Expert system of genarative CAPP system in function cutting speed.

Rahimic, Senad ; Visekruna, Vojo

1. Introduction

The manufacturing industry has been pushed to adopt more effective and efficient production strategies to meet the challenge of shorter life cycle, higher quality, lower cost, wider variety of customer demands [1]. This increased emphasis on achieving highly adaptive manufacturing to reduce manufacturing costs and to better utilize manufacturing capacity has led to a critical focus on agile manufacturing as a strategy to achieve these goals. In manufacturing, process planning is the task that transforms the design information into the manufacturing processes and determines the operation sequence.

CAPP considered a crucial link between Computer -Aided Design (CAD) and Computer -Aided Manufacturing (CAM) [2]. Research of over 30 years in CAPP has resulted in a wealth of knowledge on CAPP and many experimental and commercial CAP'P systems have been developed as a result. Computer-assisted process planning was originated in the 1960s (Niebel 1965) and since has been a very active area of research and development. During the late 1970s, the science of computer-aided process planning (CAPP) evolved into two basic approaches: variant or generative process planning (Chang 1998). Modern approaches toward CAPP include using case-based reasoning. [3]

In this paper proposed of methodology show on figure 1, which is start by geometric information about part using basic features. Target make all alternative variants process planning, as organized by sequence phase, scheduled and operations, this technological to steps different.

The module developed for searching basic mechanical features (groove revolution, slot, thread, plane, cylinder cone outer, cylinder cone inner, etc.) must has two conditions, the first condition is all features to be simple, and make together complex features defined rotation part. The second condition had defined features by every simple features to be individual.

2. An object-oriented knowledge representation scheme

Whole part considerationed, searching from plans simple features which made complex rotation part, with their individual informations possibled define process and commission technology for resumption destitute operations for to do any assignment basic features. stage enactive all features, all process and operations which can to be applicable for make with acceptable declaration quality.

We propose an appropriate knowledge representation scheme which attems to address various facets of the process planning task. We recommend object-oriented common attributes and/or their values from higher level objects. Taxonomies of features, processes, machines and tools exist in the semantic network. Inheritance is used to infer attributes and behaviour of more specific object. (Hang-Wai, L.; & Hon-Yuen, T.) a number of represent manufacturing capabilities and reasoning in process planning. Now we will explain some of these taxonomies and relations in more detail.

3. Minimum cost per piece

The average cost per piece to produce a work piece consists of the following costs: nonproductive cost per piece, machining time cost per piece, tool changing cost per piece and tooling cost per piece. Mathematically, this can be expressed as equality (1).

[c.sub.1] = [c.sub.0] [t.sub.1] + [c.sub.0][t.sub.c] + [c.sub.0][t.sub.a] ([t.sub.ac]/T) + [c.sub.a] x [[t.sub.ac]/T] (1)

Upon partially differentiating c1 with respect to v, equating to zero, and solving, we obtain the minimum unit cost cutting speed ([v.sub.min]) as follows equality (2).

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

Upon partially differentiating c1 with respect to v, equation to zero, equating to zero, and solving for v, we obtain showed on equality (3).

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)

Optimization mathematically systems, this can be expressed as systems of equality (4).

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)

The minimum cost per piece for one operation in process planning are representation sums all phases consisted in operation as inform of equality (5)

[c.sub.1o] = [f.summation over (i=1)] [c.sup.i] 1f (5)

4. Example of part

Showed on figure 3 testing rotation part for given mathematically of module. They correspond to the following types:

* Groove revolution: GRO1, GRO2.

* Slot: SLO1, SLO2, SLO3.

* Thread: THR1.

* Plane: OD1, OD3, OD5, OD7, OD10.

* Cylinder cone outer: OD2, OD4, OD6, OD8, OD9.

* Cylinder cone inner: HO1.

The production media for which the process plan is developed consisting of a group of machines-tool and fixtures that provide support for conventional machining processes. The capacities established for these manufacturing systems are the usual ones for standard quality.

Figure 3. shows the expert system for generative CAPP systems for rotating parts, using the method of object oriented programming.

