André Hürkamp

Traditional tool development processes in mechanical forming, especially for pressing tools, are often limited by time‐consuming trial‐and‐error cycles, which rely heavily on expert knowledge. These cycles can create bottlenecks and increase costs. Machine learning can be a useful tool to greatly speed up these processes by using advanced models. In this regard, previous research proposed and employed a denoising diffusion model (DDM) to inversely model effective tool surfaces from final product geometries. This model leverages spatial and temporal attention mechanisms to enhance its predictive capabilities. While DDM achieved commendable overall predictions, the accuracy in reconstructing individual denoised frames during later forming steps was lacking. To address this, the present research adopts a stepwise strategy focused on the late deformation state in forming simulations. The research utilizes pixel‐adaptive convolutional neural networks (PAC) to predict high‐resolution simulation output of various channels incorporated in DDM from coarse‐mesh results, using the final part geometry as high‐resolution guidance. This approach enables efficient and reliable estimation of these high‐resolution distributions for the late deformation stage, reducing reliance on computationally intensive full‐resolution simulations at this critical point. This work demonstrates the effectiveness of PAC for simulation refinement in late‐stage deformation and lays the foundation for future integration of PAC into the DDM framework, thereby contributing to ongoing efforts toward more efficient, data‐driven approaches in industrial tool design and manufacturing.

Especially in the automotive and aerospace sector, lightweight materials have an important role in reducing weight and enhancing overall energy efficiency and performance. Due to the combination of their high mechanical properties with low densities, fiber-reinforced thermoplastics (FRTP) offer a high lightweight potential. For automotive applications, FRTPs are usually formed into shell-like components in the thermoforming process. Here a reliable process design relies on numerical models to capture the complex forming mechanisms accurately. However, most conventional solvers provide no accurate representation of the bending behavior of FRTPs. This limitation often results in inaccurate numerical results. Therefore, this paper introduces and systematically evaluates different modeling strategies for implementation of the bending response. Both continuum and multi-ply approaches are considered under isothermal and thermomechanical conditions. Two representative FRTP systems with polyamide 6 (PA6) matrices are investigated. A glass fiber-reinforced woven fabric and a unidirectional (UD) tape. The materials are comprehensively characterized with respect to their mechanical and thermal properties. Based on this data, material models are calibrated and implemented into numerical forming simulations of a highly double-curved double dome geometry. The forming behavior of the numerical model is evaluated against experimental results with a particular focus on the shear angle distribution. The study highlights the influence of the chosen bending formulation on numerical accuracy and provides guidance for improved thermomechanical forming simulations of FRTPs.

The construction industry is facing a dual challenge: an increasing demand for new buildings on the one hand and the urgent need to drastically reduce emissions and waste on the other. One promising field of research to face these challenges comprises additive manufacturing (AM) technologies. Through these advanced methods, digital workflows between design and fabrication can be implemented to optimise the form and structure, unlocking new architectural freedom while ensuring sustainability and efficiency. However, to drive this transformation in construction, the new technologies must be investigated in large-scale applications. One of these fast-emerging AM techniques is Shotcrete 3D Printing (SC3DP). The present research documents the 1:1 scale manufacturing process, from digital to real, of a building section utilising SC3DP. A workflow and production steps, spanning from design over manufacturing to assembly, are introduced. The architectural design, reinforced by computational methods, was iteratively refined to adapt to manufacturing constraints. The paper also emphasises the importance of a digital twin in ensuring seamless data integration and real-time adjustments during construction. By incorporating reinforcement techniques such as short rebar insertion and robotic fibre winding, this study demonstrates the structural capabilities achievable with SC3DP. In summary, the implementation of comprehensive digital workflows utilising computational design, automated data acquisition and data flow, as well as robotic fabrication is presented to demonstrate the potential of AM methods in construction. Furthermore, this paper provides a perspective on potential future research paths and opportunities inherent in leveraging the innovative SC3DP technique.

Hybrid materials enable innovative and technologically advantageous lightweight design solutions, especially the efficient combination of metals and polymers is a promising approach. This contribution addresses an additive manufacturing (AM) process with polymers on the surface of foamable extruded aluminium profiles. The aim is to individualise aluminium profiles with locally applied polymer components using the foam structure as a mechanical bonding interface. The bond strength of hybrid aluminium foam-polymer composites fabricated via screw extrusion additive manufacturing (SEAM) is investigated and the influence of the key process parameters as well as the properties of the aluminium foam structures are taken into account. Additionally, X-ray microscopy is used to analyse the pore structure, evaluating the pore size distribution, the wall thickness, and the pore filling. The material- and process-dependent bond strength is determined from lap-shear and cross-tension tests. A ductile failure of the specimens was detected, which is caused by mixed failure modes such as cohesive failure and adhesive failure in the polymer and in the aluminum. By minimising the distance between the extruder nozzle and the aluminium foam, the samples with the highest adhesive strength of 5.4 MPa in the cross-tensile test and 7 MPa in the lap-shear test were produced. An inhomogeneous pore distribution shows the highest influence on the tested bond strength, which results in a large scattering of the maximum detected testing force. The aluminium foam-polymer composites show potential for overcoming the problem of joining dissimilar materials to produce hybrid structural components, which could enable further advances for these types of components.

