Intelligent Interfacing Module of Process Capability among Product and Process Development Systems in Virtual Environment Algirdas Bargelis - Petri Kuosmanen 2 - Andrius Stasiskis 3 1 University of Technology, Department of Manufacturing Technologies, Lithuania 2 Helsinki University of Technology, Department of Machine Design, Finland 3 Kaunas University of Technology, Panevezys Institute, Lithuania An intelligent interfacing module of process capability (IIMPC) between product and process design systems in virtual prototyping environment has been developed on the basis of the knowledge acquired at different organizations involved in new product development. The paper considers contradictions both, in product design procedure when seeking its best performance and in the principles of design for assembling (DFA) and design for manufacturability (DFM), whereas, when facilitating the product assembling process, the fabrication process of product parts becomes more complicated. This research can help to find the best decision of quality and lean manufacturing among available product and process alternatives. Mathematical formalization of a developed interfacing module is provided and appropriate software is created. The proposed interfacing module is being implemented for the integration of computer aided design (CAD) and computer aided process planning (CAPP) systems. IIMPC is used in industry and in study processes in universities. © 2009 Journal of Mechanical Engineering. All rights reserved. Keywords: intelligent interfacing module, product and process development, integration, process capability, quality, virtual environment 0 INTRODUCTION Manufacturing competitiveness of the 21st century is associated with computerization in the development of new products and processes and in employment of relevant information [1]. These procedures could not have been possible without integrating project management methods with concurrent engineering (CE) elements [2]. CE is oriented towards possibilities to minimize both, product and process development cost and delivery time in all stages of a product life cycle. A basic portion of the product development cycle is a conceptual design phase that greatly influences product's cost, quality, manufacturability and life cycle parameters [3] and [4]. During the product concept design phase over 5 to 10 versions are to be generated for the best solution of each product or its component. This generation concerns the design of product and process. The best solution means the lowest cost of product design and manufacturing maintaining required performance [5]. The inter-enterprise integration, when enterprises can co-operate together to develop, design, produce, and to distribute their common product, enables engineers to use the virtual prototyping environment more effectively. Engineering in virtual environment helps to save costs and time of the product and process development. The key point of engineering in virtual environment is virtual prototyping with 3D CAD systems for a new product design [6], and appropriate software for the production system design [7]. Virtual prototyping can carry out all the main functions of the new product development according to individual customer requirements in the virtual environment. The use of virtual prototyping in an organization has refocused product development philosophy incorporating the notion that products must demonstrate their value-added market capability prior to the approval of spending significant resources on product development and production. Virtual prototyping generates early product characteristics that can be compared with customer requirements and manufacturing capabilities. New intelligent support systems are necessary for a successful solution of the above-mentioned tasks. *Corr. Author's Address: Kaunas University of Technology, K^stucio 27, LT-44312 Kaunas, Lithuania, algirdas.bargelis@ktu.lt The research presented in this paper is intended for the development of a new intelligent support tool for making the best decision among available product and process alternatives. It considers how knowledge engineering of product and process development can help in creating the optimum of a production process. In this context, a developed intelligent interfacing module for product and process design raises the level of the enterprise integration, seeking minimization of new products delivery time to the market when its requirements have changed. This research is focused on the capability of various processes and suppliers located in different countries and companies to combine the product design, i.e. the number of original and standard parts, components as well as their manufacturing. An intelligent interfacing module of process capability (IIMPC) for product and process design in the virtual environment has been created on the basis of the knowledge acquired at different organizations involved in new product