptimal solutions that fail to balance the conflicting goals of energy efficiency and manufacturing quality. This study presents a novel framework for the multi-objective optimization of high-speed gear hobbing parameters by integrating an Improved Multi-Objective Grey Wolf Optimizer (MOGWO) with a decision-making technique. The primary objectives are to minimize specific machining energy consumption and to maximize a composite quality index, thereby achieving a sustainable and high-performance gear hobbing process.
Problem Formulation and Optimization Objectives
The performance of a gear hobbing operation is governed by several key input parameters. In this study, we focus on three critical variables: axial feed rate ($v_a$ in mm/min), spindle speed ($n_0$ in rpm), and feed per tooth ($d_0$ in mm). The goal is to find the optimal combination of these parameters that simultaneously satisfies two conflicting objectives derived from the gear hobbing process.
Objective 1: Minimize Specific Machining Energy (E)
Machining energy is a direct indicator of process efficiency and environmental impact. The total energy consumed during the gear hobbing process can be modeled and normalized to obtain a specific energy value. Lower energy consumption signifies a more sustainable and cost-effective gear hobbing operation.
Objective 2: Maximize Gear Quality Index (Q)
Gear quality is a composite measure reflecting surface finish, dimensional accuracy, and the absence of defects. It can be quantified using a weighted model that incorporates measurable outputs such as surface roughness ($R_a$) and profile error. A higher index $Q$ indicates superior gear quality. For multi-objective optimization, this is transformed into a minimization problem by considering $1/Q$ or a negative weight.
Thus, the formal multi-objective optimization problem for gear hobbing is defined as:
$$ \text{Minimize: } F(\mathbf{X}) = [E(\mathbf{X}), \; -Q(\mathbf{X})] $$
$$ \text{Subject to: } \mathbf{X} = [v_a, n_0, d_0] \in \Omega $$
where $\Omega$ defines the feasible operating ranges for the gear hobbing parameters based on machine and tool constraints.
The Enhanced Multi-Objective Grey Wolf Optimizer (MOGWO)
The Grey Wolf Optimizer (GWO) is a metaheuristic inspired by the social hierarchy and hunting behavior of grey wolves. The standard GWO is effective for single-objective problems. For the complex multi-objective problem of gear hobbing parameter optimization, we employ and enhance a Multi-Objective variant (MOGWO).
The wolf pack is categorized into four levels: Alpha ($\alpha$), Beta ($\beta$), Delta ($\delta$), and Omega ($\omega$). In the context of our gear hobbing optimization, the $\alpha$, $\beta$, and $\delta$ wolves represent the three best non-dominated solutions found in the current population, guiding the search towards the Pareto-optimal front.
The core hunting behavior—encircling and attacking prey—is mathematically modeled. The position update equations, which drive the optimization of gear hobbing parameters, are as follows:
$$ \vec{D} = | \vec{C} \cdot \vec{X}_p(t) – \vec{X}(t) | $$
$$ \vec{X}(t+1) = \vec{X}_p(t) – \vec{A} \cdot \vec{D} $$
Here, $\vec{X}_p$ is the position vector of the prey (the current best solution), $\vec{X}$ is the position vector of a grey wolf (a candidate set of gear hobbing parameters), and $t$ is the iteration number. The coefficient vectors $\vec{A}$ and $\vec{C}$ are calculated as:
$$ \vec{A} = 2 \vec{a} \cdot \vec{r}_1 – \vec{a} $$
$$ \vec{C} = 2 \cdot \vec{r}_2 $$
where $\vec{r}_1$ and $\vec{r}_2$ are random vectors in [0,1]. The value of $\vec{a}$ decreases linearly from 2 to 0 over the course of iterations, controlling the transition from exploration to exploitation in the parameter space of the gear hobbing problem.
