National conference on Advancements in Mechanical Engineering and Energy Environment (6-7 January, 2012), SLIET Longowal

A study on automated side-core design for die-casting dies

Abstract: Die-Casting process is capable of producing simple as well as complex shaped metal components of aluminium, zinc and magnesium alloys in large quantities. The complex shaped components have features that interfere in opening of the die and these features are called undercuts. A special tooling, called side-core, is required for moulding of undercuts. The direction of movement of side-core is different from the direction of opening of die halves. For movement of side-core additional actuating mechanism has to be provided on the die-casting machine. Thus, the design of side-core requires special effort and attention on the part of die-casting engineer. Traditional process of side-core design requires lot of human expertise and is too much time consuming. Furthermore, the process of identifying the undercuts and side-core design is iterative and prone to errors which results in higher lead times for manufacturing of die-casting components. The researchers in recent past felt the need to automate the process of side-core design so that the lead times and the dependency of the design process on human expertise could be reduced. This paper discusses the issues in automating side-core design. A brief summary of previous work is also presented that throws light on the achievements and limitations of previous research work. An algorithm has been proposed which addresses the problem of side-core design and eliminates deficiencies found in the previous systems. Firstly, the undercuts are detected from the part CAD model. Thereafter, the release direction and distance for the undercuts is calculated which is followed by design of side-core. Presently, algorithm has been tested on a few parts and it needs to be tested on more complex parts. Developed algorithm would be used for proposing a system so that process of side-core design is automated. . We are working to make this algorithm more robust and trying to implement it in the form of a system. Proposed algorithm would prove to be a good effort towards Design-Manufacturing integration of die-casting process. 1. Introduction Die-casting is the process that provides the shortest way to convert molten metal into finished product. The product of die casting process is ready to use with no or very little further machining or finishing process. A single die tool is capable of producing up to ten million of components making the process a profitable one. Quality of the components produced and the life of the die depends upon the design of the die. Therefore, the design activity for a die needs to be controlled. Generally die casting parts have features, called undercuts that require extra tooling, called side core, for their moulding as illustrated in Fig.1. The side-cores are responsible for incurring extra cost and time in die-design activity. The traditional method of designing side core manually is iterative and prone to errors. Additionally, the expertise of die-casting engineer is required at each step to realize the side-core design. The automation of side-core design is required to reduce the manufacturing lead times and cost. The following section presents the literature on automated die-design covering the topics automation of parting direction determination, undercut detection, parting line selection and side-core design.

National conference on Advancements in Mechanical Engineering and Energy Environment (6-7 January, 2012), SLIET Longowal

Figure. 1 Die-casting die with nomenclature 2. Literature review The automation of side-core design involves consideration of a number of die-design factors due to which the literature on the related factors has also been discussed. etermination of parting direction 2.1 Selection of parting direction is generally done by the die-casting engineer. Selection of parting direction affects the number of undercuts for the part to be die-cast. Following paragraphs present some of the research in automatic parting direction selection. Hui and Tan [6] presented an algorithm for mould design. The algorithm utilizes sweep operation to obtain core and cavity of the mould and uses heuristic search technique for selection of paring direction. A series of mesh points on every surface are used for estimating the blocking factor. The value of blocking factor indicates the amount of undercut in the candidate parting direction. Side-core directions are selected using side-core blocking check. Chen [3] presented an algorithm to determine parting direction. The algorithm depends on minimum volume bonding box for 3-D model of the part. Three orthogonal parting directions are selected as candidate parting directions. A fuzzy representation was used to select best parting direction based on block factor, draw distance and projected area in the candidate parting direction. Yin [14] et al. proposed a methodology for mouldability analysis of near-net-shaped parts. Freedom cones are employed to identify external and internal undercuts, and optimal parting direction with minimum number of undercuts. Khardekar et al. [7] presented an algorithm for finding the feasible mould parting directions using tessellated model of the part. Facet visibility is calculated using graphics hardware. The algorithm also detects the undercut features and performs the draft analysis. Undercut-free direction is determined using Quadtree and Convex hull intersection algorithms. 2.2 etection of undercut features Side-core is a tool to realize undercut features so automatic detection of undercuts is necessary for side-core design. Following paragraphs throw some light on previous work on automated undercut detection. Fu et al. [4] developed an algorithm for generation of optimal parting direction based on minimum number of undercut features in injection moulded parts. Undercut features are categorized based on the connectivity with the target surface and their geometrical information.