development. IIMPC considers contradictions both, in the product design procedure seeking the best performance and in the principles of design for assembling (DFA) and design for manufacturability (DFM). When facilitating the product assembling process, the fabrication process of product parts and components becomes more complicated and the problems related to the fabrication process capability can arise. Mathematical formalization of a developed interfacing module is provided and the appropriate software is created. The proposed IIMPC module is being implemented in the integration of computer aided design (CAD) and computer aided process planning (CAPP) systems. Product design • Geometry • Configuration • Materials • Parameters • Functions 1 DFM 1 PROBLEM DESCRIPTION The problem considered in this paper can be formulated as follows. A designer using CAD can provide geometric modeling of a new product. There are some additional programming tools as FEM (finite element modeling), BOM (bill of materials), DFA, DFM, etc. coming in assistance to achieve the desired accuracy, performance, functionality and productivity of a product within the budget limits of its development. A production engineer using a CAPP system has to transform the designed parameters and characteristics of a product into a suitable process. CAPP is closely related to an appropriate software such as material resources planning (MRP), enterprise resources planning (ERP), group technology for operations and process development, etc., estimating the process costs. CAD and CAPP systems are developed to operate autonomously. Various external interfaces as the connection for hooking CAD to CAPP systems are used [8]. However, these interfaces can transfer the data (geometric form, dimensions, tolerances and specification) from one system to the other seeking the integrity of both systems only. Unfortunately, they can neither evaluate possible alternatives of a product and process nor upgrade them. A developed interfacing module can test and evaluate each product and process alternative to process capability when high dimension accuracy of parts and low production cost are needed. Process capability is strongly related to product quality and costs. When it is insufficient, then an IIMPC module can suggest generating a new process with sufficient capability for each product and process alternative (Figure 1) and with minimal production costs. Interfacing between CAD and CAPP systems using IIMPC is done in a virtual environment. 1 GT DFA IIMPC FEM MRP Process design • Process plan • Operations • Machines • Tools • Costs Fig. 1. The framework of a product design for process capability The module architecture is presented in Figure 2. The starting point of an inference strategy is to place an overall functional requirement F into sub functions F1, F2 and F3. Sub function F1 is intended for the definition of material consumption. It formulates a work piece of the part as parameter P11 and material cost as parameter P12. Sub function F2 is intended for the analysis of part geometrical form, accuracy and tolerances. It examines the design feature types - P21 , quantitative - qualitative parameters - P22 and part overall dimensions - P23. Sub function F3 considers the manufacturing process of a part as quality cost - P31, manufacturing cost - P32 and machine tool and process capability indices - P33. The interrelationships among various design and process characteristics are elaborated and emphasized solving contradictions between process cost and capability. This consideration requires generating some alternatives of product and process seeking the best solution of product functionality, manufacturing cost and process capability. In modern manufacturing it is important to extend the virtual prototyping principles from a conceptual product design phase to all the other phases of a product life cycle. The major phases of the product life cycle are: conceptual and preliminary design, detailed design and integration, production and use, retirement and disposal [9]. Manufacturability of a new product in this context plays a very important role. Modeling of product manufacturability and deliverability during the preliminary stages of design is critical in achieving the reduced time to market, high quality and low cost [4] and [10]. The integration of the product-process design in the development of a production system is emphasized in research [11] which had developed a virtual model for the production system design based on technical and temporal data such as work sequences, operations, components and products delivery time, and production resources. Assuming the existing advantages and drawbacks of the product and process development in virtual environment, the IIMPC module could help minimize the product and process development time and cost. 2 IIMPC DEVELOPMENT New product design is a creative effort attempting to turn customer wishes into