Xiaosong Lan | Searcharxiv

arXiv2026

Improved Stability-Based Transition Transport Model for Airships Incorporating Wall Heating Effects

Laminar drag reduction is a critical technology for enhancing the endurance and station-keeping capabilities of airship platforms. However, existing transport-based transition models fail to account for the premature transition induced by wall heating, a limitation that significantly hinders the robust engineering application of laminar-flow technology in realistic thermal environments.To address this deficiency, this study first develops stability-based correction for transition modeling that explicitly incorporates wall-to-freestream temperature ratios. Leveraging the Falkner--Skan--Cooke (FSC) equations and linear stability theory (LST) with the $e^N$ method, we derive physics-based correlations for the transition criteria as functions of the temperature ratio, pressure gradient, and turbulence intensity. These corrections are integrated into a simplified stability-based transition transport model proposed by \citet{franccois2023simplified} and validated against the classic Schubauer and Klebanoff flat-plate experiments, demonstrating accurate prediction of transition locations under adiabatic, heated, and cooled conditions. Crucially, wind-tunnel experiments on a heated airship model show that wall-heating sensitivity is strongly influenced by local pressure-gradient variations, which is due to Reynolds-number-driven transition-location shifts. The proposed model successfully reproduces the experimentally observed transition advancement caused by wall heating. This framework, covering both heating and cooling regimes, provides a capability to support future laminar-flow control technologies based on wall-temperature modulation.

physics.flu-dyn↗

arXiv2021

Contour Loss for Instance Segmentation via k-step Distance Transformation Image

Instance segmentation aims to locate targets in the image and segment each target area at pixel level, which is one of the most important tasks in computer vision. Mask R-CNN is a classic method of instance segmentation, but we find that its predicted masks are unclear and inaccurate near contours. To cope with this problem, we draw on the idea of contour matching based on distance transformation image and propose a novel loss function, called contour loss. Contour loss is designed to specifically optimize the contour parts of the predicted masks, thus can assure more accurate instance segmentation. In order to make the proposed contour loss to be jointly trained under modern neural network frameworks, we design a differentiable k-step distance transformation image calculation module, which can approximately compute truncated distance transformation images of the predicted mask and corresponding ground-truth mask online. The proposed contour loss can be integrated into existing instance segmentation methods such as Mask R-CNN, and combined with their original loss functions without modification of the inference network structures, thus has strong versatility. Experimental results on COCO show that contour loss is effective, which can further improve instance segmentation performances.

cs.CV↗