• 대한전기학회논문지
• /
• 제38권1호
• /
• pp.22-28
• /
• 1989

This paper describes error estimate mothod for adaptive finite element analysis of two dimensional electrostatic and magnetostatic field problems. To estimate the local errors, divergence theorem is used for electrostatic field and Ampere's circuital law for magnetostatic field. To confirm the effectiveness of the proposed error estimators, adaptive finite element computations are performed using the proposed error estimators. The rates of convergence of global errors are comparable with those of existing adaptive finite element schemes which make use of field continuity conditions between element boundaries. This algorithm of error estimate can be easily implemented because of its simplicity. Especially, when the value of charge in electrostatic field and the value of current in magnetostatic field are to be figured out, this method is considerded to be preferable to other approaches.

• Journal of Magnetics
• /
• 제14권1호
• /
• pp.15-17
• /
• 2009

The magnetostatic interaction between the two magnetic nanowires was studied by using the longitudinal magneto-optical Kerr effect (MOKE). For this purpose two magnetic nanowires having different widths (400 nm, 800 nm) were fabricated on an Si substrate with electron beam lithography and the lift-off method. Magnetic hysteresis loops measured by MOKE showed double switching behavior, corresponding to the separated switching fields of each wire. The switching field of the narrow wire was greatly affected by the separation between the two wires. Based on how the switching field changes with decreasing separation, it is concluded that the magnetostatic field of the 800-nm wire strongly affects the switching of the 400-nm wire when the separation is less than $0.5{\mu}m$.

• 천문학회지
• /
• 제26권2호
• /
• pp.89-98
• /
• 1993

Magnetostatic models of starspots of late type main sequence stars$(G5V\~K5V)$ have been constructed to investigate their physical characteristics by using the similarity law suggested by Schluter and Temesvary(1958) and later employed by Deinzer(1965) and Yun(1968). The starspots are assumed to be single, circular and in horizontal magnetostatic equilibrium. In the present study we considered only those model spots whose area covers less than $12\%$ of the entire stellar surface as suggested by observations. The computed surface field strength of our model spots ranges from $10^3$\;to\;several\;10^3$ gauss and their magnetic flux is found to be $10\~100$ times that of sunspots. The field strength is sensitive to spectral type, which increases with later spectral type. In contrast to the field strength, the area of starspots depends strongly on the total magnetic flux. Finally, it is noted that the computed field strength of model spots belonging to $G0V\~G5V$ falls below the equipartition field strength at their parent stellar surface unless the coverage is less than $2\%$. This suggests that the observed spot on $G0V\~G5V$ stars is likely to be a group of small starspots.

• 대한기계학회논문집A
• /
• 제31권3호
• /
• pp.388-394
• /
• 2007

Many engineering problems require the calculation of effective material properties of a structure which is composed of repeated micro-structures. The homogenization method has been used to calculate the effective (homogenized) properties of composites and several homogenization procedures for different physical fields have been introduced. This research describes the modified homogenization technique for magnetostatic problems. Assuming that the material is periodically repeated, its effective permeability can be prescribed by calculating the homogenized magnetic reluctivity using the finite element analysis of the micro unit cell. Validity of the suggested method is confirmed by comparing the results by the energy based method as well as the widely known homogenization method.

• Journal of Magnetics
• /
• 제5권3호
• /
• pp.90-98
• /
• 2000

The change in the magnetic properties of a spin-valve multilayer with the structure IrMn (9 m)/CoFe (4 nm)/Cu (2.6 nm)/CoFe (2 nm)/NiEe (6 nm) is investigated as a function of the size and the aspect ratio. At a fixed aspect ratio (the length/width ratio) of 2, the magnetostatic interactions begin to affect the magnetic properties substantially at a spin-valve length of 5 $\mum$, and, at a length of 1 $\mum$, they become even more dominant. In the case of a fixed multilayer size (2.4 $\mum$) which is indicated by the sum of the length and the width, magnetization change occurs by continuous spin-reversal and M-H loops are characterized by no or very small hysteresis at aspect ratios smaller than unity, At aspect ratios greater than unity, magnetization change occurs by spin-flip resulting in squared hysteresis loops. A very large changes in the coercivity and the bias field is observed, and these results are explained by two separate contributions to the total magnetostatic interactions: the coercivity by the self-demagnetizing field and the bias field by the interlayer magnetostatic interaction field.

• Journal of Magnetics
• /
• 제6권4호
• /
• pp.132-141
• /
• 2001

The switching characteristics and the magnetization-flop behavior in magnetic tunnel junctions exchange biased by synthetic antiferromagnets (SyAFs) are investigated by using a computer simulations based on a single-domain multilayer model. The bias field acting on the free layer is found to be sensitive to the thickness of neighboring layers, and the thickness dependence of the bias field is greater at smaller cell dimensions due to larger magnetostatic interactions. The resistance to magnetization flop increases with decreasing cell size due to increased shape anisotropy. When the cell dimensions are small and the synthetic antiferromagnet is weakly, or not pinned, the magnetization directions of the two layers sandwiching the insulating layer are aligned antiparallel due to a strong magnetostatic interaction, resulting in an abnormal magneto resistance (MR) change from the high-MR state to zero, irrespective of the direction of the free-layer switching. The threshold field for magnetization-flop is found to increase linearly with increasing antiferromagnetic exchange coupling in the synthetic antiferromagnet. Irrespective of the magnetic parameters and cell sizes, magnetization flop does not exist near zero applied field, indicating that magnetization flop is driven by the Zeeman energy.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 B
• /
• pp.903-905
• /
• 2000

For an accurate analysis of three dimensional linear magnetostatic problems, a new boundary integral equation formulation is presented. This formulation adopts difference magnetic field concept and uses single layer magnetic surface charge as unknown. The proposed method is capable of eliminating numerical cancellation errors inside ferromagnetic materials. In additions, computing time and storage memory are reduced by 75% in comparison with the reduced and total scalar potential formulation. Two examples are given to show its efficiency and accuracy.

• 한국자기학회:학술대회 개요집
• /
• 한국자기학회 2011년도 자성 및 자성재료 국제학술대회
• /
• pp.79-80
• /
• 2011

• 대한전기학회논문지
• /
• 제38권2호
• /
• pp.100-105
• /
• 1989

This paper presents an adaptive finite element method for magnetostatic force computation using Maxwell's stress tensor. Mesh refinements are performed automatically by interelement magnetic field intensity discontinuity errors and element force errors. In initial mesh, the computed forces for different integration paths give great differences, but converge to a certain value as mesh division is performed by the adaptive scheme, We obtained good agreement between analytic solutions and numerical values in typical examples.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1991년도 추계학술대회 논문집 학회본부
• /
• pp.87-90
• /
• 1991

In the field of structural analysis so-called substructure methods have been applied extensively to solve large and complex structures by splitting them up in substructures. This substructure method is applicable to electromagnetic field analysis and highly parallel in nature. In this paper substructure FEM is implemented for magnetostatic field computation using parallel computer consists of many transputers. Parallel substructure method is a promising tool as a solution of not only compuation speed problem but also memory problem.