The Mechanisms Engineering Test Loop (METL) facility at Argonne National Laboratory plays a crucial role in advancing liquid metal technologies for sodium fast reactors. To support the facility’s operations and maintenance activities, Implexus, a digital twin consultant company, has developed a cutting-edge digital twin. This case study explores how the digital twin aids the laboratory in monitoring the test loop by providing real-time temperature visualization and analysis.
Why a Digital Twin?
According to a recent article published by the laboratory, Argonne’s METL facility is “the nation’s largest liquid metal test facility where small and medium-sized components are tested for use in advanced, sodium-cooled nuclear reactors. The facility holds 750 gallons of reactor-grade sodium that can be heated to 650 degrees Celsius and it is equipped with more than 1,000 sensors that collect diagnostic data.” The experiments conducted at METL contribute significantly to the development of advanced reactors. Derek Kultgen, Operations Manager of Argonne’s METL, connected with Implexus to develop a digital twin to pair the real time sensor data to a virtual replica of the laboratory. Live visualization of the data across the loop using a color gradient enables the METL team to accomplish their goal of more effective monitoring of the loop’s operations.
How It Works
The digital twin enables 3D visualization of both absolute and relative live temperature readings. Tracking and displaying temperatures in a digital twin provides greater clarity to operators when working remotely. The digital twin was created in Unreal Engine (UE). As a gaming engine, Unreal provides Implexus with the capacity, speed, and graphic capabilities needed to produce a high rendering level of quality (LOQ) with live data display. By articulating the space at a high LOQ, the end-user can experience the test loop in a manner similar to real space. In fact, by modeling the individual piping connections, the user has a better experience with the physical facility because they are able to turn off insulation, structure and other support layers that shield the view in the laboratory. UE plays centralized host to a variety of file types used to create the digital twin. Content from Autodesk Inventor, Autodesk Revit, Blender, and Twinmotion were compiled into UE to create the context of the twin.
Late 2022 Implexus was connected to a team developing a data visualization tool for Unreal Engine called Datasculpt. At the time, the 3D model of the test loop had been produced, and it was time to begin to connect the sensor data. There are a variety of ways to accomplish this, many of which require extensive knowledge of UE’s blueprint visual scripting system. The Implexus team members who worked on this project have backgrounds in computational design through several visual scripting tools. Visual scripting in any program has quite a similar look and feel, but in order to produce quality blueprints in Unreal, one must first learn the vocabulary of the program. However, because of the extensive community around this gaming engine, many plugins and tools are continuously being developed that remove some of this leg work for users. For example, the Datasculpt team pre-codes the functionality into their plugin and provides a simple to use interface to create visualizations. This digital twin consists of three levels. One level shows no data and instead simply provides a descriptive twin of the space color coded by element type. The second and third levels, accessed by buttons for the user to interface with, display heatmaps in absolute and relative temperatures, respectively.
At Argonne National Lab, the sensor data is managed through METL’s Supervisory Control and Data Acquisition system. All of ANL’s data is kept in-house where it is most secure, so for this project the data was accessed through HTTP requests. The twin simply reads the data output and displays the appropriate visual per the user’s interactivity. What Datasculpt offered Implexus on this project was dual-purpose. Not only does the tool offer data visualization without having to fully script in blueprints, but data management also occurs inside the plugin. Once data was accessed through parsing JSON HTTP requests, over two hundred heatmaps were generated in Datasculpt connecting each data point in the array to the partner 3D element in the UE model. More specifically each heat map contained a starting sensor datapoint, an ending sensor datapoint, and a UE color curve. Two color curves were generated to display a range of temperatures across a spectrum of minimum and maximum readings in degrees Celsius; one range for absolute and one for relative. By referencing a single-color curve per level, users can view the entire test loop holistically, clearly displaying how temperature varies across the loop. This visual helps the laboratory quickly determine if there is an outlier in one segment of the loop. The Argonne team is then able to select any single segment and display its individual temperature reading to dive deeper into a potential issue. The way these visuals were produced allows the team to quickly detect deviations from optimal operating conditions, identify hotspots or temperature anomalies, and track thermal behavior of the system over time. This empowers the METL team to proactively address potential issues, optimize system performance, and prevent failures before they occur.
The Future
Before the creation of this digital twin, Argonne’s team was able to view the data on 2-dimensional dashboards. At a baseline, by replicating the data visual in 3D, the team has a more familiar experience while interacting with the data. Accessible remotely, the twin creates flexibility for the team to instantly view the performance of the lab. In addition, the capability to view historic data begins to set a foundation for huge potential going forward. METL enables the testing of a variety of sensors to monitor plant conditions. They explore sensors that can operate while immersed in primary coolant, including sensors for rapid hydrogen detection, impurity detection, improved leak detection, and enhanced level measurement. As it evolves, the digital twin can be connected to all this sensor data providing a comprehensive look at the performance of the laboratory. As digital twins become more commonplace and Implexus’ work continues with the national laboratory, our vision for the future includes the ability to analyze data collected over time. There is power in not only viewing comprehensive live performance, but in data analysis of trends or patterns in the test loop that may provide deeper insights for the research. From the perspective of Argonne’s Derek Kultgen, the capacity and benefit of viewing virtual sensors in digital twins is only going to take off from here.
Conclusion
Through real time temperature visualization, this digital twin provides valuable insights into METL’s test loop thermal behavior. This empowers the team to make informed decisions, optimize operations, and ensure the safe and efficient functioning of the test loop. The case study demonstrates the immense value of digital twin technology in advancing liquid metal technologies for sodium fast reactors and showcases a model for how digital twins could benefit other laboratory research facilities.
Resources:
Comments