Design Methodology for Process Improvements and Innovative Light Applications

Available online at www.sciencedirect.com

ScienceDirect Procedia CIRP 39 (2016) 57 – 66

TFC 2015 – TRIZ FUTURE 2015

Design methodology for process improvements and innovative light applications Andreas Roderburga*, Jan Reyb a Vossloh-Schwabe Lighting Solutions, Carl-Friedrich-Gauß-Str. 3, Kamp-Lintfort 47475, Germany Laboratory for Machine Tools and Production Engineering (WZL) of RWTH Aachen University, Steinbachstr. 19, Aachen 52074, Germany

b

* Corresponding author. Tel.: +49-2842-980-141; E-mail address: [email protected]

Abstract This paper consists of two main parts. The first part is scientific-oriented and introduces a new design methodology approach which was originally inspired by TRIZ principles and has been developed within the former author’s research environment. The aim of the methodology is to systematically support the development of new and innovative solutions in manufacturing technologies, based on herein identified principles of process hybridization. Although the methodology was developed within the focus of manufacturing process technologies, it can be used also for the development of applications in other domains which can be traced back to the general principles of process modelling. Thus in a second part of the paper it is discussed from an industrial perspective the use of the design approach for developments in non-traditional LED-lighting applications.

© Publishedby byElsevier ElsevierB.V. B.V. This is an open access article under the CC BY-NC-ND license © 2016 2015 The The Authors. Authors. Published (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of scientific committee of Triz Future Conference. Peer-review under responsibility of Scientific committee of Triz Future Conference Keywords: Design methodology, hybrid manufacturing technology, LED lighting development, process design

1. Introduction Innovations in production technologies are a key factor for the competitiveness of enterprises. Process innovations can enable new product features or functionalities and in addition follow the goal of greater process stability or shorter production times. An innovation-oriented process development strategy can also significantly contribute to the development of technological capabilities. Compared to product innovations, innovations in manufacturing processes are usually more difficult to imitate and result in a sustainable technological lead in an international competition - especially with regard to the risk of imitation. Fundamental developments of manufacturing technologies can be implemented through a combination of different technologies in terms of an interdisciplinary development and the integration of technology knowledge from different technology domains. In the field of manufacturing technologies the technology-across development of hybrid manufacturing technologies represents an object of study in order to develop a

new systematic design approach for manufacturing process technologies. Hybrid manufacturing processes are examples of technological developments with radical step-wise improvement of capability characteristics of a conventional technology. They provide a way to overcome limitations of conventional manufacturing technologies and enable a new process chain design. By means of using hybrid technology solutions, the field of application of conventional methods can be expanded for e.g. new materials or new geometries. Lasersupported turning and milling of ceramic components are examples of such technologies. Only by the use of laser, machining of ceramic materials became possible with defined cutting edges [1]. Fig. 1 gives the definition of hybrid (manufacturing) processes according to [2]. Focussing on the combination of processes, two main principles can be distinguished. On the one hand a secondary process technology can be used to support a primary technology to overcome limits, by adding or changing mechanisms, while the main function is still fulfilled by the

2212-8271 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of Scientific committee of Triz Future Conference doi:10.1016/j.procir.2016.01.166

58

Andreas Roderburg and Jan Rey / Procedia CIRP 39 (2016) 57 – 66

primary technology. On the other hand different processes can be mixed which on their own serve the same function (maybe by different mechanisms) – e.g. joining/welding of materials but when combining the technologies a new synergy effect is generated.

(I.A) Assisted Processes

(I.B) Mixed Processes

(e.g. laser-assisted turning, vibration assisted ginding, vibration-assisted EDM)

(e.g. EDM/ECM combination)

„From your point of view, what are the preconditions for the identification of promising technology combinations?“ (No. of participants n=37)

Aim-oriented integration of expert knowledge from single technologies A systematic procedure or methodology for a purposeful combination of technologies

Hybrid Processes (I) Combination of different energy sources

necessary intermediate results within the methodology that needs to be developed.

(II) Controlled Application of Process Mechanisms, which conventionally are done in separated processes

18,9%

16,2% 13,5%

48,6%

40,5%

A holistic knowledge base about single technologies

24,3%

37,8%

Abstracted solutions and active principles of the single technologies

16,2%

35,1%

27,0%

Uniform modelling of single technologies (e.g. technology limits)

18,9%

35,1%

27,0%

Is absolutely not the case

(e.g. grind-hardening, combination of material removing and forming)

29,7%

29,7%

24,3%

21,6%

Is absolutely the case

Fig. 2. Survey results about requirements for the development of hybrid manufacturing technologies. Fig. 1. Definition of hybrid processes acc. to [2].

In order to consider the users’ requirements for the methodology, an empirical-analytic study has been carried out. Representatives from industry have been interviewed to evaluate the needs for hybrid technology developments. The questions for the survey have been derived in an expert workshop. Fig. 2 shows as a selected result the most important requirements for the identification of useful manufacturing technology integrations from the perspective of the interviewed industrial partners. This clearly shows the need for systematic support during the development of hybrid manufacturing technologies. In addition, conditions listed here indicate the

The target structure for the development of the methodology shown in Fig. 3 was derived from the parent research goal and the above research questions. Process-related aims