an economically producible product to be useful all over its life. In most design situations, compromises between product performance, cost, quality and delivery time cannot be avoided. Input data being different, variation enters into the product design. Production processes do not always make perfect products and, eventually, they introduce more variation and product defects. The capability of a process refers to its ability to meet the implementation needs of a product. Capability is not inherent to a process, but rather it depends on the designer's expectations [12]. In most cases, product implementation costs are directly related to process capability. Making the best choice of the available product and process alternatives is usually finding the trade-offs in each product life cycle stage between the product development cost, investment cost, and quality variables that are based on appropriate mathematical tools. The calculation of the product life cycle costs is a complex process [9], while production cost and process capability are just two of many factors. The research presented in this paper is devoted to a consideration of process capability aiming at minimization of both, product and process development costs and product delivery to customer time. Fig. 2. IIMPC module architecture Process capability is measured by its indices. A process capability index is a measure relating the actual performance of a process to its specified performance that depends on the traditions of enterprise and environment, peculiarities of equipment, operation, materials and people. The most popular process capability indices are Cp and Cpk [12]. Machine tool capability Cp and process capability Cpk are used to determine the work efficiency [13]. Cp is applied to determine the system's location within the tolerance limits. The size of deviations from the mean value of process dimensions will indicate how good the production is. If the system is not at the center of specification values, the trend of Cp is progressing faultily. Cpk is used to determine the average so that the system will work better within the specification limits. If the system is centralized on the target value, Cp and Cpk values will be equal. When the value of Cp and Cpk is 1, it is considered as a minimum requirement of the system for some companies. Alongside this, many companies accept greater Cp and Cpk values, for instance 2. Cp and Cpk are (0 < S < S„ S ^ min, pk defined by the following equations [12] cp = USL - LSL 6a and Cpk = min| . ( USL - X X - LSL 3a 3a (1) (2) where USL is the upper specification limit of a part, mm; LSL is the lower specification limit of a part, mm; a is the process standard deviation or overall process variability, mm; X is the mean value of the whole process parameter, mm. Process R of product P is expressed as a set of operations O R = U 0 = {0„ O2,..., ° }. (3) The value of process capability indices is calculated for each operation O, and hereby a lot of Cp for the entire process R could be expressed as follows CP =Cp (IX CP C2),.. Cp (i)}. (4) A critical operation in the set (4) is the minimum value of Cp index. On the other hand, the value of process capability indices with process costs is related to [C7 < Cp < C^, Cp ^ max' where Smax is the highest acceptable costs of an operation, Euro; C^11 = 1 and C^ = 2 are the minimal and maximal values, respectively, of the acceptable capability indices seeking the minimal process costs. In piece and serial production often Cp = 1 + 1.33, because the companies apply a cost-of-poor-quality strategy that attempts to bring costs to everyone's attention as a basis for corrective action. In mass and high-run production it is accustomed to have Cp = 2, because investments to quality costs pay for production of big volumes of parts. The purpose is to minimize the manufacturing cost reducing the quality control cost. A parametric function is developed for determining quality control operation percentage to process capability index Cp (Figure 3). The parametric function is developed in accordance to the assumption when Cp = 1 or less, then it is necessary to arrange 100% quality control of parts, and when Cp = 2 or greater - 5% quality control of parts is sufficient. The percentage value of parts quality control Y is defined as follows: Y = 100 • C (6) Fig. 3. Parametric function for definition of quality control operation percentage to Cp The work is accomplished after the systematized theoretical and experimental research based on the methods of mathematical logic, the theory of sets and the theory of chances [14]. During the concurrent product and process 4.35 or development, DFA and DFM approaches [15] are aiming at reducing process manufacturing costs S, i.e. at achieving Smin. Unfortunately, when both methods DFA and DFM are used, they frequently cause conflict situations resulting in insufficient capability of a product manufacturing process, because when