Towards Nonlinear Multimaterial Topology Optimization Using Unsupervised Machine Learning and Metamodel-Based Optimization

2015 ◽  
Author(s):  
Kai Liu ◽  
Andrés Tovar ◽  
Emily Nutwell ◽  
Duane Detwiler
Keyword(s):  
Machine Learning ◽  
Topology Optimization ◽  
Density Distribution ◽  
Compliant Mechanisms ◽  
Optimization Method ◽  
Kriging Interpolation ◽  
Unsupervised Machine Learning ◽  
Continuous Density ◽  
Structural Problems ◽  
Synthesis Strategy

This work introduces a multimaterial density-based topology optimization method suitable for nonlinear structural problems. The proposed method consists of three stages: continuous density distribution, clustering, and metamodel-based optimization. The initial continuous density distribution is generated following a synthesis strategy without penalization, e.g., the hybrid cellular automaton (HCA) method. In the clustering stage, unsupervised machine learning (e.g., K-means clustering) is used to optimally classify the continuous density distribution into a finite number of clusters based on their similarity. Finally, a metamodel (e.g., Kriging interpolation) is generated and iteratively updated following a global optimization algorithm (e.g., genetic algorithms) to ultimately converge to an optimal material distribution. The proposed methodology is demonstrated with the design of multimaterial stiff (minimum compliance) structures, compliant mechanisms, and a thin-walled S-rail structure for crashworthiness.

Topology Optimization of Compliant Mechanisms Using Hybrid Discretization Model

2010 ◽  
Vol 132 (11) ◽  
Author(s):  
Hong Zhou
Keyword(s):  
Topology Optimization ◽  
Compliant Mechanisms ◽  
Compliant Mechanism ◽  
Optimization Procedure ◽  
Optimization Method ◽  
Diagonal Element ◽  
Initial Guess ◽  
Stress Constraint ◽  
Design Variables ◽  
Hybrid Discretization

The hybrid discretization model for topology optimization of compliant mechanisms is introduced in this paper. The design domain is discretized into quadrilateral design cells. Each design cell is further subdivided into triangular analysis cells. This hybrid discretization model allows any two contiguous design cells to be connected by four triangular analysis cells whether they are in the horizontal, vertical, or diagonal direction. Topological anomalies such as checkerboard patterns, diagonal element chains, and de facto hinges are completely eliminated. In the proposed topology optimization method, design variables are all binary, and every analysis cell is either solid or void to prevent the gray cell problem that is usually caused by intermediate material states. Stress constraint is directly imposed on each analysis cell to make the synthesized compliant mechanism safe. Genetic algorithm is used to search the optimum and to avoid the need to choose the initial guess solution and conduct sensitivity analysis. The obtained topology solutions have no point connection, unsmooth boundary, and zigzag member. No post-processing is needed for topology uncertainty caused by point connection or a gray cell. The introduced hybrid discretization model and the proposed topology optimization procedure are illustrated by two classical synthesis examples of compliant mechanisms.

Dynamic Topology Optimization of Compliant Mechanisms and Piezoceramic Actuators

2002 ◽  
Author(s):  
Hima Maddisetty ◽  
Mary Frecker
Keyword(s):  
Topology Optimization ◽  
Compliant Mechanisms ◽  
Compliant Mechanism ◽  
Optimization Method ◽  
Mechanical Efficiency ◽  
Optimal Topology ◽  
Driving Frequency ◽  
Spring Stiffness ◽  
Piezoceramic Actuator ◽  
Near Resonance

A topology optimization method is developed to design a piezoelectric ceramic actuator together with a compliant mechanism coupling structure for dynamic applications. The objective is to maximize the mechanical efficiency with a constraint on the capacitance of the piezoceramic actuator. Examples are presented to demonstrate the effect of considering dynamic behavior compared to static behavior, and the effect of sizing the piezoceramic actuator on the optimal topology and the capacitance of the actuator element. Comparison studies are also presented to illustrate the effect of damping, external spring stiffness, and driving frequency. The optimal topology of the compliant mechanism is shown to be dependent on the driving frequency, the external spring stiffness, and if the piezoelectric actuator element is considered as design or non-design. At high driving frequencies, it was found that the dynamically optimized structure is very near resonance.