Result-related aims Erratic improvement in technology by changing the process of manufacture

requires

2.1. Requirements

x How does the procedure model of a design methodology for the development of integrated process solutions look like? (See chapter 2.2) x In what forms is technology knowledge available and how does it need to be worked up? (Chapter 2.2 and 3.2) x How can the limits of conventional process solutions be described sufficiently? (Chapter 3.1) x How can the limits of conventional process solutions be explained sufficiently? (Chapter 3.2) x How are integrative/interdisciplinary solutions generated and how can models of different technologies be linked for that purpose? (Chapt. 3.3) x How can solution concepts be generated, evaluated and selected? (Chapt 3.3 and 3.4)

requires

delivers

Explanation of the process limits and their relationship between cause and effect requires

supports

For the herein introduced design methodology the combination of processes and different mechanisms is targeted. When regarding activities of hybrid technology developments in the past, it can be noticed, that a design approach for a combination of solution principles from different technology domains was missing. Therefore, the aim was to develop a methodology that facilitates a specific combination of different manufacturing technologies or mechanisms of action to hybrid technology solutions. The basic hypothesis for the design methodology approach is: Through a systematic analysis of technology data and the integration of technology models of different manufacturing processes new technology solutions can be developed (more) effectively within a methodology. Thus, existing limits of production technologies can be overcome. The approach is inspired by the principles of TRIZ, understanding innovation as a new technical solution which is solving a specific problem and which is derived in a systematical way by using solution principles from other fields of application. Especially the purpose of guiding the user towards a solution which is outside of the user’s scope is common with TRIZ. Thus solutions should be found by other/interdisciplinary knowledge, linking a problem and (the best) solutions even if the solutions are not known to the user in first instance.

Based on the requirements as well as an analysis of existing general design approaches the following main research questions have been derived for the methodology approach:

Procedure model for the description of a processing logic of the methodology and its single tools

2. Objective and basic methodology approach

delivers

Identification of process limits and description with the help of a contradiction requires

delivers

Detection and conditioning of knowledge of technology

Fig. 3. Target structure of methodology development.

59

As a framework, a process model was developed based on the procedures of general problem solving methodologies, referring to Daenzer [3] and Ehrlenspiel [4]. The procedural model refers to the assumption that at the beginning of every design process there is a problem for which the root causes are identified and solutions are developed during the design process [5]. Further this work strongly intended to consider the possibility of multi-dimensionality in problem solving process, making a comprehensive design approach more complex, referring to Matchett and Briggs understanding design as “a discovery of a way to harmonize all contradictory factors and relations in a multi-dimensional situation” [6]. Hence, the framework of the basic procedural model is divided into the four main phases of problem description, problem explanation, solution design and as a last step the selection and specification of solutions. The development of the basic concept can already be regarded as a new approach to the development of manufacturing technologies that picks up the product design methodologies and adapts it for the purpose of developing hybrid manufacturing processes. Fig. 4 shows the basic scheme of this methodology. I

Problem definition

II

Problem explanation

Definition of tasks & criteria analysis „ Detection of the target criteria and

phrasing the demands „ Detection and processing of

observations/data „ Illustration and evaluation of the biand multivariate correlations

Causal modeling „ Identification of causal coherences

(mechanism of action) „ Detection of additional system values and models „ Illustration and correlation of the relation between cause and effect „ Examination of hybridizationdocking points

Problem formulation

Model detailing

„ Selection of the problem relevant

criteria and phrasing of the problem „ Definiton of the area of consideration  Local and temporal expansion  Considered system components

„ Weighting of the cause-and-effect

relations

Analysis I

Synthesis

Selection

Problem description II

Problem explanation III

Solution generation IV

Solution evaluation

Solution: Improved process technology

2.2. Basic procedural model of the methodology

Problem (Contradiction): Limitation of a process technology

Andreas Roderburg and Jan Rey / Procedia CIRP 39 (2016) 57 – 66

Information acquisition and processing Technology knowledge (problem-related and solutions-related)

Fig. 4. Main procedural structure of the methodology.

A more detailed summary of the resulting methodology shows the range of typical single steps and methods which are part of the herein presented methodology, see Fig. 5. Furthermore the methodology should be understood as a basic compatible framework in which further methods can be integrated. Moreover, several steps might not be necessary for a specific case; the methods can be adapted by linking the relevant steps individually. It is obvious that within this paper only selected parts of the whole methodology can be described more in detail. Further details are given in [7]. III

Solution design Derivation of hybrid solution concepts

„ Interdisciplinary search for solution

concepts. Use of a solution catalogue for hybrid production processes „ Expansion of potential solutions by logical connection of cause-andeffect chains  Recombination of cause-and-effect chains  Formation of parameter families „ Application of TRIZ seperation principles for solving physical contradicitions

IV

Selection and specifications of solutions Solution selection and evaluation

„ Check of exclusion criteria  Material suitability  Local and temporal expansion of

the effects „ Evaluation of solution effectiveness

and risk for undesirable effects

Specification of solution concepts „ Concretisation of solution concepts

and design of solution elements „ Improvement of solution concepts

„ Sensitivity check

„ Check for hybridization-coupling-

points

Fig. 5. Single tasks and steps of the methodology’s procedural model.

2.3. Basic approach of knowledge acquisition and modelling Knowledge about manufacturing technologies is provided in models which are explicitly described or implicitly covered by expert knowledge. A model provides general statements about elements, structure, and behaviour of a considered part of reality. It is characterised by deliberately neglecting certain attributes in order to emphasise the essential model properties that are necessary for the modelling purpose. Thus validity and quality depend largely on the model's own level of detail. In many cases, models of manufacturing processes are given as a black box image of the process. However in reality the internal structure of a manufacturing process features causeeffect relationships. Because of that, black-box-models are often unsuitable for the prediction of system behaviour when changing parameter ranges or changing the boundary conditions as it is intended in case of hybridization. In order to analyse and explain the mechanisms of a manufacturing process, Brinksmeier proposes a differentiation of the black

box model into smaller model units [8]. Referring to this proposal, sub-dividing the system into its cause-and-effect relationships increases the validity of a model. The first way of model differentiation is based on a cause-oriented consideration of the system behaviour, distinguishing different mechanisms as parallel causes for process results. For instance, during the manufacturing process of grinding, one can distinguish between mechanically and thermally induced effects and their impact on the material properties of the workpiece [9]. Another way of differentiating the process model is carried out along cause-and-effect chains, by dividing single models into several serial-linked models. Therefore in addition to process specific input and output parameters, new parameters need to be implemented (such as energy-related or material-related parameters). Fig. 6 describes those principles of process model differentiation, following the above described approach of increasing the validity of process models.

