"Cleverness is not a substitute for knowledge."

Mary McCarthy

General Design Science is not only consumer of the knowledge of other sciences, it serves also as source for deriving, ordering and providing the knowledge for other disciplines, whether in thoughts or in written processed form.

Remember that the construction of the General Design Science (i.e. on the generic plane of technical systems) was stated as first goal (see part II). As next goal, this knowledge should be transformed for different application regions. In a deductive process we can obtain (Specialized) Design Sciences for many individual kinds or types of TS and/or various additional features.

Derivations from Design Science can be useful especially from two characteristics according to our morphological diagram, figure 5--1. One derivation can be made from characteristic 3 -- recipients. Completely different knowledge systems are needed for:

• students -- as design-educational instruction, in which the design knowledge is selected and ordered according to didactic and pedagogic concepts;
• teachers or researchers/scholars -- as scientific treatments with all reasoning, justifications, substantiations and foundations;
• practitioners -- e.g., as design manuals in which especially the prescriptive statements find their place and the ordering must occur from pragmatic criteria (compare design situations, Section 7.2.2.3.5).

If today we still remark about the small practical application of Design Science, then important causes are that such derivations do not exist for practitioners to the extent as they would be needed. Some indications are given in Chapter 9.

A second derivation can be performed from characteristic 5 -- type of object. For each type of TS the corresponding knowledge can be derived and concretized. Sources are the knowledge of Design Science and the corresponding technical (branch, domain) knowledge of the area. The present knowledge, e.g., about machine elements, is joined with that of Design Science to form the Specialized Design Science for machine elements. Let us first treat the second of these classes of Design Science.

In the chapters of part II (Chapters 4 -- 7) we discussed the General Design Science, which is associated with the generic plane of technical systems. A general design process was also found (figure 7--13). However, each actual design process depends particularly on the object to be designed -- the technical system. Therefore all kinds of technical systems, as they were discussed in the Theory of Technical Systems (Section 7.1, and [214,219]), display specific design processes and demand beside the general knowledge also concrete and specific technical (branch, domain) knowledge.

The first two classes of technical systems are process systems and object (real) systems, and these influence the essence of the design processes.

In the earlier chapters we have demonstrated many classification characteristics, see figures 5--1, 7--4, 7--9and Section 7.1. In the area of object systems, to which we pay more attention, the particulars and peculiarities of an individual design process are caused especially by the following classification characteristics:

• function and mode of action (mode of operation, action principle) of the technical system (TS), features of individual specialties or industry branches;
• degrees of complexity (plant to individual component part);
• level of originality of the TS (novel design, adaptation, variant, re-use);
• number of items of the manufactured TS (mass production to single item "one-of-a-kind" production);
• size and other features of organization where TS are manufactured (TS manufactured in small, medium or large organizations).

Some examples should illustrate these characteristics:

• when designing internal combustion engines or Kaplan turbines, and many other TS-families, the phase of conceptualizing will rarely be needed, because these TS almost always use the same mode of operation. Their principle and overall organ structure is pre-determined. The situation may be different at the level of the individual partial organs, which can certainly also be newly conceptualized;
• for material processing TS and special machines, conceptualizing forms the decisive phase of the design procedure in which the quality of the TS is established;
• the structure of plant exists predominantly of available and partially of purchasable elements. Thus for designing of plant ("project engineering") the main task is the selection of elements and establishing their relationships;
• the design process for a car in mass or series production, where hundreds to thousands of pieces are manufactured, differs substantially from designing for single-item "one-of-a-kind" manufacture;
• only a part of the general design process, namely the last form-giving phase, is applied if a variant is to be designed. In this case only some of the data of an available product are changed;
• large enterprises mostly possess better technical and financial possibilities (including computer systems). The division of labor also reaches deeper, and thus also the knowledge level.

If each of these different classes of designing is taken as an example of a potential "Specialized Design Science," different suggestions for classifying characteristics logically emerge, depending on which superior criteria are chosen. One thing is certain, namely that each of these classes can allow setting up a specialized science, and therefore a specialized knowledge system. This statement is supported by knowing that in individual areas the "know-how"-information has become familiar (even if we can hardly speak of superior knowledge systems).

We recommend that the Specialized Design Sciences should be defined as completely as the General Design Science, so that they serve as many as possible of the branch experts, and can also possibly present a meaningful unit for engineering education.

From this point of view, the following hierarchy of ordering characteristics appears appropriate:

• complexity of technical systems: the difference between project engineering of plant and designing of components is large;
• branch, field of expertise: TS-families often differ greatly from each other in reference to the technical knowledge systems. The complexity level of "machines" is most applicable for this consideration;
• level of originality of technical systems: each technical system at any arbitrary complexity level and in any arbitrary branch can appear as a novel system or an adaptation (variant). According to kind of adaptation the changes can be different -- changes only in dimensions (within a given component structure), changes to the component structure (e.g., arrangement, configuration), to changes in higher (more abstract) structures (compare figure 7--3 and Section 7.4);
• kind of production: each technical system can be manufactured in mass production, as single-item production, or in other suitable quantity ranges (for instance, a plant, for which several of the elements may be mass-produced).

