Release includes the collaborative Model Manager and the risk assessing Uncertainty Quantification Module.
COMSOL Multiphysics has shifted its focus toward improving the collaboration and engineering capabilities of its simulation software with the recent release of version 6.0. The big news in this release is the addition of a file management and collaboration tool called the Model Manager and a risk assessment tool called the Uncertainty Quantification Module.
Phil Kinnane, senior vice president of Sales at COMSOL, said, “It’s going to allow you to collaborate with your colleagues such that you can even work on the same model. With these two tools, COMSOL will expand its usefulness even further in the CAE market.”
Other improvements to COMSOL 6.0 include additional add-ons, improved performance with 10x boosts in speed and memory consumption, and more efficient electromagnetic simulations for printed circuit board (PCB) design.
The Path to COMSOL 6.0
The big push behind the previous 5.X series of COMSOL releases focused on the democratization of simulation technology. This appification directive had a collaborative feel to it, in keeping with the 6.0 release, as it improved communications between simulation experts and their colleagues in and out of their organizations. Traditionally, the time these experts have is limited. However, there is no shortage of customers, coworkers and regulations that require consistent simulation results.
By “appifying” frequently run simulations, others could plug in the data they need tested and gain approved engineering results without ever having to contact the expert. For instance, if Sales gets a unique customer request, there would be no need to contact the simulation expert to test if the request is feasible or not. They could just plug the data into the app and get a preapproved response.
COMSOL has spent the last few years improving its app infrastructure, making it easier for experts to launch, share, monetize, collaborate and communicate with these apps. Kinnane explained that the big news of COMSOL 6.0, the Model Manager, is a natural progression of this shared communication. It’s a workspace that ensures the data management, version control and tracking of simulation files.
“The change from 5.6 to 6.0 is that we’re extending our engineering capabilities and movement into the market by increasing our usefulness to larger groups and teams both within and external to organizations,” Kinnane said. “The Model Manager is the collaborative part.”
Meanwhile, the other big addition in COMSOL 6.0, the Uncertainty Quantification Module, continues the company’s path towards expanding its role in the CAE market. This module extends the capabilities of COMSOL from deterministic risk analysis to the ability to apply probabilistic methods. In other words, it makes it easier for engineers to quantify uncertainty and determine safety margins.
Keep Track of Your COMSOL Data with the Model Manager
COMSOL isn’t the type of software company to come out with a new X.0 release every year. As a result, when it does issue a new release, there tends to be a big shift. In this case, it’s the Model Manager.
Kinnane remarks on the importance of this tool by saying, “The Model Manager allows you to collaborate with all of your colleagues within your organizations, industry, school or whatever grouping you put together. Instead of everyone working on separate models, projects and situations, where the only way you can share models and simulation data between each other, is through something as basic as email or trying to build a network system full of folders and sometimes using another software to control it, now there is a user interface built to control such collaborative projects within COMSOL Multiphysics.”
In other words, the Model Manager is designed to perform data management, version control and track changes. It even has advanced search capabilities. The idea is that large engineering teams can utilize the software to improve collaboration between in-house and external parties.
Using a floating network license, users can access the Model Manager from anywhere. Meanwhile, a local installation is available to help individuals manage their own modeling files.
Currently, the tool is slanted toward COMSOL files. But it can also store input data files that go into a COMSOL model. For instance, you can change a CAD file in COMSOL and that could be updated by the tool to ensure external components are using the most recent data within the design project.
“It’s all about version control,” Kinnane said. “You worked on the model using your expertise, then I came in and used mine, which improved and updated the model. Now that updated model is going to be used by four others down the hallway and they will get the latest and greatest through version control. This will cut down time to market, resources, costs and more.”
The Model Manager is rather stand-alone at this point, however that likely won’t last for long given the workflow improvement capabilities it could have connecting to third party tools—like PLM systems. As for what sets the Model Manager apart from the competition, according to COMSOL, it’s in-house compatibility.
“It’s coming from in-house,” Kinnane explained. “It suits our underlying file and workflow structure. The familiarity, the user interface—all these things are underpinning the other platform tools built into COMSOL. That will take us much further. It’s not just being tacked on, whatever new features the software has is covered by us. It’s going to be built around our structure instead of hardwiring or tacking on something that was originally built to be a plug-in.”
At the end of the day, the advantage COMSOL has is that it is developing this file management system from the ground up to work with COMSOL, and that it’s intrinsic to the software development as opposed to being a combination of once disparate tools.
Assess Your Simulation Risks with the Uncertainty Quantification Module
The Uncertainty Quantification Module helps engineers improve the accuracy, usefulness and completeness of their simulation models. It utilizes probabilistic design methods to help users assess how current unknowns could affect the performance of a final product. For instance, engineers can ask how manufacturing tolerances can affect the end product in efforts to prevent over- and under-designing the system.
The tool uses screening and sensitivity analysis to reveal which parameters are most important to the final design. This first assessment helps engineers validate assumptions and assign probability distributions to output quantities.
“Another way of looking at it is possible risk. It’s a quantification of how much uncertainty and risk you are going to take on as you simulate,” Kinnane explained. “Because of all the variations and possibilities in the world, to encapsulate them all with equations and physics would be impossible. Luckily, the majority are so insignificant that you don’t need to consider them. Yet with uncertainty quantification, you can identify critical parameters that will actually contribute to such risk.”
He added, “the Uncertainty Quantification Model is about how much we are prepared to ensure ourselves that these small insignificancies won’t build up on each other to make something less workable than we originally thought. Or are we going to come ahead, and what we thought would be a component that would survive a certain amount of time turns out to be usable for twice as long, because the probability of this is acceptable enough to be included in the overall decision-making. The module is able to put that into quantifiable terms.”
The Uncertainty Quantification Module is based on a probabilistic design as opposed to a worst-case scenario or deterministic design philosophies. If the design falls within the acceptance interval, it is considered good to go. The more computations you make, the more certainty you have in your inputs, and the narrower you can make that range of success that your products can fit into. Kinnane said, “So, rather than using gut feeling or judgment, you can actually say, here is a mathematical value that says it’s good to go based on the range of the probabilistic design.”
Jacob Yström, technology director of numerical analysis at COMSOL, said, “A strength of the Uncertainty Quantification Module is that it can be applied to any physical simulation covered by COMSOL Multiphysics. You are not limited to a certain field or application area, such as structural analysis, but can perform the same types of uncertainty analyses on applications based on acoustics, fluid flow, electromagnetics, and so on, and even when these phenomena are coupled. This makes this product wide-ranging and very powerful.”
For more on the appification of the simulation world, read COMSOL: Engineering Apps Make Simulation Tech an Advantage for All.