Andreas Roderburg and Jan Rey / Procedia CIRP 39 (2016) 57 – 66

Partial model

Partial model

Cause model

Effect model

Output parameters

Process characteristics (e.g. force, temperature, …)

e.g. thermal

Fig. 6. Resolution of a black box process model by differentiation of partial models referring to Brinksmeier [8].

The above introduced two main strategies for model differentiating into sub-models can be combined. This leads to a network of sub-models depicting the cause-effect relationships of the system. One has to consider that by this approach an arbitrarily complicated cause-effect model can be created, which might not be necessary for further design steps. However, it only serves a purpose when a sufficient level of detail is achieved for a given problem. If this is the case, only those partial models should be used that describe the essence of the considered problem in an existing process technology. Same is valid for the modelling and documentation of solution principles, which need to be modelled as differentiated as necessary to provide most of the information about the potential of the specific process solution mechanisms in a sufficiently abstracted way. 3. Results of methodology development This paper presents only selected parts of the whole methodology for each of the main phases of methodology respectively: problem description, problem explanation, solution design as well as selection and specification of solutions. Nevertheless, in order to give a better overview at the end of each of the following sub-sections a list of main specific achievements and/or main specific features of the methodology development is given. 3.1. Problem description According to TRIZ the design of a new solution is starting with a contradiction of parameters, which are describing the problem to be solved. Contradictions are resulting in trade-offs during the optimization process of an existing production system when approaching the technological limitations. Thus as a starting point of overcoming technology limitations the target is to identify the contradictions within the system. According to TRIZ one can distinguish between (physical) contradiction of single parameters and technical contradictions between several parameters. A simple correlation analysis of system behavior might give a first overview of contradictions between two parameters if those are independent from other parameters. On the other hand it can result in misleading assumptions, when not considering the degrees of freedom in an existing system and/or not considering other relevant target parameters, which can lead to multi-dimensional contradictions. Thus the target of problem description is to first

Factor loadings F1

F2

F3

F4

Factor loading diagram F5

a1j

0,939 0,337 0,004 0,062 -0,011

a2j

-0,986 0,011 -0,086 0,139 0,003

a3j

0,940 0,336 0,054 0,024 0,012

a4j

0,910 -0,233 -0,342 0,006 0,001

a5j

-0,850 0,482 -0,204 -0,061 0,000

λj

4,2899 0,513 0,169 0,027 0,000

%Var

85,8

10,3

3,4

0,6

0

Factor analysis

Fig. 7. Exemplary illustration of factor loadings of requirements and factor loading plot for the first three factors.

As a result of the factor analysis the problem is described by the distribution of single criteria within an n-dimensional room representing the n degrees of freedom. In order to define the real number of degrees of freedom for a process, it has to be considered that a process features different working points, but not all describe the problem at the “edge of process capabilities, when problem appears”. Therefore besides of the established criteria – e.g. Kaiser-criteria, Jolliffe-criteria, variance criteria or scree test – a further extraction criterion has been developed for this method in order to reduce to the relevant number of factors. The problem can be then described focused on the edge of process capabilities. This new criterion is called eigenvalue difference criterion, further explanations see [7]. The assumption is that the problem is caused by processspecific latent mechanisms, which are fully represented implicitly by the model of problem description. Thus in order to explore the relevant mechanisms in the next step of the design methodology, the problem can be reduced to a model with representative parameters, approaching an ideal typical problem. Ideal typical means a description of a model which in reality is not given due to deviations in real processes, but can be used as a pattern to characterize a specific type of problem. Such ideal typical problems are described in Fig 8. a)

b)

2. Fak tor 1

z3

2. Fak tor 1

z2

-1

z2

1

1. Fak tor

Partial model

Input parameters

Partial model e.g. mechanical

Process (system) Output parameters

Input parameters

Process (system)

consider all relevant target parameters and comprehensively describe the contradictions in a multi-dimensional way. Therefore multivariate analysis is done by means of factor analysis, see Fig. 7.

-1

1

z1

z1 -1

d)

z3

-1

c)

2. Fak tor 1

z3

2. Fak tor 1

z2 z2 -1

1

1. Fak tor

Differentiation by causal cause-and-effect relationships

1. Fak tor

Differentiation by superimposed parallel mechanisms

1. Fak tor

60

-1

1

z1 z4

z1

-1

f)

-1

Fig. 8. Examples of ideal typical aim conflicts.

