Numerical simulation is a proven tool that has been used for the design of instrumentation systems for many years. Today, numerical simulation addresses all components involved in a process, with use cases spanning pre-production, production, and diagnostics. The use of numerical simulation in physical and virtual environments—like digital twins—has made model optimization and interoperability major research objectives at CEA-List.
The use of numerical simulation is becoming more and more common across all industrial process steps and at all scales.
It is an effective tool for designing and optimizing instrumentation systems and ensuring that they perform as intended. It can also provide data to feed predictive and adaptive systems and, increasingly, machine learning and AI algorithms for diagnostics.
CEA-List is currently focusing on expanding the capabilities of numerical simulation models. The institute is developing advanced measurement and inspection techniques and updating existing models to respond to the requirements of new industrial use cases. These include real-time and near-sensor simulation, digital twins of measurement systems, and solutions to facilitate AI training for data processing applications.
Finally, CEA-List is working on model interoperability to support the development of integrated simulation chains, another decisive factor in the expansion of numerical simulation.
We are developing models specifically for new instrumentation techniques like the measurement of elastic waves in optical fibers, induction infrared thermography, and laser vibrometry, so that these techniques can be integrated into simulation tools. These new simulation capabilities will also make it easier for industrial users to begin implementing the techniques in their processes.
We are also developing transcription rules, or reference codes, for techniques that are becoming increasingly important like ultrasound, the interaction of ionizing radiation with matter, electromagnetic and thermal measurements.
Digital twins of instrumentation systems will depend heavily on simulation. To clone a system, you have to be able to reproduce the data captured in the physical world as faithfully as possible in the virtual world. Simulation can even generate data that cannot be collected in the physical world—and that that can be used to predict and optimize performance. Digital twins address all of these issues, providing input data for models and using the data produced by models to design, optimize, predict the performance of, and manage instrumentation systems.
Model interoperability is important because it will facilitate the development of numerical measurement chains that eliminate the risk of information gaps when controlling an entire manufacturing process, for example.
Giving models the ability to interact with each other and change over time also opens the door to metamodels, which can be used to very rapidly develop context-specific solutions.
CEA-List’s CIVA software is used by more than 300 companies in more than 40 countries, making it the world’s leading simulation solution for non-destructive testing (NDT). The software leverages research conducted in CEA-List labs and with academic research partners.
Learn more about CIVA software
PYRAMID, a Franco-Japanese research partnership financed by ANR, France’s national research agency, focused on assessing the extent of wall thinning in concealed piping in order to devise a strategy for pumping the water out of the Fukushima reactor. CEA-List used CIVA to model electromagnetic acoustic transducers suitable for the remote inspection of the piping.
First, the researchers made some adaptations to CEA-List’s existing simulation chain for guided wave ultrasonic inspection of pipework with complex geometries, factoring in the bends characteristic of a nuclear power plant’s piping. The presence of water in the piping was also introduced into the simulation—challenging from a technical standpoint, but necessary due to the significant impact of the water on wave propagation.
Europe :
National :
At CEA-List, we use our robust mathematical models and knowledge of physics to rapidly produce data for use cases where data may be lacking.
