M. Cristina Fernandes, A. J. M. Ferreira, A. T. Marques, A. A. Fernandes

Several methods have been proposed in the past to model product design process based either on the description of a sequence of activities or on systematic procedures that attempt to propose a more adequate set of activities.

It is not the aim of this paper to review in a comprehensive way the different methods used in practice. Nigel Cross [1] proposed an integrated systematic approach based on rational methods that cover the whole design process, from problem analysis and conceptual design to detailing. The methodology proposed, illustrated in table 1, involves seven steps in the design process, each applying a widely used rational method.

Although devised primarily for product development an attempt is made to apply this methodology, which integrates Quality Function Deployment principles, to the design of a new material.

The application of this methodology, for material design, is illustrated in the case of development of a Polymer Concrete (PC) construction material. This material is formed by selectively graded aggregates impregnated with a polymer resin system. When combined through a process of mixing, moulding and curing an extremely powerful cross linked bond is formed, giving an higher value added material. When reinforced with fibreglass, high strength, rigid and lightweight precast polymer concrete can be obtained.

Table 1 - Stages in the design process [1]

An important first step in designing is the identification of the design objectives. A list of design objectives must be obtained by questioning the prospective clients or on discussions within the material design team.

An example of design objectives is shown below for a polymer concrete construction material:

• Low price
• Lightweight
• Corrosion resistance
• Impact resistance
• Thermal resistance
• Chemical resistance
• Low water absorption
• Appearance
• Cost effectiveness
• Easy to use and transport
• Oil absorption resistance
• Colour
• Deterioration resistance
• Salt resistance
• Dimensional stability
• Manufacturing safety
• Electrical properties
• Environment conditions resistance
• Low risk to the operator
• Density
• Magnetic properties
• Texture
• Quality
• Notch toughness
• Environment impact
• Flexural strength
• Mechanical stiffness and strength

Other techniques, like "brainstorming", within a group of selected people, trying to generate a large flow of ideas to solve a specific problem, can be used.

Once the objectives are identified and classified it is necessary to order the list into sets of higher level and lower level objectives. The needs can be expressed at many different levels of detail. Table 2 illustrates an attempt to define sets of objectives: level 1 referring to higher-level objectives and level 5 to lower-level ones.

Table 2 - Ordered set of objectives for a new construction material

Although table 2 gives an idea of the objectives’ importance, it’s unpractical and does not refer the relationship between them and the hierarchical pattern of objectives and sub-objectives, making it difficult to understand by clients, managers, and members of the design team. The use of a diagram tree of objectives can illustrate hierarchical relationships and interconnections, as shown in figure 1.

The design team has to define the essential functions that a solution type will be required to satisfy. Alternative solutions that satisfy the functional requirements can be devised.

The problem level is decided by establishing a "boundary" around the coherent sub-set of functions.

It is necessary to express the overall function for the design in terms of the conversion of inputs into outputs.

Figure 2 - The "black box" systems model

At first, the overall function should be as broad as possible. In this case, in order to produce the polymer concrete ("output"), the raw materials are the "input". Inside the "black box", there are many procedures/functions to get the finished good as illustrated in Figure 3.

The overall function must be broken down into a set of essential sub-functions or sub-tasks. The analysis into sub-functions depends on different factors (for example, kinds of components available for a specific task, the necessary or preferred allocations of functions to machines and to human operators, designers experience, etc...).

Each sub-function has its own inputs and outputs and compatibility between these should be checked.

For each attribute, precise performance requirements must be stated.

Whenever possible and appropriate, a performance specification should be expressed in quantified terms, and it must identify ranges between limits. Table 5 illustrates some performance requirements without being comprehensive.

Table 5 - Performance requirements for the Polymer Concrete construction material

In determining a product specification, the relationship between characteristics and attributes should be clearly understood.

Quality function deployment method (QFD) is a comprehensive method for matching customer requirements to engineering characteristics. The buyer is one of the most important factors to determine the commercial success of a product.

QFD can be used at various stages of the design process, and it also draws upon features from several other design methods.

It’s necessary to identify customer requirements in terms of product attributes.

The method starts with the identification of customers and their own views of requirements and desired products attributes - " the voice of the customers" should be recognised and their requirements should not be the subject of "reinterpretation" by the design team, into their own perceptions of what the customers "really mean".