In strict logic, therefore, a "branch Design Science" can be considered as an element of the partial quantity of sciences within each complexity level; and also an "originality Design Science" is an element of the partial quantity of each branch, and a "kind of production Design Science" is an element of the partial quantity of each originality level.

This statement can be illustrated (see figure 8--1), by first subdividing the machine systems (crudely) according to the degree of complexity, and then associating the TS-families with the individual levels and indicating examples.

Regarding the completeness of Design Science in individual areas it must be clear that the individual states of embodiment of the ordering characteristics still make possible a general knowledge system which is broadly applicable and differs only in certain special aspects from the other knowledge systems in the region. Knowledge systems which belong as elements of several levels and kinds of sciences can emerge in this way. As example we mention the classes of "design for properties" (compare figures 7--4 and 7--5). Thus a complete Specialized Design Science can be assembled also from several rather "independent" elements, which are joined to the kernel of the branch (domain) knowledge (compare figure 8--5).

The hierarchy of technical systems can be erected according to the degree of complexity on four basic planes, even if each of these sub-divisions (units) can consist of additional levels. We are dealing with level IV -- plant, level III -- machines (instruments, devices), level II -- assembly groups and level I -- elements, components.

The best known unit of this hierarchy is presented by the machine, instrument, device and apparatus unit, which deliver to the user a complete complex of effects (outputs), and form a material unit, e.g., locomotive, vacuum cleaner, television set, house or measuring instrument. These machines can be decomposed, depending on the complexity of their structure, according to different criteria into many partial groups which display ill-defined boundaries. These partial groups again consist of assembly groups, down to elements, components.

We will now treat the specific features of the individual levels and thereby ask the question of how advantageous a Specialized Design Science can be, and at which level of the concretization it could be optimally situated.

Plant, major projects and infrastructure are technical system of the highest level of complexity (level IV). They consist of machines down to elements, components. This kind of complex technical system includes manufacturing plant, energy generating plant, chemical plant, traffic and utilities infrastructure, according to which purpose the complex (or very complex) function (capability) of the plant should serve. The modern kitchen, which prepares, stores and performs other functions for a family or a social unit, could be called a plant.

Question:
What characterizes plant (especially for designing)?

Compared with other technical systems, a plant displays several characteristic features:

• it is technically complex;
• it is costly to very costly;
• it contains several kinds of technical systems: building, machine, electrical systems, etc.;
• it combines many purchasable elements into a unity (only a small part of any plant is manufactured to order);
• in planning -- designing -- the emphasis lies clearly in conceptualizing and establishing the processes, and the arrangement (configuration) of the chosen sub-units -- machines;
• the technology of the transformation is the dominating element in reference to quality;
• the operator (factor) environment plays an unusually important role, not only technically but also socially and, in large plants, also politically;
• it is realized predominantly (or almost exclusively) in single-item (made-to-order, one-of-a-kind) production, although many of its elements can be manufactured in larger numbers of pieces.

Questions:
1. How does designing proceed? Does experience about designing/planning exist?
2. Do particular demands exist with reference to designing?

Referring to the above characteristic features, design planning or project engineering of plant is a particular process. It can appear surprising, that today the planning process is already formalized very broadly and reacts to particular demands. For instance:

• questions about potential costs are asked very early in the process;
• the location of the proposed plant must be established immediately, and the solution must react to the location and its conditions;
• because a plant influences the environment, plant building is subject to several regulations and permissions, e.g., building inspection, fire department approval, trade supervision, safety regulations, work protection;
• the planning engineer is often at the same time the realizer of the project, supervising its construction.

Planning proceeds as a rule in several stages. The general procedure usually occurs in following phases:

• pre-planning, investigative project, feasibility study, in which an approximate idea of the process, the plant structure, but also the initial costs should be established;
• elaboration of the project; this delivers accurate statements about processes (technology) of all corresponding streams (of energy, material, information; forward working and feedback control); layouts of machines, apparatus; plans for electric installation, piping, heating, ventilating, roadways, maintenance areas, etc.; a complete estimate of all costs;
• granting of building permits, and of construction contracts and building project sub-contracting (including some more detailed designing).

It is clear that plant engineering deals with a very demanding engineer work, which requires much specialized technical knowledge and is concerned with many risks (possibly including the financial risks).

Therefore it is not surprising that in this area very early attempts were made to streamline project engineering. The example of the method SLP (Systematic Layout Planning) [313] documents this effort; some of the details of this method (figure 8--2) show also the breadth of this attempt.