scholarly journals Design and Development of a Compliant Clutch Fork using Topology Optimization

2019 ◽  
Vol 8 (11) ◽  
pp. 3625-3629
Keyword(s):  
Topology Optimization ◽  
Compliant Mechanisms ◽  
Optimization Method ◽  
Automotive Application ◽  
Material Distribution ◽  
Optimum Size ◽  
Elastic Continuum ◽  
Active Research ◽  
Design Experiments ◽  
Topology Optimization Method

Compliant mechanisms and its systems are the focus of the active research. It describes a single elastic continuum used to transfer the motion and force mechanically. Their flexibility and stabilities are significant. Topology optimization Method is taken for designing the compliant mechanisms. It is a Material distribution approach for finding the optimum size and shape of the structure. The Author focused mainly on automotive application of Compliant Mechanism.i.e Design and implement of compliant clutch fork using topology optimization. Dimensional data is gathered in order to model the actual clutch fork. Compliant clutch fork designs are developed by reducing the weights compare to actual clutch fork with the help of topology optimization to get optimal compliant design. Experiments are directed to confirm the functionality of compliant clutch fork.

The Modified Quadrilateral Discretization Model for the Topology Optimization of Compliant Mechanisms

2011 ◽  
Vol 133 (11) ◽  
Author(s):  
Hong Zhou ◽  
Pranjal P. Killekar
Keyword(s):  
Topology Optimization ◽  
Compliant Mechanisms ◽  
Compliant Mechanism ◽  
Optimization Procedure ◽  
Local Stress ◽  
Optimization Method ◽  
Stress Constraint ◽  
Cell Problem ◽  
Design Variables ◽  
Design Cell

The modified quadrilateral discretization model for the topology optimization of compliant mechanisms is introduced in this paper. The design domain is discretized into quadrilateral design cells. There is a certain location shift between two neighboring rows of quadrilateral design cells. This modified quadrilateral discretization model allows any two contiguous design cells to share an edge whether they are in the horizontal, vertical, or diagonal direction. Point connection is completely eliminated. In the proposed topology optimization method, design variables are all binary, and every design cell is either solid or void to prevent gray cell problem that is usually caused by intermediate material states. Local stress constraint is directly imposed on each analysis cell to make the synthesized compliant mechanism safe. Genetic algorithm is used to search the optimum. No postprocessing is required for topology uncertainty caused by either point connection or gray cell. The presented modified quadrilateral discretization model and the proposed topology optimization procedure are demonstrated by two synthesis examples of compliant mechanisms.

The Modified Quadrilateral Discretization Model for the Topology Optimization of Compliant Mechanisms

2011 ◽  
Author(s):  
Hong Zhou ◽  
Pranjal P. Killekar
Keyword(s):  
Topology Optimization ◽  
Compliant Mechanisms ◽  
Compliant Mechanism ◽  
Optimization Procedure ◽  
Local Stress ◽  
Optimization Method ◽  
Initial Guess ◽  
Stress Constraint ◽  
Design Variables ◽  
Conduct Sensitivity Analysis

The modified quadrilateral discretization model for the topology optimization of compliant mechanisms is introduced in this paper. The design domain is discretized into quadrilateral design cells. There is a certain location shift between two neighboring rows of quadrilateral design cells. This modified quadrilateral discretization model allows any two contiguous design cells to share an edge whether they are in the horizontal, vertical or diagonal direction. Point connection is completely eliminated. In the proposed topology optimization method, design variables are all binary and every design cell is either solid or void to prevent grey cell problem that is usually caused by intermediate material states. Local stress constraint is directly imposed on each analysis cell to make the synthesized compliant mechanism safe. Genetic algorithm is used to search the optimum and avoid the need to select the initial guess solution and conduct sensitivity analysis. No postprocessing is needed for topology uncertainty caused by point connection or grey cell. The presented modified quadrilateral discretization model and the proposed topology optimization procedure are demonstrated by two synthesis examples of compliant mechanisms.

