The material workpieces are made of
When materials scientist Andreas Rucki turns the pump on, pipes start to blare behind the wall covering. This soon changes to a quiet rattle, until finally all that’s left is a gentle murmur. The result of this short-lived wind concert? There is now a high vacuum in the scanning electron microscope (SEM) in the middle of the cellar. The metal sample fixed there can now be closely examined. Or more precisely, put under an electron beam that samples the material in the sub-micron range at high resolution, in order to gain information for a digital twin of the material.
Material properties as a lever for precision
“Our customers use a complex sequence of manufacturing processes to turn raw materials into products with high added value,” says Zvi Feuer, head of Manufacturing Engineering Software in Siemens Digital Industries. “Each production step impregnates a process signature, which influences the final properties and subsequently the performance of the product. Our vision for the digital twin of the manufacturing process is therefore to provide our customerswith the relationship between process and material properties in our digital twin, both to develop innovations and to achieve a faster time-to-market. That means the concept of the digital material is a key component of our vision for digital manufacture.”
Digital twins are virtual representations of the properties and functions of components, machines, and even entire production plants. But one important aspect is usually missing: A differentiated digital representation of the material. The properties of the material are extremely important, however, since it is ultimately the material properties – in addition to changes in geometry – that determine whether a component will perform its function, and when it may need to be replaced. More and more frequently, many applications, such as robots, the power industry, vehicles, aeronautical engineering, machine tools, or mobile devices, demand innovative materials that meet sophisticated performance profiles.
Plan more accurately, manufacture more quickly, improve maintenance
A 3D model including detailed material properties capable of representing shape behavior, stability, temperature distribution, and internal stresses helps in this regard. A digital twin like this also makes it possible to simulate the condition of a machine with greater precision. That would make predictive maintenance more accurate and, if necessary, identify the causes of problems more quickly.
A better understanding of the materials means products can be simulated with greater accuracy, making production faster and improving maintenance. This helps in the case of small product runs, when the smallest possible number of prototypes must be made. It can help with monitoring of mass production. Additive manufacturing processes also benefit from a material-smart twin, since printed metals and metals made using traditional casting processes perform differently. And lastly, it improves machining – milling, turning or drilling – if, for example, the impact of the cutting edge of a milling head on the internal stresses of a workpiece must be estimated to optimize production planning.
How Siemens intends to achieve this objective
Rucki shows images from the microscope – grayscale satellite images of fine-grained landscapes with strange forms composed of particles, needles, cracks, and grooves. “Even if it looks good with the naked eye, how a material performs will often be determined at the grain boundaries you can see here. We use these details, in addition to results obtained using other methods like chemical element analysis, to characterize the material in three dimensions with maximum precision. Together with the test results from mechanical and chemical testing, we can use these to create a digital twin of the material and its properties with maximum positional accuracy.”
“For a digital twin that contains detailed information on the material properties for all stages of its lifecycle, we need a lot of data, which we gather in the lab, in factories, and from industry literature,” says materials researcher Ulrich Bast, one of Rucki’s colleagues. “But that isn’t enough: We also have to explore the dynamic laws of the material – to predict, for example, what happens if it is shaped or exposed to extreme environments. The goal is the ‘transparent product.’”
That’s why Siemens is currently developing a database that stores all this data in a structured form so it can be incorporated into existing twins. Rucki and Bast’s team is not only developing refined models that predict material behavior under all kinds of conditions, but also models that describe how to achieve particular material features.
Omar Fergani, digital manufacturing expert at Siemens Digital Industries, is very interested in these activities: “The strength of the digital twin lies in our capability to cover our customers’ complete value chain, from the idea to product performance in operation. The twin with high fidelity containing details just like a materials genome is based on the work of our researchers.”
Digital twins, expanded to include material data, could also be integrated in the Siemens Digital Industries portfolio, as a means of manufacturing products with minimum use of materials and running production facilities more efficiently. This also includes traceability of the material, for example to ensure that no raw materials from conflict areas are used. Digital twins link what used to be separate parts of the value chain even closer together, to produce an even more accurate simulation of workpieces, tools, or entire production plants. And if the materials are turned into many lines of code in the process, their origin will remain tangible: Microanalyses based on patience and refinement that draw on every possible tool to get to the heart of the tiniest samples.
Author: Hubertus Breuer
Subscribe to our Newsletter
Stay up to date at all times: everything you need to know about electrification, automation, and digitalization.