Questions:
1. Is special technical knowledge needed, and in which areas?
2. Does this knowledge already exist? If so, in which form does it exist?

To be able to fulfill the demands, a special range of knowledge had to be developed. Cost calculating and the characteristics connected with costs and calculation can serve as typical example:

The inquiry necessary for judging the economics of a project, especially of the cost emerging in the first project engineering phases, is extremely difficult. In the past, methods which allowed rough cost estimates had to be developed (compare figure 7--15). One of the developed methods consists of creating a simple model to obtain a first idea of the dimensions of the buildings, apparatus, etc. Thereby one can establish the extent of the enclosed space, and multiply this by experience numbers (e.g., cost per unit volume). Through suitable additions, the costs can be obtained for further equipment (e.g., for electrical, heating and sanitary installations). The method and the corresponding numbers form the technical knowledge of the project engineer.

To this area belongs also the knowledge about the mean values, which refer to proportions of individual types of cost contributing to the total costs. From a larger number of manufacturing plant, the following estimation values were obtained:

proportion of machines and apparatus costs about 33%, costs of buildings about 16%, engineering work about 16%, conductors and fittings about 10%, measuring and control equipment about 6,5%.

Such technical knowledge must be determined also in all other partial areas.

Questions:
1. On which plant level can the Specialized Design Science be meaningfully and efficiently set up, and serve optimally?
2. Do special demands on contents exist?
3. What is the state of the knowledge?

In the taxonomy of technical systems (as part of the Theory of Technical Systems) the related kind of systems used are summarized in hierarchically graduated higher sets (taxa). Thereby we can define species, genus, families, classes, phylums. This ordering of technical systems, analogous to the classification used botany and zoology, is suitable for all complexity levels of technical systems, as shown in the examples in figure 8--1, and defined in figure 8--4.

For plant in figure 8--1 it is shown that a concrete series size LT-3050, of the genus "Lauth three-high," belongs to the family of sheet rolling mills, that these (and others) belong to the class of rolling mills and these in turn (and others) again to the phylum of metal processing plant. We have set the question, on which level of this hierarchy it would be meaningful to situate the fundamental Specialized Design Science.

In plant construction, predominantly single pieces are manufactured, and it would surely be doubtful to construct the Specialized Design Science on the level of "species." >From this point of view, and with regard to the effectiveness of such information systems as goal, we must set the selected level as high as possibly. Here the context to concreteness (e.g., the representation of a plant) tends to be lost, and perhaps also too much of the similarity aspects. But also with these reservations, the highest plant level appears favorable for situating the Specialized Design Science, at least for a start. The knowledge system for designing can then be completed with the broadest special knowledge for individual levels (compare figure 8--5).

The state of knowledge for constructing the Specialized Design Science presents itself on this level as very favorable. Apart from the already mentioned knowledge (figure 8--2), extensive further knowledge is available in the published literature, and especially in the documentation and the basic information of known project engineering organizations.

The complexity level III of technical systems -- machine, apparatus, instrument, device (also house, bridge) -- belongs to the best known on the part of users, but also on the part of manufacturers and designers. Generally this level is supplied with the designation "product" -- something manufactured which is to be sold. Mostly these are products in commercial trading, because this unit is a useful and profitable object. For an enterprise (organization) these products form an output, which necessitates the existence of the firm. General explanations about design processes in Section 7.4 are valid as analogs for all complexity levels.

Question:
What characterizes (especially for designers) the complexity level III?

The description of this complexity level of technical systems can be established with following features:

• a technical system (TS) with capabilities which are demanded by the consumer (technical unit);
• a TS, which forms a spatial unit;
• TS with extensive diversity of types, each with very different demands;
• TS with continuously increasing requirements in depth and breadth;
• TS which are manufactured mostly in large to very large numbers of pieces;
• TS which are found mostly in a development series, have many predecessors, and at the same time are accompanied by members of a size range and/or by competing products.

Questions:
1. How does the designing of machines proceed?
2. Do certain demands on designing exist?

When designing this complexity class, the phase of conceptualizing (compare figure 7--13) is rarely needed, because mostly the new products do not change the existing mode of action (i.e. the organ structure remains unchanged). The goal is mostly a further development, therefore adaptation or transformation designing, as well as variant designing. The emphasis lies accordingly at the highest abstraction of the preliminary layout, in which the individual organs and their forms are established.

In the layout stage, designers strive on one hand to realize the new requirements, to become marketable, on the other hand to increase the fitness for manufacture, and to lower the manufacturing costs.

