Partial Differential Equations (PDEs) are fundamental to model different phenomena in science and engineering mathematically. Solving them is a crucial step towards a precise knowledge of the ...behavior of natural and engineered systems. In general, in order to solve PDEs that represent real systems to an acceptable degree, analytical methods are usually not enough. One has to resort to discretization methods. For engineering problems, probably the best-known option is the finite element method (FEM). However, powerful alternatives such as mesh-free methods and Isogeometric Analysis (IGA) are also available. The fundamental idea is to approximate the solution of the PDE by means of functions specifically built to have some desirable properties. In this contribution, we explore Deep Neural Networks (DNNs) as an option for approximation. They have shown impressive results in areas such as visual recognition. DNNs are regarded here as function approximation machines. There is great flexibility to define their structure and important advances in the architecture and the efficiency of the algorithms to implement them make DNNs a very interesting alternative to approximate the solution of a PDE. We concentrate on applications that have an interest for Computational Mechanics. Most contributions explore this possibility have adopted a collocation strategy. In this work, we concentrate on mechanical problems and analyze the energetic format of the PDE. The energy of a mechanical system seems to be the natural loss function for a machine learning method to approach a mechanical problem. In order to prove the concepts, we deal with several problems and explore the capabilities of the method for applications in engineering.
•Proof of concept for the possibility of approximating the solution of BVPs using concepts and tools coming from deep machine learning.•The energy is the basis for the construction of the loss function.•The approximation space is defined by the architecture of the neural network.•The approach is applied to several engineering problems including linear elasticity, elastodynamics, nonlinear hyperelasticity, plate bending, piezoelectricity and phase field modeling of fracture.
We investigate the merging rates of compact binaries in galaxies and the related detection rate of gravitational wave (GW) events with AdvLIGO/Virgo and with the Einstein Telescope. To this purpose, ...we rely on three basic ingredients: (i) the redshift-dependent galaxy statistics provided by the latest determination of the star formation rate functions from UV+far-IR/(sub)millimeter/radio data; (ii) star formation and chemical enrichment histories for individual galaxies, modeled on the basis of observations; and (iii) compact remnant mass distribution and prescriptions for merging of compact binaries from stellar evolution simulations. We present results for the intrinsic birth rate of compact remnants, the merging rates of compact binaries, GW detection rates, and GW counts, attempting to differentiate the outcomes among black hole-black hole, neutron star-neutron star, and black hole-neutron star mergers and to estimate their occurrence in disk and spheroidal host galaxies. We compare our approach with the one based on cosmic star formation rate density and cosmic metallicity, exploited by many literature studies; the merging rates from the two approaches are in agreement within the overall astrophysical uncertainties. We also investigate the effects of galaxy-scale strong gravitational lensing of GW in enhancing the rate of detectable events toward high redshift. Finally, we discuss the contribution of undetected GW emission from compact binary mergers to the stochastic background.
Thermokarst is the process whereby the thawing of ice-rich permafrost ground causes land subsidence, resulting in development of distinctive landforms. Accelerated thermokarst due to climate change ...will damage infrastructure, but also impact hydrology, ecology and biogeochemistry. Here, we present a circumpolar assessment of the distribution of thermokarst landscapes, defined as landscapes comprised of current thermokarst landforms and areas susceptible to future thermokarst development. At 3.6 × 10
km
, thermokarst landscapes are estimated to cover ∼20% of the northern permafrost region, with approximately equal contributions from three landscape types where characteristic wetland, lake and hillslope thermokarst landforms occur. We estimate that approximately half of the below-ground organic carbon within the study region is stored in thermokarst landscapes. Our results highlight the importance of explicitly considering thermokarst when assessing impacts of climate change, including future landscape greenhouse gas emissions, and provide a means for assessing such impacts at the circumpolar scale.
The increased use of distributed generations in distribution network has changed its characteristics from passive to active. In the present scenario, it is relevant to consider the applicability of ...some similar pricing mechanism in distribution as transmission such as nodal pricing. This paper presents new methodology based on nodal pricing for optimally allocating distributed generation for profit, loss reduction, and voltage improvement including voltage rise phenomenon. The integration of DG in distribution network may create voltage rise problem in a weak distribution network. This paper addresses voltage rise issue. The studied are based on an existing Indian rural distribution network.
We present a set of new analytic solutions aimed at self-consistently describing the spatially averaged time evolution of the gas, stellar, metal, and dust content in an individual star-forming ...galaxy hosted within a dark halo of a given mass and formation redshift. Then, as an application, we show that our solutions, when coupled to specific prescriptions for parameter setting (inspired by in situ galaxy-black hole coevolution scenarios) and merger rates (based on numerical simulations), can be exploited to reproduce the main statistical relationships followed by early-type galaxies and their high-redshift star-forming progenitors. Our analytic solutions allow one to easily disentangle the diverse role of the main physical processes regulating galaxy formation, quickly explore the related parameter space, and make transparent predictions on spatially averaged quantities. As such, our analytic solutions may provide a basis for improving the (subgrid) physical recipes presently implemented in theoretical approaches and numerical simulations and can offer a benchmark for interpreting and forecasting current and future broadband observations of high-redshift star-forming galaxies.
Background & objectives: Cognitive and other neurodegenerative conditions related to ageing have become public health priorities in low- and middle-income countries. However, contextually based, ...applied research to support the development of awareness, diagnosis and care pathways for people with dementia in South Asia is still largely undeveloped. This study was aimed to use applied research studies for dementia in South Asia as exemplars of how individual-level capacity and capability building for dementia research can be achieved.
