20005 research outputs found
Sort by
Non-iterative generation of optimized meshes for finite element simulations with deep learning
No full textAbstract The finite element method is one of the most widely used computational methods in engineering and science. It provides approximate solutions to boundary value problems. The quality of these solutions critically depends on the underlying discretization, the so-called mesh. To optimize the mesh, adaptive refinement methods have been proposed over the last years that can improve mesh quality over a series of iteration steps. Herein, we propose a novel deep learning architecture that can cut short the process of mesh optimization. This architecture exploits fundamental invariance and equivariance properties to keep the amount of training data modest. It can generate high-quality meshes for a given boundary value problem and a desired target approximation error in a direct, non-iterative way. We demonstrate the performance of our method by the application to standard two-dimensional linear-elastic elasticity problems. There, our method generates meshes that reduce the solution error by 22.6%(median) compared to uniform meshes with the same computational demand
The impact of cover geometry on evaporation suppression of partially covered water reservoirs
No full textSuppressing evaporative losses from water reservoirs has long been a challenge. While various methods have been developed to reduce evaporation, physical covers, such as floating elements, offer an efficient measure for reducing evaporative losses from open water storages. Although the impact of floating covers on evaporation reduction is well-studied, the influence of cover geometry and associated opening attributes at the surface remains underexplored. This study investigates the effect of different cover geometries with identical surface coverage fraction on the evaporation suppression efficiency of partially covered reservoirs. The results show that openings with larger perimeters lead to higher evaporation rates and lower suppression efficiency. Rectangular, diamond, triangular, and circular opening geometries resulted in 33–69 %, 36–71 %, 46–73 %, and 48–75 % reduction in evaporation, respectively, under various surface flow and air boundary conditions. We observed that water surface flow and wind speed initially promote thermal mixing and reduce evaporation, but beyond a threshold, the increased heat transfer dominates, causing the evaporation rate to rise. To bridge laboratory findings with real-world environmental conditions, a mathematical model is developed using dimensionless analysis and nonlinear regression. The model shows good agreements with field measurements obtained from small reservoirs
Intensification of wheat straw autohydrolysis at minimal water input: Advancing a novel hemicellulose-first approach
No full textThe autohydrolysis of wheat straw, as a key step in a hemicellulose-first concept, was investigated with a focus on separating xylose-based oligomers and water-soluble polymers. The feedstock preparation and the autohydrolysis with saturated steam (avoiding explosion and additional auxiliaries) are aligned to produce a hemicellulose hydrolysate rich in non-monomeric xylose at low liquid-to-solid ratios and mild reaction temperatures. For the first time, highly significant regression models for biomass solubility, yield of non-monomeric xylose, share of non-monomeric xylose and inorganics content of the hydrolysate were derived in conjunction. The findings emphasize the critical impact of dry mass content, particularly under conditions of low water input, with an optimal performance identified below 50%. High arabinoxylan solubilization (77%) and xylose recovery as non-monomers (85%) were achieved at a liquid-to-solid ratio of 1.5, without the need for extensive biomass size reduction. Wheat straw pieces larger than 1 cm with a dry mass content of 40% treated in the custom reactor at a severity factor of 3.7 (170 °C, 40 min) resulted in a hydrolysate dry mass containing 67% arabinoxylan, 8.6% glucan and only 6.8% phenolic and 2.5% inorganic compounds. Under optimized conditions, autohydrolysis with saturated steam enables controlled hemicellulose separation and the recovery of oligomers and polymers corresponding to approximately 16.4% of the biomass. The results offer valuable insights into systematic autohydrolysis interactions and provide a framework for minimizing water and energy demands - key challenges in terms of green chemistry and industrial applications
Mitigating the film-substrate decohesion in nanoporous metals
