This paper describes an experimental investigation of the dynamics of a near-blowoff, bluff-body stabilized flame. This work is motivated by a number of prior observations showing that near-blowoff flames exhibit enhanced unsteadiness. Laser sheet imaging studies and particle image velocimetry velocity field measurements show that the transient dynamics of these flames occur in two distinct stages before blowoff. The first stage is manifested by the presence of localized "holes" in the flame sheet, at locations where the instantaneous stretch rate exceeds the extinction stretch rate. During this stage, the overall flame and wake dynamics appear essentially unaltered and, moreover, the flame can persist indefinitely, although with enhanced unsteadiness. As the equivalence ratio is further decreased, the size of the flame region affected and the duration of these events increases monotonically. As the blowoff point is approached further, this leads to the second stage, large-scale alterations of the wake dynamics, violent flapping of the flame front, and even larger straining of the flame. In some cases, the flow in this second stage bears striking resemblance to the asymmetric von Kármán type flowfield.

Unilateral sinonasal pathology are common presentations but are regarded with caution as neoplastic conditions during their early stages may mimic an inflammatory pathology. The aim of the review was to analyse the varied presentations of patients with unilateral nasal mass and to identify features suggestive of neoplastic pathology. A retrospective review of all cases of unilateral nasal mass/polyp from Jan 09 to Jan 10 presenting at a tertiary care hospital were analysed. The patients were grouped as per their histopathological diagnosis as inflammatory and neoplastic. The demographic data, presenting symptoms, radiological and histopathological findings were compared between the two groups. Out of the 53 patients of unilateral nasal mass, 44 (83.1%) had inflammatory conditions and 9 (16.9%) had neoplastic conditions. Benign nasal polyp and inverted papilloma were the commonest inflammatory and neoplastic condition. Neoplastic conditions were significantly commoner in males (P = 0.0315) and in the age group above 50 years (P = 0.0046). Epistaxis and extranasal symptoms like facial pain, dental and orbital complaints were found to be significantly higher in neoplastic conditions. Neoplastic lesions of nose and paranasal sinus are one of the most challenging conditions that otolaryngologists have to diagnose and treat due to their hidden nature and late presentations. In our review neoplastic conditions were found to be higher in elderly male with epistaxis, extranasal symptoms and presence of extensive soft tissue involvement and bony destruction on CT scan. The clinician should have a high index of suspicion to rule out a neoplastic aetiology in all cases of unilateral nasal mass.

We study how visual representations pre-trained on diverse human video data can enable data-efficient learning of downstream robotic manipulation tasks. Concretely, we pre-train a visual representation using the Ego4D human video dataset using a combination of time-contrastive learning, video-language alignment, and an L1 penalty to encourage sparse and compact representations. The resulting representation, R3M, can be used as a frozen perception module for downstream policy learning. Across a suite of 12 simulated robot manipulation tasks, we find that R3M improves task success by over 20% compared to training from scratch and by over 10% compared to state-of-the-art visual representations like CLIP and MoCo. Furthermore, R3M enables a Franka Emika Panda arm to learn a range of manipulation tasks in a real, cluttered apartment given just 20 demonstrations. Code and pre-trained models are available at https://tinyurl.com/robotr3m.

Work to develop a practical, fast diagnostic technique to monitor the proximity of a combustor to blowout using measurements of the flame's acoustic signature is described. The feasibility of this approach was demonstrated on three combustors with different flame holding mechanisms that are used in most practical combustion devices: pilot, swirl, and bluff-body stabilized flames. Extensive high-speed flame images were obtained and analyzed in conjunction with simultaneous acoustic data. These analyses revealed changes in the low-frequency spectrum and/or the increased presence of time-localized and intermittent events in the acoustic data as the combustor approached blowout. Based on these observations, spectral, statistical, wavelet, and thresholding signal-processing schemes were developed for detecting blowout precursors with varying levels of time response, sensitivity, and robustness.

No abstract