Designing is also adjusted to the quantity of the product to be manufactured, and must consider the high demands of production and assembly. In most cases a prototype (or a set of prototypes) is first prepared, realized (executed, manufactured), and tested, after which the documentation for series and mass production is executed. The degree of automation of the manufacturing plant must also be taken into consideration, because the questions of raw material costs, and rational production and assembly (including robot assistance) play an especially important role. Despite the implied sequence of prototype to mass production, design and production strive towards the common and simultaneous processing of the range of problems, by using processes and methods such as "simultaneous" or "concurrent engineering," TQM and QFD (see also Chapter 1, and Section 7.1.3.6), so that problems of manufacturing are already avoided in the prototype.

A characteristic feature is that the design time is shortened as much as possible, all work proceeds under time pressure.

The design process in its structure and procedure depends on the size of the enterprise (organization) and thereby also on the available design technology.

An important element which influences designing, is the list of requirements (the requirements or design specification), which establishes the complexity of the products to be designed. Some further elements can be seen from figure 7--14.

Questions:
1. Is a special technical knowledge needed, and where is it needed?
2. Does this knowledge exist, and in which form does it exist?

From the description of the characteristic features of the machine and their design process in our last paragraphs it is clear that with respect to the emphasis of Design Science there are few methodical problems in this class. The biggest concern is the branch (domain) knowledge and, connected with the abundance of properties of the future technical system to be achieved, the knowledge about the individual properties. We refer to the knowledge in the north-western quadrant of figure 5--2 and Section 7.3, which describes this area in general. A different situation can emerge in designing and developing new products.

In the individual specialties (branches) particular demands emerge, which also require a special knowledge as peculiarity of this area. Let us take as example the food and pharmaceutical industries and their machines. For all these machines, a particular demand that occurs in no other area is made -- the possibility and need for ease of cleaning. Due to the severe hygienic regulations, no machine can reach the necessary quality without this capability. The technical knowledge to achieving friendliness for cleaning must be extensive, to produce the following capabilities of the system:

• a certain state and condition of the effector surfaces, which make cleaning possible;
• disassembly capability of the machines;
• accessibility of the areas to be cleaned and disinfected;
• a certain surface quality (no porosity which could support the growth of microorganisms);
• if possible, no complicated forms, but only simple forms;
• suitable coverage (enclosure) from outside, so that disturbances are avoided;
• no touching of sensitive metals by aggressive materials (especially foods);
• application of special sealing and containment organs.

This knowledge is mainly available, otherwise such machines with the desired qualities simply could not be built and used. However, the range of knowledge varies from maker to maker, and the form of this knowledge is not always the most favorable for designers. Its necessary revision for Design Science is no mere formal exercise. Systematic questioning reveals not only many possibilities, but also large gaps. Such a process lets the old question emerge again, namely on which taxonomic level the range of problems should be processed. Where can the information system -- the Specialized Design Science -- be set up, to not only efficiently, but also economically fulfill its function. This is discussed in the next section.

Question:
On which taxonomic plane can the fundamental Specialized Design Science be meaningfully situated, so that an optimal information system emerges?

In engineering practice, the task is frequently set to design a concrete object -- a technical system -- so that it can then be manufactured. This object can be a weaving loom, a tractor, a press, a street, a high voltage electrical transmission line, or a lathe. Figure 8--3 shows some further areas and machines to illustrate the existing diversity. In addition, the complexity levels III and IV are connected with individual branches of the economy.

The central demand for us is to prepare an information system that can be made available for designers. We proceed on the assumption that all designers must acquire a certain knowledge system. The most important element of this knowledge should be formed by the Specialized Design Science, with which we are now concerned. The Basic Specialized Design Science (BSDS -- see below) should be built on one of the planes of the taxonomy (figure 8--4). Should this happen on the plane of the "phylum" or directly on the "species" plane, where the designers work? Or should several Specialized Design Sciences (SDS) be set up, or even on all levels? And how should the specialties of the other levels be incorporated into the system, if only one or a few SDS exist?

Figure 8--5 shows a possible construction of the technical knowledge system with such structure (on a BSDS). The conventional circle (according to figure 5--2) representing the General Design Science (GDS) on the plane of the TS is projected not only onto the structure in the Specialized Design Sciences (SDS), but already contains the general knowledge and transmits it. The indicated segment in the north-west quadrant shows as an example "design for manufacture." The next level drawn in this diagram, the "phylum" level (example machine tools), is chosen as "Basic SDS." This science should be set up as completely as possible (e.g., in a fundamental book or a data base). All knowledge systems below this level, on the levels "families" to "species," expand and concretize the knowledge through corresponding "completions." The figure shows not only the completions from the TS-plane, but also those originating from the concrete organization and even the concrete case (series size, special wishes). The latter elements can already be considered on the higher planes of the hierarchy.

Questions:
1. Which are the contents areas?
2. What is the state of knowledge in individual areas?

We can assume at this stage that the structure of contents is known for the General Design Science (GDS). Figure 5--2 reflects the four fundamental quadrants. Each Specialized Design Science can be derived from this model and will contain the same parts, which are, of course, filled with the knowledge covered in the specialty.