No full textSubstrate-supported films of dealloyed nanoporous metals spontaneously develop a low-connectivity layer at the film-substrate interface, impairing interfacial adhesion strength. Comparable interfacial degradation has been reported for sintered interconnects in microelectronics. The phenomenon arises from a divergence in the diffusive, curvature-driven flux of the material, normal to the interface. This reduces the local solute fraction and eventually leads to disconnection by a Plateau-Rayleigh-type instability. Comparing dealloying and subsequent coarsening in experiments and kinetic Monte Carlo simulation, this work investigates mitigation strategies and, in particular, the effect of distinct porosity-depth profiles on the interfacial connectivity. Two design principles are suggested: First, the profile should be smooth, avoiding jumps and kinks. Second, steep porosity gradients should be placed in low-porosity regions. Exponential profiles are found preferable for avoiding degradation. In the fields of nanoporous thin films as functional materials and of interconnects in microelectronics joining, the findings provide the basis for materials design toward enhanced adhesion and lifetime
AI-driven design of high-performance optical thin film coatings for ultrafast lasers
No full textUltrafast laser systems critically depend on optical thin film coatings to control dispersion and light propagation, enabling precise shaping of light. Optical thin film coating design represents a complex inverse problem traditionally relying on computationally intensive numerical optimization methods. These conventional approaches, exemplified by the widely used Needle Algorithm [1], require significant computational resources and often rely on expert intervention [2]. We here propose a new approach to these challenges and present a physics-informed machine learning framework based on an autoencoder architecture and use it to design an ultra-broadband dispersive mirror
Cost-effective integration of CNS infrastructure for Urban Air Mobility: insights and strategies
No full textBuilding upon previous research that focused on the design and requirements of Communication, Navigation, and Surveillance (CNS) systems for Urban Air Mobility (UAM), this paper provides critical updates on CNS infrastructure while emphasizing the economic aspects of its deployment. Aligned with NASA's UAM Concept of Operations (ConOps) and tailored for potential integration within Europe, this study bridges the gap between technical advancements and financial feasibility. The paper evaluates practical deployment scenarios for CNS systems in urban environments, presenting refined designs for optimal antenna configurations that address coverage challenges specific to UAM operations. The focus then shifts to an in-depth economic analysis, including cost modeling and break-even assessments, to evaluate the financial viability of CNS deployments. Using simulation-driven data, the study examines key cost factors such as initial capital expenditure (CAPEX), operational expenditure (OPEX), and potential returns on investment. Scenarios for various urban landscapes are analyzed to provide stakeholders with actionable insights into the scalability, efficiency, and cost-effectiveness of CNS systems. The findings aim to support decision-makers in advancing UAM integration by providing a comprehensive framework for both technical and economic planning, thereby contributing to the safe, efficient, and cost-effective realization of UAM operations
Prediction of extreme vessel responses utilizing artificial intelligence
No full textSince mankind put to sea, the seakeeping capabilities of its vessels has been the decisive factor on the save voyage of their seafarers. In modern times the analysis of vessel responses to seaway by numerical calculations has developed to be the industry standard alongside model tests. At the Institute of Ship Design and Ship Safety of the Hamburg University of Technology the Software E4-ROLLS is used to perform such kind of analyses. Originating in the 1980s this piece of software has been validated through plentiful research as well as accident investigations. While sufficiently accurate and fast in its predictions on the roll motion of oceangoing ships, such calculations are currently performed only for specific loading conditions and seaways due to its requirement of some preceding calculations and manual user interaction. The current research is set out to widen the scope onto the entire operational profile of ships and all necessary seaway situations. This is achieved by first generating a large number of loading conditions for the vessel in question using a parameterized description of its deadweight items and a Monte-Carlo-based approach. Second all necessary calculations for and with E4-ROLLS are automatized to produce a large quantity of accurate data using the generated loading conditions. The third step is to find a fitting mathematical model of this data. For this methods of machine learning, as a sub-category of artificial intelligence, are implemented and tested on the achievable prediction accuracy. Currently polynomial regression is utilized in a two-step process. In the first step the seakeeping capabilities of each loading condition are modelled by a multidimensional polynomial function optimizing the polynomial degrees for a low mean quadratic error. The second step is used to extract polynomial coefficients for any in-between loading conditions. The current implementation of this process results in a rather fast calculation, while the overall mean quadratic error remains below one meter of limiting significant wave height for an exemplary ultra large container vessel. Current work focuses on refining the polynomial model. However if the accuracy cannot be improved further, other methods may be employed in the future. With the ability to calculate a real-time prediction model of the seakeeping behaviour of a ship it is further planned to incorporate this as a warning system aboard ships assisting in avoiding dangerous vessel motion