Handling large flexible components is still a challenge in many industries. Examples include the handling of fibre-reinforced plastics or the assembly of industrial-scale electrolytic cells. Difficulties often arise in the design of suitable endeffectors. Inefficient gripping point design, i.e. the total number and positioning of grippers, can lead to increased stress and deflection of the component being handled. To counteract this, endeffectors are often oversized resulting in the use of more grippers than needed. Correspondingly, heavier moving masses imply longer handling times as well as higher energy consumption. This paper presents a process for planning and optimising gripping points for large flexible components. In addition to the shape of the component, actual dynamic loads of the handling path are also taken into account. The key element to the process is an optimisation algorithm based on reinforcement learning and trained using an finite element method (FEM) simulation. After computing a desirable starting configuration, the algorithm optimises the placement of gripper positions while aiming for a reduced total number. In addition, the optimisation has prescribed handling limits, such as physical and geometric constraints, that must not be exceeded. It was shown that the algorithm satisfactorily optimises the gripping points for dynamic loads for different materials and shapes. Furthermore, it has been shown that the computation of an initial configuration yields preferable results for simple components, yet requiring optimisation in the case of more complex shapes.

In the production of hybrid composites made of thermoplast and metal sheets, certain limitations in the linear forming process can be overcome through a free kinematic forming process. This process involves distributing the thermoplast press mass using a tumbling or cyclic rolling movement of the forming tool. Due to this special kinematics, only a small part of the tool is ever in engagement with the workpiece, so the component is formed incrementally. Therefore, the final geometry does not have to be created in one stroke but can be carried out step by step in several partial strokes. In order to determine the optimal trajectory of the forming tool, the kinematic interplay between the workpiece, forming tool, and toolpath should be numerically captured in a mathematical model. The specific toolpath depends on the component being formed, as well as the actual geometry of the forming tool. In this research, the free kinematic forming process is modeled and the optimal toolpath is determined using a genetic algorithm. The optimal toolpath enables a uniform distribution of the thermoplast to a specified thickness and collisions between the upper and lower tool are avoided. This enables key insights into the free kinematic forming process and opens up further avenues of research for this innovative and future‐oriented manufacturing concept.

In the design and production of press part components, the tool development process is an essential step to fulfil the required quality criteria. Conventionally, this is achieved based on expert knowledge to design the effective surface with respect to the multitude of physical, procedural, and human influences. Thus, several iterations in the tool development process are usually required, which are costly and can lead to bottlenecks within product design cycles. To accelerate this tool design, we propose a diffusion model architecture to inversely design the necessary effective tool surface given a desired geometry of the press part. This diffusion model is able to reduce the generalization issues of classical machine learning approaches by leveraging the attention mechanism both in the spatial and temporal dimension of the underlying forming process. The applicability of a similar diffusion model has already been shown in previous applications for the inverse‐design of metamaterials and this work further demonstrates diffusion models as a suitable model candidate for the inverse‐design of 3D‐geometries. For model training, finite element simulations containing the time series of deformation states during the forming process were used. Furthermore, different geometry variations of part and tool as well as relevant press process parameters were used in the training. With the procedure demonstrated in this study, a future‐oriented support for the tool development process has been shown, enabling further developments towards a time‐ and cost‐efficient production of press tools.

Hybrid structures are increasingly relevant for lightweight design and functional components. Various manufacturing techniques for creating hybrid components exist, often focusing on combining metal and polymer components. The present study examines a novel approach for manufacturing hybrid functional structures by directly printing thermoplastic onto aluminum profiles using closed‐cell aluminum foam, with the mechanical interlocking resulting in a structural bond. The filling of pores in the aluminum foam with the polymer affects the bond strength of this hybrid compound. Within a simulation‐based process design the additive manufacturing process parameters are investigated using a computational fluid dynamics (CFD) model with Fluid‐Structure Interaction for the pore‐filling process and a finite element (FE) model to evaluate the resulting bond strength. The material throughput, the nozzle distance, and the polymer temperature are the key process parameters taken into account. A parameterized, virtual foam model is used within this framework, where a linear movement of the extruder nozzle and the resulting polymer strand is investigated. This method includes the mesh conversion from the CFD results to the FE mesh for structural analysis by exporting the iso‐surface for the polymer geometry embedded in the pores of the aluminum foam structure. This approach made it possible to investigate the phenomena that occur in the process in more detail, particularly to quantify the influence of air inclusions. Thus, a decisive contribution to the hybridization of aluminum foams has been made, which opens up further potential for research and application in the long term.