Summary of the project "Digitalization of material development along the value chain (DiMaWert)"
18.03.2025
Process heat accounts for by far the largest share of energy consumption in the manufacturing industry. In Germany, this share is 68%. Today, three quarters of process heat is generated from fossil fuels, which leads to a very high CO2 footprint. On the other hand, climate change requires efficient strategies to reduce overall energy demand and CO2 emissions - not only in Germany, but worldwide.
Digital methods, which benefit from increasing computing power, the wide availability of artificial intelligence algorithms and large amounts of material data, offer numerous powerful tools for the development of heat processes with a reduced carbon footprint. This report describes such digital methods. They are embedded in a broader approach that combines experimental and computational tools and is based on Integrated Computational Materials Engineering (ICME). The methods were developed as part of the DiMaWert project, which was carried out at the Fraunhofer Center for High Temperature Lightweight Structures HTL in Bayreuth from May 2020 to April 2024.
Figure 1.1 shows an overview of the methods used at the HTL to optimize industrial heating processes. The so-called digital furnace twin is of crucial importance for this task. It is mostly used for the systematic optimization of loading schedules and temperature cycles. If required, the composition and flow of the furnace atmosphere can also be optimized. Significant improvements to existing heating processes have been achieved. Due to the strong interaction of the parameters, conventional optimization tools are often not very efficient for industrial heating processes. With the methodology shown in Fig. 1.1, energy consumption and CO2 emissions can be drastically reduced. In addition, throughput can be increased or flexibly adapted to production requirements. A special feature of our digital oven twin is the interaction of heat management within the useful volume with the respective behavior of the batch (see Fig. 1.1). The reaction kinetics and thermodynamics of the charge are coupled with the local heat and gas transport in the furnace so that gradients can be taken into account. This coupling significantly improves the informative value of the simulation. It enables a prediction of the product quality and a significant reduction in reject rates.
In order to achieve the required accuracy, the digital furnace twin uses numerous input data from the industrial furnace, the refractory materials and the batch. The furnace data can be supplemented by mobile furnace testing devices including an autonomous sensor module (ASM) (see Fig. 1.1). As part of the DiMaWert project, special sensors were developed for harsh operating environments, which are used to monitor the structural condition, for keyhole diagnostics and for temperature or gas analysis. The material data is obtained either from databases or from high-temperature measurements on the materials in question. The latter are generally more accurate and more cost-intensive. Since a lot of data is required for a realistic simulation, the number of high-temperature measurements must be limited. This is achieved through sensitivity analyses with the digital furnace twin. Sensitivity analyses make it possible to identify those material properties that have a significant influence on the results and that are worth measuring. A further reduction in the measurement effort is achieved through adequate - preferably physical - parameterization of the material data. This parameterization enables accurate interpolation and compensates for random errors between individual data points (see Fig. 1.1).
The material data of the load is obtained using special in-situ measurement methods. These so-called thermo-optical methods (TOM) were developed to measure the reaction kinetics and other important properties during heat treatment (see Fig. 1.1). Special laboratory furnaces were built to accurately reproduce the atmospheres of industrial furnaces. Sophisticated in-situ measurement methods and reaction models have been developed for debinding, sintering and melt infiltration processes. Other heat treatments such as drying and graphitization are being worked on in follow-up DiMaWert projects.
The benefits of the digital furnace twin are closely linked to the reliability of the simulations. We have therefore attached great importance to validating the models. The reaction and thermodynamic models were validated on the TOM furnaces. As part of the DiMaWert project, new test benches were built to enable validation for the three mechanisms of heat transfer: Conduction, Radiation and Convection (see Fig. 1.1). The digital furnace twin was validated with various batch and continuous furnaces.
Thanks to the generous funding from the Bavarian State Ministry of Economic Affairs, Regional Development and Energy (€7 million), the dedicated efforts of numerous colleagues and the high level of optimization methods for industrial heating processes already achieved, the HTL is ideally equipped to improve many thermal processes. Numerous follow-up projects have already been started together with industrial partners. A major advantage of our efficiency optimization methods is that they offer probably the fastest way to reduce production costs and the carbon footprint of energy-intensive industry without having to invest in the furnace system first. Therefore, we look forward to supporting the energy-intensive industry in strengthening its competitiveness on the global market through further applications of our methods. The digital kiln twin will also be used in future projects for more precise control of the heating power when operating the kilns and for the design of new types of kilns with a minimal CO2 footprint. In view of the fluctuating energy supply from renewable energies, it enables the implementation of Demand Side Management (DSM) in industrial heating processes without deteriorating the quality of the products.
In addition to the stove, the refractory materials also have a major influence on the CO2 footprint of heating processes. On the one hand, their production is very energy-intensive. Their service life should therefore be extended. On the other hand, their properties influence the energy consumption during their use - in particular the heat capacity, thermal conductivity and emissivity. As part of the DiMaWert project, methods were also developed to evaluate existing and new refractory materials in terms of service life and thermal material properties. Applications include more efficient thermal insulation, lightweight kiln furniture and flexible resistance heaters.
The technical content of this report is organized as follows: The simulation and validation methods for the digital kiln twin are described in Chapter 2. The reaction and thermodynamic models for the batch are presented in Chapter 3. Chapter 4 discusses numerous methods for providing high-temperature material data. Methods for the design of refractory materials are described in Chapter 5.
