Design methods applied to the selection of a rapid prototyping resource

Transcription

1 Design methods applied to the selection of a rapid prototyping resource François Pérès CREATE : Rapid Prototyping Centre Laboratoire Productique Logistique Ecole Centrale de Paris, Grande Voie des Vignes Châtenay-Malabry cedex FRANCE Telephone: 13- Fax: peres@pl.ecp.fr Carmen Martin ADEPA : Agence de la Productique Département Direction Veille et Innovation 17, rue Périer 10 Montrouge FRANCE Telephone: Fax: carmen@vigie.adepa.asso.fr

2 Design methods applied to the selection of a rapid prototyping resource François Pérès CREATE : Rapid Prototyping Center Laboratoire Productique Logistique Ecole Centrale de Paris, Grande Voie des Vignes Châtenay-Malabry cedex FRANCE Telephone: 13- Fax: peres@pl.ecp.fr Carmen Martin ADEPA : Agence de la Productique Département Direction Veille et Innovation 17, rue Périer 10 Montrouge FRANCE Telephone: Fax: carmen@pl.ecp.fr KEYWORDS QFD, Concurrent Engineering, Optimisation, Design, Rapid Prototyping, Quality Management ABSTRACT The paper is divided into two main sections. We first introduce the problematic and discuss about the interest of implementing a concurrent engineering approach. We focus on the relevant rapid prototyping techniques likely to be used in order to make the design and production processes more reliable. In the second chapter we describe in generic terms the advantage of using QFD method as a decision making tool. The practical case study chosen is related to the selection of a prototyping resource. After having identified the functions to be fulfilled by the rapid prototyping resource, a technology deployment matrix is built and the principles leading to the results are commented. TOPICS Automated Manufacturing Systems Development Methodology SUBJECT AREAS Rapid Prototyping

3 1 INTRODUCTION Today in industry, the need to adapt the products to the market becomes vital. This adaptation will require to be able to satisfy contradictory parameters which are reduced time to market for a less expensive product but with a better quality. To achieve this almost unreachable goal, companies must review the way they organised the conception but also the production or industrialisation departments. Furthermore, these changes must take place in a context where the market is often unstable and the investment capacity is most of the time reduced to its maximum. The fast evolution of techniques, needs or even fashions implies to always wonder : does the product or service I propose fit the market expectations? and do I make it properly?. These answers depend essentially on technical and temporal criteria, both of them being translated into an economical point of view by the company. Time so appears as a primordial factor. Consequently, creativity must be followed by an ability to react with a time of response sufficiently short to be one of the winning argument when proposing the product. Time to market is a key element in the companies success. The conventional development chain of a product requires the realisation of sequential activities, linked together in a logical preestablished order. This serial way tends to disappear and is progressively substituted in many industries by a concurrent engineering approach which consists in placing the tasks in a parallel configuration to compare very early in the product development cycle, the objectives and expectations of the whole technical staff who contributes to the product making. The unique goal of this approach is, as soon as possible, to validate the parameters intervening in the definition, realisation and exploitation of the product. This validation will be as much easier than we can refer to physical models of the product and tooling required to its realisation. Rapid prototyping techniques allow to elaborate these physical models in a very short time. Validations made from prototypes implying the whole project team enable to quickly visualise the errors and limit

4 their consequences (time or part losses for example). Therefore, rapid prototyping is helpful in order to validate: The product: Conceptual validation Esthetical, geometrical or dimensional validation Functional validation Mechanical, thermal or any physical solicitation validation Finished machining validation The realisation mode: Tooling validation Process validation Fabrication cadence validation This new design tendencies are strongly inspired by the notion of accuracy in both physical and organisational terms. Not only the designer must make the good parts but also has to select the right way to make it. What is true for the final product is also valid for the prototype and tooling realisation. In order to achieve this goal, the designer must rely on a tool which will lead him to make the good choices allowing to deliver the prototype with the quality expected by people concerned with the product life cycle. One of the tools likely to be used in such a case is the QFD method. RAPID PROTOTYPING So far, a large amount of research has been done about prototype making. Due to these investigations, various techniques and machines have been developed (Merz ), (Nakagawa ). If research is still made on prototyping methods, more and more studies deal with rapid tooling which is an extension of prototyping concepts. In fact, classical prototyping techniques are usually not sufficient to deliver a model identical to the final product which only allows to validate the whole characteristics required in its making and use. The objective is then to make tooling which will help to realise

