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 ComputerAidedEngineering (CAE) and optimization methods have been carried
out to observe and finetune 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 CMold 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. In this paper, a featuredbased
approach to this problem is proposed. Superquadrics is presented as a tool for
recognizing the plastic part shapes and an algorithm is applied for generating
the center line of the thermal subsystem 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.
Conformal cooling channel, as the name implies, Balenciaga Bag 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
3D 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 socalled cooling zones. In the next step, a subsystem of
cooling channel is generated for each cooling zone. These subsystems of cooling
channels are further decomposed into smaller elements called cooling cells which
are easy to be analysed. Lastly, the combination process of these subsystems 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.
Nowadays, featurebased 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 subfeatures that must be recognized for the partial thermal
system designs. Feature recognition has http://www.balenciagatop.com/
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: graphbased
methods, volumetric methods and hintbased 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,
freeform surfaces are mostly used and hence, freeform 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 graphbased methods, volumetric methods
nor hintbased methods can be applied.
没有评论:
发表评论