By this approach information is reduced, but keeping the representative parameters which are influenced by the latent process mechanisms, which needs to be analysed for a holistic

61

Andreas Roderburg and Jan Rey / Procedia CIRP 39 (2016) 57 – 66

solution. Based on the representative parameters the following step of problem explanation is derived, starting with the identified n-dimensional contradiction and using knowledge/models about mechanisms for deriving an explanation model as explained further below. Summarized the main specific features and/or steps of this part of the methodology are following: x Editing and modeling of observational data in the case of nominally scaled target criteria by using a logistic regression model as a mathematical approximation for a physical saturation process. This enables the factor analysis with different scaled target criteria of a process (metric/nominal). - Not further described in this paper. x Use of factor analysis in order to better describe the limits or problems in manufacturing processes with multivariate correlations. This enables to identify and describe multidimensional contradictions. x Development of a new eigenvalue difference criterion for factor analysis (multivariate statistics) in order to determine the degrees of freedom for manufacturing processes close to the process limits. This provides an extension of the known extraction criteria in factor analysis. - Not further described in this paper. 3.2. Problem explanation The above described method for problem description is useful for the case of complicated system relations and/or in order to ensure that system behaviour is not being oversimplified in problem description. Nevertheless, as a result there might be a definition of contradiction via only one or two (but the representative) parameters. After the problem has been identified and the representative target parameters have been determined, in the step of problem explanation the target is to model cause-and-effect relations which are representing the characteristic mechanisms of the process. This causal modelling is using and combining the above described model differentiation strategies according to Fig. 6. Thus an abstract model can be built up which is explaining the cause-and-effect relations between system input parameters and contradictory output parameters see Fig. 9.

Based on the graph approach a mathematical description can be created. The method for the mathematical modelling of the cause-and-effect relations is based on graph theory [10, 11, 12]. Acc. to that a cause-and-effect diagram is regarded as a directed graph, also called a digraph. The nodes of the graph represent the parameters or variables of the system, which are represented by cause-effect relationships and related to one another by edges (arrows). To determine the variables that have a direct or indirect influence on a different variable, the reachability matrix R(D) is the mathematical model of the digraph D, calculated with the adjacency matrix A(D), the identity matrix I and the number of nodes p:

y1 Part. model

Graphical representation Cause-and-effect chain of nodes A and B:

y2

y3

x1

y3

y1

x2

Part. model

y4

Part. model

x3

y3

y1

z1

x2 x3

y4

x4 x5

Graph model for two result parameters

x1 z1

y2

Problem solution by modifying one target parameter z1 (physical contradiction)

R

A

I  A  A2  ...  A p1

B C

D

D

F

E

R( D)

E

G

F

G

Dependency diagram concerning A & B:

1

D,E, F,G

(B)

B C,(A)

0 0

A

A A ª1 B ««0 C «1 « D «1 E «1 « F «1 G ¬«1

B 0 1 0 1 1 1 1

C D E F G 0 0 0 0 0º 0 0 0 0 0»» 1 0 0 0 0» » 1 1 0 0 0» 1 1 1 0 0» » 1 1 0 1 0» 1 1 0 0 1»¼

Dependencies of result parameters:

ŸA

f C DE, F , G

ŸB

f DE, F , G

Ÿ B z f C

1

Fig. 10. Principle of deriving the hybridization coupling points.

Part. model

Abstract modelling in terms of graphs Graph model for one result parameter

Mathematical representation Matrixform Reachability matrix R(D):

Prozess Part. model

zi Part. model

(1)

By help of the graph modelling all direct and indirect influencing parameter are determined for the contradictory aim parameters. In that way system parameters can be derived which have an influence only or mainly on one of the aim parameters. Those parameters have the highest potential for integrating solutions and are called hybridization coupling points (HCPs). In the context of TRIZ, those parameters can be seen as a kind of standard parameters for process hybridization or are used as initial point for deriving a standard parameter. Fig. 10 shows the principle of finding such hybridization coupling points.

Result of the combination of model differentiation strategies

­ x1 ½ °x ° ° 2° ® ¾ °° °¯ xn °¿

I  A( D)  A( D) 2  ...  A( D) p 1

R( D )

y4

x4

x5

y2

Aim conflict

z2

Problem solution by solving aim conflict between z1 and z2 (technical contradiction)

Fig. 9. Causal modelling in the form of a digraph.

For many cases an unambitious HCP can already be found as described above, thus in the methodology at this point the next phase of solution design can be entered. For the case that such an HCP is not found directly, a further weighting of relations becomes necessary. A weighting of relations is done via elasticity values of the relations between two system parameters. For time variant relations a problem relevant working point is defined, referring to Bjørke [13]. Under those conditions the characteristics of a relationship between physical parameters in a technical system can be described by elasticity values ε calculated as follows:

H y,x

i

wy xi ˜ wxi y

(2)

62

Andreas Roderburg and Jan Rey / Procedia CIRP 39 (2016) 57 – 66

The elasticity values can be integrated within the graph theory modelling in that way that elasticity values can be used instead of 0 and 1 and further the calculation of reachability matrix can be done in the same way. The weighted adjacency matrix Ag(D) derives from as follows:

Ag

ª a11  a1n º ­a ij «   » with ® » « ¯a ij ¬«a n1  a nn ¼»

(3)

0 for not adjacent nodes H j ,i for adjacent nodes

3.3. Solution design

Fig. 11 explains the general procedure by the help of an example. The target of the analysis with the weighted relations is to identify one or more candidates for a hybridization coupling point (X1 in this example). Principle of weighting of cause-and-effect relations y1 A(D) Y Y X X X X X

Example: Model for x1→y1

y1 y1

f ( x1 )

a ˜ x1

1

2

y1,OP

x1,AP x 1

H y ,x

Example: Weighting

wy1 x1 ˜ wx1 y1

H y1 , x1 H y1 , x1

2a ˜

1

x1 2 ax1

2

H y1 , x1

AP

1

2

3

4

x 1,AP x1

y2

5

0

0

0

0

0

0

0

Y2

0

0

0

0

0

0

0

X1

1

0

0

0

0

X2

0

1

1

0

0

0

0

X3

0

0

0

1

0

0

0

0

X4

0

0

0

1

0

0

0

X5

0

0

0

1

0

0

0

(+)

