马格德堡奥托·冯·格里克大学

In this contribution, we address the challenge of non-convex optimization for parameter estimation and propose a general theoretical framework to improve global convergence and its practical implementation based on neural networks. Such inverse problems, as discussed here, are typically formulated in terms of a forward model. We interpret the set of model predictions as a Riemannian manifold and identify the origin of non-convexity in the twisted and entangled surface of the model manifold. Under assumptions of regularity, we aim to unfold and rectify the manifold using an interpretable residual autoencoder to facilitate global optimization. Crucially, this transformation must preserve the orthogonality of noisy measurements below a specified signal-to-noise ratio to support unique identification. In the theoretical limit, perfect rectification of the model manifold and its tubular neighborhood reduces the inverse problem to linear least squares. We validate our approach in a proof of concept and demonstrate its practical feasibility using a synthetic model based on a parametrized cosine series with two, three, and four parameters, confirming significant acceleration in convergence and robust global identification of parameter estimates.

The development of flame kernels in turbulence is critical in internal combustion engines and gas turbines. This study investigates the competitive effects of flame stretching/wrinkling (indicated by the integral Reynolds number Ret) and inner flame structure distortion (indicated by the Karlovitz number Ka) on extinction of premixed lean NH3/H2/air flame kernels using three-dimensional direct numerical simulations (DNS) with detailed chemistry and molecular transport models. Four DNS with varying Ret and Ka are performed. It is found that the flame kernels extinguish mainly due to inner flame structure distortion at high Ka number. The local flame reactivity is found to be controlled by preferential diffusion of H2 at both positive curvature region (due to curvature-induced flame tangential diffusion) and negative curvature region (due to strain-induced flame normal diffusion). It is shown that, at high Ka number, the intense turbulence at higher Ret leads to an earlier extinction of the flame kernel, because of a faster penetration of small-scale eddies into the flame structure. The enstrophy field represents well the suppression effect of turbulent eddies on differential diffusion. In contrast, at low Ka number, more intense flame stretching/wrinkling at higher Ret enhances the flame reactivity due to increased probabilities of positive curvature and tangential strain rate. Early flame kernel/turbulence interactions also accelerate the flame kernel development due to steeper species gradients. These results provide an improved understanding of the extinction and propagation mechanism of lean turbulent premixed NH3/H2/air flame kernels.Novelty and significanceTurbulence is commonly assumed to stretch (resulting in excessive strain rate) and wrinkle (resulting in local flame shortening) the flame at high Reynolds number, as well as to disturb the inner flame structure at high Karlovitz number. The dominant mechanism of turbulence for global flame extinction remains incompletely understood, especially for lean NH3/H2/air flames, which are highly relevant for carbon-free combustion. Through DNS of such flames at various Ka and Ret numbers, the present study systematically quantifies the effects of flame stretching and wrinkling as well as inner flame structure distortion on flame extinction (resp. enhanced burning). The results provide novel insights on what controls early flame kernel development, why the thermochemical effect (due to differential diffusion) is more important than the aerodynamic effect (due to flame stretch) for lean NH3/H2/air flames, how turbulent stretching or wrinkling enhances flame extinction (for the broken flame structure at high Ka numbers) or promotes flame burning (for the thin flame structure at low Ka numbers).

Spectroscopic ellipsometry is applied to determine the dielectric functions (DFs) epsilon and absorption coefficients alpha of orthorhombic multi-domain kappa-(InxGa1-x)(2)O-3 from 0.5 to 6.5 eV. The investigated (001)-oriented thin films with 0 <= x <= 0.175 were grown by molecular beam epitaxy on c-plane sapphire substrates. By an anisotropic analysis, we demonstrate that the in-plane DF epsilon(perpendicular to) can be decomposed into their components of the orthorhombic dielectric tensor epsilon(xx) and epsilon(yy). The resulting three characteristic transition energies (E-x, E-y,E-1, E-y,E-2) were found to shift to lower energies with increasing In-content. The results are compared to previous studies, and the origin of apparent agreements is critically examined.

The recovery of cathode materials from spent lithium-ion batteries (LIBs) is highly beneficial for environmental protection and resource circularity. Here, an efficient closed-loop process is proposed for recycling spent LiCoO2. A novel deep eutectic solvent (DES) composed of betaine hydrochloride and pyruvic acid is proposed to achieve rapid and efficient leaching of Li and Co under mild conditions. Leaching kinetics of the DES is analyzed using the shrinking core model, while the leaching mechanism is systematically investigated by integrating experimental and theoretical approaches. Co and Li are recovered from the leachate through precipitation, and the leaching performance of the regenerated DES is also examined. The electrochemical performance of the regenerated LiCoO2 is tested and compared with commercial LiCoO2. Finally, the economic and environmental analysis of the closed-loop recovery process is studied to evaluate its application prospects. This work provides valuable guidance for the comprehensive recycling and utilization of spent LIBs.

Abstract Background Previous studies have not demonstrated sufficient effects of intensive care unit (ICU) follow-up clinics on health-related quality of life. This may also be influenced by the design and implementation of these studies. Feasibility and context of such studies have yet to be sufficiently researched. It is therefore crucial to investigate these contextual factors. This study aimed to explore the experiences of patients, relatives, and healthcare professionals with both the randomised controlled trial (RCT) and the ICU follow-up clinic, to assess the feasibility and acceptability of the intervention and the study design from the perspectives of those involved. Methods This qualitative research was embedded in the process evaluation of a pilot RCT on an ICU follow-up clinic. We conducted 18 semi-structured interviews with former ICU patients, relatives and the clinical study team who participated in the PINA study and used field notes from the clinical study team. Data analysis was carried out using qualitative content analysis. Results The pilot RCT was well received by potential patients and their relatives, and the motivation to participate was high, as was the motivation of the clinical study team. The main benefits of the ICU follow-up clinic intervention named by participants were being taken seriously and receiving referrals that they would not typically get from a general practitioner. The offer of home visits for patients with long travel distances to the ICU follow-up clinic was highly appreciated. Conclusions It was feasible to integrate the perspectives of patients, relatives, and healthcare professionals within a pilot RCT on an ICU follow-up clinic. Across all stakeholder groups, the ICU follow-up clinic and the RCT trial design were considered acceptable and valuable, suggesting that such follow-up care meets patient needs and that a large-scale study appears feasible. Trial registration ClinicalTrials.gov US NLM, NCT04186468, Submission: 02/12/2019, Registration: 04/12/2019, https://clinicaltrials.gov/ct2/show/NCT04186468 .