查看更多>>摘要：This paper presents the details and the characterization of two new graphite-walled cavity ionization chambers that will constitute the first Spanish national air-kerma standard for Cs-137. Physical constants and correction factors were determined according to the recent recommendations of the ICRU 90 report and the corresponding uncertainty budgets are presented. The effect of considering one conduction electron in graphite instead of two has been studied for the first time and an increase of 0.04 % in the stopping-power ratio graphite-to-air has been found for the LMRI Cs-137-spectrum. The ratios of photon mass energy-absorption coefficients of air and graphite were calculated from the three databases recommended in ICRU 90 report and the standard deviation of the three values was only 0.02 %. The validity of the Spencer-Attix theory has been studied in detail for both chambers. The standard uncertainty of the reference air-kerma will drop from 0.76 % to 0.30 %.
查看更多>>摘要：The present work was aimed to identify and quantify the offsets biasing the thermal rock strain as measured with clip-on extensometers at a mechanical testing system (MTS) so as to increase the reliability of such measured strain. We conducted an experiment via cycling the temperature of a sample of aluminum alloy between ambient temperature and 90 degrees C under the hydrostatic confining pressure as maintained at 2 MPa. Thereby, the readouts of the extensometers appeared to be irreversible subsequent to the temperature cycle mimicking residual deformation in the sample. The numerically simulated temperature in the setup of the testing system was heterogeneous throughout such temperature cycling test showing an irreversible plug-in temperature in particular. These temperatures resulted in the apparent amplification of the circuit as constituted by the extensometer and the amplifiers as well as the relevant reading offset. Such induced offset biased the actual sample strain besides the expansion (contraction) in the metal parts of the extensometers yielding the apparent irreversible strain in the sample. We presented a workflow for quantifying the amplification- and the metal part-related offsets. The validity of the workflow was demonstrated via rectifying the strain in a rock sample that was monitored throughout a temperature cycle between 30 degrees C and 40 degrees C. Our present work would be referable for experimental researches on thermal rock strain with similar setup of measurement. The amplification- and the metal part-related offsets could be likewise estimated and excluded from the measurements conducted elsewhere before the strain in a tested sample could be obtained.
查看更多>>摘要：Deep learning (DL) is an important method in industrial fault diagnosis. However, DL's network structure needs to be designed with experience. To simplify the design of network structures, we propose the neural architecture search network with Pareto efficiency reward and insert replay buffer (NAS-PERIRB) algorithm. In this paper, the early stopping and insert replay buffer (IRB) are used to improving the training efficiency of the samples. In addition, we design the Pareto efficiency reward function to optimize the goals and design a network search space to perform effective searches. What is more, we evaluate the NAS-PERIRB under two datasets. Results show that the two datasets have reached 99% accuracy in various situations, which means the NAS-PERIRB can achieve the purpose of designing the network structure independently.
查看更多>>摘要：This study developed a simple drop-shape method to measure liquid surface tension based on curve fitting technique. In this study, the explicit lying droplet profile function in terms of surface tension was derived by solving 2-D Young-Laplace equation exactly, and the derived profile function was superior to the existing approximate profile functions. By fitting the derived profile function with the detected droplet profile image data, the liquid surface tension was then obtained. Also, the measurement deviations of the proposed method have been analyzed. The proposed method does not rely on any new instrument, it is quite different in principle and simpler than the traditional methods.
查看更多>>摘要：Thermomechanical test methods with temperature modulation (MT TMA) are increasingly used to measure a number of useful properties of materials, e.g. rheological properties of polymers. This is due to the possibility of measuring thermal effects with very high resolution. The article describes and verifies a new possibility of implementing, using this method, diagnostic task, consisting in the measurement of changes in mechanical stress states in PVD thin-film systems under the influence of thermal energy. The present paper provides a description of the measurement method and a set of results of diagnostic test, performed with its use, for the thermal stability of stresses in the substrate-coating system samples. The obtained results were also compared with the results of tests carried out using other "traditional" methods and with the results of operational tests of tools with modified, involving the same coatings, working surfaces. The results document a complete applicability of the method developed.