Temporal autoencoder for identification and predictive control of nonlinear dynamics based on Koopman operator theory
No full textThis paper presents an approach to Model Predictive Control (MPC) for nonlinear systems by combining Koopman operator theory with autoencoder networks enhanced by temporal layers for time-delay embedding. Traditional Koopman-based methods miss temporal dependencies, so we incorporate Temporal Convolutional Networks (TCN), Long Short-Term Memory (LSTM), and Gated Recurrent Units (GRU) in the autoencoder. These layers capture spatial and temporal features, improving system identification and predictive capabilities.The framework is validated on Duffing and Van der Pol oscillators, showing that temporal models, especially TCN-based, outperform non-temporal ones in prediction and control. TCN-based Koopman MPC achieves faster settling times and better control, while non-temporal models show errors and poor control. The LSTM and GRU models exhibit oscillatory control behavior. We also introduce a python package PyKoopman-AE, which facilitates Koopman operator-based system identification using both temporal and non-temporal autoencoders
How does the collar in cementless hip stems work? Comparison of the strain distribution in the cortex of the proximal femur
No full textBackground: Collared cementless hip stems have demonstrated a reduced incidence of periprosthetic femoral fractures compared to collarless counterparts. Many fractures occur during implantation, when collarless stems are seated to achieve press-fit, causing critical tensile strains in the femur. Collared stems can limit excessive seating and subsidence through calcar-collar contact. This study aimed to explain the clinically observed smaller fracture rates with collared stems by comparing strain distributions during implantation and loading between collared and collarless stems. It was hypothesized that collared stems distribute applied forces through both the collar and stem, increasing compressive axial and shear strains, allowing higher load tolerance. Methods: Seven collared and seven collarless stems were implanted with constant velocity (0.1 mm/s) in porcine femurs until failure. Two human cadaveric femurs were tested as proof of concept. Shear, axial compressive and tangential tensile strains were compared alongside fracture patterns, subsidence and forces. Findings: Collared stems in porcine femurs resisted approximately twice as much force until failure occurred (collared: 4187 N, collarless: 1980 N; p < 0.001), with similar tangential tensile strains (1 % to 1.4 % p = 0.805) and subsidence of 1.6 mm for collarless and 1.1 mm for collared stems at different failure forces (p = 0.288). Axial compressive strain was heavily increased by 1147 % with collared stems (collared: 1.2 %, collarless: 0.1 %; p = 0.026). Human femurs exhibited similar trends. Interpretation: During loading, the collar prevents periprosthetic femoral fractures by increasing axial compressive strains instead of causing critical excessive tangential tensile strains (hoop strains) that can result in fractures
Wertstoffe aus Weizenstroh – Konzeptentwicklung und -bewertung im Bioraffinerie-Kontext
No full textIn this study, a process concept for wheat straw is developed that enables an effective valorization from hemicellulose and can potentially be implemented in biorefineries. A techno-economic analysis of existing utilization methods is used to derive feasible alternatives and their requirements. The concept and the experimental investigations are geared towards energy-efficient and resource- efficient feedstock utilization with hydrothermal pretreatment. To evaluate the concepts based on mass and energy balances, the experimental results are transferred to suitable process models and optimization potentials are identified.Im Rahmen dieser Arbeit wird ein Nutzungskonzept für Weizenstroh erarbeitet, das eine effektive Wertschöpfung aus der Hemicellulose ermöglicht und potenziell in Bioraffinerien umsetzbar ist. Über eine techno-ökonomische Analyse bestehender Nutzungsverfahren werden umsetzbare Alternativen und ihre Anforderungen abgeleitet. Die Konzeptionierung und die experimentellen Untersuchungen sind auf eine energieeffiziente und ressourcenschonende Rohstoffnutzung mit einer hydrothermischen Vorbehandlung ausgelegt. Zur Bewertung der Konzepte anhand von Massen- und Energiebilanzen werden die experimentellen Ergebnisse in geeignete Prozessmodelle übertragen und mögliche Optimierungspotenziale identifiziert