Y1

0

0

0

0

0

0

0

Y2

0

0

0

0

0

0

0

X1

2

0

0

0

0

X2

0

2

1

0

0

0

0

X3

0

0

0

-3

0

0

0

X4

0

0

0

1

0

0

0

X5

0

0

0 0,5 0

0

0

0

(+)

(+)

x1

0

Ag(D) Y1 Y2 X1 X2 X3 X4 X5

1

hier gilt:

H y1 , x1

2

2

Y1

(-) (+)

x3

x2 (+)

Y1 Y2 X1 X2 X3 X4 X5

Y1

0

0

0

0

0

0

0

Y2

0

0

0

0

0

0

0

X1

2

0

0

0

0

0

0

X2

0

2

1

0

0

0

0

X3

0

0

0

-2

0

0

0

X4

0

0

0

1

0

0

0

X5

0

0

0 0,5 0

0

0

+2

A

y1 +2

+1

+0,5

+1

+0,5

E

Y1 Y2 X1 X2 X3 X4 X5

Y1

1

0

0

0

0

0

0

Y2

0

1

0

0

0

0

0

X1

2

0

1

0

0

0

0

X2

2

2

1

1

0

0

0

X3

-4 -4 -2

-2

1

0

0

X4

2

2

1

0

1

0

X5

1

1 0,5 0,5 0

0

1

1

X4

2

X5 -4

-2

X3

X1

2 -2

1

E

F

G

B

2

Secondary technology in order to influence parameter C e.g. by heating: - Laser - Induction - Spark erosion -…

C

2 1

D

 0,5

2

e.g.: Laser integration

1

E

F

G

Fig. 12. Principle of hybridization illustrated within the cause and effect diagram.

X2

x5

x4

Potential of hybridization

A

C

x5

x4

Y2 4 +1

x3

 0,5

2

x2

-3

x3

D

-

+2

x1

0

2 1

C

x2

-2

B

2

y2

Weighted influences on Y1 und Y2

+2 +1

x1

If one or more potential hybridization coupling points could be identified, those are the origin for further steps with the target to change the mechanisms of the process and to solve the problem. In the case of hybridization the mechanisms of the process are changed by adding a new source of energy thus the identified HCP is modified or controlled in such a way that the contradiction in the target parameters is solved. Figure 12 shows the principle of hybridization within the cause-andeffect diagram.

x5

x4

Analysis of influences on conflicting aim parameters y1 y2

Ag

x Extension of models from graph theory: Weighting of relationships through the determination of elasticity values εj,i within the process operating point. x Development of algorithms and mathematical rules for the determination of all influences on the target parameters and for the identification of hybridization coupling points. - Not further described in this paper.

4

Y1

-4

In order to provide technology across solution principles, the knowledge about existing process solutions is been documented in an abstract way, similar to the principle of standard solutions in TRIZ. For hybrid manufacturing processes solution principles has been modelled by abstract cause-and-effect chains and listed in a specific solution catalogue, Fig. 13 shows a screenshot of an extract of the catalogue (developed in German language). Standard Parameter Modifikation

Since this requires a graph without feedback loops, in the case of system feedbacks the models can be converted. This is not described further in the paper on hand, but can be looked up in [7]. Further in the case that it is needed, a way of using fuzzy information is described, thus the deviations and also more detailed case-differentiations can be taken into account. Summarized the specific features and/or steps of this part of the methodology are the following: x Development of a modeling approach for the mapping of cause-and-effect relationships as an explanation model using model differentiation strategies (see Chapter 2.3). x Mathematical modeling of cause-and-effect relationships by the use of graph theory.

Pivot-Tabelle

Fig. 11. Principle of weighted influence analyzis

Filter Kraft reduzieren

Filter: Auswahl des gewünschten Effektes

FunktionsträgerAnzahl von Ketten ID technologie Ketten ID EDM EDM-002-1 EDM-002-2 Laser-001-1 Laser-001-2 EDM-001-1 Laser-005-1 Laser-003-1 Laser-003-2 Kondukt-001-1 Induktion-001-1 Laser-002-1 Laser-002-2 Laser-008-1 Gesamtergebnis

Standard Parameter Modifikation Temperatur Härte Kraft Verschleiß Temperatur Härte Kraft Zeitspanungsvolumen Temperatur Härte Kraft Verschleiß Temperatur Härte Kraft Rissbildung Temperatur Härte Kraft Umformbarkeit …

erhöhen reduzieren reduzieren reduzieren erhöhen reduzieren reduzieren erhöhen erhöhen reduzieren reduzieren reduzieren erhöhen reduzieren reduzieren reduzieren erhöhen reduzieren reduzieren reduzieren …

Laser

Konduktion

Induktion

1 1 1 1 1 1 1 1 1

3

Funktionsträgertechnologie EDM EDM EDM EDM EDM EDM EDM EDM Laser Laser Laser Laser Laser Laser Laser Laser Laser Laser Laser Laser …

1 1 1 8

1

Anwendung/Verfahren EDM-unterstütztes Schleifen EDM-unterstütztes Schleifen EDM-unterstütztes Schleifen EDM-unterstütztes Schleifen EDM-unterstütztes Schleifen EDM-unterstütztes Schleifen EDM-unterstütztes Schleifen EDM-unterstütztes Schleifen Laserunterstützes Drehen Laserunterstützes Drehen Laserunterstützes Drehen Laserunterstützes Drehen Laserunterstützes Drehen Laserunterstützes Drehen Laserunterstützes Drehen Laserunterstützes Drehen Laserunterstützte inkr. Blechumformung Laserunterstützte inkr. Blechumformung Laserunterstützte inkr. Blechumformung Laserunterstützte inkr. Blechumformung …

