The elaboration of the process drawings within the generative CAPP systems.

Doicin, Cristian ; Ionescu, Nicolae ; Tonoiu, Sergiu 等

1. INTRODUCTION

All the software systems, developed in the field of generating the manufacturing process plans, use two main methods: a variant method and a generative one (Allada & Anand, 1995). A method to create a generative CAPP system, based on an algorithm composed of four main stages was developed by the authors, as follows: 1) Define the initial data, necessary to generate the variants of process plans, by creating the solid model of the part and after that, by filling in with technological data; 2) Define the constraints and preliminary establish the feasible variants of process plans, by preliminary establish the raw stock dimensions and intermediate dimensions, the machining parameters and the time required for machining, and by preliminary generating the set of variants of process plans satisfying specific criteria; 3) Establish the feasible variants of process plans and select the optimum one; 4) Design the jigs and fixtures used in the optimum variant of process plan.

In accordance with the above-mentioned stages, TehnoCIN, a generative CAPP system was created by the authors. The system is designed for revolution parts and is composed from 3 modules: The Module for Identifying the Parts' Shape--MGF (corresponding to the stage 1); The Module for Generating Variants of Process Plans--MGPT (corresponding to the stages 2 and 3); The Module for Designing the Jig and Fixtures--MGES (corresponding to the stage 4);

The paper presents the final step of the third stage of the method, those of elaborating the process drawings for all machining operations. The block which implements this step uses AutoLISP[R] and ARX[R] routines which automatically generate the process drawings, within an Autodesk Mechanical Desktop[R] environment.

2. CONCEPTS AND DEFINITIONS

In the process of developing the TehnoCIN system we've used some old and new concepts (with some particular definitions given by the authors) to define the constructive-technological entities (or part components) describing the machined part. We named these components Distinctive Features (Doicin, 2001; 2002).

Let us consider that a Simple Distinctive Feature (shortly, Distinctive Feature, abbreviated DF) is the simplest constructive piece of a part--made from a surface or a group of surfaces that can be machined together--in which a part can be divided (the atom of the part).

Related to this definition, other two important concepts can be described (Doicin, 2002): Complex Distinctive Feature (CDF), as the distinctive feature composed by several neighbouring distinctive features (a group of adjacent DF's), connected by precedence relations and having a certain role in the activities of defining the process plans variants, and Dominant Distinctive Feature (DDF), as those distinctive feature within a Complex Distinctive Feature having the biggest dimensions, and which gives the complex feature the main characteristics from constructive point of view.

3. ROUTINES AND FUNCTIONS

The routines for generating the process drawings are written by the system in a specific file, having a structure defined by two types of functions: a) Generative functions, defined before the system starts to run, as the core of the graphical representation kernel of the TehnoCIN, which are able to draw standard graphical components of the drawings (points, lines, squares, rectangles etc.) and b) Transfer functions, which depend on the currently analysed variant of process plan, and contain the lists with 3D point coordinates of the part, extracted from the database of the CAD model of the part.

The generative functions create the process drawings. The transfer functions are automatically generated by the CAPP system for each variant of process plan. They have the role of transferring the data--calculated or extracted from the database (Doicin, 2001; Doicin & Tonoiu, 2002)--to the generative functions. The routines contain technological information about machine-tools, tooling etc. and also create the list with 3D point coordinates required in order to draw the part having machined surfaces corresponding to the current operation. In this way, the generative functions will use the 3D lists of points made by the transfer functions in order to generate the 2D drawing of the part, as main component of the process drawing.

4. THE TRANSFER MATRIX

Prior to generate the transfer functions, a matrix (named Transfer Matrix) containing the intermediate dimensions during the machining phases is generated. The transfer matrix contains all intermediate dimensions important during the manufacturing process and is generated by passing the following steps: 1) Corresponding to each DDF contained by each CDF, the intermediate dimensions (the dimensions of the machined distinctive feature after each machining sequence) (Vlase, 1996), are calculated. The TehnoCIN system uses a database containing the types of machining phases able to generate all the technical possible types of DDFs. The calculus of the intermediate dimensions takes into account the longest succession of machining sequences of the same type identified in the above-mentioned database for each distinctive feature; 2) After that, the intermediate dimensions are extracted from the database and assigned to each distinctive feature within all the CDFs defining the analysed part; 3) The complete information regarding the intermediate dimensions for all feasible variants of process plans developed for the current analyse part are now available. All these data are stored in a matrix with intermediate dimensions (Transfer Matrix), built in accordance with the structure described in the table 1.

It is considered that each machining phase is defined by two important parameters (Doicin, 2001): a) the type of the machining phase (i.e. turning, milling, boring etc.) and b) the nature of the machining phase (i.e. roughing, finishing etc.).

In the transfer matrix, only those cells corresponding to the feasible machining sequences are storing data (i.e., the distinctive feature DF1 will be machined by a succession of only MP1 and MP2 machining sequences).

The implemented method is subject to improve because it is not able yet to offer precise values for simple distinctive features of type chamfer and fillet. This is the result of the fact that these kind of features are "borrowing" their tolerance and limits from the associate DDF, usually, the cylinder contained by the current CDF. It is considered that the entire part is composed by 3n simple distinctive features (DFs). For each machining sequence corresponding to a distinctive feature, the nominal value of the intermediate dimensions is retained, together with the upper and lower tolerance limits. This data will be used to generate the process drawings.