5 prototypes in a way that will enable to validate not only the industrialised product but also its elaboration mode. More and more techniques use the speed of prototype making to accelerate the tooling realisation. Prototype must then be high quality products to fulfil the constraints demanded by the tooling environment. This quality requirement can be achieved through the use of the proper techniques and by carrying out the process according to determined rules..1 Main Rapid Prototyping techniques Stereolithography: Under action of a laser, a photosensitive liquid resin is solidified by a chemical transformation. The beam of light emitted by the laser is conducted on the resin surface by a set of dynamic mirrors. The move of these mirrors, driven by CAD software will make the light runs on the resin surface along the trajectory corresponding to the considered section. In its wake, the laser polymerizes the resin and only the exact section remains solidified. After realization of a section, the platform supporting the object being made goes down into the resin at a depth dependent on layer thickness (usually from 0.07 mm to 0.7 mm). Stacking-up of layers leads to a 3 dimensional part. Solid Ground Curing: is an other method based on resin polymerization. Unlike stereolithography, the process do not use a laser but a powerful ultra violet lamp coupled with a masking device. With this method, all points of a same section are simultaneously solidified. The mask, which is a negative of the section, is a glass plate covered on appropriate zones with a black electrostatic toner so as photocopier machines (ionography). Selective Laser Sintering: requires a thermal laser used to fuse a powder material mixed with a binder. On laser wake, powder is heated slightly above its melting point and agglomerate when cooling. A thermal treatment can be useful to improve physical properties of the part and reduce porosity. Prototype materials used depend on the powder, usually made from plastic, sand, ceramic or metal components.

6 3D Printing: is very similar to the previous technique. Its principle is based on powder agglomeration through deposition of binding droplets on the section points. When finished, the part is larger than it should be in order to compensate for the important shrinkage due to sintering. Here again, a thermal treatment will be required. 3D printing is sometimes used in ceramic molds making (Cf. 1.). Fused Deposition Modeling: which has been developed in 1 uses a head mounted on a 3 axes CNC machine to depose a fused thread on the part being made. Solidification is instantaneous when bringing the thread into contact with the previous section. Thread materials used are wax, nylon, polypropylene, ABS, This process is quite fast and cheap and can be helpful to make empty parts. Laminated Object Manufacturing: does not use material change of state. Sheets are cut out, piled up and stuck. Cutting out can, according to the techniques, be the last operation. A thin sheet of paper covered by a polypropylene film is deposed on the previous section and pressed under temperature. Heat makes the film melt which sticks the paper sheet. A laser cuts out the outline of the considered section at a depth corresponding to the thickness of the sheet. The final part is very similar to wood. Stratoconception : is a process very simple to implement. Actually, it does not need specific machines. Various sections are cut out in a plate through techniques such as water cutting, laser cutting or even milling. Assembly is realized by sticking, fusion or locating. In order to save up materials, an optimization is done to select the most adapted cutting plan. Almost all solid materials can be used in stratoconception. Moreover, as conventional machines are sufficient, large 3 Dimensional prototypes can be made. The techniques presented so far are all based on matter adding. They are the most common methods used to realize, in a brief period, complex prototypes. Let us just mention here that an other way to make prototypes consists in removing matter. Techniques such as high speed machining or hot thread cutting are also widely used in industry.

7 . Application to Rapid Tooling Up to now, the techniques we have exposed are most of the time unable to make prototypes in the same material than will be the final product. This is not a problem when controlling characteristics such as geometry, volume, aspect or parameters like aptitude to be machined or accessibility and modularity for maintenance but very often prototyping requires to be able to validate other properties of the product. Most of the time, features like mechanical or thermal resistance, ability to be integrated in a given context and to be assembled or faculty to run properly in a nominal mode of use must be checked before industrialization and serial tooling realization. Rapid tooling techniques allow to make prototypes that will make easier those validations i.e. in a material identical to the final product or very similar on the ground of the characteristics to be controlled. Furthermore, rapid tooling is very useful when several parts must be tested to carry through the validation process (when using destructive control for instance). Most of rapid tooling techniques enable to realize a number of products superior to unity. Finally, beyond the product validation, the serial realization mode requires to be controlled to verify its capacity to deliver conform products in term of quality and cadence as prescribed in the specifications. Some rapid tooling methods make possible this form of validation. More precision will be found in (Pérès ). 3 QFD The purpose of this paper is not to remind the establishment rules of the quality matrices which are the basis of the QFD method. We advice the reader to refer to (Lin & al. ), (Mallon & al. 3), (Jacobs & al ) to find more precision on the topic. The benefit of using QFD in this field is clearly linked to its capacity to deliver parameters guiding the decision making. The abundance of processes we have described on the previous section requires when using or even better when deciding to purchase resources, to think about its adequation with the real needs of the company in the field of rapid prototyping. Taking the decision is not easy and the QFD methodology we propose here, aims at simplify somewhat the problem by proposing a way to quantify the appreciation level of the different processes respect to the required functions.