The enhanced MOGWO algorithm incorporates an external archive to store the non-dominated Pareto-optimal solutions found during the search for the best gear hobbing parameters. A leader selection mechanism based on crowding distance is used to choose $\alpha$, $\beta$, and $\delta$ wolves from this archive, promoting diversity in the solutions. The flowchart of the proposed algorithm for gear hobbing optimization is summarized below:
Algorithm: Enhanced MOGWO for Gear Hobbing Parameter Optimization
1. Initialize the population of grey wolves (candidate parameter sets $\mathbf{X_i}$).
2. Initialize the external archive for non-dominated gear hobbing solutions.
3. Calculate objective values $E$ and $Q$ for each wolf.
4. While $t <$ Max number of iterations:
4.1. Select the archive leaders ($\alpha$, $\beta$, $\delta$) for gear hobbing guidance.
4.2. For each wolf, update its position using the $\alpha$, $\beta$, $\delta$ position update equations.
4.3. Update the objective values for the new positions (new gear hobbing parameters).
4.4. Update the external archive with new non-dominated solutions.
4.5. $t = t + 1$.
5. End While
6. Output the Pareto-optimal set from the archive.
Decision Making using TOPSIS
The enhanced MOGWO algorithm yields a set of non-dominated Pareto-optimal solutions for the gear hobbing process. To select a single, best-compromise solution for practical implementation, the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) is employed. TOPSIS selects the alternative that is closest to the ideal positive solution and farthest from the ideal negative solution. The steps for our gear hobbing parameters are:
1. Construct the decision matrix from the Pareto set.
2. Normalize the decision matrix.
3. Determine the weighted normalized matrix (assigning equal weight of 0.5 to both energy and quality objectives for gear hobbing).
4. Identify the ideal best ($A^+$) and ideal worst ($A^-$) solutions.
5. Calculate the Euclidean distance of each alternative gear hobbing parameter set from $A^+$ and $A^-$.
6. Calculate the similarity score (relative closeness, $RC_i$) for each solution:
$$ RC_i = \frac{S_i^-}{S_i^+ + S_i^-} $$
where $S_i^+$ is the distance to the ideal worst and $S_i^-$ is the distance to the ideal best. A higher $RC_i$ is better.
7. Rank the gear hobbing solutions based on $RC_i$.
Experimental Setup and Results
The proposed methodology was validated through a case study on a YKS3112CNC7 high-speed gear hobbing machine. The workpiece was a small-modulus gear with the following specifications:
| Parameter | Value |
|---|---|
| Module | 6.5 mm |
| Number of Teeth | 48 |
| Pitch Diameter | 312 mm |
| Pressure Angle | 22° |
| Face Width | 32 mm |
| Helix Angle | 60° |
The initial process parameters for the gear hobbing operation were set as: spindle speed $n_0 = 1200$ rpm, axial feed $v_a = 0.12$ mm/rev. The feasible ranges for optimization were defined based on machine capabilities.
The enhanced MOGWO was run to generate a Pareto-optimal front. The relationship between the input gear hobbing parameters and the objectives was analyzed. It was observed that a higher axial feed rate ($v_a$) generally led to lower energy consumption but could compromise quality. Conversely, a very high spindle speed ($n_0$) increased energy consumption significantly and could also degrade gear quality due to increased heat generation. This analysis underscores the need for a multi-objective approach to balance these trade-offs in gear hobbing.
TOPSIS was then applied to the obtained Pareto set with equal weighting for energy and quality. The top-ranked solutions for gear hobbing are presented below:
| Rank | Similarity Score ($RC_i$) | Optimal Parameter Set: $\mathbf{X}=[v_a, n_0, d_0]$ |
|---|---|---|
| 1 | 0.0443 | P3: [72.48 mm/min, 822 rpm, 0.085 mm] |
| 2 | 0.0405 | P7: [70.26 mm/min, 816 rpm, 0.086 mm] |
| 3 | 0.0368 | P2: [75.10 mm/min, 830 rpm, 0.084 mm] |
| 4 | 0.0221 | P5: [68.50 mm/min, 810 rpm, 0.087 mm] |
| 5 | 0.0168 | P8: [69.80 mm/min, 815 rpm, 0.086 mm] |
The very low similarity scores (all below 0.05) indicate that the selected optimal gear hobbing parameter sets are all very close to the ideal solution, demonstrating the high stability and effectiveness of the optimization and decision-making framework.
