Automatically generating conformal cooling channel design for plastic injection molding.

Park, Hong Seok ; Pham, Ngoc Han

Abstract: This paper proposes a method for developing a conformal cooling system for injection molding that facilitates rapid temperature change at the mold surface with the minimum cycle time. A plastic part with the complex shape is decomposed into simpler surfaces. The cooling channels for individual surfaces are first obtained and then combined to form the conformal cooling channel system for the entire part.

Keywords: Injection molding, Conformal, Feature recognition, Solid Freeform Fabrication

1. INTRODUCTION

The constant temperature mold of molding plastic parts with high precision contours is of significance in determining not only the productivity of the injection molding process but also the product quality. A solution to this challenge is the rapid thermal response molding process in which uniform temperature overall the mold part ensures the product quality by preventing differential shrinkage, internal stress and mold release problems (Li, 2001). Many Computer-Aided-Engineering (CAE) and optimization methods have been carried out to observe and fine-tune the influences of the thermal system (Park et al., 1998). The results of these research works are obtained by using thermal analysis modules of commercial CAE packages such as C-Mold or Moldflow which are based on the initial designs generated by the human. By given an initial thermal configuration design, efficiency and quality of the molded part can be predicted before an actual plastic mold is manufactured. One more necessary step for the complete automation in the molding thermal system is to generate the initial design for the conformal cooling channels. In this paper, a featured-based approach to this problem is proposed. Super-quadrics is presented as a tool for recognizing the plastic part shapes and an algorithm is applied for generating the center line of the thermal sub-system of each individual surface. Finally, these sub sets of center lines are combined to create a unique center line which is the guide line for generating the cooling channel of the thermal system.

2. INFORMATION

Conformal cooling channel, as the name implies, refers to the channels that conform to the surface of the mould cavity. Conformal cooling channels have demonstrated simultaneous improvement in production rate and part quality as compared with conventional production tools. In the previous researches, cooling line design and fabrication have been confined to relatively simple configuration, primarily due to the limits of the fabrication method used to make tools, but also due to the lack of appropriate design methodology. Emergence of Solid Freeform Fabrication processes with the ability to fabricate 3-D feature with almost arbitrary complexity is exceedingly useful to mould design process (Xu et al., 2001). The remaining problem to be solved is how to optimize the design process of the thermal system. In this paper, a systematic method for designing cooling channel is proposed. Firstly, the feature recognition algorithm is applied to identify and decompose the moulded part into manageable sections so-called cooling zones. In the next step, a sub-system of cooling channel is generated for each cooling zone. These sub-systems of cooling channels are further decomposed into smaller elements called cooling cells which are easy to be analysed. Lastly, the combination process of these sub-systems is done to create a complete conformal cooling system for the whole plastic part based on the constraints of the combination algorithm and design rules.

3. FEATURE SEGMENTATION AND FEATURE RECOGNITION

Nowadays, feature-based modeling has been a standard for 3D designs. Most of the complex shapes are obtained by synthesizing from sets of simple features. This design strategy is not sensitive to the part geometry; therefore, it keeps the simplicity of the design routine no matter how complicated the geometry of the part is. For the same purposes of simplicity and efficiency, the molded part is segmented into sub-features that must be recognized for the partial thermal system designs. Feature recognition has drawn much attention from researchers and been proposed in literatures (Lentz et al., 1993). The majority of these has based on machining feature recognition techniques which can be classified in three categories: graph-based methods, volumetric methods and hint-based methods. Although recent machining feature recognition technique can be a good solver for parts with complicated intersecting feature, this technique is not appropriate for detecting shape feature for thermal system design of plastic products. In plastic products, free-form surfaces are mostly used and hence, free-form features have to be processed. Furthermore, a shape feature in a plastic part may blend smoothly to another feature and the boundaries between features can not be explicitly defined. With these two reasons, neither graph-based methods, volumetric methods nor hint-based methods can be applied.

