MD simulation of a copper rod under thermal shock

In this paper, thermoelastic problem of onedimensional copper rod under thermal shock is simulated using molecular dynamics method by adopting embedded atom method potential. The rod is on axis x, the left outermost surface of which is traction free and the right outermost surface is fixed. Free boundary condition is imposed on the outermost surfaces in direction y and z. The left and right ends of the rod are subjected to hot and cold baths, respectively. Temperature, displacement and stress distributions are obtained along the rod at different moments, which are shown to be limited in the mobile region, indicating that the heat propagation speed is limited rather than infinite. This is consistent with the prediction given by generalized thermoelastic theory. From simulation results we find that the speed of heat conduction is the same as the speed of thermal stress wave. In the present paper, the simulations are conducted using the large-scale atomic/molecular massively parallel simulator and completed visualization software.

Payload Encoding Representation from Transformer for Encrypted Traffic Classification

Traffic identification becomes more important,yet more challenging as related encryption techniques are rapidly developing nowadays.Unlike recent deep learning methods that apply image processing to solve such encrypted traffic problems,in this pa⁃per,we propose a method named Payload Encoding Representation from Transformer(PERT)to perform automatic traffic feature extraction using a state-of-the-art dynamic word embedding technique.By implementing traffic classification experiments on a pub⁃lic encrypted traffic data set and our captured Android HTTPS traffic,we prove the pro⁃posed method can achieve an obvious better effectiveness than other compared baselines.To the best of our knowledge,this is the first time the encrypted traffic classification with the dynamic word embedding has been addressed.

A Mesh Adjustment Scheme for Embedded Boundaries

An adaptive meshing technique and solution method is proposed in which a two-dimensional body-fitted multi-block mesh is locally adjusted to arbitrarily embedded boundaries that are not necessarily aligned with the mesh.Not only does this scheme allow for rapid and robust mesh generation involving complex embedded boundaries,it also enables the solution of unsteady flow problems involving bodies and interfaces moving relative to the flow domain.This scheme has been implemented within a block-based adaptive mesh refinement(AMR)numerical framework which can ease computational expense while maintaining a detailed representation of the embedded boundary and providing an accurate resolution of the spatial characteristics of the fluid flow.Rigid body motion and evolving motion due to physical processes are considered.A block-based AMR level set method is used to deal with evolving embedded boundaries.Numerical results for various test problems are presented to verify the validity of the scheme as well as demonstrate the capabilities of the approach for predicting complex two-dimensional inviscid and laminar fluid flows.