Considerations and use of the genetic algorithm in optimization of standard fixture components.

Krsulja, Marko ; Barisic, Branimir ; Car, Zlatan 等

1. INTRODUCTION

Mechanical standard components for factory automation are commonly used in today modern factories. They consist of components for workpiece holding, supporting, positioning, locating and components for tool guiding. Fig. 1 shows obtained design for fixture. The main problem is the supplementary time, difficulties in handling of the workpieces, obtaining optimal solution of automation.

[FIGURE 1 OMITTED]

Nowadays stability in milling process should be performed in the planning phase where it is much easier to implement changes. It is easier to control material and process specifications by the use of optimization processes. Artificial intelligence can be used to perform backward computing to redefine the milling process. Product on market is under severe pressure concerning costs and it is required to be precise with narrowed tolerances. Locating system can be a bottleneck in small and medium size production if not planned carefully. By adequate positioning of locators and clamps the mechanical properties of material can be affected. And systems of a higher generation of automation can be implemented. Reducing the error caused by elastic deformation of loaded fixture-work piece. Wrong positioning of the workpiece may cause dimensional problems and deformation in machining.

The clamping fixture construction is usually adapted to specific situation. In order to create a good accuracy of the milled part the interaction of tool work piece and locating system must be taken in account. The process stability and sensitivity analysis should be conducted before starting of the production. Positioning and clamping are the basic functions of fixtures, and supplementary time includes the time needed for the manipulation of workpiece and tools. Fixture clamping sequence also influences the part location errors (Anand & Shreyes 2003). The optimum layout of devices and machines in the manufacturing systems is one of the basic requirements for the design of flexible manufacturing systems (FMS), since good solutions in designing such a system are prerequisite for its efficient functioning and low costs of operation (Ficko et. Al. 1998). It has been estimated that 20% to 50% of the manufacturing costs are due to handling of workpieces: proper planning of devices can reduce the above costs for 10% to 30% (Tompkins et. Al. 1996). The FMS include devices and machines which usually do not have firmly determined dimensions and distances (Heragu & Kusiak 1988). The layout can be optimized with the respect of the relations with the neighboring devices, cost function, mutual dependencies (pin positioning, workpiece orientation, immediate vicinity of clamps, tool path, workpiece geometry etc.).

2. FIXTURES AND MILLING

In milling operations fixtures are used for locating and holding of the workpiece.

2.1 Fixtures

Fixtures are components used in a milling operation for control of the workpiece movement. Wrong clamping or pin placement can influence the final workpiece accuracy. For a relatively rigid workpice the localized elastic deformations cause it to undergo rigid body translations and rotations which alter its location with respect to the cutting tool. (Bo Li & Shyres N. Mekolte 1998). It is important to reduce the displacement that rises under the influence of bigger cutting forces. In this paper a model is presented that a process engineer or fixture designer can use to understand the impact of fixture sequence on workpiece location error. In Fig. 2 components used in fixtures are shown.

2.2 Milling

Milling is a process of metal removal; principle is that the multiple cutting edges rotate in a spindle. Teeth of a milling cutter alternately engage and disengage from the workpiece. Milling machines use machine tools that can rigidly hold and rotate a cutter while feeding a workpiece into the cutter. The workpiece is fixed with standard components on the milling table. During process a consideration has to be taken for influence of milling tool on the fixture (force, feeds, chatter, vibrations etc.). Rigidity of the fixture setup influences the feed rate of milling process.

[FIGURE 2 OMITTED]

3. GENETIC ALGORITHM

We used the traveling salesmen method for optimizing the tool path. Parameters ware inserted in the algorithm, TSP method was selected, and city positions were defined. Number of used cities was in example 1 (34 cities), example 2 (25 cities), example 3 (34 cities). The program went trough simulation and listed results. The obtained path like in Fig. 3 represented the best tool path. In the Fig. 4 it is shown that with the complex shape (number of cities) a bigger number of generations are needed to find an optimal tool path. Today computers can solve up to 100 cities in a very short time, so it is not a problem. Also with a bigger tool head the number of cities can be reduced. Together the cities represented a mesh for obtaining the desired shape. The best tour represents the best tool path; a tour is a closed path which traverses each city only once. The problem is to find a tour of minimum length, for different city configurations (shape). For genetic algorithm we used: population size of 10000, mutation of 3%, maximum generation 10000000 (ending condition), nearby cities 5, group size 5, nearby city odds 90%, and random seed 0. One of the problems in the generation of final shape was the selection of nearby city odds (80%, 85%, 90%) and group size (3, 4, 5) that resulted in different shapes. Different results for same problem, but a simple solution is to take fixed parameters in order to get uniform results in experiments.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

4. RESULTS AND DISCUSION

Obtained optimal clamping mesh with placed fixture components is shown in Fig. 5. Under the workpiece are located three supporting pins. The force enters and continues from point 4.4 to 5.4 and continues to 5.1 the pin L1, L2 and L3 have been accordingly placed, in order to minimise initial influence of tool contact forces. Other directions have been balanced with clamps C1 and C2.

5. CONCLUSION

Process planning contains careful setup planning with consideration of manufacturing features. Mechanical standard components for factory automation affect the overall cost and quality of part machining. Tool movement influences the final quality; an approach has been presented for constraining the workpiece which can help the optimal layout. The result is a better clamping setup. Different criterions could be used to find the optimal solution but fixed conditions and different examples of similar geometry ware used.

6. REFERENCES

Ficko, M.; Brezocnik, M. & Balic, J. (2005). On designing of layout of flexible manufacturing systems, Proceedings of the 16th International DAAAM Symposium 2005, Katalinic, B. (Ed.), pp. 115-116, ISBN 3-901509-46-1, Opatia, Croatia, October 2005, DAAAM international, Vienna

Heragu, S. Kusiak, A. (1988). Machine layout problem in flexible manufacturing systems, Operations research, Vol. 36, pp. 258-268, ISSN: 0030-364X

Tompkins, J. A.; White, J. A.; Bozer, Y. A.; Frazelle, E. H.; Tanchoco, J.M. & Trevino, J. (1997). Facilities Planning, John Wiley & Sons, ISBN 0-471-41389-5., New York

Bo, L. Shryes N. Melkote (1998). Improved workpiece location accuracy trough fixture layout optimization, International Jurnal of Machine Tools, Manufacture Vol. 39, (1999) 871883

Anad R., Shreyes N. M. (2003). Analysis of the effect of fixture clamping sequence on part location errors, International Jurnal of Machine Tools, Manufacture Vol. 44, (2004) 373-382