To develop an accurate idea of the contents, we briefly describe these fundamental knowledge elements, and machine tools should serve as example of the product families.

A further remark is needed to the terminology: we will use the term "Basic Specialized Design Science (BSDS)" consistently as a technical term (terminus technicus), it designates a knowledge system for designing at a specific complexity level and in a branch (specialty).

Theory of Phylum of Technical Systems (TS-phyla, e.g., Machine Tools -- south-west)

Starting from the "map" in figure 5--2, we ask about the contents of the south-west quadrant. In accordance with the Theory of Technical Systems, the quadrant contents can be specified through those part segments which are shown in figure 7--9. The machine tools are analyzed and described from all these points of view. Thereby the typical functions and features of this phylum are found and all solutions cataloged. This analysis is understandably further deepened, if we consider and process the classes within the phylum, in our case the lathes, milling machines or planing/shaping machines. For these, the new knowledge is captured in the completions, or if this would be beneficial, the basic plane could be put on one of the deeper planes. Thereby we could generate the Basic Specialized Design Science (BSDS) for lathes, milling machines and further classes, instead of only for machine tools. Particular attention should be given to the specific properties of the phylum (or the class), for instance in our case to precision, stiffness or stability. The Basic Specialized Design Science could also be set up at two levels, e.g., phylum and classes.

Theory of Design Processes for TS-Phylum (south-east)

The theory in this sector should describe the special design processes for creating the phylum and the classes. It will lean largely on the general treatment in the General Design Science (GDS) and emphasize only the specific conditions of the field, be it in the area of the designers, their working methods, representation methods, working means, management or working conditions in designing a product phylum or a class.

The statements in the two northern prescriptive quadrants are considerably more concrete in comparison with the two southern theoretical quadrants.

Technical Knowledge about Objects of designing (north-west)

Whereas in the theory of the TS-phylum the structures, laws and forms of construction, properties and all other laws are described, the contents of this quadrant are directed towards the provision of practical knowledge. The question to be answered is: "How do I achieve or realize this structure, property and further features?"

Especially the fitness of the proposed TS for different life phases should be supported through suitable instructions which directly serve the creating designers. The instructions as "know-how" of the designers are generated as individual, almost independent knowledge segments. Some of these are fairly well known, such as design for manufacture and assembly (e.g. [48,49,80,81]), and may even to some extent be computer-supported [80,81]. Others wait for a first formalization, such as design for maintenance, the environment, among others. If we say, that they wait for a first formalization, this does not mean that no relevant knowledge exists. It is usually found in other areas, or in a different form that is not so appropriate, directly usable or understandable for the designers. The large step will be fully done if (for instance) the "design for manufacture" knowledge is derived from manufacturing technology.

Concerning the distribution of this kind of knowledge on the individual hierarchical planes, this is formed unlike the relevant theory. A much more concrete statement can already be made on the upper, more general levels, because the majority of the knowledge is universally valid. On the lower planes only specialties are added, such as instructions for machine tool beds made of reinforced concrete as part of the "design for material and manufacturing" area, which are certainly not treated in any other specialty.

Technical Knowledge about Designing (north-east)

In this north-east quadrant we expect to find the practical knowledge about designing, the processes. At present, the predominant share of this knowledge is formed by methodical references: the procedural models recommended to be followed, and also the methods which support the completion of individual design steps. Additional partial segments should give practical references for application of the working means (e.g., computers), management or working conditions.

This knowledge area, in comparison with the others, has been filled in recent years mostly through design research. Unfortunately, this happened mainly on a general level, where (because of attempted universality) no sufficiently concrete methodologies can be formulated. The general references and rules of conduct (compare figure 7--14) are to be adjusted not only for the branches, domains and specialties, but also for well defined conditions of a certain enterprise organization, or even for one particular design group or certain designers. As the practice shows, this adjustment is not as simple or trivial as we may initially suppose. This is certainly one of the elements, which can explain the present lack of broad application the design methodology in engineering practice.

The close relationship of the working methods with the object and/or with the level of the branch (domain) is already evident in figure 5--4. For instance many design characteristics are already defined for an individual branch at the start of a design process, i.e. designing begins from the organ structure (or even only within the component structure) which is determined for the TS-class (compare Section 7.4). This fact, well known in the practice, has not up to now been clearly and explicitly expressed in the theory.

Summary

• State of the Knowledge

  The presented model of the Design Sciences at different levels of abstraction, which appear as elements of the holistic knowledge basis for designers, shows clearly the advantages of such an information system in comparison with the present practice. The impression could almost emerge that no large difference exists between these systems. That is deceiving, because many of the knowledge elements are certainly available. Their displacement and/or transformation brings order, which is indispensable not only for the theory but also for educational instruction and practice.