simplifying the assembling process, a designer reduces the number of product parts inducing the other parts to become more complicated. The solution of these conflict situations and search of the best version require engineers to generate a vital number of product and process alternatives checking their Cp and S. Manufacturing costs S of product Pj without material cost, set up time and overheads by production time consumption L are expressed as Lm = Vp ■ C. S = ej |X 4+X LkjDj \ (7) where A¡ is the assembling operation time of product Pj, h; ej is the production volume of product Pj; Lkj is the predicted manufacturing time of product part k, h; D is the number of parts k in product Pj, is the number of assembling operations; r is the number of different parts in product Pj. Product assembling operation time can be expressed as an abstraction function A = fi(Lkj, P, ^ q, e), (8) where P is the product type; r is the number of assembling parts; q is the product qualitative parameters; e is the high-run, serial or piece production. L = Lh + Lm , (9) where Lh is the handling time in h of an operation, which is conditionally constant and depends on an operation, machine tool type, material profile and part dimensions; Lm is the machining time in h of an operation. According to the research [5] and [10], where V is the removed material volume in mm from a work piece during part machining operation; p is the slope of a regression trend line; C is the intercept of a regression trend line. The effect of material, manufacturing accuracy and production volume is estimated by correction coefficients [5] and [10]. IIMPC is constructed in the product and process development domain. The best available practice, experience and traditions of a simultaneous product and process design in different countries and companies have been used for this purpose. The knowledge was acquired and research was done on the integrated knowledge-based inter-discipline study program on the web site of geographically dispersed organizations [16]. Another approach of the process quality estimation by applying control charts has been used in research [17], regrettably, it evaluates neither the manufacturing cost nor quality cost. The framework of consideration dependence among process capability indices, standard deviation a and manufacturing cost S as well as the product characteristics as class, part type, and design features with their quantitative -qualitative parameters has been developed (Figure 4). Statistical standard deviation a of the data sets collected from factory processes was the starting point for predicting the process capability indices [18]. The considered parts might be minimum either of two different companies or of two different machine tools. The IIMPC module has been developed on the software level and appropriate database (DB). The first version of software has been programmed applying Visual Basic 6.0 programming language and Structural Query Language (SQL). Fig. 4. The structure of a frame for possessed a and S in different manufacturing systems The developed software generates available process alternatives of a part with manufacturing cost S and capability indices. The software window for data input is presented in Figure 5; the results obtained when applying the developed software are illustrated in a case study. The developed IIMPC module has been tested and validated in the laboratory of Integrated Manufacturing Engineering of Kaunas University of Technology (KTU) and in the department of Machine Design of Helsinki University of Technology (TKK). During the development procedure the IIMPC module was verified by a number of process plan alternatives with different Cp and S for various products and components. 3 CASE STUDY A sequence of IIMPC module work has shown that it is aiming at the optimal Cp index and manufacturing cost S of a process in the early product design stage. A typical mechanical part -gear pump housing with seven various design features has been taken as a sample. The work piece of housing is made from cast iron. It was machined using various chip removing operations such as milling, turning, drilling and grinding. These operations for high-run and serial production have been investigated in two different medium-size manufacturing systems. The manufacturing system "A" runs its business on order handled piece and serial production with various NC and CNC machine tools, while the manufacturing system "B" has some specialpurpose machines and production lines, and CNC machine tools. The manufacturing system "A" can survive without its own design department; however, it always experiences troubles related to the manufacturing processes when seeking less costs. The manufacturing system "B" develops new products itself, it can implement DFA and DFM approaches to a new product development procedure. Sometimes both companies co-operate in the production of mechanical parts and receiving mutual