COMSOL Day: Optimization
See what is possible when multiphysics simulation is combined with optimization
The full potential of multiphysics simulation goes beyond using single calculations to verify prototype specifications: Innovative engineering organizations use simulation to optimize their designs for performance and efficiency. For example, topology optimization can be used generate novel geometry early in the design phase, or shape optimization can be used to refine an existing design in a systematic way, considering several situations and objectives in every iteration. In parallel to this, parameter estimation might be combined with experimental data to ensure the accuracy of material parameters for realistic model predictions. The process of optimization can be complex, involving nonlinear phenomena with often surprising results. State-of-the-art optimization technology provide easy solutions to common engineering challenges, but great flexibility also allows for solving specialized problems in fields where the systematic use of optimization is less established.
COMSOL Day: Optimization will feature two keynote presentations by experienced COMSOL® users and includes a full program describing optimization in COMSOL Multiphysics®, with a focus on the features of the Optimization Module. Join us for this free, 1-day online event to learn how to bring the power of optimization to your multiphysics simulations.
To start, we will briefly discuss the format of the day and go over the logistics for using GoToWebinar.
Design engineers are using the optimization methods available in the COMSOL® software for much more than just mechanical analysis; the software’s unique capabilities are being used to optimize devices and processes that involve a wide range of physics, including acoustics, electromagnetics, fluid flow, heat transfer, and more.
In this session, we will discuss the latest optimization technology that is driving this trend as well as the wide variety of applications for its use. We will introduce the different types of optimization available in the COMSOL® software and give an overview of new optimization functionality in version 6.1, including:
- Milling constraints for topology optimization
- Gradient-based optimization of eigenvalue problems
- Improved preservation of continuity in shape optimization
DENSO, a major parts supplier for the automotive industry, develops and manufactures electric motors in house. In this keynote talk, Sukhwa Jung will discuss how DENSO is using COMSOL Multiphysics® to conduct topology and shape optimizations of permanent magnet electric motors. Jungh will explain DENSO's motivation for employing optimization, present the results of the shape and topology optimizations, and compare them to results from academic literature.
Extracting material and model parameters from experimental data is an important step when modeling and simulating, for example, batteries, non-Newtonian flow, and nonlinear structural materials. COMSOL Multiphysics® includes dedicated functionality for parameter optimization, parameter estimation, and validation for models that include arbitrary, coupled physics phenomena, or multiphysics.
In this session, we will demonstrate parameter estimation and parameter optimization and how they can be used with a wide range of combinations of physics phenomena using different solvers. We will demonstrate how to:
- Estimate and validate material and model parameters from experimental data
- Optimize geometric parameters with respect to arbitrary objective functions
Shape optimization can be used to deform the boundaries of an existing design. It makes it possible to start with a general outline of an expected design — the algorithm will adjust the position, orientation, and shape of each boundary. In COMSOL Multiphysics®, this optimization method is compatible with any type of physics and combination of physics phenomena, that is, multiphysics.
In this session we will demonstrate:
- Freeform shape optimization
- Optimization by the simple translation of objects
- Stress and fatigue optimization
Learn the fundamental workflow of COMSOL Multiphysics®. This introductory demonstration will show you all of the key modeling steps, including geometry creation, setting up physics, meshing, solving, and postprocessing.
Nanophotonic devices have found use in a wide range of applications, from data transfer and processing to sensing, with more potential applications appearing continuously. One type of nanophotonic device is the optical cavity, which can store light both spatially and temporally and is critical for a range of applications relying on a strong interaction between photons and electrons.
In this keynote talk, Rasmus Ellebæk Christiansen of DTU will present topology optimization strategies for designing photonic devices. In particular, he will present the recent study in which he and his colleagues applied inverse design with fabrication constrains to design an optimized optical cavity capable of simultaneously strongly confining light spatially and temporally. The design optimization was followed by the fabrication and experimental validation of the device. In the study, published on nature.com, record-breaking performance was observed experimentally, owing in-part to the inclusion of manufacturing constraints in the optimization methodology.
Topology optimization gives extreme design freedom where any shape within the design space is allowed. Nowadays, topology-optimized designs are more easily realized thanks to additive manufacturing methods. In COMSOL Multiphysics®, this optimization method is compatible with any type of physics phenomena as well as arbitrary multiphysics combinations. In this session, we will show you how to get started with topology optimization and will cover:
- Topology optimization with manufacturing constraints
- The verification of body-fitted meshes
The Uncertainty Quantification Module uses specialized solver studies and postprocessing tools to expand your models and produce more encompassing, accurate, and useful versions of your model. Through applying probabilistic design, this enables you to look at questions such as how manufacturing tolerances affect the intended performance of the final product and how to avoid preventing over- and underdesigns of devices and processes. In this session, we will investigate the new Uncertainty Module’s application to:
- Screening and sensitivity analyses
- Uncertainty propagation and reliability analyses
- Use within any kind of physics simulation, such as structural, chemical, acoustics, fluid flow, and electromagnetics applications