Andersen, Jens E. T.Moemedi, KeaboletseKatse, Kebabonye
9页
查看更多>>摘要：A comprehensible method validation is required for analytical chemistry and across the disciplines, in order to produce reliable levels of uncertainty of measurements to perform modeling, testing and comparison. New guidelines of quality assurance serve as an aid to determine the uncertainty of measurement with focus on precision rather than focusing on trueness. A method validation of high-resolution continuous source flameatomic spectrometry (HR-CS FAAS) for analysis of iron and copper using the principle of pooled calibrations (PoPC), that is not part of contemporary guidelines of quality assurance (QA), was used to deliver consensus values to fulfil the requirements to scientific methodology. Improved equations of the PoPC were developed for both the homoscedastic scenario and the heteroscedastic scenario. The analysis of both iron and copper could be allocated to the homoscedastic scenario with unexpected high levels of uncertainties that might be related to both human error and sizeable detector instability.
查看更多>>摘要：Because the gear inspection process is time-consuming and the equipment wears easily, gear machining and inspection data collection are costly; therefore, it is difficult to collect sufficient gear machining data for deep learning training. To solve this problem, this study proposes an attention and adversarial transfer learning method based on ResNet with adaptive Coral loss (A(2)ResNet-aCoral) for training small-sample datasets. The method integrates sample-spatial attention and dynamic-channel attention to enhance the feature extraction ability. To realize the transfer ability, an adversarial domain structure is constructed to induce the network to extract the homogeneous features of the source and target domains. Moreover, an adaptive correlation alignment loss is proposed to shorten the distance between the source and target domains. The results of a case study indicate that A(2)ResNet-aCoral has good transfer ability on small sample gear hobbing signals, and the accuracy can reach 92.71%.
查看更多>>摘要：Uncertainties in the driving energy transferred to the split spoon sampler (Esampler) cause inaccuracies of the standard penetration test (SPT) N values. In this study, an energy monitoring module (E-module) is developed to evaluate Esampler and correct N. To demonstrate the application of the E-module, SPTs were performed at three study sites where the grounds are mainly composed of silty sands, and the transferred energy at the head of the driving rod (Ehead) and Esampler were evaluated. In addition, static penetration resistances were measured to validate the energy corrections of N. The experimental results and analyses indicate that the N corrected using Esampler exhibits an improved relationship with static penetration resistances compared with that corrected using Ehead because the Esampler directly affects the penetration of the split spoon sampler. Therefore, the evaluation of Esampler using the E-module and the correction of N may be used to obtain improved subsurface characterizations.
查看更多>>摘要：Extraction of weak bearing failure signals remains a thorny issue in press lines. An adaptive enhanced morphological gradient operator (EMGO) fault feature diagnosis method based on third-order cumulant diagonal slice spectrum (AEMGO-TOCSS) is developed. Firstly, an EMGO is proposed to enhance the filtering ability of the operator according to the different features of the signal extracted by the basic morphological operator. Then, on account of the vital significance of structural element selection in filtering, a new adaptive feature energy per-mutation entropy (FEPE) selection strategy is put forward. Finally, the denoising performance of TOCSS is used to further improve the feature extraction ability of the EMGO operator for faulty information. Both simulation and experimental results verify that the EMGO is more effective and accurate than other morphological operators in identifying weak fault features of rolling bearings, and that it presents a broad application potential in practical engineering.
查看更多>>摘要：Considering the problem that the ultra-wideband detection system is susceptible to be jammed, the mechanism of narrowband suppressive jamming on ultra-wideband radio impulse detection radar is mathematically analyzed at first in this paper. As typical suppressive jamming, the noise amplitude modulation signal is selected as research object, and the average signal-to-jamming is defined as an evaluation index. Simulation results show that the center frequency of jamming signal has the most significant influence on the impulse signal among a series of optimal jamming parameters. Afterwards, combined with genetic algorithm, an optimized singular spectrum decomposition method is proposed to realize adaptive extraction of the impulse signal. Error of the reconstructed signal based on proposed method is much smaller than that of threshold methods. The results provide a reference for subsequent anti-jamming and waveform optimization.
NSTL
Elsevier