Structural Design And Path Planning In Additive Manufacturing

Additive manufacturing has been applied in various fields such as industry,aerospace,biomedical engineering and so on.At the same time,structural design has also been vigorously developed.The widespread attention to structural design also pro-vides more diverse inputs for additive manufacturing(3D printing).Structural design answers the question what to print,and additive manufacturing solves the problem how to print.These two technologies complement and promote each other,which are two vital parts in industrial production.This dissertation will focus on the structural opti-mization and path planning methods in additive manufacturing.There are many optimization problems in the field of structural design and this dissertation focuses on stress-based topology optimization and the design of differen-tiable microstructures.Topology optimization with stress constraints is a very important branch in the field of structural design,which gains fast developed in recent decades.However,there is no unified solution algorithm,which is mainly due to the three ma-jor difficulties:locality,singularity and high non-linearity.In chapter 3,the alternating direction method of multipliers based on the Lagrange multiplier method is proposed.The optimization model has two sets of design variables,that are the stress and density of elements.The algorithm introduces stress as a design variable and imposes an equal constraint on the model of stress calculation.The equal constraint is then placed as a soft constraint in the objective function using the augmented Lagrange method,and the corresponding stress and density variable are optimized alternately.The method is solved separately using finite element analysis and isogeometric analysis tools.Numeri-cal examples demonstrate the effectiveness of the proposed algorithm on two numerical solutions.The design of periodic microstructures is a basic problem in the field of structural design.In recent years,functionally graded materials(FGMs)whose the physical prop-erties show gradient changes with the increasing of volume have become a hot topic.In this dissertation,the concept of"differetialable microstructure" is put forward based on FGMs.Differentiable microstructure is controlled by continuous parameters,its vol-ume and physical properties are continuously changing with these parameters,and its physical properties(such as elastic tensor,Poisson's ratio)are required to be as close as possible to the upper limit of Hashin-Shtrikman curve.The basic idea of this disser-tation is to construct a physical field through parameters,and each cross section of the physical field corresponds to one microstructure.If the physical field is continuous,the volume is continuous and naturally satisfies the requirement of connectivity.In order to design the microstructure with optimal mechanical properties,this dissertation takes the heat conductivity as design variables and the sum of the physical properties of the key microstructure as the objective function.After that,we could obtain the physical field(temperature field)by solving the heat conduction equation.Numerical experiments demonstrate the effectiveness and robustness of the proposed algorithm.Structural design provides a variety of options for additive manufacturing inputs.When the inputs of additive manufacturing are available,Printing efficiency is a prob-lem worthy of attention.Here,we focus on the path planning problem.At present,there is a lack of complexity domains in most path planning algorithms.The generated paths are universal.However,complex structures are common in nature,such as bones,corks,hives,corals,etc.For the model with complex slices,the existing filling path will appear discontinuous and low efficiency.In this dissertation,the path planning prob-lem of complex structures is studied,and the complex models are divided into complex boundary models and complex topology models.The idea of"divide and conquer" is adopted for path planning.Firstly,the complex model is simplified,and the purpose of path filling is achieved through path filling for simple areas and then global connection.The path is measured by printing cost(time,material),path segmentation,sharp turn ratios,infill ratios,and visual effect.The proposed algorithm is compared with the tradi-tional path planning methods(Zigzag path,contour parallel path and global continuous Fermat spiral path).Experimental results show that the proposed algorithm is superior to other methods in material cost,printing time and structural stability.