MCADCafe Industry Predictions for 2024 – Comsol

By Ed Fontes, Chief Technology Officer, COMSOL

Trends in Computer-Aided Engineering: A Multiphysics Perspective

Ed Fontes

In the last decades, modeling and simulation (M&S) has become essential for understanding, predicting, and optimizing the behavior and design of processes and devices. Moore’s law, new numerical methods, and new software have facilitated this development.

M&S was once exclusive to R&D departments at large companies, universities, and governmental labs but has since spread to the design departments of both large and small organizations. The use of models has also changed. For decades, models were used in the exploratory and development stages of a project. Today, models can be incorporated into devices and processes to continuously adapt to, extract information from, and control them during operation. For example, models can be incorporated into digital twins in order to determine when a wing should be replaced in a fighter jet or when a wind turbine needs service.

Due to this expanding scope of use, the expectations for computer-aided engineering (CAE) software have grown together with the demand for M&S. Although CAE software has become more user friendly, most products still require some degree of knowledge about mathematical and numerical modeling. Ease of use — in terms of assisting engineers and scientists with modeling issues such as consistent initial conditions, boundary conditions, loads, and constraints — is an important aspect of CAE software design. Additionally, making user interfaces more self-guiding is something that most software developers are spending considerable resources on, partly driven by increased demand but also by the scarcity of M&S specialists. In this context, the use of bots may become more important to guide users in the general modeling process and give instructions for specific modeling steps.

However, these efforts are probably not enough to meet the large demand for M&S. An efficient alternative way to meet this demand is to provide tools that make it easy for modeling specialists to create simulation apps that can be used by a larger group of engineers and scientists, such as product specialists. These simulations apps are based on validated numerical models but provide user interfaces tailored for specific purposes. Since product specialists often have a deeper knowledge and understanding of the device or process being investigated, making apps available to them makes the use of simulations more efficient. In addition, using simulation apps means that running simulations does not become the bottleneck in R&D, design, and maintenance projects, which also frees time for modeling and simulation specialists. For example, by running a simulation app incorporated into a digital twin, maintenance engineers can see when a fighter jet wing needs to be replaced without having to consult a modeling specialist.

The extended use of M&S also increases the requirements of solution speed and portability. New methods for lumped models, model order reduction, and surrogate models are able to generate almost instantaneous results while the models remain compact and portable. Numerical methods such as proper orthogonal decomposition, reduced basis methods, as well as advanced function approximations like DNNs and Gaussian processes make models fast and compact enough to be incorporated into system models, uncertainty quantification analysis, simulation apps, and digital twins.

At COMSOL, our mission is to provide easy-to-use modeling and simulation software. Using partial differential equations (in mathematical models) and solving them (during simulations) is still the best way of understanding physics phenomena in engineering and science. This means that we have to efficiently formulate and solve the right equations for coupled physics phenomena, i.e., multiphysics, in order to provide highly accurate simulations of real-world applications. The need to implement new descriptions of physics phenomena, numerical methods, optimization algorithms, numerical solvers, and other basic concepts will not go away. In a way, we could call this need an everlasting trend.

COMSOL is a registered trademark of COMSOL AB.

About Author:

Ed Fontes is the chief technology officer at COMSOL. He has been with COMSOL since 1999, and was previously the lead developer for the CFD, heat transfer, and chemical engineering products. He received his PhD in chemical engineering from the Royal Institute of Technology, Stockholm.

Category: Predictions

This entry was posted on Thursday, January 25th, 2024 at 1:40 pm. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.