Researchers at the Fraunhofer Institute for Mechanics of Materials IWM in Freiburg have developed a simulation model that identifies potential weak points in forged components at an early stage. Such weak points are frequently caused by incomplete grain reformation, known as recrystallization. Integrated into a continuous digital process, this model forms the basis of a virtual forging lab that helps save material and energy in forging production.
The forging industry currently finds itself under considerable pressure: energy prices are rising, climate policy requirements are becoming stricter, and the shift toward sustainably produced "green steel" calls for a rethink of production processes. New manufacturing methods for green steel result in greater fluctuations in chemical composition and altered material properties, making forging behavior harder to predict. This is particularly critical for safety-relevant components, where a flawed microstructure can not only compromise quality but, in the worst case, lead to serious component failure. As a result, the industry finds itself caught between rising sustainability demands and consistently high quality standards.
Where Potential Remains Untapped
Turning a blank into a finished component involves several energy-intensive production steps: heating, forming, intermediate annealing, and re-forming. Each of these steps requires precise coordination of numerous parameters, including temperature control, degree of deformation, holding times, and the press forces applied. In practice, manufacturers have so far relied primarily on years of accumulated experience and costly trial-and-error test series. This approach reaches its limits, however, where fluctuating material quality introduces growing uncertainty. While efforts to replace physical testing with simulation are not new, they have often fallen short in the past due to a lack of models tailored to specific materials, insufficient reliable data, and the absence of practical ways to translate simulation results into concrete process instructions.
Why the Material Itself Is the Real Challenge
At the heart of the problem lies the behavior of steel during hot forming: its internal structure is far from static, undergoing continuous change throughout the process. Deformation initially hardens the material, while recrystallization simultaneously and continuously produces new, stress-free grains. As this new grain structure continues to grow, it can coarsen again — a process that can be slowed by extremely fine precipitate particles. The strength of these competing mechanisms depends heavily on temperature, degree of deformation, and the specific chemical composition of the steel.
Understanding and controlling this interplay of physical, chemical, and mechanical effects is what makes forging technology genuinely challenging — as is the ability to precisely link forging press settings to the development of the material's microscopic structure.
A Digital Lab for a Traditional Industry
This is where Fraunhofer IWM's new approach comes in: a practice-oriented digital process that makes the complex internal processes within the material computationally tractable. At its core is a physically grounded materials model specifically geared toward recrystallization. This so-called mean-field model combines two key strengths: because it is built on fundamental thermodynamic principles, it delivers reliable results even under the variable conditions of real-world forging processes. At the same time, its computational efficiency is designed so that even components weighing several tons can be modeled on a computer within a manageable timeframe. Because the model is based on real physical relationships rather than purely on empirical values, its predictions remain robust even with new alloys or changed process parameters.
The process starts with what is known as a material map: laboratory tests provide the key material parameters, such as flow behavior and grain size development under defined conditions. Building on this, a finite element simulation calculates, for every individual point in the component, the temperatures and forces it is exposed to over time. These results then feed into the mean-field model, which uses them to derive the development of the material's microstructure. The end result is a three-dimensional representation of the component showing exactly where, internally, which grain sizes can be expected. This makes it possible to compare and evaluate different manufacturing variants virtually — including concrete recommendations for avoiding defects.
Practical Test Confirms the Model's Performance
The method proved its practical value in the EU project "AID4GREENEST," whose goal is to facilitate the processing of sustainably produced steel, which naturally shows greater variation in chemical composition. As part of the project, the complete forging process of a 22-ton turbine shaft made of high-strength steel was fully simulated on a computer — a multi-hour process involving several forming and reheating steps.
"Our simulation correctly identified critical zones in advance," explains Dr. Maxim Zapara, Team Leader for Bulk Metal Forming at Fraunhofer IWM. "For the ends of the shaft, an undesirably coarse-grained structure was predicted, since the material in these areas was not sufficiently worked during the final, decisive forging step."
Notably, the model did not stop at predicting the weak points — it also identified their cause. At the core of the shaft, sufficient deformation during the final forging step led to a complete, fine-grained reformation of the structure. At the ends, however, this crucial "reset" did not take place. This computer-based prediction was subsequently confirmed through material analysis of the actual forged component.
Zapara adds: "This use case shows that the model not only reflects material behavior but can also uncover process errors before they occur in actual production. We can now test various approaches virtually: a simple adjustment to the final forging step in the simulation directly resulted in a completely fine-grained, defect-free component. In practice, this saves substantial amounts of material and energy."
What This Means for the Future
The digital forging lab marks a shift in focus: away from after-the-fact quality control and toward forward-looking planning of the entire process. Defects no longer need to be identified only in the finished component — they can be detected and avoided at the computer stage. Instead of months-long physical test series for new materials, hundreds of scenarios can now be tested virtually within a matter of hours.
A deeper understanding of how machine settings and material properties interact opens up further optimization potential: heating cycles can be shortened, and press forces reduced. More precise, near-net-shape production, in turn, lowers energy consumption, machine utilization, and the effort required for subsequent post-processing.
Looking further ahead, every forged component could eventually receive its own "digital passport" — a complete record of its internal structure that documents its quality and safety throughout its entire life cycle, from production to end use.
The research results were developed as part of the EU Horizon Europe project AID4GREENEST (Grant Agreement No. 101091912).