The relative importance of attributes must be determined. Techniques of rank- ordering or point-allocation can be used to help determine the relative weights that should be attached to the various attributes. Table 6 illustrates importance of attributes identified for polymer concrete.

The attributes of competing products, must be evaluated.

Performance scores for competing products and the design team’s own product (if a version of it already exists) should be listed against the set of customer requirements. Judgements about competitor products are always made by customers.

Since we have to compete against existing products in the market, the design team has to ensure that its products will satisfy customer requirements better than the competitor products.

Table 6 - Importance of attributes

A matrix of product attributes against engineering characteristics is drawn.

It included all the engineering characteristics that influence any of the product attributes and ensure that they are expressed in measurable units.

The attributes form the rows of the matrix and the characteristics form the columns.

Each cell of the matrix represents a potential interaction or relationship between an engineering characteristic and a customer requirement. The strength of the relationships can be indicated by either symbols or numbers; using numbers has some advantages, but can introduce a spurious "accuracy".

The design team works methodically through the matrix and records in the matrix the relationship wherever it occurs and the strength of that relationship.

The identification of any relevant interactions between engineering characteristics is quite important. The "roof" matrix of the "house of quality" provides this check, but may be dependent upon changes in the design concept.

The interaction between engineering characteristics can be either positive or negative.

The simple way of checking these interactions is to add another section to the interaction matrix, which is added on top of the existing matrix.

As it provides a triangular - shaped "roof" to the matrix, the resulting diagram is often referred to as the "house of quality".

Many assumptions may have to be made about the final design when completing the "roof" matrix, and it should be remembered that changes in the design concept result in changes in these interactions.

Target figures to be achieved by the engineering characteristics, using information from competitor products or from trials with customers, should be set.

In this step of the procedure, the team determines the targets that can be set for the measurable parameters of the engineering characteristics in order to satisfy customer requirements or to improve the product over its competitors. A detailed investigation of competitor products may be necessary.

Figure 5 illustrates the interaction matrix for the polymer concrete under study.

The generation of alternative solutions for a product is the key aspect of designing.

Most of the designing is a variation from or a modification to an already existing product. Clients and customers usually prefer improvements rather than novelties.

Making variations on established themes is an important feature of design activity. It is also the way in which much creative thinking actually develops.

This creative reordering is possible because even a relative small number of basic elements or components can usually be combined in a large number of different ways.

The morphological chart method encourages the designer to identify new combinations of elements or components.

The main objective is to widen the search for potential new solutions.

Covering different analysis (morphological analysis), this method is a systematic attempt to analyse the form that a product might take.

Different combinations of sub-solutions can be selected from the chart, leading to new solutions that have not previously been identified.

It’s important to make a list of the features or functions that are essential to the product.

It’s necessary to try to establish essential aspects which must be incorporated in the product.

In the morphological chart method they are sometimes called the design "parameters".

In this method, for each feature or function of the list, the means by which it might be achieved.

The list of means can include not only the existing conventional components or sub-solutions of the particular product, but also new ones that might be feasible.

After the means were defined, the identification of feasible combinations of sub-solutions was proceeded.

For any product or material, the complete range of possible combinations can be a very large number.

If the total number of possible combinations is not too large, then it might be possible to list each combination and to set out the complete range of solutions.

There are many ways of doing it:

· choose only a restricted set of sub-solutions

· identify the non-feasible sub-solutions, or incompatible pairs of sub-solutions.

Since the alternative designs are created, it’s necessary to make a decision of choice between alternative sub-solutions or alternative features that might be incorporated into a final design. Choices can be made on different ways. The evaluation of alternatives can only be done by considering the objectives that design are supposed to achieve.

In order to assess and compare the performances of alternative designs across the whole set of objectives, a weighted objectives method is applied, using differentially weighted objectives.

This method assigns numerical weights to objectives and numerical scores to the performances of alternative designs measured against these objectives.

The final step in the weighted objectives method is to calculate and compare the relative and "utility values" of the alternative designs. Utility values are obtained multiplying the weight value of each objective by the score allocated to each parameter.

The wisest alternative has the highest sum value; the comparison and discussion of utility value profiles may be a better design aid than simply choosing the "best".

These utility values are then used as a basis of comparison between the alternative designs.

[1] Nigel Cross, Engineering Design Methods. Strategies for product design. John Wiley & Sons , 2nd Ed., 1996.