Using CAD software in this paper is Solid Works as an example of the rotation parts on figure 3 [5]. Graph of function for minimum cost per piece for all possible variants process planning shown in figure 4.

Showed on figure 5 testing for testing rotation part, we get for all possible variants for process planning in function cutting speed and minimum cost per piece and number workpieces.

On graph showed function for minimum cost per piece for different machining regime possible variants process planning shown in figure 6 [7].

On the table I: showed minimum cost per piece for variants with different machining regime in function of number workpieces and given coefficient rate of dependent [7].

On figure 7 rate variant 1 and variant 4, in function on numbers workpieces, we can conclusion to there mathematically dependence which is declare through coefficient showed on equality (6).

k = [C.sub.1] - [C.sub.4]/[C.sub.4] (6)

5. Conclusion

In this paper proposed an expert system suitable for generative the process plan of machining parts. It is a proposal in which the peripheral processes needed for these parts are taken into account and they enable the development of alternative process plans.

The process plans offered by the system constitute all of the alternatives for the sequence of phases, guaranteed by a high degree of optimization with regard to cost and number of phases. These alternatives explicitly include feasible alternatives for machines and, tools for operations (depending on the type of machine for each phase).

These characteristics are largely based on the methodology proposed, the functional structure, the use of general information models and the general functional procedures. All these factors working together give the system its qualities. The development of the CAPP expert system based on this proposal has demonstrated the system's feasibility and its optimal qualities.

The target is connecting CAD and CAM system to realization flow over information's by design of product then to its manufacturing. The development of the CAPP system based on this proposal has demonstrated the system's feasibility and efficient.

The process plans offered in CAPP system alternative variants, which are composed of sequence and phases. The optimization of the process variant depends of the minimal cost per peace and optimal speed of cutting.

DOI: 10.2507/27th.daaam.proceedings.017

5. References

[1] Ben-Arieh, D.; Gutin, G.; Penn, M.; & Zverovitch, Y.; Process planning for rotational parts using the generalized travelling salesman problem, int. j. prod. res., 2003, vol. 41, no. 11, 2581-2596

[2] Gonzalez, F.; & Rosado, P.; General and flexible methodology and architecture for CAPP: GF-CAPP system, int. j. prod. res., 2003, vol. 41, no. 12, 2643-2662

[3] Kusiak, A.; Decomposition in Data Minning: An Industrial Case Study, IEEE Transactions on electronics packagin manufacturing, 2000

[4] Hang-Wai, L.; & Hon-Yuen, T.; Object-oriented analysis and design of computer aided process planning system, Int. Journal CIM--vol 13, 2000

[5] Rahimic, S. & Visekruna, Model minimalnog jedincnog troska varijante tehnoloskog procesa, CIMOSov dan raziskav 2006--CRD '06, Koper, 17. november 2006

[6] Rahimic, S., Visekruna V & Balic J., Optimization of variant process palnning for genarative CAPP system, 11th International Scientific Conference on Production Engineering -CIM2007, Croatian Association of Production Engineering, Zagreb, 2007

Caption: Fig. 1. Assignment of alterative processes and operations

Caption: Fig. 2. The process planning semantic network

Caption: Fig. 3. Example Expert System for generative CAPP system

Caption: Fig. 4. Drawing model with features

Caption: Fig. 5. Graph of function for minimum cost per piece for cutting

Caption: Fig. 6. Shown all possible variants

Caption: Fig. 7. The considered variant 1-4.

Caption: Fig. 8. The coefficient dependence between variants. Table 1. Achieving coefficient rate of dependent number workpieces V1 V4 k 28,83 34,27 0,16 3 12,95 16,17 0,20 5 9,84 12,28 0,20 10 7,51 9,37 0,20 15 6,73 8,4 0,20 20 6,34 7,91 0,20 35 5,84 7,29 0,20 50 5,64 7,04 0,20 75 5,49 6,84 0,20 100 5,41 6,75 0,20 250 5,27 6,57 0,20 400 5,24 6,53 0,20 500 5,22 6,51 0,20 800 5,21 6,49 0,20 1000 5,2 6,48 0,20 > 3000 5,18 6,46 0,20