• Form of the Knowledge

  In constructing the Specialized Design Science, we should not forget its form. We have already repeatedly underlined this aspect and validated it with examples (figures 7--21, 7--22, 7--23). With the example of the knowledge about representation (engineering drawing) in figure 8--6 we again want to show the importance of a clear formulation and presentation of a knowledge complex, to aid judgment and to copy.

The complexity level of lowest technical systems is formed by assembly groups, mechanisms, organisms (II) and elements, components (I), i.e. the component structure. The functions of these technical systems are often problematic, and an exact coincidence "elementary function = elementary technical system" does not and cannot exist. Especially for the formation of assembly groups, organizational aspects (e.g., sequence of assembly) can predominate over functional aspects. We treat both of these complexity levels at the same time, because their range of problems is closely related and they are both represented in engineering by the discipline of "machine elements."

Question:
What is characteristic (especially with regard to designing) for these technical systems?

The characteristic features in the class for complexity levels I (elements) and II (assembly groups) can be listed as follows, whereby the varying and abundant forms of embodiment of individual features are typical:

• technical systems (TS) for which the task that they have to fulfill in many different conditions is simple, but that are not exactly designed for a specified block of functions (therefore not as organs, function carriers, or functional units);
• TS which occur as "species" in many different machines (i.e. gear wheels, bearings, shafts, levers);
• TS which occur either in mass production (screws, rolling bearings, seals) or in single-item or small series production (special casings or shafts);
• TS that are the originators of all properties of the higher technical systems. Their form, raw materials, dimensions, surface qualities and arrangement in space (configuration) are those design properties which determine all other properties.

Questions:
1. How does designing proceed?
2. Do special demands on designing exist?

Designing the assembly groups and elements (components) contains concrete decisions about all design properties. Therefore this phase has an especially large importance for the quality of the products (compare figure 8--7). Design methodology should therefore offer exact procedures for their individual (different) classes. A very concrete set of technical knowledge must accompany the concrete methodical references. We will return to this aspect later.

Specific procedural plans can be derived (compare figure 7--14) with ever more concrete contents on different taxonometric planes. On the upper plane of the machine elements -- components, figure 8--8 shows a procedural plan as example, which on one hand is founded on the general procedural model (figure 7--13), on the other hand defines how the individual design properties and characteristics can be progressively established. The assignment of the individual steps to the generally defined stages is obvious here. The elements and components become definitive and complete after a set of repetitions and iterations (at least twice: in the preliminary layout and in the definitive layout), by repetition of the steps (iteration of values of properties) in the flow chart. The necessity to approach the goal by repetitions (iterative procedure), is conditioned by the complex relationships among the design properties. They are not only in a complex relationship to the demands, but are also mutually in different relationships. Designers must work first with assumptions, which can only be defined in the second (or third) process. The flow chart shows a possibility for optimization, namely the work divided into certain blocks with internal feedback (return) loops, to avoid too many repetitions of the full procedure.

Figure 8--9 presents another flow chart, which represents a procedural plan for laying out of hydrodynamic bearings (a family of connections). We do not want to enter into details of the procedure in this context. If we now compare this process with that in figure 8--8, or even with the general procedure in figure 7--13, the difference can clearly be seen: on the general plane of generic TS, mostly extensive and generalized processes appear; in contrast, the flow chart of figure 8---9 specifies rather simple operations, such as "determine the Sommerfeld number." (Note and caution: the definition of Sommerfeld number in figure 8--9 is conventional for most of Europe. The inverse definition is used in other world regions. The scales and shapes of the value diagrams in the literature may therefore be different.) The references are much more concretized, and therefore also the objectivity of the decision, because the decision criteria are concrete and simple. For instance, "if the dimension e is numerically smaller than 0.6, then ... ." It is clear, that such an instruction could already be executed by the computer, for this statement is algorithmizable.

With these examples we have shown the particular situation at this complexity level, and the analysis has demonstrated the differences between the procedural models and a procedural plan by juxtaposing them. We could also recognize how important the design methodology for this area. The almost exclusive concentration on the level of machines (complexity level III) of the existing design methodology, as reported in the literature, has only little justification. This is probably another facet of the big picture of reasons why design methodology has lacked broad applications in engineering design practice.

Questions:
1. Is a special technical knowledge needed, and where is it needed?
2. Does this knowledge exist, and in which form?

The example of procedure for laying out of hydrodynamic bearings (figure 8--9) is not only instructive for the methodology but also for the relation with knowledge. We learn from this example in which quantity special knowledge is needed for the procedure and for each individual step.

