The field of predictive maintenance is rapidly evolving, with more companies recognizing the benefits of identifying potential equipment failures before they happen. Traditional maintenance processes rely on regularly scheduled maintenance, which can be time-consuming, costly and inefficient. However, by using simulation and virtual twins of processes, companies can plan their maintenance regimes more accurately and efficiently.
Dassault Systèmes is one company contributing to the predictive maintenance field. As a leading provider of 3D design software, simulation applications and digital transformation solutions, Dassault Systèmes’ products have been crucial assets to the engineering and manufacturing sectors.
The company’s comprehensive suite of simulation software including CATIA and its 3DEXPERIENCE platform are enabling engineers to model complex systems, simulate operational performance and optimize designs for reliability and serviceability. While the CATIA and 3DEXPERIENCE software systems aren’t specifically designed for predictive maintenance, they are examples of available tools for design and engineering teams seeking a broad approach to simulation, potentially incorporating them in predictive maintenance strategies. Meanwhile, Dassault offers its SIMULIA and DELMIA brands targeting customers who need to identify maintenance problems before they materialize.
Understanding simulation for predictive maintenance
Virtual models of real-world processes allow companies to simulate different operating scenarios and predict how equipment will perform under various conditions. Digital twins enhance the virtual models by creating a digital replica of machines, providing more detailed and accurate representations of how the systems work in the real world. This higher level of detail is what yields a simulation powerful enough to evaluate performance and detect potential issues before they become reality.
Simulation for predictive maintenance is especially crucial for sectors where safety of machine operation is paramount, such as in the aerospace industry. Using digital twin simulations allows companies to assess various flight scenarios that put stress on an aircraft’s components, allowing them to proactively replace parts before they fail. Other companies that depend on reliability of machinery to maintain operations can use predictive maintenance to avoid unexpected downtime and the related financial losses.
Specialized software, such as platforms offered by Dassault Systèmes and others, are necessary for companies wanting to harness predictive maintenance to get a digital twin of equipment and/or processes created. The replica will then typically either be fed real-time process data or connect to machine sensors and IoT devices to put the digital twin into simulated action.
Tools for predictive maintenance
Finite Element Analysis (FEA) technology simulates the behavior of structures and components under various loads and conditions. FEA can help predict failures and identify areas of stress concentration that may lead to maintenance issues.
For companies that need help identifying potential issues with flow pressure, and temperature of fluids and gases that could cause maintenance problems, Computational Fluid Dynamics (CFD) simulates how substances will behave around equipment.
Electromagnetic Simulation technology simulates the behavior of electromagnetic fields and their interactions with equipment. It can help predict failures and identify potential sources of interference or other maintenance issues.
According to Dassault Systèmes EMEA Industry Process Consultant Manager Ivan Cabañas, simulation technologies allow maintenance engineers to model and analyze equipment behavior in a virtual environment before any physical maintenance is performed. This can help identify potential issues before they occur and allow for more efficient maintenance planning and execution. Additionally, simulation can help optimize maintenance schedules and improve equipment performance by identifying opportunities for optimization.
Ansys, another software developer offering commercial finite element analysis (FEA) programs, developed the system to be optimized for AMD Instinct accelerators, the newest data center GPUs from AMD. These accelerators provide exceptional performance for data centers and supercomputers to help solve the world's most complex maintenance problems.
To optimize for the AMD accelerators, Ansys developed APDL code in Ansys Mechanical to interface with AMD ROCm libraries on Linux, which support performance and scaling on the accelerators.
"Today's large, complex engineering challenges require quick, predictively accurate simulations that scale," said Brad McCredie, corporate vice president, Data Center and Accelerated Business Unit at AMD. "The collaboration between Ansys and AMD can enable a speed boost for some applications, enabling our joint customers to run complex structural simulations."
This collaboration resulted in a solution that Ansys' says significantly speeds up simulation of large structural mechanical models—between three and six times faster for Ansys Mechanical applications using the sparse direct solver.
“Ansys’ collaboration with AMD will help enable mutual customers to leverage cutting-edge GPU hardware for Ansys Mechanical applications in the data center, both on-premises and in the cloud to reduce time to market and deliver more optimal solutions," said Shane Emswiler, senior vice president of products at Ansys. “This work is well-aligned with our high-power computing strategy to invest deeply in GPUs as an emerging, sustainable, powerful technology for Ansys simulations.”
The nuts and bolts of simulation for predictive maintenance
Digital twin simulations can be valuable for all sorts of companies, especially those that rely on complex equipment and systems as part of their core operations. Manufacturers that use complex machinery to produce their products can benefit greatly from virtual twins of processes. These virtual models can be used to simulate the behavior of equipment under different conditions and identify potential issues before they occur. Energy and transportation companies also use the technology for greater efficiency and safety of equipment.
Companies that deploy predictive maintenance technologies can expect to see reduced downtime by catching critical maintenance issues earlier, reduced costs from minimizing the need for unnecessary or missed maintenance, and a lower risk of unexpected equipment failures.
“While it is true that larger manufacturing companies might be better suited to implementing these types of technologies in their value chain, smaller companies can also benefit from the numerous advantages of predictive maintenance through implementation of these solutions at a smaller scale,” said Cabañas. “What is most important here is the complexity of the product, the system, or the process that the company develops, and not so much the size of the company itself.”
Implementing simulation-based predictive maintenance requires data collection from equipment, an analytics platform for data analysis, accurate simulation tools and integration with existing maintenance systems. Companies must gather structured sensor data and other relevant information to feed analytics platforms for data processing and machine learning. They must employ simulation tools to model equipment performance under various scenarios.
Successful adoption entails integrating this process with current maintenance systems, developing workflows and training personnel. A cloud-based platform can facilitate these aspects while minimizing initial investments by leveraging cloud infrastructure instead of on-premises hardware.
To successfully adopt a simulation-based predictive maintenance process, companies should define clear objectives and strategies for their models, build a multidisciplinary team of data scientists, engineers and maintenance personnel, start with a small pilot project, and continuously monitor and refine the process. By outlining specific data, simulations and success metrics, creating a skilled team, testing the process in a pilot project and making ongoing adjustments based on performance metrics and feedback, companies can effectively implement the process and enjoy benefits such as reduced downtime, improved accuracy and cost savings.
This story is one in a series underwritten by AMD and produced independently by the editors of engineering.com. Subscribe here to receive informative infographics, handy fact sheets, technology recommendations and more in AMD’s data center insights newsletter.