Mobile Edition
Print
Get the Mag
Weekly Updates1. Introduction
Process planning is the systematic determination of the steps by which a product is manufactured. It is a key element that bridges activities in design and manufacturing. In the past decades, process planning and its automation enablers have been extensively studied and significant progress has been reported (Maropoulos, 1995). Many approaches to process planning have been suggested including conceptual process planning, setup planning and detailed process planning; see Fig. 1. Conceptual process planning includes engineering feature recognition, process selection and machine/tooling selection. Detailed process planning includes fixture design, quality-assurance-strategy selection and cost analysis.
[FIGURE 1 OMITTED]
Setup planning constitutes a critical component that connects conceptual process planning and detailed process planning. Conceptual process planning provides qualitative information to setup planning, including designated features, selected processes and datum scheme constraints. The purpose of setup planning is to arrange manufacturing features into an appropriate sequence of setups in order to ensure product quality and productivity (Huang and Liu, 2003). A setup plan is comprised of setup formation, datum scheme selection and setup sequencing (Huang, 1998). It defines a series of datum/fixturing schemes for a Multi Stage Machining Process (MMP), as shown in Fig. 1. However, the setup plan obtained from those traditional methods provides limited detailed information to subsequent planning activities in process planning.
Product quality is one of the main concerns of setup planning. A well-defined setup plan should be able to satisfy quality specifications under normal manufacturing conditions. Product quality is affected by the outcome of setup planning since the series of datum and fixtures defined by a specific setup plan may introduce errors which will propagate along the machining stages and accumulate in the final product. Different setup plans specify different datum/fixturing schemes, lead to different variation propagation scenarios, and result in different product quality. Thus, one of the major tasks in setup planning is to identify the optimal setup from multiple alternatives to ensure product quality.
Some research has been conducted in quality-assured setup planning, addressing issues in setup formation, datum scheme selection and setup sequencing. Zhang et al. (1996) proposed principles for achieving tolerance control proactively via appropriately grouping and sequencing features according to their tolerance relationships. Mantripragada and Whitney (1998) presented the "datum flow chain" concept to relate datum logic explicitly with Key Product Characterstics (KPCs) tolerances and assembly sequences. Quantitative approaches were also developed to evaluate variation stack-up associated with different process design. Rong and Bai (1996) presented a method to verify machining accuracy corresponding to fixture design. Song et al. (2005) developed a Monte Carlo simulation-based method to analyze the quality impact of production planning. Xu and Huang (2006) modeled the simulated quality distributions in multiple attribute utility functions. In addition to the simulation-based approaches, analytical methods have also been used to investigate the interactions between product quality and process variability. For a given setup plan, Stream of Variation (SoV) methodologies (Shi, 2006) and state space modeling techniques have been developed to model the dimensional variation propagation along different setups (Hu, 1997; Jin and Shi, 1999; Ding et al., 2002; Zhou et al., 2003; Huang et al., 2007a; Huang et al., 2007b).
Cost-effectiveness is another critical concern in setup planning. It can be evaluated in terms of Cost Related to Process Precision (CRPP), such as the cost to achieve necessary fixture precision to satisfy product quality requirements. The precision refers to the inherent variability in an MMP and CRPP is the cost to achieve a required precision level to ensure product quality requirements. The CRPP is assumed to be inversely proportional to the necessary process precision. Corresponding to different setup plans, different process precisions are required and thus different costs are incurred. Therefore, setup planning should be a discrete constrained optimization procedure. Ong et al. (2002) considered various cost factors in the optimization index, including the cost of machines and fixtures. However, these cost factors are not directly linked with process precision.
It is desirable that the optimal setup plan is the one that satisfies the product quality specification using relatively imprecise fixtures and machines to minimize the CRPP. However, setup plans developed solely based on principles and experience can be very conservative. Although they are generally feasible with respect to the quality consideration, cost-effectiveness may not be optimal. For instance, in order to ensure the final product quality, engineers tend to conservatively select unnecessarily precise fixtures and thus cause unnecessary CRPP. This is especially true for the upstream stages of an MMP where there are no techniques to evaluate variation propagation. Furthermore, in order to automate process planning, it should be easy to integrate the outcomes of the setup planning procedure with other detailed process planning activities, e.g., fixture design. Fixture layout design for a particular setup is critical input data for setup planning, whereas the setup planning results determine an MMP whose fixture system should be optimized at the process level. However, although the fixture layout design has been successfully investigated at both the single-stage level (Cai et al., 1997) and process level (Kim and Ding, 2004), effective setup/fixture planning studies are still required. This is because qualitative-principle-based setup planning provides limited potential for specifying quantitative precision requirements of fixture design. In addition, conservative process precision requirements will make the designed fixture unnecessarily expensive. This functional limitation of conventional setup planning significantly hinders the implementation of process planning automation.
Existing setup planning approaches are summarized in Table 1. As can be seen, most reported research has focused on the evaluation of setup plan alternatives. Some work exists that uses qualitative or simulation-based evaluation of product quality to perform optimal setup planning. Although simulation provides an effective strategy to compare alternative setup plans in terms of their output product quality, it consumes a substantial amount of time and computational resources.
This paper adopts an integrated setup/fixture planning strategy to process planning. It focuses on the systematic development of a cost-effective, quality-assured setup planning, which is a fundamental enabler to integrated setup/fixture planning. Because of the complexity of the integrated problem and the overwhelming computational requirements, an iterative approach is appropriate. As illustrated in Fig. 2, the stage/setup level optimal fixture layouts for all candidate datum schemes are first determined and fixed. In each stage, different datum scheme options may be assigned with different fixture layouts. These stage/setup level fixture layouts are the inputs to the setup planning, together with information on the feature representation, design specification, constraints on datum scheme and setup sequence. As shown in Fig. 2, the development of the proposed setup planning consists of three steps.
[FIGURE 2 OMITTED]
1. Candidate setup formations and datum schemes are formulated based on input information. Their potential variation stack-up can be analytically predicted by the soV model.
2. Based on those candidate setups defined in step 1, the setup planning is formulated as a sequential decision making process on an optimal series of setups that cost-effectively satisfies product quality specifications. A cost criterion is defined to evaluate the optimality of candidate setup plans under the constraints of product quality specifications.
3. Dynamic Programming (DP) is used to solve the optimal sequential decision-making problem and generate the optimal setup plan, which provides setup information for subsequent activities in process planning. Based on an analytical quality evaluation strategy, the proposed optimal setup planning methodology will be effective and efficient. When the optimal setup plan is determined, the approach of Kim and Ding (2004) can be applied to achieve a process level optimal fixture layout, which will be used to update the stage/setup level fixture layouts for repeating the iterative optimization procedure.
The remainder of this paper is organized as follows. The SoV-based optimal setup planning methodology is introduced in Section 2. Section 3 presents a case study in which the proposed approach is used to generate a setup plan for MMPs. Conclusions are drawn and areas of future work are discussed in Section 4.
2. Quality-assured cost-effective setup planning
The design specifications of a machined product are often satisfied by machining operations performed in a series of stages. In each stage, a set of features are generated with a specific setup. The dimensional precision of the final product is affected by three major variation sources in the machining operations.
1. Machine and cutting tool, which refers to the random deviation of the cutting tools from their nominal paths.
2. Fixture, which refers to the random deviation of the fixture locators from their nominal positions.
3. Datum, which refers to the random deviation of the datum features, generated in previous stages, from their nominal positions and/or dimensions.
FEED