On a generalized algorithm of the technical product conceptual design.

Diaconescu, Dorin ; Jaliu, Codruta ; Neagoe, Mircea 等

1. INTRODUCTION

The conceptual design denomination is derived from the older notion of product concept, used for the result of this design. The new terminology promotes the denomination of the product principle solution instead of the concept notion; momentary both notions coexist in literature.

Based on a comparative analysis of the algorithms that are proposed in literature (see references), the authors propose a new modeling variant, for the conceptual design modeling, whose algorithm generalizes the existing algorithms.

2. COMPARATIVE ANALYSIS OF CONCEPTUAL DESIGN MODELS

Several models for conceptual design modeling are proposed in literature. The most representative of them belong to Cross, Ulrich & Eppinger, Dieter, Pahl & Beitz and VDI. From the comparative analysis of these models, the following useful conclusions come out:

2.1. On the requirements list elaboration

Unlike the first three mentioned models (Cross, 1994; Ulrich & Eppinger, 1995; Dieter, 2000), in which the requirements list is considered a step of the conceptual design, the requirements list is input entity in the German models (Pahl & Beitz, 1995; VDI, 1997); therefore, in the German view, the requirements list is a distinct activity, relatively autonomous.

2.2. On the strictness of terminology

From the analysis of the used terminology in the mentioned models, some interesting observations come out:

* Cross's model (Cross, 1994) uses a terminology of general type, like: objective, problem and sub-problem, function and sub-function, requirement, characteristic, alternative (variant), sub-solution, solution etc.

* The models of the American authors, Ulrich & Eppinger (Ulrich & Eppinger, 1995) and Dieter (Dieter, 2000), use a more elaborated terminology which has also specialized notions, like: target specification, product concept, concept development, the concept of a sub-problem, integrated solution, project development etc.

* The German models use the most rigorous terminology, and between them, the VDI model (VDI, 1997) has the most accurate form. Beside the general notions, there are used strictly specialized notions as: structure of functions, solving principle, physical effect, effect carriers, configurations of the effect carriers, solving structure, principle solution, effects plan, configuration plan etc.

2.3. On the models' common denominator

Though they have different formulations, all the mentioned design models can be reduced, through simplification, to the same common denominator: the algorithm for solving cycle of a technical problem. This algorithm can be found in the simplified variant of the type: global problem [right arrow] sub-problems [right arrow] sub-solutions [right arrow] global solution in the structure of each design model that was previously mentioned.

3. GENERALIZED ALGORITHM

Starting from the product design variant modeled in Fig. 1, a new variant that models the product conceptual design is proposed in Fig. 2, derived from the previous mentioned models by generalization. The structure of the generalized algorithm from Fig. 2 is centered on an adequate base of information, starts from a requirements list and contains:

1. Global function statement. Result: The global function,

2. Global function description by structures of sub-functions. Result: The structure of the global function,

3. Solving structures generation by: solving the sub-functions, sub-solutions composition and elimination of the inadequate solutions. Result: Solving structures,

4. The best solving structure selection by evaluation.

Result: The product principle solution (the concept).

The inverse connections between steps (Fig. 2), which are necessary in the iterative optimization and in readapting the requirements list, can be made directly, through the peripheral information flow, or indirectly through the information base.

Among the four steps (Fig. 2), the solving structures generation (step 3) represents the key step; for this step, illustrated shaded in Fig. 2, a general algorithm was detailed in Fig. 3. This algorithm (Diaconescu et al., 2008), derived from the homonymous algorithm proposed by Ulrich and Eppinger (Ulrich & Eppinger, 1995) considering the VDI terminology, is centered on an adequate information base, starts from a known structure of sub-functions and contains (see Fig. 3):

1. Grouping of the sub-functions into: a) functions of U.S.P. type (Unknown Solving Principles) and b) functions of K.S.P. type (Known Solving Principles).

Results: Functions of U.S.P. type and functions of K.S.P. type.

2. Solving principles synthesis (for U.S.P. functions), i.e. synthesis of physical effects and configuration of effects carriers. Result: New solving principles.

3. Identification of known solutions. Result: existent solutions.

4. Generation of the solving structures, which contains: 4.1 solving structural variants generation and 4.2 pre- establishing and keeping the variants whose technical features satisfy quantitatively the requirements list. Result: solving structures.

In the case of the K.S.P. functions, the identification of the existent solutions is based on two main sources: the technical literature and the existent technical systems. For the U.S.P. functions, the physical effects and/or the configuration of the effects' carriers can be found by: conventional methods (as: analysis of the natural systems, models analogy etc.), intuitive methods (as: Delphi method, Brainstorming, etc.) and logical methods (as: search by using catalogues of physical effects, the variation of the effects' carriers configuration for known solutions, etc.).

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

The sub-solutions are usually combined by using morphological matrixes (VDI, 1997).

4. CONCLUSIONS

The following conclusions can be formulated based on the algorithms from Fig. 1, 2 and 3:

a. Each step from Fig. 1 represents a distinct activity, relatively autonomous, with its own notions and methods.

b. According to Fig. 2 and Fig. 3, the global function of a product has elements of novelty if: [b.sub.1]) it has at least one U.S.P. function and/or if: [b.sub.2]) at least one original connection interferes in the global function structure.

c. In terms of novelty, more design types can come out from the global function of a product (Pahl & Beitz, 1995): 1) Original design, 2) Adaptive design and 3) Variant design; in each of them, the principle solution contains: 1) elements of originality at least in the plan of effects or at least in the plan of connections; 2) elements of novelty only regarding the configuration of the effects carriers; 3) known effects, known effects carriers configurations and known connections; therefore, the variant design is a form of design through similitude (by the "model").

[FIGURE 3 OMITTED]

5. REFERENCES

Cross, N. (1994). Engineering Design Methods, J. Wiley & Sons, ISBN 0-4719-4228-6, New York

Diaconescu, D. et al. (2008). Products' Conceptual Design, Transilvania University Publishing House, ISBN 978-973598-230-0, Brasov

Dieter, G. (2000). Engineering Design, McGraw Hill, ISBN 007-116204-6, Boston

Pahl, G. & Beitz,W. (1995). Engineering Design, Springer, ISBN 3540504427, London

Ulrich, K. & Epinger, S. (1995). Product Design and Development, McGraw-Hill Inc. ISBN 0-07-113742-4, New York

VDI--Verein Deutscher Ingenieure (1997). Richtlinien 2221 and 2222