benefit. Firstly, a 3D CAD model of a chosen mechanical part was created using the standard CAD system. The second step of IIMPC module work is the housing data extraction from the 3D CAD model. The extraction of the data is performed at the interactive regime applying the developed software data input window (Figure 5). The results of the second step module work are the process plan prediction of a housing and definition of statistical process standard deviation aa for each design feature, and then calculation of capability index Cp. The results of a second step module work are presented in Table 1 for both companies. Process standard deviation is obtained from the machine tool capability Fig. 5. Software data input window study using the statistical data of previously produced parts and design features according to the methodology described in research [18]. The Cp index was defined applying the expression (1), upper and lower specification limits of each design feature and the statistical process standard deviationca value. The third step of an IIMPC module work is defining the material resources. The part manufacturing cost S and quality cost as the results of software third step processing are shown in Tables 2 and 3. The results have been analyzed by a designer and collected into the data base if they are suitable; conversely, the changes of the product or process design are to be made. Software programming mistakes are found and removed during the procedure of software development in the KTU laboratory of Integrated Manufacturing Engineering, while software validation in two Lithuanian manufacturing companies of production mechanical components has been performed. The manufacturing cost S of each design feature and operation applying expressions (7), (9) and (10) and control percentage by applying expression (6) are predicted and presented in Tables 2 and 3. The manufacturing system "B" has higher Cp values and lower total manufacturing costs S compared to the manufacturing system "A" because of better tooling and quality management. The implementation of special-purpose machines, production lines and multiple drilling devices as well as investments to quality assurance helps in reaching the target values which are important to both, customers and producers. The data in Tables 1, 2 and 3, shows that increasing Cp value by one hundredth, it is possible to decrease the total manufacturing cost S by one percent. 4 CONCLUSIONS AND FURTHER WORK Growing complexity of new products and stiff competition in marketplaces enhance the demand to minimize the product and process development costs and delivery time in all stages of a product life cycle. A proposed intelligent IIMPC module for the product and process design will raise the level of activity integration in the organization and will reduce the risk of implementing new processes and operations. It is shown that the analysis of capability and manufacturing cost helps determine the possibilities of manufacturing within the tolerance limits and engineering specifications. Capability and manufacturing cost analysis yields the information on the changes and tendencies of the system during production. The improved intelligent support for modeling the concepts in the virtual environment of the manufacturing domain has been emphasized. Knowledge engineering is based on the research done on the integrated knowledge-based inter-discipline study program for geographically dispersed organizations. It has been shown that proper research can eliminate the shortage of appropriate engineering knowledge and experience. The appropriate software has been programmed for the confirmation of theoretical consumptions. Regrettably, the suggested approach has some limitations, the main one being a relatively Table 1. Statistical standard deviation aa of serial and high-run production DF Operation Tolerance (mm) Serial - MS "A" High-run - MS "B" 1 Milling 0.200 0.0162 2.058 0.0152 2.193 2 Milling 0.120 0.0136 1.471 0.0125 1.600 Grinding 0.020 0.0025 1.333 0.0022 1.515 3 Milling 0.220 0.0222 1.652 0.0183 2.004 4 Turning 0.080 0.0094 1.418 0.0089 1.498 Grinding 0.025 0.0031 1.344 0.0029 1.437 Precise grinding 0.011 0.0014 1.309 0.0013 1.410 5 Drilling 0.110 0.0092 1.993 0.0081 2.263 6 Drilling 0.090 0.0079 1.899 0.0075 2.000 Tapping - - - - - 7 Countersink 0.210 0.0206 1.699 0.0157 2.229 Tapping - - - - - Table 2. Manufacturing and quality control costs for serial production (Manufacturing system "A") DF Operation Serial production (quantity 30) Lh (h) V (mm3) Lm (h) S (h) Control percentage (%) Control time ( h) Total costs (h) 1 Milling 0.083 3.1-104 0.133 0.216 4.3 0.0013 0.2173 2 Milling 0.083 7.5-104 0.203 0.786 18.7 0.0056 0.2916 Grinding 0.083 3.9103 0.21 0.293 28.6 0.0086 0.3016 3 Milling 0.083 6.9-103 0.083 0.166 11.3 0.0034 0.1694 4 Turning 0.102 2.3-103 0.067 0.169 21.9 0.0110 0.1800 Grinding 0.102 2.01-102 0.121 0.223 27.6 0.0138 0.2368 Precise grinding - 4-101 0.082 0.082 31.0 0.0155 0.0975 5 Drilling 0.07 4.56-103 0.103 0.173 5.0 0.0008 0.1738 6 Drilling 0.07 5.6-103 0.196 0.266 6.1 0.0010 0.2670 Tapping - - - - - - - 7 Countersink 0.075 1.96103 0.067 0.142 10 0.0025 0.1445 Tapping - - - - - - - Z 2.516 0.0635 2.5795 Table 3. Manufacturing and quality control costs for serial production (Manufacturing system "B") Serial production (quantity 30) DF Operation Lh (h) V (mm3) Lm (h) S (h) Control percentage, (%) Control time (h) Total costs (h) 1 Milling 0.075 3.1104 0.124 0.199 3.3 0.0010 0.2000 2 Milling 0.075 7.5-104 0.105 0.180 12.9 0.0039 0.1839 Grinding 0.075 3.9-103 0.165 0.240 16.4 0.0069 0.2469 3 Milling 0.075 6.9-103 0.071 0.146 4.9 0.0015 0.1475 Turning 0.092 2.3-103 0.060 0.152 17.2 0.0086 0.1606 4 Grinding 0.092 2.01102 0.116 0.208 20.7 0.0104 0.2184 Precise grinding - 4-101 0.080 0.080 22.4 0.0112 0.0912 5 Drilling 0.063 4.56-103 0.045 0.108 2.9 0.0005 0.1085 6 Drilling 0.063 5.6103 0.062 0.125 4.9 0.0008 0.1258 Tapping - - - - - - - 7 Countersink 0.067 1.96103 0.057 0.124 3.1 0.0008 0.1248 Tapping - - - - - - - Z 1.562 0.0456 1.6076 narrow area of manufacturing systems, products and processes to be applied to. Future work will focus on the expansion of the variety of data and features in the developed module, the number of product types, processes, operations, and in particular, at aiming to overcome the existing limitations of the proposed approach. 5 ACKNOWLEDGEMENTS The research was supported by EC Leonardo da Vinci Project No LT/02/B/F/PP- 137022 "Integrated Knowledge-based InterDiscipline Study Program on the Web Site". It has been conducted at the Mechanical Engineering Faculties of Kaunas University of Technology (Lithuania) and Helsinki University of Technology (Finland). 6 REFERENCES [1] Semolič, B. Šostar, A. (2007) Network organizations - a new paradigm of the 21st century. Strojniški vestnik - Journal of Mechanical Engineering, vol. 53, no. 3, p. 193-211. [2] Kušar, J., Bradeško, L. Duhovnik, J., Starbek, M. (2008) Project Management of product development. Strojniški vestnik -Journal of Mechanical Engineering, vol. 54, no. 9, p. 588-606. [3] Bargelis, A. (1998) The Modeling structure of intelligent manufacturing system for mechanical components. Proceedings of the 2nd International Conference on Integrated Design and Manufacturing in Mechanical Engineering (IDMME'98), Compiegne, France, May 25-28. 1998. [4] Gebresenbet, T., Jain, P.K., Jain, S.C. (2002) Preliminary manufacturability analysis using feature-function resource considerations for cylindrical machined parts. Int. J. Computer Integrated Manufacturing, vol 15, p. 361-378. [5] Bargelis, A., Hoehne, G., Česnulevičius, A. (2004) Intelligent functional model for costs minimization in hybrid manufacturing systems. Informatica, vol. 15, p. 3-22. [6] Chua, C.K., The, S.H., Gray, R.K.L. (1999) Rapid prototyping versus virtual prototyping in product design and manufacturing. The International Journal of Advanced Manufacturing Technology, vol. 15, p. 597603. [7] Wiendahl, H.P., Fiebig, T.H. (2003) Virtual factory design - a new tool for a cooperative planning approach. Int. J. Computer Integrated Manufacturing, vol. 16, p. 535-541. [8] Rehg, A., Kreaber, H.W. (2001) Computer integrated Manufacturing. Columbus, Ohio. [9] Buede, D.M. (2000) The engineering design of systems: models and methods. John Wiley & Sons, Inc., New York, ISBN 0471250317. [10] Bargelis A., Mankute R. (2005) Internet web-based integration of process and manufacturing resources planning, in A. Dolgui et al (Eds), Supply Chain optimization: product/process design, facility location and flow control. Applied Optimization, Springer, vol. 94, p. 233-246. [11] Martin, P., D'Acunto, A. (2003) Design of a production system: an application of integration product - process, Int. J. Computer Integrated Manufacturing, vol., p. 509-516. [12] Oakland, j.s. (1999) statistical process control, 4th ed. Butterworth-Heinemann, UK. ISBN 07-50640987. [13] Motorcu, A.L., Gullu, A. (2006) Statistical process control in machining, a case study for machine tool capability and process capability. Materials and Design, vol. 27, p. 364-372. [14] Ayyub, B.M. (2001) Elicitation of expert opinions for uncertainty and risk. CRC Press, New York, ISBN 1-58488-286. [15] Boothroyd, g. et al. (2002) product design for manufacture and assembly - Marcel Dekker, New York, ISBN 0-8247-04062-9. [16] Bargelis, A. (2004) Web - based integration for knowledge sharing in hybrid manufacturing systems. Proceedings of the 10th ICPQR International Conference on Productivity and Quality Research, Miami, Florida, USA, February, 16-19., 2004. [17] Dudek-Burlikowska M. (2005) Quality estimation of process with usage control charts type x-r and quality capability of process Cp and Cpk, Journal of Materials Processing Technology, 162-163(2005), p. 736-743. [18] Bauer, L.W. (2002). Process capability modeling - GE global research. Report no. 2002GRCO38.