AI & ML
Mesh Sensitivity of Different Boundary Treatments in High-Order Compact Gas-Kinetic Scheme for Hypersonic Heat-Flux Prediction
arXiv cs.NA0 views
arXiv:2611.24392v1 Announce Type: cross Abstract: Accurate wall-heat-flux prediction in hypersonic flow depends on both the wall treatment and the available near-wall resolution. This work compares three complete implementations of a gas-kinetic scheme (GKS) for Mach~8.03 laminar flow over an isothermal circular cylinder: a linear Green--Gauss scheme (GG), a compact ghost-cell scheme (GC), and a pointwise one-sided compact scheme (POS). In POS, the stationary isothermal-wall conditions are imposed at the two Gaussian points used by the kinetic wall-flux quadrature. A reference mesh and six meshes in the uniform reduction (UR), log-layer unpacked (LLU), and log-layer packed (LLP) families are used to assess heat-flux sensitivity. At 24 digitized experimental angles, GG has the lowest root-mean-square difference on the reference, UR, and LLP families, whereas POS is closest on the two LLU meshes; GC is not the lowest-RMS scheme on any tested mesh. Hence no treatment is uniformly superior. On the present high-aspect-ratio cylinder grids, GG is the more robust choice except under LLU node redistribution, for which POS is preferable. Uniform reduction produces the largest heat-flux response for all three implementations. The comparison concerns complete schemes and does not isolate the pointwise constraints from differences in reconstruction and geometry representation.
Source:https://arxiv.org/abs/2611.24392PreviousAG-BFGS: Foundations for Accelerated Geometry LearningNextConsistent interfacial flow numerics - Part II: The momentum convection term
Related posts
How to detect user frustration in chatbot conversations
0dev.to
Mình Đã Biến Gemini Thành Trợ Lý Cá Nhân Như Thế Nào (Lịch, Drive, Gmail và Những Lần Suýt Toang)
0dev.to
Im looking fr any free ai models / tools are there?, leaVE UR Lights💡
0dev.to
I checked 10 SEO agency websites for AI-search readiness. The real bugs were the boring ones.
0dev.to

How much GPU memory does an LLM need? The whole sum, worked out on real cards
0dev.to

Routing Upscaling Jobs by Output Size, Not Input Size
0dev.to