Two different kinds of the necessary organized knowledge exist:

1. The "engineering knowledge," referring to physics, chemistry, and others areas of knowledge, which must be available for the acquisition of the function, reliability, strength and all further properties of the technical system. Beside the relationships of individual sizes, e.g., in scientific and empirical formulae, other data and values (both science-based and heuristic) must exist as individual characteristic magnitudes so that the parts can be laid out. These are for instance the permissible stress values, the possible wall thickness of a casting, the normally used maximum flow rate, and others.
2. A further kind of knowledge emerges especially in the connection with considerations about procedure and CAD. This should give designers and the computer additional items of information. These can be catalogs of the solutions of individual problems emerging in the procedure of designing, but also masters (compare Section 8.4) of individual documents of the methodical series, such as a list of requirements (design specification), function structure, sketch of principles (organ structure, concept), among others. The master list of requirements for bearings (part 1), and masters of TS-structures (part2), as shown in figure 8--10 can serve as an example of this form of process knowledge.

The "engineering knowledge" for individual families was explored, researched, published and taught as part of the discipline "machine elements" for almost hundred years. It is just over 100 years since C. Bach (Die Maschinenelemente -- The Machine Elements, Stuttgart 1891) published a new view about the calculation of machine elements (and wrote: "... the method of relationship numbers has had its foundation removed.") The contents of machine elements has since then remained quite stable in its structure, but in newer works we can find methods and values for individual areas and individual operations. Altogether, knowledge is amply available for this area -- even at times in a form suitable for designers. Many computer programs exist, and gradually also knowledge-based systems are being developed which are capable of form-giving for whole organs in dialogue with the designer. The program CADOBS for "rotating shafts and bearing arrangements" is quoted here as an example [152].

The situation of the second kind of knowledge is not nearly as satisfactory. Some areas have been processed, but more in the form of examples and not as systematic knowledge.

Question:
On which taxonomic planes can the Basic Specialized Design Science be meaningfully situated, so that an optimal information system emerges at this level of complexity?

In the treatment of the higher complexity levels, the principle of a taxonomy has been presented and explained using the example of machine tools. The order at the complexity levels I and II discussed here can be set up on similar way. As an example for this we show one of the best known groups of the components (machine elements) -- connections. Let us follow the origin of this hierarchy in figure 8--11, but first we present a terminological remark: figure 8--11 uses the term "organ" as the "means to solve a problem" (compare figure 7--3), something that fulfills a certain function --- a function carrier. Although the terms "organs" and "components" (elements) are not identical, the diagram is valid for both.

Now we can discuss figure 8--11, and the scheme and examples of connections. "Connections" in general are placed on the phylum-plane, with other organs that carry different functions. The question about the degrees of freedom of the connection gives rise to the origin of several classes: the first with no degrees of freedom (the fixed connections), then those with 1 or 2 degrees of freedom, among others. According to the direction of the degree of freedom, a relative axial motion (guidance) or rotational motion (bearing) of a part can exist. The latter needs also the prevailing force direction as radial or axial (thrust) bearing. These lower classes of connections are then arranged according to the kind of friction: the family of rolling bearings with rolling friction, the family of sliding bearings with sliding friction. Consideration of lubrication of the lower classes yields dry bearings, hydrodynamic and hydrostatic bearings. The way in which lubricant is supplied bring pressurized or self-supplying bearings, to which all the further lower classes belong. The form the lubrication gap defines the species of the sliding bearing connections: circular bearing, two-arc bearing, or segmented bearing (fixed or pivoted segments). These can be associated again with individual sizes or concrete type sizes. Let us remember here the flow chart, figure 8--8, which partly algorithmizes this step.

The hierarchy described above has many applications. This hierarchy is not only an ordering, it lets us also recognize that such a diagram demarcates the possible step-wise solution path for designing of connections. We again refer to the possibility that such a representation, accompanied by precise conditions for the steps downward (compare figure 8--9), forms the basis for a computer program.

If we now put the question where the basic specialized design knowledge should be situated in this hierarchy, then the experience in the area of the machine elements can help us. The published literature has in practice been divided into two main streams. One presents works about machine elements (e.g. [110,146,236,377,395,406]), i.e. the highest phylum of general technical systems of the first complexity level. The other treats the connecting elements, bearings [66,158,185], i.e. the phylum, class or family plane. The knowledge on the lower planes is concretized through the corresponding completions (TS). We consider it expedient to take over this practice for the construction of the knowledge system for designing, which brings many advantages.

That would mean, in such a model, that we should process the machine elements into a Basic Specialized Design Science (BSDS), and to plan and work out the TS-completions systematically. With regard to contents we would be concerned about working out the part about designing, because the currently available publications treat this problem very sparsely, and about completing further parts about machine elements.

Harmonizing the view about the systematics and other areas of machine elements with the Theory of Technical Systems will produce no revolutionary changes, contrary to initial expectations. It is important as introduction for a uniform set of ordering systems, whereby the general order is facilitated.

