Search

Bend bars made from FV566 martensitic stainless steel were extracted from the root of ex-service power plant turbine blades and representative notches introduced. Constant amplitude cyclic loading, and with overload cycles of OLR of 1.56 applied every 150 baseload cycles, were applied to the samples until a crack of 1 mm was observed. A ‘lifetime extension strategy’ was applied, involving grinding away existing short cracks, followed by T0 shot peening on the top surface of the sample. Fatigue cyclic loading was recommenced with initial strain range loading conditions until failure. A lifetime prediction method combining the Coffin-Manson and Paris Law relationships were used to investigate the efficacy of the ‘lifetime extension strategy’ for industrial use. The ‘lifetime extension strategy’ increased the life of U-notched samples tested and a minimum increase in strain range following the application of the ‘lifetime extension strategy’ was estimated. It is expected that the application of a’ lifetime extension strategy’ such as the one tested may provide some additional service life to existing in-service turbine blades.

This paper proposes a transfer learning approach to recalibrate our previously developed Wheel Odometry Neural Network (WhONet) for vehicle positioning in environments where Global Navigation Satellite Systems (GNSS) are unavailable. The WhONet has been shown to possess the capability to learn the uncertainties in the wheel speed measurements needed for correction and accurate positioning of vehicles. These uncertainties may be manifested as tyre pressure changes from driving on muddy and uneven terrains or wheel slips. However, a common cause for concern for data-driven approaches, such as the WhONet model, is usually the inability to generalise the models to a new vehicle. In scenarios where machine learning models are trained in a specific domain but deployed in another domain, the model's performance degrades. In real-life scenarios, several factors are influential to this degradation, from changes to the dynamics of the vehicle to new pattern distributions of the sensor's noise, and bias will make the test sensor data vary from training data. Therefore, the challenge is to explore techniques that allow the trained machine learning models to spontaneously adjust to new vehicle domains. As such, we propose the Recalibrated-Wheel Odometry neural Network (R-WhONet), that adapts the WhONet model from its source domain (a vehicle and environment on which the model is initially trained) to the target domain (a new vehicle on which the trained model is to be deployed). Through a performance evaluation on several GNSS outage scenarios - short-term complex driving scenarios, and on longer-term GNSS outage scenarios. We demonstrate that a model trained in the source domain does not generalise well to a new vehicle in the target domain. However, we show that our new proposed framework improves the generalisation of the WhONet model to new vehicles in the target domains by up to 32%., arXiv admin note: text overlap with arXiv:2104.02581

This paper reports results of phase transformation experiments in the α/β alloy Ti-6Al-4V from α→β→α on heating below and near the β-transus using in situ neutron diffraction. The development and evolution of crystallographic texture during the α→β→α phase transformation was determined. Annealing at a sub-transus temperature 1562K (850°C) shows similar α-phase and β-phase pole figures from room temperature up to 1562K (850°C), with a slight increase in the intensity of the β-phase pole figures, showing that the β-phase texture strengthens during heating. During cooling, the β→α transformation occurs very quickly and the initial alpha-texture obtained at room temperature is significantly weakened by the heat-treatment. Strengthening of the β-phase texture is observed on heating the sample from [1163K to 1253K (890°C to 980°C), respectively] and a Burgers relationship ((0002) α/(110) β and {11-20} α/{111} β) is seen with the α-phase texture at room temperature; but no evidence of variant selection was observed during α→β transformation. During the β→α phase transformation on cooling, the Burgers relationship holds. This indicates a texture memory effect due to the growth of the primary alpha phase present at high temperature.

Over the past decade, several researchers have presented various optimisation algorithms for use in truss design. The no free lunch theorem implies that no optimisation algorithm fits all problems; therefore, the interest is not only in the accuracy and convergence rate of the algorithm but also the tuning effort and population size required for achieving the optimal result. The latter is particularly crucial for computationally intensive or high-dimensional problems. Contrast-based Fruit-fly Optimisation Algorithm (c-FOA) proposed by Kanarachos et al. in 2017 is based on the efficiency of fruit flies in food foraging by olfaction and visual contrast. The proposed Spontaneous Fruit Fly Optimisation (s-FOA) enhances c-FOA and addresses the difficulty in solving nonlinear optimisation algorithms by presenting standard parameters and lean population size for use on all optimisation problems. Six benchmark problems were studied to assess the performance of s-FOA. A comparison of the results obtained from documented literature and other investigated techniques demonstrates the competence and robustness of the algorithm in truss optimisation., Comment: Presented at the International conference for sustainable materials, energy and technologies, 2019

Mucosal-associated invariant T (MAIT) cells are innate-like T cells abundant in humans that can be activated in a TCR-independent manner by inflammatory and antiviral cytokines. In humans, the capacity for TCR-independent activation is functionally linked to a transcriptional program that can be identified by the expression of the C-type lectin receptor, CD161. In addition to MAIT cells, it has been demonstrated that a subset of γδT cells expresses CD161 and can be activated by TCR-independent cytokine stimulation. In this study, we sought to clarify the nature of cytokine-responsive human γδT cells. We could link CD161 expression on Vδ2+ versus Vδ1+ γδT cells to the observation that Vδ2+ γδT cells, but not Vδ1+ γδT cells, robustly produced IFN-γ upon stimulation with a variety of cytokine combinations. Interestingly, both CD161+ and CD161− Vδ2+ γδT cells responded to these stimuli, with increased functionality within the CD161+ subset. This innate-like responsiveness corresponded to high expression of PLZF and IL-18Rα, analogous to MAIT cells. Vδ2+ γδT cells in human duodenum and liver maintained a CD161+ IL-18Rα+ phenotype and produced IFN-γ in response to IL-12 and IL-18 stimulation. In contrast to MAIT cells, we could not detect IL-17A production but observed higher steady-state expression of Granzyme B by Vδ2+ γδT cells. Finally, we investigated the frequency and functionality of γδT cells in the context of chronic hepatitis C virus infection, as MAIT cells are reduced in frequency in this disease. By contrast, Vδ2+ γδT cells were maintained in frequency and displayed unimpaired IFN-γ production in response to cytokine stimulation. In sum, human Vδ2+ γδT cells are a functionally distinct population of cytokine-responsive innate-like T cells that is abundant in blood and tissues with similarities to human MAIT cells.

In this study, residual stress fields, including the near-surface residual stresses, were determined for an Al7050-T7451 sample after laser peening. The contour method was applied to measure one component of the residual stress, and the relaxed stresses on the cut surfaces were then measured by X-ray diffraction. This allowed calculation of the three orthogonal stress components using the superposition principle. The near-surface results were validated with results from incremental hole drilling and conventional X-ray diffraction. The results demonstrate that multiple residual stress components can be determined using a combination of the contour method and another technique. If the measured stress components are congruent with the principal stress axes in the sample, then this allows for determination of the complete stress tensor.