CHAPTER 7 CONCLUSIONS AND SCOPE FOR FUTURE RESEARCH

Transcription

1 97 CHAPTER 7 CONCLUSIONS AND SCOPE FOR FUTURE RESEARCH 7.1 CONCLUSIONS This research work focused on issues related to selective assembly of products comprising multiple components each having one or more critical quality characteristics. As the individual components are usually produced by different operations/processes, the dimensional distributions of them are invariably different and when such components are assembled at random the quality characteristics of final assemblies exhibit variations. While random assembly is adapted in most of the manufacturing industries, which facilitates precision assemblies out of relatively less precision components, often it leads to problems like wide variation of assembly tolerance, generation of surplus parts and reduction in manufacturing system efficiency. Another common practical issue with assembly precision is, when the components are produced with high accuracy, the manufacturing cost would go up and there will be a drastic increase in the manufacturing cost of the product. In this research work, two different selective assembly issues (as detailed in chapter 4 and chapter 5) have been described to represent and cover the wide range of selective assembly problems. As these problems are NP hard, the following intelligent search heuristics have been developed to approximately solve these problems and got nearer-to-optimal solutions. 1. a modified particle swarm optimization algorithm for multiple characteristic radial assembly problem to minimize assembly tolerance variation

2 98 2. an ant colony algorithm based metaheuristic for single characteristic radial assembly problem to minimize manufacturing cost When the above said heuristics are tested with experimental data sets, their performances are found better than the existing other solution methodologies of selective assembly reported in the literature. It is inferred that selective assembly methodology proposed in this research work for assemblies is highly suitable compared to existing assembly methodologies. The sensitivity analysis done on the performance of the developed heuristics in terms of number of selective assemblies, number of quality characteristics proves that the developed heuristics can be successfully applied to any kind of problems and of any problem sizes with appropriate modifications/fine tuning if required. 7.2 FUTURE RESEARCH DIRECTIONS As a continuation of research in this area, the following research problems are submitted for the pursuers of further research. i) Assembly problems with multiple matching criteria need to be addressed. For example, in the two-component assembly shown in Figure 7.1, the cylinder is inserted into groove until it makes contact on the two sides of the groove. For each set of components, the distance B needs to accommodate the current value of dimensions A, R and θ.

3 99 Figure 7.1 ii) Two-component assembly of more than one quality characteristics For assemblies which are based on certain matching features, for example matching two different orifices that is, for control flow and reduced turbulence, different assembly methods are needed that exploit the structure under various characteristic values of each assembly part. iii) Real time on-line assembly for multi-component parts, in an assembly line is another issue for future research. In this case, the unassembled residuals for each assembly part would affect the performance of the system over time which needs to be addressed. iv) Modeling of assemblies that requires adjustments during the actual assembly operation is a possible topic for future research. Examples of this type of assembly are automobile body panels and aircraft wings. v) In addition to the assembly characteristic variation, manufacturing cost optimization, consideration of other practical issues such as inventory, scheduling, and capacity issues would greatly enhance its functionality.

4 100 vi) Formulation of models and design of tools specific to unique assembly problems can be researched. Figure 7.2 shows the assembly of an internal combustion engine. The compression ratio should be kept constant at a certain value. Due to the deviation caused by manufacturing process of crankshaft, connecting rod, piston, and cylinder, the clearance varies between the actual value and design value when the piston arrives at the upper dead point. As the clearance affects the compression ratio directly, the tolerance of clearance as closing link is 0.27 mm. The assembly parts are listed in Table 7.1. The dimensional chain is shown in Figure 7.3. The objective function may be expressed by the following equation: ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) (7.1) requirement: The objective function is subject to the following technical (7.2) The constraints of eccentricity may be represented as (7.3) (7.4) (7.5)

5 101 Figure 7.2 The assembly of an internal combustion engine and dimensional chain Figure 7.3 Main parts of internal combustion engine

6 102 Table 7.1 Dimension and tolerance of mating components of Internal Combustion Engine Sl Quality characteristic Dimension Tolerance No (mm) (mm) 1 Distance between cylinder top surface and centerline of cylinder spindle x 1 = T 1 = ± Eccentricity between centerline of cylinder spindle and axis diameter x 2 = 0 T 2 = ± Center distance between crankshaft spindle and crankshaft axle x 3 = 50 T 3 = ± Eccentricity between centerline of big hole and centerline of crankshaft x 4 = 0 T 4 = ± Center distance between centerline of big hole and centerline pinhole x 5 = 174 T 5 = ± Eccentricity between centerline of pinhole and centerline of piston pinhole x 6 = 0 T 6 = ± Distance between piston top surface and centerline of piston pinhole x 7 = 69 T 7 = ± Diameter of crankshaft spindle x 21 = 70 T 21 = ± Diameter of cylinder spindle x 22 = 70 T 22 = ± Diameter of crankshaft axle x 41 = 57 T 41 = ± Diameter of connecting rod big hole x 42 = 57 T 42 = ± Diameter of connecting rod pinhole x 61 = 32 T 61 = ± Diameter of piston pinhole x 62 = 32 T 62 = ± Distance between piston top surface and cylinder top surface x Σ = 0.7