A numerical dynamic-aerodynamic interface for simulating the separation dy­n­a­m­ic­s of co­n­­s­t­r­a­i­ned strap-on boosters jettisoned in the atmosphere is presented. A 6-DOF multi body dynamic solver ،using Constraint For­ce Equation Methodology is coupled with a numerical time dependent Euler flow solver. An automatic dyna­m­i­­c mesh updating proc­e­d­ure is employed using smoothing and l...

This article aims to investigate the accuracy and dimensional stability of bonded metal CFRP (Carbon Fibre Reinforced Plastic) adherends. The motivation behind this study was assess suitability for optical devices through evaluation precision bonding technology. A binocular selected as a reference device. technological sample designed, with required total runout key dimensions 0.05 mm. underwen...

We propose a method for human action recognition from still images that uses the silhouette and the upper body as a proxy for the pose of the person, and also to guide alignment between instances for the purpose of computing registered feature descriptors. Our contributions include an efficient algorithm, formulated as an energy minimization, for using the silhouette to align body parts between...

Accurate metering of agricultural water is becoming increasingly important world wide. Surprisingly, the dynamics of water in small irrigation canals are not well understood and this limits the accuracy of flow measurements in these channels. To help improve the accuracy of flow data, we performed detailed measurements of the velocity distribution for a variety of flow conditions in a several s...

Efficient maintenance and upgrading of a building during its lifecycle are difficult because a cementitious finish uses materials and parts with low disassembly properties. Additionally, the reuse and recycling processes during building demolition also present numerous problems from the perspective of environmental technology. In this study, an induction heating (IH) method was used to disassem...

We train a kernel SVM for each action class. The SVM kernel is a convex combination of base kernels, which capture different visual cues: HOG, SIFT, color, pose, object detection scores. Some of these cues are computed at various relative locations of the provided human bounding box, yielding a total of 20 kernels. We optimize the weights for kernel combination using randomized grid search. Thi...

Very often, in order to optimized welded constructions, some of their parts are made of heterogeneous steels. For example, microalloyed steels are in use for cars chassis making and high alloyed steels for different constructions making. From that reason, it’s not rare to find welded joints between ferrite microalloyed and austenitic highalloyed steels. As a rule, these joints are welded by usi...