8 According to the product typologies and the class of validation, functions delivered by the rapid prototyping process are as many as diversified. A short list of the main functions characterising the real expectations in rapid prototyping terms is coming next. According to the context, the rapid prototyping machine will be used to : F 1 : Improve reactivity with respect to market fluctuations F : Help creativity and development of new concepts in design F 3 : Simplify communication between customer and supplier F : Optimise the integration of the product in its environment F : Anticipate fabrication, maintenance, control or conditioning problems F : Make rapid tests on the product to verify some of its characteristics (ergonomic, aerodynamic, ) F 7 : Form a part which will be used as a pattern in order to make a tooling in a very short time F : Manufacture directly a mold for plastic injection or metal casting Respect to these functions likely to be delivered by the process, we propose the deployment of consistent technologies able to satisfy the requirements. The processes we will consider here are those we have been talking about so far. Namely : P 1 : Stereolithography (STE) P : Laminated Object Manufacturing (LOM) P 3 : Laser Sintering (LS) P : 3 Dimensional Printing (3DP) P : Solid Ground Curing (SGC) P : Fused Deposition Modelling (FDM) P 7 : Stratoconception (STR) P : High-Speed Machining (HSM) A first matrix (Cf. Table 1) called technology deployment matrix is used to connect the different processes P i and the main functions F j they allow to satisfy. Consequently, the dimension of the matrix will be i.j. The cells that make correspond P and F are filled with a mark X ij revealing the aptitude of a process i to deliver the correspondent function j. Marks can be, 3, 1 or 0 according to the strength of the relation respectively large, medium, weak or null.

9 Let us precise immediately that the matrix is wanted to be realistic. Although it has been built on our large experience of rapid prototyping techniques it is nevertheless the result of subjective considerations proper to the principles of QFD notation. P1 P P3 P P P P7 P STE LOM LS 3DP SGS FDM STR HSM Σ F1 Reactivity ,00 F Creativity ,00 F3 Communication ,00 F Integration ,00 F Anticipation ,00 F Testing ,00 F7 Modeling ,00 F Tooling ,00 Σ 3,00 30,00 7,00,00 3,00 7,00 0,00,00 X,0 3,7 3,3 3,0,7 3,3,0 3,0 σ 3,7 3,7 3,3,7 3,3,3 0,7,7 Table 1 : Technology deployment matrix At this level, the signature of each process reveals its capacity to fit the different functions as we can see on Figure 1. STE 1 LOM 1 LS 1 3DP SGC STR FDM HSM Figure 1 : Technology signatures The global performance of a single process is obtained by adding all the values of the corresponding column in Table 1. Performance repartition on the different functions reveals the process polyvalence which can be expressed by the calculation of the mean and the variance of dispersion on the various functions. The horizontal summation characterises the potential of response to the considered function by the integrity of rapid prototyping techniques. However, acquisition of a rapid prototyping process is often motivated by a polymorph and plurifunctional need. Difficulty then lies in the research

10 of a compromise giving an optimal satisfaction level on the desired functions. Investigation carried out to find the best process can be helped by using the QFD matrix represented on Table. P M P7 M S P W M S P M M M S P M W W M M P3 W M N W M S P N W M N W M S P1 P1 P P3 P P P P7 P K STE LOM LS 3DP SGC FDM STR HSM Σ F1 Reactivity 0,37 0,37 1,10 0,37 3,30 1,10 1,10 0,37 3,30 11,00 F Creativity 0,7 0,7,00 0,7,00 0,7,00,00 0,00 1,00 F3 Communication 0, 1,33 1,33 1,33,00 1,33 1,33 1,33 0,00 1,00 F Integration 0,0,71 0,00 0,00 0,00,71,71,71,1 1,00 F Anticipation 0,3 3,3 3,3 3,3 0,00 3,3 1,1 1,1 0,00 1,00 F Testing 0,3,7 1, 0,3 0,3,7,7 1, 0,3 1,00 F7 Modeling 0,1 1,3 1,3 0, 0,00 1,3 0, 0,1 0,00,00 F Tooling 0,17 0,17 0,00 1,7 0,00 0,17 0,00 0, 1,7,00 Σ 1,7 10,77,33 13,3 1, 13,,1 13, X 1, 1,3 1,0 1,73 1, 1, 1,3 1, σ 1, 1,03 1,0, 1, 1,0 0,,7 Table : QFD of RP technologies Entries of the matrix are the same as previously but a column is added and filled with a coefficient K representing the relative importance I of each function related to the relative weight of the process on the considered function: K = i I j X ij. The new matrix values are pondered with these coefficients: X = K. X i, j. IJ j ij In the example of Table, the selected values representing the relative functions importance are all deliberately non valueless. In this case, the user would run the rapid prototyping resource he owns to satisfy the main functionalities we have already presented. Most often, the first ponderation is the result of a value analysis carried out on the branches of a tree released by a functional decomposition. The roof of the quality house inform on the compatibility between technologies. According on the strength of the compatibility which can be strong, medium, weak or null, the cell positioned at the intersection of the two compared processes is filled with the letter S, M, W or N. It is obvious that incompatibilities appear but the diversity of processes allow various combinations. It seems rather