Gesamtergebnis 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 13

UW-Ketten KettenID Pos. EDM-002-1 1 EDM-002-1 2 EDM-002-1 3 EDM-002-1 4 EDM-002-2 1 EDM-002-2 2 EDM-002-2 3 EDM-002-2 4 Laser-001-1 1 Laser-001-1 2 Laser-001-1 3 Laser-001-1 4 Laser-001-2 1 Laser-001-2 2 Laser-001-2 3 Laser-001-2 4 Laser-002-1 1 Laser-002-1 2 Laser-002-1 3 Laser-002-1 4 … …

Fig. 13. Extract of solution catalogue for hybrid manufacturing technologies

63

Andreas Roderburg and Jan Rey / Procedia CIRP 39 (2016) 57 – 66

Extended solution design by recombination of cause –and-effect chains Direct Solution

Extendention of solution space

T1

B

A

T2

C

B

E

T3

D

Indirect solutions

B

Extended solution space C

B

A

Function carrier technology

Physikal contradiction

Laserl Optic

Removed material (by etching)

Modified workpiece material

D

Mechanisms

Removal rate as low as possible

Removal rate as high as possible

Focus point of the laser beam Etch bath

Resolution of contradiction by seperation in space Material removal No materia removal

Fig. 15. Principle of Separation in Space for In-Volume Laser-assisted Etching (ISLE).

Summarizing, the specific features and/or steps of this part of the methodology are the following:

A T..

T2 T3

Example for Separation in Space

Source of prictures: Propawe, Lecture Lasertechnique II, RWTH Aachen University 2008

C

T1

contradiction becomes obvious within the etching mechanism for material removal: On the one hand the material removal is needed for the purpose of in-volume structurization, on the other hand material removal is not allowed where the material needs to be remained. In the cause-and-effect modelling it can be derived the linked contradiction that the etching sensitiveness of the material needs to be high and low at the same time. Here the separation in space principle can be used in combination with the solution catalogue. Laser can be used to increase the sensitiveness of material (activation principle), if the material is transparent for the laser, the laser can prepare the material in three dimensions thus the material is removed by etching only at the locations where it is required.

Principle of In Volume Selective Laser Etching (ISLE)

This abstract description of technology is a result of mapping each individual production technologies by means of the above-described cause and effect modelling. The set of mechanisms of action is an abstract solution, characterized by a high degree of generality and thus applicable to different technology disciplines. Within this collection of solutions, a targeted search for certain mechanisms of action can be conducted, e.g., where these mechanisms cause a desired effect on the influencing variables via the HCP. As part for the filing of such a collection of solutions, the concept of the design or solution catalogues has experienced a broad application in the field of engineering design methodologies. Compared to other solution catalogues, this specific solution catalogue for hybrid processes specifically considers causal relations in terms of cause-and-effect chains. Each cause-and-effect chain considers the n different standard parameters which are influenced and the carrier technologies which are used in existing process solutions. Further this causal modelling enables to combine different cause-and-effect chains and thus creates more new solution approaches by recombination of different solution principles as shown in Fig. 14.

…

Function carrier technology Hybridization-couplingpoint (A) System parameters (B,C,D,E)

Functions

Fig. 14. Principle for additional solution search by the recombination of known cause and effect chains.

Within the here described methodology part of solution design established methods can be integrated, both from creative and systematic solution design. The TRIZ method of the four separation principles is given here only as an example among others, which is appropriate for use in combination with the herein described methodology, especially when it has not been possible to identify clear hybridization coupling points in the steps before. In that case the parameters from the causeand-effect diagram can be used in order to check for physical contradictions (which might cause a technical contradiction in the output parameters). In that case the separation principles can help to design a hybrid solution in which by the combination of technologies a separation in room, time, system level and/or by change of conditions can be realized. As an example in volume selective laser etching (ISLE) is shown. Initially within conventional etching a contradiction appears when specific structures should be generated, because a local 3-dimensional etching is not possible. The physical

x Transferring the principles of solution catalogs and standard solutions (TRIZ) from product design theory to the development of hybrid manufacturing processes. An abstract solutions catalog for solution principles in hybrid manufacturing processes has been created. x Extension of the “conventional” solution catalogs concepts: Instead of only single parameter entries, cause-and-effect chains are described. This enables a derivation of new solution concepts through logical linking and recombination of cause-and-effect relations. x Extension of the methodology by the aim-oriented integration of TRIZ separation principles for a resolution of physical conflicts within the cause-and-effect model. 3.4. Selection and specification of solutions This last section of methodology is transferring the solutions from the idea/concept state into the state of evaluation of concepts for concrete development projects. This phase is mainly evaluating the effectiveness of solution concepts and its risks (side effects) in order to priories and/or excludes solution approaches in an early stage. In that regard, the methodology is differing from a basic design paradigm in TRIZ whereby the best solution principles can/should be found directly, while creating multiple different ideas which might not work should be avoided. The deviation from this TRIZ paradigm is based on

64

Andreas Roderburg and Jan Rey / Procedia CIRP 39 (2016) 57 – 66

Material applicability Is the proposed hybridization technology applicable to the material which needs to be processed?

Extention of the desired effect in time and space Is a superposition of the procesesses possible in the required time- and space-related extent?

