RESUMO
Data-driven analysis for damage assessment has a large potential in structural health monitoring (SHM) systems, where sensors are permanently attached to the structure, enabling continuous and frequent measurements. In this contribution, we propose a machine learning (ML) approach for automated damage detection, based on an ML toolbox for industrial condition monitoring. The toolbox combines multiple complementary algorithms for feature extraction and selection and automatically chooses the best combination of methods for the dataset at hand. Here, this toolbox is applied to a guided wave-based SHM dataset for varying temperatures and damage locations, which is freely available on the Open Guided Waves platform. A classification rate of 96.2% is achieved, demonstrating reliable and automated damage detection. Moreover, the ability of the ML model to identify a damaged structure at untrained damage locations and temperatures is demonstrated.
Assuntos
Aprendizado de Máquina , Ultrassom , Algoritmos , Computadores , Ondas UltrassônicasRESUMO
In this manuscript, we explore the sensor properties of epitaxially grown graphene on silicon carbide decorated with nanolayers of CuO, Fe3O4, V2O5, or ZrO2. The sensor devices were investigated in regard to their response towards NH3 as a typical reducing gas and CO, C6H6, CH2O, and NO2 as gases of interest for air quality monitoring. Moreover, the impact of operating temperature, relative humidity, and additional UV irradiation as changes in the sensing environment have been explored towards their impact on sensing properties. Finally, a cross-laboratory study is presented, supporting stable sensor responses, and the final data is merged into a simplified sensor array. This study shows that sensors can be tailored not only by using different materials but also by applying different working conditions, according to the requirements of certain applications. Lastly, a combination of several different sensors into a sensor array leads to a well-performing sensor system that, with further development, could be suitable for several applications where there is no solution on the market today.
RESUMO
Gases, such as nitrogen dioxide, formaldehyde and benzene, are toxic even at very low concentrations. However, so far there are no low-cost sensors available with sufficiently low detection limits and desired response times, which are able to detect them in the ranges relevant for air quality control. In this work, we address both, detection of small gas amounts and fast response times, using epitaxially grown graphene decorated with iron oxide nanoparticles. This hybrid surface is used as a sensing layer to detect formaldehyde and benzene at concentrations of relevance (low parts per billion). The performance enhancement was additionally validated using density functional theory calculations to see the effect of decoration on binding energies between the gas molecules and the sensor surface. Moreover, the time constants can be drastically reduced using a derivative sensor signal readout, allowing the sensor to work at detection limits and sampling rates desired for air quality monitoring applications.
RESUMO
Dedicated methods for quantification and identification of reducing gases based on model-based temperature-cycled operation (TCO) using a single commercial MOS gas sensor are presented. During high temperature phases the sensor surface is highly oxidized, yielding a significant sensitivity increase after switching to lower temperatures (differential surface reduction, DSR). For low concentrations, the slope of the logarithmic conductance during this low-temperature phase is evaluated and can directly be used for quantification. For higher concentrations, the time constant for reaching a stable conductance during the same low-temperature phase is evaluated. Both signals represent the reaction rate of the reducing gas on the strongly oxidized surface at this low temperature and provide a linear calibration curve, which is exceptional for MOS sensors. By determining these reaction rates on different low-temperature plateaus and applying pattern recognition, the resulting footprint can be used for identification of different gases. All methods are tested over a wide concentration range from 10 ppb to 100 ppm (4 orders of magnitude) for four different reducing gases (CO, H2, ammonia and benzene) using randomized gas exposures.
