Due to manufacturing tolerance and deterioration during operation, fan blades in the same engine exhibit geometric variability. The absence of symmetry will inevitably exacerbate and contribute to the complexities of running geometry prediction as the blade variability is bound to be amplified by aerodynamic and centrifugal loading. In this study, we aim to address the fan blade untwist related phenomenon known as alternate passage divergence (APD). As the name suggests, APD manifests as alternating passage geometry (and hence alternating tip stagger pattern) when the fan stage is operating close to/at peak efficiency condition. APD can introduce adverse influence on fan performance, aeroacoustics behaviour, and high cycle fatigue characteristics of the blade. The main objective of the study is to identify the parameters contributing to the APD phenomenon. In this study, the APD behaviours of two transonic fan blade designs are compared.

Due to imperfection in the manufacturing process and in-service wear, fan blades in a turbofan engine do not have the same geometry. This lack of symmetry inevitably leads to difficulties in predicting the fan blades' running geometry in an assembly as blade variability is amplified by the aerodynamic and centrifugal loading. This variability can lead to an aeromechanical phenomenon termed “alternate passage divergence” (APD). As the name suggests, it manifests itself as alternating geometric and aerodynamic patterns in a fan assembly during operation. APD can potentially influence the fan performance, the stability, and the multiple pure tone (MPT)/“Buzz-Saw” noise and is therefore an important area of research. For this study, the APD phenomenon is purposely triggered on a transonic fan blade. A full-assembly computational domain with one mis-staggered blade is used to examine the whole assembly performance and the untwist behavior with APD. In particular, the influence of fan blade stiffness on the APD behavior is examined. The behavior of the current blade is compared with that of a blade used in a precursor study, and it is found that, under certain conditions, the blades show similar behavior even though they have distinctly different geometry features. This illustrates that it is important to understand the phenomenon as the accurate prediction of running geometry is vital at early design stage.

In this study, the long-term mortality in the National Lung Screening Trial (NLST) was investigated using a deep learning-based method. Binary classification of the non-lung-cancer mortality (i.e. cardiovascular and respiratory mortality) was performed using neural network models centered around a 3D-ResNet. The models were trained on a participant age, gender, and smoking history matched cohort. Utilising both the 3D CT scan and clinical information, the models can achieve an AUC of 0.73 which outperforms humans at cardiovascular mortality prediction. The corresponding F1 and Matthews Correlation Coefficient are 0.60 and 0.38 respectively. By interpreting the trained models with 3D saliency maps, we examined the features on the CT scans that correspond to the mortality signal. By extracting information from 3D CT volumes, we can highlight regions in the thorax region that contribute to mortality that might be overlooked by the clinicians. Therefore, this can help focus preventative interventions appropriately, particularly for under-recognised pathologies and thereby reducing patient morbidity.