Evaluation of solution effectiveness and risk of undesired side effects high

gy A Technology

moderate

Solution Effectiveness

Technology B

Low invest Mid invest

low

Recursive process of solution specifications (referring to Pugh)

the assumption that when selecting abstract solution principles the complete information about effectiveness and risks of those solution principles is not available until the later steps of technology developments. Also when selecting the potential carrier technologies as described before, boundary conditions about the specific application are mostly not considered. Solutions concepts might need to be further evaluated, optimized and selected in a recursive process. For the hybridization of processes the proposed criteria for early evaluation of solution concepts are related to the material usability and the extension of effects in time and space. It is also taken into account the possibility of information incompleteness (e.g. about harmful side effects) which might need to be identified and solved in a recursive development process referring to Pugh’s design approach [14].

Technology C High invest high

moderate

high

Risk for harmful side effects

Fig. 16. Procedure for the evaluation and pre-selection of solutions.

The resulting evaluation of technologies considers the ideality of solution in terms of technical effects. Costs/invests are not further considered in the methodology yet. At this point the herein described methodology is linked to subsequent methodologies of technology planning and evaluation (e.g. technology roadmapping). Here also the availability and maturity of carrier technologies is relevant, thus as a result of the above described methodology different solution concepts might be interesting for further developments. 4. Applicability of methodology for non-traditional lighting development Within the originally considered field of manufacturing processes the application of the methodology has been validated via use-cases shown within previous publications, see

[7, 15]. Instead of giving another use-case from manufacturing technologies, within this paper the transferability of the above introduced methodology to other disciplines – where also aspiring for process developments – is discussed. Motivated by the recent author’s profession within the field of lighting technology developments, the above introduced design approach is evaluated concerning its use in non-manufacturing related developments using the example of biological active LED lighting activities. This example describes a research domain looking for innovative applications of solid state lighting, especially enabled by the tremendous evolution of components for LED lighting technologies during the last decade. While those developments on technology side offer new opportunities in the control of various different light characteristics – such as individualized spectrum design or space and time differentiated light signal modulation – on the application side today the market opportunities are only roughly surmised, the full potential of applications seems to be not known by far yet. Here the biological active LED lighting begins, describing a group of applications in which a specific lighting system is modifying the results of a bio-chemical process in such a way that limitations/contradictions of the conventional process (without LED system integration) can be overcome. Hence, LED light systems take over new functions by the controlled & aim-oriented influencing of the biological process and thus interact as a new “hybridization tool” in the biological process. To evaluate the potentials of using the above described methodology, an exemplary biological process is described by means of a generalized plant growth process. Fig. 17 shows (in a simplified way) examples of the light-related input characteristics of a plant growth process, selected process mechanisms and examples of relevant result parameters. Further Fig. 17 shows the sensitiveness of typical photoreceptors of a plant depending on the wavelength of the light. Each of those receptors initiates/controls specific plant growth mechanisms through the reception of light. It can be seen that different light wavelength activate different photoreceptors. By understanding the dependencies of the growth output characteristics from the separated activation of photoreceptors by light control, effects can be accelerated or even specific new effects can be reached such as e.g.: x x x x x x x

Higher growth rates More sugar, more vitamins, better taste More sprout shoots, more flowers, more fruits In-time controlled flowering In-time controlled ripeness Better pest resistance, better yield etc.

Andreas Roderburg and Jan Rey / Procedia CIRP 39 (2016) 57 – 66

Light characteristics

Process mechanisms

• • •

Intensity Spectra Light signal characteristics

• • • • •

•

Control over time and space

Chlorophyll a Phycocyanin

Photosynthesis Photomorphogenis Photoperiodism Phototropism …

Chlorophyll b Phytochrome Pr

Process output Quality • Size and Shape • Flowering • Rooting • Pollination • Pest resistance • Color • Taste • … Cost-rel. parameters • Energy saving • Yield • ROI

Carotenoids Phytochrome Pfr

Phycoerythin

1

Relative Sensitivity

0,9 0,8

hybridization coupling points in biological processes similar to the approach for manufacturing processes. A brought analysis of light-related mechanisms in biological processes has been started in order to create a collection of principle solutions. Such a “solution catalogue for hybridization of biological processes via light characteristics” also needs to represent cause-and-effect relations as abstract descriptions of solution principles. Those are representing the expert knowledge about the biological mechanisms in an abstract and compressed manner, thus it can be used for a transfer of solution principle in terms of interdisciplinary developments. Further the need for time and space separated effects in biological processes is obvious and by far not covered by suitable lighting solutions yet. Here the separation principles of TRIZ can be used and are technically realized by individualized LED lighting systems – adaptable in space and time – in order to solve physical contradictions.

0,7 0,6

5. Summary and Outlook

0,5 0,4 0,3

0,2 0,1 0 350

400

450

500

550

600

650

700

750

Wavelength [nm]

Fig. 17. Principle model of a plant growing process and its dependencies from light characteristics

Similar principles and strong dependencies from light characteristics are known also for other biological processes in animals, micro-organisms and even human beings. In case of humans, hormone production (e.g. melatonin) can be influenced. This leads to several application targets such as better learning, working or sleeping. Further various medical applications are considered. Those applications are grouped very often within the so called research field of “human centric lighting”. Today there are many interdisciplinary research projects ongoing in that field, involving experts from technical, biological and medical institutions. For instance VosslohSchwabe Lighting Solutions is investigating LED system solutions for human centric lighting applications within a cooperation research project named by the acronym OLIVE. In this project common research is currently carried out with several project partners with the target of optimized light systems for improvements in performance capability and health of humans. Medical examinations with new LED technology prototypes are carried out by the project partner Charité hospital in Berlin, in order to understand the effect of light characteristics onto the biological processes of humans. The time-related influence of different light characteristics on the biological processes (chronobiological effects) are analyzed and will be used to develop smart lighting systems for future medical applications and applications in daily people’s environment. However the solutions in detail will look like, also here the integration of new technological solutions can be regarded as a hybridization of (biological) processes. Thus as an interim conclusion it is deduced that the use of the methodology for hybrid manufacturing technology developments and its principles can help to create new lighting solutions for biological processes. It can help to identify