5. THE PROCESS DRAWINGS ELABORATION

In order to elaborate one process drawing, it is necessarily to respect the following algorithm:

1. For each i operation in the current analysed variant of process plan, all the feasible machining sequences are skimmed through (each machining sequence is made over one simple distinctive feature, DF);

2. For each machining sequence, the DF to which the sequence is assigned has to be identified. If the shape of the same distinctive feature is repeatedly modified by the same type of machining sequences (having different natures--i.e. roughing turning and finishing turning) only the last machining sequence (finishing turning) is retained in a data-vector;

3. For each Pj machining sequence within the i operation, the values of the dimensions and of the tolerance limits corresponding to the assigned distinctive features are extracted from the Transfer Matrix;

4. The previously extracted values are assigned to the Pj machining sequence;

5. The complex distinctive feature containing the previously assigned DF (step 2)--machined during currently analysed machining sequence--is identified;

6. The succession of steps from 2 to 5 is repeated until all the machining sequences within the current operation are analysed;

7. All the simple distinctive feature compounding the machined part are grouped by their originate CDF;

8. All Complex Distinctive Features are ordered by the axial coordinates of a so called characteristic point (Doicin, 2001; 2002), unique defined for each CDF;

9. The list containing geometric information is generated;

10. The steps from 1 to 9 are repeated for all the operations within the current variant of process plan.

11. The steps from 1 to 10 are repeated for all the feasible variants of process plan generated by the TehnoCIN CAPP system.

The previously determined data are exported as lists. For each feasible variant of process plan, such a list is defined. The list contains point coordinates, used to draw all the process drawings for those variant.

For revolution parts, a variant of process plan composed from k operations was considered. First two of them, Cutting and Machining Planar surfaces are not analysed by the algorithm. Thus, by first operation the TehnoCIN will understand the third operation of the process plan.

By running the routines, the process documentation--containing all the description of the machining operations within all the feasible variants of process plans--will be automatically generated.

6. CONCLUSION

The paper brings in some important theoretical contribution related to the elaboration of the process drawings within generative CAPP systems.

Thus, the general algorithm of generate the process drawings is presented. This is based on a large amount of data and complex calculus for establish the dimensions and the tolerances of the surfaces machined in each machining sequence, and to find the coordinates of all the points defining the process drawings.

The algorithms are programmed using AutoLISP[R] and ARX[R] routines, automatically generated by the CAPP system, depending on the shape of the current analysed part.

The generative method assumes that a specific routine is written to draw each process drawing. All the routines related with the operations belonging to the same variant of process plan are written in a single file. This will be loaded and ran within an AutoCAD[R] or Autodesk Mechanical Desktop[R] working session.

The result is represented by the process specifications / process drawings, the documents required in order to machine the raw stock and to obtain the finished part.

7. REFERENCES

Allada, V.; Anand, S. (1995). Feature-based modeling approaches for integrated manufacturing: state-of-art survey and feature research directions, International Journal of Computer Integrated Manufacturing, vol. 8, no. 6, November 1995, pp. 411-440, ISSN 0951-192X.

Allada, V.; Anand, S. (1996). Machine Understanding of Manufacturing Features, International Journal of Production Research, Vol. 34, No. 7, July 1996, pp. 1791-1819, ISSN 0020-7543;

Doicin C.V. (2002). Concepts regarding the Description of the Revolution Parts through Constructive--Technological Entities, Proceedings of the International Conference on Integrated Engineering, G. Draghici & S. Tichkiewitch (Ed.), pp. 94-98, ISBN 973-8247-92-6, Timisoara, April 2002, Ed. Politehnica, Timisoara;

Doicin C.V.; Tonoiu S. (2002). The codification and transfer of the data within the CAPP generative systems, Proceedings of the 4th Workshop "Human Factor and Environmentalist", Branko Katalinic & Emil Wessely (Ed), pp. 27-28, Vienna-Kosice, December 2002, DAAAM International Vienna, Vienna-Kosice;

Vlase A. (1996). Machine Building Technology (in Romanian), E.T., Bucharest, ISBN 973-648-228-7, 1996;

Tab. 1. The Transfer Matrix

 Machining phase DF1 Upper tol. Lower tol.
 (type, nature) limit, As limit, Ai ...

MP 1 ([t.sub.1], [n.sub.1]) Dim.1 As1 Ai1
MP 2 ([t.sub.1], [n.sub.2])
MP 3 ([t.sub.2], [n.sub.1]) Dim.2 As2 Ai2
...
MP k ([t.sub.j], [n.sub.i])

 Machining phase [DF.sub.3n] As Ai
 (type, nature)

MP 1 ([t.sub.1], [n.sub.1]) Dim1 As1 Ai1
MP 2 ([t.sub.1], [n.sub.2]) Dim2 As2 Ai2
MP 3 ([t.sub.2], [n.sub.1]) ... ... ...
... Dimg Asg Aig
MP k ([t.sub.j], [n.sub.i])