Computational Design Of Functional Cavity Shapes

Cavity shape and shell model are widely used in industrial manufacturing,aerospace,medical devices,product design and other fields.Modifying the structure and shape by hand of human designers using computer-aided design(CAD)software is often cumbersome and repetitive,which consumes a lot of time,because it is difficult for designers to predict the physical properties of the visual appearance of models.On the other hand,due to the advancement of additive manufacturing technology,models with complex topology and shape can be fabricated more conveniently and fast,thus the shape design of model has greater freedom.At the same time,the development of 3D printing technology enables non-expert users to print out small-scale models,which also need to achieve certain functionalities.In this paper,the parametric design of cavity model is studied,and physical simulation model is established to predict the physical properties of the model,such as structural strength,standing,buoyancy,roly-poly and rotation stability.According to the simulation of physical properties of the model,the automatic iterative optimization design is carried out.Specifically,this paper proposes two types of representation to parameterize the shell cavity shape models: the scalar field on the surface and the function representation.Based on the representation of scalar field,the shell model is parameterized into scalar field on the model surface,and applied to the structural strength optimization of the model to design the inner surface shape of the model.The thickness parameters of the model are optimized according to the stress analysis results of the model,and the final optimized manufacturing model is obtained.The parametric method of cavity model based on function representation introduces the implicit representation of geometric model to describe the input shape.Through the Boolean operation between shapes,the shell cavity model can be expressed as a continuous differentiable function.The function representation of the model defines a continuous smooth surface,which is more convenient in manufacturing.The introduction of Boolean function makes the internal surface fusion of the model not need additional special treatment.This property expands the design space of the model,and can get more novel topology changing shapes,with better physical performance to provide the required functions.At the same time,different from the previous empirical optimization related work,the modeling based on function representation in this paper can provide optimization Gradient,which can rely on mature parameterized optimization operators to reduce the number of iterations to the optimal structure.Based on the parameterized representation,this paper proposes a mesh-free simulation method to predict the physical properties of the entity model,which effectively avoids the high-cost grid re division operation and makes the optimization process more efficient.Compared with the previous work,the parametric method proposed in this paper can get a more novel and efficient structure,and has higher accuracy and calculation speed.The effectiveness and efficiency of the method are verified through the verification of the optimization model of solid manufacturing.