The objective of the herein described work was to create a methodology for the design of hybrid manufacturing processes as a specific form of soaring developments in manufacturing technologies. As a result, a methodology has been created which meets the special requirements of a suitable analysis and modeling of cause-and-effect relationships in production processes. Further the methodology promotes interdisciplinary (integrative) designs of hybrid process solutions. Thus a gap has been closed within the methodology landscape for the development of technical systems specifically to the area of production technologies. Parts of the methodology show close affinity to TRIZ principles. Those are regarded as methodology developments driven and inspired by a generic TRIZ approach. The designed methodology is generically applicable for different technologies. At the same time the methodology can be configured individually for specific applications. In designing new manufacturing processes, the methodology gives the possibility of steady expansion and specification of the knowledge base about solution principles and mechanisms in hybrid or non-hybrid technologies. Hence, the work in hand provides a new and valuable tool for future developments of advanced manufacturing opportunities and serves the interdisciplinary search for new high-performance processes. Within this paper for first time the methodology has been introduced and discussed in the context of non-traditional LED lighting applications, showing that the herein described methodology and its principles can be transferred to other domains outside of manufacturing domain. In that way it has been indicated that the methodology can be applied when the objective of improvement is described by a process which can be modified by the integration of new technologies. Based on this methodology, further research is ongoing in the field of manufacturing technology integration. While the scope of the herein described methodology is focused on the technological aspects of process combination, one has to be aware that this represents an early technology concept development. Further it has to be considered that new processes – as targeted here – are requiring complex machine solutions. Manufacturing technology integration may give new

65

66

Andreas Roderburg and Jan Rey / Procedia CIRP 39 (2016) 57 – 66

opportunities to generate added value in products or reduce the process chain through elimination of secondary processing times and transportation times. However, availability and maturity of technology components needs to be evaluated, further such technologies are in most cases more expensive than conventional technology equipment. Thus, the positive effects of advanced manufacturing technology integration must be pondered over the greater costs in an efficiency model. Today’s models are not giving sufficient evaluation. Because of this the co-author of this paper currently works on the subsequent project „Multi-Technology Production Systems“ within the cluster of excellence „Integrative Production Technology for High-Wage Countries“, see [16]. Acknowledgements The authors would like to thank the German Research Foundation DFG for the support of the depicted research within the Cluster of Excellence "Integrative Production Technology for High-Wage Countries" [16]. Further details of the herein described methodology development have been published in [7] and some basics of this thesis have been pre-published in [15]. The author would like to thank the German Federal Ministry for Education and Research (BMBF) for the support of the project “Optimierte Lichtsysteme zur Verbesserung von Leistungsfähigkeit (OLIVE)” mentioned in Chapter 4. References [1] Klocke F, Zeppenfeld C, Pampus A, Mattfeld P. Fertigungsbedingte Produkteigenschaften - FePro. Apprimus Verlag, Aachen, 2008.

[2] International Academy for Production Engineering (CIRP): 5th CWG meeting on Hybrid Processes. January Meetings, Paris, January 25-27, 2012. [3] Daenzer W F, Büchel A.. Systems Engineering. Verlag für industrielle Organisation, Zürich, 2002. [4] Ehrlenspiel K. Integrierte Produktentwicklung – Denkabläufe, Methodeneinsatz, Zusammenarbeit. 4th Ed. Carl Hanser Verlag, München, 2009. [5] Hubka V. Theorie der Konstruktionsprozesse – Analyse der Konstruktionstätigkeiten. Springer-Verlag, Berlin/Heidelberg, 1976. [6] Matchett E, Briggs A H. Practical design based on method (fundamental design method). In: Gregory, S. A. (Ed.) The design method. Butterworth, London, pp. 183-199, 1966. [7] Roderburg A: Methodik zur Entwicllung hybrider Fertigungstechnologien. Dissertation of RWTH Aachen University. Aachen: Apprimus Verlag; 2013. [8] Brinksmeier, E.; Aurich, J.C.; Govekar, E.; Heinzel, C.; Hoffmeister, H.W.; Peters, J.; Rentsch, R.; Stephenson, D. J.; Uhlmann, E.; Weinert, K.; Wittmann, M.: Advances in Modeling and Simulation of Grinding Processes. Annals of the CIRP Vol. 55, Nr. 2, S. 667-696, 2006. [9] Nachmani Z. Randzonenbeeinflussung beim Schnellhubschleifen. Dissertation RWTH Aachen. Apprimus Verlag, Aachen, 2008. [10] Diestel R. Graph Theory. 3rd Edition, Springer, Heidelberg, 2006. [11] Harary F, Norman R, Cartwright, D. Structural Models: An Introduction to the Theory of Directed Graphs. John Wiley & Sons, New York, 1965 [12] Harary F. Graph Theory. Addison-Wesley, Reading, Massachusets, 1969. [13] Bjørke Ø. Manufacturing Systems Theory – A Geometric Approach to Connection, Tapir, 1995. [14] Pugh S. Concept Selection – A method that works. Proceedings of the International Conference on Engineering Design ICED. Rome, 1981. [15] Brecher C. (Ed.). Integrative Production Technology for High-Wage Countries. Springer, Berlin, 2012. [16] www.production-research.de, last visit 1st Aug 2015