The treatment of Design Science with regard to the complexity levels was useful, but it has not led us to an information system for designers.

In engineering practice, designers always have the task to design a concrete object -- a product, an artifact -- we use the inclusive term "technical system" (TS). This may not only be a machine in general, but possibly a certain boiler, or a particular steam turbine, or a bridge, or a house, a transformer, a high voltage instrument, a crane or a personnel elevator. The enumeration of these particular objects (TS) would certainly be very long, even endless. This enumeration would be different if we consider more generally bridges, cranes, elevators, or even lifting equipment in general: here we are already on the plane of the product families (compare figure 8--4), namely the branches or specialties. The kinship of individual products in a product family can be different. It begins mostly with the common function (or functions), and especially with partial functions, and continues with the application of similar modes of action, manufacturing technologies, raw materials, ways of constructing, forms and a series of further historically conditional features. This is the reason why the knowledge of these areas can be developed in similar ways, can be jointly set up, and why also the transfer of experiences can be at all possible.

For designers an expedient field of activity emerges with a product family, on which they bring their experiences to bear. The broader and deeper the kinship of the products, the more favorable can be the exploitation of the available technical knowledge in individual families.

The present structure of the branches, specialties and/or product families was generated pragmatically both in the theory (discipline books, compare the titles in figure 7--1), as well as in practice. In the theory, the function (purpose) or the action principle, and possibly other features, have prevailed as ordering principles. The diagram in figure 2--1 indicates some ordering possibilities in a branch (specialty). The formation of the branches in practice was influenced sometimes by production kinship (e.g., instrumentation mechanics), which for Design Science is not a favorable kinship.

In any case the existing order is not homogeneous, and cannot be taken over without correction. On the other hand, otherwise introduced information systems exist, as well as a certain order among the quantity of the addressees to which Design Science turns, and which have a clear task field.

It cannot become the goal of Design Science to set up and recommend a certain branch order. The goal can only be to collect the necessary knowledge of the specialty for the design process, to organize and to process it, so that it is available in favorable form.

If we remain at the complexity level III -- machines, then the analyses of designing and technical knowledge in Section 8.1.2 have been comprehensively exercised, and can serve as basis for all specialties on this plane. A new element will likely be the level of originality, i.e. possibly new problems will emerge in the development of novel products, which are unknown for the daily cases. Also the number of items to be manufactured can be an influential factor.

At the complexity levels I and II, a series of specialties emerges beside the connections (see figure 8--11), but also in the lower levels of the family. The fact that mass production is possible, and very demanding manufacturing technologies are necessary (e.g., for rolling bearings), allows a deeper specialization. At a factory for rolling bearings, the technical and economic parameters now depend more on the construction of the manufacturing processes than on the construction of the bearing as product. The considerations about design process and technical knowledge, which were treated in Section 8.1.3, form also the basis for all components-related specialties, just as for the machines.

The most important ordering features which have caused us to treat the Specialized Design Sciences were degrees of complexity and product families. Of those classification characteristics enumerated in the beginning, three still remain: level of originality, number of pieces and organization size.

When considering the degree of complexity and the product families we mentioned everywhere further influencing factors, and figure 8--5 indicates also the corresponding completions in the whole system. Therefore we will not enter into further details.

One type of technical system which is very special, and has up to now remained without explicit treatment, although we have spoken repeatedly about its range of problems and task. It deals with processes as systems of operations, and with the range of problems of designing of processes, as distinct from designing TS-object systems. Especially in connection with plant we mentioned the importance of the process and the technology.

Some areas exist in which this range of problems already exhibits independent works, such as Technologische Betriebsprojektierung (Technological Works Project Engineering) by W. Rockstroh [368]. The boundaries are very fuzzy, and often the whole plant project engineering task is covered. Also in the book SLP [313] referred to above, the process is presented as "R-routing" and arranged among the five influencing factors (compare figure 8--2).

The combination of designing with some specialties such as chemical engineering appears at least unusual for a mechanical engineer. But also in such specialities, the task of the chemical engineer, namely "designing the process," has been acknowledged and integrated into education. In the research towards Design Science, chemical engineers have also participated, e.g., S.A. Gregory [175,177].

In the previous sections of this chapter we have analyzed the formation of the Specialized Design Sciences at various levels of complexity of the TS, in several branches (TS-families), and also at various levels of abstraction.

In order to increase understanding of the differences between the various kinds of SDS, figure 8--12 contrasts the statements of the General and the Specialized Design Sciences. The selection of statements should, where possible, reflect the whole breadth of the knowledge, and thus contribute to better understanding. The individual statements are augmented by references to suitable examples, mostly diagrams. These figures are in part published within this book, or were published in the Theory of Technical Systems [214,219], Engineering Design [228], or Theorie der Konstruktionsprozesse (Theory of Design Processes) [202].