An Evolutionary Soft-Add Topology Optimization Method for Synthesis of Compliant Mechanisms With Maximum Output Displacement

2017 ◽  
Vol 9 (5) ◽  
Author(s):  
Chih-Hsing Liu ◽  
Guo-Feng Huang ◽  
Ta-Lun Chen
Keyword(s):  
Topology Optimization ◽  
Compliant Mechanisms ◽  
Volume Fraction ◽  
Compliant Mechanism ◽  
Optimization Method ◽  
Computational Time ◽  
Maximum Output ◽  
Spring Stiffness ◽  
Output Displacement ◽  
Topology Optimization Method

This paper presents an evolutionary soft-add topology optimization method for synthesis of compliant mechanisms. Unlike the traditional hard-kill or soft-kill approaches, a soft-add scheme is proposed in this study where the elements are equivalent to be numerically added into the analysis domain through the proposed approach. The objective function in this study is to maximize the output displacement of the analyzed compliant mechanism. Three numerical examples are provided to demonstrate the effectiveness of the proposed method. The results show that the optimal topologies of the analyzed compliant mechanisms are in good agreement with previous studies. In addition, the computational time can be greatly reduced by using the proposed soft-add method in the analysis cases. As the target volume fraction in topology optimization for the analyzed compliant mechanism is usually below 30% of the design domain, the traditional methods which remove unnecessary elements from 100% turn into inefficient. The effect of spring stiffness on the optimized topology has also been investigated. It shows that higher stiffness values of the springs can obtain a clearer layout and minimize the one-node hinge problem for two-dimensional cases. The effect of spring stiffness is not significant for the three-dimensional case.

Electro-thermal compliant mechanisms design by an evolutionary topology optimization method

2013 ◽  
Vol 30 (7) ◽  
pp. 961-981 ◽  
Author(s):  
Rubén Ansola ◽  
Estrella Veguería ◽  
Javier Canales ◽  
Cristina Alonso
Keyword(s):  
Topology Optimization ◽  
Compliant Mechanisms ◽  
Optimization Method ◽  
Topology Optimization Method

Dynamic Topology Optimization of Compliant Mechanisms and Piezoceramic Actuators

2004 ◽  
Vol 126 (6) ◽  
pp. 975-983 ◽  
Author(s):  
Hima Maddisetty ◽  
Mary Frecker
Keyword(s):  
Topology Optimization ◽  
Compliant Mechanisms ◽  
Compliant Mechanism ◽  
Optimization Method ◽  
Mechanical Efficiency ◽  
Optimal Topology ◽  
Driving Frequency ◽  
Spring Stiffness ◽  
Piezoceramic Actuator ◽  
Near Resonance

A topology optimization method is developed to design a piezoelectric ceramic actuator together with a compliant mechanism coupling structure for dynamic applications. The objective is to maximize the mechanical efficiency with a constraint on the capacitance of the piezoceramic actuator. Examples are presented to demonstrate the effect of considering dynamic behavior compared to static behavior and the effect of sizing the piezoceramic actuator on the optimal topology and the capacitance of the actuator element. Comparison studies are also presented to illustrate the effect of damping, external spring stiffness, and driving frequency. The optimal topology of the compliant mechanism is shown to be dependent on the driving frequency, the external spring stiffness, and whether the piezoelectric actuator element is considered design or nondesign. At high driving frequencies, it was found that the dynamically optimized structure is very near resonance.

Sign in / Sign up

Export Citation Format

Share Document