RESUMO
Shape Memory Alloys (SMA) using elastocaloric cooling processes have the potential to be an environmentally friendly alternative to the conventional vapor compression based cooling process. Nickel-Titanium (Ni-Ti) based alloy systems, especially, show large elastocaloric effects. Furthermore, exhibit large latent heats which is a necessary material property for the development of an efficient solid-state based cooling process. A scientific test rig has been designed to investigate these processes and the elastocaloric effects in SMAs. The realized test rig enables independent control of an SMA's mechanical loading and unloading cycles, as well as conductive heat transfer between SMA cooling elements and a heat source/sink. The test rig is equipped with a comprehensive monitoring system capable of synchronized measurements of mechanical and thermal parameters. In addition to determining the process-dependent mechanical work, the system also enables measurement of thermal caloric aspects of the elastocaloric cooling effect through use of a high-performance infrared camera. This combination is of particular interest, because it allows illustrations of localization and rate effects - both important for efficient heat transfer from the medium to be cooled. The work presented describes an experimental method to identify elastocaloric material properties in different materials and sample geometries. Furthermore, the test rig is used to investigate different cooling process variations. The introduced analysis methods enable a differentiated consideration of material, process and related boundary condition influences on the process efficiency. The comparison of the experimental data with the simulation results (of a thermomechanically coupled finite element model) allows for better understanding of the underlying physics of the elastocaloric effect. In addition, the experimental results, as well as the findings based on the simulation results, are used to improve the material properties.
Assuntos
Ligas , Modelos Químicos , Temperatura Alta , TitânioRESUMO
AIMS: To study whether coexpression of the two hypoxia-related proteins hypoxia-inducible factor (HIF)-1α and glucose transporter (GLUT)-1 has prognostic relevance in oral squamous cell carcinomas (OSCCs). METHODS AND RESULTS: Eighty-two OSCC samples were analysed for expression levels of HIF-1α and GLUT-1 by immunohistochemistry. Protein expression was assessed with an immunoreactive score system, and the correlations between gene expression and both clinical and pathohistological parameters were examined. Overexpression of either GLUT-1 or HIF-1α was associated with poor disease-specific survival in OSCC patients. Multivariate Cox proportional-hazards regression analysis revealed that increased expression of HIF-1α was significantly associated with disease-specific survival (relative risk = 3.24, P = 0.024), as compared with the group with a low level of expression. Coexpression of HIF-1α and GLUT-1 was additively and significantly associated with adverse prognoses in patients with OSCC. Patients whose tumours had increased levels of expression of both HIF-1α and GLUT-1 were found to have a 5.13-fold increased risk of tumour-related death (P = 0.017). CONCLUSIONS: Coexpression of high levels of HIF-1α and GLUT-1 is significantly correlated with prognosis in OSCC patients, suggesting that the coexpression of these proteins can be used as both an early diagnostic and independent prognostic marker.
Assuntos
Carcinoma de Células Escamosas/secundário , Transportador de Glucose Tipo 1/metabolismo , Subunidade alfa do Fator 1 Induzível por Hipóxia/metabolismo , Neoplasias Bucais/patologia , Adulto , Idoso , Idoso de 80 Anos ou mais , Biomarcadores Tumorais/metabolismo , Carcinoma de Células Escamosas/metabolismo , Carcinoma de Células Escamosas/mortalidade , Terapia Combinada , Feminino , Alemanha/epidemiologia , Humanos , Linfonodos/patologia , Metástase Linfática , Masculino , Pessoa de Meia-Idade , Neoplasias Bucais/metabolismo , Neoplasias Bucais/mortalidade , Modelos de Riscos Proporcionais , Taxa de Sobrevida , Adulto JovemRESUMO
The critical molecular regulator of hypoxia is the hypoxia-inducible factor 1 alpha (HIF-1alpha). The prognostic impact of this regulator protein in oral squamous cell carcinomas (OSCC) has not been comprehensively investigated. The aim of this study was to analyze the expression of HIF-1alpha in 82 patients with OSCC and to correlate it with their disease-specific survival. Immunohistochemical staining for HIF-1alpha was performed on 82 OSCC specimens using a standard immunoperoxidase technique. The expression of HIF-1alpha was correlated with poor disease-specific survival for OSCC patients. Patients with negatively or weakly HIF-1alpha-expressing tumors had a survival rate of 80%, whereas the survival decreased to only 33.6% in case of moderate or strong expression. In multivariate Cox regression analysis, we found a 3.5-fold increased risk of tumor-related death when HIF-1alpha was strongly expressed (p=0.016) compared to negative or weak expression of HIF-1alpha. We suggest HIF-1alpha is an independent prognostic marker in OSCC. Immunohistochemical detection of HIF-1alpha appears to be useful in the diagnosis of OSCC and to provide prognostic information in addition to TNM stage and histological grade.