首页> 外文会议>IEEE International Magnetics Conference >Study on Calculation Methods of AC Loss for a HTS Magnet with Iron Core.
【24h】

Study on Calculation Methods of AC Loss for a HTS Magnet with Iron Core.

机译:铁芯高温超导磁体交流损耗的计算方法研究。

获取原文
获取外文期刊封面目录资料

摘要

The great advantage of HTS coils in magnet is that it can provide large excitation in a limited space. However, under high level excitation condition, especially in the case of fast adjusting process, AC losses will occur and lead to reduction of thermal stability [1], [2]. Therefore, it is necessary to calculate AC loss of HTS coils quickly and accurately. The homogenization method based on H formulation basically meets the calculation requirements of AC loss for a thousand-turn magnet with simple structure [3]. However, for magnets with non-linear ferromagnetic materials such as HTS controllable reactors [4], [5], the non-linear saturation in ferromagnetic domains makes it difficult to calculate AC loss rapidly and accurately with homogenization methods. Three simplified calculation methods of AC loss for HTS magnets have been proposed in this paper. In order to verify the validity of the simplified algorithm quickly and effectively, a two-dimensional axisymmetric model is adopted. The key to reducing the nonlinearity is to quickly calculate the magnetic permeability distribution in the core region. Three methods have been proposed to simplify the calculation. The first method is A+H formulation method. In the A formulation, set the same resistivity through the current flowing area, apply a uniform current density excitation and set BH magnetic properties in the core. In the H formulation, set the nonlinear resistivity decided by E-J characteristic through the current flowing area [6], apply a total current constraint and the magnetic permeability of the core region is from the real-time calculation of the A formulation results. The coupling between the two formulations does not occur during the solution, but the permeability of the core in the H formulation is provided by the calculation of the A formulation. The second method is magnetic permeability transfer of the core region method. Core area is divided into different regions. The magnetic permeability in all the core regions is solved in the A formulation, which is made into a data table. Then the magnetic permeability is applied in the H formulation model as a known item. The third method is A formulation coupling with H formulation method (A&H formulation). A formulation and H formulation share the same model. The PDE module (the control formulation is the H formulation) only contains the superconducting domain and part of the air domain around the superconducting domain. All the domains are contained in the magnetic field module (the control formulation is the A formulation). The core domain is described by the BH curve. The HTS coils are excited by the uniform current density. The section boundary of air domain is shared with the PDE module to transfer the magnetic field strength to the PDE module. The example model is a small iron-containing superconducting magnet, which is modeled with a homogenization method. Considering that the H-formulation method is widely used and supported by a large number of experiments [7], [8], the present example uses the H-formulation results as the benchmark in the error analysis of each method. Fig.1 shows the calculation time with different methods. In the linear discrete model, A+H formulation solves two physical formulations for the whole domain, which has the highest degree of freedom and a 13.5% increase in the number of degrees of freedom compared with other linear discrete models. The A-formulation coupling with H-formulation adopts quadratic discrete and has the highest degree of freedom, which also has the longest solution time. Fig.2 shows the average loss of each method. The average loss of each model with linear discreteness is not much different from that of the H-formulation model and the maximum deviation is 0.527%. The error of the third method using the quadratic discrete is larger, ranging from 11.59% to 20.47%. Both of them are larger than the H-formulation calculation results. In summary, the first method is the first choice when we calculate AC Loss for a HTS Magnet with Iron Core. If the degree of freedom is too large in the model, we can choose the second method. If the AC loss is still difficult to calculate, we need to choose the third method at the expense of a little accuracy. This paper presents three simplified calculation methods of AC loss for HTS magnets, which make it possible to calculate the AC loss of the core-containing superconducting magnet rapidly and accurately. The first two methods have high precision, but they are not suitable for the large-scale model. The third method can calculate the large-scale model quickly, but there is a small amount of error.
机译:HTS线圈在磁体中的最大优势在于,它可以在有限的空间内提供较大的励磁。但是,在高水平励磁条件下,尤其是在快速调节过程中,会发生交流损耗并导致热稳定性降低[1],[2]。因此,有必要快速,准确地计算出高温超导线圈的交流损耗。基于H公式的均质化方法基本可以满足结构简单的一千匝磁铁交流损耗的计算要求[3]。然而,对于具有非线性铁磁材料的磁体,例如HTS可控电抗器[4],[5],铁磁域中的非线性饱和使得很难通过均质化方法快速准确地计算出交流损耗。本文提出了三种简化的高温超导磁体交流损耗计算方法。为了快速有效地验证简化算法的有效性,采用了二维轴对称模型。减少非线性的关键是快速计算核心区域的磁导率分布。已经提出了三种方法来简化计算。第一种方法是A + H配制方法。在A配方中,在电流流动区域中设置相同的电阻率,施加均匀的电流密度激励,并在铁心中设置BH磁性能。在H公式中,通过电流区域设置由E-J特性决定的非线性电阻率[6],施加总电流约束,并且核心区域的磁导率是根据A公式的实时计算得出的。固溶过程中不会发生两种配方之间的偶合,但通过A配方的计算可提供H配方中岩心的渗透性。第二种方法是磁芯区域的磁导率转移方法。核心区域分为不同区域。用公式A求解所有核心区域的导磁率,并将其制成数据表。然后,将磁导率作为已知项应用于H配方模型中。第三种方法是将A配方与H配方方法结合使用(A&H配方)。配方和H配方共享相同的模型。 PDE模块(控制配方为H配方)仅包含超导域和围绕超导域的空气域的一部分。所有域都包含在磁场模块中(控制配方为A配方)。核心域由BH曲线描述。 HTS线圈被均匀的电流密度激励。空气域的截面边界与PDE模块共享,以将磁场强度传递到PDE模块。示例模型是一个小型的含铁超导磁体,使用均质化方法对其进行了建模。考虑到H公式化方法已被广泛使用并得到大量实验的支持[7],[8],本示例将H公式化结果用作每种方法的误差分析的基准。图1显示了不同方法的计算时间。在线性离散模型中,A + H公式解决了整个域的两种物理公式,与其他线性离散模型相比,它具有最高的自由度,并且自由度数增加了13.5%。 A配方与H配方的耦合采用二次离散,具有最高的自由度,并且求解时间也最长。图2显示了每种方法的平均损失。具有线性离散性的每个模型的平均损失与H公式模型的平均损失相差不大,最大偏差为0.527%。使用二次离散的第三种方法的误差较大,范围为11.59%至20.47%。两者都大于H公式的计算结果。总而言之,当我们计算带铁芯的高温超导磁体的交流损耗时,第一种方法是首选。如果模型中的自由度太大,则可以选择第二种方法。如果仍然很难计算出交流损耗,则我们需要选择第三种方法,但会降低精度。本文介绍了三种用于高温超导磁体的交流损耗的简化计算方法,这使得可以快速,准确地计算含芯超导磁体的交流损耗。前两种方法具有较高的精度,但不适用于大规模模型。第三种方法可以快速计算大型模型,但是误差很小。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号