Robustness of numerical methods has attracted an increasing interest in the community of computational fluid dynamics. One mathematical aspect of robustness for numerical methods is the positivity-preserving property. At high Mach numbers or for flows near vacuum, solving the conservative Euler equations may generate negative density or internal energy numerically, which may lead to nonlinear instability and crash of the code. This difficulty is particularly profound for high order methods, for multi-material flows and for problems with moving meshes, such as the Lagrangian methods. In this paper, we construct both first order and uniformly high order accurate conservative Lagrangian schemes which preserve positivity of physically positive variables such as density and internal energy in the simulation of compressible multi-material flows with general equations of state (EOS). We first develop a positivity-preserving approximate Riemann solver for the Lagrangian scheme solving compressible Euler equations with both ideal and non ideal EOS. Then we design a class of high order positivity-preserving and conservative Lagrangian schemes by using the essentially non-oscillatory (ENO) reconstruction, the strong stability preserving (SSP) high order time discretizations and the positivity-preserving scaling limiter which can be proven to maintain conservation and uniformly high order accuracy and is easy to implement. One-dimensional and two-dimensional numerical tests for the positivity-preserving Lagrangian schemes are provided to demonstrate the effectiveness of these methods.

No abstract uploaded!

This paper presents a method for generative design of decorative architectural parts such as corbel,moulding and panel, which usually have clear structure and aesthetic details. The method is composed of two components: offline learning and online generation. The offline learning trains a 2D CurveInfoGAN and a 3D VoxelVAE that learn the feature representations of the parts in a dataset. The online generation proceeds with an evolution procedure that evolves to product new generation of part components by selecting, crossing over and mutating features, followed by a feature-driven deformation that synthesizes the 3D mesh representation of new models. Built upon these technical components, a generative design tool is developed, which allows the user to input a decorative architectural model as a reference and then generates a set of new models that are “more of the same” as the reference and meanwhile exhibit some “surprising” elements. The experiments demonstrate the effectiveness of the method and also showcase the use of classic geometric modelling and advanced machine learning techniques in modelling of architectural parts.

It is shown that if a function$u$satisfies a backward parabolic inequality in an open set Ω∉$R$^{$n$+1}and vanishes to infinite order at a point ($x$_{0}·$t$_{0}) in Ω, then$u(x, t$_{0})=0 for all$x$in the connected component of$x$_{0}in Ω⌢($R$^{$n$}×{$t$_{0}}).

For a self-similar measure in d-dimensional Euclidean space with overlaps but satisfies the so-called bounded measure type condition introduced by Tang and the authors, we set up a framework for deriving a closed formula for the Lq-spectrum of the measure for nonnegative q. The framework allows us to include iterated function systems that have different contraction ratios and those in higher dimension. For self-similar measures with overlaps, closed formulas for the Lq-spectrum have only been obtained earlier for measures satisfying Strichartz second-order identities. We illustrate how to use our results to prove the differentiability of the Lq-spectrum, obtain the multifractal dimension spectrum, and compute the Hausdorff dimension of the measure.