A well connected, locally-oriented and efficient multi-scale topology optimization (EMTO) strategy
نویسندگان
چکیده
Multi-scale topology optimization (a.k.a. micro-structural optimization, MTO) consists in optimizing macro-scale and micro-scale simultaneously. MTO could improve structural performance of products significantly. However, a few issues related to connectivity between micro-structures high computational cost have be addressed, without resulting loss performance. In this paper, new efficient multi-scale (EMTO) framework has been developed for purpose. Connectivity is addressed through adaptive transmission zones which limit A pre-computed database used speed up the computing. Design variables also chosen carefully include orientation enhance EMTO successfully tested on two-dimensional compliance problems. The results show significant improvements compared mono-scale methods (compliance value lower by 20% simplistic case or 4% more realistic case), demonstrate versatility EMTO.
منابع مشابه
A framework for efficient large scale equation-oriented flowsheet optimization
Despite the economic benefits of flowsheet optimization, many commercial tools suffer from long computational times, limited problem formulation flexibility and numerical instabilities. In this study, we address these challenges and present a framework for efficient large scale flowsheet optimization. This framework couples advanced process optimization formulations with state-of-the-art algori...
متن کاملEfficient Combination of Topology and Parameter Optimization
This paper presents a combination method of Particle Swarm Optimization (PSO) and topology optimization. With this method a better result can be achieved compared with the sequential application of these two optimization methods. It inherits the ability in finding global optimum from PSO and also suits for discretized design domain. Some special schemes are used in order to provide higher compu...
متن کاملEfficient Non-linear Optimization via Multi-scale Gradient Filtering
We present a method for accelerating the convergence of gradient-based nonlinear optimization algorithms. We start with the theory of the Sobolev gradient, which we analyze from a signal processing viewpoint. By varying the order of the Laplacian used in defining the Sobolev gradient, we can effectively filter the gradient and retain only components at certain scales. We use this idea to adapti...
متن کاملEFFICIENT SIMULATION FOR OPTIMIZATION OF TOPOLOGY, SHAPE AND SIZE OF MODULAR TRUSS STRUCTURES
The prevalent strategy in the topology optimization phase is to select a subset of members existing in an excessively connected truss, called Ground Structure, such that the overall weight or cost is minimized. Although finding a good topology significantly reduces the overall cost, excessive growth of the size of topology space combined with existence of varied types of design variables challe...
متن کاملLocally Connected Deep Learning Framework for Industrial-scale Recommender Systems
In this work, we propose a locally connected deep learning framework for recommender systems, which reduces the complexity of deep neural network (DNN) by two to three orders of magnitude. We further extend the framework using the idea of recently proposed Wide&Deep model. Experiments on industrial-scale datasets show that our methods could achieve good results with much shorter runtime.
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Structural and Multidisciplinary Optimization
سال: 2021
ISSN: ['1615-1488', '1615-147X']
DOI: https://doi.org/10.1007/s00158-021-03048-1