Performance Comparison and Analysis
To benchmark the performance of the proposed enhanced MOGWO, its results were compared against those obtained using a standard Multi-Objective Particle Swarm Optimization (MOPSO) algorithm under the same conditions and constraints for the gear hobbing problem. The key comparison metric was the specific machining energy consumption of the best-compromise solution selected by TOPSIS from each algorithm’s Pareto set.
| Optimization Algorithm | Selected Optimal Parameters | Specific Machining Energy (E) |
|---|---|---|
| Enhanced MOGWO (Proposed) | $v_a=70.26$, $n_0=816$, $d_0=0.086$ | 150.27 MJ |
| Standard MOPSO (Traditional) | $v_a=72.48$, $n_0=822$, $d_0=0.085$ | 178.58 MJ |
The results show a clear superiority of the proposed method for gear hobbing optimization. The solution obtained from the enhanced MOGWO achieved a 15.58% reduction in specific machining energy consumption compared to the solution from the traditional MOPSO algorithm, while maintaining a high gear quality index. This significant reduction highlights the improved exploitative search capability and convergence efficiency of the enhanced MOGWO in navigating the complex parameter space of gear hobbing towards more energy-efficient regions.
Influence of Parameters on Gear Hobbing Objectives
A sensitivity analysis was performed to understand the individual impact of each gear hobbing parameter on the two optimization objectives. The analysis was conducted by varying one parameter at a time within its feasible range while keeping the others constant at their optimal values from the MOGWO solution.
| Parameter | Trend | Effect on Machining Energy (E) | Effect on Quality Index (Q) | Remarks for Gear Hobbing |
|---|---|---|---|---|
| Axial Feed ($v_a$) | Increase | Decreases significantly | Decreases gradually | Major lever for energy saving, but requires careful control to avoid quality loss. |
| Spindle Speed ($n_0$) | Increase | Increases sharply | Increases then decreases | High speeds increase energy and heat, potentially degrading quality after an optimum point. |
| Feed per Tooth ($d_0$) | Increase | Decreases moderately | Decreases | Affects chip load; a balance is needed to reduce energy without causing excessive tool wear or poor finish. |
This analysis confirms the strong, non-linear interactions between the parameters in the gear hobbing process and justifies the necessity of a sophisticated multi-objective optimization approach to find the best compromise.
Conclusion
This research successfully developed and applied an enhanced Multi-Objective Grey Wolf Optimizer integrated with TOPSIS for the optimization of high-speed gear hobbing process parameters. The primary goals were to minimize specific machining energy and maximize gear quality—two inherently conflicting objectives in the gear hobbing process.
The proposed enhanced MOGWO algorithm effectively generated a well-distributed Pareto-optimal front, capturing the trade-off relationship between energy consumption and quality in gear hobbing. The TOPSIS method provided a robust mechanism for selecting a single, best-compromise solution from the Pareto set. The optimal gear hobbing parameters identified through this framework resulted in a substantial 15.58% reduction in energy consumption compared to a solution derived from a traditional optimization algorithm, while ensuring high gear quality, as evidenced by the very low TOPSIS similarity scores (all below 0.05).
The study demonstrates that intelligent, algorithm-driven optimization is crucial for advancing sustainable and high-performance manufacturing practices like gear hobbing. The proposed framework offers a practical and effective tool for process engineers to determine optimal gear hobbing parameters, leading to significant energy savings and consistent product quality without the need for extensive and costly trial-and-error experiments. Future work may involve integrating real-time monitoring data to create a dynamic optimization system and expanding the objective set to include tool wear and production rate for the gear hobbing process.