In order to represent the feature template, the shape component must be able to cover a wide range of shapes that are commonly found in injection mould design. Furthermore, an algorithm must be able to recognize the shape from the plastic part. Super-quadrics is proposed as a method for the shape component because it satisfies both requirements. It is found that in almost all system designs, super-quadrics has the ability to represent the shapes of the plastic parts. A super-quadrics is given by an implicit equation as follows:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

Where:

[a.sub.1], [a.sub.2], [a.sub.3] define the size; [a.sub.4], [a.sub.5] define the shape of the super-quadrics. To determine if a given point (x, y, z) belongs to a superquadrics or not, a function q is defined as follows:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

The function q equals or is less than or greater than zero if the given point (x, y, z) is on, inside or outside the super-quadrics. This super-quadrics will help find the approximate shapes of the sets of points. This super-quadrics fitting problem has been formulated as a non-linear least square minimization problem which can be solved by Levenberg-Marquardt method (Press et al., 1997).

4. GENERATING GUIDE LINES FOR

CONFORMAL COOLING CHANNEL

After the molded part is segmented into zones so-called cooling zones, they must be recognized which type of super-quadrics they belong to. In the next step, every single cooling zone is divided one more time into appropriate sized regions so-called cooling cells which are thermal-treated by a single channel. After that, each cooling cell is rectangle-meshed into n rows and m columns. The size of a rectangular mesh are to be chosen small enough so that every cooling cell formed by 4 points [P.sub.i,j], [P.sub.i,j+1], [P.subn.i+1,j] and [P.sub.i+1,j+1] is approximately considered as a rectangle to its own local coordinate. According to this assumption, sets of new points are formed.

A set of points which belong to i-th row are called [P.sub.i]. Every two sets of [P.sub.i] and [P.sub.i+1] are taken to form a new set of [P.sub.i]' of which coordinate is given by:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)

(n-1) new sets of points will be formed and this routine is iterated (n-1) times until only one set of points is obtained. Finally, a smoothing algorithm, which is spline algorithm, is applied to this set of points to create a conformal guide line for generating the cooling channel for this cooling cell. An example is carried out to prove the conformal characteristics of the proposed method as shown in Fig. 2. A cooling zone which is rectangle-meshed into 5 rows and 9 columns is considered. This cooling zone has 5 rows supposed as 5 sets of points. Coordinates (x, y, z) of points are obtained for the calculation routine which is iterated until only one set of points are attained, i.e. 4 iterations. Lastly, a spline line is created with the control points which are the unique sets of points obtained above. With this approximation algorithm, creating the guide line for the whole cooling system not only is sensitive to part geometry but also ensures the conformal characteristics of the cooling channel.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

5. COMBINATION

This is the final step in automatically generating the thermal channel for injection plastic molding. The purpose of combining different partial guide lines into a unique set of guide lines is to generate the complete cooling system for the plastic mould. If adjacent guide lines are to be merged, they must have the same relative directions and radius of curvature, to within a specified tolerance. This problem has been discussed in many previous literatures known as Linking Adjacent Segment problem. The remaining problem is how many guide lines should be linked together. The combination is based not only on the restrictions of algorithm but also the design rules including the design for coolant pressure drop, design for coolant temperature uniformity, design for sufficient part cooling, design for uniform cooling and design for mold strength. Those present a constraint on the length and the complexity of the cooling channels. The efficient and uniform control of the mold temperature through conformal cooling makes the first step toward the dynamic thermal management of the injection molding process for quality and productivity.

6. CONCLUSION

This paper presents a systematic method for designing conformal cooling channels which adds one more step to the fully automatically generating injection molding design. This technique combined with Solid Freeform Fabrication processes can create molding tooling with complex channels, offering the potential for substantial improvement in production rate and part quality.

7. REFERENCES

Lentz, D.H. & Sowerby, R. (1993) Feature Extraction of Concave and Convex Regions, Computer-Aided Design, Vol. 25 pp. 421-437.

Li, C.L. (2001). A feature-based approach to injection mould cooling system design, Computer-Aided Design, Vol. 33 pp. 1073-1090.

Park, S.J. & Kwon, T.H. (1998). Optimal Cooling System Design for the Injection Molding Process, Polymer Engineering and Science, Vol. 38, No.9.

Press, W.H, Flammery, B.P., Teukolsky, S.A., Vettering, W.T. (1997). Numerical recipes in C: The Art of Scientific Computing, 2nd Ed, Cambridge University Press, New York.

Xu, X., Sachs. E. & Allen, S. (2001), The Design of Conformal Cooling Channels in Injection Molding Tooling, Polymer Engineering & Science, Vol. 41, No. 7.