11 logical since overlapping between functions satisfied by each technology are not many as we can notice on Figure where 7 of the elementary surfaces relative to the efficiency zones of the different technologies are visible. Tool i ng ng Testing Reactivity ST E Creativity SGC HSM LS Communi cation LOM FDM ST R Integration 3DP Anticipation Figure : Signature of functions fulfilled by RP Taking into account the whole techniques, the potentiality of a function to be satisfied differ according to the considered function. Figure 3 shows for instance that rapid prototyping contributes indeed to amplify the function of "creativity" but in a smaller proportion (0%) than interest brought by rapid prototyping techniques on the function "anticipation". The same observation can be made on the function "integration" for the sensitive nature of used materials contrasts with the characteristics required to integrate the prototype in technical environment mechanically solicited. We will not explore further this track since the reflection is oriented around the methodology of technological comparison and not around the technologies themselves. Rapid prototyping function integration 10% 13% 1% % 1% 1% % 17% Reactivity Creativity Communication Integration Anticipation Testing Modeling Tooling Figure 3 : Capability to satisfy function The aptitude to satisfy functions by the totality of the techniques is a relevant parameter in the construction of the QFD matrix. It is used in fact as a basis to compare the different technologies. By

12 relating each elementary potential of a technique to the entire potentiality affected to rapid prototyping technologies, a scale is established which allow to compare the interest level of the technology on the considered function. CONCLUSION The purpose of this paper was to focus on the benefit of using the QFD method to choose a rapid prototyping technology. After having introduced the technologies, the approach implemented consisted first of all in identify the main functions corresponding to the most frequent services required from a rapid prototyping mean. These functions were deployed on technologies and quantified according to the response brought by each individual technology on the considered function. The results can be used directly for a single function use. Then, the expression of requirement was taken into account by the way of a matrix making appear a double ponderation. The calculation of the corresponding coefficients allow to simultaneously ponder importance of one technology respect to the other on a given function and the value of a function respect to the other on a determined technology. Methodology remaining unchanged, user will be able to manipulate this kind of matrix to select the most suitable rapid prototyping resource to the need he will have expressed. BIBLIOGRAPHY (Lin ) "Intelligent quality function deployment system in concurrent engineering environment", Z.-H. Lin, A. Che, Xi'an Jiatolong Univ. China, Intelligent Systems in Design and Manufacturing, Proceedings of Spie International Symposium, Boston, Massachusetts, - November, 1 (Mallon 3) "QFD : a system for meeting customers'needs", J.-C. Mallon, E. Mulligan, Journal of Construction Engineering & Management, 11, 3, pp.1-31, 13 (Jacobs ) "Improving communication and Decision-Making within Quality Function Deployment", S. Jacobs, S. Kethers, Proceedings of Concurrent Engineering : Research and Application, Pennsylvania, August 1,

13 (Merz ) "Shape deposition manufacturing", R. Merz & al., Proceedings of the solid freeform fabrication symposium, University of Texas, Austin, p 1-, 1 (Nakagawa ) "Laser stereolithography and its applications to forming industries", T. Nakagawa, J. Wei, Institute of Industrial Science, University of Tokyo, CMET inc., 1 (Pérès ) "Rapid tooling from prototypes: the cool spray moldmaking method", F. Pérès, A. Mofakhami, Proceedings of Spie International Symposium, Boston, Massachusetts, - November, 1 (Cabrera ) Fabrication, outillage et prototypes rapides : le choix de la technologie appropriée, M. Cabrera, M. Shellabear, èmes assises européennes du prototypage rapide, Paris, 1 BIOGRAPHY François PERES is a production Engineer. Following a Ph.D. on Maintenance Management in Bordeaux I University is now at Ecole Centrale Paris (ECP). Manager of a prototyping center (CREATE), head of the production and maintenance at ECP, its main research concerns the concurrent engineering aspects on both management and fabrication levels. Carmen MARTIN is an industrial engineer graduated from the CPS (Centro Politecnico Superior de Zaragoza). She is at the moment working as a design development manager in the ADEPA company in Paris combining this activity with a Ph.D. in the Productic Logistic Laboratory of Ecole Centrale de Paris.