National conference on Advancements in Mechanical Engineering and Energy Environment (6-7 January, 2012), SLIET Longowal

Lu et al. [8] proposed a methodology to detect interference elements and release direction in die-casting and injection mould components. Interference element indicator, which is a count of the number of rays from a test point striking against the component body, is used to recognize the interference element or the undercut. Yin et al. [15] presented a methodology for geometric mouldability analysis by addressing the issues like parting line determination, undercut detection, etc. Common boundary of core and cavity moulded regions is taken as the parting line. Undercuts are recognized from concave features using volume-based feature recognition method. However, the implementation of the system is limited to polyhedral parts with maximum edge loop only. Nee et al. [11] proposed a system for determination of parting direction in plastic injection mould design. Parting direction which maximizes the number of undercuts that can be moulded by core and cavity is selected as the best parting direction. Undercuts are categorized as external and internal undercuts that are further classified as inside and outside undercuts. However the application of the system is limited to parts which are having unique outer edge loop. Zhang et al. [16] proposed a methodology for undercut and parting direction determination from moulded parts which are having free-form surfaces. Concave Region representation (CR-rep) is constructed from B-rep model to extract depressions and protrusions. Depressions and protrusions are defined on the basis of CR-rep and continuity (C0 and C1). Bassi et al. [2] presented a system to recognize protrusion and depression features along-with intersecting depression features. The system first classifies the faces of the part on the basis of accessibility. Thereafter, the features are classified based on geometry, topology and accessibility. Sweeping and regularized Boolean operations are used to analyse the faces. System also presents methodology to determine the release direction of undercut features. Madan [10] presented a methodology to recognise the intersecting undercut features. Further, the algorithms are presented to identify the intersecting features into separable regions. The algorithm uses the concept of feature recognition to first recognise the undercut features and then division of feature is done based on accessible direction for parts of the feature. The algorithm is successfully implemented on intersecting features with L and T shapes. 2.3 Side-core design Banerjee et al. [1] developed algorithms for automated design of side actions for injection moulded parts. The algorithms work on the faceted part as input and calculate the retraction space for the undercut feature that has been recognised earlier. The algorithm presents a novel approach in understanding the moulding of undercuts using the side-cores which could be utilised in parting line selection. Lu et al. [9] proposed a method to detect the interference elements and their release direction for die-casting and injection moulding components. Three dimensional ray method has been employed to detect the interference elements. Release direction of an interference element is calculated by first detecting the release direction of all the points, curves and surfaces of the interference element. The methodology is helpful in determination of release direction of the interference elements upon which side-core design could be built upon. The methodology is limited to simple undercuts, intersection undercuts and tapered undercut features have not been taken into account. Fu et al. [5] presented concept of visibility demouldability and mouldability to define the surfaces and then determining the most preferred moulding direction for the undercut features. The undercuts that are mouldable in same are grouped in same side-core and main body of the side-core is then generated. The system presented could not design side core for the intersecting and undercut features.

National conference on Advancements in Mechanical Engineering and Energy Environment (6-7 January, 2012), SLIET Longowal

Ye et al. [13] used graph based methods for recognition of undercut features and classified these into depressions and protrusions. The proposed method is capable of handling intersecting undercut features. The side-cores are created by trimming the bounding box of the undercut feature. The core moment is only in a direction perpendicular to the principal parting direction. Scanlan [12 ] studied the effect of injection on side core. The study presents the fundamentals of side-core performance. The effect of length of core on its deflection due to the injection pressure was studied. The engineering relationships of stress and strain on side-cores were also reported. The study is limited to the functioning of the side-core rather than the designing of side-core. 3. Research gaps and objectives of research The literature review reveals that the available systems for side-core design have some limitations. The following are some of the grey areas for automated side-core design  side-core design for intersecting undercut features  design of side-core for tapered undercut feature and  consideration of side-core performance during the injection of metal the algorithm presented in this paper is an attempt to address the gaps in previous research. The proposed system is capable of detecting the undercut features and then classifying them into depressions and protrusions. Then the methodology developed by Madan [] is used to detect intersecting features and then divide into separate regions. Finally, the release direction of the separate regions is calculated and side-core is designed by extruding the contour of the undercut loop in release direction. 4. Automated side-core design This section presents the algorithm developed for automated side-core design. The algorithm begins with the determination of non-convex regions of the part model. The undercut regions are then found from non convex regions by checking the overlap of surfaces in parting direction. Undercuts are then classified into simple and intersecting undercuts. Intersecting undercuts are separated into simple undercuts and the release direction for each undercut is calculated. Finally the surfaces of undercut are determined and filled to get the side-core. Adjustment in length of side core is made to compensate for strain during injection of the metal into the die. The algorithm is illustrated with the help of a flow chart in Fig. 2. 5. Results and conclusions The proposed algorithm has been tested on some industrial parts having moderate complexity and results obtained are encouraging. After a few moderations the algorithm would be applied to parts with more complexity. Presently, the algorithm involves a lot of human intervention at various stages. Automation has been done for determining undercut features. The full automation could be done further to reduce the need of human intervention.

