• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.17 no.3
• /
• pp.139-170
• /
• 2013

We extend a p-hierarchical decomposition of the second degree finite element space of N$\acute{e}$d$\acute{e}$lec for tetrahedral meshes in three dimensions given in [1] to meshes with hexahedral elements, and derive p-hierarchical decompositions of the second degree finite element space of Raviart-Thomas in two dimensions for triangular and quadrilateral meshes. After having proved stability of these subspace decompositions and requiring certain saturation assumptions to hold, we construct a local a posteriori error estimator for fem and bem coupling of a time-harmonic electromagnetic eddy current problem in $\mathbb{R}^3$. We perform some numerical tests to underline reliability and efficiency of the estimator and test its usefulness in an adaptive refinement scheme.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2006.11a
• /
• pp.464-473
• /
• 2006

This paper summarises theoretical and experimental work on the feedback control of sound radiation from honeycomb panels using piezoceramic actuators. It is motivated by the problem of sound transmission in aircraft, specifically the active control of trim panels. Trim panels are generally honeycomb structures designed to meet the design requirement of low weight and high stiffness. They are resiliently-mounted to the fuselage for the passive reduction of noise transmission. Local coupling of the closely-spaced sensor and actuator was observed experimentally and modelled using a single degree of freedom system. The effect of the local coupling was to roll-off the response between the actuator and sensor at high frequencies, so that a feedback control system can have high gain margins. Unfortunately, only relatively poor global performance is then achieved because of localisation of reduction around the actuator. This localisation prompts the investigation of a multichannel active control system. Globalised reduction was predicted using a model of 12 channel direct velocity feedback control. The multichannel system, however, does not appear to yield a significant improvement in the performance because of decreased gain margin.

• Fibers and Polymers
• /
• v.2 no.4
• /
• pp.203-211
• /
• 2001

Residual stresses were predicted by a flow analysis in the mold cavity and residual stress distribution in the injection molded product was measured. Flow field was analyzed by the hybrid FEM/FDM method, using the Hele Shaw approximation. The Modified Cross model was used to determine the dependence of the viscosity on the temperature and the shear rate. The specific volume of the polymer melt which varies with the pressure and temperature fields was calculated by the Tait\`s state equation. Flow analysis results such as pressure, temperature, and the location of the liquid-solid interface were used as the input of the stress analysis. In order to calculate more accurate gap-wise temperature field, a coordinate transformation technique was used. The residual stress distribution in the gap-wise temperature field, a coordinate transformation technique was used. The residual stress distribution in the gap-wise direction was predicted in two cases, the free quenching, under the assumption that the shrinkage of the injection molded product occurs within the mold cavity and that the solid polymer is elastic. Effects of the initial flow rate, packing pressure, and mold temperature on the residual stress distribution was discussed. Experimental results were also obtained by the layer removal method for molded polypropylene.

• Nuclear Engineering and Technology
• /
• v.56 no.2
• /
• pp.611-623
• /
• 2024

The ISP-47 TOSQAN experiment was analyzed with containmentFOAM which is an open-source CFD code based on OpenFOAM. The containment phenomena taking place during the experiment are gas mixing, stratification and wall condensation in a mixture composed of steam and non-condensable gas. The k-ω SST turbulence model was adopted with buoyancy turbulence models. The wall condensation model used is based on the diffusion layer approach. We have simulated the full TOSQAN experiment which had a duration 20000 s. Sensitivity studies were conducted for the buoyancy turbulence models with SGDH and GGDH and there were not significant differences. All the main features of the experiments namely pressure history, temperature, velocity and gas species evolution were well predicted by containemntFOAM. The simulation results confirmed the formation of two large flow stream circulations and a mixing zone resulting by the combined effects of the condensation flow and natural convection flow. It was found that the natural convection in lower region of the vessel devotes to maintain two large circulations and to be varied the height of the mixing zone as result of sensitivity analysis of non-condensing wall temperature. The computational results obtained with the 2D mesh grid approach were comparable to the experimental results.