Simulation is a standard practice at many design engineering firms, but many of those organizations still struggle to efficiently manage and reuse simulation data.
In some industries, companies may find themselves recreating simulations using data or versions of software that are several years old for compliance purposes. More commonly, companies want to revisit existing simulations to improve an iteration without having to rerun the simulation, or to create new derivative designs based on existing products and data.
In either case, they often struggle to use existing data to inform the next design or assembly without asking their analysts to spend valuable time reproducing simulations that have already been completed.
To overcome this issue, companies need tools that go beyond standard product data management (PDM) systems. “A design may have multiple configurations and changes, and just one variant may have hundreds of simulations applied to it,” says Matteo Nicolich, enterprise solutions product manager at ESTECO. “A PDM system cannot handle those simulations. That is why companies need a specific solution for managing simulation data.”
But many companies aren’t even trying to wrangle their simulation data at all. “Simulation is managed in silos that are not only related to the analysts themselves, but also to the different physics that are addressed by them,” says Dominique Lefebvre, product director at ESI Group. “There is a relatively weak link between PLM/PDM at large and simulation.”
Once a solution for a design has been validated, it is considered finished and everyone moves on, which is not necessarily optimal when it comes to leveraging all of the data generated to get to that point. The data is not optimally stored for reuse or searchability. “Most engineering groups don’t do much to formally manage data,” says Bruce Hart, business development director for Dassault Systèmes SIMULIA. “Often that data may be managed by an individual analyst, or even stored on a local disk or thumb drive. That not only isolates the data; it also isolates the simulation function in the company.”
More companies are turning to simulation data management (SDM) because they recognize the value of intellectual property reflected in those simulation models. That has led them to seek out solutions to help store the data, make it searchable and apply version control to it.
FMI (Future Market Insights) forecasts that the simulation and test data management market will reach $480.6 million in revenue by the end of 2026, with a compound annual growth rate of 12.5%.
“Model-based systems and digital twins and other use cases require a strong foundation for simulation data and process management,” says Sanjay Angadi, director of product management at ANSYS. “It’s really a foundational block for enabling the use cases that customers will get into in the future.”
Simulation Data Management Challenges
As more non-analysts are asked to take part in simulation, having a better handle on searching for and accessing existing simulations can help make running basic tests easier for engineers and other stakeholders—the data can be used to create templates and best practices, as well as automate some of those functions so it is easier for non-experts.
However, simulations generate large amounts of data that rely on a variety of tools and different analysts’ interpretations. “It’s an iterative process, and simulation may produce many models, with one that is useful for real results,” Hart says. “The inputs into that model can vary considerably, and analyst assumptions vary considerably. How do you store that?”
The simulation data is also underused, because only a fraction of the results are ultimately used to optimize a design.
Not being able to locate simulation data after the fact can result in wasted time looking for data, or time spent recreating models. “We were working with a medical equipment supplier that had to resurrect old content, and they literally had to go on a year-long review to pore through that data and cobble together the information that led to that design and those decisions,” Hart says. “This is frustrating for analysts as well as engineers.”
Another challenge is that simulation models may have been generated by various software tools from different vendors over lengthy periods of time. “It’s not only the digital information that needs stored, but you may also need specific hardware and software that were used for that specific simulation,” Nicolich says.
“As soon as there is an engineering or design change, you have to redo the first model, redo the second model and retest the physics,” Lefebvre says. “Keeping track of engineering changes is often not automated or a single process.”
Analysts also have a unique role in the organization, so there are cultural and organizational challenges as well. “There is also the sheer complexity of the simulation data and tools that you have to deal with,” Angadi says.
Emerging Simulation Data Management Solutions
Companies tend to go through an evolution when it comes to simulation management. Storing and organizing the data is usually just the first step. “These environments are just holding tanks for data, but don’t do much to help you get your hands on the models you want to look for,” Hart says. “What’s missing is the analyst input into the data. You have to build analyst confidence that the model has been built with certain assumptions and techniques to developing the physics, so they can trust the data that is available.”
Simulation management solutions can help formally manage that data, make it searchable and traceable, and provide insight into every aspect of simulation—geometries, meshes/models, input/output parameters, test conditions, supporting material, the tools used and the results.
According to Lefebvre, better leveraging simulation data can help improve data analytics and design space exploration activities as well, but only if the entire process of creating, storing and managing the simulation data has been automated.
“You have to manage not just the data, but the process,” Lefebvre says. “You check a given model against different physics, and it is difficult to keep track of all of these simulations related to the same product.”
In the case of ESI Group, its Visual DSS tool helps manage processes and models, including assigning tasks and better managing how design changes are communicated to different stakeholders. The company’s Vdot project management tool also provides SDM.
A variety of other types of tools are also available that can help companies get a better handle on simulation results. ANSYS Enterprise Knowledge Manager is a web-enabled solution for sharing and managing simulation data and creating workflows.
“When simulations are uploaded, the data is checked into our data management system,” Angadi says. “We extract metadata from those files, which then becomes searchable anywhere in the organization.”
Analysts can also create simulation workflows to establish standard processes that help improve management of the data and collaboration.
MSC Software’s SimManager manages large volumes of complex simulation data by automatically capturing information and allowing engineers to manage collaborations based on that data. PDTec’s SimData Manager tracks CAD geometries and simulation models to make it easier to find the correct analysis results, while also updating analysts about CAD model changes. Altair offers the Simulation Manager web-based portal that allows access to simulation data and provides traceability.
ESTECO’s VOLTA simulation data management offering is a browser-based system that organizes and accesses analysis models and results.
“They can keep the definition for the model and simulation, and the key results information,” Nicolich says. “They can quickly use the key results for retrieving information about how the results were obtained.”
The ESTECO tool can work with simulation tools from multiple vendors. “We manage and store the information about these processes, and they can be reproduced as many times as you want very quickly,” Nicolich says. “That provides a direct benefit in democratizing simulation. The system can recognize a complex simulation chain without having the analyst regenerate the model from the past. You can store the properties and then reproduce the results.”
Dassault’s SIMULIA Simulation Lifecycle Management is integrated with the company’s design platform. “In the 3DEXPERIENCE platform we capture that data automatically, so we have information about codes being run, what data was run, etc.,” Hart says. “The next layer is to capture some information about how the model was created, so the next person downstream can make sense of the data and reuse it.”
An analysis by CIMdata (sponsored by Dassault) found that companies could reduce costs by up to 90% by reducing the effort required for data acquisition, model creation, solving, reporting and other tasks. Analysis model reuse could potentially reduce costs by as much as 60%, and companies could further cut costs through reducing the need for prototypes and tests and reducing rework.
As simulation becomes more democratized and is used earlier in the design process, SDM will be even more critical.