National conference on Advancements in Mechanical Engineering and Energy Environment (6-7 January, 2012), SLIET Longowal

Input part model Determine the non convex surfaces of the part

Is non-convex surface overlapped in parting direction? Yes Surface is part of undercut feature Connect the neighbouring surfaces to form the undercut regions

No Mouldable surface

Is there any convex edge inside the undercut? Yes Undercut feature is an intersecting one Separate the intersecting-undercut feature into the simple components Determine the release direction of each undercut feature Select the surfaces of the undercut to form side-core by filling the space Adjust the core length for strain during injection of metal into the die Final side-core design

No A simple undercut feature

Fig. 2. Flow chart for automated side core design

National conference on Advancements in Mechanical Engineering and Energy Environment (6-7 January, 2012), SLIET Longowal

6.

Acknowledgement:

This paper was part of the research project titled 'Automated Design of Die from part Product Model'. Authors are thankful to All India Council for Technical Education (AICTE) which has funded this project under Research Proposal Scheme vide grant no. 8023/BOR/RID/RPS146/2008-09 dated March 12, 2009 to Mechanical Engineering Department of SLIET, Longowal, Punjab, India. 7. References [1] Banerjee, A. G., and Gupta, S. K., 2007, Geometric algorithms for automated edesign of side actions in injection moulding of complex parts, Computer Aided Design, 39, pp. 882-897. [2] Bassi, R., Reddy, N.V., and Bedi, S., 2010, Automatic recognition of intersrscting features of side core design in two piece permaneent molds, Int J Adv Manuf Technol., 50, pp. 421-439. [3] Chen, Y.H., 1997 Determining parting direction based on minimum bonding box and fuzzy logics, Int. Journal of Machine Tools and Manufacture, 37 (9), pp. 1189-1199. [4] Fu, M.W., Fuh, J.Y.H., and Nee, A.Y.C., 1999, Generation of optimal parting direction based on undercut features in injection moulded parts, IIE Transactions, 31, pp. 947-955. [5] Fu, M.W., 2007, The application of surface demoldability and mouldability to sidecore design in die and mold CAD,40, pp. 567-575. [6] Hui, K.C., Tan, S.T., 1992 Mould design with sweep operations- a heuristic search approach, Computer-Aided Design, 24 (2), pp. 81-92. [7] Khardekar, R., Burton, G., and McMains, S., 2006, Finding feasible mold parting directions using graphics hardware, Computer-Aided Design, 38, pp. 327-341. [8] Lu, H.Y., and Lee, W.B., 1999, Detection of interference elements and release direction in die-cast and injection-moulded components, Proc Instn Mech Engrs, 214, pp. 431-441. [9] Lu, H.Y., and Lee, W. B., 2000, detection of interference elements and release directions in die-cast and injection-moulded componenets.214(B), pp. 431-441. [10] Madan, J., Singh, A., and Kumar, S., 2008, Recognition of intersecting features and identification of separable regions for design-manufacturing integration, proceedings of international conference on digital factory, pp. 60-67. [11] Nee, A.Y.C., Fu, M.W., Fuh, J.Y.H., Lee, K.S., and Zhang, Y.F., 1997, Determination of optimal parting directions in plastic injection mold design, Annals of the CIRP, 46 (1), pp. 429-432. [12] Scanlan, M., 2001 Mold side-actions: how, why and when they work, website of www.pfa-inc.com/pdfs/Moldmakingarticle2partsweb.pdf, surfed on 27 June, 2011. [13] Ye, X.G., Fuh, J.Y.H., and Lee, K.S., 2004, Automatic Undercut Feature Recognition for Side Core Design of Injection Molds, J. Mech. Des., 126 (3), pp. 519-526. [14] Yin, Zhou-Ping, Ding, H., and Xiong, You-Lun, 2000, Mouldability analysis for near net shaped manufacturing parts using freedom cones, Int J Adv Manuf Technol, 16, pp. 169-175. [15] Yin, Zhou-Ping, Ding, H., Li, Han-Xiong, and Xiong, You-Lun, 2004, Geometric mouldability analysis by geometric reasoning and fuzzy decision making, ComputerAided Design, 36 (1), pp. 37-50. [16] Zhang, C., Zhou, X., and Li, C., 2010, Feature extraction from freeform molded parts for moldability analysis, Int J Adv Manuf Technol., 48, pp. 273-282.