Methods: Using Theory of Change as a framework, we embedded capacity and capability building into the studies through six domains: people (human resources), research integrity and governance, study delivery skills, international collaborative working, patient and public involvement (PPI) (awareness raising, stigma and health literacy) and development of 'pathways'. For each aspect, development goals were defined and how they would be achieved.
Results: New principal investigators, research assistants (including outcome raters), study coordinators and intervention practitioners were trained across eight study sites in India, Pakistan, and Bangladesh, for dementia research. Training was delivered at study start, and through booster sessions, using workshops, face-to-face sessions, online training and video-link sessions. International collaborations were fostered, leading to a proposal for international funding. Each study site co-created PPI events to raise awareness and to inform the research. The recruitment pathways and study logistics fostered the development dementia diagnosis and care pathways.
Interpretation & conclusions: Embedding capacity and capability building in applied dementia research in South Asia fosters the sustainability of dementia research, which is essential in developing diagnostic and care pathways.
Circuit quantum electrodynamics has proven to be a powerful tool to probe mesoscopic effects in hybrid systems and is used in several quantum computing (QC) proposals that require a transmon qubit ...able to operate in strong magnetic fields. To address this we integrate monolayer graphene Josephson junctions into microwave frequency superconducting circuits to create graphene based transmons. Using dispersive microwave spectroscopy we resolve graphene's characteristic band dispersion and observe coherent electronic interference effects confirming the ballistic nature of our graphene Josephson junctions. We show that the monoatomic thickness of graphene renders the device insensitive to an applied magnetic field, allowing us to perform energy level spectroscopy of the circuit in a parallel magnetic field of 1 T, an order of magnitude higher than previous studies. These results establish graphene based superconducting circuits as a promising platform for QC and the study of mesoscopic quantum effects that appear in strong magnetic fields.
Recurrent pregnancy loss (RPL) in polycystic ovary syndrome (PCOS), which occurs in ∼50% of total pregnancies is a frequent obstetric complication. Among the several hypotheses, insulin resistance ...(IR), obesity and hyperhomocysteinemia (HHcy) play significant role/s in RPL. This study was conducted to assess the link between elevated levels of homocysteine and IR in PCOS-associated women with RPL in Kolkata, India. A retrospective study was conducted of one hundred and twenty six PCOS women (<30 years) who experienced two or more spontaneous abortions during the first trimester presenting to Institute of Reproductive Medicine (IRM) in Kolkata during the period of March 2008 through February 2011. One hundred and seventeen non-PCOS subjects with matching age range were randomly chosen as controls. Incidence of HHcy and IR was 70.63% (n = 89) and 56.34% (n = 71), respectively, in RPL-affected PCOS population which was significantly higher (p<0.04; p<0.0001) when compared to the non-PCOS set (HHcy: 57.26%; IR: 6.83%). Rates of miscarriage were significantly higher (p<0.008; p<0.03) in hyperhomocysteinemia-induced miscarriage when compared to the normohomocysteinemic segment (PCOS: 70.63% vs.29.36% & non-PCOS: 57.26% vs. 42.73%) along with the insulin resistant (p<0.04; p<0.0001) population (PCOS: 70.63% vs. 56.34% & non-PCOS: 57.26% vs. 6.83%) in both groups. A probabilistic causal model evaluated HHcy as the strongest plausible factor for diagnosis of RPL. A probability percentage of 43.32% in the cases of HHcy- mediated RPL suggests its increased tendency when compared to IR mediated miscarriage (37.29%), further supported by ROC-AUC (HHcy: 0.778vs. IR: 0.601) values. Greater susceptibility towards HHcy may increase the incidence for miscarriage in women in India and highlights the need to combat the condition in RPL control programs in the subcontinent.
Sclerotinia stem rot is an economically important disease of canola (Brassica napus) and is caused by the fungal pathogen Sclerotinia sclerotiorum. This study evaluated the differential gene ...expression patterns of S. sclerotiorum during disease development on two canola lines differing in susceptibility to this pathogen. Sequencing of the mRNA libraries derived from inoculated petioles and mycelium grown on liquid medium generated approximately 164 million Illumina reads, including 95 million 75-bp-single reads, and 69 million 50-bp-paired end reads. Overall, 36% of the quality filter-passed reads were mapped to the S. sclerotiorum reference genome. On the susceptible line, 1301 and 1214 S. sclerotiorum genes were differentially expressed at early (8-16 hours post inoculation (hpi)) and late (24-48 hpi) infection stages, respectively, while on the resistant line, 1311 and 1335 genes were differentially expressed at these stages, respectively. Gene ontology (GO) categories associated with cell wall degradation, detoxification of host metabolites, peroxisome related activities like fatty acid ß-oxidation, glyoxylate cycle, oxidoreductase activity were significantly enriched in the up-regulated gene sets on both susceptible and resistant lines. Quantitative RT-PCR of six selected DEGs further validated the RNA-seq differential gene expression analysis. The regulation of effector genes involved in host defense suppression or evasion during the early infection stage, and the expression of effectors involved in host cell death in the late stage of infection provide supporting evidence for a two-phase infection model involving a brief biotrophic phase during early stages of infection. The findings from this study emphasize the role of peroxisome related pathways along with cell wall degradation and detoxification of host metabolites as the key mechanisms underlying pathogenesis of S. sclerotiorum on B. napus.
PDF
