Abstract
Effective selection of parts from diverse sources—reused, remanufactured, and new—is critical for ensuring the quality and cost-effectiveness of remanufactured products. This research proposes a multi-objective optimization model that addresses the impact of dimensional tolerances on quality loss and total cost. By implementing a hierarchical accuracy grading mechanism, the method ensures that remanufactured assemblies meet or exceed the precision of original equipment manufacturer (OEM) standards. An Enhanced Cooperative Co-evolutionary Algorithm (ECCA) is employed to navigate the trade-offs between cost and quality. The approach is validated through a case study of a machine tool feed box, demonstrating significant improvements in dimensional accuracy while maintaining cost efficiency.
1. Introduction
Remanufacturing plays a pivotal role in the circular economy by extending the lifecycle of industrial components. However, the inherent heterogeneity of parts—stemming from varying wear levels and recovery sources—presents a significant challenge for assembly. Unlike traditional manufacturing, remanufacturers must integrate reused parts, remanufactured parts (repaired to spec), and new components. Achieving the required dimensional accuracy in the final assembly (the "dimension chain") while minimizing costs requires a sophisticated selection strategy.
2. Methodology
The proposed framework utilizes a three-tier accuracy grading mechanism (Levels I, II, and III). Level III corresponds to standard OEM specifications, while Levels I and II represent higher precision achieved through advanced remanufacturing techniques. This allows for "upgrading" product quality during the remanufacturing process.
The selection process is structured across three levels: product, component, and part. As illustrated in Figure 1, the system evaluates the accuracy constraints and cost implications at each level to ensure the final assembly meets the design tolerance ($T_{0def}$).
The optimization model balances two primary objectives:
- Quality Loss Function: Measures the deviation from ideal precision across the assembly's dimension chain.
- Total Cost Function: Accounts for disassembly, cleaning, testing, surface repair, machining, and purchasing new components.
To solve this multi-objective problem, the researchers developed an Enhanced Cooperative Co-evolutionary Algorithm (ECCA). This algorithm decomposes the complex dimension chain into sub-problems (increasing and decreasing links), optimizing them in parallel to find the global optimum.
3. Case Study and Results
The model was applied to the remanufacturing of a machine tool feed box transmission shaft assembly. The assembly consists of five components, including gears, shafts, and spacers, with a critical axial clearance requirement of 0.01 mm to 0.20 mm.
The optimization process generated a Pareto front, representing the set of optimal trade-offs between total cost and accuracy loss. As shown in Figure 2, the "optimal solution" identifies a configuration that significantly reduces accuracy loss without an exponential increase in cost.
In the optimized configuration, the system selected high-precision "Level I" remanufactured parts for the shaft and gears. This resulted in an axial clearance of 0.0102 mm—significantly tighter than the average new OEM product—at a total cost of 29.7 yuan.
4. Conclusion
This research demonstrates that remanufactured products can achieve superior dimensional accuracy compared to new products through strategic part selection. By leveraging a diverse inventory of parts and advanced optimization algorithms like ECCA, manufacturers can deliver high-quality, cost-effective products that support sustainable industrial practices. Future work will expand this model to include surface integrity and long-term reliability metrics.
Summary updated